glew_clinical studies in medical biochemistry 3rd ed

Clinical Studiesin Medical Biochemistry,

Third Edition

Robert H. GlewMiriam D. Rosenthal,

Editors

OXFORD UNIVERSITY PRESS

Clinical Studies in Medical Biochemistry

This page intentionally left blank

CLINICAL STUDIES

IN MEDICAL BIOCHEMISTRY

Third Edition

Edited by

Robert H. Glew

and

Miriam D. Rosenthal

12007

1Oxford University Press, Inc., publishes works that further

Oxford University’s objective of excellencein research, scholarship, and education.

Oxford New YorkAuckland Bangkok Bogotá Buenos Aires Cape Town Chennai

Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi

São Paulo Shanghai Singapore Taipei Tokyo Toronto

With offices inArgentina Austria Brazil Chile Czech Republic France Greece

Guatemala Hungary Italy Japan Poland Portugal SingaporeSouth Korea Switzerland Thailand Turkey Ukraine Vietnam

Copyright © 2007 by Oxford University Press

Published by Oxford University Press, Inc.198 Madison Avenue, New York, New York, 10016

www.oup.com

Oxford is a registered trademark of Oxford University Press

All rights reserved. No part of this publication may be reproduced,stored in a retrieval system, or transmitted, in any form or by any means,

electronic, mechanical, photocopying, recording, or otherwise,without the prior permission of Oxford University Press.

Library of Congress Cataloging-in-Publication DataClinical studies in medical biochemistry / edited by Robert H. Glew and Miriam D. Rosenthal.—3rd ed.

p. ; cm.Includes bibliographical references and index.

ISBN-13: 978-0-19-517687-2 (cloth)ISBN-10: 0-19-517687-1 (cloth)

ISBN-13: 978-0-19-517688-9 (paper)ISBN-10: 0-19-517688-X (paper)

1. Clinical biochemistry—Case studies. I. Glew, Robert H. II. Rosenthal, Miriam D.[DNLM: 1. Biochemistry—Case Reports. 2. Laboratory Techniques and Procedures—Case Reports.

3. Metabolic Diseases—Case Reports. QU 4 C641 2006]RB112.5.C57 2006

612'.015—dc22 2005054659

1 3 5 7 9 8 6 4 2

Printed in the United States of Americaon acid-free paper

www.oup.com

During the last 25 years, medical schools haveextensively integrated the teaching of basicscience, including biochemistry, into the clini-cal world. It is now rare to find a medicalschool curriculum in which biochemistry istaught in isolation as a vast array of reactions,chemical mechanisms, and metabolic path-ways dissociated from the normal and patho-physiological processes involved in humanhealth and disease. Despite the current diver-sity in form of medical curricula—which rangefrom traditional lectures, through various mix-tures of lectures and small group formats, tothe “new pathway,” which relies primarily onsmall group, problem-based learning—the con-tent is now solidly oriented in a clinical con-text. As such, many courses in biochemistry,particularly those located in or associated withmedical schools, have a need for teaching ma-terials in which biochemical concepts and par-ticulars are articulated and developed throughpresentations of specific examples of humandisease.

From its inception, this book was designed tomeet this growing need. Both the first edition,published in 1987, and the expanded, revised1997 edition have served effectively as compan-ion texts to many of the standard textbooks ofbiochemistry, particularly at institutions thathave retained much of the traditional lectureformat.At the same time, experience has shownthat the 8- to 14-page chapters that constitutethis collection of actual case reports are suffi-ciently comprehensive and self-contained tostand on their own.As such, the book has beenand can be used as the primary resource in “bio-chemistry of disease” courses at the advancedundergraduate level or in masters or doctorategraduate programs.

The chapters in the current edition have beensubstantially revised to incorporate new ad-vances, particularly in molecular biology and ingene therapy, and to integrate more coherentlyinto a comprehensive presentation of medicalbiochemistry. In addition, new chapters havebeen developed to expand the scope of thebook, including collagen structure (osteogenesisimperfecta) and mitochondrial metabolism (mi-tochondrial myopathy) and reverse cholesteroltransport (Tangier disease).A new chapter on hy-perhomocystinemia provides discussion of re-cent insights on the effects of impairedmetabolism of sulfur-containing amino acid me-tabolism on vascular disease.There is also morecoverage of nutritional biochemistry, includingnew chapters on protein-calorie malnutrition,obesity, vitamin A deficiency, calcium deficiencyrickets,and iron metabolism (hemochromatosis.)

Each chapter begins with a detailed case re-port that includes the relevant history, perti-nent clinical laboratory data, and physicalfindings. In some cases, the patient about whomthe chapter is developed was the same casethat was the first of its kind described in themedical literature; in others, a more recent caseis utilized to discuss advances in our under-standing of the pathophysiology underlying en-hanced diagnosis or therapy. The contributorsto this book have been careful to define med-ical terms with which the readers might be un-familiar and to minimize their need to resort toa medical dictionary. The case presentation isfollowed by a brief Diagnosis section, which in-cludes a brief discussion of differential diagno-sis and criteria needed for establishing thediagnosis. In addition, this section of each chap-ter usually explains the principles behind keylaboratory and diagnostic tests.

Preface

The Biochemical Perspectives section formsthe heart of each chapter and is the longest sec-tion; it goes into considerable detail in explainingthe fundamental defect that lies at the core ofthis case. Incorporating recent advances in mo-lecular biology and human genetics, this sectionprovides molecular biological as well as classicbiochemical-enzymological explanations of per-tinent physiological mechanisms.The fourth sec-tion of each chapter,Therapy, provides a conciseaccount of how the disease in question istreated. If applicable, this section also incorpo-rates discussion of current and experimentaltherapies based on molecular approaches.

Each chapter ends with a list of key primaryand perhaps secondary references and a set ofquestions designed to test the reader’s compre-hension of the case in all its dimensions. Thequestions also serve to stimulate group discus-sions if the book is used in a small group or tu-torial setting.

The diseases and disorders chosen for discus-sion and the order of presentation parallel sub-ject matter taught in most first-year medicalbiochemistry. Chapters in the first part of thebook, Nucleic Acids and Protein Structure, illus-trate the relationships of protein structure andfunction with respect to collagen (OsteogenesisImperfecta) and hemoglobin (Sickle Cell Ane-mia). The chapters Fragile X Syndrome andHereditary Spherocytosis discuss key aspects ofDNA and protein structure and their respectiverole in chromosomal and cytoskeletal structure.The chapter cardiac troponin and myocardial in-farction provides an up-to-date demonstration ofthe usefulness of both structural proteins andenzymes as markers of cardiovascular disease,while the chapter α1-Antitrypsin Deficiency dis-cusses the important role of endogenous en-zyme inhibitors.

The second section of the book is Fuel Me-tabolism and Energetics. Important pathwaysand enzymes involved in fuel utilization are dis-cussed in the chapters Pyruvate DehydrogenaseComplex Deficiency, Mitochondrial En-cephalomyopathy, and Systemic Carnitine Defi-ciency. The role of gluconeogenesis in glucosehomeostasis is illustrated by a discussion in thechapter Neonatal Hypoglycemia.

An expanded section, Intermediary Metabo-lism, constitutes the third part of the book. Dis-orders of glucose and fatty acid metabolism arediscussed in the chapters Glucose 6-Phosphate

Dehydrogenase Deficiency, Biotinidase Defi-ciency, and Adrenoleukodystrophy. Catabolismof essential amino acid skeletons is discussed inthe chapters Phenylketonuria and HMG-CoALyase Deficiency.The chapters Inborn Errors ofUrea Synthesis and Neonatal Hyperbilirubine-mia discuss the detoxification and excretion ofamino acid nitrogen and of heme.The chapterGaucher Disease provides an illustration of therange of catabolic problems that result in lyso-somal storage diseases. Several additional chap-ters deal with key aspects of intracellulartransport of enzymes and metabolic intermedi-ates: the targeting of enzymes to lysosomes(I-Cell Disease), receptor-mediated endocytosis(Low-Density Lipoprotein Receptors and Famil-ial Hypercholesterolemia) and the role of ABCtransporters in export of cholesterol from thecell (Tangier disease).

The fourth section deals with various as-pects: Digestion, Absorption, and NutritionalBiochemistry. The chapter Obesity considerscurrent problems with respect to the ever-increasing incidence of imbalance betweenenergy intake and utilization. Key problems ofundernutrition are discussed in the chaptersProtein-Energy Malnutrition and Vitamin ADeficiency in Children. The chapters LactoseIntolerance, Pancreatic Insufficiency, andAbetalipoproteinemia focus on the biochemi-cal processes underlying food digestion andabsorption. Calcium Deficiency Rickets, Vita-min B12 Deficiency, and Hemochromatosis pro-vide discussions of absorption and utilizationof vitamin D, vitamin B12, and iron, respec-tively.

The last section, Endocrinology and Integra-tion of Metabolism, includes chapters on hor-monal regulation of energy metabolism (Type IDiabetes) and steroid hormone metabolism(Congenital Adrenal Hyperplasia).

This book could not have been put togetherwithout the assistance of the skilled and patientinvestigators who contributed chapters to thisthird edition of Clinical Studies in Medical Bio-chemistry; many have first-hand experiencewith the clinical disorders they describe. Fur-thermore, most of the authors of these chaptersare themselves engaged in educating medicalstudents.Whatever success this book enjoys,weowe to these contributors and to our skilled ed-itors at Oxford, Jeffrey House and at BythewayPublishing, Kim Hoag.

vi Preface

Contributors xi

Part I Nucleic Acids and Protein Structure and Function

1. Fragile X Syndrome 3

YUJI YOKOYAMA, SHINSUKE NINOMIYA, AND KOJI NARAHARA

2. Sickle Cell Anemia 17KEITH QUIROLO

3. Osteogenesis Imperfecta 30ARMANDO FLOR-CISNEROS AND SERGEY LEIKIN

4. α1-Antitrypsin Deficiency 42SARAH JANE SCHWARZENBERG AND HARVEY L. SHARP

5. Cardiac Troponin: Clinical Role in the Diagnosis of Myocardial Infarction 54FRED S. APPLE AND ALLAN S. JAFFE

6. Hereditary Spherocytosis 66HIROSHI IDEGUCHI

Part II Fuel Metabolism and Energetics

7. Pyruvate Dehydrogenase Complex Deficiency 77PETER W. STACPOOLE AND LESA R. GILBERT

8. Mitochondrial Encephalomyopathy, Lactic Acidosis, and Strokelike Episodes (MELAS):A Case of Mitochondrial Disease 89FRANK J. CASTORA

9. Systemic Carnitine Deficiency: A Treatable Disorder 101ERIC P. BRASS, HARBHAJAN S. PAUL, AND GAIL SEKAS

Contents

10. Neonatal Hypoglycemia and the Importance of Gluconeogenesis 107IAN R. HOLZMAN AND J. ROSS MILLEY

Part III Intermediary Metabolism

11. Glucose 6-Phosphate Dehydrogenase Deficiency and Oxidative Hemolysis 123CATHERINE BURTON AND RICHARD KACZMARSKI

12. Biotinidase Deficiency: A Biotin-Responsive Disorder 134BARRY WOLF

13. Adrenoleukodystrophy 144MARGARET M. MCGOVERN

14. Low-Density Lipoprotein Receptors and Familial Hypercholesterolemia 152MARINA CUCHEL AND DANIEL J. RADER

15. Tangier Disease: A Disorder in the Reverse Cholesterol Transport Pathway 159LIEN B. LAI, VIJAYAPRASAD GOPICHANDRAN, AND VENKAT GOPALAN

16. Gaucher Disease: A Sphingolipidosis 167WILLIAM C. HINES AND ROBERT H. GLEW

17. I-Cell Disease (Mucolipidosis II) 181JAMES CHAMBERS

18. Inborn Errors of Urea Synthesis 195PRANESH CHAKRABORTY AND MICHAEL T. GERAGHTY

19. Phenylketonuria 204WILLIAM L. ANDERSON AND STEVEN M. MITCHELL

20. HMG-CoA Lyase Deficiency 217VIRGINIA K. PROUD AND MIRIAM D. ROSENTHAL

21. Hyperhomocysteinemia 226ANGELA M. DEVLIN AND STEVEN R. LENTZ

22. Neonatal Hyperbilirubinemia 234JEFFREY C. FAHL AND DAVID L. VANDERJAGT

Part IV Digestion, Absorption, and Nutritional Biochemistry

23. Obesity: A Growing Problem 245MIRIAM D. ROSENTHAL AND LAWRENCE M. PASQUINELLI

viii Contents

24. Protein-Energy Malnutrition 255VIJAYAPRASAD GOPICHANDRAN, LIEN B. LAI, AND VENKAT GOPALAN

25. Lactose Intolerance 266MARCY P. OSGOOD AND ABIODUN O. JOHNSON

26. Pancreatic Insufficiency Secondary to Chronic Pancreatitis 278PETER LAYER AND JUTTA KELLER

27. Abetalipoproteinemia 290PAUL RAVA AND M. MAHMOOD HUSSAIN

28. Vitamin B12 Deficiency 300DOROTHY J. VANDERJAGT AND DENIS M. McCARTHY

29. Vitamin A Deficiency in Children 313NUTTAPORN WONGSIRIROJ, EMORN WASANTWISUT, AND WILLIAM S. BLANER

30. Calcium-Deficiency Rickets 323DOROTHY J. VANDERJAGT AND ROBERT H. GLEW

31. Hereditary Hemochromatosis 335SCOTT A. FINK AND RAYMOND T. CHUNG

Part V Endocrinology and Integration of Metabolism

32. Type I Diabetes Mellitus 345SRINIVAS PANJA, ARUNA CHELLIAH, AND MARK R. BURGE

33. Congenital Adrenal Hyperplasia: P450c21 Steroid Hydroxylase Deficiency 360GERALD J. PEPE AND MIRIAM D. ROSENTHAL

Index 369

Contents ix

WILLIAM L.ANDERSON, PH.D.Department of Biochemistry and Molecular

BiologySchool of MedicineUniversity of New MexicoAlbuquerque, New Mexico

FRED S.APPLE, PH.D.Department of Laboratory Medicine and

PathologyHennepin County Medical CenterMinneapolis, Minnesota

WILLIAM S. BLANER, PH.D.Department of MedicineCollege of Physicians and SurgeonsColumbia UniversityNew York, New York

ERIC P. BRASS, M.D., PH.D.Department of MedicineCenter for Clinical PharmacologyHarbor-UCLA Medical CenterLos Angeles, California

MARK R. BURGE, M.D.Department of Internal MedicineSchool of MedicineUniversity of New MexicoAlbuquerque, New Mexico

CATHERINE BURTON, M.A., MRCPDepartment of HaematologyUniversity CollegeLondon, United Kingdom

FRANK J. CASTORA, PH.D.Division of BiochemistryDepartment of Physiological SciencesEastern Virginia Medical SchoolNorfolk, Virginia

PRANESH CHAKRABORTY, M.D.,FRCPSC, FCCMGChildren’s Hospital of Eastern OntarioUniversity of OttawaOttawa, Ontario, Canada

JAMES CHAMBERS, PH.D.The Brain Research Laboratory of

BiochemistryDivision of Life SciencesUniversity of Texas at San AntonioSan Antonio, Texas

ARUNA CHELLIAH, M.D.University of New Mexico Health Sciences

CenterAlbuquerque, New Mexico

RAYMOND T. CHUNG, M.D.Hepatology CenterMassachusetts General HospitalBoston, Massachusetts

MARINA CUCHEL, M.D., PH.D.Department of MedicineUniversity of PennsylvaniaPhiladelphia, Pennsylvania

Contributors

xi

ANGELA M. DEVLIN, PH.D.Department of PediatricsBritish Columbia Research Institute for

Child and Women’s HealthUniversity of British ColumbiaVancouver, British Columbia, Canada

JEFFREY C. FAHLDepartment of PediatricsSchool of MedicineUniversity of New MexicoAlbuquerque, New Mexico

SCOTT A. FINK, M.D., M.P.H.Division of GastroenterologyDepartment of MedicineBrigham and Women’s HospitalBoston, Massachusetts

ARMANDO FLOR-CISNEROS, M.D.Bone and Extracellular Matrix BranchNational Institute of Child Health and

Human DevelopmentNational Institutes of HealthBethesda, Maryland

MICHAEL T. GERAGHTY, M.B., MRCPI,FACMG, FRCPSC, FCCMG

Children’s Hospital of Eastern OntarioUniversity of OttawaOttawa, Ontario, Canada

LESA R. GILBERT, R.N.Department of MedicineUniversity of FloridaGainesville, Florida

ROBERT H. GLEW, PH.D.Department of Biochemistry and

Molecular BiologySchool of MedicineUniversity of New MexicoAlbuquerque, New Mexico

VENKAT GOPALAN, PH.D.Department of BiochemistryCollege of Biological SciencesOhio State UniversityColumbus, Ohio

VIJAYAPRASAD GOPICHANDRAN, MBBSDepartment of BiochemistryCollege of Biological SciencesOhio State UniversityColumbus, Ohio

PAUL HARMATZ, M.D.Clinical Research CenterChildren’s Hospital and Research Center

at OaklandOakland, California

WILLIAM C. HINESDepartment of Biochemistry and Molecular


IAN R. HOLZMAN, M.D.Department of PediatricsMount Sinai School of MedicineNew York, New York

M. MAHMOOD HUSSAIN, PH.D.Departments of Anatomy, Cell Biology,

and PediatricsState University of New York DownstateMedical CenterBrooklyn, New York

HIROSHI IDEGUCHI, M.D.Department of Laboratory MedicineSchool of MedicineFukuoka UniversityFukuoka, Japan

ALLAN S. JAFFE, M.D.Cardiovascular DivisionDepartment of Internal Medicine andDepartment of Laboratory Medicine

and CardiologyMayo ClinicRochester, Minnesota

ABIODUN O. JOHNSON, M.B., B.S., M.D.Department of PediatricsTexas Tech University Health Sciences

CenterAmarillo, Texas

xii Contributors

RICHARD KACZMARSKI, M.D., FRCP,FRCPATH

Department of HaematologyHillingdon HospitalUxbridge, United Kingdom

JUTTA KELLER, M.D.Department of MedicineIsraelitic HospitalHamburg, Germany

LIEN B. LAI, PH.D.Department of BiochemistryCollege of Biological SciencesOhio State UniversityColumbus, Ohio

PETER LAYER, M.D., PH.D.Department of MedicineIsraelitic HospitalHamburg, Germany

SERGEY LEIKIN, PH.D.Section on Physical BiochemistryNational Institute of Child Health and

Human DevelopmentNational Institutes of HealthBethesda, Maryland

STEVEN R. LENTZ, M.D., PH.D.Department of Internal MedicineCarver College of MedicineUniversity of IowaIowa City, Iowa

DENIS M. MCCARTHYDepartment of Internal MedicineSchool of MedicineUniversity of New MexicoAlbuquerque, New Mexico

MARGARET M. MCGOVERN, M.D., PH.D.Department of Human GeneticsMount Sinai School of MedicineNew York, New York

J. ROSS MILLEY, M.D., PH.D.Division of NeonatologyDepartment of PediatricsUniversity of Utah School of Medicine

STEVEN M. MITCHELLSchool of MedicineUniversity of New MexicoAlbuquerque, New Mexico

KOJI NARAHARADepartment of PediatricsOkayama Red Cross HospitalOkayama, Japan

SHINSUKE NINOMIYADepartment of PediatricsOkayama University Medical SchoolOkayama, Japan

MARCY P. OSGOOD, PH.D.Department of Biochemistry and

Molecular BiologySchool of MedicineUniversity of New MexicoAlbuquerque, New Mexico

SRINIVAS PANJA, M.D.School of MedicineUniversity of New MexicoAlbuquerque, New Mexico

LAWRENCE M. PASQUINELLI, M.D.Department of PediatricsEastern Virginia Medical SchoolNorfolk, Virginia

HARBHAJAN S. PAUL, PH.D.Biomed Research & Technologies, Inc.Wexford, Pennsylvania

GERALD J. PEPE, PH.D.Department of Physiological SciencesEastern Virginia Medical SchoolNorfolk, Virginia

VIRGINIA K. PROUD, M.D.Department of PediatricsEastern Virginia Medical SchoolNorfolk, Virginia

KEITH QUIROLO, M.D.Department of HematologyNorthern California Sickle Cell CenterChildren’s Hospital and Research Center


Contributors xiii

DANIEL J. RADER, M.D.University of Pennsylvania

Medical CenterPhiladelphia, Pennsylvania

PAUL RAVA, B.S.Departments of Anatomy, Cell Biology,

and PediatricsState University of New York DownstateMedical CenterBrooklyn, New York

MIRIAM D. ROSENTHAL, PH.D.Department of Physiological SciencesEastern Virginia Medical SchoolNorfolk, Virginia

SARAH JANE SCHWARZENBERG, M.D.Department of PediatricsUniversity of Minnesota School

of MedicineMinneapolis, Minnesota

GAIL SEKASBiomed Research & Technologies, Inc.Wexford, Pennsylvania

HARVEY L. SHARPDepartment of PediatricsUniversity of Minnesota School

of MedicineMinneapolis, Minnesota

PETER W. STACPOOLE, M.D., PH.D.General Clinical Research CenterUniversity of FloridaGainesville, Florida

DAVID L.VANDERJAGT, PH.D.Department of Biochemistry and Molecular


DOROTHY J.VANDERJAGT, PH.D.Department of Biochemistry and Molecular


ELLIOTT VICHINSKY, M.D.Northern California Sickle Cell CenterChildren’s Hospital and Research Center


EMORN WASANTWISUT, PH.D.Institute of NutritionMahidol UniversityBangkok, Thailand

BARRY WOLF, M.D., PH.D.Connecticut Children’s Medical CenterUniversity of Connecticut School of

MedicineHartford, Connecticut

NUTTAPORN WONGSIRIROJ, M.A.Institute of Human NutritionCollege of Physicians and SurgeonsColumbia UniversityNew York, New York

YUJI YOKOYAMA, M.D.Department of PediatricsOkayama University Medical SchoolOkayama, Japan

xiv Contributors

Part I

Nucleic Acids and ProteinStructure and Function

CASE REPORT

The patients were brothers, aged 2 years 9months and 1 year 8 months, and were referredto our hospital for evaluation of developmentaldelay.

The Elder Brother

The elder brother was born at term when hismother was 22 years old and his father 33.Pregnancy and delivery were uneventful, andbirth weight was 3260 g. Neonatal screeningfor metabolic diseases and hypothyroidism wasunremarkable. However, gross psychomotor re-tardation became apparent; head control wasachieved at the age of 5 months, sitting un-aided at 12 months, and walking began at 20months, but the child had not acquired anymeaningful words at the time of examination(2 years 9 months).

Physical examination revealed a hyperactiveboy with a height of 94.5 cm, weight of13.2 kg, and head circumference of 47.8 cm(Fig. 1-1a). His face was somewhat long andsquare with a high forehead and large, promi-nent ears. The left lower incisor was absent,and the testes were not enlarged. Psychometrictesting revealed developmental quotient to be38% that of a normal child of the same age. Thechild exhibited unique behavioral abnormali-ties characterized by hyperactivity, short atten-tion span, poor eye contact, and excessivewithdrawal response to strange people or envi-ronments; however, these behaviors did notmeet the Diagnostic and Statistical Manualof Mental Disorders, Fourth Edition (DSM-IV)

criteria for childhood autistic disorders. Elec-troencephalography (EEG) and acoustic brain-stem response were normal.

The Younger Brother

The younger brother was born at the 36th weekof gestation with a birth weight of 2780 g. Mod-erate psychomotor retardation was noted; heachieved head control at the age of 4 months,sat unaided at 10 months, and could stand upholding onto furniture at 20 months but had notacquired any meaningful words at the time ofexamination (1 year 8 months).

Physical examination revealed a hyperactiveboy with normal height and head circumfer-ence. His face was somewhat square with aprominent forehead,everted ears,and absent leftlower incisor, like his elder brother (Fig. 1-1b).Developmental quotient was assessed as 47% ofnormal. He had a short attention span but nosocial aversion or hand mannerisms.At the ageof 20 months, EEG and acoustic brainstem re-sponse were unremarkable; however, from 18months, massive epileptic discharges had be-come evident in the bilateral parieto-occipitalregions during sleep.

The Family

Investigation of the patients’ family (Fig. 1-2) re-vealed the presence of mental retardation inone maternal aunt ( II-6, closed circle). She hadbeen born at term with a birth weight of3080 g. The pregnancy had been complicatedby upper gastrointestinal radiographic expo-sure in the first trimester. While her motor

1

Fragile X Syndrome

YUJI YOKOYAMA, SHINSUKE NINOMIYA, and KOJI NARAHARA

3

development was reported to be almost normal,speech was grossly delayed, and she had at-tended special educational schools. Anticon-vulsant drugs were administered for EEGabnormalities. Menstruation started at the ageof 13 years and remained regular. Physical ex-amination at age 20 years revealed a shy girl

with mental retardation. Height and head cir-cumference were normal although she had asomewhat long face with prognathism; how-ever, her ears were not malformed. Speech dis-turbance was characterized by lack of fluency,echolalia, and inappropriate grammar. Psycho-metric testing demonstrated an intelligence

4 NUCLEIC ACIDS AND PROTEIN STRUCTURE AND FUNCTION

Figure 1-1. The two patients, the older brother, aged 2 years 9 months (left), and theyounger brother, aged 1 year 8 months (right), showing the square faces, high forehead,and large everted ears.

Figure 1-2. Family pedigree of the patients. Solid squares and circles indicate subjectswith mental retardation, and circle with hatched lines denotes a person with borderlineintelligence.

quotient (IQ) of 45 (normal range, 77–124). Theremaining maternal siblings appeared to be in-tellectually normal, although psychometric testswere not performed. Other than for one uncle(II-1) who left high school to work as a truckdriver, no family history of education problemswas reported. The mother’s IQ was assessed at92 on the Wechsler Adult Intelligence Scale. Thematernal grandmother (I-2) was the sixth of aseven sibship and denied the presence of men-tal retardation in her brothers and sisters.

The mother became pregnant again 6 monthsafter the first visit regarding her two sons. Sherefused prenatal fetal diagnosis on religiousgrounds, and a female infant weighing 3462 gwas delivered at term by cesarean section. Psy-chomotor development of this infant was almostnormal with the exception of speech; she beganto speak a few meaningful words at 2 years 10months.Examination revealed normal height andhead circumference. She had a somewhat squareface but otherwise appeared normal. Her intelli-gence was assessed at 85% that of a normal child.

DIAGNOSIS

The two probands presented with moderate-to-severe developmental retardation not associated

with any recognizable malformation syndromes.Dysmorphic features included a long, squareface, large prominent ears, and prominent fore-head. The elder brother exhibited unique be-havioral abnormalities and appeared to displayautistic traits.Borderline mental retardation wasalso evident in the younger sister.Although theexistence of mental retardation in one maternalaunt was difficult to explain, the familial diseasewas thought to be consistent with X-linked in-heritance with low penetrance (the frequencywith which a heritable trait is manifested bythose carrying the affected gene) in females.Be-cause fragile X syndrome represents about 40%to 50% of all forms of X-linked familial mentalretardation, a diagnostic workup was initiated.

Cytogenetic studies have indicated that frag-ile X syndrome is associated with a rare fragilesite at Xq27.3 (Fig.1-3) and is caused by a muta-tion in the fragile X mental retardation 1(FMR1) gene. Although diagnosis of the disor-der should be based on molecular studies,cytogenetic studies are useful to exclude subtleabnormalities of sex and autosomal chromo-somes frequently associated with nonspecificmental retardation. The fragile site, fra(X)(q27.3), can be induced under specific cultureconditions (e.g., use of a medium deficient infolate or supplemented with methotrexate,

Fragile X Syndrome 5

Figure 1-3. Partial karyotypes of Giemsa-stained human chromosomes showing variousmanifestations of the fragile X site (arrows): a chromatid break (a), an isochromatid gap(b), a chromosome break (c), and endoreduplication (d).

fluorodeoxyuridine, or excess thymidine duringthe final 24 hours of culture to disturb folatemetabolism).As shown in Table 1-1, the two pa-tients ( III-1 and III-2), sister ( III-3), mother (II-3), and aunt (II-6) all expressed fra(X)(q27.3),whereas the remaining family members did notdemonstrate this fragile site.

To determine whether selective inactivationof X chromosomes could be related to pheno-typic differences in the female carriers withfra(X)(q27.3), we studied DNA replication pat-terns of the X chromosome using bromod-eoxyuridine (thymidine analogue) incorporationduring the final 7 hours of incubation. Thesestudies demonstrated that the ratio of inactivatedlate-replicating X chromosomes (those incorpo-rating large amounts of bromodeoxyuridine intoDNAs) carrying fra(X)(q27.3) did not differ sig-nificantly among the three female carriers.

The fragile X mutation involves expressionof a CGG repeat stretch in the 5' encoding re-gion of FMR1. Hypermethylation (a chemicalalteration of DNA induced by methylation ofdeoxycytidine to 5-methyldeoxycytidine in theCG sequence where cellular regulation of geneexpression or X inactivation is believed to oc-cur) of a regulatory CpG island (a DNA regionof high deoxycytidine and deoxyguanosinecontent linked by a phosphodiester bond lo-cated near the protein-coding region of a geneand related to its regulation of gene expression)just upstream of the CGG repeat segment andnonexpression of FMR1 protein.

We investigated the length of the CGG repeatsegment and the methylation status of the CpGisland in the patients’ family using Southernblot hybridization with pPCRFX1 (a kind ofPfxa3) as a DNA probe that detects restrictionfragments containing the CGG repeat (Fig. 1-4).Digestion with a methylation-insensitive restric-tion enzyme, EcoRI generates a 5.2-kilobase(kb) fragment containing the mutable region.This fragment may be further digested by amethylation-sensitive restriction enzyme EagIinto 2.4-kb and 2.8-kb fragments if the EagI siteis not methylated. Normal males demonstrate a2.8-kb band, whereas normal females will havea 2.8-kb band and a 5.2-kb band. The two pa-tients ( III-1 and III-2) exhibited an indistinctsmear band (i.e., a dispersed expansion) rang-ing from 6 to 9 kb (0.7- to 4-kb expansion of theCGG repeat segment), and the aunt ( II-6) alsohad an indistinct smear band ranging from 6 to9 kb in addition to the 2.8- and 5.2-kb normalbands. The grandmother (I-2) and mother dis-played additional 2.9-kb and 3.8-kb bands, re-spectively. Furthermore, the maternal uncle(II-1) had a 3.1-kb band instead of a normal 2.8-kb band. Intriguingly, the sister ( III-3) exhibitedan additional 4.0-kb band. The other familymembers showed normal blotting patterns.From the results of this analysis, the two pa-tients and the aunt were diagnosed as havingfragile X syndrome, and the mother, sister, un-cle, and maternal grandmother as fragile X pre-mutation carriers.


Table 1-1. Cytogenetic Studies of the FamilyIntelligence Fra(X)(q27.3) Ratio of Inactivated

Subject Sex Age (yr) Quotient Expression (%) fra(X) Positive X (%)

I-1 M 59 Normal* 0 ND

I-2 F 53 Normal 0 ND

II-1 M 27 Normal 0 ND

II-2 M 35 Normal 0 ND

II-3 F 25 92 9 48

II-4 F 23 Normal 0 ND

II-5 M 21 Normal ND ND

II-6 F 20 45 19 53



III-1 M 2 38 28 ND

III-2 M 1 47 32 ND

III-3 F 2 85 18 58

ND, not determined.

*Normal range: 77–124.

MOLECULAR PERSPECTIVES

Fragile X Syndrome

The incidence of mental retardation is wellknown to be higher in males than in females. In1943, Martin and Bell (1943) reported a familialmental retardation consistent with X-linked in-heritance, and this family has now been demon-strated to be the first example of individualswith fragile X syndrome. Increasing interest inthe etiological and biological mechanisms ofmental retardation has stimulated much re-search in this area in the decades since the late1960s. In 1969,Lubs (1969) described a markerX (a constriction in the distal long arm of the Xchromosome) present in affected males and ob-ligate carrier females of a family with X-linkedmental retardation; however, this finding wasnot readily reproduced by standard cytogenetictechniques. It was not until 1977 that Suther-land (1977) reported that the induction of rarefragile sites, including marker X, is dependenton the composition of medium used for cultur-ing peripheral blood lymphocytes. This condi-tion was termed fragile X syndrome because,among all rare fragile sites, only the expressionof the fragile in band Xq27.3 (FRAXA) is associ-ated with clinical disease (Sutherland andRichards, 1994).

The establishment of cytogenetic methodsfor detecting fragile X syndrome promptedepidemiologic studies of its prevalence in vari-ous populations. The syndrome occurs in allethnic groups and affects approximately 1 in1250 males and 1 in 2500 females, accountingfor around 20% of all familial mental retarda-tion. This estimated prevalence is comparableto that of Down syndrome (1 in 800–1000) asa cause of mental retardation. However, deter-mination of frequency in these studies hasbeen based on cytogenetic detection, and thetrue prevalence may be even higher using anewly available molecular test for populationscreening.

The phenotypic features of fragile X syn-drome in relation to puberty are summarized inTable 1-2. Although the syndrome is apparentfrom birth in affected patients, it is difficult todiagnose during early infancy.Prepubertal malestend to exhibit only nonspecific clinical find-ings, and characteristic physical features becomeobvious with age. Typical postpubertal maleswith fragile X syndrome exhibit a clinical triadof the so-called Martin-Bell syndrome: mentalretardation, long face with large everted ears,and macro-orchidism (abnormally large testes).Other craniofacial features include prominentjaw, large forehead, and relative macrocephaly.Additional features are suggestive of connective


Figure 1-4. Southern hybridization analysis of the family using double digestion ofDNAs with EcoRI and EagI.

tissue dysplasia: hyperextensible joints, mitralvalve prolapse (allowing retrograde flow intothe left atrium), and dilatation of the ascendingaorta (aortic root dilatation). Developmental de-lay and mental retardation are the most signifi-cant and prominent symptoms of fragile Xsyndrome. Most male patients have IQ scores inthe 20 to 60 range, with an average of 30 to 45.In particular,prepubertal boys with fragile X ex-hibit characteristic behavioral abnormalities, in-cluding hyperactivity, short attention span,emotional instability, hand mannerisms, andautistic features.

The physical and behavioral features of thedisease in female patients are usually milderthan in affected males. Somatic features may beabsent or mild, although the faces of mentallyretarded females tend to resemble those ofmale patients with advancing age. The intelli-gence deficit of female patients is less severe,with most patients having mild-to-borderlinemental impairment. There is evidence of an in-crease in psychological and psychiatric prob-lems among female patients.

The inheritance pattern of fragile X syn-drome is unusual. While about 80% of maleswho inherit the mutation exhibit mental retar-dation and a more or less definitive phenotype,the remaining 20% of carrier males are pheno-typically normal. Such clinically normal hemizy-gous males are termed transmitting malesbecause the mutation is transmitted throughtheir unaffected daughters to grandchildren,

who often manifest this syndrome. The risk ofmental retardation in grandchildren is 74% formales and 32% for females but is much loweramong siblings of transmitting males (18% formales and 10% for females). Male offspring ofmentally impaired carrier mothers have a higherrisk of mental retardation (100%) than do fe-male offspring (76%). The large variation in riskof mental retardation in fragile X families con-taining transmitting males cannot be explainedby classic genetics and is termed the Shermanparadox after its discoverer (Sherman et al.,1985).

Because the cytogenetic approach is of lim-ited value in detecting transmitting males andcarrier females, efforts to identify and character-ize a putative fragile X gene were undertaken inmany molecular genetic laboratories. The associ-ation of the fragile site Xq27.3 with this form ofX-linked mental retardation suggested the puta-tive gene is located at or near the fragile X site.Tarleton and Saul (1993) described how posi-tional cloning of the fragile X site was achieved.In addition to conventional analysis of restrictionfragment length polymorphisms, new moleculartools,such as pulsed field gel electrophoresis andthe yeast artificial chromosome, were used to de-fine and isolate this region. The yeast artificialchromosome can accommodate large DNA frag-ments from species other than yeasts, facilitatingthe cloning of a gene of interest,while restrictionfragment length polymorphisms provide usefulmolecular landmarks on chromosomes, thereby


Table 1-2. Clinical Features in Males with Fragile X SyndromePrepubertal

Birth weight: a mean at approximately the 70th percentileHeight: mostly between 50th and 97th percentilesHead circumference: slightly increasedDevelopmental delay: sit alone at 10 months, walk at 20.6 months, first meaningful words at 20 monthsAbnormal behavior: hyperactivity, hand mannerisms, excessive shyness, tantrum, autism

PostpubertalMental retardationCraniofacial features: prominent forehead, prominent jaw, large prominent ears, long faceMacro-orchidism

Additional featuresOrthopedic: joint hyperextensibility, flat feet, torticollis (a contracted state of the neck), kyphoscoliosis (backward and

lateral curvature of the spinal column)Ophthalmologic: myopia and strabismusCardiac: mitral valve prolapse and dilatation of ascending aortaDermatologic: fine velvety skin with striaeGenitourinary: cryptorchidism (failure of the testes to descend into the scrotum) and inguinal herniaOthers: epilpsy, hyperreflexia, gynecomastia (excessive development of the male mammary glands)

enabling segregation analysis and risk assess-ment of the probability of inheriting a disease(linkage analysis).

The FMR1 Gene

In 1991, several groups of investigators re-ported almost simultaneously that the mutationresponsible for fragile X syndrome was an ex-pansion of the trinucleotide sequence CGG (orCCG) within a gene termed fragile X mental re-tardation 1 (FMR1) (Oberle et al., 1991;Verkerket al., 1991;Yu et al., 1991). The FMR1 gene en-compasses 38 kb on the X chromosome at theposition of the fragile site and it comprises 17exons. The triplet repeat of sequence CGG lieswithin the 5' untranslated region of the firstexon, 69 base pairs (bp) upstream from theinitiation codon and 250 bp downstream fromthe CpG island regulatory gene (Fig. 1-5).

This microsatellite repeat is polymorphic innormal humans, ranging from 6 to 52 repeats,with a mean of 30. In affected patients withfragile X syndrome, however, this repeat con-tains many times the normal number of tripletrepeats: between 230 and several thousand

copies of CGG.When the trinucleotide repeatsexceed 230 copies, they are chemically modi-fied in such a way that the FMR1 gene will nolonger function. The deoxycytidines withinthe repeats become methylated, producing 5-methyldeoxycytidines.These methylation eventsextend upstream into the regulatory CpG is-land, which is normally unmethylated, and pre-vent the gene from being expressed.Virtually allaffected patients lack detectable FMR1 mRNA,and the loss of FMR1 function as a result of thesuppression of transcription is believed to be thecause of fragile X syndrome. Three instances ofnon-CGG mutations of the FMR1 gene, includingdeletions of the FMR1 locus and a missense mu-tation involving the critical domain of FMR1 inpatients with apparent fragile X syndrome, haveprovided further supporting evidence for thishypothesis. It should be emphasized that a muta-tion resulting from triplet expansion has notbeen recognized as a cause of human geneticdisease.

Although its complete sequence is known,the exact function of the FMR1 gene has notyet been defined. The FMR1 gene has proper-ties of a housekeeping gene; it is expressed in


Figure 1-5. Diagram of the fragile X mental retardation (FMR1) gene with restrictionmap and FMR1 probes used for diagnostic Southern blots. The circle indicates the CpGisland, and the box represents the first exon. The dark region shows the location oftriplet repeats.

diverse tissues and exhibits DNA sequencesthat are highly conserved in other species.Alter-nate splicing produces a considerable numberof mRNA molecules.As would be expected, thegene is most intensely transcribed in both thebrain and testes. Its protein product (fragile Xmental retardation protein), which is predomi-nantly cytoplasmic, has multiple functional do-mains, including two types of RNA-bindingdomain, a nuclear export signal, and a nuclearlocalization signal (Eberhart et al., 1996; Siomiet al., 1993). Extinction of an interaction be-tween fragile X mental retardation protein anda subset of brain mRNA in neurons is thoughtpotentially to play an important role in the neu-rological manifestation of fragile X syndrome.Moreover, autopsies of patients with fragile Xhave revealed defects in neurite density andmorphology (Irwin et al., 2001), suggesting thatFMR1 may play a role in neurite branching.

Drosophila has proven to be a good modelof fragile X syndrome since this species con-tains a single homolog of FMR1, dfxr (alsocalled dfmr1).DFXR, the protein product of thedfxr gene, is expressed in brain neurons but notin glia. Loss of DFXR function results in markedloss of neurite extension, and irregular branch-ing, and axon guidance defects in dorsal clusterneurons.The lateral neurons show variable de-fects in extension and guidance (Morales et al.,2002). The dfxr mutant alleles were found tocause failure of adult eclosion and disorderedcircadian rhythm (Dockendorff et al., 2002).DFXR constructs a complex that includes tworibosomal proteins, L5 and L11, along with 5SRNA, and Argonaute 2 (AGO2), which is an es-sential component of the RNA-induced silenc-ing complex that mediates RNA interference(RNAi) in Drosophila (Hammond et al., 2001)and a Drosophila homolog of p68 RNA helicase(Dmp 68) (Ishizuka et al., 2002). It is possiblethat RNAi and DFXR-mediated translational con-trol pathways intersect, and that RNAi-relatedmachinery plays an important role in the con-trol of neural function.

Fragile X families exhibit two types of FMR1gene mutation. The repeat expansion of morethan 230 copies with subsequent methylationof the CpG island is referred to as a full muta-tion. All males and about half of the femaleswho carry full mutations have mental retarda-tion. Mosaic males with full mutations are al-most always affected to the same extent as fullyaffected males, while mosaic females vary in

clinical phenotype. The mosaic state is thoughtto reflect different degrees of expansion orDNA methylation in somatic cells. The othermutation, in which the repeat ranges from 50 to230 copies, is termed a premutation. Becausepremutations are not methylated and are tran-scriptionally active, phenotypic abnormalitiesdo not occur in any male or female carriers withthis type of mutation. However, it should be un-derstood that no precise number of copiesmarks the transition from the normal chromo-some to premutation or from premutation tofull mutation. In general, geneticists have agreedto define a copy number between 50 and 230as premutation and one of more than 230 as fullmutation.

The most prominent characteristic of theCGG repeat is the variation in its length. Be-cause expansion occurs after conception, therange of repeat expansion varies in differentcells from the same tissue in the same affectedperson.This variation is particularly prominentwhen the expanded repeat is transmitted frommother to child.When women transmit the re-peat to offspring of either sex, the sequenceusually increases in size (although it has beenknown to decrease); however, when transmit-ted by males, sequence size either remains con-stant or decreases. As males do not transmitmore than 230 copies of the repeat, theirdaughters do not have fragile X syndrome.Thismeans that even an affected male with a fullmutation in nearly all of his cells may be essen-tially within the premutation range with re-spect to the repeat number in his sperm. Nonew mutation from the normal number of re-peats has been seen in fragile X syndrome, anda complete family investigation always identi-fies a premutation in one of the ancestral gen-erations. It is likely that small premutationsmay have segregated through many genera-tions before a further repeat expansion oc-curred.

The Sherman paradox (Sherman, 1991; Sher-man et al., 1985) was resolved by analyzing theFMR1 gene in fragile X families with transmit-ting males (Fu et al., 1991). Transmitting malesalways have premutations, and the daughters oftransmitting males inherit about the same num-ber of CGG repeats as found in their fathers.The premutations become unstable after ooge-nesis (the process of gamete formation) in thedaughters, leading to full mutations with overseveral hundred CGG repeats in their offspring.


Because the mothers of transmitting maleshave copy numbers of CGG repeats in thelower end of the carrier range (50–70), broth-ers of transmitting males are much less likely tohave full mutations than premutations.Premuta-tions larger than 80 CGG repeats, however, al-most always expand into the full mutationrange when passed through mothers.Therefore,the Sherman paradox indicates that the variationin the propensity of premutations to become fullmutations may be related to the size of the pre-mutation and the gender of the carrier.

The fragile X site is expressed when the CGGrepeat is expanded to a copy number higherthan 230. The expression of the fragile site isthought to be the result of an incomplete DNAreplication in the expanded region caused bydepletion of intracellular pools of dCTP anddGTP under specific culture conditions. How-ever, the enormous expansion of CTG tripletsin myotonic dystrophy, another genetic diseasecharacterized by trinucleotide repeat expan-sion, has never been associated with any visiblefragile site, suggesting that the nucleotide com-position of the amplified repeats is also crucialto the expression of the fragile site. Unlike CTGrepeats, CGG repeats undergo methylation,which might stabilize tetraplex DNAs formedby CGG tracts. These stable tetrahelical struc-tures could suppress transcription, replication,

and chromatin condensation, leading to genera-tion of the fragile site.

Now that the molecular basis of the fragile Xsyndrome has been defined and characterized,exclusion of this disorder on clinical or cytoge-netic grounds is no longer warranted. Once achild is identified with this syndrome, familymembers should be evaluated to detect individ-uals at risk of having affected children and to fa-cilitate decisions about future reproduction.

Southern Blotting

Molecular diagnosis of fragile X syndrome isnow possible using Southern hybridization andPCR methods (Brown et al., 1993; Rousseauet al., 1991). Southern hybridization is the diag-nostic method of choice because it can deter-mine the extent of CGG repeat expansion aswell as the methylation status of the CpG is-land. The choice between restriction enzymeand probe depends on the diagnostic informa-tion expected (Table 1-3). Cleavage with PstIand hybridization with a Pfxa3 probe is suitablefor detecting small premutation alleles. To ex-amine the methylation status and CGG repeatlength simultaneously, double digestion with amethylation-sensitive enzyme, such as BssHII orEagI, can be used (see Fig. 1-5).A 5.2-kb band isobserved from the inactive X, and two smaller


Table 1-3. RFLP Patterns by Southern Blot Analysis from Normal Individuals, Premutation Carriers, andPatients Affected with Fragile X Syndrome*

DNA Digestion With EcoRI + EagI or BssHII Hybridization With PstI

pE5.1 Pfxa3 or StB12.3 Pfxa3

Normal (6–50 repeats)Male 2.4 + 2.8 2.8 1.0Female 2.4 + 2.8 + 5.2 2.8 + 5.2 1.0

Premutation carriers(50–230 repeats)Male 2.4 + (2.9–3.3) (2.9–3.3) (1.1–1.6)Female 2.4 + 2.8 + (2.9–3.3) + 5.2 2.8 + (2.9–3.3) + 5.2 1.0 + (1.1–1.6)

Patients withfull mutations(> 230 repeats)Male >5.7 >5.7 >1.6Female 2.4 + 2.8 + 5.2 + >5.7 2.8 + 5.2 + >5.7 1.0 + > 1.6

Mosaic patientsMale (2.9–3.3) + >5.7 (2.9–3.3) + >5.7 (1.1–1.6) + >1.6Female 2.4 + 2.8 + (2.9–3.3) + 5.2 + >5.7 2.8 + (2.9–3.3) + 5.2 + >5.7 1.0 + (1.1–1.6) + >1.6

RFLP, restriction fragment length polymorphism.

*Sizes of bands are expressed in kilobases.

bands (2.8 and 2.4 kb) are observed from theactive X of the female and the single X of a nor-mal male. As the CGG repeat lies in the 2.8-kbband, males with premutations show a bandslightly larger than 2.8 kb, corresponding to anincrease in the repeat length. Males with fullmutations demonstrate a band larger than5.2 kb, reflecting the methylated and expandedFMR1 mutation. Females with premutations ex-hibit the three bands seen in the normal femalepattern (unmethylated active state of 2.4 and2.8 kb and methylated, inactive state of 5.2 kb)plus one additional premutation band, whichsometimes merges into the normal bands. Infull-mutation females, the expanded CGG re-peat is always overmethylated, and a smearband in excess of 5.7 kb can be seen in addi-tion to the normal female bands. The interpre-tation of data in mosaic female patients ismore complex because the pattern of bandsreflects the methylated and unmethylatedstates of both normal and abnormal X chromo-some alleles.

Polymerase Chain Reaction

The PCR approach is particularly useful when amore accurate determination of CGG repeatnumbers is necessary in normal or premutationcarriers. Initial attempts to analyze the fragile Xmutation by PCR were not successful owing tothe difficulty in amplifying DNA regions with ahigh CG content, the preferential amplificationof the smallest allele in females, and the failureto amplify full mutations. These disadvantageshave been partially overcome by the substitutionof 7-deaza-dGTP for dGTP, the use of improvedprimers, and the introduction of sequencingacrylamide gels.The advantages of PCR are thatit is rapid and requires only minimal amounts ofDNA. It will likely become the technique ofchoice in the diagnosis of fragile X syndrome ifa method that can reliably amplify full muta-tions is devised.

PRENATAL DIAGNOSIS

Because no effective therapy is available, prena-tal diagnosis of fragile X is of prime importancein pregnancies of female carriers who are atrisk of having affected children. Cytogeneticanalysis no longer has a place in the prenatal di-agnosis of fragile X syndrome. Prenatal diagno-sis can be accomplished by analyzing DNA

obtained by chorionic villus (a villus on the ex-ternal surface of the chorion:fetal tissue) sam-pling using Southern blot analysis or, morerecently, using PCR, which can detect the num-ber of CGG repeats.

Male fetuses with 50 to 230 copies of the re-peat should be asymptomatic, whereas thosewith more than 230 copies will have fragile Xsyndrome.Female fetuses with 50 to 230 copiesalso will be asymptomatic; however, it is diffi-cult to predict the extent of mental retardationin female fetuses with more than 230 copies ofthe repeat. Although hypermethylation of theCpG island is a poor prognostic indicator, itis not always present in DNA extracted fromchorionic villus samples (Sutherland et al.,1991). Empiric data showing that female carri-ers with full mutations have nearly a 50% riskof mental impairment should be consideredreliable.

GENETIC DISEASES ASSOCIATEDWITH DYNAMIC MUTATIONS

Fragile X syndrome was the first of 12 humangenetic diseases in which dynamic mutation ofthe trinucleotide repeat was identified as thecause (Table 1-4). In these diseases, the se-quence of the trinucleotide repeat and the ef-fect of the expansion on the function of thegene in which it resides can differ. Genetic an-ticipation (a phenomenon in which the diseasehas an earlier age of onset and becomes in-creasingly severe in succeeding generations) isa common feature in these diseases and can beexplained by the expansion of the repeatwhen transmitted from parent to child. Gen-der bias regarding the parent contributing themost severe form of the disease is evident insome of these disorders. For example, the formof myotonic dystrophy that is apparent frombirth occurs only in children who have inher-ited the mutation from their mother. In con-trast, the juvenile-onset forms of Huntingtondisease and spinocerebellar ataxia type I de-velop primarily when the mutation is transmit-ted from the father. It should be noted thatexpression of another fragile site, FRAXE (frag-ile site, X chromosome, E site), has a similar ge-netic mechanism to fragile X syndrome: anexpansion of the CGG repeat and methylationof the CpG island, resulting in mental retarda-tion. In contrast to fragile X syndrome, the re-peat number in the FRAXE can expand or


Table 1-4. Genetic Diseases Associated with Dynamic MutationsSex Bias of Parent No. of Copies

Chromosome Repeated Contributing Normal No. Associated with Disease Location Sequence Severe Form of Copies the Diseases Gene Product

Fragile X syndrome Xq27.3 CGG Maternal 6–50 Premutation: 50–230 FMR1Full mutation: 230–2000

Fragile XE syndrome Xq28 CGG (—) 6–25 Premutation: 25–200 FMR2Full mutation: >200

Spinobulbar muscular atrophy Xq11-12 CAG ? 11–31 40–62 Androgen receptor

Huntington disease 4p16.3 CAG Paternal 9–37 Premutation: 30–38 HuntingtinFull mutation: 37–121

SCA1 6p22-23 CAG Paternal 25–36 43–81 Ataxin 1

SCA2 12q24.1 CAG Paternal 15–24 35–59 Ataxin 2

Machado-Joseph disease 14q32.1 CAG Paternal 13–36 68–79 Ataxin 3

SCA6 19p13 CAG ? 5–20 21–30 Ataxin 6

SCA7 3p12-21.1 CAG ? 7–18 37–200 Ataxin 7

DRPLA 12p13.31 CAG Paternal (mainly) 7–23 49–75 Atrophin

Friedreich’s ataxia 9q13 GAA Maternal 30–40 200–900 Frataxin

Myotonic dystrophy 19q13.3 CTG Maternal 5–35 Premutation: 50–80 Myotonin protein Full mutation: 80–2000 kinase

DRPLA, dentatorubral-pallidoluysian atrophy; SCA, spinocerebellar ataxia.

contract and is equally unstable when passedthrough the mother or father.

The molecular mechanism of repeat expan-sion in fragile X syndrome is not known. Link-age analyses of microsatellite markers flankingthe CGG repeat have suggested a founder effectin fragile X syndrome: Numerous full fragile Xmutations are derived from a few ancestral pre-mutations that could increase in the geneticpool due to their relative stability and selectiveneutrality. There is other evidence that the CGGrepeat of the FMR1 gene in normal individualsexhibits AGG interruptions, and that repeatswith documented unstable transmission havelost AGG interruptions (Eichler et al.,1994). Thissuggests that either DNA sequences flanking therepeat or variations in the repeat itself are in-volved in the mutation mechanism. The massiveexpansion of triplet repeats associated with

fragile X syndrome, when transmitted from aparent with more than approximately 80 copiesof the repeat, cannot be explained by simple re-combination. Okazaki fragment slippage (thetendency for a single-strand DNA with free endscaused by two breaks slides along a templatestrand, resulting in a greater likelihood of muta-tion after DNA replication) has been proposedas a possible mechanism for such rapid expan-sion (Fig. 1-6) (Richards and Sutherland, 1994).

THERAPY

Because no specific treatment for fragile Xsyndrome is available, medical, physical, andoccupational interventions are directed towardalleviating neurological and behavioral manifes-tations of the disorder (Hagerman, 1989). It is


Figure 1-6. Models for instability and hyperexpansion involving Okazaki fragmentslippage. For copy number below 80 (n = a), only one single-stranded break is likely tooccur within the repeat during replication (a). Slippage of the elongated strand duringpolymerization can result in the addition or deletion of a few copies (y). For copynumber above 80 (n = b), it is possible that two single-stranded breaks occur withinthe repeat in the process of replication (b). The strand between these breaks is not an-chored at either end by unique sequence and is therefore free to slide during poly-merization, enabling the addition of many more copies (z) than were present in theoriginal sequence (b) pending the outcome of the repair process. From Richards andSutherland (1994).

also important for parents to have contact withother fragile X families for further support andinformation. Medical management of fragile Xsyndrome includes pharmacological treatmentfor specific behavioral problems and follow-upof frequently encountered complications. Folicacid supplementation is no longer recom-mended for treatment of the intellectual andbehavioral deficits in fragile X syndrome sinceseveral studies have found that it is of no bene-fit. Central nervous system stimulants, such asmethylphenidate and dextroamphetamine,haveproved to be effective in improving the atten-tion span and learning performance of some hy-peractive fragile X children.

Educational interventions can be institutedafter diagnosis of the disorder have been madeand extensive genetic counseling of the familyhas been initiated. Teachers and therapistsshould create an educational program in keep-ing with neuropsychological characteristics offragile X patients. Fragile X patients have moredifficulty with auditory processing than with vi-sual processing, which relates to their atten-tional problems, impulsivity, and distractibility,thereby validating the use of central nervoussystem stimulants in fragile X patients. Calmingtechniques, such as deep breathing, relaxation,and music therapy, are sometimes effective inavoiding emotional upsets and outbursts innew situations or confusing circumstances. Thegoal of speech and occupational therapies is tohelp fragile X patients reach their intellectualpotentials.

QUESTIONS

1. What clinical features do prepubertalmale patients affected with fragile X syn-drome have?

2. Why is it important for medical personneland scientists to understand the molecu-lar basis of fragile X syndrome?

3. What is the Sherman paradox in fragile Xsyndrome, and how can this paradox beresolved on a molecular basis?

4. How would you go about informing themother (II-3) and the uncle (II-1) of thepatients regarding their risk of having chil-dren affected with fragile X syndrome in afuture pregnancy?

5. Both the aunt ( II-6) and the sister ( III-3)of the patients had fragile X expressionand apparent full mutation. Why was the

phenotype of the sister much milder thanthat of the aunt?

6. What other human genetic diseases havebeen attributed to dynamic mutation of atrinucleotide repeat?

Acknowledgment: We would like to thank Dr. Grant R.Sutherland (Center for Medical Genetics, Department ofCytogenetics and Molecular Genetics,Women’s and Chil-dren’s Hospital,Adelaide, South Australia) for reading themanuscript and for permission to use Figure 1-6.

BIBLIOGRAPHY

Brown WT, Houck GE, Jeziorowska A, et al.: Rapidfragile X carrier screening and prenatal diagno-sis using a non-radioactive PCR test. JAMA270:1569–1575, 1993.

Dockendorff TC, Su HS, McBride SMJ, et al.:Drosophila lacking dfmr1 activity show defectsin circadian output and fail to maintain courtshipinterest.Neuron 34:973–984,2002.

Eberhart DE, Malter HE, Feng Y, et al.: The fragile Xmental retardation protein is a ribonucleopro-tein containing both nuclear localization andnuclear export signals.Hum Mol Genet 5:1083–1091, 1996.

Eichler EE, Holden JJA, Popovich BW, et al.: Lengthof uninterrupted CGG repeats determines insta-bility in the FMR1 gene. Nature Genet 8:88–94,1994.

Fu YH, Kuhl DPA, Pizzuti A, et al.: Variation of theCGG repeat at the fragile X site results in gene-tic instability: resolution of the Sherman para-dox. Cell 67:1047–1058, 1991.

Hagerman R: Behaviour and treatment of the frag-ile X syndrome, in Davies KE (ed):The Fragile XSyndrome. Oxford, UK, Oxford University Press,1989, pp. 56–75.

Hammond SM, Boettcher S, Caudy AA, et al.: Arg-onaute 2, a link between genetic and biochemi-cal analyses of RNAi. Science 293:1146–1150,2001.

Irwin SA, Patel B, Idupulapati M, et al.: Abnormaldendritic spine characteristics in the temporaland visual cortices of patients with fragile-Xsyndrome: a quantitative examination. Am JMed Genet 98:161–167, 2001.

Ishizuka A, Siomi M, Siomi H: A Drosophila fragileX protein interacts with components of RNAiand ribosomal proteins. Genes Dev 16:2497–2508, 2002.

Lubs HA: A marker X chromosome. Am J HumGenet 21:231–244, 1969.

Martin JP, Bell J: A pedigree of a mental defectshowing sex-linkage. J Neurol Psychiatry 6:151–154, 1943.

Morales J, Hiesinger PR, Schroeder AJ, et al.:Drosophila fragile X protein, DFXR, regulates



neuronal morphology and function in the brain.Neuron 34:961–972, 2002.

Oberle I, Rousseau F, Heitz D, et al.: Instability of a550-base pair DNA segment and abnormalmethylation in fragile X syndrome. Science 252:1097–1102, 1991.

Richards RI, Sutherland GR: Simple repeat DNA isnot replicated simply. Nature Genet 6:114–116,1994.

Rousseau F, Heitz D, Biancalana V, et al.: Direct diag-nosis by DNA analysis of the fragile X syndromeof mental retardation. N Engl J Med 325:1673–1681, 1991.

Sherman S: Epidemiology, in Hagerman RJ, Silver-man AC (eds): Fragile X Syndrome: Diagnosis,Treatment, and Research. Baltimore, MD, JohnsHopkins Press, 1991, pp. 69–97.

Sherman SL, Jacobs PA, Morton NE, et al.: Furthersegregation analysis of the fragile X syndromewith special reference to transmitting males.Hum Genet 69:289–299, 1985.

Siomi H, Siomi MC, Nussbaum RL, et al.: The pro-tein product of the fragile X gene, FMR1, has

characteristics of an RNA-binding protein. Cell74:291–298, 1993.

Sutherland GR: Fragile sites on human chromo-somes: demonstration of their dependence onthe type tissue culture medium. Science 197:265–266, 1977.

Sutherland GR, Gedeon A, Kornman L, et al.: Prena-tal diagnosis of fragile X syndrome by direct de-tection of the unstable DNA sequence. N Engl JMed 325:1720–1722, 1991.

Sutherland GR, Richards RI: Dynamic mutations.Am Sci 82:157–163, 1994.

Tarleton JC, Saul RA: Molecular genetic advancesin fragile X syndrome. J Pediatr 122:169–185,1993.

Verkerk AJMH,Pieretti M,Sutcliffe JS, et al.: Identifi-cation of a gene (FMR-1) containing a CGG re-peat coincident with a break-point clusterregion exhibiting length variation in fragile Xsyndrome. Cell 65:905–914, 1991.

Yu S, Pritchard M, Kremer EJ, et al.: Fragile X geno-type characterized by an unstable region ofDNA. Science 252:1179–1181, 1991.

CASE HISTORY

The patient was an 11-year-old girl diagnosedat birth by newborn screening as having hemo-globin F and hemoglobin S.She was subsequentlydetermined to be hemoglobin SS. She had along history of complications of her sickle celldisease, beginning with splenic sequestrationat 6 months of age.At that time, she was treatedwith transfusions and eventually required apartial splenectomy. She was admitted to thehospital at 19 months of age with acute chestsyndrome as a complication of respiratory syn-cytial viral infection.

Over the next several years, she had re-current episodes of reactive airway disease.Atthe age of 4 years, she had a life-threateningepisode of acute chest syndrome requiring ad-mission to the intensive care unit and exchangetransfusion. She was subsequently transfusedwith red blood cells monthly for 6 months toprevent recurrence. Two years later, she wasagain admitted to the intensive care unit withacute chest syndrome. During this admission,she was found to have Streptococcus pneumo-niae sepsis and pneumonia. She again receivedRBC transfusions monthly for 6 months. Follow-ing this course of transfusion therapy, she wasoffered therapy with hydroxyurea, but this ther-apy was never instituted.

She had been well until the evening prior toadmission, when she developed a fever withoutother symptoms. On admission to the emer-gency room, her parents gave a history of acough, decreased physical activity, and fever athome of 38.8°C (normal is 37°C).She was admit-ted to the emergency room and seen immedi-

ately. Her only prescribed medications were250 mg penicillin twice a day,1 mg folic acid perday, and albuterol using a metered dose inhalergiven as two inhalations on an as-needed basis.

Her admitting examination was performed at1230 hours. Her examination revealed an alertbut toxic-appearing young girl with slightly la-bored and rapid breathing, decreased activity,and intermittent sleepiness. She was asking forfood and drink.Her initial temperature was 38°Corally, her heart rate was 158 beats per minute,her respiratory rate was 24 to 28 breaths perminute (normal is 12–20 breaths per minute),and her blood pressure was 90/40 mmHg (nor-mal for sickle cell disease is 104 to 110/60 to74 mmHg). Her oxygen saturation on room airwas 99% (normal is 98% to 100%).

On physical examination, she had ictericsclera, dry mucus membranes, shotty anteriorcervical adenopathy, and an erythematous pos-terior pharynx. Her chest was clear to ausculta-tion; her heart had a regular rate and rhythmwith a grade II/VI systolic ejection murmur (aflow murmur is common in children, particu-larly in those who have anemia due to the re-quired increase in blood flow to maintainnormal tissue oxygenation). Her abdomen wasdiffusely tender with the liver edge palpable2 cm below the right costal margin. Her skinhad a fine scarlatiniform rash (an exanthemconsisting of a generalized erythematous erup-tion of small bright red macules, commonlycaused by infection with streptococcal organ-isms); she was pale and appeared slightly jaun-diced. With the exception of her lethargy, herneurological examination was normal.

Her initial laboratory evaluation showed a

2

Sickle Cell Anemia

KEITH QUIROLO

17

hematocrit of 17.5% (normal is 30%–43%) withhemoglobin of 62 g/L (normal for this patientwas between 60 and 70 g/L; normal for childrenof her age is 115 to 135 g/L), with a reticulocytecount of 12.9% (normal for age is 1.18%–3.78%).Her total leukocyte count was 9.1 × 106/L (nor-mal is 4.5–13.5 × 106/L). The differential in-cluded 46% neutrophils and 25% band forms(normal is 35%–54% neutrophils, 3%–5% bands).Her platelet count was 1860 × 109/L (normal is1300 to 4000 109/L). She had normal elec-trolytes, a serum glucose of 2.8 mmol/L (normalis 4.1–5.9 mmol/L), bicarbonate of 15 mmol/L(normal is 22–29 mmol/L), serum urea nitrogenof 8.2 mmol/L (normal is 2.1–7.1 mmol/L),and a creatinine of 115 mmol/L (normalis 62–115 mmol/L). Her unfractionated (i.e.,total) bilirubin was 32.5 mmol/L (normal is3–22 mmol/L). A urinalysis revealed a specificgravity of 1.005 (normal is 1.002–1.030), tracebilirubin, and a normal microscopic examina-tion.Additional laboratory tests included a typeand crossmatch for 3 units of leukopoor (bloodfiltered after collection to remove white bloodcells), phenotypically matched (blood matchedfor red cell antigens in addition to the usual ABOblood groups) packed red blood cells.

A chest radiograph was obtained; it revealedatelectasis ( loss of lung volume due to collapseof the lung) of the right lung base and mild car-diomegaly. Due to her tender abdomen, she alsohad a radiograph of her abdomen, which re-vealed a mass in the right upper quadrant com-patible with hepatomegaly.

She was treated with a 20-mL/kg bolus oflactated Ringer’s solution for her hypotensionand was given ceftriaxone (75 mg/kg) for pre-sumed sepsis. She appeared so ill that she wasalso given vancomycin (10 mg/kg) in the eventshe had a bacterial infection resistant to thecephalosporin. Shortly after receiving antibi-otics, she became more hypotensive and re-quired further intravenous fluids; her oxygensaturation decreased to 92% on room air. Anarterial blood gas was obtained that showed apH of 7.29 (normal is 7.35–7.45), an arterialoxygen pressure (pO2) of 6 kPa (normal is11.1–14.4 kPa), a pCO2 of 4.30 kPa (normal is4.26–5.99 kPa), bicarbonate of 15 mmol/L (nor-mal is 22–29 mmol/L), and an oxygen saturationof 76%. She was immediately placed on oxygenby nasal cannula and transferred to the inten-sive care unit. The time was now 1500 hours,which was 2.5 hours after her initial examina-tion in the emergency department.

On arrival to the intensive care unit, she waslethargic but sitting up and able to cooperatewith the staff. Her oxygen saturation suddenlydecreased to 67%, and she became pale and un-responsive. A femoral access line was placed;she was intubated and placed on a ventilatorwith 100% oxygen.The packed red blood cellsordered in the emergency room were available,and she was given a blood transfusion as a rapidbolus to increase her perfusion and blood pres-sure; her hemoglobin was 17 g/L on the bed-side blood-monitoring device. It was noted thatshe had an enlarging right upper quadrantmass. With the blood transfusions and 100%oxygen via endotracheal tube her arterial bloodgas had a pH of 7.29, an O2 of 6 kPa, a pCO2 of4.30 kPa, and bicarbonate of 15 mmol/L. Afterintubation, her blood gas improved, but shecontinued to become more acidotic: pH 7.22, apO2 of 70.4 kPa, a pCO2 of 2.80 kPa, and a bi-carbonate of 9 mmol/L. A blood count aftertransfusion revealed a hemoglobin of 81 g/L, aplatelet count of 17,000, and a leukocytecount of 12.7%. Her glucose was 2.8 mmol/L,and her calcium was 2.00 mmol/L (normal is2.15–2.50 mmol/L).

She began bleeding from all venipuncturesites, and bruising was noted on her trunk andlower extremities. Her blood pressure couldnot be maintained with intravenous fluids con-sisting of fresh frozen plasma, cryoprecipitate,platelet concentrates, and normal saline. Shewas begun on vasopressors along with volumesupport. She was noted to have no urine outputafter a Foley urinary catheter was placed. At1800 hours, 6 hours after presenting in theemergency department, she expired from sep-tic shock.

It was noted that she had gram-negative rodson her blood smear; her blood culture grewpenicillin-sensitive Streptococcus pneumoniae,in 6 hours. At autopsy, she was noted to haveStreptococcus pneumoniae endocarditis of theright ventricle, focal ischemia of the left ventri-cle, bilateral pleural effusions, hepatic conges-tion with thrombosis, renal congestion, bilateraladrenal hemorrhage, and necrosis. Death wasdue to septic shock from Streptococcus pneu-moniae.

DIAGNOSIS

The diagnosis of sickle cell disease is based onthe identification of the hemoglobin type found


in the patient’s red cells, usually determinedat the time of newborn screening and subse-quently confirmed by follow-up testing. Opti-mally, parental hemoglobin samples are used toestablish the diagnosis with certainty. Whenonly one parent is available, DNA methods canbe used to establish the hemoglobin mutationspresent in the newborn.

Newborn screening for sickle cell diseasewas implemented on a large-scale basis in theUnited States in the early 1990s. In some states,ethnic groups were targeted for hemoglobinscreening, historically missing about 20% ofnewborns with sickle cell disease. Prior to thattime, an individual with sickle cell disease wasdiagnosed due to a symptom of the diseaseprompting the individual’s physician to investi-gate sickle cell disease as a possible diagnosis.This is still the case in areas where newborntesting is not performed.

Currently, almost all states in the UnitedStates and some members of the EuropeanUnion require newborn screening for hemoglo-binopathies, including sickle cell disease. Ablood spot is obtained by the Guthrie methodof blood collection on filter paper from a new-born. The dried blood can be used for high-per-formance liquid chromatography (HPLC) or forisoelectric focusing as the initial screening test.Both of these methods use the molecularcharge on the hemoglobin molecule to differen-tiate between normal and variant hemoglobins.States performing hemoglobinopathy screeningrequire a second test to confirm and refine theinitial diagnosis. This confirmatory testing in-cludes hemoglobin electrophoresis, β-globinchain analysis by DNA testing such as reversedot-blot, amplification refractory mutation sys-tem, DNA sequencing, or α-gene mapping as re-quired to make an accurate diagnosis.

The separation of hemoglobin species de-pends not only on the net charge of the mole-cule but also on the ability of the hemoglobinspecies to migrate through the media with theapplication of an electric field.Therefore, use ofvaried pH and media will separate hemoglobinsby charge and migration. Many hemoglobinspecies have similar net charge and size and mi-grate together. Generally, more than onemethod is needed for definitive identification ofhemoglobin type.The two most useful solid me-dia for hemoglobin electrophoresis are cellu-lose acetate at a pH of 8.2 to 8.6 and citrateagar at a pH of 6.0 to 6.2 (Fig. 2-1).

Isoelectric focusing uses a pH gradient to

separate hemoglobins at their isoelectric points.The isoelectric point is the point at which thehemoglobin molecule has no net charge.Thereare polyacrylamide or cellulose acetate gels em-bedded with amphoteric molecules with var-ied isoelectric points that create a pH gradientacross the gel. When hemoglobins are mi-grated across the gel in an electric field, theyare held at their isoelectric points. The hemo-globins are then stained and can be read withrespect to standard hemoglobins as well asquantitated using a densitometer. Althoughmost hemoglobins are sharply separated usingthis method, there are some hemoglobins thatcomigrate together even in this system.

HPLC consists of a negatively charged station-ary absorbent column over which the hemoglo-bin solution is passed. Within the column, thedifferent hemoglobin species are separated bycharge.The hemoglobin is then eluted by an in-creasingly positive buffer competing with thehemoglobin for sites on the absorbent.The netcharge on the hemoglobin molecule determinesthe elution time. The system is automated andcomputerized, and the hemoglobin can be iden-tified and quantitated. This is the most com-monly used method for mass screening ofnewborns in the United States.Variant hemoglo-bin species detected by HPLC are confirmed byother methods.

BIOCHEMICAL PERSPECTIVES

Sickle cell disease was first diagnosed in West-ern medicine in 1904 when a medical intern,Ernest Irons, observed “pear shaped elongatedforms” on the blood smear while investigating apatient who had pneumonia. Dr. Irons was serv-ing at the Presbyterian Hospital in Chicago. Thepatient was Walter Clement Noel, a dental stu-dent from Barbados, who had been ill for amonth with respiratory symptoms. The attend-ing physician, James Herrick, had an interest inblood and cardiovascular diseases. Dr. Herrickfollowed this patient for the next 2.5 years. In1910, he published a brief article describing theblood findings for Mr.Noel in the Archives of In-ternal Medicine (Herrick, 1910); it was the firstcharacterization of sickle cell disease in a jour-nal publication.After finishing dental school, Dr.Noel returned to Grenada and died 9 years laterfrom acute chest syndrome. Serjeant (2001) haswritten a review of the history and medical ad-vances in the treatment of sickle cell disease.

Sickle Cell Anemia 19

In 1922,V. R. Mason (Mason, 1992) publisheda case review in the Journal of the AmericanMedical Association entitled “Sickle Cell Ane-mia,” and the homozygous condition has sincethen been referred to by the description of theshape of the red cells seen by Dr. Irons a decadeearlier. In his review article, Dr. Mason promul-gated the misconception that this disease wasexclusively seen in persons of African origin. In1923, Sydenstricked and colleagues reviewedthe cases of two children with sickle cell dis-ease and observed the blood smears of Cau-casian and African Americans and concluded,with Mason, that sickle cell anemia was a condi-tion peculiar to people of African descent. Neel(1949) reviewed blood smears of families withsickle cell disease over a 2-year period and cor-rectly concluded that sickle cell anemia was adisease with Mendelian inheritance.

Also in 1949, Pauling, Itano, Singer, and Wellspublished a now-famous article in the journalScience: “Sickle Cell Anemia: A Molecular Dis-ease.” Itano, working in Pauling’s laboratory,

used electrophoresis to separate hemoglobinsfrom individuals who had clinical evidence ofsickle cell disease, related individuals who hadabnormal hemoglobin electrophoresis withoutthe disease, and individuals who had normal he-moglobin. He showed that there was a slightcharge difference between these three hemo-globins. Pauling, who had a research interest inantigen–antibody reactions, deduced that theremust be a conformational change in the hemo-globin molecule that caused the molecules toalign and change the shape of the red cell. In1956 Ingram was able to separate hemoglobin Afrom hemoglobin S and showed that hemoglo-bin S had a positive charge relative to hemoglo-bin A. In 1957, Ingram published a reportshowing that there was more valine and lessglutamic acid in hemoglobin S. Twenty yearslater, Morotta showed that this was consistentwith a one-nucleotide change of an adenine fora thymine, a residue in the β-globin gene: GAGto GTG, leading to the amino acid substitutionpredicted by Ingram. The mutation was shown


Figure 2-1. Diagnostic testing. The three upper figures are cellulose acetate, citrateagar, and isoelectric focusing; high-performance liquid chromatography is shown on thebottom. High-performance liquid chromatography is used in most states for newbornscreening. The result is confirmed by a combination of electrophoretic methods andDNA studies.

to be at the position of the sixth amino acid onthe β-globin chain.

In the late 1940s and 1950s, researchers inAfrica searched for kindred having sickle celldisease but were unable to find evidence of a fa-milial pattern of inheritance. In 1949 and againin 1956, Lehmann and Raper described a com-munity in Uganda in which they predicted 10%of the population would have sickle cell anemia.However, they could find no children or adultswith the disease.Other scientists studying popu-lations in different areas of Africa confirmed thisobservation. Later, it became evident that theseearly studies in Africa sampled populations afteryoung children, like the child in the case study,had died. The infant mortality rate and life ex-pectancy of indigenous Africans was so shortand infant death was so common that it was notimmediately apparent that those affected bysickle cell disease had not survived childhood.

Although the homozygous condition forsickle cell disease leads to early mortality, theheterozygous condition confers a positive sur-vival advantage.This is termed a balanced poly-morphism. Many hemoglobinopathies, includingsickle cell disease, occur in areas where Plas-modium falciparum malaria is common. Het-erozygosity for sickle hemoglobin S with normalhemoglobin A confers a selective advantage tochildren living in these areas. These childrenhave lower rates of infection and less para-sitemia than children with normal hemoglobin.There have been many theories concerning thisphenomenon, including one that contends thatthe dehydration and sickling of hemoglobin AS-containing red cells kills infecting trophozoites,changes adhesion molecules for P. falciparum,and changes cytokine production, leading to amoderation of malaria in AS individuals. How-ever, a definitive pathophysiological mechanismfor this effect has not been determined. Theprotective effect of the sickle gene is most com-pelling between the ages of 2 and 16 months,when parasitemia and anemia are most preva-lent and when children have lost the protectionof maternal antibodies but have not developednatural immunity.

The mutation for sickle hemoglobin oc-curred at least three times in Africa and once onthe Indian subcontinent.The disease has spreadfrom Africa to the Mediterranean, areas ofTurkey, North and South America, the Caribbe-an, and the United Kingdom. The sickle hemo-globin combined with common hemoglobinvariants in those areas: β-thalassemia in the

Mediterranean, α-thalassemia and hemoglobinC in Africa, and hereditary persistence of fe-tal hemoglobin in North Africa and India.These combinations have created diverse pre-sentations of this disease, the most severe be-ing hemoglobin SS and hemoglobin Sβ zerothalassemia.

An understanding of sickle cell disease re-quires an understanding of the hemoglobingene clusters occurring on chromosomes 11and 16 and their expression, as well as thestructure and function of the hemoglobin mole-cule. Max Perutz is most responsible for eluci-dating the structure and function of thehemoglobin molecule. Dr. Perutz together withSir John Kndrew received the Noble Prize in1962 for their research on hemoglobin.

Current understanding of the hemoglobingene clusters required the work of so many sci-entists that there is not a readily identifiable sci-entist whose work brought understanding totheir structure and function. In the 1970s, therewas a burst of activity related to determiningthe structure of hemoglobin genes using South-ern blotting and gene cloning. Pioneers in thehemoglobin field had characterized the struc-ture of the hemoglobin protein and determinedthe function and organization of the hemoglo-bin genes without knowing the exact locationor structure of the genes. This knowledgegreatly facilitated the efforts of the later molec-ular biologists in their research on the β-globingene structure and function.A historical reviewof this period of hemoglobin discovery waspublished by Weatherall (2001).

Two similar gene clusters code for the twoglobin proteins (Fig. 2-2). The β-globin genecluster is found on the short arm of chromo-some 11 (11p15.4), and the α-globin gene clus-ter is found on the short arm of chromosome 16(16p13.3). The globin gene clusters are highlyconserved, probably arising from duplicationand unequal crossing over within these regions.The genes are similar all vertebrates, includinghumans.

In both the α- and β-genes there are three ex-ons, or coding regions, and two introns or inter-vening sequences. Within the β-globin genecluster, there are five functional genes and onepseudogene.Within the α-gene cluster, there arethree functional genes and two pseudogenes.The α-gene mutations most commonly involvedeletions, duplications, and triplications. In con-trast, in the β-gene,point deletions predominate.In both gene clusters, the genes are arranged in


the order of expression during fetal develop-ment from the 5' end to the 3' end. During fe-tal development, there are two β-like globinswitches, epsilon to gamma and gamma to beta;and one α-like globin switch, zeta to alpha, inearly fetal development. During these switches,the hemoglobin produced changes from theembryonic Gower 1 and 2 (ζ2ε2 and ζ2α2) andPortland (ζ2γ2) hemoglobins to fetal (α2γ2) he-moglobin. After birth, hemoglobin switches toadult hemoglobin:A (α2β2) and A2 (α2δ2) or, inthe case of sickle cell disease, S (α2βs

2).Promoter regions control each globin gene.

The globin gene promoter regions share con-served sequences that bind transcription fac-tors. The entire β-globin cluster is controlledupstream by a locus control region consistingof five hypersensitive sites. The α-globin clusteris controlled by a similar hypersensitive region.There are numerous erythroid-specific and non-

specific trans-acting factors not coded on ei-ther chromosome 11 or chromosome 16 thatregulate hemoglobin gene expression. An exam-ple of an erythroid-specific trans-acting tran-scription factor is the erythroid Krupple-likefactor. This protein is a member of the zinc fin-ger proteins and binds to CACC motifs in thepromoter region of the β-globin gene. It is a spe-cific activator of the β-globin gene and may beinvolved in the globin switch from γ- to β-globinexpression during the transition from fetal toadult hemoglobin.The most important general-ized regulator of gene function for hemoglobinF synthesis on trans-acting chromosomes is atXp22.2 on the X chromosome.

A dominant theme in the treatment of sicklecell disease is the manipulation of the hemoglo-bin switch from fetal to adult hemoglobin. If theswitch did not occur at all or if the γ-globingene did not completely switch to β-globin


6

10

20

30

40

50

β-LCR5

Chromosome 11

HS-40

Chromosome 16

12

Perc

enta

ge o

f tot

algl

obin

syn

thes

isS

ite o

fer

ythr

o-po

iesi

s

Cel

lty

pe

18

Yolk sac Spleen

ε

ε

ζ

α

26 30 36 1Birth

6 12 18 24

Postnatal age (weeks)Postconceptual age (weeks)

30 36 42 48

Gγ Aγ

γ

γ

ψβ β

β

β

δ

5

ζ2 ψζ1 ψα2 ψα1 α2 α1 θ

4 3 2 1

Megaloblast Macrocyte Normocyte

Bone marrowLiver

Figure 2-2. α- and β-globin gene clusters. Gene arrangement on the two clusters andtheir ontogenic order of expression. The regulatory locus control region (LCR) and thehypersensitive region (HSR) are 5' to the clusters. The middle graph shows relativelevels of hemoglobin and hemoglobin switching. The second bar shows the sites ofhematopoiesis through development. Not shown is hemoglobin A2, which appears nearthe 12th week after birth and plateaus soon after birth, reaching about 2.5% of the totalhemoglobin. Reproduced with permission from Weatherall (2001).

after birth, then the symptoms of sickle cell dis-ease would be greatly attenuated. An under-standing of hemoglobin maturation has greatlyadvanced the search for treatment modalitiesand the basic science of hemoglobinopathies,including sickle cell disease.

Each individual globin chain envelopes andstabilizes the oxygen-binding heme moiety.Theglobin chains interact with each other underthe influence of heterotophic ligands: hydro-gen ions, carbon dioxide,chloride ions, and 2,3-bisphosphoglycerate. All of these ligands, alldifferent (heterotrophic), have a stabilizing ef-fect on the deoxyhemoglobin species, therebydecreasing oxygen affinity and shifting the oxy-gen equilibrium curve to the right. Oxygen andcarbon monoxide (CO) are considered ho-motrophic ligands; the binding of oxygen orCO to one of the hemes in the hemoglobinmolecule increases the affinity of the otherthree for oxygen, shifting the oxygen equilib-rium to the left. Hemoglobin influences the sol-ubility of carbon dioxide in plasma by therelease of protons (Bohr effect) when hemo-globin is deoxygenated. Deoxyhemoglobin isalso a carbon dioxide transporter.When carbondioxide is bound covalently to α-amino groupson the globin chains, protons are released, andchloride ions are drawn into the cell. 2,3-bisphosphoglycerate is synthesized in the redcell and stabilizes deoxyhemoglobin.

The interaction between hemoglobin andthese heterotrophic ligands changes hemoglo-bin affinity for oxygen and alters the shape ofthe molecule through what are called allostericeffects. Allostery refers to the ability of a ligandto change the structure of an enzyme by bind-ing to a site distant from the active site of theenzyme, or in this case the heme moiety of he-moglobin. Most proteins that exhibit allosteryalso exhibit cooperativity. In the presence ofthese heterotropic ligands, the hemoglobinmolecule is in the deoxygenated or “tensestate” (T structure), whereas in the presence ofoxygen the hemoglobin molecule is in the “re-laxed state” (R structure). The cooperative ef-fect of hemoglobin refers to the progressivelyincreased oxygen affinity when the second orthe third oxygen molecule is bound to the he-moglobin. It is at this point that the hemoglobinmolecule switches from the T to the R structure(Fig. 2-3). The classic allosteric enzyme exhibit-ing cooperativity is described by the familiarsigmoid plot of the oxygen dissociation curve(Fig. 2-4).

The polymerization of sickle hemoglobinonly occurs with the hemoglobin in the T, de-oxygenated, state. Heterotrophic ligands in-crease polymerization of sickle hemoglobin. Inthe absence of oxygen,only 1 in 3 million mole-cules of hemoglobin are in the R state, a condi-tion that greatly increases the probability ofpolymerization within the red cell.

Besides combining with oxygen, hemoglobinalso binds CO and nitric oxide (NO). NO is apotent vasodilator.During hemolysis, there is anincreased level of CO production. The role ofCO and heme oxygenase in sickle cell diseasehas not been studied. However, NO has beenstudied extensively, and its effects in sickle celldisease are just being appreciated. The concen-trations of both NO and its precursor, arginine,are low in patients with sickle cell disease. Re-placing arginine has a beneficial effect in sicklecell disease.

The polymerization of deoxygenated sicklehemoglobin is the pathognomonic event insickle cell disease (Fig. 2-5). The requisite fea-tures are hemoglobin in the deoxygenated Tstate and increased concentration of hemoglo-bin within the red cell. Other factors favoringpolymerization are low pH, increased tempera-ture, and decreased availability of oxygen.

Reflecting on the scenario of the case his-tory, one can infer what was occurring at themolecular level during the child’s illness. Ouryoung patient presented with a history of fever,dehydration, metabolic acidosis, and relative hy-poxia due to anemia and pneumonia. All ofthese factors contribute to an increase in sicklehemoglobin polymerization. Due to acidosis,her hemoglobin remained in the T state with adecreased ability to take up oxygen in the pul-monary capillaries. Regional hypoxia leads toV/Q mismatch, a condition in which areas ofthe lung with decreased oxygenation due to at-electasis or other disease process experience areflexive decrease in blood flow in the pul-monary vasculature. She then developed hemo-globin polymerization in her lungs, a uniquecomplication of sickle cell disease, namely theacute chest syndrome.

In dilute solutions, both hemoglobin A andhemoglobin S have identical oxygen-bindingcurves. At the concentrations of hemoglobinoccurring within the red cell, the solubility ofhemoglobin S is decreased.This decreased solu-bility increases the possibility of polymerizationwhen the red cell enters the microcirculationand releases oxygen, increasing the hemoglobin


in the T state within the cell, therefore favoringpolymerization. During the journey of red cellsthrough the relatively hypoxic microcircula-tion, there is an increase in T-state hemoglobinand an increase in the possibility of nucleationformation of sickle hemoglobin. Polymerizationbegins with homogeneous nucleation of indi-vidual hemoglobin molecules with a uniformdelay time. This nucleation progresses to a het-erogeneous process with new nuclei for poly-merization occurring on the surface of the

existing polymer, leading to a stochastic in-crease in growth of the polymers.At low oxy-gen tension, sickle hemoglobin and sicklehemoglobin polymers are in the T state, bind-ing oxygen with low affinity. At low oxygentension, cooperativity is lost, shifting theoxygen-binding curve to the right. This shift isactually beneficial to the patient since in-creased amounts of oxygen become availableto the tissues.

Under normal conditions, only about 5% of


a

b

DEOXY

Aβ2

α138α241β197α244

α238α141

α144β297

F

BC

D

E

F

α1

α1 β1

β2

α2

α2

β1

F

F

E

Deoxyhemoglobin

H

A

H

G

GBC

OXY15

β2

α138

α141α144

β297

α238

α241α244

β197

α1 α2

β1

CC

B

FA

A

G

G B

F

FE

H

D

H

E

F

E

EH

AG

B

E

F

B H

GA

B

E

E

D

CB

D

F

C

38

41 44

97

97

41

3844

β1

β2 α2

Oxyhemoglobin

α1

97

41

41

44

44

38

38

97

B

GE

F

C

E

B

BE

D

GA

A

H

H B E

C

Figure 2-3. Hemoglobin transition states T (tense): deoxygenation state; R (relaxed):oxygenated state. (a) Viewed perpendicular to the axis showing rotation along this axisand sliding of the alpha and beta chains over each other. (b) Viewed from the top show-ing narrowing of the central cavity with oxygenation. Sickle hemoglobin polymerizationcan only occur in the T state. Reproduced with permission, Mathews,Van Holde,Ahern,1999. © Irving Geis.

sickle red cells actually undergo polymer forma-tion while traversing the microcirculation. Im-peding polymer formation are the variability oftransit time, the variability of hemoglobin con-centration in individual red cells, and the het-erogeneity of the concentration of hemoglobinF in the red cells. Many of the sickled cells re-gain their normal shape after reoxygenation;however, some cells become irreversibly sick-led cells (ISC). These ISC can be those that aresubjected to repeated cycles of polymer forma-tion or can be formed after one cycle of poly-merization. These ISC are dense cells with amean hemoglobin concentration as high as500 g/L. ISC tend to be younger red cells with ashortened life span. In the case history, ouryoung patient had severe dehydration, leadingto red cell dehydration, which resulted in in-creased polymerization of hemoglobin withinthe red cell.

The molecular contacts between hemoglobinS occur in axial and lateral planes.The β6 sicklemutation is involved in the lateral contacts be-tween the β-globin chains. The unoccupiedspace is taken up by water, thereby giving riseto the formation of hydrogen bonds in the freespace between the β-globin molecules.The axialcontacts are much more complex. Seven doublestrands make up each hemoglobin fiber, whichhas a helical arrangement with a periodicity of22 Å (Fig.2-6).Both hemoglobin A and hemoglo-bin C can copolymerize with hemoglobin S dueto the fact that these two hemoglobins havecharged amino acids at the β6 site.

Neither hemoglobin F nor hemoglobin A2copolymerize with sickle hemoglobin, and bothinhibit sickling. Hemoglobin A2 differs from he-moglobin A by 12 amino acids, but only onechange, at the amino acid site δ87 (Glu to Thr)on the β-chain, inhibits copolymerization. He-moglobin F differs from hemoglobin A by 39amino acids, at least 2 of which are involved inthe inhibition of copolymerization, γ80 (Asp toAsn) and γ87 (Glu to Thr). The mechanism ofhemoglobin F inhibition of polymerization isnot completely understood since there may beother of the 39 amino acid changes involved ininhibition.

Polymerization and sickle hemoglobin affectthe red cell membrane, which in turn interactswith the microvascular epithelium and molecu-lar environment within the circulatory systemto account for the pathognomonic changes ofsickle cell disease. Sickle hemoglobin itselfcauses oxidative damage to the cell membraneby creating hemichrome (an oxidation productof methemoglobin) that can be seen micro-scopically within the red cell as Heinz bodieson the inner membrane. During polymeriza-tion, the hemoglobin fibers cause red cellmembrane-cytoskeleton uncoupling, which re-sults in the release of microvesicles of red cellmembrane and free hemoglobin into the cir-culation. The altered membrane damages theproteins responsible for maintaining the asym-metry of the lipid membrane. This results inphosphatidylserine exposure to the circula-tion and activation of the coagulation cascade,


2000

25

50

Oxy

gen

Sat

urat

ion

(%)

100

75

40P50

60

Partial pressure of Oxygen (mmHg)

80 100

The curve has a sigmold shapebecause of positive cooperativity.

Mixed venous point

Arterial point

Standard ConditionsTemp = 37°CpH = 7.40BE = 0

Figure 2-4. Oxygen dissociation curve for adult hemoglobin (HbA) demonstrating thecharacteristics of allosteric positive comparativity.Right shift of the curve caused by het-erotrophic ligands and increased temperature leads to decreased oxygen affinity.

measured by an increase in markers of fibrinoly-sis, and d-dimer formation.

In addition, vascular endothelial cells be-come activated in patients with sickle cell dis-ease, resulting in the exposure of adhesionmolecules, in particular, vascular cell adhesionmolecule 1 (VCAM-1), which is present on thesurface of the vascular endothelium in sicklecell disease. An integrin on the red cell, desig-nated very late activation antigen 4, binds with

VCAM-1, causing upregulation of this adhesionmolecule. Soluble VCAM-1 is found in increasedamounts in plasma during inflammation andsickle vaso-occlusive episodes. VCAM-1, inter-cellular adhesion molecule-1 ( ICAM-1), andE-selectin are all molecules found during in-flammation and upregulated in sickle cell dis-ease, and all these have been implicated in thevascular endothelial damage occurring duringsickle cell vaso-occlusion.

Sickle cell disease can be characterized as astate of abnormally activated vascular endothe-lium that promotes increased adhesion of thered cells and leukocytes as well as a procoagula-tant state. The catastrophic events described inthe case history were the direct result of sepsis,hepatic sequestration of red blood cells, andacute chest syndrome. Most of the chronic ef-fects of sickle cell disease can be explained byinflammation, coagulopathy, and arteriolar ob-struction. Membrane disruption and adhesionto endothelial surfaces disrupts the red cellmembrane, leading to hemolysis and anemia.Vaso-occlusion is a complex event involving en-dothelial activation, leukocyte and red cell ad-hesion, hemoglobin polymerization, vesselocclusion, and tissue damage from necrosis. Inthis case, pain was not a feature of the child’spresentation, illustrating that even without painthe effects of sickle cell disease can be severe.Had this child survived her infection and he-patic sequestration, she would have been at riskfor vessel occlusion within the cerebral arteriesand stroke. As the life span of patients withsickle cell disease increases, they become atrisk for pulmonary hypertension, renal failure,bone necrosis and osteoporosis,brain injury, co-agulopathy, red cell destruction, vessel obstruc-tion, and the effects of inflammation.

THERAPY

Therapy for sickle cell disease has changeddramatically since the mid-1990s. Prior to the1980s therapeutic interventions for sickle celldisease consisted of supportive care duringacute illness, opioids for pain management,and occasional transfusions for severe anemiaor life-threatening complications.At that time,sickle cell disease was considered a pediatricdisease as there were few children who survivedinto adulthood. In 1986, the Penicillin Prophy-laxis Study was conducted, providing evidencethat early intervention with penicillin prevented


T

GAG(glutamic acid)

GTG(valine)

Hemoglobin Ssolution

Hemoglobin Scell

Sickledcells

Vaso-occlusion

Cellheterogeneity

Hemoglobin Spolymer

Oxygenated Deoxygenated

β-Globin gene(sixth codon)

Figure 2-5. Pathologic changes in sickle cell dis-eases. Vaso-occlusion is a combination of exposureof phosphatidylserine on the red cell membrane,activation of vascular endothelial cells, activation ofleukocytes, and a state of increased coagulability.From Steinberg,1999.Copyright © 1999,Massachu-setts Medical Society. All rights reserved.

80% of life-threatening infections by Streptococ-cus pneumoniae. Penicillin was subsequentlyestablished as a therapy for newborns and chil-dren with sickle cell disease. This interventionmade sickle cell disease and, by default, all he-moglobinopathies eligible for inclusion in new-born screening programs. Penicillin therapy andnewborn screening ushered in a new era oftreatment for patients with sickle cell disease.

In 1984,hydroxyurea was shown to be effec-tive in increasing fetal hemoglobin levels inpatients with sickle cell disease and was thefirst accepted therapy to treat the basic patho-

physiology of sickle cell disease, namely, hemo-globin polymerization. Hydroxyurea is a drugthat inactivates ribonucleoside reductase andblocks the synthesis of deoxynucleotides, thusinhibiting DNA synthesis. Hydroxyurea is ab-sorbed from the gastrointestinal tract and hasa half-life of about 2 hours in the circulation.After trials were conducted in adults to deter-mine efficacy and in children to determinesafety it was found that this drug has numerouseffects in addition to increasing fetal hemoglo-bin that were not generally appreciated in theearly studies.


Figure 2-6. Electron micrograph of individual fibers of sickle hemoglobin. Individualstrands of polymerized hemoglobin in red cell are shown.The inset shows an electronmicrograph of an individual fiber with a periodicity of 22 Å. The model reveals the sevendouble strands making up each hemoglobin S fiber with one pair detailing the helicalarrangement of the fibers.Adapted with permission from Josephs (1999).

Hydroxyurea therapy results in a decreasedleukocyte count, a decrease in markers of in-flammation, a decrease in endothelial adhesionmarkers, an increase in NO, and with pro-longed use, an increase in hemoglobin concen-tration. Hydroxyurea has been shown todecrease the incidence of most complicationsof sickle cell disease, with the possible excep-tion of first stroke. Hydroxyurea therapy re-quires close monitoring because of the sideeffects that can occur with this therapy. Hy-droxyurea therapy may increase the possibilityof malignancy, birth defects or other complica-tions; however, as of 2004, these effects werenot reported in patients with sickle cell disease.

Other chemotherapeutic agents have beenused in sickle cell disease and are currently intrials. These agents include magnesium, clotri-mazole, and other novel inhibitors of membranetransport used to induce red cell hydration anddecrease the concentration of hemoglobin;arginine-containing compounds to increase sub-strate for the production of NO; compoundsthat decrease cell adherence; and agents to in-crease fetal hemoglobin.Combination therapy isunder investigation, such as with hydroxyureaand magnesium.

For many years, blood transfusion has been atherapy for children and adults with sickle celldisease. Prior to the 1980s, due to the lack ofavailability of blood products and the standardof care at that time, transfusion was used infre-quently and generally only for catastrophiccomplications of this disease. During the 1980s,the risk of infection through transfusion was sohigh that transfusion continued to be used infre-quently.When reliable testing for infectious dis-eases (e.g., HIV and hepatitis) in blood productsbecame available, the use of red cell transfusionbecame standard of care for complications ofsickle cell disease.

In 1988, transfusion to prevent stroke be-came the standard therapy after the Stroke Pre-vention Trial in Sickle Cell Disease (the STOPtrial). Transfusion therapy decreases morbidityin acute chest syndrome and surgery, and de-creases the recurrence of stroke in sickle celldisease.By reducing the red cells containing he-moglobin S to 30% or less of the total hemoglo-bin, stroke risk is decreased dramatically.Reducing hemoglobin S to 50% can decreasethe morbidity of the disease. An emerging andsevere complication of sickle cell disease, andof hemolytic anemias in general, is pulmonaryhypertension. It is not clear that transfusion

alone can reduce the incidence or severity ofthis complication.

An unavoidable complication of blood transfu-sion is iron accumulation in body tissues, calledtransfusion-induced hemosiderosis. There areno excretory pathways to eliminate excess iron,and accumulation results in organ damage andfailure. The challenge in transfusion therapy isthe reduction of iron overload in transfused pa-tients with the administration of chelator drugssuch as desferrioxamine. Chelating agents mustbe given daily by subcutaneous injection overhours. Exchange blood transfusion by erythrocy-tapheresis can delay or prevent the accumula-tion of iron in some patients and is the standardof care for transfusion in sickle cell disease.Blood donors and patients with sickle cell dis-ease are generally ethnically diverse, and the redcell antigens (other than ABO and Rh) presenton the majority of donor red cells occur at differ-ent frequencies than in most recipients. Patientswith sickle cell disease are typed for these minorred cell antigens and receive antigen-compatibleblood beyond the usual red cell ABO, Rh identifi-cation, decreasing antibody formation againstthe transfused red cells.

Stem cell transplantation using bone mar-row,peripherally collected stem cells, or umbili-cal cord blood has been used to cure sickle celldisease (Walters et al., 2000).There is an ongo-ing effort to increase the availability of stemcells with sibling umbilical cord blood collec-tion and unrelated cord blood banking. Thereare efforts to reduce the toxicity of bone mar-row transplantation with the use of nonmye-loablative therapies and new therapies forgraft-versus-host disease. Studies are ongoing tomodify T-effector cells involved in graft-versus-host to make this complication less likely.Therecan be significant morbidity from this therapy,and it is generally reserved for those patientswho have severe disease due to stroke, severeacute chest syndrome, or other chronic compli-cations. The patients who have the best out-comes are those under the age of 2 years.Usually, these children have not declared them-selves to be patients with severe disease meet-ing the criteria for stem cell transplant.

QUESTIONS

1. Patients who have hemoglobin SS as wellas a mutation that produces hereditarypersistence of fetal hemoglobin have few


symptoms of sickle cell disease. Explainwhy this is the case.

2. An infant has 96% hemoglobin F and 4%hemoglobin S at birth.What are the possi-ble diagnoses? What tests can be done toconfirm the newborn screening test?

3. Couples who each have sickle cell traithave three children;none of their childrenhave sickle cell disease.They tell you theydo not want to have another child be-cause they know there is a one in fourpossibility for their children to have sicklecell disease, and they already have threeunaffected children.What is the probabil-ity their next child will have sickle cell dis-ease? If they have another child and thischild has sickle cell disease, then what isthe probability that one of the other chil-dren will be an HLA match?

4. Describe the concept of allostery as it ap-plies to hemoglobin and to regulatory en-zymes.

5. A mother complains to you at a clinic visitthat her 10-year-old child, homozygous forhemoglobin S, still has enuresis at night.Can you explain to her, and to yourself,why this might be the case? What otherorgan dysfunction might be expected in achild this young?

6. Why does deoxygenated hemoglobin Spolymerize? Besides polymerization, whatare the causes of the clinical manifesta-tions of sickle cell disease?

7. How could the child in the case studyhave died from pneumococcal sepsis ifthis bacterium was sensitive to penicillin?

8. If both hemoglobin A and hemoglobin Shave the same oxygen-carrying capacityand exhibit the same oxygen saturationcurves when at low hemoglobin concen-trations, then why does hemoglobin Shave a decreased oxygen-carrying capac-ity at the concentrations found in the redcell? Trace the path of a red cell in sicklecell disease from the lungs to the capillarybed and back to the lungs, noting the he-moglobin changes that are likely to occurduring the journey.

BIBLIOGRAPHY

Bain BJ: Haemoglobinopathy Diagnosis. Black-well Science, Oxford, UK, 2001.

Claster S, Vichinsky EP: Managing sickle cell dis-ease. Br Med J 327:1151–1155, 2003.

De Franceschi L, Corrocher R: Established and ex-perimental treatments for sickle cell disease.Haematologica 89:348–356, 2004.

Herrick JB: Peculiar elongated and sickle shapedred blood corpuscles in a case of severe anemia.Arch Intern Med 6:517–521, 1910.

Jison ML, Munson PJ, Barb JJ, et al.: Blood mononu-clear cell gene expression profiles characterizethe oxidant, hemolytic and inflammatory stressof sickle cell disease. Blood 104:270–280, 2004.

Josephs R: Research on sickle cell hemoglobin, vir-tual tour of sickle hemoglobin polymerization.Laboratory for Electron Microscopy at the Uni-versity of Chicago, 1999. Available at: http://gingi.uchicago.edu/sc2-tour1.htm, 2004.

Lenfant C, National Institutes of Health, NationalHeart Lung and Blood Institute: The Manage-ment of Sickle Cell Disease. 2002. PublicationNo. 02-2117. Available at: www.nhlbi.nih.gov/health/prof/blood/sickle/sc_mngt.pdf, 2004.

Locatelli F, Stefano PD: New insights intohaematopoietic stem cell transplantation for pa-tients with haemoglobinopathies. Br J Haema-tol 125:3–11, 2004.

Manci EA, Culberson DE,Yang YM, et al., and Inves-tigators of the Cooperative Study of Sickle CellDisease:Causes of death in sickle cell disease: anautopsy study. Br J Haematol 123:359–365,2003.

Mason VR: Sickle cell anemia. JAMA 79:1318–1320;1922.

Neel JV:The inheritance of sickle cell anemia. Sci-ence 110:64–65, 1949.

Old JM: Screening and genetic diagnosis of haemo-globin disorders. Blood Rev 17:43–53, 2003.

Roberts I: The role of hydroxyurea in sickle celldisease. Br J Haematol. 120:177–186, 2003.

Serjeant GR:The emerging understanding of sicklecell disease. Br J Haematol 112:3–18, 2001.

Serjeant GR, Serjeant BE: Sickle Cell Disease. 3rded. Oxford University Press, New York, 2001.

Steinberg MH, Forget BG, Higgs DR, et al.: Disor-ders of Hemoglobin. Cambridge UniversityPress, Cambridge, UK, 2001.

Vichinsky EP, Neumayr LD, Earles AN, et al.:Causes and outcomes of the acute chest syn-drome in sickle cell disease. National AcuteChest Syndrome Study Group. N Engl J Med342:1855–1865, 2000.

Walters MC, Storb R, Patience M, et al.: Impact ofbone marrow transplantation for symptomaticsickle cell disease: an interim report. Multi-center investigation of bone marrow trans-plantation for sickle cell disease. Blood 95:1918–1924, 2000.

Weatherall DJ:Towards molecular medicine: remi-niscences of the haemoglobin field, 1960–2000.Br J Haematol. 115:729–738, 2001.


http://gingi.uchicago.edu/sc2-tour1.htm

http://gingi.uchicago.edu/sc2-tour1.htm

www.nhlbi.nih.gov/health/prof/blood/sickle/sc_mngt.pdf

www.nhlbi.nih.gov/health/prof/blood/sickle/sc_mngt.pdf

CASE REPORTS

Patient 1

The patient is a 1.5-year-old white female admit-ted to the NIH Clinical Center for evaluation ofbone deformities in the lower extremities, gen-eralized osteopenia, and a recent left femur frac-ture. She was the product of an uncomplicated,full-term pregnancy delivered by cesarean sec-tion and was born with normal Apgar scoresand a birth weight of 2954 g (normal for full-term newborns).A prenatal ultrasound done at35 weeks of gestation suggested shortening andbowing of femurs, tibias, and fibulas and rhi-zomelic proportions (shortening of proximallimb segments) of the upper extremities. Herperinatal hospital course was uncomplicated.However, she was noted to have bilateral clavi-cle fractures at birth.

The past medical history was significant for adislocated left elbow at 6 months of age that re-solved spontaneously, left femur fracture at 16months of age that occurred while she was try-ing to pull up to stand,and chronic sinusitis dueto an underdeveloped ethmoidal-sphenoid si-nus (air-filled cavity in the skull behind thebridge of the nose). Her psychosocial develop-ment appeared normal. She was able to crawland scoot but could not cruise yet.

On physical examination, her height was63.6 cm (<3rd percentile for chronological ageand 50th percentile for a child aged 4 months).Her weight was 7.2 kg (<3rd percentile forchronological age and 50th percentile for achild aged 6 months), and head circumferencewas 46.2 cm (45th percentile). Her skull shape

was normal with slightly triangular facies and aflattened midface. Her anterior fontanelle wasopen; measuring 3 × 2 cm. Sclera hue was lightblue with normal reflexes and extraocularmovements. Her ears were in normal positionand shape, and her oral examination revealedseven erupted gray-translucent pointed teeth.Her heart, lungs, and abdominal examinationwere normal. Her spine was straight, and herupper extremities had no major deformities. Ex-amination of her lower extremities revealedmild bowing of both femurs and anterior bow-ing of her tibias. Both upper and lower extremi-ties had normal range of motion with nomuscle contractures. Her neurological examwas entirely normal.

Her initial x-rays (babygram) revealed os-teopenia ( low bone mineral density) through-out the bony structures.There were deformitiesof multiple thoracolumbar vertebral bodies(bony segments of upper and lower spine), andher ribs appeared thin.The proximal left femurand distal right femur showed mild anterolat-eral bowing.Anterior bowing of both tibias andfibulas was also seen. A bone densitometrystudy of the L1–L4 spine done at 3 years of agewas 6.66 standard deviations below the meanfor children of the same age.

Patient 2

The second child was a 9-month-old white fe-male admitted to the NIH Clinical Center forevaluation of multiple long-bone fractures. Shewas born to a 31-year-old mother with a previ-ous spontaneous miscarriage. The patient wasthe product of a normal pregnancy born at 38

3

Osteogenesis Imperfecta

ARMANDO FLOR-CISNEROS and SERGEY LEIKIN

30

weeks of gestation with pronounced cran-iotabes (softening of the skull bones) and alarge posterior fontanelle. The delivery wasvaginal with normal Apgar scores and a birthweight of 2800 g. At 4 weeks of age, the babywas admitted to the local hospital for irritabilityand inconsolable crying. Physical examinationat that time revealed an infant with swollen bi-lateral upper leg areas and a flaccid right arm.Abone survey revealed generalized osteopenia,bilateral femur and right tibia midshaft frac-ture, right humerus fracture, a right claviclefracture with new callous formation, and an oldposterior fifth rib fracture. Her past medicalhistory was significant for poor weight gainand growth, constipation, and an umbilical her-nia.At 5 months, she fractured her left humerus,and 3 months later she fractured both femursagain.

On physical examination, her weight was5.8 kg (<3rd percentile for chronological ageand 50th percentile for a child aged 4 months),her height was 63 cm (<3rd percentile forchronological age and 50th percentile for achild aged 4 months), and her head circumfer-ence was 45 cm (50th percentile). Her headshowed a flattened occiput with frontal bossing,a triangular face with small chin and a wide-open anterior fontanelle measuring 4 × 6 cm(abnormally enlarged for age). Her sclerae wereblue, and her ears were low set and posteriorlyrotated. Her heart and lungs were normal. Herchest had a pectus carinatum shape (protrud-ing chest), and her abdomen showed a smallumbilical hernia. Examination of her upper ex-tremities revealed moderate bowing of rightand left humerus and left ulna. Her lower ex-

tremities showed pronounced bowing of rightand left femurs and distal left tibia.

Her x-ray series revealed generalized under-mineralization throughout the bony structures.Her ribs were thin, and the spine showed thepresence of multiple compression fractures atthe thoracolumbar level. There was markedbowing of all extremities with evidence of oldfractures in femurs, humerus, and left ulna. Abone mineral density study of the lumbar verte-bral bodies (L1–L4) performed at 5 years of agerevealed a value of 6.57 standard deviations be-low the mean for children of the same age.

DIAGNOSIS

The clinical presentation and the radiologicalabnormalities seen in both patients suggested amoderately severe form of osteogenesis imper-fecta (OI) with moderate bone fragility in pa-tient 1 and a severe, progressively deformingform of OI in patient 2. OI—also known as brit-tle bone disease—is a relatively rare (∼1 per10,000 for all types of OI), heterogeneous groupof heritable disorders (Table 3-1) characterizedby bone fragility, usually accompanied by otherconnective tissue abnormalities (Byers andCole, 2002).

Many epidemiological, clinical, genetic, andbiochemical studies have been conducted sincethe first widely recognized descriptions of OI inthe 18th century. The disease has been shownto be caused by defective synthesis of type Icollagen, a triple helical protein made from twoα1(I)- and one α2(I)-chains with fibers thatform the structural frame of bone matrix.

Osteogenesis Imperfecta 31

Table 3-1. Clinical Heterogeneity and Inheritance of Osteogenesis ImperfectaOI Type Clinical Features Inheritance

I Normal stature; little or no deformity; blue sclerae; hearing ADloss in 50% of individuals; dentinogenesis imperfecta rare but may distinguish a subset

II Lethal in the perinatal period; minimal calvarial mineralization; thin AD (new mutations) with beaded ribs; marked long bone deformity; platyspondyly parental mosaicism

III Progressively deforming bones, usually with moderate deformity at AD with parental mosaicism;birth; extreme short stature; sclerae variable in hue, often lighten autosomal recessive (rare)with age; common dentinogenesis imperfecta and hearing loss

IV Mild-to-moderate bone deformity; variable short stature; white or AD with parental mosaicismblue sclerae; common dentinogenesis imperfecta; hearing loss in some patients

V Bone fragility; mild-to-moderate short stature; no dentinogenesis Probably ADimperfecta; radioulnar synostosis; hyperplastic callus formation

AD, autosomal dominant.

Abnormal electrophoretic mobility confirmedthe presence of defects in type I collagen fromboth patients (see next section). SubsequentDNA sequencing revealed a glycine-to-serinesubstitution at helical position 238 in the α2(I)-chain from patient 1 and a similar substitutionat helical position 193 in the α1(I)-chain frompatient 2.

Sillence Classification of Osteogenesis Imperfecta

Although the majority of OI cases are causedby type I collagen mutations, current approachto clinical diagnosis of different OI forms is stillbased on family history and physical and radi-ographic findings rather than biochemi-cal analyses or underlying pathophysiologicalmechanisms.The original Sillence classification(Sillence et al., 1979) distinguishes four differ-ent types of OI.

OI type I is the mildest and most commonform of the disease, with autosomal dominantmode of inheritance. Affected individuals usu-ally have blue sclerae and mild-to-moderatebone fragility prior to puberty. Fractures rarelyoccur in utero. During infancy and childhood,the fractures are related to moderate trauma.The height of individuals with type I OI is usu-ally normal or near normal. About half haveearly hearing loss that usually starts in the lateteens and may progress to severe loss by adult-hood.Type I OI is subdivided based on the ab-sence (IA) or presence (IB) of dentinogenesisimperfecta. Additional clinical findings for OItype I are mitral valve prolapse, easy bruising,and hyperextensibility of large joints.

OI type II is the most severe form of the dis-ease, leading to fetal demise or death in theperinatal period. It is caused mostly by new,dominant mutations. The small rate of recur-rence of this phenotype among siblings ismost consistent with parental mosaicism. In-fants are usually stillborn or born prematurelywith both weight and length small for theirgestational age. If they survive birth, then af-fected infants usually die of respiratory failurewithin the first 2 months of life. Type II OI ischaracterized by triangular face, flat midface,small beaked nose, and bluish-gray sclerae.Thehead is large for body size with an extremelysoft calvarium and a large anterior and poste-rior fontanelle. The thorax is deformed andsmall with a narrow apex. Radiographic exami-nation reveals multiple in utero fractures in

various stages of healing. The long bones areosteopenic, cylindrical, and deformed withvery little cortex.

OI type III or progressively deforming OI(Fig. 3-1B, 3-1D) is the most severe form of thedisease compatible with life beyond infancy.The majority of cases appear to result fromdominant mutations, but rare autosomal reces-sive cases have also been reported.Affected in-dividuals are recognized after birth because oflong bone fractures and cranial abnormalities(as seen in patient 2). Many individuals withtype III OI die in childhood or in adulthooddue to respiratory, cardiac, or neurologicalcomplications.Those who survive exhibit grad-ual deformity of the long bones and spine withpronounced vertebral compression. Patientswith type III OI have extreme growth defi-ciency; adults have the height of prepubertalchildren. Clinically, these individuals usuallyhave macrocephaly (increased head circumfer-ence) and a triangular, flat midface. The sclerais initially bluish and whitens with age. Radi-ographically, there is generalized osteoporosis,wormian bones (irregular plates of bone inter-posed in the sutures between the large cranialbones), and “codfish” vertebra. As childhoodprogresses, a flared long bone metaphysis devel-ops, and the long bones appear thin, twisted,and cylindrical.

OI type IV (Fig. 3-1A, 3-1C) envelops a largespectrum of moderately severe forms of domi-nantly inherited disorders characterized by mild-to-moderate bone fragility. Some children withtype IV OI have prenatal fractures and deformityat birth, while others have mild-to-moderatebowing. During childhood, they may have sev-eral fractures per year. The fracture frequencyusually declines after puberty. In addition to os-teoporosis, these individuals have modeling ab-normalities of the long bones, and about onethird of patients develop progressive scoliosis.Clinically, these patients have variable shortstature, grayish to normal sclerae, and macro-cephaly. Similar to type I, OI type IV is also sub-divided into type IVA and IV B based ondentinogenesis imperfecta (abnormal dentin for-mation).The clinical and radiological findings forpatient 1 are most consistent with OI type IV.

Additional OI types

Since the original classification was proposed,25 years of studies revealed difficulties in sepa-rating the wide and nearly continuum spectrum


of OI phenotypes into four distinct categories.Most commonly, patients not fitting more nar-rowly defined types I–III are diagnosed withtype IV OI, resulting in a rather broad range ofclinical and radiological findings within this cat-egory. However, several new OI types were pro-posed based on differences in bone histologyobserved within the Sillence type IV group(Rauch and Glorieux, 2004).

OI type V is now widely recognized as a dis-tinct OI phenotype with characteristic clinicaland radiological features, such as predispositionto formation of hypertrophic callus at sites offractures or surgical interventions, early calcifi-cation of the interosseous membrane of theforearm, and appearance of dense metaphysealbands in radiographs. Patients have moderate

bone fragility, white sclerae, and no dentinogen-esis imperfecta. Their histological examinationreveals a characteristic, meshlike bone lamella-tion pattern.Type V OI has an autosomal domi-nant inheritance pattern, but no evidence oftype I collagen abnormalities has been found.

Unusual histological findings of increased os-teoid formation and abnormal,“fish scale” bonelamellation pattern were used to suggest a newtype VI OI (Rauch and Glorieux, 2004). So far,these findings were reported in fewer than adozen patients, and the mode of inheritance ofthe disorder was not established.Another pecu-liar disorder with recessive inheritance (unusualfor type IV OI) was observed in a community ofNative Americans in northern Quebec (Rauchand Glorieux, 2004). In these patients, bone


Figure 3-1. (A), (B) Phenotypic characteristics of osteogenesis imperfecta (OI): (A) OItype IV and (B) OI type III. (C ), (D) Radiographs of the long bones of the lower limbs:(C) patient 1 with type IV OI and (D) patient 2 with the deforming type III OI.

fragility was accompanied by prominent short-ening of proximal limb segments (rhizomelia)and a decrease in the femoral neck shaft angle(coxa vara).The mutation was localized to chro-mosome 3p22-24.1, outside the loci for type Icollagen genes. The authors of this study pro-posed to classify this disorder as type VII OI.There is little doubt regarding the merits ofgrouping OI cases with distinct histological andgenetic findings. However, so far these caseshave not been demonstrated to have distinctclinical and radiological features common formany unrelated patients, which could be usedas phenotype criteria following the tradition es-tablished by the original Sillence classification.Therefore, the available information is still notsufficient for determining whether thesegroups truly represent newly recognized dis-tinct Sillence types of OI.

BIOCHEMICAL PERSPECTIVE

Since OI is a dominant genetic disorder, a curefor it would require repair, removal, or inacti-vation of the affected genes—a daunting tasklikely to require many years of research(Prockop, 2004). However, the same mutationsoften cause drastically different OI phenotypes,for instance, several patients with the sameGly238Ser substitution in α2(I)-chain as patient1 but different symptoms were described, in-cluding severe type III OI and moderate type IVOI. One case was even originally classified asmild type I OI. Large variation in the OI severitywas also found in unrelated patients with iden-tical mutations, within affected families, and inthe Brtl mouse model of OI (Forlino et al.,1999). These observations suggest that betterknowledge of OI biochemistry underlying thephenotype variability might lead to simplerpharmacological treatments, reducing severityof the disease. Some progress in understandingmolecular origins of OI has been made, butmuch remains to be discovered before suchtreatments become a reality.

Osteogenesis Imperfecta Mutations

The overwhelming majority of OI cases arecaused by mutations in COL1A1 and COL1A2encoding the α1(I)- and α2(I)-chains of typeI collagen, respectively. Several hundred differ-ent OI mutations were described (see, e.g.,an online mutation database at www.le.ac.uk/

genetics/collagen/; Dalgleish, 1998). Based ontheir effect on collagen synthesis, these muta-tions can be separated into three groups:COL1A1 and COL1A2 “null allele”and structuralmutations.

COL1A1 null allele mutations do not pro-duce mutant collagen. Instead, the affected al-lele fails to produce functional α1(I)-chains, andcells make about half the normal amount of col-lagen. Insufficient synthesis of normal collagenis believed to be the primary cause of type I OI(Byers and Cole, 2002).

COL1A2 null allele mutations produce colla-gen that has three α1(I)-chains rather than twoα1(I)- and one α2(I)-chains (50% of all type Icollagen in heterozygous and 100% in homozy-gous cases).The α1(I)-homotrimer does have a“normal”role,but only as a minor collagen in fe-tal tissues. Mutations that result in exclusivesynthesis of α1(I)-homotrimers cause rare re-cessive forms of OI and Ehlers-Danlos syn-drome (Schwarze et al., 2004).

Structural mutations in COL1A1 or COL1A2produce abnormal type I collagen with substitu-tions,deletions,or insertions of amino acids.Mu-tations of this group usually cause type II, III, orIV OI (Byers and Cole, 2002). However, patientswith phenotype closer to type I OI were also re-ported (e.g., one with the same Gly238Ser sub-stitution as patient 1). By far the most commontype of structural mutations (>80%) is substitu-tions of “obligatory”glycines as in patients 1 and2; these mutations are the focus of this chapter.

Collagen Structure and Stability

The obligatory glycines are located in every thirdposition of the collagen triple helix.They are re-quired for maintaining the structural integrity ofmature type I collagen molecule, which consistsof a long triple helix (1014 amino acid) flankedby short (11–26 amino acids), nonhelical pep-tides. Preponderance of glycine substitutionsindicates that disruptions of the triple helixstructure are the primary cause of moderate, se-vere, and lethal OI. However, the relationship be-tween structural defects and OI symptoms is notstraightforward, as illustrated by studies of colla-gen thermal stability.

In particular, the triple helix irreversibly de-natures within minutes at 41°C–42°C unless itis incorporated into fibers, where it becomessubstantially more stable (Leikina et al., 2002).Structural defects caused by OI mutations re-duce the denaturation temperature Tm, some by


www.le.ac.uk/genetics/collagen/

www.le.ac.uk/genetics/collagen/

as much as 5°C–6oC (Fig. 3-2).At least in model,collagenlike peptides, larger changes in Tm cor-relate with larger structural disruptions of thetriple helix (Beck et al., 2000).The change in Tm

might also directly affect fibrillogenesis and ma-trix formation (e.g., if Tm drops below normalbody temperature and molecules denature be-fore they have a chance to form fibers). Thus,one could expect Tm changes to correlate withthe severity of OI,but no such correlations werefound (Bachinger et al.,1993). It was argued thatthe Tm values traditionally measured by triple-helix susceptibility to proteolytic enzymesmight be unreliable.However,our studies of sev-eral dozen mutations by differential scanningcalorimetry (DSC) did not produce any correla-tions as well (DSC is widely accepted as themost reliable and accurate technique for Tm

measurement). For instance, despite the pheno-type differences, DSC tracings showed the same2°C change in Tm for both of our patients (Fig.3-2).A smaller Tm change (≤1°C) is observed in a

patient with an α1(I)-Gly832Ser substitution,and a much larger change (∼5°C) is seen in a pa-tient with an α1(I)-Gly25Val substitution, al-though both patients have type IV OI.

Chain Synthesis and Association

To understand how structural defects in collagencould cause OI, we need to follow the collagenfiber synthesis pathway (Fig. 3-3). It starts fromsynthesis of precursor (procollagen) chains.Asso-ciation of two pro-α1(I)- and one pro-α2(I)-chainsat their C-propeptides initiates procollagen fold-ing, which proceeds in a zipperlike manner fromC- to N-terminal end of the molecule (Engel andProckop,1991).These processes occur inside theendoplasmic reticulum (ER) and are assisted by avariety of ER chaperones and enzymes (Lamandeand Bateman,1999).

Substitutions of the obligatory glycines do notaffect collagen chain synthesis and association.Thus, heterozygous substitutions in α2(I) should


Figure 3-2. Analysis of collagen thermal stability by differential scanning calorimetry(DSC). DSC measures the heat of denaturation on temperature variation at a constantscanning rate (here 0.125°C/min). DSC thermogram of normal (control) collagen has asingle denaturation peak with the apparent denaturation temperature Tm ≈ 42°C. Addi-tional peaks with lower Tm in DSC tracings from patients 1 and 2 are associated with mu-tant molecules. Patient 2 collagen molecules with one and two mutant α1(I)-chains havea similar Tm, so only one broadened mutant peak is observed.The fraction of mutant mol-ecules can be estimated from the area under the corresponding peak (e.g., 45% ± 5% mu-tant in patient 1 and 70% ± 5% mutant in patient 2). Some mutations reduce the apparentTm to normal core body temperature or even below it (e.g., Tm ≈ 37.5°C for α1(I)-Gly25Val substitution). Other mutations have minimal effect on Tm (e.g., only a singlebroadened peak of mutant and normal collagen is observed for α1(I)-Gly832Ser).

Figure 3-3. Collagen matrix synthesis pathway.Type I collagen precursor (procollagen) is a heterotrimer of two pro-α1(I)- and one pro-α2(I)-chains.It is synthesized and folded within endoplasmic reticulum (ER) (steps a through c).Procollagen folding is triggered by association of fully synthesizedchains at C-propeptides (b). It proceeds in a zipperlike manner from C- to N-terminal end (c).Chain synthesis,association,and folding processes are as-sisted by chaperones and are accompanied by hydroxylation of proline residues within Xaa-Pro-Gly triplets, hydroxylation of some lysine residues,and glycosylation of hydroxylysine residues. Folded procollagen is transported from ER through Golgi and secreted from cells (d). Mature collagen isa 300-nm long triple helix flanked by short (11–26 amino acids) nonhelical peptides. It is formed by cleavage of C- and N-propeptides from procolla-gen by specialized proteases (e).Propeptide cleavage triggers self-assembly of collagen triple helices into fibrils ( f ).The fibrils bind proteoglycans andother molecules (g) and assemble into a complex fiber network,which serves as a matrix for bone and other connective tissues.

generate 50% molecules with one mutant chainand 50% normal molecules (assuming equalexpression of both alleles). Mutations in α1(I)should also generate 50% molecules with onemutant chain, but only 25% normal moleculesand 25% molecules with two mutant chains.

Interestingly, glycine substitutions in α1(I)generally lead to more severe OI than similarsubstitutions at the same site in α2(I). This ef-fect might be caused by a larger amount of de-fective collagen produced by α1(I)-mutationsand could explain the more severe phenotypeof patient 2 compared to patient 1. However, itcannot explain the phenotype variation in caseswith the same α2(I)-Gly238Ser substitution.

Procollagen Folding

Procollagen folding is preceded by hydroxyla-tion of prolines in Y positions of sequentialGly-Xaa-Yaa triplets and is accompanied by hy-droxylation and glycosylation of some lysineswithin unfolded regions of procollagen chains.Folding of collagen molecules with glycine sub-stitutions proceeds normally from the C-termi-nus up to the mutation site, where it stops(Engel and Prockop, 1991). Folding resumesonly after a conformation of the chains accom-modating the mutation is found, and helix for-mation is renucleated.The resulting prolongedexposure of unfolded chains on the N-terminalside of the mutation to ER enzymes leads to ex-cessive hydroxylation and glycosylation of lysineresidues. This posttranslational overmodifica-tion is the hallmark of structural mutations. It isusually detected by slower migration of collagenchains and their CNBr fragments in gel elec-trophoresis (Fig.3-4).

The overmodification of collagen from pa-tients 1 and 2 is weak because it occurs only onthe N-terminal side of the mutation. Mutationslocated closer to the C-terminal end lead to sub-stantially larger overhydroxylation and overgly-cosylation. In principle, this overmodificationgradient could explain milder OI symptoms ob-served for N-terminal glycine substitutions(e.g., almost no lethal substitutions in the N-ter-minal quarter of collagen were reported, andmany of these cases were classified as type IOI). However, no consistent severity gradientand no correlations of symptoms with over-modification were found for mutations locatedfurther toward the C-terminal end. Thus, therole of overmodification in OI phenotype stillremains unclear.


Figure 3-4. Gel electrophoresis of (A) collagenchains and (B) α1(I) CNBr peptides (B) and (C )location of different cyanogen bromide (CNBr)fragments on α1(I)-chain. Posttranslational over-modification can be detected from slower migra-tion of mutant chains in (A) or asymmetric shapeof CB spots in (B).The spots correspond to pep-tides obtained by CNBr digestion of α1(I)-bandsand electrophoresis in a second, perpendicular di-rection. Slower migration of overmodified, mutantα1(I)-chains is clearly visible in the Gly997Ser lanein (A) but barely detectable in patients 1 and 2due to substantially weaker overmodification (typ-ical for mutations in the first 200–300 amino acidsfrom the N-terminal end). Correspondingly, allα1(I)-CNBr peptides from Gly997Ser collagen ap-pear to be overmodified. In collagen from patient2, only CB 8+5 has the characteristic asymmetricshape. Since only the fragments located on theN-terminal side of the mutation are overmodified,analysis of CNBr fragments allows estimation ofapproximate mutation location. For instance,Gly997Ser mutation at the C-terminal end of themolecule causes overmodification of all CNBrpeptides, while Gly193Ser mutation in patient 2causes overmodification of CB 8+5 but not CB 6,CB 7, CB 3, or CB 8.

Secretion

After completion of folding,procollagen is trans-ferred from ER, transported through the Golgisystem, and secreted from cell (Kadler, 1994).Abnormal protein is retained within ER and tar-geted for degradation. Some mutant molecules

are recognized as abnormal and destroyed,whileothers escape the quality control system andare secreted. The outcome varies from signi-ficant retention to complete secretion depend-ing on the mutation. The retention anddegradation of mutant molecules puts an extrastress on collagen-producing cells, potentiallyexplaining their reduced activity and viabilitycommonly observed in OI.

Characterization of intracellular retention/degradation is important but often challenging.Unlike Gly → Cys substitutions, which can beselectively labeled by [35S]-Cys, chains withGly → Ser substitutions are difficult to distin-guish by simple biochemical assays. Sometimesan increased ratio of type III/type I collagen infibroblast culture media is used as an indicatorof abnormal secretion of type I molecules, butthe results should be confirmed by other meth-ods. For patients 1 and 2, the type III/type I ra-tio was within the “normal” range of 10%–15%.To confirm the secretion and to estimate the re-tained/degraded fraction, we used the changein Tm of mutant molecules. From deconvolutionof mutant and normal peaks in DSC thermo-grams, we found that more than 80% of mutantmolecules were secreted (Fig. 3-2).

Extracellular Processing

Secreted procollagen undergoes additional pro-cessing by specialized enzymes that cleave N-and C-propeptides and convert some lysine andhydroxylysine residues into reactive allysyl alde-hydes (Kadler, 1994). Most glycine substitutionsdo not affect these reactions. However, severalcases with slow or incomplete cleavage ofN-propeptides due to altered conformation ofthe cleavage site were reported.The conforma-tional change can be caused by mutations inthe adjacent N-terminal part of the triple helixas well as by long-range structural rearrange-ments (such as a chain register shift) inducedby glycine substitutions located even severalhundred residues away. Delayed or incompleteN-propeptide cleavage leads to incorporationof uncleaved molecules, limiting the thicknessof collagen fibers. The resulting reduction inthe fiber mechanical strength might causemore severe joint laxity and hyperextensibilitythan usually observed in OI (resembling typeVII Ehlers-Danlos syndrome associated withN-propeptide cleavage site mutations).

In vitro testing of fibroblast procollagenfrom patients 1 and 2 showed normal cleavage

by purified bovine N-proteinase, indicating thatthese mutations do not induce long-range struc-tural changes propagating all the way to theN-terminal end of the triple helix.

Intermolecular Interactions and Fibrillogenesis

Propeptide cleavage triggers collagen self-assembly into fibers.The fibers are stabilized bycross-linking of allysyl aldehydes with lysineresidues on adjacent collagen helices.They bindproteoglycans and other matrix molecules andorganize into a complex three-dimensional net-work serving as a scaffold for bone mineraliza-tion (Kadler, 1994).

All information needed for fiber self-assemblyis encoded within the collagen triple helix. Forinstance, a solution of purified helices in a phys-iological buffer forms nativelike fibers whenheated to body temperature. Structural defectsintroduced by mutations might disrupt interac-tions that govern the self-assembly process,potentially explaining altered in vitro fiber for-mation observed for OI collagens. However, be-cause collagens from only one control and onepatient with a given mutation are typically com-pared in such studies, effects of different gene-tic background of the two individuals cannotbe excluded (e.g., variations in the sites and theextent of posttranslational modification). A sys-tematic study of interactions between mutanttype I collagen helices has so far been performedonly for the Brtl mouse model of OI with anα1(I)-Gly349 → Cys substitution (Kuznetsova etal., 2004).A wide variation of in vitro fibrilloge-nesis of tissue collagens from animals with thesame genotype but different genetic back-ground was observed, but no statistically signifi-cant difference between the mutant and wildtype was found.The absence of changes in inter-actions between type I collagen molecules wasconfirmed by direct measurement of intermole-cular forces. Still, it remains unclear whetherthis is just a peculiar feature of this particularmutation.

In vivo, type III, type V, and other collagensare also incorporated into fibers, although insmaller quantities (dependent on the tissue). In-teractions of type I collagen with them appearsto be important for formation and functionalproperties of the fibers. Glycine substitutionsmight affect these interactions directly as wellas by altering the composition of the collagenmix secreted by collagen-producing cells. An


elevated content of type III collagen was ob-served in tissues and cell cultures from OI pa-tients, but no systematic studies of interactionsbetween type III and type I collagens and theirrole in OI phenotype were reported. In the Brtlmice, we found substantial variations in thecontent of type III collagen in different tissuesand fibroblast cultures, but no statistically sig-nificant correlations with the animal pheno-type.

In addition to other collagens, at least severaldozen matrix molecules are known to interactwith type I helices. Some understanding ofthese interactions (e.g., likely locations of bind-ing sites) has been developed (Di Lullo et al.,2002). Disruption of these interactions by OImutations within the binding sites might ex-plain, for instance, the appearance of distinctlethal and nonlethal regions of glycine substitu-tions in the α2(I)-chain (Forlino and Marini,2000). However, no direct confirmation of thishypothesis has so far been reported. For in-stance, collagen-proteoglycan interactions arebelieved to play an important role in bone for-mation and mineralization. Based on the pro-posed binding site map (Di Lullo et al., 2002),the α2(I)-Gly238 → Ser mutation of patient 1 islocated within a region where no sites havebeen discovered so far. The α1(I)-Gly193 is lo-cated within a proposed decorin-binding re-gion. Its substitution might contribute to moresevere symptoms of patient 2, but without cor-roborating evidence, this is just a speculation.

Molecular Mechanisms of Pathology

In summary, substitutions of obligatory glycineresidues alter the structure of collagen triplehelix. It has now been widely accepted thatstructural changes in type I collagen affect os-teoblast function by causing reduced thermalstability, slower folding, posttranslational over-modification, and intracellular retention/degra-dation of procollagen molecules.These changesalso affect interactions of type I collagen mole-cules with each other, with other collagens, andwith other matrix molecules.Abnormal activityand viability of osteoblasts and abnormal extra-cellular interactions of secreted mutant colla-gens lead to abnormal bone matrix formationand bone fragility. This picture of moleculardefects contributing to OI is reasonably clear.However, no consistent correlations of OI phe-notype and severity with any specific defectshave been found so far.

Animal Models

Probably one of the most important factors hin-dering studies of the phenotype-genotype rela-tionship has been the lack of good animalmodels. Presently, only one nonlethal mouse OImodel (the Brtl mouse) with a Gly349 → Cyssubstitution knocked into one col1a1 allele isavailable (Forlino et al., 1999). The initial bio-chemical and biophysical study of collagenfrom this model (Kuznetsova et al., 2004) re-vealed significant variations of posttranslationalovermodification (hence, likely the rate of fold-ing), intracellular retention and degradation,and intermolecular interactions of OI collagenswithin the same genotype. The properties ofcollagen purified from cell cultures and differ-ent tissues originating from the same animalwere significantly different. For instance, femurcollagen had substantially larger extent of post-translational modification, higher Tm and differ-ent in vitro fibrillogenesis kinetics compared toskin or fibroblast collagen. Furthermore, a sig-nificant animal-to-animal variation, exceedingthe difference between mutant and wild-typeanimals, was observed.

Thus, many of the biochemical abnormalitiesare tissue specific and are significantly affectedby the genetic background of the animal, stress-ing the need for systematic biochemical analy-sis of collagen from many individuals (animals)with the same or similar mutations and differ-ent genetic background. Only then can wehope to understand the genotype-phenotyperelationship,determine the molecular origins ofphenotype variability within the same OI geno-type, and rationally design effective pharmaco-logical treatments.

THERAPY

In general, the mainstay conventional clinicalmanagement of OI is to facilitate the achieve-ment of gross motor skills, including ambula-tion and head control, and to maximize skillsfor independent living. In patients with type IIIand IV, this is best accomplished by early inten-sive physical rehabilitation and by orthopedicintervention as needed. Gross motor skills aredelayed in OI due to muscle weakness and jointlaxity. Patients with moderate OI who have po-tential for independent walking should haveprotected ambulation with a combination ofbracing, surgical correction, and physical ther-apy beginning as early as possible.


Patients with OI should be under the care ofan orthopedic surgeon with experience in man-agement of OI. Bone fractures and deformitiescan be managed by a combination of splinting,cast immobilization, and intramedullary rod-ding to provide anatomic positioning of ex-tremities and to prevent loss of function. Theaim of surgery is to correct bowing and inter-rupt the fracture and refracture cycle.The clas-sical surgical approach is to perform multipleosteotomies with realignment of long bone sec-tions and fixation with intramedullary rods (By-ers and Cole, 2002).

The search for beneficial pharmacologicaltreatment of OI with the goal to reduce therate of fractures and promote longitudinalgrowth continues. Several uncontrolled studiesof bisphosphonates as a potential monother-apy for OI have caused great excitement in theOI community (Glorieux et al., 1998). Cyclicpamindronate infusions have been widely usedas an off-label treatment of OI. This drug is apotent nitrogen-containing bisphosphonatethat is internalized by osteoclasts. It inhibitsenzymes of the mevalonate pathway, therebypreventing the biosynthesis of isoprenoidcompounds that are essential for the posttrans-lational modification of small GTP-binding pro-teins.The inhibition of protein prenylation anddisruption of key regulatory proteins explainthe loss of osteoclast activity. Most of the expe-rience with these compounds is derived fromstudies in postmenopausal osteoporosis.Whenused in OI, this drug would not affect the syn-thesis and deposition of abnormal collagen.Thus, patients with OI might have quantita-tively more bone after treatment, but theirbone would not be expected to have betterstructural quality than before drug administra-tion. Uncontrolled studies of pamindronate inchildren with OI reported increased vertebralheight and bone mineral density, decreasedfractures, and improved ambulation (Glorieuxet al., 1998).Conversely, a recent 2-year placebo-controlled study using the oral bisphosphonateolpandronate reported an increase of spinalbone mineral content but no beneficial effectson the functional outcome in the olpandronategroup compared to controls (Sakkers et al.,2004). Moreover, the long-term effects of thisdrug are still under investigation. Thus, thequestion of whether bisphosphonates will alterthe natural history of OI remains unresolved.More controlled trials are essential to deter-mine if increased bone density is associated

with stronger bone, or if increased density cor-relates with increased brittleness.

QUESTIONS

1. A couple with a history of two pregnan-cies resulting in children with type II OIsought genetic counseling prior to subse-quent pregnancy.Which is the most likelyexplanation of their pregnancy history?

2. What information and recommendationswould you give to the couple described inquestion 1?

3. The blue coloration is an indication of ab-normally thin sclerae.Why is it commonlyobserved in type I OI?

4. Examination of type I collagen from a pa-tient with type IV OI revealed normalelectrophoretic mobility but altered ther-mal stability. Which region of the mole-cule is likely to contain the mutation?

5. Why do mutations in C-propeptides causesignificant posttranslational overmodifica-tion of type I collagen?

6. In 1984, a homozygous patient with a re-cessive type III OI was described. Hiscells synthesized nonfunctional α2(I)-chains, and his type I collagen secretedby cells consisted only of that from α1(I)-homotrimers. Recently, several patientswith homozygous and compound het-erozygous null allele mutations inCOL1A2 were described.Their cells pro-duced unstable COL1A2 transcripts anddid not synthesize α2(I)-chains. Theircells also secreted only α1(I)-collagenhomotrimers, but the phenotype of thesepatients had prevalent features of theEhlers-Danlos syndrome rather than OI.What could be the molecular reason forthis drastic phenotype difference?

7. Why does a reduction of collagen thermalstability by 6°C prevent incorporation ofthe corresponding mutant molecules intobone matrix?

8. A teenager with a history of fractures butno skeletal deformities and no acute dis-tress was diagnosed with moderate typeIV OI. After learning that radiographicanalysis revealed low bone mineral den-sity in their child, the parents inquiredabout a possibility of bisphosphonatetherapy. How would you respond to theirinquiry?


Acknowledgments: We are grateful to Joan C. Marini forstimulating discussions and critical comments. The gelsin Figure 3-4 were kindly provided by Wayne A. Cabral.

BIBLIOGRAPHY

Bachinger HP, Morris NP, Davis JM:Thermal stabil-ity and folding of the collagen triple helix andthe effects of mutations in osteogenesis imper-fecta on the triple helix of type I collagen. Am JMed Genet 45:152–162, 1993.

Beck K, Chan VC, Shenoy N, et al.: Destabilizationof osteogenesis imperfecta collagen-like modelpeptides correlates with the identity of theresidue replacing glycine. Proc Natl Acad SciUSA 97:4273–4278, 2000.

Byers PH, Cole WG: Osteogenesis imperfecta, inRoyce PM, Steinmann B (eds): Connective Tis-sue and Its Heritable Disorders. Molecular, Ge-netic, and Medical Aspects, 2nd edition. NewYork:Wiley-Liss, 2002, pp. 385–430.

Dalgleish R: The Human Collagen Mutation Data-base 1998. Nucleic Acids Res 26:253–255, 1998.

Di Lullo GA, Sweeney SM, Korkko J, et al.: Mappingthe ligand-binding sites and disease-associatedmutations on the most abundant protein in thehuman, type I collagen. J Biol Chem 277:4223–4231, 2002.

Engel J, Prockop DJ:The zipper-like folding of col-lagen triple helices and the effects of mutationsthat disrupt the zipper. Annu Rev Biophys Bio-phys Chem 20:137–152, 1991.

Forlino A, Marini JC: Osteogenesis imperfecta:prospects for molecular therapeutics. MolGenet Metab 71:225–232, 2000.

Forlino A, Porter FD, Lee EJ, et al.: Use of theCre/lox recombination system to develop anon-lethal knock-in murine model for osteogen-

esis imperfecta with an alpha1(I) G349C substi-tution.Variability in phenotype in BrtlIV mice. JBiol Chem 274:37923–37931, 1999.

Glorieux FH, Bishop NJ, Plotkin H, et al.: Cyclic ad-ministration of pamindronate in children withsevere osteogenesis imperfecta. N Engl J Med339:947–952, 1998.

Kadler K: Extracellular matrix 1: fibril-forming col-lagens. Prot Profile 1:519–638, 1994.

Kuznetsova NV, Forlino A, Cabral WA, et al.: Struc-ture, stability and interactions of type I collagenwith GLY349-CYS substitution in α1(I)chain ina murine osteogenesis imperfecta model. Ma-trix Biol 23:101–112, 2004.

Lamande SR, Bateman JF: Procollagen folding andassembly: the role of endoplasmic reticulum en-zymes and molecular chaperones. Semin CellDev Biol 10:455–464, 1999.

Leikina E,Mertts MV,Kuznetsova N,et al.: Type I col-lagen is thermally unstable at body temperature.Proc Natl Acad Sci USA 99:1314–1318,2002.

Prockop DJ: Targeting gene therapy for osteogene-sis imperfecta. N Engl J Med 350:2302–2304,2004.

Rauch F, Glorieux FH: Osteogenesis imperfecta.Lancet 363:1377–1385, 2004.

Sakkers R, Kok D, Engelberg R, et al.: Skeletal ef-fects and functional outcome with olpadronatein children with osteogenesis imperfecta: a 2year randomised placebo control study. Lancet363:1427–1431, 2004.

Schwarze U, Hata R, McKusick VA, et al.: Rare auto-somal recessive cardiac valvular form of Ehlers-Danlos syndrome results from mutations in theCOL1A2 gene that activate the nonsense-mediated RNA decay pathway.Am J Hum Genet74:917–930, 2004.

Sillence DO, Senn A, Danks DM: Genetic hetero-geneity in osteogenesis imperfecta. J Med Genet16:101–116, 1979.


CASE REPORT

The patient was a 35-year-old white male withα1-antitrypsin deficiency. He received a com-bined liver-kidney transplant for cirrhosiscomplicated by portal hypertension, renal in-sufficiency secondary to membranoproliferativeglomerulonephritis, and combined restrictiveand obstructive pulmonary disease at age 18years.

Jaundice and pruritus were observed at age6 weeks and resolved spontaneously after ap-proximately 2 months. He was hospitalized forpneumonia at age 20 months, and an enlargedliver was noted. A percutaneous needle biopsyspecimen from the liver was interpreted toshow postnecrotic cirrhosis, although reevalua-tion of the biopsy specimen showed the pres-ence of globules that were periodic acid–Schiff(PAS) positive, diastase resistant (Fig. 4-1). Hewas then referred to our institution at age 2.5years for liver transplantation.

At the time of the initial evaluation, the phys-ical examination revealed a well-developedyoung boy in no acute distress.He was not jaun-diced; however, it was notable that he had aprotuberant abdomen with a liver edge palpa-ble 3 cm below the right costal margin in themidclavicular line.The liver was nontender andhard.The spleen was palpable 6 cm below theleft costal margin. No other physical abnormali-ties were noted.

The child’s laboratory tests revealed a nor-mal hematological picture except for a plateletcount of 122,000/mm3 (below normal). He hadnormal liver enzymes except for a serum glu-tamic oxaloacetic transaminase level of 197

4

α1-Antitrypsin Deficiency

SARAH JANE SCHWARZENBERG and HARVEY L. SHARP

42

units/mL (mildly increased). Both his bloodurea nitrogen (BUN) and creatinine levels werenormal. The patient’s serum protein elec-trophoresis was abnormal, with a low serumalbumin of 2.9 g/dL (normal 3.3–4.6 g/dL) andan α1-globulin band that was barely visible. Be-cause of this last finding, protease inhibitor (PI)phenotyping was done on the child and hisfamily. The child’s phenotype was PIZZ, whileboth parents had the heterozygote, that is, thePIMZ phenotype. The nomenclature of the PItypes is based on the electrophoretic mobilityof the various PI types at pH 4.9. PiMM repre-sents the homozygous normal allele; letters al-phabetically before M designate anodalvariants, and those after M designate cathodalvariants.

The child’s subsequent course was one ofgradual hepatic deterioration.At age 3 years, hewas noted to have ascites (intra-abdominalfluid accumulation). This progressed slowly un-til the age of 6 years, when severe ascites andperipheral edema necessitated the initiation ofspironolactone (a potassium-sparing diuretic).Several admissions to the hospital were re-quired over the next 6 years for ascites withscrotal edema. Serum albumin values were per-sistently low, less than 2.0 g/dL. During thistime, the patient also had two episodes of pri-mary peritonitis (intraperitoneal infection) andone episode of α-streptococcal sepsis.

By age 11 years his renal function decreased,as measured by a creatinine clearance of71 mL/min (normal is 105 mL/min). He alsodeveloped protein-losing nephropathy with a24-hour urinary protein excretion of 600 mgthat increased to 14 g after albumin infusions

α1-Antitrypsin Deficiency 43

(normal urinary protein excretion is100 mg/24 h). Because of abnormal coagula-tion studies, renal biopsy was deferred; how-ever, the clinical picture was consistent withthe membranoproliferative glomerulonephritisseen in α1-antitrypsin deficiency.

At age 12, the patient was admitted withacute chest pain from a left spontaneous pneu-mothorax (air within the pleural cavity).This re-quired hospitalization and chest tube insertion,but he recovered without sequelae. After theresolution of this problem, pulmonary functiontesting revealed findings of both severe airwayobstruction and destruction of alveolar lung tis-sue, consistent with emphysema. No furtherpulmonary problems occurred until the patientwas age 16 years, when he developed occa-sional episodes of bronchospasm (spasmodiccontraction of the smooth muscles of thebronchus). Pulmonary function studies at thattime, though improved from those immediatelyfollowing his pneumothorax, still revealed com-bined obstructive and destructive lung disease.

Also at about age 12 years, the patient began

to experience episodes of acute hepatic en-cephalopathy (graded onset of coma broughton by circulating false neurotransmitters). Thefirst of these episodes was the most severe.Thepatient entered the hospital in a confused anddisoriented state and with an elevated ammo-nia level. The immediate problem was easilycontrolled with neomycin (a nonabsorbableantibiotic that kills intestinal, ammonia-formingbacteria) enemas. During this admission, gas-trointestinal bleeding developed, exacerbatingthe hyperammonemia (plasma ammonia of 450µmol/L; normal < 35 µmol/L).The patient grad-ually slipped into grade IV hepatic coma withfundoscopic changes consistent with increasedintracranial pressure. Coagulopathy (increasedbleeding tendency) prevented placement of anintracranial pressure monitor. He was treatedempirically for presumed cerebral edema asso-ciated with acute hepatic encephalopathy withhyperventilation, a mannitol drip, and barbitu-rate coma.The patient recovered without neu-rological sequelae. Despite protein restriction(1.0 g protein per kilogram body weight per

Figure 4-1. Photomicrograph of periportal hepatocytes of liver from patient in casereport. Note variation in size and presence of cytoplasmic globules. The stain used isperiodic acid–Schiff after diastase.

day), the patient had two other milder episodesof hyperammonemia over that year. He contin-ued on limited protein intake, neomycin, andlactulose, with periodic monitoring for sub-clinical hepatic encephalopathy by electroen-cephalography.

Despite these complications of α1-antitrypsindeficiency, he maintained an active life, attend-ing school and camping with his parents. By theage of 16 years, his renal condition had deterio-rated considerably, with a creatinine clearanceof only 23 mL/min. He was therefore acceptedas a candidate for a combined liver-kidney trans-plant.After a 2-year wait for an immunologicallycompatible donor, the transplant was success-fully performed at age 18 years at the Universityof Minnesota Hospitals. He completed highschool and was employed full time in goodhealth for over a decade.

Seventeen years after his transplant, he wasfound to have cirrhosis from hepatitis C, likelyacquired from the many blood transfusions herequired prior to and during his transplanta-tion. He died awaiting a second hepatic trans-plant.

DIAGNOSIS

This patient illustrates a complicated clinicalcourse of α1-antitrypsin deficiency. Our patienthad liver disease that presented during infancyand developed into hepatic cirrhosis. He exhib-ited most of the complications of cirrhosis, in-cluding portal hypertension with ascites,hyperammonemia, malnutrition, and varicealhemorrhage. These complications of cirrhosisare not unique to α1-antitrypsin deficiency, butit is important to note the potential severity ofthe liver disease associated with this condition.

α1-Antitrypsin deficiency is suspected inthree clinical situations:

1. Cholestasis in infancy (see Fig. 4-2).2. Cirrhosis of undetermined etiology at any

age.3. Emphysema early in life, especially if pre-

dominantly basilar.

Liver disease is commonly associated with α1-antitrypsin deficiency and may develop atany age. Approximately 10% to 20% of α1-antitrypsin-deficient infants with the pheno-type PIZZ are first seen for neonatal cholestaticliver disease, as was the child in this casereport. Conjugated hyperbilirubinemia and he-

patomegaly are noted in the first month of life.The clinical course in the first year of life maybe mild or severe. Most children improve withtime and resolve their liver disease by 1–2years of age. Jaundice after 6 months of age isusually an ominous finding, suggesting a signifi-cant deterioration in clinical status within 1year. A decrease in hepatic synthetic capacityaccompanies this deterioration, manifest by adecrease in coagulation factors synthesized bythe liver.

Most children with α1-antitrypsin deficiencyare normal in the newborn period. Approxi-mately 50% of people who are PIZZ do notever manifest liver disease. In the other half,liver disease develops insidiously over manyyears, presenting at some point in adulthood ascirrhosis. Clinically, cirrhosis associated withα1-antitrypsin deficiency is similar to that seenin other forms of childhood liver disease. Mal-nutrition, coagulopathy, and complications ofportal hypertension, including splenomegaly,variceal hemorrhage, and ascites, develop to avarying degree. In the absence of infection oran episode of dehydration, both of which mayprecipitate hepatic deterioration, a patientmay survive for years with cirrhosis and ade-quate hepatic function. Hepatocellular carci-noma is more common in individuals withPIZZ-associated α1-antitrypsin deficiency.Whileit is often a complication of long-standing cir-rhosis, carcinoma can develop in individualswith α1-antitrypsin deficiency who do not havecirrhosis. It is most commonly seen in olderadults with this disease.

While liver disease may become manifest atany age, from infancy to extreme old age,emphy-sema is extremely rare in childhood, presentingmore commonly in the adult α1-antitrypsin-deficient individual. Indeed, it represents themost common manifestation of the deficiency.The emphysema associated with α1-antitrypsindeficiency (representing about 2% of all casesof emphysema) develops at a relatively early age(in the third or fourth decade of life) with anequal distribution between men and women.The majority of young symptomatic PIZZ indi-viduals with emphysema have a history of ciga-rette smoking. Abstention from cigarettes maydelay the disease onset by 20 years. In a study of22 nonsmokers homozygous for α1-antitrypsindeficiency living in areas free of urban air pollu-tion, it was found that the onset of emphysemawas later than in smokers (51.4 years for smok-ers compared to 71.0 years for nonsmokers).



More nonsmokers were completely free ofsymptoms or had only mild symptoms in theirsixth or seventh decades.

Emphysema usually presents with shortnessof breath, dyspnea, and chronic cough. Pneu-mothorax may result from the bursting of anemphysematous bleb. The emphysema associ-ated with α1-antitrypsin deficiency is indistin-guishable from nonfamilial forms, althoughchronic bronchitis and the associated cough oc-cur less frequently than in other forms of em-physema.

There is a wide range of disability associatedwith the pulmonary complications of α1-antitrypsin deficiency, from completely asymp-tomatic individuals to chronic pulmonarycripples. Most patients develop chronic ob-structive pulmonary disease. Unfortunately,once emphysema becomes symptomatic in theα1-antitrypsin-deficient individual, it usually

pursues a relentless course. A Danish registrystudy reported that the life expectancy of PIZZsmokers is 52 years and that of those whonever smoked is 69 years.

Renal disease is seen in 17% of infants withα1-antitrypsin deficiency. It causes massive pro-tein loss, hypoalbuminemia, and renal failure.The kidney disease is an immunological disor-der occurring only in patients with liver dis-ease, resulting in a membranoproliferativeglomerulonephritis with immunoreactive α1-antitrypsin antigen present, as well as comple-ment and immunoglobulins. It should not beconfused with the nonspecific spotty asympto-matic glomerulonephritis or hepatorenal syn-drome, commonly seen in cirrhosis.

Vascular conditions have been associatedwith α1-antitrypsin deficiency, including panni-culitis (an ulcerative, necrotizing skin condi-tion) and cerebral aneurysm, among others.

Figure 4-2. Photomicrograph of a portion of a hepatocyte from a cholestatic infantwith PIZ phenotype.Amorphous α1-antitrypsin (AT) is in the transitional zone betweenrough and smooth endoplasmic reticulum. B, bile; T, triglyceride droplet.

It is controversial whether heterozygote(PIMZ) individuals are at risk for liver or lungdisease. Some studies have suggested that PIMZindividuals are overrepresented among patientswith chronic end-stage liver disease. PIMZ indi-viduals are also overrepresented in diagnosticliver biopsies with cirrhosis as apposed toonly mild fibrosis. It is equally unclear whethercarriers of the Z allele are at increased risk ofemphysema if they smoke. Large prospectivestudies of heterozygotes would solve thesedilemmas.

The diagnosis of α1-antitrypsin deficiencymay be suspected during direct observationof the cellulose acetate serum protein elec-trophoresis, in which the α1-globulin band issmall or undetectable. As α1-antitrypsin repre-sents 90% of the total α1-globulin peak, depres-sion of this peak usually represents deficiencyof that protein. Serum levels of α1-antitrypsinare low in PIZZ individuals, less than 15% ofnormal levels. Serum quantitative tests do notpermit accurate diagnosis and genetic counsel-ing; therefore, abnormal levels of α1-antitrypsinon screening examination should lead to morespecific diagnostic tests.This is true of individu-als with intermediate levels of α1-antitrypsin aswell. While it is likely that a patient with aserum α1-antitrypsin level of 30%–80% is un-likely to have PIZZ, they may well have α1-antitrypsin deficiency predisposing them tolung disease. Appropriate counseling and dis-ease management requires knowledge of thepatient’s specific phenotype. The diagnostictest of choice is PI typing, in which isoelectricfocusing is used to separate the various α1-antitrypsin species in the individual’s serum bycharge differences. Comparison with sera ofknown PI type permits identification of thephenotype of the individual.

In patients with evidence of significant he-patic involvement, liver biopsy is important inthe evaluation of α1-antitrypsin-associated liverdisease as it facilitates a more accurate progno-sis. The patient with α1-antitrypsin deficiencymay have one of several histological pictures. Inthe neonate with α1-antitrypsin deficiency, theliver may have evidence of fibrosis or cirrhosis,and some specimens will demonstrate paucity ofthe intrahepatic bile ducts. However, the usualhistological finding in the neonate is prolifera-tion of the intrahepatic bile ducts that simulatesthe histological picture of extrahepatic biliaryatresia.There are usually few giant cells (multinu-cleated hepatocytes) seen. α1-Antitrypsin defi-

ciency is therefore an important diagnosis toexclude in the evaluation of the child sus-pected of extrahepatic biliary atresia since itmay mimic this disease clinically and histologi-cally. Patency of the extrahepatic biliary treemay be demonstrated by radionuclide scanning,detection of bile in duodenal contents, orcholangiography.

Periodic acid Schiff (PAS)-positive, diastase-resistant globules representing the retained α1-antitrypsin protein are found in periportalhepatocytes on light microscopy. In the PAS/di-astase study, periodic acid is used to oxidizebonds in sugars in liver tissue, allowing them toreact with Schiff reagent to produce a magentacolor. Glycogen is the material most commonlystained this way, but unsecreted α1-antitrypsinwill also stain with PAS. To allow the α1-antitrypsin to be seen, the glycogen is first di-gested with diastase. The subsequent PAS testthen stain magenta the globules of α1-antitrypsin.These globules represent the abnor-mal α1-antitrypsin not transported from theendoplasmic reticulum. These inclusions arefairly pathognomonic of the deficiency state.They can be verified with α1-antitrypsin anti-body stains and are also seen in heterozygoteindividuals (PIMZ).

Although the average child with α1-antitrypsin deficiency needs no more pul-monary evaluation than careful periodicexaminations, PIZZ adults with symptomaticpulmonary disease require more specific evalu-ation. The first step in evaluating patients sus-pected of having emphysema is the chestroentgenogram and pulmonary function test-ing. The chest x-ray film may show hyperinfla-tion, flattened diaphragms, the presence ofbullae, and narrowing of the pulmonary arter-ies.As basal emphysema is more common in α1-antitrypsin deficiency, the lower zones of thelungs are more commonly involved. Even withnormal x-ray findings, pulmonary function testsmay reveal early emphysematous changes. Theα1-antitrypsin-deficient patient may have a de-crease in forced expiratory volume in 1 secondand forced expiratory flow with an increase intotal lung capacity and functional residual ca-pacity. These findings are probably the resultsof a combination of obstructive lung diseasewith a decrease in the elasticity of the lung.Computed tomography of the lung is more sen-sitive in demonstrating early evidence of lungdestruction. Some authors suggest this modal-ity provides the best prognostic information.


Hypercarbia on examination of arterial bloodgases is nonspecific and generally a late finding.

Renal disease is detected through urinanaly-sis done yearly on PIZZ individuals with liverdisease. Detection of proteinuria would lead toexamination of a quantitative 24-hour urine forprotein. If confirmation of the diagnosis is es-sential, then renal biopsy would be necessary.


α1-Antitrypsin deficiency is an inborn error ofmetabolism predisposing to emphysema; it wasidentified in the early 1960s by Sten Erikssonand C.-B. Laurell (Laurell and Ericksson, 1963).In 1969, Harvey Sharp recognized the associa-tion of hepatic cirrhosis and α1-antitrypsin defi-ciency. The natural course of the liver diseasewas difficult to define at first as most patientswere referred because of serious liver disease.Several decades ago,Tomas Sveger and his asso-ciates (Sveger and Thelin, 2000) began an ongo-ing study of the natural history of this diseasein 1976, identifying all deficient children in a2-year cohort of Swedish infants. This difficultstudy has allowed a more thorough and bal-anced clinical picture of liver disease associatedwith α1-antitrypsin deficiency. During the1960s and 1970s, many investigators examinedthe lung disease of α1-antitrypsin deficiency. Amajor contribution was made by C. Larsson in1978, who convincingly showed that cigarettesmoking was an independent risk factor in thedevelopment of lung disease.

Sequencing of the human α1-antitrypsin genein 1984 initiated an explosion in the study ofthis disease. Transgenic mice were created tostudy the effects of α1-antitrypsin deficiency,and cell culture studies of the regulation of thegene became possible. In 1992, D.A. Lomas andcolleagues (Lomas et al., 1992) showed that theaccumulation of abnormal α1-antitrypsin in theER of the liver was the result of polymerizationof α1-antitrypsin.This discovery set the stage forthe study of the effects of this polymerized ma-terial on the liver, which it is hoped will allowthe mechanism of liver disease associated withα1-antitrypsin deficiency to be determined.

α1-Antitrypsin is a 52-kd glycoprotein pro-duced primarily not only by the hepatocyte,butalso by macrophages. Serum α1-antitrypsin isderived almost exclusively from the liver. Thefunction of this serine protease is to protect tis-sues from proteolytic enzymes released during

the normal inflammatory response. Althoughα1-antitrypsin has some activity against mostserum proteases, its primary protease target iselastase, with some activity against cathepsinG and proteinase 3. The mechanism of inhibi-tion is an irreversible reaction between α1-antitrypsin and elastase at the reactive center ofα1-antitrypsin, a methionine residue at position358. This residue is called the reactive center.Elastase, which cleaves proteins at methionylresidues, recognizes α1-antitrypsin as a sub-strate and attempts to cleave it. In doing so, theprotease is trapped by the α1-antitrypsin mole-cule, and a drastic conformational change of theantiprotease crushes the protease, rendering itinactive and preparing it for clearance from theserum. α1-Antitrypsin represents the primaryregulating factor for elastase. Deficiency of thisantiprotease results in excessive pulmonaryelastase activity.

This antiprotease is encoded by a 12.2-kb(kilobase) gene located on the long arm ofchromosome 14, the protease inhibitor or PI lo-cus.The gene consists of seven exons and six in-trons, with the transcriptional start site varyingdepending on the cell type in which transcrip-tion occurs. The regulation of cell-specific ex-pression is complex. Hepatocytes producepredominantly a single 1.6-kb transcript, al-though they can produce small amounts oftranscripts characteristic of the monocyte cellline when stimulated by interleukin 6. Mono-cytes and macrophages use variable splicing ofthe first exon to produce four different tran-scripts, with lengths varying from 1.8 to 2.0 kb.

The α1-antitrypsin promoter contains a con-sensus TATA box, a B-recognition element fortranscription-activating factor IIB, a hepatocytenuclear factor 1 site, and two non-tissue-specificregions that increase transcription.There is a 3'enhancer region with five potential binding sitesfor transcription factors.The specific factors thatbind in this area remain to be described.

α1-Antitrypsin is an acute phase reactant,which means that synthesis of the protein in-creases during inflammation, including malig-nancy, bacterial infections, and severe burns. Italso rises during certain changes in hormonalconditions, as in pregnancy and during treat-ment with estrogen-containing birth controlpills. The increase in serum α1-antitrypsin ininflammation is regulated primarily at the levelof gene transcription by cytokine mediatorsof inflammation, predominantly interleukin 6.Inflammatory-mediated increased transcription


is regulated by an interaction (as yet unspecified)between the 3' enhancer and the promoter re-gions of the α1-antitrypsin gene.

Sequence similarities to other serine PIs es-tablished the existence of a large evolutionar-ily related gene family, the SERPINs (serineprotease inhibitors). Members of this genefamily are found in plants and viruses as wellas animals. Some examples of SERPINs includeα1-antichymotrypsin, antithrombin III, α2-antiplasmin, thyroid- and corticosteroid-bindingproteins, and ovalbumin.This gene family likelyhas a common ancestry, evolving through geneduplication and mutation. Functionally, there isbroad diversity, from functional PIs to regula-tors of the clotting cascade and hormone-binding proteins.

During protein synthesis, the nascent transla-tion product of the α1-antitrypsin gene is co-translationally translocated to the ER, where thesignal peptide is cleaved and high mannose gly-cosylation residues are added (Fig. 4-3). As theglycoprotein completes its transit through theER to the Golgi apparatus, the terminal man-nose residues of the oligosaccharide moiety arecleaved, and secondary glycosylation residues,including sialic acid, are added.The mature gly-coprotein is packaged in the Golgi and secretedinto the serum. The secreted protein containsvariability in the secondary carbohydrate sidechains, which produces a microheterogeneity,causing the glycoprotein to resolve into multi-ple bands when subjected to acid protein elec-trophoresis.

The mechanism by which Z-α1-antitrypsinis transported out of the ER and degraded hasbeen of interest to investigators. Since intracel-lular trapping of α1-antitrypsin is a significantfeature in the pathophysiology of the liver dis-ease, understanding the degradation of themolecule might provide avenues to correctthe defect. It appears that Z-α1-antitrypsin isbound to calnexin (a chaperone molecule thattransitions abnormally folded molecules to adegradation pathway) longer than it is to M-α1-antitrypsin, which would increase degradationof the Z molecule. The Z-α1-antitrypsin that issecreted has achieved a normal molecular con-formation, suggesting that in the ER, chaperonemolecules are achieving their role as mediatorsof correct conformation. It appears that the ma-jority of Z-α1-antitrypsin is degraded, probablyby the proteasome, and that this requires retro-grade transport of the protein back into the cy-tosol. Some investigators have suggested that

individuals who develop liver disease have a de-creased capacity to clear Z-α1-antitrypsin fromthe ER.

The PI locus is highly pleomorphic, withmore than 75 allelic variants identified. Thenomenclature of the PI locus is based on theelectrophoretic mobility of the α1-antitrypsinvariants at pH 4.9. PIM represents the normalallele; it is actually composed of four M alleles,M1–M4. Letters alphabetically before M desig-nate anodal variants, and those after M designatecathodal variants.The majority of these variantalleles produce no change in α1-antitrypsinserum levels or function. Null alleles exist thatproduce no detectable serum α1-antitrypsin.Since both alleles are expressed codominantly,heterozygosity for the PI locus may be identi-fied electrophoretically. This is important forfamily studies of α1-antitrypsin-deficient pa-tients. Identity of unusual allelic variants may beconfirmed by DNA sequencing.

Although many variant forms of the antipro-tease exist, the most common alleles associatedwith its deficiency are the S and Z variants,bothproducing proteins that migrate cathodal to thenormal protein. Homozygosity for these vari-ants (PIZZ and PISS genotypes), as well as com-pound heterozygosity (PISZ), are all associatedwith deficient serum levels of α1-antitrypsin.PIZZ individuals have approximately 10% to20% of the normal level of α1-antitrypsin, whilePISZ individuals have approximately 35% of thenormal level.The rare individual with homozy-gosity or compound heterozygosity of the nullgene with one of these alleles has severe α1-antitrypsin deficiency.

These abnormal PI alleles are most com-monly found in individuals of European de-scent. Approximately 3% of Europeans areheterozygotes for the Z allele, while about 7%carry the S gene.The Z allele tends to be foundmore commonly in northern Europeans, whilesouthern Europeans tend to have the S allele at aslightly higher frequency.The incidence of seri-ous α1-antitrypsin disease is about 1 in 2000among individuals of northern European extrac-tion. The Z variant, which has been studiedmost extensively, has a single base-pair muta-tion leading to the production of a protein inwhich lysine is substituted for glutamic acid atposition 342. Primary glycosylation residues areadded to the Z-variant polypeptide in the ER,but subsequent transport out of the ER is im-paired.This trapping of the relatively insolubleZ-α1-antitrypsin is the result of the tendency of


Figure 4-3. (Top) Synthetic pathway of α1-antitrypsin diagramming the translocation ofthe nascent polypeptide chain into the endoplasmic reticulum (ER), with addition thereof high-mannose glycosylation residues to the newly folded protein. The protein, carry-ing the “primitive” glycosylation residues, translocates to the Golgi apparatus, where thehigh-mannose residues are cleaved, and sialic acid residues are added. The protein, withmature glycosylation residues, is then secreted. CHO, carbohydrate. (Bottom) Processingof glycosylation residues. High-mannose residues (Man), attached to an oligosaccharidebackbone consisting of N-acetylglucosamine residues (GlcNAc) and glucose residues(Glc) attached to the amino acid asparagine (Asn),comprise the initial glycosylated protein,on the left. In the Golgi, mannose residues are cleaved and further N-acetylglucosamine,galactose (Gal), and sialic acid (Sia) residues are added.The dotted line illustrates the alter-native bi- and triantennary side-chain structures of α1-antitrypsin.

49

Z-α1-antitrypsin to undergo spontaneous poly-merization when present in high concentrationsin the ER. The main β-sheet of the Z-α1-antitrypsin molecule can open spontaneously,allowing the reactive loop of a second Z-α1-antitrypsin molecule to insert itself into theopening. Polymers of Z-α1-antitrypsin are cre-ated with a stable structure that resists the nor-mal degradative processes. Thus, the PIZZindividual accumulates large amounts of endo-plasmic reticular α1-antitrypsin protein that isnot released into the circulation, thereby ac-counting for the markedly reduced serum levelsof α1-antitrypsin. The Z protein that is releasedinto the serum has the ability to inhibit elastaseand has a half-life similar to that of the M protein.Carriers for this allele (PIMZ) have serum α1-antitrypsin levels approximately 60% of normal,with small α1-antitrypsin-containing globulesvisible in their hepatocytes. Other rare pheno-types associated with both low serum levelsof α1-antitrypsin and hepatic globules includePIMMalton and PIMDuarte.

The S variant has a mutation at amino acidposition 264,where valine is substituted for glu-tamic acid. The S-α1-antitrypsin appears to havenormal protease-inhibitory activity but is de-graded intracellularly prior to secretion. Thenull alleles (PIQ) produce no immunologicallyor functionally active protein; there are severalsuch allelic variants described, generally involv-ing deletion of large portions of the α1-antitrypsin gene.

The pathogenesis of the lung disease in PIZZindividuals is fairly well established. Deficiencyof α1-antitrypsin produces an imbalance in theratio of protease to PI, resulting in emphysema.Both alveolar macrophages and polymorphonu-clear leukocytes contain proteases, includingneutrophil elastase, cathepsin G, and proteinase3.Neutrophil elastase is released in large quanti-ties in the lung and has as its substrate elastin, acomponent of the extracellular matrix, as wellas other proteins. During the normal inflamma-tory response in the lung, as might occur withinfection or cigarette smoking, these phago-cytic cells migrate to the lung alveoli, releasingtheir proteases. In the lungs of a normal individ-ual, α1-antitrypsin is the major inhibitor of neu-trophil elastase and keeps elastase activitylocalized to the site of the inflammation, allow-ing it to scavenge damaged proteins and gram-negative bacterial cell walls, thereby protectingalveolar tissue. In the lungs of PIZZ individuals,an imbalance exists in the protease/PI ratio,

allowing elastase uninhibited access to alveolarand pulmonary connective tissue, producingthe emphysematous lesion.

Cigarette smoking markedly increases the riskof emphysema in PIZZ homozygotes. Cigarettesmoke causes the release of neutrophil chemo-tactic factor, increasing the number of protease-containing neutrophils in the lung. Release ofelastase from these cells is increased in the pres-ence of cigarette smoke. Cigarette smoke oxi-dizes the reactive-site methionine in the smallamount of Z-α1-antitrypsin that escapes the liver,destroying its inhibitory capacity for elastase.Thus, emphysema associated with α1-antitrypsindeficiency is not simply due to a genetic defect,but requires an environmental stimulus as well.

There remains controversy regarding thepathophysiology of liver disease associated withα1-antitrypsin deficiency. Studies in transgenicanimals and clinical analysis of human diseasesuggested that the globules of Z-α1-antitrypsintrapped in the ER of the liver either directlydamage the hepatocyte or facilitate damage byinfection or toxins. The mechanism for this in-jury is unclear.

It has been shown in cell culture models ofα1-antitrypsin that Z-α1-antitrypsin in the ERcauses expression of a novel set of stress geneswith products that have the role of restoring ERfunction to normal. Z-α1-Antitrypsin inducestwo signal transduction pathways: the ER over-load response and unfolded protein responsepathways.Via these transduction pathways, thepresence of Z-α1-antitrypsin, as opposed to M-α1-antitrypsin, induces an hepatic inflammatoryresponse. It is thought that this ER-specificstress response results in hepatic injury.

An alternative hypothesis is that ER retentionof Z-α1-antitrypsin results in autophagy, specifi-cally of hepatic mitochondria. The basis for thishypothesis is the increase in autophagosomesin cells engineered for inducible expression ofZ-α1-antitrypsin.The mutant protein, along withthe chaperone molecule calnexin, can be foundin these autophagosomes by immune electronmicroscopy. It is postulated that mitochondrialdysfunction results from the damage to the mi-tochondria in the PIZZ liver, leading to the he-patic injury.

THERAPY

Conventional medical management of emphy-sema consists of supportive care, including early


antibiotic treatment of all pulmonary infections.Patients are immunized against influenza virusand Streptococcus pneunoniae. If,on pulmonaryfunction testing, any reversibility of the airwayobstruction can be effected with bronchodila-tors, these may be used. As the pathogenesis ofemphysema associated with α1-antitrypsin defi-ciency is related to cigarette smoking, it is impor-tant to counsel the patient to stop smoking.Thisis the only measure known to improve life ex-pectancy for these patients.

Fractionation techniques have made it possi-ble to recover active α1-antitrypsin from blood.Use of this product for intravenous replace-ment therapy in deficient individuals hasshown that it is possible to increase levels inthe serum to those of PISZ heterozygotes whoexperience no increase in pulmonary diseaseover the general population. Pulmonary lavageof patients transfused with this productshowed that functional α1-antitrypsin reachesthe alveolar structures. The Food and Drug Ad-ministration has approved weekly administra-tion of purified serum-derived α1-antitrypsin toPIZZ and PI null individuals with pulmonarydisease.Although serum levels of α1-antitrypsinincrease to those believed to be protective, ithas not been possible to show clinical improve-ment. Furthermore, viral transmission via bloodproducts is a significant risk factor.

The use of recombinantly produced α1-antitrypsin would reduce the risk of transfusion-related viral disease; however, its half-life is tooshort to produce adequate serum levels withoutdaily infusion. Studies have shown that the re-combinant product can be administered as anaerosol. It diffuses across the respiratory epithe-lium, enters the lung lymph, and eventuallyreaches the systemic circulation. Unfortunately,it is not known whether this therapy has any in-fluence on the development or progression ofemphysema. Further carefully controlled, multi-center trials are necessary.

Lung transplantation has been used in thetreatment of lung disease associated with α1-antitrypsin deficiency. Survival is the same asfor patients undergoing the procedure forchronic obstructive pulmonary disease.As withany transplantation, lifelong immunosuppres-sion is necessary.

Liver disease is managed conventionally,maintaining appropriate nutritional supportand managing the complications of portal hy-pertension and hepatic failure as they occur.We counsel heterozygote parents of a PIZZ

child to stop cigarette smoking because of thedanger passive cigarette smoke represents tothe lungs of their PIZZ child. Definitive therapyof the liver disease is limited to successful livertransplantation. There is no evidence of recur-rence of the liver disease after successful livertransplant. Theoretically, one would assume thatliver transplantation,by providing normal serumlevels of α1-antitrypsin, would prevent develop-ment of the lung disease; however, studies nec-essary to prove this hypothesis have not beendone. No data support a role for α1-antitrypsinreplacement therapy in liver disease.

Some investigators have speculated that itshould be possible to solubilize intracellular Z-α1-antitrypsin, allowing it to complete its secre-tory pathway. If achieved, this would ameliorateboth the pulmonary and the hepatic disease as-sociated with PIZZ phenotype. Phenylbutyratewas suggested as one possible therapy, but ithas been ineffective for this purpose. Helen Par-frey and colleagues (Parfrey et al., 2004) haveshown in vitro that an oligopeptide that specif-ically binds to Z-α1-antitrypsin can inhibit itspolymerization. The peptide can dissociatefrom the binary complex, allowing the Z-α1-antitrypsin to have full antiprotease activity.This and other novel mechanisms of promotingsecretion rather than degradation of the Z-α1-antitrypsin may allow medical therapy of thiscondition.

Gene therapy offers the hope that the dis-ease might eventually be completely cured.Thehuman α1-antitrypsin gene has been success-fully introduced into rat hepatocytes, whichwhen transplanted into a rat liver,expressed theprotein in small quantities.This strategy wouldmost likely be successful in preventing the de-velopment of emphysema. However, it is un-clear whether it would cure the liver disease. Ifliver disease is dependent on the presence ofunsecreted Z-α1-antitrypsin, then adding thenormal α1-antitrypsin gene would not stop itsdevelopment or progression. Investigators inthis field are working to develop techniques oftargeted homologous recombination to replacecompletely the α1-antitrypsin exon containingthe PIZ mutation. Patients thus treated wouldexpress only the M-α1-antitrypsin.

There is a phase 1 trial (trials that examine thetoxicity of the proposed treatment and deter-mine an appropriate dose of the treatment) of anintramuscularly delivered α1-antitrypsin gene ina modified adeno-associated virus vector. It willbe some time before efficacy data are available;



however, studies using this vector in animalssuggested that it may allow prolonged gene ex-pression at levels sufficient to supply normalserum levels of α1-antitrypsin. It also inducesminimal levels of inflammation. This strategymay allow prevention and treatment of lung dis-ease but is unlikely to affect liver disease.

We recommend that all relatives of a patientwith α1-antitrypsin deficiency be PI typed be-cause relatives of the proband may have clini-cally unsuspected PIZZ liver or lung disease.For future counseling of the family, it is impor-tant to identify the PI type of the patient accu-rately and to identify heterozygotes. Althoughmeasurement of α1-antitrypsin immunologi-cally or functionally is accurate, there may besome difficulty identifying PIMZ heterozy-gotes with these methods since they have α1-antitrypsin levels that may rise to normal levelsduring inflammation. One must therefore relyon PI typing, which should be done by an insti-tution familiar with the many α1-antitrypsinalleles.

Neonatal screening for α1-antitrypsin defi-ciency has been debated for years. Swedenscreened a cohort of 200,000 infants born in1972–1974. Subsequent evaluation of thesechildren has shown that there was little psy-chosocial impact on the family from the identi-fication of a child as α1-antitrypsin deficientthrough screening.A slight increase in maternalanxiety was the only negative impact found.Discrimination by insurance companies andemployers remains a theoretical risk. Impor-tantly, adolescents with α1-antitrypsin deficiencyidentified through screening were significantlyless likely to smoke than their peers.This find-ing, combined with the capacity to preventpassive smoke exposure in children with α1-antitrypsin deficiency, has led many investiga-tors to call for screening programs in countrieswith large numbers of affected births.

It is possible to offer parents who are knownheterozygotes for the Z allele low risk early di-agnosis of α1-antitrypsin deficiency in utero.Prenatal diagnosis can be done on fibroblastscollected by amniocentesis or by chorionic vil-lus biopsy using polymerase chain amplificationand analysis of the mutated region of the gene.Counseling of the families of a PIZZ individualrequires a careful explanation of the principlesof codominant inheritance. Parents who areboth PIMZ heterozygotes have a one-in-fourrisk of having a PIZZ child in any subsequent

pregnancy.The risk of serious liver disease in aPIZZ child is small, and the risk of lung diseasecan be significantly modified by avoiding ciga-rette smoking.Thus, there are few medical rea-sons to recommend in utero diagnosis ofα1-antitrypsin deficiency, although ultimatelythe decision should be left to the family.

QUESTIONS

1. Describe the range of possible clinicalsigns and symptoms of α1-antitrypsin defi-ciency in a newborn.

2. The mother of the child in the case reportwas PIMZ. Discuss the likely results of aserum α1-antitrypsin level and a liverbiopsy in the mother. If her serum α1-antitrypsin level was found to be 75% ofnormal, then how would you explain it?

3. If the parents of the proband sought gene-tic counseling from you in a subsequentpregnancy, then what information wouldyou give them?

4. If a drug existed that specifically stimu-lated the transcription of messenger ri-bonucleic acid for α1-antitrypsin in theliver, then would this be a reasonable ther-apy for PIZZ individuals? Why or why not?Would your recommendation be differentfor the PISS patient?

5. As the patient described reaches maturity,what medical advice would you give himto help him protect his lungs?

BIBLIOGRAPHY

Carrell RW Lomas DA:α1-antitrypsin deficiency—amodel for conformational disease. N Engl J Med346:45–53, 2002.

Graziadei IW, Joseph JJ, Wiesner RH, et al.: In-creased risk of chronic liver failure in adultswith heterozygous α1-antitrypsin deficiency. He-patology 28:1058–1063, 1998.

Ibarguen E, Gross CR, Savik SK, et al.: Liver diseasein α1-antitrypsin deficiency: prognostic indica-tors. J Pediatr 117:864–870, 1990.

Kalsheker N, Morley S, Morgan K: Gene regulationof the serine proteinase inhibitors α1-antitrypsinand α1-antichymotrypsin. Biochem Soc Trans30:93–98, 2002.

Laurell CB, Ericksson, S: The electrophoretic α1-globulin pattern of serum in α1-antitrypsindeficiency.Scand J Clin Lab Invest 15:132,1963.

Lomas DA, Evans DL, Finch JT, et al.: The mecha-nism of Z α1-antitrypsin accumulation in theliver. Nature 357:605–07, 1992.

Needham M, Stockley RA: α1-Antitrypsin defi-ciency. 3: Clinical manifestations and natural his-tory. Thorax 59:441–445, 2004.

Parfrey H, Dafforn TR, Belorgey D, et al.: Inhibitingpolymerization: new therapeutic strategies for Zα1-antitrypsin-related emphysema. Am J RespirCell Mol Biol 31(2):133–9, 2004.

Perlmutter DH: α1-Antitrypsin deficiency. SeminLiver Dis 18:217–225, 1998.

Perlmutter DH: α1-Antitrypsin deficiency: liverdisease associated with retention of a mutantsecretory glycoprotein in the endoplasmic retic-ulum. Methods Mol Biol 232:39–56, 2003.

Seersholm N, Kok-Jensen A: Clinical features andprognosis of life time non-smokers with severeα1-antitrypsin deficiency. Thorax 53:265–268,1998.

Silverman GA, Bird PI, Carrell RW, et al.: The ser-pins are an expanding superfamily of struc-turally similar but functionally diverse proteins.Evolution, mechanism of inhibition, novel func-tions, and a revised nomenclature. J Biol Chem276:33293–33296, 2001.

Stecenko AA, Brigham KL: Gene therapy progressand prospects: α1-antitrypsin. Gene Ther 10:95–99, 2003.

Sveger T, Thelin T: A future for neonatal α1-antitrypsin screening? Acta Paediatr 89:628–631, 2000.


5

Cardiac Troponin: Clinical Role in the Diagnosis of Myocardial Infarction

FRED S.APPLE and ALLAN S. JAFFE

54

CASE REPORT

The patient, a 63-year-old Caucasian female, washospitalized on 4 April 2002 though 10 April2002 for a non-ST segment elevation myocardialinfarction (non-Q-wave MI per chart documen-tation). She had a negative adenosine stress testafter the initial event.Her serum cardiac-specifictroponin I (cTnI) concentration 24 hours afterher onset of chest pain was 1.4 µg/L (upperlimit of normal is 0.3 ng/mL), and her creatinekinase (CK) MB level was 12.5 µg/L (upper limitof normal 6.0 ng/mL).Three days post-event hercTnI level was 0.5 µg/L and her CK-MB levelwas 4.5 µg/L (Fig. 5-1). MB refers to one of theisoenzyme forms of CK found in serum. Theform of the enzyme that occurs in brain (BB)does not usually get past the blood-brain barrierand therefore is not normally present in theserum. The MM and MB forms account for al-most all of the CK in serum.Skeletal muscle con-tains mainly MM, with less than 2% of its CK inthe MB form. MM is also the predominant myo-cardial creatine kinase and MB accounts for10%–20% of creatine kinase in heart muscle.

At 72 hours after presentation, the patientexperienced new-onset chest pain, described asa burning pain in the left shoulder, arm,and epi-gastrium.The electrocardiogram (ECG) demon-strated only nonspecific T-wave abnormalitiesand was not different from the one obtained atthe time of her initial presentation. Normal si-nus rhythm was now present. Nitroglycerinprovided some relief. Based on new symptoms,along with recurring T-wave abnormalities and

an increasing cTnI,diagnosis of reinfarction (ex-tension of her initial event) was made.The cTnIconcentration on the day of her suspected rein-farction (day 4) was increased to 1.8 µg/L witha corresponding CK-MB value of 13.6 µg/L. Car-diac catheterization revealed a 95% distal leftanterior descending stenosis, a 95% mid-rightcoronary artery narrowing, and a 90% occludedcircumflex proximally. Stents were placed inboth the distal and proximal right coronary ar-tery.The rest of her hospital stay was unevent-ful, and she was discharged home on day 7. At3-month follow-up, the patient was participat-ing in a cardiac rehabilitation program and do-ing well.

DIAGNOSIS

The term acute myocardial infarction (AMI)is defined as an imbalance between myocar-dial oxygen supply and demand, resulting ininjury to and the eventual death of myocytes.When the blood supply to the heart is inter-rupted,“gross necrosis” of the myocardium re-sults. Abrupt and total loss of coronary bloodflow usually results in a clinical syndromeknown as ST segment elevation AMI (STE AMIor Q-wave MI; diagnostic by electrocardio-gram) because of the characteristic electrocar-diographic changes that occur. Partial loss ofcoronary perfusion, if severe, also can lead tonecrosis, which is generally less extensive, andthis type of infarction is usually termed non–STelevation myocardial infarction (NSTEMI or

Cardiac Troponin in MI Diagnosis 55

non–Q-wave MI; not diagnostic by electrocar-diogram). There is considerable overlap be-tween the pathophysiology and the pattern ofnecrosis of the two entities. Other events of lessseverity may be missed entirely or if detectedmay be diagnosed as angina that can range fromstable to unstable.

The term acute coronary syndrome (ACS)encompasses most of the patients defined so farin this chapter who present with unstable is-chemic heart disease. Most of these syndromesoccur in response to an acute event in the coro-nary artery when circulation to a region of theheart is obstructed. If the obstruction is highgrade and persists, then necrosis usually ensues.Since necrosis takes some time to develop, it isapparent that therapy, including opening theblocked coronary artery in a timely fashion, of-ten can prevent some of the death of myocar-dial tissue. These syndromes are usually, but notalways, associated with chest discomfort.

Previously, the diagnosis of AMI establishedby the World Health Organization required atleast two of the following criteria: a history ofchest pain, evolutionary changes on the ECG, orelevations of serial cardiac biomarkers (initiallydefined as a twofold increase of total serum CKor CK-MB). However, it was rare for a diagnosisof AMI to be made in the absence of biochemi-cal evidence of myocardial injury.A 2000 Euro-pean Society of Cardiology/American Collegeof Cardiology (ESC/ACC) consensus conferencehas codified the role of biomarkers, with spe-cific focus on cardiac troponins, by advocating

that the diagnosis be based on evidence of myo-cardial injury based on biomarkers of cardiacinjury in the appropriate clinical situation.

For these guidelines, either of the followingcriteria satisfies the diagnosis for an acute,evolving, or recent MI.The first is a typical riseor gradual fall of cardiac troponin, or morerapid rise and fall of CK-MB, with at least one ofthe following: (1) ischemic symptoms; (2) de-velopment of pathologic Q waves on the ECG;(3) ECG changes indicative of ischemia (ST seg-ment elevation or depression); or (4) coronaryartery intervention (e.g., coronary angioplasty).For the second, there should be pathologic find-ings of an AMI as identified at autopsy. Theguidelines recognized the reality that neitherthe clinical presentation nor the ECG had ade-quate sensitivity and specificity, but that the tro-ponin markers, in particular, could provideboth. These guidelines do not suggest that all el-evations of these biomarkers should elicit a di-agnosis of AMI, only those associated with theappropriate clinical (ischemic presentation)and ECG findings. When elevations of cardiactroponin are observed that are not due to acuteischemia, the clinician is obligated to search foranother etiology for the cardiac injury.

Patients with ACS can be categorized intofour groups. First, there is the group of patientswho present early to the emergency room,within 0 to 4 hours after the onset of chestpain, and who lack diagnostic ECG evidence ofAMI. These patients require rapid laboratorytesting for evidence of cardiac injury. Thus, use-ful laboratory markers of cardiac injury arethose that are released rapidly from the heartand are highly specific for cardiac myoctyedamage. These assays must be rapid and sensi-tive enough to detect even the small changeswithin the reference interval that can occur inblood early after the onset of symptoms.

The second patient group presents 4 to 48hours after the onset of chest pain but withoutdiagnostic evidence of AMI by ECG. This groupof patients also requires serial monitoring ofcardiac biomarkers and ECG changes.

The third group is patients who present stilllater,more than 48 hours after the onset of chestpain, and also lack diagnostic ECG changes. Theideal biomarker of myocardial injury for thisgroup would have to be one that persists in thecirculation for several days to provide a late di-agnostic time window. The shortfall of such amarker might be its inability to distinguish re-current injury from the prior, older injury.

Figure 5-1. Time-course of changes in serum car-diac troponin I and creatine kinase MB (CK-MB)following myocardial infarction and subsequentreinfarction during hospitalization. Cardiac-specifictroponin I (cTNI), open squares; CK-MB, filled cir-cles. Reprinted from Apple and Murakami (2005).

The fourth group is those who present to theemergency department at any time after the on-set of chest pain with clear ECG evidence ofAMI. In this group, detection with serum bio-markers of myocardial injury is not necessary ini-tially. Many of these patients may qualify forreperfusion therapy at a time before blood mark-ers of cardiac injury have increased, and therapyshould not be withheld if these criteria are met.Subsequently, specific and sensitive myocardialmarkers could be employed to monitor the suc-cess of reperfusion during the 60- to 90-minuteperiod after therapy.Rapid assays providing earlyserial values followed by interpretation of themarkers’ patterns of appearance are often help-ful in determining subsequent management.


Cardiac Troponin I and T

The contractile proteins of the myofibril in-clude three troponin regulatory proteins. Thetroponin complex includes three protein sub-units, troponin C (the calcium-binding compo-nent), troponin I (the inhibitory component),and troponin T (the tropomyosin-binding com-ponent). The subunits exist in a number of iso-forms. The distribution of these isoforms variesbetween cardiac muscle and slow- and fast-twitch skeletal muscle. Only two major iso-forms of troponin C are found in human heartand skeletal muscle. These are characteristic ofslow- and fast-twitch skeletal muscle. The heartisoform is identical with the slow-twitch skele-tal muscle isoform. Isoforms of cardiac-specifictroponin T (cTnT) and cTnI also have beenidentified and are the products of unique genes.All cardiac troponins are localized primarily inthe myofibrils (94%–97%), with a smaller cyto-plasm fraction (3%–6%).

Cardiac troponin subunits I and T are en-coded by different genes than the respectiveskeletal muscle isoforms and have differentamino acid sequences, giving them unique car-diac specificity. cTnI has never been shown tobe expressed in normal, regenerating, or dis-eased human or animal skeletal muscle. By con-trast, small amounts of cTnT are expressed asone of four identified isoforms in skeletal mus-cle during human fetal development, in regener-ating rat skeletal muscle, and in diseased humanskeletal muscle. cTnT isoform expression hasbeen demonstrated in skeletal muscle specimens

obtained from patients with muscular dystrophy,polymyositis, dermatomyositis, and end-stage re-nal disease. Thus, care is necessary to choose an-tibody pairs for cardiac assay use that do notdetect the isoforms reexpressed in noncardiactissue. The commercial assay used in clinicalpractice only detects the heart cTnT form.

Cardiac troponin I exists as a part of the tro-ponin T-I-C ternary complex as a structural andregulatory component of the myofibril. Follow-ing myocardial injury, multiple forms of cardiactroponins are elaborated both in tissue and inblood (Fig. 5-2). These include the T-I-C ternarycomplex, IC binary complex, free I, and multiplemodifications of these three forms resultingfrom oxidation, reduction, phosphorylation, anddephosphorylation, as well as both C- and N-ter-minal degradation. What is elaborated likely re-flects the nature of the injurious stimulus, bloodflow that determines how long the protein re-mains in the tissue prior to reaching the circula-tion, the timing of the insult (i.e., forms maychange as the tissue damage evolves), and per-haps genetics. Depending on which fragmentsare elaborated, the selection of antibodies usedto detect cTnI (i.e., different antibody configura-tions) can lead to substantially different recogni-tion patterns. It is now clear that assays need tobe developed with the antibodies that recognizeepitopes in the stable region of cTnI and ideallydemonstrate an equimolar response to the differ-ent cTnI forms that may circulate in the blood.

Creatine Kinase Isoenzymes and Isoforms

Three cytosolic isoenzymes (CK-MM, CK-MB,CK-BB) and one mitochondrial isoenzyme (CK-Mt) of CK have been identified. Three differentgenes have been identified that encode for andare specific for CK-M, CK-B, and CK-Mt sub-units.Although CK-MM is predominant in bothheart and skeletal muscle, CK-MB has beenshown to be more specific for the myocardium,which contains 10% to 20% of its total CK activ-ity as CK-MB, compared to amounts varyingfrom 0% to 7% in skeletal muscle.

Early studies involving animal hearts or speci-mens obtained at autopsy from human heartssuggested a uniform distribution of CK-MB rang-ing from 5% to 50% of the total CK activity. How-ever, it has been shown by Ingwall andcolleagues (1985) that the proportion of CK-MBwas 6% to 15% lower in the surrounding normalareas of tissue than in infarcted myocardium in


humans. When studied more completely in hu-mans, CK-MB concentrations ranged from 15% to24% of total CK in myocardial tissue obtainedfrom patients with left ventricular hypertrophy(LVH) due to aortic stenosis, from patients withcoronary artery disease (CAD) without LVH, andfrom patients with CAD and LVH due to aorticstenosis. In contrast, patients with normal leftventricular tissue had a low percentage of CK-MB(<2%).These data suggest that changes in the CKisoenzyme distribution are dynamic and occur inhypertrophied and diseased human myocardium.Diseased cells also have less total CK per cell.

Normal skeletal muscle usually contains verylittle CK-MB.Levels as high as 5%–7% have beenreported in some muscles, but less than 2% ismuch more common. Severe skeletal muscle in-jury following trauma or surgery can lead to ab-solute elevations of CK-MB above the upperreference (normal) limit of CK-MB in serum. In-creases in serum total CK and CK-MB in severalpatient groups often present a diagnostic chal-lenge to the clinician. Persistent elevations ofserum CK-MB resulting from chronic muscledisease occur in patients with muscular dystro-phy, end-stage renal disease, and polymyositis (agenerative disease of skeletal muscle) as well asin healthy subjects who undergo extreme exer-cise or physical activities.The increase in serumCK-MB in runners, for example, may be related

to the adaptation by the skeletal muscle duringregular training and after acute exercise, result-ing in increased CK-MB tissue concentrations.The mechanism responsible for increased CK-MB in skeletal muscle following chronic muscledisease or injury is thought to be due to the re-generation process of muscle, with reexpres-sion of CK-B genes similar to those found in theheart, thus giving rise to increased CK-MB levelsin skeletal muscle. Thus, skeletal muscle can be-come like heart muscle in its CK isoenzymecomposition, with up to 50% CK-MB in somepatients with severe polymyositis.

IMMUNOASSAY PERSPECTIVES FOR MONITORING CARDIACBIOMARKERS

Cardiac Troponins

The first assay (a radioimmunoassay) that mea-sured cTnI used polyclonal anti-cTnI antibodies.The first monoclonal enzyme-linked immunosor-bent assay, anti-cTnI antibody-based immunoas-say, was described by Bodor and co-workers(1992). Numerous manufacturers have now de-veloped monoclonal antibody-based diagnosticimmunoassays for the measurement of cTnI inserum.Assay times range from 5 to 30 minutes.


Figure 5-2. Schematic of cardiac troponin I and T release following myocardial cellnecrosis into the circulation, demonstrating the multiple forms that exist in the blood.cTnI, cardiac-specific troponin I; cTnT, cardiac-specific troponin T.


As shown in Table 5-1, over a dozen assays havebeen cleared by the Food and Drug Administra-tion (FDA) for patient testing within the UnitedStates on central laboratory and point-of-care(POC) or near-bedside testing platforms. In addi-tion to these quantitative assays, several assayshave been FDA-cleared for the qualitative deter-mination of cTnI.

In practice, two obstacles limit the ease forswitching from one cTnI or cTnT assay to an-other. First, there is currently no primary refer-ence cTnI material available for manufacturers touse for standardizing their assays. Second, be-cause of the different epitopes recognized by thedifferent antibodies used, assay concentrationsfail to agree. While standardization of assays re-mains elusive, harmonization of cTnI concentra-tions by different assays has been narrowed froma 20-fold difference to a 2- to 3-fold difference.

Several adaptations of the Roche DiagnosticscTnT immunoassay have been described, result-ing in an FDA-cleared third-generation assayavailable worldwide. Two monoclonal anticar-diac troponin T antibodies are used in the third-generation assay. Skeletal muscle TnT is nolonger a potential interferent, as was found inthe first-generation enzyme-linked immunosor-bent cTnT assay. In contrast to cTnI, no stan-dardization bias exists for cTnT because thesame antibodies (M11, M7) are used in both the

central laboratory and POC quantitative andPOC qualitative assay systems.

In 2001, quality specifications for cardiactroponin assays were published. These specifi-cations were intended for use by the manufac-turers of commercial assays and by clinicallaboratories utilizing troponin assays.The over-all goal was to attempt to establish uniform cri-teria so all assays could objectively be evaluatedfor their analytical qualities and clinical perfor-mance. Both analytical and preanalytical factorswere addressed. The following recommenda-tions have been proposed:

1. The antibody specificity (which epitopelocations are identified) needs to be de-lineated. Epitopes located on the stablepart of the cTnI molecule should be apriority.

2. Assays need to clarify whether differentcTnI forms (i.e.,binary vs.ternary complex)are recognized in an equimolar fashion bythe antibodies used in the assay.Specific rel-ative responses need to be described forthe following cTnI forms: free cTnI; the I-Cbinary complex; the T-I-C ternary complex;and oxidized, reduced, and phosphorylatedisoforms of the three cTnI forms.

3. The effects of different anticoagulants onbinding of cTnI need to be addressed.

Table 5-1. Cardiac Troponin Assays Cleared by the Food and Drug AdministrationAssay LLD 99th ROC 10% CV

Abbott AxSYM ADV* 0.02 0.04 0.4 0.16

Architect* 0.009 0.012 0.3 0.032

Bayer ACS 0.03 0.1 1.0 0.35

Centaur 0.02 0.1 1.0 0.35

Beckman Access* 0.01 0.04 0.5 0.06

Biosite Triage* 0.19 0.19 0.4 0.5

Dade RxL* 0.04 0.07 0.6–1.5 0.14

CS* 0.03 0.05 0.6–1.5 0.06

DPC Immulite 0.1 0.2 1.0 0.6

i-STAT i-STAT 0.03 0.08 ND 0.1

Ortho Vitros 0.02 0.08 0.4 0.12

Roche Elecsys† 0.01 0.01 0.1 0.03

Reader 0.05 <0.05 0.1 NA

Tosoh AIA* 0.06 0.06 0.31–0.64 0.06

LLD, lower limit of detection; 99th, 99th percentile reference limit; ROC, receiver operator characteristic curve optimizedcutoff; 10% CV, lowest concentration to provide a total imprecision of 10%.

*Second generation.

†Third generation

NA, not available

4. The source of material used to calibratecTn assays, specifically for cTnI, should betraceable.

While clinicians and laboratorians continueto publish guidelines supporting turnaroundtimes (TATs; defined as the time from blooddraw to reporting of the result to a health careprovider) of less than 60 minutes for cardiacbiomarkers, the largest TAT study published todate demonstrated that TAT expectations arenot being met in a large proportion of hospitals.A survey study of 7020 cardiac troponin and4368 CK-MB determinations in 159 hospitalsdemonstrated that the median and 90th per-centile TATs for troponin and CK-MB were as fol-lows, respectively: 74.5 min, 129 min; 82 min,131 min.Less than 25% of hospitals were able tomeet the TAT in less than 60 minutes. Prelimi-nary data has shown that implementation ofPOC cardiac troponin testing can decrease TATsto less than 30 minutes in cardiology criticalcare and short-stay units. These data highlightthe continued need for laboratory services andhealth care providers to work together to de-velop better processes to meet a TAT that is lessthan 60 minutes as requested by physicians.

Reference Intervals for CardiacTroponins and Creatine KinaseIsoenzymes

If possible, each laboratory should determine a99th percentile of a reference group for cardiactroponin assays using the specific assay used inclinical practice or validate the assay based onfindings in the literature. Further, acceptable im-precision (coefficient of variation, %CV ) of eachcardiac troponin assay (as well as for CK-MBmass assay) has been defined as 10% or lowerCV at the 99th percentile reference limit. Unfor-tunately, the majority of laboratories do not havethe resources to perform adequately poweredreference interval studies. Therefore,clinical lab-oratories have to rely on the peer-reviewed pub-lished literature to establish reference intervals.When reviewing reference studies, caution mustbe taken when comparing the findings reportedin the manufacturer’s approved package insertswith the findings reported in journals because ofdifferences in total sample size, distributions bygender and ethnicity, age ranges, and the statisticused to calculate the 99th percentile given.

There is no established guideline set tomandate a consistent evaluation of the 99th

percentile reference limit for cardiac troponins.The largest and most diverse reported refer-ence interval study to date showed plasma (he-parin; used for anticoagulation of blood) 99thpercentile reference limits for eight cardiac tro-ponin assays (seven cTnI, one cTnT) and sevenCK-MB mass assays (Table 5-2). These studieswere performed in 696 healthy adults (agerange 18 to 84 years) stratified by gender andethnicity. The data demonstrate several issues.First, two cTnI assays showed a 1.2- to 2.5-foldhigher 99th percentile for males versus females.Second, two cTnI assays demonstrated a 1.1- to2.8-fold higher 99th percentile for African Amer-icans versus Caucasians.Third, there was a 13-fold difference between the lowest versus thehighest measured cTnI 99th percentile limit.The lack of cardiac troponin assay standardiza-tion (there is no primary reference materialavailable) and the differences in antibody epi-tope recognition between assays (different as-says use different antibodies) give rise tosubstantially discrepant results.

Although many studies have addressed the to-tal imprecision of cTn assays, the manufacturers’package inserts prefer to publish imprecisiondata primarily based on within-run or within-dayprecision.Again, there is no consistent specifica-tion regarding the precision value that shouldbe reported in the package insert. Publishedfindings demonstrating the total imprecision for13 commercial assays have indicated none ofthe assays were able experimentally to achieve a10% CV (total imprecision) at their 99th per-centile cutoff.

Therefore, to avoid the potential for false-positive diagnostic criteria based on cTn moni-toring at the 99th percentile, a group of expertsin both the laboratory medicine and cardiologycommunities has endorsed the concept that un-til cardiac troponin assay imprecision improvesat the low concentrations, the lowest concentra-tions to attain a 10% CV (CV is the same as pre-cision) should be used as a modified ESC/ACCdiagnostic cutoff for detection of myocardial in-jury. This concept has been endorsed by severalcardiology and laboratory medicine groups.The ultimate goal will be to have all cTn assaysattain a 10% CV at the 99th percentile referencelimit.This approach should reduce false-positiveanalytic results from lack of imprecision valuesbetween the 10% CV cutoff and the 99th per-centile. However, all biomarker increases abovethe 99th percentile should be interpreted cau-tiously, especially in the high-risk patient, and



followed with serial samples over a 6- to 12-hourperiod after presentation.

Myocardial Infarction Detection Ratesand Cardiac Troponins

Advances in diagnostic technology for the de-velopment of improved low-end analyticaldetection of cardiac troponins have impactedthe prevalence of AMI detection.Accumulatingdata suggest that the more sensitive cardiactroponin tests result in greater rates of MI diag-nosis and greater rates of cardiac troponin pos-itivity, compared to CK-MB. Smaller MIs will bedetected. Clinical cases that were earlier classi-fied as unstable angina will be given a diagno-sis of MI (due to an increased cTn), and nowprocedure-related troponin increases (i.e., fol-lowing angioplasty) will be labeled MI.

The importance of small troponin increaseshas been confirmed by their association with apoor prognosis. Based on several studies thatcompared CK-MB and cardiac troponin assaysin patients with ACS, a substantial increase inrate of MIs ranging from 12% to 127% was de-tected. In one study of 1719 patients with ACSpresenting to rule in/rule out MI, a subset (5%)of cTnI-negative but CK-MB-positive patients re-vealed the potentially underlying false-positiveMI rate when using CK-MB as a standard for MIdetection. This was likely due to release of CK-MB from skeletal muscle, in the absence of myo-cardial injury. Further, a subset (12%) of

cTnI-positive, CK-MB-negative patients demon-strated a subset of MIs that would not havebeen detected without cardiac troponin moni-toring.All these data taken together support theimplementation of cardiac troponin in place of,not in combination with, CK-MB. This fact willthen have an impact on the prognosis of AMIoverall.

Creatine Kinase MB

CK-MB can be measured in numerous ways. Im-munoassays developed in recent years have im-proved on the analytical and clinical sensitivityand specificity of the earlier immunoinhibitionand immunoprecipitation assays. These assaysnow: (1) measure CK-MB directly and providemass measurements, (2) are easily automated,and (3) provide rapid results (≤30 minutes).Mass assays reliably measure low CK-MB con-centrations in both samples with low total en-zyme activity (<100 U/L) and with high totalenzyme activity (>10,000 U/L). Furthermore,nointerferences from other proteins have beendocumented. The majority of commerciallyavailable immunoassays that use monoclonalanti-CK-MB antibodies are the same as thoselisted in Table 5-2 for cardiac troponin assays.Excellent concordance has been shown be-tween mass concentration and activity assays.A primary reference material is commerciallyavailable to assist in harmonization. If used forassay standardization, then this material allows

Table 5-2. Heparin-Plasma 99th Percentile Reference Limits (µg/L) by Gender and Race for CardiacTroponin Assays Cleared by the Food and Drug Administration

n* Abbott Beckman Dade OCD Roche† n‡ Tosoh n§ Bayer n� DPC

All 696 0.8 0.08 0.06 0.10 <0.01 473 0.07 403 0.15 281 0.21

Male 315 0.8 0.10 0.06 0.11 <0.01 223 0.07 187 0.17 115 0.21

Female 381 0.7 0.04 0.06 0.09 <0.01 250 <0.06 216 0.14 166 <0.2

P — .739 .034 .985 .017 .534 — .521 — .441 — .21

Caucasian 400 0.8 0.07 0.04 0.11 <0.01 215 <0.06 193 0.17 166 0.21

African- 218 0.5 0.08¶ 0.03 0.10 <0.01 196 0.17 156 0.17 91 <0.2American

*Number of samples tested in the Abbott, Beckman, Dade-Behring, OCD, and Roche assays.

†The Roche assay is the only Cardiac-Specific troponin T assay on the market; all other assays are for Cardiac-Specific troponin I.

‡Number of samples tested in the Tosoh assay.

§Number of samples tested in the Bayer assay.

�Number of samples tested in the DPC assay.

¶Significantly different (P = .05) from Caucasians based on mean concentrations.

DPC = Diagnostics Products Corporation

OCD = Ortho-Clinical Diagnostics

concentrations to be reported within 20% ofeach other.

As has been recognized for years for total CK,all CK-MB assays demonstrate a significant 1.2-to 2.6-fold higher 99th percentile for males ver-sus females. Several assays showed higher, up to2.7-fold, concentrations for African-Americansversus Caucasians. These data demonstrate thatclinical laboratories must consider establishingdifferent CK-MB reference cutoffs for at leastmen versus women.

CLINICAL UTILIZATION OF CARDIAC BIOMARKERS

Use in Patients with Acute Coronary Syndrome

The ideal marker of myocardial injury should:(1) provide early detection of injury, (2) allowrapid diagnosis of cardiac injury, (3) serve as arisk stratification tool in patients with ACS, (4)assess the success of reperfusion after throm-bolytic therapy, (5) detect reocclusion andreinfarction, (6) determine the timing of an in-farction as well as infarct size, and (7) detectprocedural-related perioperative MI during car-diac or noncardiac surgery.At present, the per-fect biomarker to satisfy all these needs doesnot exist. It is the function of the laboratory toprovide advice to physicians about cardiac bio-marker characteristics.

Patients present to emergency departmentsor other primary care providers with a multi-tude of clinical signs and symptoms for whichthe differential diagnosis of AMI (heart attack)is considered. Figure 5-3 demonstrates the com-plete spectrum of clinical presentations of sucha patient. This entire spectrum of clinical pre-sentations has been designated ACS.The corner-stone of the redefinition of MI is predicated oncardiac biomarkers, specifically cTnI or cTnT.The following are designated as biochemical in-dicators for detecting myocardial necrosis:

1. A maximal concentration of cTnI or cTnIexceeding the decision limit, defined asthe 99th percentile of values for a refer-ence control group, on at least one occa-sion during the first 24 hours after theindex clinical event.

2. A maximal value of CK-MB (preferablymass) exceeding the 99th percentile ofvalues for a reference control group ontwo successive samples or a maximal


value exceeding twice the upper refer-ence limit during the first hours after theindex clinical event.Although the consen-sus document states values for troponinand CK-MB should rise and fall, either arising or a falling pattern should be con-sidered diagnostic.Values that remain ele-vated without change are rarely due to MI.

3. In the absence of availability of a cardiactroponin or CK-MB assay, total CK greaterthan two times the upper reference limitmay be employed.

In addition to the ESC/ACC consensus docu-ment for redefining MI, the ACC/AmericanHeart Association guidelines for managementof unstable angina recommend monitoringcardiac troponin in patients with ACS as a wayof differentiating unstable angina (defined aswhen cardiac troponin is within the 99th per-centile reference limit) and non–ST segment-elevation MI (defined as when cardiac troponinis increased above the 99th percentile refer-ence limit).

Several markers should no longer be used toevaluate cardiac disease, including aspartateaminotransferase, total CK, total lactate dehy-drogenase (LDH), and LDH isoenzymes. Due totheir wide tissue distribution, these markershave poor specificity for the detection of car-diac injury. Because total CK and CK-MB haveserved as standards for so many years, some lab-oratories may continue to measure them to al-low for comparisons to cardiac troponin overtime, before discontinuing use of CK and CK-MB. In addition, the use of total CK in develop-ing countries may be the preferred or onlyalternative for financial reasons. However, itshould be clear that, for monitoring ACS pa-tients to assist in clinical classification, cardiactroponin is the preferred biomarker.

For the majority of patients, blood should beobtained for testing at hospital admission (0hours), at 6 to 9 hours, and again at 12 to 24hours if the earlier specimens are normal andthe clinical index of suspicion is high. For pa-tients in need of an early diagnosis that wouldparallel a rapid triage protocol, a rapidly appear-ing biomarker such as myoglobin has been sug-gested to be added to serial cardiac troponinmonitoring. In practice, it appears that the ma-jority of hospitals throughout the world do notuse these markers.

Several general clinical impressions can bemade regarding cTnI and cTnT. First, the early


release kinetics of both cTnI and cTnT are simi-lar to those of CK-MB after AMI, with increasesabove the upper reference limit seen at 2 to 6hours. The initial increase is due to the 3% to6% cytoplasm fraction of troponin (CK-MB is100% cytoplasmic). Second, cTnI and cTnT canremain increased up to 4 to 14 days after AMI.The mechanism is likely the ongoing release oftroponin from the 94% to 97% myofibril-boundfraction since the half-life in clearance studiesof either the native protein or of complexes isin the range of 2 hours. The long interval of car-diac troponin increase has resulted in its utiliza-tion in place of the LDH isoenzyme assay in thedetection of late-presenting AMI patients. Third,the very low to undetectable cardiac troponinvalues in serum from patients without cardiacdisease (normal, healthy reference population)permits use of lower discriminator concentra-tions compared to CK-MB for the determinationof myocardial injury.Finally, cardiac tissue speci-ficity of cTnI and cTnT should eliminate falseclinical impression of AMI in patients with in-creased CK-MB concentrations following skele-tal muscle injuries.

Clinical use of the percentage relative index[%RI; %RI = (CK-MB mass/Total CK activ-ity × 100)] or %CK-MB (CK-MB activity/TotalCK × 100) aids in the interpretation of CK-MBconcentrations for the detection of AMI. While

not absolute, an increased %CK-MB or %RIpoints to the heart as the source of CK-MB inserum. However, the %RI and %CK-MB shouldnot be used for interpretation when the totalCK activity remains within the reference inter-val.Any concomitant skeletal muscle injury willdecrease the sensitivity of the relative index forthe detection of cardiac events.

As increases in cardiac troponin detect anyform of myocardial injury, nonischemic mecha-nisms of injury are also responsible for cardiactroponin release from the heart, causing in-creases in circulating troponin. Table 5-3 showsa list of potential etiologies that have been re-sponsible for increases in non-ischemic damageto the heart.Thus,whenever cardiac troponin ismonitored, it is important to follow the serialpattern of a rising or a falling pattern of the bio-marker. An increased cTn pattern that remainsrelatively unchanged and is not indicative ofthis serial trend is likely not an MI.

Strategies for the Role of CardiacTroponin for Risk Assessment

Numerous prospective and retrospective clini-cal studies have evaluated and compared theutility of measurements of cTnI and cTnT for riskstratification or clinical outcomes assessmentof patients with ACS with possible myocardial

Figure 5-3. Complete spectrum of acute coronary pathophysiological process from ini-tiation of atherosclerosis to cell death. Biomarkers released at different stages include in-terleukin 6 (IL-6),myeloperoxidase (MPO),soluble CD40 ligand (sCD40L),high-sensitivityC-reactive protein (Hs-CRP), ischemia-modified albumin (IMA), cardiac troponins I and T(cTnT and cTnI, respectively), B-type natriuretic peptide (BNP), and N-terminal proBNP(NT-proBNP). Of these, only the troponins and BMPs are myocardial specific.


ischemia in the emergency department. Patientspresenting with a complaint of chest pain orother symptoms suggesting ACS have been as-signed to blood-sampling protocols includingonly a single draw at presentation as well as toseveral serial blood samplings over a 12- to 24-hour period following presentation. Overall, inthe approximately 18,000 patients studied, at 30days the odds ratio for an adverse outcome was3.4 for increased troponin. As both cTnT andcTnI offer powerful risk assessment, cTn moni-toring needs to be included in current practiceguidelines not only regarding diagnosis andmanagement of ACS patients, but also as usefulrisk stratification tools. It is recommended todraw two samples on patients with ACS who donot rule in for AMI: one at presentation and oneat 6 to 9 hours following presentation. This willallow for an increase in either cardiac troponinto occur above baseline in a patient presentingwith a very recent acute coronary lesion. How-ever, it should be noted that a normal cardiactroponin does not remove all risk.

Several studies have now documented thatassays with lower limits of detection are able to

identify more patients with ACS with poor prog-nosis who may be candidates for early invasiveprocedures. In one representative study (Vengeet al.,2000), two assays were compared to assessclinical performance in unstable patients withCAD. While both assays showed patients withnormal cTnI levels had a significantly betterprognosis than patients with increased levels, acohort of 11% of patients (n = 98) with a poorprognosis was identified only by the second-generation assay with a lower limit of detection.Invasive treatment only reduced clinical eventsin the group of patients with increased cTnI.Thus, each troponin assay, I or T, needs to evalu-ate the stratification of patients at low-end con-centrations to avoid the potential of analyticalinaccuracies leading to inappropriate manage-ment decisions and therapy. These high-risk pa-tients have been shown to benefit fromaggressive therapies, including low molecularweight heparin, IIb/IIIa glycoprotein platelet in-hibitors, and an early interventional strategy.

Clinicians are often confronted with a clinicalhistory of a patient without overt CAD and a lowprobability of myocardial ischemia.However,as a

Table 5-3. Elevations of Troponins without Overt Ischemic Heart DiseaseTrauma (including contusion, ablation, pacing, cardioversion)

Congestive heart failure, acute and chronic*

Aortic valve disease and hypertrophic obstructive cardiomyopathy with significant left ventricular hypertrophy*

Hypertension

Hypotension, often with arrhythmias

Postoperative noncardiac surgery patients who seem to do well*

Renal failure*

Critically ill patients, especially with diabetes, respiratory failure*

Drug toxicity, such as adriamycin, 5-fluorouracil, herceptin, snake venoms*

Hypothyroidism

Coronary vasospasm, including apical ballooning syndrome

Inflammatory diseases such as myocarditis (e.g., with parvovirus B19, Kawasaki disease, sarcoidosis, smallpoxvaccination, or myocardial extension of bacterial endocarditis)

Postpercutaneous coronary intervention patients who appear without complication*

Pulmonary embolism, severe pulmonary hypertension*

Sepsis*

Burns, especially if total burn surface area >30%*

Infiltrative diseases, including amyloidosis, hemochromatosis, sarcoidosis, and scleroderma*

Acute neurological disease, including cerebrovascular accident, subarachnoid bleeds*

Rhabdomyolysis with cardiac injury

Transplant vasculopathy

Vital exhaustion

*Designations imply prognostic information has been reported.

precautionary reflex, serial cardiac biomarkers,specifically cTn, are ordered. The 20% of sus-pected ACS patients who clinically do not rulein for MI but display an increased cTn representthose nonischemic pathologies such as my-ocarditis,blunt chest trauma,or chemotherapeu-tic agents for which the mechanisms of injuryare well defined (Table 5-3) as well as the unex-pected finding of myocardial injury, for whichpatients have been shown to have increasedcTn, but the mechanism of release is not clear.These observations have led to important andnovel investigations involving patients with non-ischemic heart disease and the role for cardiactroponin as a diagnostic and prognostic tool.The conditions shown in Table 5-3 that are indi-cated by an asterisk demonstrate that, in addi-tion to cardiac troponin being indicative ofcardiac injury, the data have also indicated thatincreased cardiac troponin is useful as a prog-nostic tool for assessing risk of death and MI.

Monitoring Reperfusion FollowingThrombolytic Therapy

Release of cTnI or cTnT from myocardium intothe blood following AMI and after the washoutthat accompanies successful reperfusion gener-ates an excellent signal compared to no de-tectable baseline levels prior to myocardialdamage.The initial rapid release of cardiac tro-ponin subunits I and T following successfulreperfusion is most likely derived from the solu-ble cytosolic myocardial fraction (6% cTnT, 3%cTnI).The clinical utility of cardiac biomarkersfor monitoring reperfusion following throm-bolytic therapy has not gained favor as a rou-tine form of testing for determining the successor failure of reperfusion therapy because it can-not distinguish TIMI 2 from TIMI 3 flow, whichis a critical issue in regard to prognosis. (TIMI isthe timed intervention in myocardial infarction.TIMI 2 and 3 refer to the extent of flow throughcoronary vessels, with TIMI 2 referring to par-tial flow and TIMI 3 to complete flow.)

It is accepted that the kinetics of myocardialprotein appearance in the circulation dependson infarct perfusion. Early reperfusion causesan earlier increase above the upper referencelimit and an earlier and greater enzyme peak af-ter reperfusion. However, once the peak has oc-curred, there is no difference in the time ofclearance of enzymes. In addition, enhancedwashout identifies whether an artery is patentor closed but cannot distinguish between nor-

mal and abnormal coronary perfusion, which isanother key prognostic parameter. Further, itis difficult to assess the amount of irreversiblemyocardial injury by infarct sizing because ofthe large variability in the amount of enzymewashout that appears after reperfusion.

Strategies Using Multimarkers

There is a growing body of evidence suggestingthat different cardiac biomarkers provide inde-pendent and complementary information aboutpathophysiology, diagnostics, prognostics, andresponse to therapy in patients with ACS. Thus,it is probable that multimarker strategies or bio-chemical profiling may be used to characterizeindividual patients presenting with ACS. For ex-ample, in one multicenter study, cTnI, CK-MB,and myoglobin multimarker analysis identifiedpositive patients earlier and provided a betterrisk stratification for 30-day mortality than cen-tral laboratory analysis of CK-MB alone (20% vs.3%, respectively).

Estimation of Infarct Size

Older studies have shown that one can use theintegrated values for total or CK-MB to estimatethe biochemical extent of infarction. Furtherstudies verified that the amount of cardiac dam-age is the primary determinant of prognosis.Such determinations have been correlated withmorphological infarct size. Some have used thepeak CK-MB value as a surrogate for the inte-grated data. Reperfusion changes the release ra-tio (the percentage of marker that appears inthe blood relative to the amount depleted frommyocardium),making infarct sizing problematicin the modern era.Additional data from both ex-perimental and patient-related data have sug-gested that the 72-hour troponin measurementcorrelates with scintigraphically-determined in-fract size. The data are stronger for troponin Tthan for I, although the principles are probablysimilar for both analytes.At present, it is not rec-ommended that serial monitoring of cardiactroponin or CK-MB be used for infarct sizing.

Reinfarction

Figure 5-1 demonstrates the serial patterns forcardiac troponin I and CK-MB both during theinitial infarct and reinfarction, as describedfor the initial case presentation. Although thenumber of reinfarction cases presented in the


myocardial reinfarction. Clin Chem 51:460–463,2005.

Apple FS, Quist HE, Doyle PJ, et al.: Plasma 99thpercentile reference limits for cardiac troponinand creatine kinase MB mass for use with Euro-pean Society of Cardiology/American College ofCardiology consensus recommendations. ClinChem 49:1331–1336, 2003.

Apple FS, Wu AHB, Jaffe AS: European Society ofCardiology and American College of Cardiologyguidelines for redefinition of myocardial infarc-tion: how to use existing assays clinically andfor clinical trials. Am Heart J 144:981–986,2002.

Bodor GS, Porter S, Landt Y, et al.: Development ofmonoclonal antibodies for an assay of cardiactroponin-I and preliminary results in suspectedcases of myocardial infarction. Clin Chem38:2203–2214, 1992.

Fishbein MC,Wang T, Matijasevic M, et al.: Myocar-dial tissue troponins T and I: an immunohisto-chemical study in experimental models ofmyocardial ischemia. Cardiovasc Pathol 12:65–71, 2003.

Ingwall JS, Kramer MF, Fifer MA, et al.:The creatinekinase system in normal and diseased human my-ocardium.N Engl J Med 313:1050–1054,1985.

Jaffe AS,Landt Y,Parvin CA,et al.:Comparative sensi-tivity of cardiac troponin I and lactate dehydroge-nase isoenzymes for diagnosing acute myocardialinfarction.Clin Chem 42:1770–1776,1996.

Jaffe AS, Ravkilde J, Roberts R, et al.: It’s time for achange to a troponin standard. Circulation102:1216–1220, 2000.

Katrukha AG, Bereznikova AV, Filtaov VL, et al.:Degradation of cardiac troponin I: implicationfor reliable immunodetection. Clin Chem 44:2433–2440, 1998

Katus HA, Looser S, Hallermayer K, et al.: Develop-ment and in vitro characterization of a new im-munoassay of cardiac troponin T. Clin Chem38:386–393, 1992

Lin JC,Apple FS, Murakami MM, et al.: Rates of pos-itive cardiac troponin I and creatine kinase MBamong patients hospitalized for suspected acutecoronary syndromes. Clin Chem 50:333–338,2004.

Ottani F, Galvani M, Nicolini A, et al.: Elevated car-diac troponin levels predict the risk of adverseoutcome in patients with acute coronary syn-dromes. Am Heart J 40:917–927, 2000.

Panteghini M, Gerhardt W,Apple FS, et al.: Qualityspecifications for cardiac troponin assays. ClinChem Lab Med 39:174–178, 2001.

Venge P, Lagerquist B, Diderholm E, et al. on behalfof the FRISC II Study Group: Clinical perfor-mance of three cardiac troponin assays in pa-tients with unstable coronary artery disease (aFRISC II Substudy).Am J Cardiol 89:1035–1041,2000.


literature is fewer than 20, the findings demon-strate that CK-MB analysis is no longer clinicallyrelevant, or cost-effective, in the differentialdiagnosis of myocardial reinfarction in patientswith ACS when cardiac troponin monitoring isavailable. We encourage others to test this hy-pothesis to be able to dispel the theoretical ra-tionale for use of CK-MB testing in addition tocTn testing, expediting the cost-effective adap-tation favoring only cTn monitoring in testingfor MI or reinfarction.

QUESTIONS

1. Diagram the rising and falling serial bio-marker profiles for cardiac troponins com-pared to CK-MB.

2. Define the ESC/ACC consensus guidelinesfor the redefinition of AMI.

3. Review the role of cardiac biomarkers forinfarct sizing and monitoring the successof reperfusion following therapy.

4. List the multiple forms of cardiac tro-ponin I that may exist in the circulation.

5. Define the two major concerns that havebeen responsible for substantial concen-tration differences when measuring car-diac troponin I by different commercialimmunoassays.

6. Explain how the determination of refer-ence intervals can have an impact on thenumber of MI cases that are defined in agiven population of patients with chestpain.

7. Explain the role of cardiac troponin deter-minations in risk-stratifying patients whopresent with ACS.

BIBLIOGRAPHY

Alpert JS,Thygesen K,Antman E, et al.: Myocardialinfarction redefined—a consensus document ofthe Joint European Society of Cardiology/Ameri-can College of Cardiology Committee for the re-definition of myocardial infarction. J Am CollCardiol 36:959–959, 2000.

Apple FS: Tissue specificity of cardiac troponin I,cardiac troponin T, and creatine kinase MB. ClinChim Acta 284:151–159, 1999.

Apple FS, Falahati A, Paulsen PR, et al.: Improveddetection of minor ischemic myocardial injurywith measurement of serum cardiac troponin I.Clin Chem 43:2047–2051, 1997.

Apple FS, Murakami MM: Cardiac troponin and cre-atine kinase MB monitoring during in-hospital

CASE REPORT

A 25-year-old Japanese man was admitted to thehospital because of fever, sore throat, and gen-eral malaise. He had been febrile for severaldays and had noticed that the color of his urinewas unusually dark. On admission, his tempera-ture was 37.8°C. The physician noted anemicconjunctiva (delicate membrane lining theeyelids and covering the eyeballs), injectedpharyngeal mucosa (mucous membrane), cer-vical lymphadenopathy (swelling of the lymphnodes), and mild splenomegaly (enlargementof the spleen).

Hematological data showed moderate anemia(hemoglobin 95 g/L; normal is 136–172 g/L), in-creased reticulocytes (7.8%; normal is <2.4%),and leukocytosis (leukocytes 12 × 109/L; nor-mal is 3.2–9.8 × 109/L) with increased lympho-cytes (62% of total leukocytes).The blood filmshowed moderate anisocytosis (variability ofsize of erythrocytes), increased spherocytes(small, globular erythrocytes without the usualcentral pallor), and a scatter of atypical lympho-cytes.

The urine examination revealed no abnor-mality except an increase of urobilinogen. Theerythrocyte sedimentation rate was increased(45 mm per hour; normal is <20 mm per hour).

Biochemical data for his serum were as fol-lows: total bilirubin 99.2 µmol/L (normal is 3.4–20.5 µmol/L); conjugated bilirubin 15.4 µmol/L(normal is 0–6.8 µmol/L); alanine aminotrans-ferase 125 U/L (normal is 0–40 U/L); and lac-tate dehydrogenase 570 U/L (normal is 120–250 U/L).

The serological examinations for Epstein-

Barr virus confirmed that he had infectiousmononucleosis. He was kept in bed for 2 weeks,by which time the signs of inflammation and he-patocellular damage entirely disappeared. How-ever, mild anemia (hemoglobin 110–120 g/L)with reticulocytosis (5.0%–6.0%), increasedserum unconjugated bilirubin,and splenomegalystill remained, suggesting the presence of per-sistent hemolysis. The physician therefore per-formed further examinations to confirm thediagnosis of hemolytic anemia.

Bone marrow aspirates from the sternumshowed marked erythroid hyperplasia, with agranuloid to erythroid ratio of 1:1.5. Theantiglobulin (Coombs) test failed to detect au-toantibodies to red blood cells. The sugar-watertest and Ham test, the diagnostic tests for parox-ysmal nocturnal hemoglobinuria, were nega-tive. The levels of both hemoglobin F andhemoglobin A2 were normal, and hemoglobinelectrophoresis showed no abnormally migrat-ing hemoglobins. The isopropanol test did notreveal any unstable hemoglobins. In contrast,the osmotic fragility of the patient’s erythrocyteswas increased, and it was enhanced by incubat-ing the cells at 37°C for 24 hours. These resultswere consistent with a diagnosis of hereditaryspherocytosis (HS).

Abdominal ultrasound revealed moderatesplenomegaly and gallstones. The patient hadan episode of transient jaundice of unknownetiology at the age of 22 years. His parents hadbeen in apparent good health. However, his fa-ther had also experienced repeated episodes ofmild jaundice since childhood. His father’s ery-throcytes were also spherical and osmoticallyfragile.

6

Hereditary Spherocytosis

HIROSHI IDEGUCHI

66

From these findings, he was finally diagnosedas having (1) HS with gallstones and (2) infec-tious mononucleosis. The patient underwentsplenectomy and cholecystectomy at the age of26 years. An enlarged spleen weighing 530 gand a gallbladder containing numerous smallstones were removed. Two months after the op-eration, all previous findings reflecting acceler-ated hemolysis disappeared.

DIAGNOSIS

The most important diagnostic features of HSare as follows: (1) congenital hemolytic anemia;(2) microspherocytosis on the peripheral bloodfilm; (3) increased osmotic fragility, particularlyin incubated red cells; and (4) negative antiglob-ulin (Coombs) test.

A typical HS patient is relatively asympto-matic, and mild jaundice may be the only symp-tom of the disease. Anemia is usually mild or

even absent owing to compensatory erythroidhyperplasia in bone marrow. Spherocytosis, ormore correctly microspherocytosis, is a char-acteristic feature of the stained blood films(Fig. 6-1). The microspherocytes are small cells,usually with perfectly round contours, that stainrelatively densely with May-Grunwald-Giemsastain. The reticulocyte count is usually in-creased (5%–20%) and remains high through-out the patient’s life unless splenectomy hasbeen carried out.

The osmotic fragility test is the most sensi-tive and useful test for the diagnosis of HS (Fig.6-2).Although the fresh red cells of about 20%of HS patients have a normal or near-normal os-motic fragility, the test performed in cells incu-bated at 37°C for 24 hours is more oftenpositive in association with an increased rate ofspontaneous hemolysis (autohemolysis).

Patients with HS are occasionally subject to“crises,” a sudden exaggeration of their anemia.Two types of crises, hemolytic and aplastic, are

Hereditary Spherocytosis 67

Figure 6-1. Photomicrographs of blood films of a normal subject (a) and a patient withHS (b). Small, densely staining microspherocytes can be seen in (b).

known.As in the patient described in the casereport, hemolytic crises usually occur withcommon viral syndromes and are characterizedby a manifestation of accelerated destruction ofred cells, such as a transient increase in jaun-dice, anemia, reticulocytosis, and splenomegaly.On the other hand, aplastic crises are broughtabout by a temporary failure of erythropoiesis,mostly caused by human parvovirus B19 infec-tion (erythema infectiosum). During the aplasticphase, the counts of red cells and reticulocytesrapidly fall, and marrow erythroblasts disappear.

Aplastic crises are less frequent but are usuallymore serious, and severe life-threatening ane-mia can result.

Gallstones are a common complication ofHS, and most patients eventually develop them;they are rarely found in children, and their inci-dence increases with age. Since gallstone colic(severe convulsive abdominal pain) and chole-cystitis (inflammation of the gall bladder) arequite common, it is desirable that patients withHS should be periodically examined by abdomi-nal ultrasound.


Figure 6-2. Osmotic fragility test (Parpart method). A typical osmotic fragility curve offresh (left) or 24-hour incubated (right) erythrocytes from a patient with HS is shown.The hatched area represents the normal range.

Figure 6-3. Structure of the red cell membrane (a) and the hexagonal lattice structureof the membrane skeleton (b).


History

HS was first described in 1871 by Vanlair andMasius (Vanlair and Masius, 1871). They re-ported a young woman who had repeated at-tacks of abdominal pain and jaundice and foundthat some of her erythrocytes were sphericaland much smaller than normal. In 1907, Chauf-fard (Chauffard, 1907) demonstrated increasedosmotic fragility of erythrocytes as the hallmarkof the disease.A membrane lesion was first sug-gested by the observation of Bertles in 1957(Bertles, 1957) that HS red cells are unusuallypermeable to sodium ions. Since then, many ab-normalities have been reported in HS red cells,but it is now clear that HS is a consequence ofheterogeneous defects in the red cell mem-brane proteins.

Red Cell Membrane Organization

The red cell membrane has been well charac-terized biochemically, and the topological or-ganization of various lipids and proteins hasbeen delineated (Fig. 6-3a). The major lipidcomponents are phospholipids and choles-terol, and their composition is responsible forthe fluidity of the membrane matrix in whichtransmembrane proteins reside. The mem-brane proteins are conventionally separated inthe laboratory by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Fig. 6-4;Table 6-1).They are classified into two groups,integral and peripheral proteins. The integralproteins, such as band 3 and glycophorin C(Fig. 6-3), penetrate into or are embedded inthe lipid bilayer and are tightly bound to themembrane lipids through hydrophobic inter-actions. In contrast, the peripheral proteinssuch as spectrin are located on the cytoplas-mic surface of the lipid bilayer and associatedwith each other to form a flexible, filamentousnetwork, generally referred to as the mem-brane skeleton; they are responsible for redcell shape and deformability.

The major components of the membraneskeleton are spectrin, actin, and protein 4.1.Spectrin is a highly flexible, rodlike moleculecomposed of two nonidentical polypeptides: α-spectrin and β-spectrin.These chains are alignedside by side in the form of a αβ-heterodimer,andspectrin heterodimers in turn join head to headto form (αβ)2-tetramers. The tail ends of spec-

trin tetramers are associated with short actinfilaments. Although the spectrin–actin interac-tion is itself weak, each spectrin–actin junctionis greatly stabilized by the formation of a ternarycomplex with the protein 4.1.Thus, six spectrintermini complex with each actin oligomer toform a network with an approximately hexago-nal lattice (Fig. 6-3b).

The major integral proteins include band 3(anion exchanger 1) and the glycophorins (Fig.6-3). These proteins span the membrane andhave distinct structural and functional do-mains. The two clearly established functions ofband 3 are: the (1) mediation of the Cl−–HCO3

−

exchange across the membrane, which is es-sential for the transport of CO2 from the tis-sues to the lungs; and (2) anchorage of theunderlying membrane skeleton to the mem-brane by binding to ankyrin (band 2.1) andpossibly protein 4.2.


Figure 6-4. Sodium dodecyl sulfate polyacry-lamide gel electrophoresis pattern of normal redcell membrane proteins with Coomassie bluestaining.

The four sialic acid-rich glycoproteins (gly-cophorins A, B, C, and D) also belong to theclass of integral proteins. The presence of sialicacids imparts a strong net negative charge tothe erythrocyte surface; this is functionally im-portant in reducing the interaction of red cellswith one another as well as with other bloodcells or the vascular endothelium.

Molecular Lesions in HereditarySpherocytosis Erythrocytes

As knowledge of the membrane skeleton andits contribution to red cell shape and stabilityemerged during the 1970s, investigators soughtevidence of a skeletal protein defect in HS.The search was given impetus by the finding ofa marked spectrin deficiency in the commonhouse mouse. The red cells of these mice werespherocytic and osmotically fragile, and theyspontaneously vesiculated in the circulation.Thus, investigators have focused on spectrin asthe possible primary lesion of HS.

In 1982, Agre and coworkers (Agre et al.,1982) first described two patients with severeHS inherited autosomal recessively and whosered cells had only about half of the normalspectrin content. However, in the majority ofHS patients, spectrin deficiency may not repre-sent the primary molecular defect since recentcytogenetic studies and subsequent biochemi-cal studies have led to the identification of dis-

tinct molecular defects in other membrane pro-teins (Table 6-2).

It is now clear that a deficiency or dysfunc-tion of ankyrin, the protein that anchors thespectrin-based skeleton to band 3 (Fig. 6-3),may represent a common membrane lesion inHS. The synthesis of ankyrin and its assemblyon the membrane were reduced, and as a con-sequence, assembly of spectrin was also im-paired despite normal spectrin synthesis. In1996, Jarolim et al. showed that a deficiency ofankyrin (and its related spectrins) was presentin 60% of 166 unrelated patients with HS.

Partial deficiency of band 3 was reported inabout 25% of patients with HS who presentedwith a phenotype of a mild-to-moderate domi-nantly inherited HS. Nearly 50 different band 3mutations associated with HS have been re-ported.

A deficiency of protein 4.2 has also beennoted in several patients with recessively in-herited HS, particularly in the Japanese popula-tion.Although the physiological role of protein4.2 remains to be established, this protein isthought to stabilize the binding of band 3 tospectrin mediated by ankyrin.

Because of the heterogeneous molecularnature of HS (Table 6-2), the “HS genes” can beassigned to several chromosomes: chromosome1 (α-spectrin), chromosome 8 (ankyrin), chro-mosome 14 (β-spectrin), chromosome 15 (pro-tein 4.2), and chromosome 17 (band 3). The


Table 6-1. Major Red Cell Membrane ProteinsMolecular Peripheral (P)

Band Designation Weight or Integral (I) Function

Band 1 Spectrin-α 240 K P Membrane skeleton

Band 2 Spectrin-β 220 K P Membrane skeleton

Band 2.1 Ankyrin 210 K P Anchoring skeleton to bilayer

Band 3 Anion exchanger 95 K I Anion transport; binding sites for ankyrin, protein 4.2

Band 4.1 Protein 4.1 80 K P Membrane skeleton; association with GPC

Band 4.2 Protein 4.2 72 K P Stabilizing ankyrin–band 3 interaction (?)

Band 5 Actin 43 K P Membrane skeleton

Band 6 G3PD 35 K P Glyceraldehyde 3-phosphate dehydrogenase

Band 7 Stomatin 31 K I Regulating monovalent cation transport (?)

GPA Glycophorin A 36 K IMNSs blood group*

GPB Glycophorin B 20 K I

GPC Glycophorin C 32 K IGerbich blood group; association with protein 4.1

GPD Glycophorin D 23 K I

*MN and Ss are antigens on human red blood cells. Each of the antigens M,N,S, and s are identifiable by reaction with specific anti-sera: anti-M, anti-N, anti-S, and anti-s. These glycophorin A and glycophorin B antigens are polypeptides are coded by genes onchromosome 4.

Table 6-2. Molecular Basis of Hereditary SpherocytosisResponsible Chromosome Phenotype and Clinical Gene Locus Molecular Lesion Inheritance Expression Prevalence

Spectrin-α 1 Spectrin deficiency AR Severe Rare

Spectrin-β 14 Defective binding to protein 4.1 AD Mild to moderate ∼10%Spectrin Kissimmee (202Trp→Arg)Others

Ankyrin 8 Ankyrin and spectrin deficiency AD or AR Mild to severe 50%–60%Ankyrin gene deletionPromoter region mutationsAnkyrin Stuttgart (329: 2-nt deletion)Ankyrin Einbeck (572: 1-nt Insertion)Ankyrin Marburg (797/798: 4-nt deletion)Ankyrin Bovenden (1436Arg→ter)Others

Band 3 17 Band 3 deficiency AD Mild to moderate 20%–30%Band 3 Boston (285Ala→Asp)Band 3 Coimbra (488Val→Met)Band 3 Kyoto (490Arg→Cys)

(Bicêtre I)Band 3 Prague II (760Arg→Gln)Band 3 Birmingham (834His→Pro)Band 3 Philadelphia (837Thr→Met)Others

Protein 4.2 15 Protein 4.2 deficiency AR Mild to moderate Rare; more commonProtein 4.2 Nippon (142Ala→Thr) in JapanProtein 4.2 Fukuoka (119Trp→ter)Protein 4.2 Tozeur (310Arg→Gln)Protein 4.2 Shiga (317Arg→Cys)Others

AD, autosomal dominant; AR, autosomal recessive.

inheritance is both autosomal dominant (in themajority of patients with HS) and autosomal re-cessive.

Several rapid and sensitive methods havebeen developed to detect mutations in cDNAand genomic DNA. These include ribonuclease(RNase) protection analysis, denaturing gra-dient gel electrophoresis, and single-strandconformation polymorphism analysis. Thesemethods can be used in conjunction with thePCR technique to detect small deletions or in-sertions and single base substitutions.

RNase protection is used to detect the RNase-sensitive site in a synthetic RNA probe that hasbeen hybridized with the DNA fragment to beexamined. If a particular mutation lies withinthe DNA fragment, then the RNA probe is mis-matched at the site of the mutation whereRNase digestion occurs.

Denaturing gradient gel electrophoresis isused to detect the change in mobility of DNAfragments run on a polyacrylamide gel formedwith an increasing gradient of the denaturant(urea, formamide). It is based on the principlethat different DNA fragments have differentpoints in the gradient where they begin to un-dergo regional denaturation, and the rate of mi-gration is markedly slowed ( Fig. 6-5a).

Single-strand conformation polymorphismanalysis is the more popular method becauseof its simplicity and wide applicability to manydifferent genes. This method makes use of

sequence-dependent folding of single-strandedDNA, which alters the electrophoretic mobilityof the fragments, to detect sequence differencesbetween closely related molecules. In brief,regions of DNA thought to contain mutationsare amplified using the PCR technique and ren-dered single stranded by heating in a denaturingbuffer. The separated strands are then fraction-ated on polyacrylamide gels under conditionsthat may resolve two molecules differing by asingle base (Fig. 6-5b). If by one of these meth-ods the DNA fragment appears to be abnormal,then the specific change can be determined bysubsequent nucleotide sequence analysis.

Pathophysiology of Hemolysis in Hereditary SpherocytosisErythrocytes

The importance of the spleen in the pathophys-iology of the hemolysis of HS has been substan-tiated. Two factors determine the selectivedestruction of the HS cells in the spleen: (1)poor HS red cell deformability,which is a reflec-tion of a decreased surface-to-volume ratio re-sulting from the loss of membrane; and (2) theunique anatomy of the splenic vasculature,which acts as a “microcirculation filter.” Asshown in Table 6-2, the underlying molecularbasis of HS is heterogeneous, and the primarymolecular lesion in HS is likely to involve severalmembrane proteins, including spectrin, ankyrin,


Figure 6-5. Rapid and sensitive methods for detecting mutations in DNA. (a) Denatur-ing gradient gel electrophoresis. (b) Single-strand conformation polymorphism.

band 3,and protein 4.2. These molecular lesionsmay finally lead to lipid bilayer destabilizationand microvesiculation, resulting in surface areadeficiency and formation of poorly deformablespherocytes (Fig. 6-6).

The principal sites of red cell entrapment inthe spleen are openings in the endothelial wallof splenic sinuses, where blood enters the ve-nous circulation. In contrast to normal disco-cytes, which have an abundant surface thatallows red cells to deform and pass through nar-row slits, the HS red cells lack the extra surfaceneeded to permit deformation. Consequently,the nondeformable spherocytes accumulate inthe red pulp and become grossly engorged.Once entrapped by the spleen, HS red cells in-cur additional damage. Because of the stagnantcirculation, lactic acid accumulates, and the ex-tracellular pH falls, probably to between 6.5 and7.0. The intracellular pH also declines, inhibitinghexokinase and phosphofructokinase, the rate-limiting enzymes of glycolysis and thereby re-tarding glucose utilization and ATP production.

Contact of red cells with macrophages mayinflict additional damage on the red cell mem-brane.Thus, the spherocytes in the splenic cordsare severely stressed by detaining in a metaboli-cally threatening environment and are subjectto further loss of surface area and an increase in

the density of the cells. This process is knownas splenic conditioning (Fig. 6-6). The condi-tioning effect of the spleen is likely to representa cumulative injury.Conditioned red cells appearas microspherocytes in the peripheral circula-tion and are particularly susceptible to recaptureand destruction in the spleen and other parts ofthe reticuloendothelial system.

THERAPY

Patients with mild cases of HS often do not needany treatment. However, these patients shouldbe watched carefully for the development ofhemolytic or aplastic crisis. Splenectomy is thetreatment of choice in moderate-to-severe HScases. In general, splenectomy is indicated inpatients who are continuously anemic or whohave a history of gallstone colic or repeatedcrises. The clinical results of splenectomy forHS are almost uniformly excellent. However,splenectomy in very young children shouldbe postponed to later in childhood becausesplenectomized infants are more susceptible toserious and potentially lethal infections than areolder children and adults.At the time of splenec-tomy, it is important to identify and remove anyaccessory spleen; otherwise, the operation will


Figure 6-6. Pathophysiology of hemolysis of HS red cells.

result in an unfavorable outcome. Patients whoexperience an episode of gallstone colic orcholecystitis may have to undergo simultaneouscholecystectomy.

Within days following splenectomy, jaundicefades, the hemoglobin concentration rises,and red cell survival usually returns to nearnormal. Although the number of peripheralblood microspherocytes remains unchanged,the morphological features of accelerated ery-thropoiesis are not observed. Blood transfusionis rarely indicated except during aplastic crisis.At such times, red cell replacement may be life-saving.

QUESTIONS

1. What are the essential findings for theclinical diagnosis of HS?

2. Patients with HS may occasionally un-dergo two types of crises. Describe themechanisms causing these crises.

3. Explain why HS red cells are hemolyzedin vivo principally in spleen.

4. Describe the structural organization ofthe membrane skeleton that is a major de-terminant of red cell shape and deforma-bility.

5. Hereditary spherocytosis is a group of dis-orders caused by heterogeneous intrinsicdefects of the red cell membrane pro-teins. Describe the rapid and sensitivemethods to detect mutations in genomicDNA and complementary DNA.

BIBLIOGRAPHY

Agre P, Orringer EP, Bennett V: Deficient red-cellspectrin in severe recessively inherited sphero-cytosis. N Engl J Med 306:1155–1161, 1982.

Becker PS, Lux SE: Hereditary spherocytosis andhereditary elliptocytosis, in Scriver CR, BeaudetAL, Sly WS, Valle D (eds): The Metabolic andMolecular Basis of Inherited Disease. 7th ed.,Vol. 3. McGraw-Hill, New York, 1995, pp.3513–3560.

Bertles JF: Sodium transport across the surfacemembrane of red cells in hereditary spherocyto-sis. J Clin Invest 36:816–824, 1957.

Chauffard A: Pathogéne de l’ictére congenital del’adulte. Semin méd (Paris) 27:25–29, 1907.

Dacie J: Hereditary spherocytosis (HS), in Dacie J(ed): The Haemolytic Anaemias. 3rd ed.,Vol. 1.Churchill Livingstone, Edinburg, Scotland, 1985,pp. 134–215.

Gallagher PG, Jarolim P: Red cell membrane disor-ders, in Hoffman R, Benz EJ, Shattil SJ, et al.(eds): Hematology Basic Principles and Prac-tice. 3rd ed. Churchill Livingstone, 2000, pp.576–610.

Hassoun H, Palek J: Hereditary spherocytosis: a re-view of the clinical and molecular aspects ofthe disease. Blood Rev 10:129–147, 1996.

Jarolim P, Murray JL, Rubin HL, et al.: Characteriza-tion of 13 novel band 3 gene defects in heredi-tary spherocytosis with band 3 deficiency.Blood 88:4366–4374, 1996.

Palek J, Sahr KE: Mutations of the red cell mem-brane proteins: from clinical evaluation to de-tection of the underlying genetic defect. Blood80:308–330, 1992.

Vanlair CF, Masius JB: De la microcythémie. BullAcad R Méd Belg 5, 3rd series:515–613, 1871.


Part II

Fuel Metabolism and Energetics

CASE REPORT

Identical twins with the fictitious names Annand Elizabeth were born 6 weeks prematurelyto a healthy 26-year-old mother whose preg-nancy was uneventful. The family history wasnoteworthy for epilepsy in a paternal aunt,congenital deafness in another paternal aunt,and possible mental retardation in a maternalcousin. The twins’ sister was a healthy 3-year-old girl with the same biological father. Thebirth weight was 3 pounds 9 ounces for Annand 3 pounds 12 ounces for Elizabeth. Apgarscores (numerical expressions of the conditionof a newborn infant; based on assessment ofheart rate, respiratory effort, muscle tone, reflexirritability, and color) 1 and 5 minutes afterbirth were 8 and 9 for both girls (maximumscore is 10). Other than prematurity, neitherchild evidenced physical or laboratory abnor-malities, and they were discharged home within2 weeks (for Ann) and 3 weeks (for Elizabeth),both apparently in good health. The mothersubsequently remarried and gave birth to a son,who has remained well through age 6 years.

During the 3 months following their birth,both sisters sustained recurrent upper respira-tory illnesses. These were more severe in Eliza-beth, who required frequent hospitalizations.Physical examination of both infants during thisperiod disclosed poor feeding habits, failure tothrive, floppiness, microcephaly (small headsize), and delayed developmental milestones.They were below the fifth percentile for weightand height. These findings were more pro-

nounced in Elizabeth, who also demonstratedocular hypertelorism (abnormal increase inintraorbital distance) and pseudoathetosis (invol-untary movements) of the lower face and hands.Magnetic resonance imaging (MRI) of Eliza-beth’s head revealed mild frontal atrophy and in-creased T2 signals in the periventricular andsubcortical white matter. These signals are in-dicative of scarring or abnormalities in myelina-tion.A pediatric neurologist diagnosed probablecerebral palsy in Elizabeth at age 1.5 years.

Formal neurobehavioral evaluation of bothchildren at age 3 years 10 months generatedthe results summarized in Table 7-1. Furthertesting of Elizabeth showed profound hypoto-nia (decreased muscle tone), electrophysiologi-cal evidence of peripheral neuropathy, normalsomatosensory evoked potentials, and a normalelectroencephalogram recorded during sleep.

At age 2 years, the twins were placed on adiet comprised of approximately 60% total calo-ries as (mainly saturated) long-chain fatty acids(see “Therapy” section). They tolerated this “ke-togenic diet” well, gained weight, and showedmild psychomotor improvement.At age 3 years9 months, they enrolled in a controlled clinicaltrial of oral dichloroacetate (DCA), an activatorof the pyruvate dehydrogenase complex (PDC).Venous blood concentrations of lactate, ob-tained initially after an overnight fast and then2 hours after a liquid meal containing 40%of calories as carbohydrate, were 1.2 mmol/Land 2.8 mmol/L, respectively, in Ann and1.6 mmol/L and 5.7 mmol/L, respectively, inElizabeth (normal venous blood lactate after

7

Pyruvate Dehydrogenase Complex Deficiency

PETER W. STACPOOLE and LESA R. GILBERT

77

fasting 0.4–1.0 mmol/L). The cerebrospinalfluid lactate concentration, obtained prior tocarbohydrate ingestion, was approximately5.6 mmol/L (normal approximately 1 mmol/L)in both children.

DIAGNOSIS

The twins were referred subsequently to ametabolic specialist because of the suspicionof an inborn error of metabolism. Biochemicaltesting revealed each had a hyperchloremic(increased blood chloride concentration)metabolic acidosis that was more profound inElizabeth. Serum levels of glucose and livertransaminases were normal. Urinary organicacids revealed modestly increased concentra-tions of lactate and ketone bodies. Blood sam-ples and fibroblasts from skin biopsies fromboth girls were sent to an established diagnos-tic laboratory for genetic mitochondrial dis-eases. Tests of respiratory chain complexenzymatic activities were normal.

Results of the measurement of PDC enzymeactivity are summarized in Table 7-2. These datarevealed severely reduced PDC activities inboth freshly isolated peripheral blood lympho-cytes and cultured fibroblasts in Ann and aneven more striking reduction in enzyme activi-ties in cells from Elizabeth. In each case, the de-fect in PDC activity could be traced to

a deficiency in the α-subunit of the first (E1)enzyme (pyruvate dehydrogenase; pyruvate de-carboxylase) of the complex.


Pyruvate Dehydrogenase Complex

PDC is a multienzyme complex located in theinner mitochondrial membrane. Under aerobicconditions, PDC catalyzes the rate-determiningreaction for the oxidative removal of glucose,pyruvate, and lactate and helps sustain the tri-carboxylic acid cycle by providing acetyl-CoA(Fig. 7-1). Reducing equivalents, in the form ofNADH and FADH2, are generated by reactionscatalyzed by the PDC and by various dehydro-genases in the tricarboxylic acid cycle and pro-vide electrons to the respiratory chain foreventual reduction of molecular oxygen to wa-ter and synthesis of ATP.

The PDCs of eukaryotes are the largest(Mr ∼ 107) multienzyme complexes known.They are distinguished not only by size,but alsoby their morphology and unusual structural bi-ology features. The complex is entirely nuclearencoded and consists of three basic functionaltypes of catalytic protein (Fig. 7-2), a propertyshared with the other two mammalian α-ketoacid dehydrogenases that use α-ketoglutarateor branched-chain ketoacids as substrates, re-spectively. Pyruvate dehydrogenase (E1; EC1.2.4.1) is a heterotetrameric (α2β2) α-ketoaciddecarboxylase that irreversibly oxidizes pyru-vate to acetyl-CoA in the presence of thiaminepyrophosphate (TPP) and catalyzes the subse-quent reductive acetylation of the lipoyl moietyof dihydrolipoamide transacetylase (E2; EC2.3.1.12). Lipoyl groups (Lip) that are covalentlylinked to E2 facilitate transfer of both protonsand acetyl groups between the differ-ent component enzymes of the PDC. The result-ing acetyl-lipoamide intermediate reacts with

78 FUEL METABOLISM AND ENERGETICS

Table 7-1. Neurobehavioral Examinationof Twins (Aged 3 years 10 months)

Developmental Age (months)Test Ann Elizabeth

Vineland* 13–21 6–11

Bayley† mental 19 7

Bayley† motor 24 8

*Based on parental assessment of skills of communication, dailyliving, socialization, and motor function.†From Black and Matula, (2000).

Table 7-2. Residual Activity of the Pyruvate Dehydrogenase Complex (PDC) inBlood Lymphocytes or Cultured Skin Fibroblasts

Residual Activity (% of normal)Enzyme Component Ann Elizabeth

PDC (total activity) 55% ( lymphocytes) 18% (lymphocytes)28% (fibroblasts) 14% (fibroblasts)

E1 (pyruvate dehydrogenase) 34% (fibroblasts) 10% (fibroblasts)

E2 (dihydrolipoyl transacetylase) Normal Normal

E3 (dihydrolipoyl dehydrogenase) Normal Normal

coenzyme A (CoASH) to form acetyl-CoA andreduced lipoamide [Lip(SH)2]. NAD+ is the finalelectron acceptor, and reducing equivalents aretransferred to the respiratory chain via NADH.

The central feature of the complex in eukary-otic cells is the 60-mer E2 core, which is highlyconserved among species. The core is arrangedas a pentagonal dodecahedron, a cagelike struc-ture that serves as a scaffold about which theother components are organized (Fig. 7-3).Mammalian PDCs require a binding protein(BP;previously termed protein X) to anchor thedehydrolipoyl dehydrogenase dimer (E3; EC1.8.1.4) to the core. The relative stochiometry

for the structural components of bovine kidneyPDC is approximately 22 molecules of E1,about 6 molecules of E2, and about 6 moleculesof E3. Thus, each multienzyme complex con-tains potentially 60 centers for acetyl-CoA syn-thesis. Highly flexible structural domains in thePDC promote the channeling of intermediatesof catalysis between the active sites and con-tribute to the remarkable structural and func-tional organization of the complex.

The PDC is a “junction box” for cellular me-tabolism and energetics in which substrates maybe diverted toward anabolic pathways (e.g., car-bohydrate synthesis) or catabolic reactions and

PDC Deficiency 79

Figure 7-1. Pathways of fuel metabolism and oxidative phosphorylation. Pyruvate maybe reduced to lactate in the cytoplasm or may be transported into the mitochondria foranabolic reactions, such as gluconeogenesis, or for oxidation to acetyl-CoA by the pyru-vate dehydrogenase complex (PDC). Long-chain fatty acids are transported into mito-chondria,where they undergo β-oxidation to ketone bodies (liver) or to acetyl-CoA (liverand other tissues). Reducing equivalents (NADH, FADH2) are generated by reactions cat-alyzed by the PDC and the tricarboxylic acid (TCA) cycle and donate electrons (e−) thatenter the respiratory chain at NADH ubiquinone oxidoreductase (Complex I) or at succi-nate ubiquinone oxidoreductase (Complex II). Cytochrome c oxidase (Complex IV) cat-alyzes the reduction of molecular oxygen to water, and ATP synthase (Complex V)generates ATP from ADP. Reprinted with permission from Stacpoole et al. (1997).

ATP synthesis (oxidative phosphorylation). It isnot surprising, therefore, that the complex ishighly regulated (Fig. 7-2). The PDC undergoesrapid posttranslational changes in catalytic ac-tivity by reversible phosphorylation of up tothree serine residues on the E1α protein, ren-dering the complex inactive.Phosphorylation ismediated by pyruvate dehydrogenase kinase(PDK), a dimeric protein. In mammals, at leastfour isoenzymes (PDK1–PDK4) are expressedin a tissue-specific manner. PDK2 is expressed

ubiquitously, whereas tissue expression of theother isoforms is more restricted. It is note-worthy that PDK4 is highly expressed in tis-sues (e.g., heart, skeletal muscle, liver, kidney,and pancreatic islets) involved in fuel home-ostasis.

PDK activity is enhanced by increases in theratios of NADH/NAD+ or acetyl-CoA/CoA and isinhibited by pyruvate or ADP.Pyruvate dehydro-genase phosphatase (PDP) catalyzes the de-phosphorylation of E1α and activates the PDC.


Acetyl-SCoA + NADH + H+ + CO2

NAD+

NADH2

[FADH2]

[LipS2]

[TPP]

[Lip(SH)2]

[CH3CH-TPP]

[CH3C-SLipSH]CH3C-SCoAAcetyl-CoA

CoASH

[FAD]

CH3CCO2HPyruvate

phosphatase

PDH(inactive)

PDH(active)

ATPkinase

(−)

DCA

CO2

E1 E2

E3

O

Pyruvate + NAD+ + CoASHNet reaction:

O

OH

O

PO4-3

Figure 7-2. Reactions of the pyruvate dehydrogenase (PDH) multienzyme complex(PDC). Pyruvate is decarboxylated by the PDH subunit (E1) in the presence of thiaminepyrophosphate (TPP). The resulting hydroxyethyl-TPP complex reacts with oxidizedlipoamide (LipS3), the prosthetic group of dehydrolipoamide transacetylase (E2), to formacetyl lipoamide. In turn, this intermediate reacts with coenzyme A (CoASH) to yieldacetyl-CoA and reduced lipoamide [Lip(SH)2]. The cycle of reaction is completed whenreduced lipoamide is reoxidized by the flavoprotein, dehydrolipoamide dehydrogenase(E3). Finally, the reduced flavoprotein is oxidized by NAD+ and transfers reducing equiva-lents to the respiratory chain via reduced NADH. PDC is regulated in part by reversiblephosphorylation, in which the phosphorylated enzyme is inactive. Increases in the in-tramitochondrial ratios of NADH/NAD+ and acetyl-CoA/CoASH also stimulate kinase-mediated phosphorylation of PDC. The drug dichloroacetate (DCA) inhibits the kinaseresponsible for phosphorylating PDC, thus “locking” the enzyme in its unphosphory-lated, catalytically active state. Reprinted with permission from Stacpoole et al. (2003).

PDP is stimulated by calcium and magnesium.Like PDK, PDP is a dimer, but it is more looselybound than the kinase to the complex. Mam-mals have two PDP isoforms, each of whichcontains both regulatory and catalytic subunits.PDP1 is highly expressed in cardiac and, to alesser degree, skeletal muscle. PDP2 is presentin both oxidative (heart, kidney) and lipogenic(liver, adipose) tissues.

From this discussion, it follows that the activ-ity of the PDC tends to be directly associatedwith high rates of ATP turnover or high concen-trations of pyruvate, that is, conditions duringwhich the oxidative removal of glucose, lactate,and pyruvate is accelerated. In contrast, PDCactivity tends to be inversely associated with di-version of these substrates toward gluconeoge-nesis. A reciprocal relationship exists in sometissues between the oxidation of carbohydrateand long-chain fatty acids that is mediated, inpart, by the ratio of phosphorylated/unphos-phorylated PDC. This phenomenon, termed theglucose–fatty acid cycle, is best demonstrable

in tissues that normally exhibit high rates ofoxidative metabolism.An exception is the cen-tral nervous system (CNS), which cannot me-tabolize fatty acids and relies almost exclusivelyon the oxidation of glucose or lactate for energyunder fed conditions or on the combination ofcarbohydrate and ketone bodies (acetoacetate,β-hydroxybutyrate) during fasting.

It is not surprising, therefore, that the activityof the PDC is high in brain, where the complexexists predominantly in its unphosphorylatedform. In contrast, the PDC is more highly regu-lated in many other tissues to meet fluctuatingmetabolic demands. For example, enzyme activ-ity in liver is highest in the fed state, duringwhich acetyl-CoA can be utilized for lipogene-sis. In contrast, hepatic PDC activity is sup-pressed during fasting, and pyruvate is divertedtoward gluconeogenesis via oxaloacetate for-mation. In the resting state, myocardial cellspreferentially oxidize long-chain fatty acids, andthe PDC is mainly phosphorylated and inactive.However, during myocardial work, the ratio ofunphosphorylated to phosphorylated PDC isincreased, as is the proportion of energy de-rived from aerobic glucose metabolism.

Congenital Disorders of the Pyruvate Dehydrogenase Complex

The PDC is the subject of intense scrutiny be-cause of its pivotal role in fuel metabolism andits association with numerous acquired andcongenital disorders. Descriptions of proven orputative acquired deficiencies of the complexmay be found elsewhere; this chapter focuseson the genetics, biochemistry, clinical presenta-tion, and course of congenital defects in thePDC. Over 200 cases of PDC deficiency havebeen reported, and many other cases have beendiagnosed but remain unpublished. The diagno-sis in most patients has been based on demon-strating reduced total catalytic activity of thecomplex or in one of its component enzymes.

For diagnostic purposes, these assays are ap-plied most commonly to freshly isolated pe-ripheral blood lymphocytes, fresh or frozenskeletal muscle (usually quadriceps),or primarycultures of skin fibroblasts. The procedure isbased on the TPP/NAD+/CoASH-dependent de-carboxylation of 1-14C-pyruvate, with residualactivity of total PDC, E1, E2, and E3 expressedindividually as nanomoles product/minute incu-bation at 37°C/milligram total cell protein.Preincubation of cells with DCA (which inhibitsPDK) or fluoride (which inhibits PDP) can be

PDC Deficiency 81

Figure 7-3. Cutaway model of the fully assembledpyruvate dehydrogenase complex (PDC) viewedon its threefold axis. The dehydrolipoyl dehydroge-nase (E3) homodimer (black) is docked into thepentagonal opening of the core (gray). The bindingprotein (BP)-E3 components associated with thecore are not revealed in this model.The inner link-ers (dotted) bind pyruvate dehydrogenase (E1)(light gray) to the dihydrolipoyl transacetylase (E2)scaffold (gray).The E1-binding site of the E2 innerlinker is located about 50 Å above the scaffold, asindicated by the asterisk and serves as the anchorfor the lipoyl domains to pivot.The swinging armpivots about a position that is approximately 50 Åfrom the E1, E2, and E3 active sites. Reprinted withpermission from Zhou et al. (2001). Copyright2001, National Academy of Sciences, U.S.A.

used to examine activation/inactivation of thecomplex. Results obtained in patient cellsshould be interpreted relative to clinical pre-sentation and to enzyme activities obtained inboth concurrently assayed and previously mea-sured (historical) control cell samples. In manycases, a definitive diagnosis requires repeatedassay of cells from a single tissue sample orfrom different tissue types (e.g., lymphocytesand fibroblasts).

Several additional strategies have been em-ployed as aids in the diagnosis of PDC defi-ciency. These include: (1) mutational analysis,based on sequencing of genomic DNA, comple-mentary DNA, or both; (2) visualization of theproteins of the complex by electrophoresis/immunoprecipitation or electrophoresis/im-munoblotting techniques; and (3) most re-cently, an immunocytochemical method forassaying the content of E1α protein in individ-ual patient cells.

E1 Defects

Approximately 90 percent of biochemicallyproven causes of PDC deficiency are due todefects in the E1α subunit, which is on the Xchromosome. Both hemizygous males and het-erozygous females are affected by this condi-tion, such that E1α deficiency is representedequally between the sexes.Most cases appear tobe new mutations. In other cases, the mothermay exhibit little or no clinical phenotype.Three general classes of mutations in the E1αgene have been identified (Okajima et al., 2005):

1. Missense point mutations, resulting in lossof PDC catalytic activity without reductionof E1α protein or mRNA (stable, im-munoreactive, catalytically impaired en-zyme).

2. Deletions or exon-skipping mutations, re-sulting in significant loss of E1α mRNAand protein (diminished E1α synthesis).These cells will likely also exhibit loss ofimmunoreactive E1β protein since it hasbeen shown that the presence of bothsubunits is required for stability of theirnormal heterotetrameric structure.

3. Deletions, missense, or nonsense muta-tions, resulting in loss of immunoreactiveE1α and, consequently, E1β without lossof E1α mRNA (unstable E1α).

In general, the clinical manifestations of E1αdeficiency are more severe in boys than in girls.

Greater variability of expression of the defect infemales is due to random inactivation ( lyoniza-tion) of one X chromosome during early embry-onic development. Although all tissues fromaffected females contain cells that express themutant gene, the relative distribution of normaland mutated cells among tissues is stochastic.Therefore, the clinical phenotype reflects pri-marily the combined effects of three indepen-dent factors:

1. The severity of the particular mutation,meaning the level of residual PDC activity,if any, that is retained by cells harboringthe mutant allele

2. The degree to which X inactivation (in fe-males) involves the mutant allele in agiven tissue or organ

3. The susceptibility of a particular tissue tofactors 1 and 2 above, meaning the depen-dency of that tissue on the PDC for sup-porting its energy needs

This concept helps rationalize the clinical, bio-chemical, and anatomical heterogeneity of pa-tients with PDC deficiency (Table 7-3) and thefact that the reported frequency of this X-linkeddisease is similar between genders. It also ex-plains the particular vulnerability of the CNS toE1α mutations, given the rate-limiting quality ofPDC for fuel metabolism in nerve tissue andexperimental evidence that the complex mustoperate at near capacityi.e., predominantly un-phosphorylated) for normal brain function. Inheterozygous female individuals, even a modestreduction of PDC activity in brain cells may limitnormal function, whereas more severe muta-tions or a pattern of X inactivation favoring ex-pression of the mutant allele will cause moresevere CNS consequences. Such variable expres-sion of the defect within the brain may also ac-count for the anatomical location and extent ofCNS structural abnormalities associated withPDC deficiency and observed by brain imaging,although this hypothesis has not been ad-dressed experimentally.

The qualitatively and quantitatively variableimpact of the mutation also helps us to appre-ciate the challenge in establishing an unequivo-cal biochemical diagnosis, in which variabilitywithin tissues (e.g., cultured fibroblast linesderived from separate biopsies) and among tis-sues (e.g., skin, muscle, leukocytes) of the en-zyme deficiency is common. Although anindividual may exhibit striking lactic acidosisand CNS clinical and neuroimaging findings


Table 7-3. Reported Clinical, Metabolic, and Anatomical Manifestations of PyruvateDehydrogenase Complex DeficiencyFinding Comment

ClinicalFailure to thriveDevelopmental delay Most common early manifestationsHypotoniaMental retardation Almost universalSeizures Common; includes West syndromeChoreoathetosisMyoclonic jerksDecerebrate posturingAtaxiaDiminished deep tendon reflexesPeripheral neuropathy Formal nerve conduction testing requiredCortical blindnessOphthalmoloplegiaPtosisDysphagiaAbnormal electrocardiogram Tachycardia, conduction abnormalitiesRespiratory disturbances Tachypnea; awake or sleep apneaCyanosis

MetabolicSerum or plasma

↑ Lactate* Highly variable↑ Pyruvate Highly variable↓ pH (venous) Common, but usually mild↓ pH (arterial) Uncommon, unless in metabolic crisis↑ β-Hydroxybutyrate From poor feeding or ketogenic diet↓ Bicarbonate↑ Chloride Mild hyperchloremic, metabolic acidosis commonAnion gap↓ Creatinine Low muscle mass↑ Ammonia↑ Alanine From block in pyruvate oxidation↓ Protein↑ Uric acid

Urine↑ Lactate

Most common abnormalities↑ Ketones↓ Creatine clearance↑ Calcium↑ Protein↑ Amino acids↑ Uric acids↓ FeNa and FePO4

†

Fanconi syndrome‡

Cerebrospinal fluid↑ Lactate Often higher than corresponding blood level

AnatomicCephalic

MicrocephalyFacial dysmorphysim§ Many featuresOptic atrophyBrain dysgenesis� Many features

�

�

�

(Continued)

83

consistent with PDC deficiency, the inability tobiopsy the most affected tissues leads to a de-pendence on measuring enzyme activity in al-ternative cell types in which PDC activity is inthe normal range. For similar reasons, it may beextremely difficult to prove biochemicallywhether an apparently healthy female is a car-rier of the mutation. Furthermore, the antena-tal diagnosis of E1α deficiency in male fetusescan be performed using currently availablemeasurements of enzymatic activity and im-muno(cyto)chemical analyses described in the“Diagnosis” section. However, in the case of fe-male fetuses, normal enzymatic or immuno-chemical findings do not exclude a diagnosis ofE1α deficiency.

It might be assumed that the amount ofresidual E1α activity measured in cells (fibrob-lasts, leukocytes) would correlate more closelywith clinical phenotype in males who are hem-izygous for a mutation than in females, in whomthe biochemical expression of the defect de-pends on the degree of lyonization. However,there is generally poor correlation betweenmeasured enzyme activity and clinical presenta-tion and course, except for newborns of eithersex with marginal residual activity, in whom ful-minant lactic acidosis usually portends deathwithin days or a few weeks of birth.Affected pa-tients who survive the neonatal period may stillexhibit early clinical manifestations (Table 7-4),intermittent or persistent hyperlactatemia, and


Table 7-3. (Continued)Finding Comment

OtherShort stature Almost universalShort fingers/armsSimian creasesRenal dysmorphysim Variable

*Persistently or intermitantly normal or moderately elevated (≤ 5 mmol/L) in stably ill patients.†Fractional excretion of sodium and phosphorus.‡Based on the presence of at least three of the following findings: generalized aminoaciduria, glucosuria,increased FePO4, decreased blood bicarbonate concentration.§May include narrow head, frontal bossing, low anterior hairline, upslanting eyes, wide nasal bridge, upturnednose, long philtrum, flared nostrils, low set ears, micrognathia.�May include Leigh’s syndrome; hypogenesis/agenesis of corpus callosum; cerebral atrophy; ventriculomegaly;ectopic olivary nuclei; hypoplasia of basal ganglia, cerebellum, and brainstem; cystic changes in cerebrum, basalganglia, and brainstem.

Table 7-4. Naturally Occurring Compounds of Unproven Efficacy AdministeredSingly or in Combination to Patients with Pyruvate Dehydrogenase Complex(PDC) Deficiency

Compound Purported Mechanism

Carnitine Raises CoASH/acetyl-CoA ratio

Lipoic acid Antioxidant and PDC cofactor

Coenzyme Q Antioxidant and respiratory chain component

Vitamins C and K Facilitate electron transport by respiratory chain

Thiamine* PDC cofactor

Riboflavin Precursor of FAD (PDC cofactor)

Biotin† Pyruvate carboxylase cofactor

Vitamin E Antioxidant

CoA, coenzyme A.

*Rare cases of thiamine-responsive PDC deficiency have been well documented.†Biotin is a cofactor of various carboxylases, but its effectiveness in pyruvate carboxylase deficiency isunproven. However, it has been used effectively in cases of biotinidase deficiency, a vary rare cause ofcongenital lactic acidosis.

CNS structural abnormalities, but their illnesstends to be more protracted, and the measuredbiochemical defect tends to be less severe.

It is important to emphasize that the concen-tration of blood lactate is a relatively insensitiveindicator of disease severity. Basal (postpran-dial, resting) lactate levels may be normal oronly minimally elevated in clinically severely af-fected patients unless a secondary stress (i.e.,energy demand) such as a respiratory infectionor an asthmatic attack precipitates life-threatening acid–base decompensation. Otherpatients with similar measurable degrees ofresidual PDC activity and symptomatology mayhave persistent moderate (2–5 mmol/L) hyper-lactatemia, often in association with decreasedblood bicarbonate and increased blood chlo-ride levels, a mildly increased anion gap (usuallydefined as the difference, in mmol/L, betweenthe sum of the serum concentrations of sodiumand potassium minus the serum concentrationsof chloride and bicarbonate; the normal aniongap range is 12–14 mmol/L) and a modestlydepressed venous pH. In stably ill patients,the circulating concentration of pyruvate usu-ally mirrors that of blood lactate, so the lac-tate/pyruvate ratio remains normal. The CSFlactate level is often higher than the correspon-ding blood concentration and probably reflectsthe extreme vulnerability of the CNS to PDC de-ficiency. Urinary lactate is frequently presentand is the principal organic acid anion excretedin PDC deficiency.

Three cases of defects in E1β have been re-ported. This condition is inherited in an autoso-mal recessive manner.

E2, E3 and Binding Protein Defects

A small number of patients have been de-scribed in whom there is good evidence for adefect in the E2 (dihydrolipoyl transacetylase)or E3 (dihydrolipyl dehydrogenase) gene or inthe HsPDX1 gene encoding the PDC bindingprotein (BP). Clinical phenotype varies widelyand includes intermittent or persistent hyper-lactatemia and a spectrum of cognitive functionfrom normal to profound psychomotor retarda-tion with Leigh syndrome (a genetically hetero-geneous condition characterized bydevelopmental delay or regression, lactic acido-sis, and focal, symmetrical, necrotic, and de-myelinating lesions of the basal ganglia of thebrain; also called subacute necrotizing en-cephalomyelopathy).

No clinical signs distinguish patients withthese defects from each other or from patientswith E1α deficiency. The diagnosis is based ondetermination first of a decrease in overall PDCactivity, followed by enzymatic evidence of adeficiency in E2 or E3 or reduced expression ofE2, E3, or BP on immunoblotting, with normalamounts of the other proteins, including E1α.Patients with E3 mutations manifest combinedα-keto acid deficiency, with elevated concentra-tions of pyruvate, lactate, α-ketoglutarate, andbranched-chain amino acids (valine, leucine,isoleucine) in blood and increased excretion ofpyruvate, lactate, α-hydroxybutyrate, α-ketoglutarate, and α-hydroxyisovalerate inurine. Chorionic villus sampling was employedto diagnose a female fetus with a large homozy-gous deletion found at the 5' end of theHsPDX1 coding sequence. This and other casesof defective BP result in low residual activity ofPDC and illustrate the essential role of the BP inpreserving the overall structural and functionalintegrity of the complex.

Pyruvate Dehydrogenase PhosphataseDefects

At least four cases of PDC deficiency have beendescribed in which the intrinsic defect lay not inthe complex per se but in one of the two gene-tically distinct isozymes required to dephospho-rylate the serine residues of the E1α subunit.The clinical signs of PDP deficiency do not dis-tinguish it from primary defects of E1α, E1β, E2,E3, or BP. The most telling biochemical abnor-mality reported to date has been the inability toactivate PDC in cells cultured with DCA. How-ever, neither direct measurements of PDP activ-ity nor mutations in the PDP gene have beenreported for any published case of phosphatasedeficiency.

THERAPY

Treatment of most patients with genetic mito-chondrial diseases has been disappointing andhas usually been approached in a sporadic, un-controlled manner. There is no proven therapyfor patients with PDC deficiency.Current strate-gies rely on nutritional or pharmacological in-terventions or both to improve patient qualityof life. Recent studies have also begun to ad-dress the potential role of gene transfer for E1αdefects.

PDC Deficiency 85

Nutritional “cocktails” are commonly pro-vided patients with PDC deficiency or othermitochondrial diseases (Table 7-4).While thesevary in composition, they are generally predi-cated on the belief that (1) the components aresafe and (2) they may interdict at various pointsin the process of oxidative phosphorylation toenhance energy production or mitigate the ac-cumulation of reactive oxygen species gener-ated in abnormal amounts by a compromisedrespiratory chain. However, use of two nutri-tional interventions, so-called ketogenic dietsand high doses of thiamine (vitamin B1), aremore strongly anchored to sound biochemicalrationales and have been investigated widely, al-beit anecdotally.

Ketogenic Diets

Ketogenic diets have been used for almost acentury in the treatment of epilepsy in chil-dren. Although a compelling biochemical ra-tionale for this intervention has never beenwell defined, the usual ketogenic diet forepileptics consists of a 4:1 caloric ratio of fatsto carbohydrates and proteins. Ketogenic dietshave also been employed to treat certain severeacquired or congenital errors of intermediarymetabolism. Patients with PDC deficiency donot oxidize carbohydrates efficiently;hence, thepyruvate derived from glycolysis is more likelyto be reduced in the cytoplasm to lactate. In-deed, carbohydrate-containing meals may exac-erbate or precipitate life-threatening lacticacidosis in patients with severe PDC deficiency.This has led to the widespread use of high-fatdiets that induce ketosis and provide an alterna-tive source of acetyl-CoA, particularly for theCNS, which can utilize ketone bodies for en-ergy metabolism (Fig.7-1).Case reports of a fewchildren with PDC deficiency whose clinicalcourse improved dramatically while following ahigh-fat diet are consistent with this postulate.

Unfortunately, ketogenic diets have neverbeen applied in a consistent or controlled man-ner to patients with PDC deficiency, leading toconsiderable variation in both the quality andquantity of fat calories provided to patients. Ingeneral, however, published reports in whichthe dietary composition has been specified typ-ically include a caloric distribution of 55% to80% fat, up to 25% carbohydrate, and 10% to20% protein. Strong proponents of ketogenic di-ets for PDC deficiency advocate a fat intake of

at least 75% of total calories. Traditionally, a highproportion of the fatty acids provided by such aregimen has been saturated fatty acids.

Despite the logic of ketogenic diets for PDCdeficiency on biochemical grounds, there areseveral theoretical concerns regarding theirlong-term use. For example, diets in which thefat intake is relatively low (≤65%) may containan amount of protein (to accommodate caloricneeds) that exceeds the age-adjusted recom-mended dietary allowance. The acid load fromsuch diets could exacerbate the congenitalacid–base disorder in PDC-defective individualsand could contribute to the renal hyperfiltra-tion reported to occur in some patients. Keto-genic diets may also cause hypercalciuria andincrease the risk of bone demineralization inPDC deficiency.

Last,diets habitually high in fat can exert pro-found effects on both carbohydrate and lipidmetabolism. Oxidation of long-chain saturatedfatty acids increases the intramitochondrial ra-tios of acetyl-CoA/CoASH and NADH/NAD+ andconsequently the activity of PDKs that re-versibly phosphorylate and inhibit the PDC(Table 7-3). Therefore, diets high in saturatedfats might further dampen any residual PDC ac-tivity in affected patients. In contrast, polyunsat-urated fatty acids, particularly those of theomega-3 class, are reported to have opposite ef-fects on PDC activity. Moreover, investigationsin animals and humans have demonstrated thatconsumption of diets enriched in saturated fatprovoke deleterious effects on circulating con-centrations of lipids and lipoproteins and in-hibit the action of insulin on peripheral tissues.In comparison, diets high in mono- or polyun-saturated fatty acids tend to improve lipid me-tabolism and insulin action.

In summary, cogent reasons exist for incor-porating ketogenic diets into the standard ofcare for PDC deficiency. Equally compelling rea-sons exist for undertaking a prospective, rigor-ously controlled evaluation of the long-termbenefits and risks of such regimens, particularlyregarding their nutrient composition.

Thiamine

Very high doses of thiamine, sometimes ex-ceeding 2 g/day, are reported to benefit somepatients with PDC deficiency. TPP is an obli-gate cofactor for the E1 component of thePDC, and two molecules are bound to each


α2β2 tetramer. According to a model of theTPP-binding site for E1, the pyrophosphatemoiety of TPP is bound by means of a singlecalcium ion to a site formed by closure of aflexible loop of the α-subunit, whereas the thi-azolium ring of TPP is bound to a site locatedon the β-subunit.

Most early clinical descriptions of apparentthiamine-responsive PDC deficiency were notcharacterized biochemically to ascertain truethiamine dependence. In subsequent reports,immunochemical analyses have demonstratedvaried patterns of α- and β-subunit expression,and in vitro studies of cultured cells havesometimes found altered E1 enzyme kinetics(high Km, low Vmax) for TPP. When moleculargenetic analyses have been undertaken, differ-ent mutations have been identified within theconserved TPP-binding motif that are consid-ered to lead to diminished binding affinity forTPP or to decreased stability of the α2β2

tetramer.

Dichloroacetate

DCA is the only drug that has been investigatedin detail for the treatment of PDC deficiency. Itinhibits PDKs and thus maintains E1α in its un-phosphorylated, catalytically active form. DCAis a potent lactate-lowering agent, primarily byvirtue of its effect on the PDC.The rationale forits use in PDC deficiency is based on the expec-tation it would stimulate residual enzyme activ-ity in patients with either partial PDC activityor enzyme heterozygosity. In the latter case, ac-tivation of E1 by DCA in normal cells mighthelp decrease lactate accumulation in adjacentdefective cells and improve overall tissue or or-gan energy metabolism.

Based on promising anecdotal reports, a ran-domized, double-blind, placebo-controlled trialof DCA was conducted in 43 children with con-genital lactic acidosis due to deficiencies inPDC or one or more respiratory chain com-plexes or to a mutation in mitochondrial DNA(mtDNA). Comparisons were made after 6months of 12.5 mg/kg DCA, administered twicedaily by mouth or feeding tube, or placebo.Allpatients received thiamine (1 mg/kg/day). DCAwas well tolerated and significantly decreasedblood lactate concentrations following carbohy-drate meal challenges but did not alter variousclinical indices of neurological or neurobehav-ioral function or the frequency or severity of

intercurrent illnesses or hospitalizations. Al-though DCA may cause reversible peripheralneuropathy and hepatotoxicity, the frequencyof these adverse events did not differ betweentreatment groups. Following the double-blindphase of this trial, 6 patients have continued re-ceiving open-label DCA for up to 7 years with-out toxicity. Four subjects have PDC E1αdeficiency, and their clinical course has re-mained stable. It is unknown whether DCAmay be particularly effective in this or othersubgroups of patients with congenital lacticacidosis.

Gene Transfer

Human gene therapy for genetic mitochondrialdiseases is in its preclinical infancy and has fo-cused mainly on overcoming defects inmtDNA. However, the central role of the PDCin cellular energetics and the relatively highfrequency of PDC deficiency as a cause of con-genital lactic acidosis make the complex an at-tractive target for gene therapy. Furthermore,since all the PDC components are nuclear en-coded, evolutionary mechanisms already existfor their importation from the cytoplasm intothe mitochondria, unlike proteins normally en-coded by mtDNA.

Studies using recombinant adeno-associatedvirus as a vector to deliver full-length E1α withits own mitochondrial targeting presequencehave demonstrated the potential utility of thisstrategy for defects in PDC. Transduction ofcells with the recombinant adeno-associatedvirus E1α vector resulted in its localization inmitochondria and in partial restoration of PDCactivity in cultured fibroblasts from a male pa-tient with E1α deficiency. It is unknownwhether effective gene transfer of wild-typeE1α will depend on the type of mutation har-bored by defective cells or whether enhance-ment of enzyme activity can be achieved invivo, initially in an animal model of the disease.This last goal is particularly challenging due tothe difficulty in generating viable offspringfrom animals with targeted inactivation of theE1α gene. Resolving these problems raises theintriguing possibility of eventually combininggene and pharmacological therapy for E1α defi-ciency, in which successful gene transfer wouldbe followed by administration of DCA in thehope of maximally dephosphorylating thetransgene.

PDC Deficiency 87

QUESTIONS

1. Based on the genetics of PDC, how couldthe phenotypic heterogeneity betweenidentical twins Ann and Elizabeth be ex-plained?

2. Why are some tissues more vulnerable toinhibition of PDC activity, and how doesthis selective tissue vulnerability explainthe cardinal clinical manifestations of thedisease?

3. What are the common classes of muta-tions in the PDCA1 gene that give rise toPDC deficiency?

4. Why might the measurement of PDC ac-tivity vary markedly among cells from thesame patient?

5. What is the biochemical rationale for ke-togenic diets in treating PDC deficiency,and what are the potential caveats associ-ated with their use?

6. What is the primary mechanism of actionof dichloroacetate, and why might pa-tients with E1α deficiency be responsiveto this drug?

BIBLIOGRAPHY

Black MM, Matula K: Essentials of Bayley Scales ofinfant development. II Assessment. New York,NY,Wiley, 2000.

Brown RM, Fraser NJ, Brown GK: Differentialmethylation of the hypervariable locus DXS255on active and inactive X chromosomes corre-lates with the expression of a human X-linkedgene. Genomics 7:215–221, 1990.

Fuerhlein BS, Rutenberg MS, Silver JN, et al.: Differ-ential metabolic effects of saturated versuspolyunsaturated fats in ketogenic diets. J ClinEndocrinol Metab 89:1641–1645, 2004.

Holness MJ, Sugden MC: Regulation of pyruvatedehydrogenase complex activity by reversible

phosphorylation. Biochem Soc Trans 31:1143–1151, 2003.

Kerr DS: Treatment of congenital lactic acidosis:review. Intern Pediatr 10:75–81, 1995.

Lib MY, Brown RM, Brown GK, et al.: Detection ofpyruvate dehydrogenase E1α subunit deficienciesin females by immunohistochemical demonstra-tion of mosaicism in cultured fibroblasts. J His-tochem Cytochem 50:877–884,2002.

Lissens W, De Meirleir L, Seneca S, et al.: Mutationsin the X-linked pyruvate dehydrogenase (E1)αsubunit gene (PDHA1) in patients with a pyru-vate dehydrogenase complex deficiency. HumMutat 15:202–219, 2000.

Okajima K, Mewhort LZ, Lusk MM, et al.: Relation-ships of clinical and biochemical manifestationsto genotype analysis in patients with pyruvatedehydrogenase complex deficiency. MolecularGenetics and Metabolism 84:232–33, 2005.

Robinson BH: Lactic academia: disorders of pyru-vate carboxylase and pyruvate dehydrogenase,in Scriver CR, Beaudet AL, Sly WS,Valle D (eds),The Metabolic and Molecular Bases of Inher-ited Disease. 8th ed. McGraw-Hill, New York,2001, pp. 2275–2296.

Stacpoole PW, Barnes CL, Hurbanis MD, et al.:Treatment of congenital lactic acidosis withdichloroacetate: a review. Arch Dis Child 77:535–541, 1997.

Stacpoole PW,Owen R,Flotte TR: The pyruvate de-hydrogenase complex as a target for gene ther-apy. Cur Gene Ther 3:239–245, 2003.

Weber TA, Antognetti MR, Stacpoole PW: Caveatswhen considering ketogenic diets for the treat-ment of pyruvate dehydrogenase complex defi-ciency. J Pediatr 138:390–395, 2001.

Wexler ID, Hemalatha SG, McConnell J, et al.: Out-come of pyruvate dehydrogenase deficiencytreated with ketogenic diets: studies in patientswith identical mutations. Neurology 49:1655–1661, 1997.

Zhou ZH, McCarthy DB, O’Connor CM, et al.: Theremarkable structural and functional organiza-tion of the eukaryotic pyruvate dehydrogenasecomplexes. Proc Natl Acad Sci USA 98:14802–14807, 2001.


CASE REPORT

A 17-year-old Caucasian female was referred toa neurologist following a 3-day episode of se-vere headache accompanied by repeated vom-iting. She complained of weakness in her armsand legs and of a general feeling of fatigue,especially after minor amounts of exercise,which left her “feeling exhausted.” She indi-cated that she had suffered from migraineheadaches “since she was little,” and duringher recent episodes she had auditory and vi-sual hallucinations. Her speech was slurred,and she had slight paralysis of the right side(hemiparesis).

Her birth history was normal. She was theproduct of a nonconsanguineous marriage andweighed 3850 gm after an unremarkable preg-nancy, labor, and delivery. However, several mo-tor and mental developmental milestones wereabnormal. She was of short stature (<25th per-centile) and low weight (<15th percentile).Since age 8 years, she had been identified aslearning impaired and had attended special ed-ucation classes throughout elementary and sec-ondary school.

On physical examination, the patient wasslightly febrile (38.4°C). Serum concentrationsof lactate (24.6 mg/dL) and pyruvate (3.8 mg/dL) were elevated (normal values are 5–18 mg/dL and 0.55–1.0 mg/dL, respectively). Her lac-tate:pyruvate ratio was elevated as well (34:1;normal range 10:1 to 20:1). Analysis of lumbar

8

Mitochondrial Encephalomyopathy,Lactic Acidosis, and Strokelike Episodes (MELAS):

A Case of Mitochondrial Disease

FRANK J. CASTORA

89

cerebrospinal fluid (CSF) showed an abnormallyhigh protein concentration of 0.66 mg/dL (nor-mal range 0.18–0.58 mg/dL) and increased lac-tate (3.7 mmol/L; normal is < 2.8 mmol/L) andpyruvate (284 µmol/L;normal is 8–150 µmol/L).An electroencephalogram indicated moderateslowing of conduction in the right hemisphere,and a computed tomographic (CT) scan showeda hypodensity in the right temporoparietaland occipital cortices.On audiometric examina-tion, she was found to have decreased per-ception of high tones and moderate bilateralsensorineural hearing loss. Several of her ma-ternal relatives (mother, grandmother, and ma-ternal uncle) also had some degree of hearingloss.

In addition to ptosis of the left eye (Fig. 8-1),bilateral external ophthalmoplegia (paralysisof some or all of the muscles of the eye) andproximal muscle weakness were observedduring the neurological examination. Magneticresonance imaging (MRI) of the brain re-vealed signal intensity and diffusion coefficientchanges compatible with acute ischemic infarctas well as some degree of cerebellar and cere-bral atrophy.

The presence of lactic acidosis, bilateralhearing loss, progressive muscle weakness, andstrokelike episodes was suggestive of a mito-chondrial disorder. DNA extracted from periph-eral blood cells from the patient, her mother,grandmother, mother’s brother, and symptom-less younger sister was positive for a mutation

at nucleotide position 3243 of the mitochond-rial DNA (mtDNA) genome.

DIAGNOSIS

Mitochondrial encephalomyopathy, lactic acido-sis, and strokelike episodes (MELAS) is a pro-gressive neurodegenerative disorder with highmorbidity and mortality. Early death is a com-mon outcome. The reported ages of patientswith MELAS at the time of death range from 10to 35 years, but patients alive in their 50s alsohave been described. Most deaths are due tomedical complications; a few are a result of sta-tus epilepticus (an epileptic seizure that lastsmore than 30 minutes or a constant or near-constant state of having seizures without gain-ing consciousness between them).

The typical presentation of patients withMELAS includes mitochondrial encephalomy-opathy (a disorder causing both brain and mus-cle cells to function abnormally), lactic acidosis,and strokelike episodes. In addition, neurosen-sory hearing loss and diabetes mellitus aresometimes present as part of this syndrome. Itis also not unusual to find multiple organ sys-tems involved, including the central nervoussystem, skeletal muscle, eye, cardiac muscle,and, more rarely, the gastrointestinal system.The physical examination will often include anassortment of nonspecific findings such asshort stature, muscle weakness, and cognitiveand psychiatric problems (e.g., learning difficul-ties,dementia,psychosis,personality disorders).Cardiomyopathy with signs of congestive heartfailure may also be observed in the physicalexamination. On ophthalmologic examination,some patients may present with ophthalmople-gia, ptosis, and pigmentary retinopathy (heredi-tary, progressive degenerative disease of the eyemarked by atrophy and pigment changes in the

retina, constriction of the visual field, and even-tual blindness).

Strokelike episodes are the hallmark featureof this disorder. In many patients with MELAS,presentation occurs with the first stroke-likeepisode, usually when an individual is aged4–15 years. Episodes initially may manifest withvomiting and headache that may last up toseveral days. These patients also may be af-fected with episodes of seizures and visual ab-normalities followed by hemiplegia (total orpartial paralysis of one side of the body).Seizure types may be tonic-clonic or myoclonic.Migraine or migrainelike headaches observed inthese patients also may reflect the strokelikeepisodes. Pedigrees of patients with classicMELAS identify many members whose onlymanifestations are migraine headaches.

Some patients may experience hearingloss, which may accompany diabetes. Usually,type 2 diabetes is described in individuals withMELAS, although type 1 or insulin-dependentdiabetes also may be observed. Palpitationsand shortness of breath may be present insome patients with MELAS secondary to car-diac conduction abnormalities such as Wolff-Parkinson-White syndrome. Acute onset ofgastrointestinal manifestations (e.g., acute onsetof abdominal pain) may reflect pancreatitis,ischemic colitis, and intestinal obstruction.Numbness, tingling sensation, and pain in theextremities can be manifestations of peripheralneuropathy. Some patients may have the pre-sentation of Leigh syndrome (i.e., subacutenecrotizing encephalopathy).

A CT scan or MRI of the brain following astrokelike episode reveals a lucency (an area ofluminosity) that is consistent with infarction.Later, cerebral atrophy and calcifications maybe observed on brain imaging studies. The vas-cular territories of focal brain lesions and theprior medical history of these patients differsubstantially from those of typical patients withstroke. Serial MRI studies often demonstratelesion resolution, differentiating these lesionsfrom typical ischemic strokes. An electroen-cephalogram is often performed when seizuresare a concern. This is especially necessary inMELAS since patients occasionally have in-tractable status epilepticus as a terminal condi-tion. Mental deterioration usually progressesafter repeated episodic attacks. Psychiatricabnormalities (e.g., altered mental status,schizophrenia) may accompany the strokelikeepisodes. The encephalopathy may progress to


Figure 8-1. Ptosis (droopy eyelids). The patientshown here has left eye ptosis, a drooping of theleft upper eyelid,most likely due to paralysis of thethird nerve or to loss of sympathetic innervation.

dementia, and eventually the patient may be-come cachectic and die. Another cause ofhigh mortality is the less-common feature ofcardiac involvement. Causes can be a hyper-trophic cardiomyopathy and conduction abnor-malities, such as atrioventricular blocks orWolff-Parkinson-White syndrome.

When a patient is suspected of having a mito-chondrial disorder such as MELAS, the physicianwill usually combine clinical, biochemical, andmorphological information to establish a diag-nosis. Often, the diagnosis of a mitochondrialdisorder is obtained by a three-tiered testing ap-proach. The initial, noninvasive screening willinclude basic glucose and electrolyte analysis,and blood counts. In addition, screening for avariety of metabolic disorders that could giverise to similar symptoms will be performed.These include: thyroid and parathyroid disorders,Cushing syndrome, inflammatory myopathies(immune cells invading muscle tissue), en-docrinopathies (diabetes), and collagen vasculardiseases.Blood and urine ammonia,ketones, andlactic acid levels will also be determined.

If the history, physical findings, and labora-tory results are suggestive but not conclusiveof a particular mitochondrial disease (e.g.,MELAS), then a second tier of tests is per-formed. These include: blood and CSF lactateand pyruvate, as well as the lactate/pyruvate ra-tio; timed or random measurement of aminoacids in blood, urine, and CSF; organic acids inurine and CSF; and ketones and free and totalcarnitine in blood and urine.

Lactic acidosis is a very important feature ofthis disorder. After first eliminating the morecommon causes of lactic acidosis, such as tissuehypoxic-ischemic injury, hyperglycemia, and hy-poglycemia, the assessment of lactic acidosis inMELAS patients can be critically important inreaching the correct diagnosis of this disorder.In general, lactic acidosis does not lead to sys-temic metabolic acidosis, and it may be absentin patients with significant CNS involvement. Insome individuals with MELAS, acid levels maybe normal in blood but elevated in CSF. In respi-ratory chain defects, the ratio between lactateand pyruvate is high. Characteristics of lacticacidosis in MELAS are unique. Arterial lactateand pyruvate are high, and CSF lactate also maybe high. Lactate and pyruvate may increase sub-stantially with exercise. In concert with the ab-solute increases in lactate and pyruvate, thelactate/pyruvate ratio usually is increased aswell. Unlike tissue-injury lactic acidosis where

an increased ratio is coincident with decreasedO2 saturation, the increased lactate/pyruvate ra-tio of MELAS patients is observed at normal O2

saturation.Positron emission tomographic (PET) stud-

ies may reveal a reduced cerebral metabolicrate for oxygen utilization, while single-photonemission computed tomographic (SPECT) stud-ies can ascertain strokes in individuals withMELAS by using a tracer, N-isopropyl-p-[123-I]-iodoamphetamine. The tracer accumulates inthe parieto-occipital region, and it can delineatethe extent of the lesion. SPECT studies are usedto monitor the evolution of the disease.

At this point, if the results have suggested aspecific mitochondrial disorder (e.g., MELAS),then mtDNA isolated from blood cells can betested for any known mutations associated withthe suspected disorder. Unfortunately, manydisease-causing mutations are not detectable inmtDNA isolated from blood cells (because ofthe rapid turnover in these cells, defective mito-chondria are often lost). Therefore, inconclusiveresults may warrant further testing.

This third tier of diagnostic tests includes re-peat testing for some of the same compounds asbefore but now under different conditions (e.g.,after a fast or during an illness). In addition,a skinor muscle biopsy will be performed.As these lastprocedures are very invasive, the costs and risksmust be compared to the benefits gained fromestablishing the diagnosis.Furthermore,althoughmost hospitals can perform many of the basicdiagnostic tests, few are equipped to performall the metabolic or molecular genetic tests re-quired. Therefore, material from the skin or mus-cle biopsy must be appropriately prepared andshipped to the performing laboratory to main-tain the integrity of the sample.

The biopsy material is used to obtain bio-chemical, morphological, and genetic infor-mation to establish a diagnosis. Mitochondriaisolated from the fresh muscle biopsy or fromcells (fibroblasts) cultured from the skin biopsycan be used to measure respiratory chain activ-ity by polarography. This procedure measuresO2 consumption as different respiratory chainsubstrates are added to the mitochondrial sus-pension.

A recurring feature of MELAS and manyother mitochondrial disorders is the presenceof ragged red fibers that are demonstrable aftertreating the muscle cells with modified Gomoritrichome stain (Fig. 8-2). The muscle biopsy ispositive for ragged red fibers in at least 85% of

MELAS 91

MELAS cases. Ragged red fibers, common toMELAS, myoclonic epilepsy with ragged redfibers, Kearns-Sayre, and overlap syndromes, re-flect proliferation of abnormal mitochondriaunder the sarcolemma. However, some patientswith MELAS do not develop such a proliferationof abnormal subsarcolemmal mitochondria, so anegative muscle biopsy finding, particularlyearly in the course of disease progression, doesnot preclude consideration of this syndrome.

Enzyme histochemical staining of musclefibers can identify abnormal levels of respira-tory Complexes II, IV, and V, while specificimmunohistochemical stains can be used to

evaluate specific respiratory chain subunits(e.g., subunit I or II of cytochrome c oxidase).Muscle fibers from patients with MELAS will of-ten show significantly increased staining forsuccinate dehydrogenase (respiratory ComplexII) as well as reduced staining for cytochromec oxidase (respiratory Complex IV), althoughnormal or excessively positive staining for cy-tochrome oxidase can be seen in some raggedred fibers (Fig. 8-2).

Fresh or even frozen muscle can be used todetermine electron transport chain (ETC) ac-tivity by spectrophotometric assays. Patientswith MELAS have been found to have marked


Figure 8-2. Histochemical staining of muscle fibers. Left: Gomori trichome stain show-ing abnormally proliferating ragged red fibers (arrows). Center: Cytochrome oxidase(respiratory Complex IV) staining showing fibers with reduced or absent staining (long,thin arrow) and fibers with increased cytochrome oxidase staining (short, thick arrow).Right:Succinate dehydrogenase (respiratory Complex II) staining with arrows indicatingfibers with increased succinate dehydrogenase.

deficiency in the activity of complex I of therespiratory chain. Some patients with the disor-der have a combined deficiency of Complex Iand Complex IV. Morphologically abnormal mi-tochondria, some with paracrystalline bodies, aswell as abnormally proliferating mitochondriacan also be visualized by electron microscopy.

Molecular analysis of mtDNA from musclebiopsies often provides a definitive diagnosis ofMELAS. Individuals with more severe clinicalmanifestations of MELAS generally have greaterthan 80% mutant mtDNA in postmitotic tissuessuch as muscle. In approximately 80% of MELAScases, the responsible mutation is an A → Gbase substitution at nucleotide position 3243.A smaller subset of MELAS patients (7.5%)possesses a different point mutation, a T → Ctransition at nucleotide position 3271. At leasteight additional point mutations and one four-base pair deletion mutation also have been de-scribed.


Mitochondrial myopathies are a heterogeneousgroup of disorders that usually involve multipleorgan systems and that display a wide varietyof symptoms, including muscle weakness,retinopathy, deafness, seizure, cardiac conduc-tion defects, cardiomyopathy, and lactic acido-sis. The earliest report of a mitochondrialdisorder can be traced to Theodur Leber’s de-scription in 1871 of a patient suffering from bi-lateral vision loss and retinopathy (Howell,1999), although at that time it was not knownto be a mitochondrial disease.The first diseaseproposed to be a mitochondrial disorder wasreported in 1962 when Luft and coworkers de-scribed a female with polydipsia (abnormalthirst), polyphasia (extreme talkativeness), ex-cessive sweating, and a basal metabolic ratetwice the norm. It was 26 years later that thefirst reports of evidence associating mtDNAmutations with two mitochondrial diseases,Leber’s hereditary optic neuropathy (Wallace etal., 1988) and Kearns-Sayre syndrome (Lesti-enne and Ponsot, 1988), were published.

The mitochondrion, in addition to being the“powerhouse of the cell” (because it generatesmore than 90% of the ATP used by the cell), isalso the site of fatty acid oxidation, the tricar-boxylic acid (TCA) cycle, electron transport,and amino acid metabolism. Central to the uti-lization of fuel molecules—carbohydrates, pro-

tein, and fats—is the catabolism of these largemacromolecules into smaller molecules thatserve as substrates for the TCA cycle. Carbohy-drates and fats are metabolized to acetyl-CoA,which is also used in the TCA cycle to con-dense with oxaloacetate to form citrate. Pro-tein degradation also results in the generationof acetyl-CoA (from metabolism of 11 of theamino acids) as well as in amino acids thatserve as precursors for several of the TCA cycleintermediates.

In addition to the obvious role in the ca-tabolism of macromolecules, the TCA cycleprovides numerous intermediates for anabolicreactions, such as the synthesis of porphyrinsfrom succinyl-CoA,purines from α-ketoglutarate,pyrimidines from fumarate and oxaloacetate,and proteins from amino acids derived from ox-aloacetate, fumarate, and α-ketoglutarate.

The oxidation of the fuel molecules that enterthe TCA cycle results in the release of carbondioxide and the formation of the reduced elec-tron carriers NADH and FADH2.To maximize therecovery of energy (in the form of ATP) fromthese fuel molecules, the electrons carried byNADH and FADH2 are transferred to the electrontransport chain (ETC), a series of five complexesembedded in the inner mitochondrial mem-brane.The ETC is designed to recover the energyof the various fuel molecules by coupling thetransfer of electrons down the ETC to the syn-thesis of ATP, the energy currency of the cell, in aprocess termed oxidative phosphorylation (OX-PHOS).A diagram of the components of the OX-PHOS system is shown in Figure 8-3.

Electrons enter the ETC at respiratory Com-plexes I and II. The electrons from NADH enterat respiratory Complex I (RC I, NADH dehydro-genase) with the concomitant oxidation ofNADH to NAD+. The electrons carried byFADH2 are transferred to RC II (succinate dehy-drogenase) as the FADH2 is oxidized to FAD andsuccinate is reduced to fumarate. These elec-trons from RC I and II are transferred to thequinone form of coenzyme Q (CoQ), which de-livers them to RC III (UQ-cytochrome c reduc-tase). Cytochrome c then accepts the electronsfrom RC III, and the reduced cytochrome c isreoxidized as it delivers the electrons to RC IV,cytochrome c oxidase. The electrons are thenused by RC IV to reduce molecular oxygen towater.

At RC I, III, and IV, the transfer of electrons isaccompanied by the pumping, by these com-plexes,of H+ in the mitochondrial matrix across

MELAS 93

the impermeable mitochondrial inner mem-brane into the intermembrane space. The in-creased accumulation of H+ on the cytosolicside of the inner membrane establishes both aproton and electrical gradient. Ultimately, thegradient potential is used to drive H+ back intothe mitochondrial matrix through a channel inthe ATP synthase (RC V).This flow through thesynthase provides the energy necessary for thesynthesis of ATP by the condensation of inor-ganic phosphate with ADP.

As seen in Figure 8-4, disruption in the flowof electrons through the ETC can lead to an in-crease in NADH depending on the location ofthe block in electron transport.The increase inNADH will shut down the TCA cycle via spe-cific inhibition of key TCA enzymes by NADH.The buildup of NADH will thus result in deple-tion of NAD+ stores. How are the NAD+ levelsrestored? The NADH is used to reduce pyruvateto lactate, as indicated in the following diagram:

development of lactic acidosis as elevated lev-els of lactate create lower pH in the cell.

Another consequence of the failure of elec-trons to move completely through the ETC isthe increased release of reactive oxygen species(ROS) such as superoxide anion (O2

−) in themitochondrion. Such ROS are produced as oxy-gen molecules react with electrons at the CoQor RC I stage. The superoxide and other ROS de-rived from it (H2O2 and OH•) cause damage tonearby proteins, membranes, and nucleic acids.

Approximately 1000 proteins comprise themitochondrion; the majority are encoded ongenes located on nuclear DNA. In fact, as seenin Figure 8-5, the mtDNA encodes only 13 pro-teins. These mtDNA-encoded proteins are theseven subunits (ND1, 2, 3, 4, 4L, 5, and 6) of theNADH-dehydrogenase (RC I); one subunit (cy-tochrome b) of RC III; three subunits (CO I, II,and III) of cytochrome c oxidase (RC IV); andtwo subunits (A6 and A8) of the ATP synthase(RC V).All of these proteins are components ofthe ETC or the ATP synthase involved inOXPHOS. In addition to these 13 protein-coding genes, the mtDNA encodes 22 mito-chondrial transfer ribonucleic acids (tRNAs)and two ribosomal RNA (rRNA) molecules (thelarge 16S rRNA and the small 12S rRNA).

As can be seen in Table 8-1, the respiratorycomplexes of the OXPHOS system are multi-subunit complexes, all except RC II havingsubunits encoded by both nuclear and mito-chondrial genes.

Mutations in the nuclear genes that specifymitochondrial proteins can lead to mitochond-rial disorders that obey Mendelian genetics. For


Figure 8-3. The components of oxidative phosphorylation located in the inner mem-brane of the mitochondria. ADP, adenosine diphosphate; ATP, adenosine triphosphate;NADH, nicotinamide adenine dinucleotide.

For patients with MELAS and other mito-chondrial disorders, this reaction can result in

example, some mitochondrial disorders result-ing from large and or multiple deletions inmtDNA, such as Kearns-Sayre syndrome andchronic progressive external ophthalmoplegia,are found in some cases to be autosomal domi-

nant, indicating a nuclear-encoded gene as thedefect responsible for the production of deletedmtDNA. On the other hand, a number of mito-chondrial disorders resulting from mutations inmtDNA will usually appear sporadic or follow

MELAS 95

Figure 8-4. The effect of the MELAS mutation on mitochondrial function is to interferewith function of respiratory Complex I or I and IV, leading to increased levels of NADHand thus also of lactate. CoQ, coenzyme Q; MELAS, mitochondrial encephalomyopathy,lactic acidosis, and strokelike episodes; RC, respiratory complex.

Figure 8-5. The human mtDNA molecule (16,569 bp) showing the 13 protein-codinggenes, 2 rRNA genes, and the 22 tRNA genes as well as the D-loop region containing theheavy strand and light strand promoters PH and PL, respectively, and the heavy strand ori-gin of replication OH.Also shown is the light strand origin of replication OL and the lo-cation of the A3243G MELAS (mitochondrial encephalomyopathy, lactic acidosis, andstrokelike episodes) mutation within the transfer RNALeu gene. Cyt, cytochrome.

matrilinear inheritance. A typical pedigree of afamily with MELAS is shown in Figure 8-6. Notethe strict maternal inheritance. Only offspringof affected mothers are themselves affected;none of the affected males pass on their mito-chondrial disease.

Maternal inheritance is one of several inter-esting features of mitochondrial genetics.Thereare three other aspects of mitochondrial inheri-tance that are important.The first is replicativesegregation leading to heteroplasmy (the pres-ence of both mutant and wild-type mtDNA mol-ecules within a cell or tissue). During celldivision, mitochondria distribute to daughtercells according to their position in the cell dur-ing cytokinesis. Thus, if mutant mitochondriaare concentrated in a particular area of the cy-toplasm, then these mutant mitochondria maylocate as a group in one of the daughter cells.Thus, a parental cell containing only 10% mu-tant mitochondria may give rise to a daughtercell with 20% mutant mitochondria after a sin-gle cell division. In this way, the mutant load ina muscle or brain cell may steadily increase dur-ing the natural aging process.

The second is the threshold effect describingthe different degrees of dependency of varioustissues on OXPHOS. For example, the brain and

CNS have the highest demand of all the cells inthe body for OXPHOS to provide ATP for nor-mal function. Cardiac and skeletal muscle havethe next highest need for OXPHOS, while theendocrine system and skin have lesser require-ments for OXPHOS. Thus, as mitochondrialfunction fails due to deleterious mutations inmtDNA, often the earliest manifestations ofsuch mitochondrial distress is seen in neurolog-ical and muscular abnormalities.

The third aspect is the high error rate andlack of efficient DNA repair systems, leading toa high mutation rate for mtDNA. It is estimatedthat mtDNA is 7–17 times more prone to muta-tion than somatic DNA. These aspects of mito-chondrial biogenesis lead to the production ofmutant mtDNA, contributing to the overall de-gree of heteroplasmy in human tissues.

MELAS is a mitochondrial genetic disease.Atleast 10 different causative point mutationshave been described in mtDNA, 5 of which arelocated in the same gene, the tRNALeu (UUR)

gene. The most common mutation, initially re-ported in 1990 by Goto and coworkers (Gotoet al., 1990) and found in 80% of individualswith MELAS, is an A → G transition at nu-cleotide pair (np) 3243 (A3243G) in the dihy-drouridine arm of the tRNALeu (UUR) gene. Anadditional 7.5% have a heteroplasmic T → Cpoint mutation at np 3271 (T3271C) in the ter-minal nucleotide pair of the anticodon stem ofthe tRNALeu (UUR) gene. Other mutations in thisgene known to be associated with MELAS areA3252G, C3256T, and T3291C. In addition, thereare several other mutations in this gene that arelinked to other mitochondrial diseases havingcharacteristics different from the group defin-ing the MELAS syndrome.

Since both normal and abnormal mitochon-dria may be present in tissues (heteroplasmy),


Figure 8-6. Four generations of a family with the mitochondrial disorder MELAS (mito-chondrial encephalomyopathy, lactic acidosis, and strokelike episodes). Note that onlyfemale affected members will pass on the mutation to their offspring.

Table 8–1. Nuclear and Mitochondrial DNA-Encoded Subunits of the OXPHOS ComplexesRespiratory Nuclear MitochondrialComplex DNA DNA

I 39 7

II 4 0

III 10 1

IV 10 3

V 14 2

the clinical presentation of MELAS can be het-erogeneous. Measurements of respiratory en-zyme activities in intact mitochondria haverevealed that more than one half of the patientswith MELAS may have Complex I or ComplexI + IV deficiency. The occurrence of strokelikeepisodes, sensorineural hearing loss, pigmen-tary retinal degeneration, heart involvement,and recurrent vomiting or abdominal pain canvary significantly among individual cases.

In the United States, as in the worldwidepopulation, the frequency of the A3243G MELASmutation is 16.3 per 100,000. Males and femalesare equally affected. This high prevalencesuggests that mitochondrial disorders may con-stitute one of the largest diagnostic categoriesof neurogenetic diseases among adults.

The presence of this most abundant MELAS-associated mutation, the A3243G mutation, canbe quickly determined using a combined PCRamplification of mtDNA followed by treatmentof the PCR product with the restriction enzymeApa I.As seen in Figure 8-7, a 372-base pair (bp)PCR product would be generated from this re-gion of the mtDNA molecule. The cleavage ofthis 372-bp product by Apa I (which recognizesand cleaves at the sequence 5'-GGGCCC-3') into263- and 109-bp fragments would indicate thepresence of the A3243G mutation.The intensityof the DNA bands can be quantitated and re-lated to the proportionate amounts of wild-typeand mutant mtDNA present in the sample.Thisassay, therefore,not only detects the presence ofthe mutation but also measures the degree ofheteroplasmy.This is very useful because oftenthe level of heteroplasmy will correlate with theseverity of the symptoms observed and can bepredictive of the effective response to treatmentprotocols. All of the family members tested inFigure 8-7 possessed the A3243G mutation.Thesample from the proband had the highest pro-portion of mutant mtDNA, 71%.

Mutations in tRNALeu (UUR) may be expectedto have an important effect on protein synthesisin mitochondria. The MELAS-associated humanmitochondrial tRNALeu (UUR) mutation causesaminoacylation deficiency and a concomitantdefect in translation initiation (Borner et al.,2000). The expected result would be a deficit ingeneral mitochondrial protein synthesis with re-sulting deficiencies of multiple OXPHOS com-ponents. Patients with MELAS disorder havebeen found to have a marked decrease in the ac-tivity of respiratory Complex I with a secondaryreduction in the activity of Complex IV.

It is interesting to note the high number ofmutations in this region of the mtDNA. There isa further interesting aspect of this particular re-gion of the mtDNA that might explain the ex-traordinary number of pathologic mutationsconcentrated in this length of DNA sequence.In addition to encoding the tRNALeu(UUR) gene,this region of mtDNA codes for a signal respon-sible for terminating transcription through thetwo rRNA genes. As seen in Figure 8-8, thetRNALeu(UUR) gene follows immediately afterthe 16S large rRNA gene.

Although transcription of the heavy strand ofmtDNA initiates from the heavy strand pro-moter (PH) and proceeds completely aroundthe circular mtDNA molecule, this full-lengthtranscript is only made about 10% of the time.The other 90% of the time a truncated tran-script from the heavy strand promoter termi-nates after the 16S rRNA gene. How doesthis happen? The nucleotides marked with anasterisk in Figure 8-8 form a transcriptiontermination signal that, when bound by a mito-chondrial termination factor, causes the mito-chondrial RNA polymerase to dissociate fromthe mtDNA and cease transcription. It is pro-posed that alterations in the DNA sequence ofthis tRNA gene therefore not only may affectthe role of the tRNA in protein synthesis but

MELAS 97

Figure 8-7. MELAS (mitochondrial encephalomy-opathy, lactic acidosis, and strokelike episodes) as-say using restriction enzyme Apa I to detect thepresence of mutant mtDNA in skeletal musclebiopsies. GM, grandmother; M, mother; P, proband;U, maternal uncle. The uncut PCR product is372 bp. The presence of the MELAS mutation re-sults in cleavage into 263- and 109-bp fragments.

also may interfere with proper metabolism ofmitochondrial RNA.

THERAPY

Although there is no cure for MELAS or any ofthe mitochondrial diseases, palliative treat-ments are available while others are being de-veloped to alleviate symptoms and to slow theprogression of the disease (Shoffner and Wal-lace, 1994). In general, the responses to treat-ment seem to be better for those patients withless-severe disease. Modifying the diet, addingvitamins and other supplements, and minimiz-ing or avoiding external stresses are importantapproaches to managing the patient withMELAS. Although the long-term effects of di-etary manipulations are unknown, it is impor-tant for the MELAS patient to maintain a goodnutritional status, to remain properly hydrated,and to avoid fasting. Treatment must be tailoredto each patient as the effectiveness of thesetreatments varies on an individual basis.

The goal of the metabolic therapies is to in-crease the production of ATP and interrupt

the progression of this mitochondrial disorder.Metabolic therapies used for the managementof MELAS include: CoQ10, idebenone, dichloroac-etate (DCA), carnitine, menadione, phyllo-quinone, ascorbate, riboflavin, nicotinamide,succinate, and creatine monohydrate. A lack oflong-term follow-up has hampered evaluationof these drug therapies.

The raison d’être for each of these metabolictherapies varies. CoQ10 is a naturally occurringsubstance that stabilizes respiratory complexeslocated in the mitochondrial inner membrane.It also scavenges free radicals and acts as a po-tent antioxidant. Treatment with CoQ10 hasbeen reported to reduce serum lactate levelsand to improve muscle weakness in somepatients with MELAS. Idebenone (an analogueof coenzyme Q10 capable of crossing the blood-brain barrier) may stimulate OXPHOS functionin the mitochondria of brain cells as well as in-crease cerebral metabolism. It also stimulatesmitochondrial respiratory activity by protectingmitochondrial membranes from lipid peroxi-dation. However, CoQ10 has been shown tomediate mitochondrial uncoupling through theproduction of superoxide free radicals, so pro-longed exposure to elevated concentrations ofCoQ10 or idebenone may in fact lead to in-creased free-radical damage and compromisedmitochondrial function.

DCA is an inhibitor of pyruvate dehydrogenasekinase, the enzyme that regulates pyruvate dehy-drogenase (PDH) activity by phosphorylation ofthe E1 subunit, leading to a decrease in PDH activ-ity (see Chapter 7). DCA has been used to lowerlactate levels in some patients with MELAS.Stimu-lation of PDH activity reduces the release oflactate from peripheral tissues and enhances itsmetabolism by the liver. Since prolonged use ofDCA may lead to sensory neuropathy, treatmentmust be closely monitored.

Supplementing the diet with carnitine maystimulate the uptake of long-chain fatty acidsinto affected cells. Fatty acid oxidation withinthe mitochondrion will generate acetyl-CoA,which when oxidized in the TCA cycle will pro-duce NADH and FADH2 to feed into the ETC.Carnitine also shuttles potentially toxic fattyacid catabolic by-products out of the mitochon-drial matrix to the kidney for excretion in theurine.

Succinate is an intermediate formed duringthe metabolism of acetyl-CoA that has enteredthe TCA cycle. This intermediate can donate


G

GG

G

3243

GC

C C

A A AA

A

TT

T

A

T

A

C

C

GG

GG

C

C

CC

CC

CCCT TT

T

T

TGGGA A

AA

CGT

TA

TATAATA

A-OH3′

5′

TCG*

****************

**

** * * * * * *

A TA TA

A

A

Anticodon

AA

A

TT

T

Figure 8-8. The mitcochondrial tRNALeu(UUR)

gene. The diagram shows the position of theA3243G mutation. The nucleotides marked withan asterisk are involved in termination of tran-scription.

electrons directly to RC II (succinate dehydro-genase). Succinate treatment has been used fora patient with MELAS, who showed significantimprovement, suffering fewer strokelikeepisodes when treated with 6 g/day sodiumsuccinate. Creatine monohydrate also has beenused in some patients, and an increase in mus-cle strength in high-intensity anaerobic and aer-obic activities has been reported. This effectmay be related to increased intracellular crea-tine/phosphocreatine content, which may beinvolved in maintaining cellular ATP and instabilizing the mitochondrial permeability tran-sition pore, with subsequent reduction of neu-ronal death due to apoptosis.

Menadione (vitamin K3),phylloquinone (vita-min K1), and ascorbate (vitamin C) have beenused to donate electrons to cytochrome c.For example, ascorbate is oxidized to dehy-droascorbate as it uses its electrons to reducecytochrome c directly.The dehydroascorbate isquickly reduced to ascorbate in the mitochon-drion by NADH or FADH2. Menadione appearsto improve cellular phosphate metabolism andto enhance electron transfer after a respiratoryComplex I block.

Riboflavin (vitamin B2) has been reported toimprove the exercise capacity of a patient withComplex I deficiency.After conversion to flavinmonophosphate and FAD, riboflavin functionsas a cofactor for electron transport in ComplexI, Complex II, and electron transfer flavopro-tein. Nicotinamide has been used becauseComplex I accepts electrons from NADH andultimately transfers electrons to Q10.

Medications such as β-blockers, calciumchannel blockers, digoxin, and amiodaronecan be used to control cardiac conductionabnormalities (arrhythmias), and a pacemakermay be inserted to combat heart failure. Thegeneral supportive care measures used inacute stroke syndromes also should be fol-lowed. Death in patients with MELAS is usuallythe result of cardiac failure, pulmonary embo-lus, or renal failure.

Status epilepticus can occasionally be fatal;therefore, seizures should be treated aggres-sively. Seizures can be managed with antiepilep-tic medications such as carbamazepine. Thecommon antiepileptic drug valproic acid is con-traindicated because it depletes the body of car-nitine, which may in fact exacerbate symptoms.The potential risk of stroke may be reduced us-ing appropriate medication. The complications

of diabetes can often be managed through di-etary and nutritional modification and by ap-propriate medications.

In patients with MELAS and other mitochon-drial disorders,moderate treadmill training mayresult in improvement of aerobic capacity anda drop in resting and postexercise lactate lev-els. Concentric exercise training (by whichmuscle cells shorten to exert force on an ob-ject, as in lifting a weight) also may play an im-portant role since after a short concentricexercise training, a remarkable increase report-edly occurs in the ratio of wild-type to mutantmtDNAs and in the proportion of muscle fiberswith normal respiratory chain activity (Taivas-salo et al.,1999). If conditions such as cardiomy-opathy (a structural or functional disease ofheart muscle marked especially by hypertrophyof cardiac muscle) are present, then exerciseshould be restricted. A mild degree of aerobicactivity may lead to an improvement of aerobiccapacity; however, strenuous exercise may leadto rhabdomyolysis (the destruction of skeletalmuscle tissue accompanied by the release ofmuscle cell contents into the bloodstream).

QUESTIONS

1. How can you explain the observationthat, even within the same family, pa-tients with the A3243G mutation maymanifest very different severity of symp-toms?

2. Why would a physician recommendavoiding fasting or to eat multiple smallermeals during the day rather than the tradi-tional three meals?

3. Some pediatric patients with MELAS reactbadly when they are immunized againstcommon childhood diseases. Can you sug-gest a reason why vaccination may exacer-bate MELAS symptoms?

4. Which of the respiratory complexeswould you expect to be least affected bymutations in mtDNA?

5. Although many mitochondrial diseases,such as MELAS, will involve multiple or-gans, neurological or neuromuscular ab-normalities are often the earliest signs ofdisease.Why?

6. Why does the A3243G mutation associ-ated with MELAS lead to reduced activityof respiratory complexes I and IV, while

MELAS 99

the point mutation at 11778 (within theND4 gene) associated with Leber’s heredi-tary optic neuropathy affects only respira-tory complex I?

7. Vitamin C (ascorbate) enhances the ab-sorption of iron from the intestines. Forthis reason, it is recommended that vita-min C not be administered to a patientwith MELAS immediately prior to or sub-sequent to a meal of red meat or otherfoods rich in iron.What is the biochemicalbasis for the concern about increased ironavailability to a patient with MELAS?

BIBLIOGRAPHY

Borner GV, Zeviani M, Tiranti V, et al.: Decreasedaminoacylation of mutant tRNAs in MELAS butnot in MERRF patients. Hum Mol Genet9:467–475, 2000.

Goto Y, Nonaka I, Horai S: A mutation in thetRNA(Leu)(UUR) gene associated with the

MELAS subgroup of mitochondrial encephalomy-opathies. Nature 348:651–653, 1990.

Howell N: Human mitochondrial diseases: answer-ing questions and questioning answers. Int RevCytol 186:49–116, 1999.

Lestienne P,Ponsot G:Kearns-Sayre syndrome withmuscle mitochondrial DNA deletion. Lancet1:885, 1988.

Luft R, Ikkos D, Palmieri G, et al.: A case of severehypermetabolism of nonthyroid origin with adefect in the maintenance of mitochondrial res-piratory control: a correlated clinical, biochemi-cal, and morphological study. J Clin Invest41:1776–1804, 1962.

Shoffner JM, Wallace DC: Oxidative phosphoryla-tion diseases and mitochondrial DNA muta-tions: diagnosis and treatment. Annu Rev Nutr14:535–568, 1994.

Taivassalo T, Fu K, Johns T, et al.: Gene shifting: anovel therapy for mitochondrial myopathy.Hum Mol Genet 8:1047–1052, 1999.

Wallace DC, Singh G, Lott MT, et al.: DNA mutationassociated with Leber’s hereditary optic neu-ropathy. Science 242:1427–1430, 1988.


CASE REPORT

A 3.5-year-old boy presented to the emergencyroom with hypoglycemia. The patient’s historywas remarkable for multiple episodes of alteredconsciousness resulting in hospital admissionssince age 3 months. These episodes were char-acterized by depressed mental status or comaand hypoglycemia (blood glucose typically15 mg/dL, normal 75–105 mg/dL). Seizure activ-ity was noted on several admissions. On evalua-tion, the patient was consistently noted to haveproximal muscle weakness, hepatomegaly, con-gestive heart failure,elevations of serum transam-inases (aspartate aminotransferase [AST], andalanine aminotransferase [ALT]) to greater than2000 IU/L (normal values 10–30 IU/L and 10–40IU/L for AST and ALT, respectively) and elevatedserum creatine phosphokinase to greater than1500 IU/L (normal <175 IU/L).

The patient was born to nonconsanguineousparents from Chihuahua, Mexico, after an un-complicated pregnancy and delivery.A brotherhad died at 3 months of age of a “liver problem”after an unexplained coma. The parents andthree other siblings were well.

Developmental history was remarkable fordelayed milestones and growth retardation(weight and height below the third percentile).

Previous diagnostic evaluation had included anelectroencephalogram, brain scan, and chromo-somal studies, which were unremarkable. Alsonormal were the plasma levels of electrolytes,calcium, phosphorus, magnesium, bilirubin, thy-roxine, thyroid-stimulating hormone,and growth

9

Systemic Carnitine Deficiency:A Treatable Disorder

ERIC P. BRASS, HARBHAJAN S. PAUL, and GAIL SEKAS

101

hormone.Results of total serum protein determi-nation, serum electrophoresis, cerebrospinalfluid studies, and studies of immune functionwere normal. Between acute episodes, bloodsugar, ammonia, ALT, AST, and creatine phospho-kinase were all normal. A urinary organic acidprofile was nondiagnostic.A muscle biopsy hadbeen recently performed and microscopic exam-ination revealed large amounts of neutral lipids.A liver biopsy also showed severe but nonspe-cific fatty changes.

As part of the evaluation during the currenthospital admission, a 32-hour fasting study wasperformed. During the fasting period, the pa-tient had no nausea or cramps. At the start of thefast, his blood glucose was 91 mg/dL, but by 24hours it had fallen to 66 mg/dL and at 32 hourshad fallen to 35 mg/dL. The fast was terminatedat this point, and the patient was given glucose.During the fast, plasma triglyceride levels rosefrom 66 to 126 mg/dL, and free fatty acids in-creased from 0.1 to 2.0 mEq/L. Serum ALT rosefrom 36 to 1450 IU/L,with no increase in serumcreatine phosphokinase or aldolase activities.Remarkably, acetoacetate and β-hydroxybutyrateconcentrations did not increase during the fast.The patient’s prefast serum carnitine concentra-tion was markedly reduced (4.8 µmol/L vs.control range 30–45 µmol/L). Carnitine mea-surements were performed on the previouslyobtained liver and muscle biopsies and alsoshowed low carnitine content (Table 9-1).

The patient was placed on a high-carbohydrate, low-fat diet and was treated withoral L-carnitine (2.0 g/day). After 3 months of

treatment, the patient was symptomaticallyimproved. While receiving carnitine therapy,laboratory studies showed an increase in theserum carnitine concentration to 18.5 µmol/Land a dramatic increase in urinary carnitine ex-cretion (Table 9-1).

DIAGNOSIS

The patient’s history strongly suggested a sys-temic disease based on the multisystem involve-ment (liver, heart, and skeletal muscle). Theunknown disease in the patient’s brother sug-gested the possibility of an inherited disease.Therecurrent episodes of hypoglycemia suggested ametabolic basis for the clinical manifestations.

The fasting study that was performed pro-vided a functional evaluation of the patient’sintegrative metabolic regulation. The develop-ment of hypoglycemia without increases inacetoacetate or β-hydroxybutyrate concentra-tions, pointed to a defect in hepatic fatty acidoxidation. Although, as discussed in the nextsection, there are many potential etiologies forimpaired hepatic fatty acid oxidation, the ex-tremely low serum, liver, and muscle carnitineconcentrations strongly suggested carnitine de-ficiency as the primary etiology.

Carnitine is present in biological systems asboth carnitine and acylcarnitines generated intissues (see next section). Carnitine deficiencymay be a primary defect due to a genetic de-fect in carnitine transport systems or may besecondary to other metabolic derangements.Normal carnitine homeostasis requires reab-sorption of carnitine in the renal tubule via aspecific transport protein. This same transportprotein is responsible for the accumulation ofcarnitine in heart and skeletal muscle. If thistransport system is not functional, then carni-tine cannot reach tissues, and primary carnitine

deficiency results. Secondary carnitine defi-ciency results from the abnormal accumula-tion of acylcarnitines due to the accretion ofmetabolic intermediates associated with anunderlying metabolic defect (for example,propionyl-CoA accumulation in propionicacidemia). Thus, in secondary carnitine defi-ciency, the low carnitine concentration is asso-ciated with a relative increase in acylcarnitines.Also, urinary organic acid analysis will typicallyresult in diagnosis of the primary underlyingmetabolic disease.

The approach to diagnosis of a patient withsuspected carnitine deficiency should be sys-tematic and include

1. Case history: Frequency of illness,provocative factors, spectrum of sympto-matology

2. Family history: Complete four-generationpedigree with all diseases noted

3. Physical examination: Cardiac, neurologi-cal, muscle strength, liver

4. Laboratory studies: Serum transaminases,blood glucose, plasma and urine carnitineand acylcarnitines, urinary organic acids

5. Metabolic studies: Response to fasting(glucose, free fatty acids, acetoacetate, β-hydroxybutyrate), response to medium-chain triglycerides

6. Tissue biopsy: Light and electron mi-croscopy, tissue carnitine content, othermetabolic enzymes as indicated

7. Genetic studies: Culture of fibroblasts forpotential genetic and function studies, ge-netic testing for specific mutations

In obtaining the clinical history, particularattention should be paid to situations that mayprovoke episodic symptoms. Skipping meals,mild viral illnesses, and exercise are examplesof metabolic stressors that may result in symp-tomatic decompensation. Importantly, results of


Table 9-1. Patient Carnitine Levels in Muscle, Liver, Serum, and Urine*Control Patient Before Patient After

Analysis Values Therapy Therapy

Muscle total carnitine 2.5 0.02 0.2(mmol/kg wet weight)

Liver total carnitine 0.9 0.04 0.5(mmol/kg wet weight)

Serum carnitine (µmol/L) 33 4.8 18.5

Urine carnitine (µmol/24 hours) 100 40 1500

*Values are not from a single case but represent a composite from varied case reports.

each step of the evaluation should guide thespecifics of subsequent steps to ensure efficientand definitive evaluation.


Fatty acid oxidation is a multistep process re-quiring orchestration of reactions in the cyto-plasm and mitochondria (Fig. 9-1). Free fattyacids enter the cell and are activated to theircoenzyme A (CoA) thioesters in the reactioncatalyzed by fatty acyl-CoA synthetase:

Fatty acid + ATP + CoASH→ Fatty acyl-CoA + AMP + PPi

The mitochondrial inner membrane is imperme-able to fatty acyl-CoA. To be transported into mi-tochondria, the acyl moiety must first beesterified to L-carnitine (hereafter referred tosimply as carnitine) to form the correspondingfatty acylcarnitine. This reaction is catalyzed bycarnitine palmitoyltransferase I (CPT I) localized

on the outer surface of the mitochondrial innermembrane (Fig. 9-2):

Fatty acyl-CoA + Carnitine

� Fatty acylcarnitine + CoASH

A specific transport protein, the carnitine-acylcarnitine translocase, moves the fatty acyl-carnitine into the mitochondrial matrix whilereturning carnitine from the matrix to the cyto-plasm. Once inside the mitochondria, anotherenzyme, carnitine palmitoyltransferase II (CPTII), located on the matrix side of the mitochon-drial inner membrane, catalyzes the reconver-sion of fatty acylcarnitine to fatty acyl-CoA.Intramitochondrial fatty acyl-CoA then under-goes β-oxidation to generate acetyl-CoA.Acetyl-CoA can enter the Kreb’s cycle for completeoxidation or, in the liver, be used for the synthe-sis of acetoacetate and β-hydroxybutyrate (ke-tone bodies).

Dysfunction of any of the steps in Figure 9-1,or in the enzymes of β-oxidation, can resultin impaired fatty acid oxidation and clinical

�

System Carnitine Deficiency 103

Figure 9-1. Role of carnitine in fatty acid oxidation. Long-chain fatty acids are activatedas the thioester of CoA on the cytoplasmic side of the mitochondrial membrane. The fattyacyl group is then transferred to form the corresponding carnitine ester in a reaction cat-alyzed by carnitine palmitoyltransferase I (CPT I). The acylcarnitine then enters the mito-chondrial matrix in exchange for carnitine via the carnitine-acylcarnitine translocase. Theacyl group is transferred back to CoA in the matrix by carnitine palmitoyltransferase II(CPT II). The intramitochondrial acyl-CoA can then undergo β-oxidation.

disease. Interestingly, the transmitochondrialtransport system depicted in Figure 9-1 is onlyrequired for the complete oxidation of long-chain fatty acids (12 carbon atoms or longer).Shorter-chain fatty acids are transported into mi-tochondria, can be directly activated to the acyl-CoA within the mitochondrial matrix, and thencan undergo β-oxidation. Thus, the ability todifferentially oxidize long-chain versus shorterfatty acids can be used in vitro or as a diagnos-tic test clinically to identify defects in thecarnitine-dependent reactions of fatty acid oxi-dation.

It is clear from this discussion that carnitineis required in humans for the oxidation oflong-chain fatty acids. In humans, carnitine isderived from both dietary sources and endoge-nous biosynthesis. Meat products, particularlyred meats, and dairy products are important di-etary sources of carnitine. Since biosynthesiscan meet all physiological requirements, carni-tine is not an essential nutrient. Premature in-fants are an exception to this rule as they lacka mature biosynthetic system and have limitedtissue carnitine stores. As many infant formu-las, particularly those based on soy protein, arelow in carnitine, premature infants receiving asignificant part of their nutrition from suchformulas may be susceptible to carnitine defi-ciency.

Carnitine biosynthesis utilizes the essentialamino acid lysine, with terminal methyl groupsdonated by S-adenosylmethionine. Only lysineincorporated into proteins is a substrate for themethylation reaction. In humans, the final reac-tion in the biosynthetic pathway, catalyzed by acytosolic hydroxylase, occurs in liver and kid-ney but not in cardiac or skeletal muscle. Thecarnitine requirement of these tissues is met bycarnitine transported to them via the plasma

from sites of biosynthesis or after intestinal ab-sorption of dietary carnitine. Carnitine is accu-mulated in muscle against a concentrationgradient with normal carnitine concentrationsin skeletal muscle approximately 100-fold thosein plasma. To prevent renal losses of carnitineduring this trans-tissue transportation, 95% ofthe carnitine filtered by the kidney is reab-sorbed into the plasma by the renal tubule. Themembrane transport of carnitine into both mus-cle and the renal tubule is mediated by asodium-dependent transport protein, organiccation transporter 2 (OCTN-2).OCTN-2 is essen-tial for normal carnitine homeostasis. If OCTN-2is defective, then the kidney will be unable to re-absorb the carnitine delivered to it. Thus, thebulk of any carnitine synthesized or obtainedfrom dietary sources will simply be lost in theurine, resulting in systemic carnitine deficiency.Any residual carnitine in the plasma will be in-effectively taken up by critical tissues such as theheart and skeletal muscle. Mutations in OCTN-2are responsible for clinical syndromes of primarycarnitine deficiency such as the case presentedin this chapter.

The clinical sequelae of OCTN-2 dysfunctionand carnitine deficiency are readily predictedbased on the biochemistry discussed here. Fattyacids are a major source of ATP in cardiac andskeletal muscle. Thus, carnitine deficiency willresult in cardiac impairment (cardiomyopathy)and generalized skeletal muscle weakness. Un-der conditions of stress, skeletal muscle integritymay be compromised, resulting in significantmuscle injury (rhabdomyolysis).Between meals,fatty acids are also a major source of energy forthe liver. Under these conditions, fatty acid oxi-dation is required to support the critical func-tions of the liver, including gluconeogenesis andureagenesis.


Figure 9-2. Carnitine palmitoyltransferase reaction. Palmitoyl-CoA is shown as a proto-typic substrate. Carnitine palmitoyltransferase I (CPT I) and carnitine palmitoyltrans-ferase II (CPT II) are shown illustrating the direction of the reaction catalyzed by eachenzyme during physiological fatty acid oxidation.

During starvation, the liver also uses acetyl-CoA generated from fatty acid oxidation tosynthesize ketone bodies as a fuel source forthe central nervous system. Thus, hypoglycemiamay be a prominent clinical sign as peripheraltissues use glucose in the absence of the abilityto burn fat, and the liver is unable to synthesizenew glucose to meet the demand.The systemicdependence on fatty acid oxidation is greatestduring periods of caloric deprivation, and de-layed feedings are often a cause of episodicsymptoms in patients with carnitine deficiency.CNS symptoms, including coma or seizures,may result from the hypoglycemia, absence ofketone bodies, and hyperammonemia (resultingfrom impaired ureagenesis). These manifes-tations of systemic carnitine deficiency areusually expressed early in life.As fatty acids de-livered to tissues cannot be oxidized, they areesterified to form triglycerides, which accumu-late in tissues and are evident on light mi-croscopy of biopsied tissue. The case presentedin this chapter demonstrates many of these clin-ical features, including an abnormal response toa controlled diagnostic trial of fasting.

THERAPY

Understanding the biochemical bases for carni-tine deficiency and the resulting clinical syn-drome forms the foundation for the two majorapproaches to therapy: avoidance of biochemi-cal stressors and carnitine supplementation.

Biochemical stress can be minimized by us-ing frequent feedings to minimize dependenceon fatty acid oxidation, particularly for the liver.Meals should have a high-carbohydrate, low-fatcontent. Medium-chain triglycerides (syntheticor derived from coconut or palm kernel oils)can be used as these lipids can be oxidized in-dependent of carnitine. These steps are particu-larly important when any external metabolicstress, such as a viral illness, is present.

Carnitine supplementation is also critical inthese patients. In acute severe metabolic crises,carnitine can be administered intravenously.Oral carnitine supplementation can be used forchronic maintenance therapy. Note that as theprimary transport defect will still be present,carnitine supplementation cannot normalizethe patient’s carnitine homeostasis.

In the present case, as soon as the diagnosisof carnitine deficiency was considered, the pa-tient was given a high-carbohydrate, low-fat diet

containing 20% of the calories as protein. In ad-dition, the patient was given a total of 2.0 g ofcarnitine per day, which was administered inthree equal doses. Carnitine therapy resulted indramatic clinical and biochemical changes.

After 3 months of therapy (Table 9-1), serumlevels of carnitine, though still lower than thecontrols, had increased markedly from thepretherapy levels. Muscle carnitine levels also in-creased but remained well below normal. De-spite this, clinical muscle strength and tone wereremarkably improved.Carnitine levels in the liver(where carnitine transport is not dependent onOCTN-2) increased more dramatically than thosein muscle. Despite the low serum carnitine con-centration, urinary carnitine losses were dramati-cally increased in the patient while on therapy ascompared with controls.This reflects the contin-ued dysfunction of the renal OCTN-2 and thusthe continued urinary wasting of carnitine.

In addition to the biochemical changes,the patient experienced dramatic clinical im-provements. Hepatomegaly was no longer evi-dent after 2 weeks of carnitine therapy, andhis transaminase levels returned to normal.Neurological disturbances, including hyper-reflexia of the lower extremities, bilateral Babin-ski responses, and walking disability due toweakness and ataxia disappeared after 2 weeksof therapy. His muscle strength increased from2/5 to 4/5 and he could run for the first time inhis life. Sensory nerve conduction in the leftperoneal nerve, virtually unmeasurable beforetherapy, was normal. Somatosensory evoked po-tentials, strikingly abnormal before therapy,showed marked improvement after 3 months oftreatment. The cardiomyopathy was largely re-solved.

After establishing the diagnosis and institut-ing carnitine therapy, the prognosis in suchcases is generally quite favorable.Cases of carni-tine deficiency have been reported since themid-1980s, and limited long-term follow up ofthese patients is encouraging.

QUESTIONS

1. What are the reasons for lipid accumula-tion in the liver of this patient?

2. Why was the patient unable to produceketone bodies after 24 hours of fastingprior to carnitine therapy?

3. What is the cause of hypoglycemia ob-served in this patient?

System Carnitine Deficiency 105

4. Why did skeletal muscle carnitine contentremain below normal despite high-dosecarnitine therapy in this patient?

5. What is the biochemical basis for the dif-ferential handling of medium-chain versuslong-chain fatty acids in liver?

BIBLIOGRAPHY

Bremer J: Carnitine—metabolism and function.Physiol Rev 63:1420–1480, 1983.

Cederbaum SD, Koo-McCoy S, Tein I, et al.: Carni-tine membrane transport deficiency: a long-termfollow up and OCTN2 mutation in the first docu-mented case of primary carnitine deficiency.MolGen Metab 77:195–201, 2002.

Kerner J, Hoppel C: Genetic disorders of carnitinemetabolism and their nutritional management.Annu Rev Nutr 18:179–206, 1998.

Rebouche CJ, Seim H:Carnitine metabolism and itsregulation in microorganisms and mammals.Annu Rev Nutr 18: 39–61, 1998.

Stanley CA,Treem WR, Hale D, et al.: A genetic de-fect in carnitine transport causing primary car-nitine deficiency. Prog Clin Biol Res 321:457–464, 1990.

Tang NLS, Ganapathy V, Wu X, et al.: Mutations ofOCTN2, an organic cation/carnitine transporter,lead to deficient cellular carnitine uptake in pri-mary carnitine deficiency. Hum Mol Genet8:655–660, 1999.

Treem WR, Stanley CA, Finegold DN, et al.: Primarycarnitine deficiency due to a failure of carnitinetransport in kidney, muscle and fibroblasts. NEngl J Med 319:1331–1336, 1988.

Wang Y,Ye J, Ganapathy V, et al.: Mutations in theorganic cation/carnitine transporter OCTN2 inprimary carnitine deficiency. Proc Natl Acad SciU.S.A. 96:2356–2360, 1999.


CASE REPORT

Female infant L. weighed 3.98 kg at birth to a29-year-old mother after a 35-week pregnancy.Mrs. L. was an insulin-requiring diabetic whohad had type 1 diabetes since she was 15 yearsold and had not yet shown signs of vascular dis-ease. Her pregnancy had been complicated bymultiple episodes of hyperglycemia (excesssugar in the blood) and glycosuria (presence ofabnormal amounts of sugar in the urine). Herprenatal care was only episodic because herhusband had been part of a layoff at work, andthe family was without health insurance.

On the day of delivery, Mrs. L. had a routinenonstress test (a test that records changes in fe-tal heart rate with fetal movement and is usedto assess whether the fetus is suffering from alack of oxygen or blood flow to the brain) be-cause she noticed that her fetus had decreasedmovement over the preceding 2 days. The non-stress test showed a flat baseline heart ratewithout any variation, a finding suggesting thatthe fetus was at risk for asphyxia ( lack of oxy-gen leading to an accumulation of lactic acidand carbon dioxide in the blood).Following therupture of membranes, a blood sample takenfrom the fetal scalp revealed an acidotic pH of6.9 (normal > 7.3). Because these findings wereclear evidence of fetal distress, an immediate ce-sarean section was performed.

The infant’s Apgar scores (scale 0–10, with10 being best) of 2 at 1 minute and 6 at 5 min-utes suggested depression of neurological andcardiopulmonary function. The physicians

10

Neonatal Hypoglycemia and the Importance of Gluconeogenesis

IAN R. HOLZMAN and J. ROSS MILLEY

107

resuscitated the infant, who was then taken tothe neonatal intensive care unit. She had a nor-mal pulse, respiratory rate, and rectal tempera-ture.The infant’s head circumference and lengthwere both normal for her gestational age; how-ever, she was clearly macrosomic ( large bodyand organs) and plethoric (ruddy with an in-creased RBC mass) and had a protuberant, firmabdomen and a disproportionate amount of adi-pose tissue around the upper body. The liveredge was 4 to 5 cm below the right costal mar-gin; the kidneys were large bilaterally but ofnormal shape and position, and the initial car-diac examination revealed no abnormalities.The pulses were symmetric but difficult to feelin all extremities. There were no other congeni-tal anomalies.

When the infant was aged 30 minutes, thehematocrit was 0.63 (normal is 0.45–0.62), andthe serum glucose was 2.5 mmol/L (normal is2.8–3.3 mmol/L). An intravenous line was es-tablished.The initial intravenous fluid was 10%dextrose in water and was administered at arate of 10 mL/hour or a total of 60 mL/kg/day(4 mg/kg/minute of glucose). Several minuteslater, the infant’s blood glucose concentrationwas 1.2 mmol/L, and 10 mL of 25% dextrose inwater was given via intravenously over a 5-minute period (125 mg/kg/minute).

The first arterial blood gas determination at45 minutes of life revealed a pH of 7.18 (normalis 7.30–7.35), a PO2 of 6.3 kPa (normal is6.7–9.3 kPa), a PCO2 of 6.1 kPa (normal is4.7–6.0 kPa), and a bicarbonate concentrationof 17 mmol/L (normal is 18–21 mmol/L), with

a base excess (anion gap; see chapter 7), of(−11 mmol/L.A chest radiographic film showedthat the lungs were clear, and the heart waslarge. One hour and 10 minutes after birth, theinfant’s blood sugar was 21.0 mmol/L (after in-fusion of 10 mL of 25% dextrose noted above).

About 45 minutes later, the blood sugar was1.8 mmol/L, and 8 mL of 10% dextrose wasgiven intravenously over 5 minutes (40 mg/kg/minute).The intravenous infusion was changedto 12.5% dextrose in water at a rate of 13 mL/h(78 mL/kg/day or 6.8 mg/kg/min of glucose).Within 10 to 15 minutes, the serum glucoselevel had fallen to between 0 and 1.4 mmol/L,and an additional 4 mL of 12.5% dextrose wasgiven intravenously. A catheter was then in-serted into the umbilical vein to lie in the lowerpart of the right atrium, and an infusion of 15%dextrose was begun.

By 5 hours of age, this infant was receivingglucose at a rate of 13.5 mg/kg/min. The bloodglucose concentration continued to be lessthan 1.1 mmol/L,and the infant was given 15 mgof hydrocortisone every 6 hours. Echocardiog-raphy showed decreased left ventricular con-tractility and hypertrophy of the ventricularseptum but no other structural abnormalities.These findings were consistent with the car-diomyopathy often seen in infants of diabeticmothers.

By 12 hours of age, the infant had seizure ac-tivity that included rhythmic movement of bothupper extremities, hiccupping, and repetitivechewing movements.An electroencephalogramwas grossly abnormal, showing paroxysmalbursts of activity (seizures) and periods of sup-pression. Phenobarbital was given to alleviatethe seizure activity.

Eventually, the infant’s need for hydrocorti-sone began to decrease so that by the ninthday of life she no longer required steroids, andshe maintained serum glucose concentrationsabove 2.8 mmol/L with a glucose infusion ofless than 7 mg/kg per minute.Subsequently,glu-cose control was no longer difficult, and no fur-ther medications were given for this problem.

Sonar examination of this infant’s head at 1week of age showed an infarction of the rightparieto-occipital area and hemorrhage withinthe right ventricle. This brain damage was man-ifested clinically: Infant L. had great difficultywith sucking and swallowing. She was also dif-fusely hypotonic (decreased muscle tone),withfewer movements of the lower extremitiesthan of the upper extremities. Hearing tests re-

vealed severe bilateral central hearing loss.At age 1 month, the infant was discharged to atransitional care center, where she could re-ceive developmental stimulation, and the par-ents could be educated regarding how to feedher.

DIAGNOSIS

This case exemplifies many of the problems clas-sically associated with infants born to diabeticmothers (Cowett, 1998). Inadequate maternalglucose control during pregnancy generates po-tential metabolic difficulties in the infant (seeBiochemical Perspectives). Meticulous atten-tion to diabetic control during pregnancy hasmarkedly improved the outlook for these in-fants in terms of both morbidity and mortality.Indeed, in many high-risk centers, the outlookfor a diabetes-associated pregnancy is indistin-guishable from that of a normal pregnancy.

The size of this infant at birth (nearly 4 kg in apreterm infant),however,categorizes her as largefor gestational age, a finding characteristic of af-fected infants of diabetic mothers.This infant istherefore likely to have the other manifestationsassociated with this pathophysiology. Indeed,many characteristics of this infant are typical fora child born to a mother with inadequatelytreated diabetes mellitus. Specifically, the macro-somia (large organs), round facies, plethora, andoverall weight, length, and head circumferenceare all consistent with poorly controlled mater-nal diabetes mellitus that has significantly af-fected fetal growth and metabolism.

As the end of gestation approaches, the fetusof a diabetic mother is more likely to be poorlyoxygenated in utero. The markedly decreasedscalp pH and low Apgar scores are convincingevidence that this infant’s supply of oxygenwas insufficient to meet her need.The etiologyof the hypoxemia is not always apparent at thetime of birth.Any mechanical disruption of thefetal-placental circulation (e.g., abruption orseparation, placenta previa or placenta locatedover the cervical os, or cord accidents) couldcompromise fetal oxygen supply, as could afetal infection. In the present case, however,there was no evidence of such problems. Con-sequently, it seems more likely that the fetusexhibited the chronic oxygen deficit thatoccurs in infants of diabetic mothers and be-comes more severe as the pregnancy pro-gresses.


Hypoglycemia beginning soon after birth istypical for infants of diabetic mothers. Indeed,the physicians caring for this child should haveexpected this problem because the infant wasso conspicuously affected by her mother’s dia-betes. Fortunately, the severity of the presentcase, including the need for glucocorticoids,hasbecome rare since the advent of more sophisti-cated maternal care during pregnancy. Seizuresare an unfortunate and serious complicationof hypoglycemia, but it is not possible to de-termine whether the lack of oxygen duringprenatal life, the hypoglycemia, or a combina-tion of both is responsible for the infant’s braindamage.

BIOCHEMICAL PRESPECTIVES

To understand the etiology of hypoglycemia inthe newborn, it is first necessary to know theprinciples of fetal metabolism that form the ba-sis for newborn physiology and pathology. Aunique characteristic of the fetus is the need forthe continual provision of substrates across thematernal and placental circulations for growthand energy. A second unique characteristic ofthe fetus is its biochemical development so thatit no longer needs this constant substrate sup-ply (i.e., is capable of independent existence) atbirth.

The primary metabolic substrates the fetusreceives from the mother are glucose, lactate,amino acids, essential fats, free fatty acids (pri-marily stored and not oxidized), and oxygen.Each of these is required for three ongoing pro-cesses: energy metabolism, growth, and prepa-ration for extrauterine life. The provision ofglucose is especially important because of itscentral role in fetal brain metabolism; cerebralglucose utilization proceeds under most condi-tions at a rate of about 5 mg/min/0.1 kg of tis-sue. The brain of the term human fetus weighsapproximately 0.4 kg and thus would requireapproximately 20 mg/kg per minute of glucosefor brain use alone.

Most of this glucose is needed for cerebraloxidative metabolism, but there is a further re-quirement for the synthesis of macromolecules.Other fetal tissues are less dependent on glu-cose as an oxidative fuel but still require glu-cose for growth. In addition, tissues such asliver, lung, heart, and skeletal muscle use glu-cose to form glycogen, and adipocytes distrib-uted throughout the fetus convert glucose into

lipid (triglycerides). To provide glucose for allthese purposes, the fetus must receive glucoseat a rate of 6 to 8 mg/kg per minute.

What system is in place to ensure that thefetus receives this constant supply of glucose?The transport of glucose from the maternal tothe fetal circulation occurs by facilitated diffu-sion across the placental syncytiotrophoblastlayer from maternal blood to the fetal capillar-ies. The carrier-mediated transport occurs viaa glucose transporter named GLUT1 (one of alarge family of GLUT proteins that are responsi-ble for glucose transport into and out of mam-malian cells). GLUT1 is present in the majorityof fetal cells, and because it has a very high glu-cose affinity, it ensures maximal glucose trans-port to the rapidly growing fetal cells.A secondglucose transporter, GLUT3, is also present inplacental villi, and its concentration decreasesas term approaches. It is primarily found in vas-cular endothelium and may also play a role inglucose transport from mother to fetus (Sim-mons, 2004).

This system of membrane transport equili-brates glucose concentrations across cell mem-branes but will not move sugar against aconcentration gradient; therefore, the amountof glucose received by the fetus depends on theconcentration gradient between the fetus andmother. Consequently, under circumstances ofincreased maternal glucose concentration, thefetomaternal glucose concentration gradientwill increase, and more glucose will be deliv-ered to the fetus.Conversely, if the maternal glu-cose concentration falls, then the fetomaternalglucose concentration gradient will decrease,and the fetus will receive less glucose. In preg-nancies complicated by maternal diabetes, thelarger placentas have an increase in GLUT1 pro-tein, consistent with the increase in glucosetransport to the fetus (Simmons, 2004).

The maternal glucose concentration is ex-tremely critical in determining fetal glucosesupply, and a normal fetal endocrine environ-ment is important in modulating the way the fe-tus uses this supply of glucose. The endocrinemilieu of the fetus provides an anabolic envi-ronment within which the fetus can producecomplex macromolecules from simpler precur-sors.Although the placenta is essentially imper-meable to maternal insulin, the fetal pancreasreleases insulin in response to increases in fetalblood glucose concentration.

The pattern of insulin effects on the fetusmakes sense if one knows which tissues have

Neonatal Hypoglycemia and Gluconeogenesis 109

receptors to insulin and realizes that the effectsof activation of these receptors differ in varioustissues. Insulin is the major anabolic hormonethat affects fetal growth. Because the fetus inlate gestation begins to develop insulin-sensitiveglucose transporters (GLUT4) in both fat andmuscle, one of the major effects of fetal insulinis to regulate the use of glucose by these twotissues.Thus, in muscles, both skeletal and car-diac, higher insulin concentrations promote in-creased glucose uptake and oxidation as well asthe storage of glucose in the form of glycogen.In adipose tissue, insulin promotes lipogenesis.

In addition, there are numerous effects of in-sulin, generally anabolic, not mediated throughits well-known effects on glucose transport.For example, in muscle, insulin decreasesglycogenolysis and proteolysis. Lipolysis is inhib-ited by insulin in adipocytes. The effect of in-sulin on hepatic tissue is to promote increasedglycogen production and protein synthesis andto inhibit glycogenolysis, proteolysis, lipolysis,and gluconeogenesis (Fig.10-1).

Other hormones, notably glucagon and epi-nephrine, are important counterregulators ofthe action of insulin in postnatal life; however,both fetal glucagon and epinephrine concen-trations are normally quite low through fetallife and are even lower in infants of diabeticmothers. Thus, as the action of insulin is pre-dominant, the endocrine milieu of fetal life isuniquely anabolic. A number of other growthfactors are also believed to play a role in the ex-cessive growth of infants of diabetic mothers.These include a family of insulin-like growthfactors and their binding proteins as well asleptin, all of which are elevated in these in-fants.

The most obvious perinatal event is the lossof the constant supply of nutrients to the in-fant. At birth, the serum glucose concentrationof a normal infant is approximately 4.2 mmol/L(the majority of glucose in the blood is locatedoutside the red blood cells and thus is inserum), and the serum volume 45 mL/kg. There-fore, the total amount of available circulatingglucose is 0.19 mmol/kg.At a utilization rate of0.03 mmol/kg/min, the fetus receives enoughglucose to support 7 minutes of normal metab-olism. Obviously, continued existence dependson the ability of the infant to produce glucoseor alternative fuels from endogenous sources.This is especially crucial for human brain cells,which have been shown to require some glu-cose constantly as a source of energy. Glucosetransport into human brain cells is not regu-

lated by insulin (although insulin may play arole in the expression of some glucose recep-tors in the brain); rather, it occurs by facilitatedtransport. Therefore, unlike muscle and fatcells, in which high insulin concentrations cancontinue to maintain glucose entry in the faceof hypoglycemia, in brain cells the lower theserum concentration of glucose, the less glu-cose the brain cells receive. Consequently, pro-longed periods without glucose may produceirrevocable brain damage.

The consequences of severe hypoglycemiato the newly born infant are significant; there-fore, late fetal and early neonatal metabolism isspecifically adapted to alleviate these effects.Initially, there is a precipitous decline of bloodglucose in the newborn infant (Fig. 10-2).Whenthe blood glucose pool is abruptly depleted, theinitial source for replacement is from hepaticglycogen.The liver of the newborn infant con-tains about 0.012 kg of glycogen, representingsufficient glucose for about 12 hours of normalglucose use. Mobilization of these stores is de-pendent on a rapid change in the neonatal hor-monal environment (Cowett, 2004). Over thefirst hours, there is a marked decline in the fetalinsulin concentration and a striking rise in fetalglucagon concentration (Fig. 10-3).

Other counterregulatory hormones arealso affected by birth. Plasma epinephrineand norepinephrine concentrations increasemarkedly, as does the concentration of corti-sol. These rapid changes in hormone levels setin motion a cascade of enzymatic reactionsthat generate active phosphorylase, which inturn catalyzes the degradation of glycogen toglucose 6-phosphate. These same hormonalchanges are likely also responsible for the re-markable postnatal induction of glucose 6-phosphatase, the enzyme responsible for thefurther metabolism of glucose 6-phosphate toglucose.

Because infants normally do not receive an ad-equate supply of exogenous glucose (from feed-ing) for the first days of life, however, hepaticglycogen stores are soon exhausted, and the glu-cose pool must be replenished from other en-dogenous sources (e.g., gluconeogenesis).

The only other endogenous source for glu-cose production is the generation of glucosefrom gluconeogenic amino acids. By late fetallife, all the enzymes needed for gluconeogenesisare present; however, because there is an easilyobtainable source of glucose (i.e., across the pla-centa) and the fetal endocrine environment isnot suitable for gluconeogenesis, the pathways


for the production of newly formed glucose arenot used.Shortly after birth, the fall of the bloodglucose concentration and the rise in concen-trations of glucagon, catecholamines, and possi-bly the corticosteroid hormones serve as strong

stimulants for the initiation of gluconeogenesis.These factors increase the synthesis and activityof cytosolic phosphoenolpyruvate carboxy-kinase, an enzyme that is directly related tohepatic gluconeogenic capacity.This activity of


Glucose-6-phosphatase

Glucose-6-phosphate

Glyceraldehyde 3-phosphate

3-Phosphoglycerate

1,3-Bisphosphoglycerate

2-Phosphoglycerate

Fructose-6-phosphate

Fructose 1,6-phosphate

Fructose 1,6-Diphosphate Aldolase

Triosephosphate IsomeraseGlyceraldehyde Phosphate Dehydrogenase

Phosphoglyceromutase

Phosphoenolpyruvate GlucoseGlycogenicamino acids

Malate Dehydrogenase

Carboxykinase

Enolase

Fructose Diphosphatase

Glucosephosphate Isomerase

GlucosePi

CO2

NADHNAD+

GDPGTP

Phosphoenolpyruvate

Oxaloacetate

Oxaloacetate

Pyruvate

Pyruvate

PhenylalanineTyrosine

AlanineThreonine

SerineCysteine

Fumarate

Succinyl CoA

α-Ketoglutarate

Succinate

Acetyl CoA

PyruvateDehydrogenase

CO2

CO2

Malate

Pi

ADPATP

ADPATP

Acetyl CoAPyruvate Carboxylase

Malate DehydrogenaseNADHNAD+

Malate

NADH

Insulin

AMPCitrate, 3PGInsulin

IsoleucineMethionine

Valine

Glutamate

Isocitrate

TRICARBOXYLICACID

CYCLE

Citrate

ArginineHistidine

GlutamineProline

LeucineLysine

Tryptophan

Acetyl CoA

AspartateAsparagine

NAD+

Figure 10-1. Enzymatic pathways for glucose synthesis from amino acids or pyruvate inmammalian liver. Enclosed in the boxes are the glucogenic amino acids with arrows indi-cating the points where carbon skeletons from these amino acids enter the pathways ofgluconeogenesis or the tricarboxylic acid cycle.Bracketed next to the rate-controlling en-zymes for gluconeogenesis are some of the substances that increase (↑) or decrease (↓)the activity of these enzymes. 3PG, 3-phosphoglycerate.

this enzyme, which is minimal in the near-termfetus, rises remarkably postnatally (Fig. 10-4).Amino acids derived from protein breakdownin skeletal muscle provide the most importantsource of gluconeogenic precursors. Thissource of glucose allows the newborn to main-tain normal blood glucose concentrations dur-ing the period before the onset of adequatefeeding (and adequate exogenous carbohy-drates).

In addition, hepatic fatty acid oxidation is alsorequired to sustain gluconeogenesis. These fattyacids may be obtained from exogenous feedingor metabolism of fatty acids released from en-dogenous lipid stores. β-Oxidation of fatty acidsprovides the acetyl-CoA needed to activate mito-chondrial pyruvate carboxylase and the NADHused as the substrate in the reaction catalyzed byglyceraldehyde 3-phosphate dehydrogenase inthe direction of gluconeogenesis (see Fig.10-1).


0

20

80

40

60

120

100

180

140

160

220

200

1

Pla

sma

Glu

cose

(mg/

dL)

2 3 4

(52)

(52)

*

(51)

(51)

(49)

(69)

(40)

(55)

(55)

(35)

(26)

6

Age in Hours

( ) Number of Samples

* Mean and 95% ConfidenceInterval

12–2425–48

49–7273–96

97–168

Figure 10-2. Decline in plasma glucose over the first hour of life in healthy term infantsand its rapid increase by 3 hours without any intervention.Reproduced with permissionfrom Srinivasan et al. (1986).

Glu

cago

n (n

g/m

L) Insulin (µU/m

L)0.2

0.4

0.6

0.8

1.0

0

20

40

60

80

100

0

0 3 6

Insulin

Glucagon

24 Hours−24 1

Figure 10-3. Postnatal changes in newborn rat plasma insulin and glucagon concentra-tions. Reprinted with permission from Girard et al. (1992).

Although glycogenolysis and gluconeogene-sis represent the only pathways that the new-born can call on to produce glucose fromendogenous sources, glucose usage could beminimized, and thus the blood glucose levelmaintained, if other substrates were to supplythe metabolic needs of tissues. The newbornhas a potential source of energy in the lipidstores present primarily in white adipose tissue(Girard et al., 1992). Indeed, among species, thehuman is unique in having a body fat contentof about 16% of body weight.These stores arepresent almost entirely in the form of triacyl-glycerols, which contain significant amountsof palmitic and oleic acid. Mobilization ofthese stores in adipose tissue is controlled by alipase that is regulated by the hormonal envi-ronment.

Insulin is probably the most important in-hibitor of lipolysis. In contrast to adults, in whomcatecholamines represent the most importantstimulators of lipolysis, thyrotropin (TSH) is themost important stimulator of lipolysis in thenewborn. Plasma free fatty acid concentrationsrise markedly in the first hours after birth in re-sponse to a marked increase in the TSH concen-tration and a fall in the insulin concentration.The fatty acids released from lipid stores are oxi-dized by some extrahepatic tissues (e.g., heartand skeletal muscle, kidney, intestine, and lung).Because the respiratory quotient (the ratio ofcarbon dioxide production to oxygen use) fallsfrom a value of 1.0 (showing that carbohydrateoxidation is the primary source of energy) to avalue of 0.8 to 0.9 (showing increasing oxida-tion of protein or fatty acids) at 2 to 12 hours ofage, at a time when protein catabolism is usuallyinsignificant, fatty acid oxidation must represent

a significant energy source in the early postna-tal period.

Fatty acids are metabolized differently in theliver than they are in other tissues (Fig. 10-5).Only a small portion of the fatty acids is oxi-dized totally to the level of CO2 and water inliver. Hepatic tissues convert most of the fattyacids they oxidize to ketone bodies, which arereleased into the bloodstream and carried toperipheral tissues, most notably the brain,where they are used as metabolic fuels.Becauseglucagon stimulates ketogenesis and insulin in-hibits ketogenesis, the hormonal changes thatoccur at birth would be expected to promoteketone body formation; however, blood ketonebodies remain quite low during the first hoursafter birth, reaching a maximal value only after1 to 2 days. The reason for this delay in ketonebody formation may be the low concentrationsof carnitine in the human newborn liver. Carni-tine is needed to transfer long-chain fatty acidsinto the mitochondria, where β-oxidation oc-curs (Fig. 10-5). Because newborn infants havea limited capacity for carnitine synthesis, exoge-nous carnitine must be supplied through thediet before ketogenesis can proceed optimally.Thus, although cerebral oxidation of ketonebodies occurs shortly after birth and sparessome of the cerebral glucose requirement, therestill is a major need for glucose as an energysource in the brain.

With the above discussion in mind, it is possi-ble to predict at least some of the consequencesto the fetus and newborn of an elevated mater-nal blood glucose concentration, such as occursin diabetes mellitus (Cowett, 1998). As the ma-ternal blood glucose level increases, the gradientbetween fetal and maternal blood glucose


PE

PC

K a

ctiv

ity (

U/g

live

r)G

luconeogenesis(µm

ol/h/100 g)

1

2

3

4

5

0

10

20

0

0Birth

3 6

Gluconeogenesis

PEPCK

24 Hours−24 1

Figure 10-4. Postnatal changes in newborn rat liver cytosolic phosphoenolopyruvatecarboxykinase (PEPCK) activity and gluconeogenic rate in vivo from [14C-lactate].Reprinted with permission from Girard et al. (1992).

concentrations increases, causing more glucoseto be transported to the fetus. With the conse-quent rise in the fetal blood glucose concentra-tion, the fetal pancreas releases more insulin(fetal hyperinsulinism). Fetal skeletal and car-diac muscles now become exposed to higherblood concentrations of both glucose and in-sulin. In response to the consequent increase inthe rate of glucose transport, muscles metabo-lize more glucose and store more of it as glyco-

gen. The fetal liver also has a higher than usualglycogen content. In addition, the activities ofthe key gluconeogenic enzymes (see Fig. 10-1),phosphoenolpyruvate carboxykinase in particu-lar, are depressed. The hyperinsulinemia stimu-lates fat cells throughout the body, includingsubcutaneous adipocytes to deposit lipid at amaximal rate, causing the fetus to weigh morethan would be expected based on its length orhead circumference.


LCFAcyl-CoA

LCFAcyl-CoA

Octanoyl-CoA

CirculatingKetone Bodies

Acetoacetate

Pyruvate Pyruvate Acetyl-CoA

Acyl-Carnitine

Acetyl-CoA

Malonyl-CoA

LCFA

CoA

Acetoacetate

Acetoacetyl-CoA

CitricAcidCycle

Citrate

CO2

CitrateHydroxymethylGlutaryl-CoA

3-Hydroxybutyrate 3-Hydroxybutyrate

LCFA

CoA1

8

7

2

3

4

5

6

Octanoate

Palmitoylcarnitine

LCFA VLDL

Triacylglycerol

Glycerol-3Phosphate

−CPT II

CPT I

Figure 10-5. Intrahepatic metabolism of free fatty acids (FFA). CPT I, CPT II, carnitinepalmitoyltransferase I, II, respectively; LCFA, long-chain fatty acid;VLDL, very low-densitylipoprotein. 1, Long-chain acyl-CoA synthase; 2, acetoacetyl-CoA thiolase; 3, hydrox-ymethylglutaryl-CoA synthase; 4, hydroxymethylglutaryl-CoA lyase; 5, 3-hydroxybutyratedehydrogenase; 6, acetyl-CoA carboxylase; 7, fatty acid synthase; 8, glycerolphosphateacyltransferase; Reprinted with permission from Girard et al. (1992).

In addition, the metabolic effects of insulinultimately increase fetal oxygen consumption.The circulation of the fetus is unique; specifi-cally, because blood in the fetal descendingaorta contains a portion of venous return aftertissue perfusion, the increased rate of consump-tion of oxygen by fetal tissues decreases thearterial oxygen concentration of the fetus. Thefetal tissues supplied by this circulation (whichinclude most of the mass of the fetus) are thenat risk for inadequate tissue oxygenation.

There are two consequences of such tissuehypoxia. First, as renal oxygenation becomes in-adequate, the fetal kidney produces erythropoi-etin, which in turn stimulates the productionof new erythrocytes. This response leads to anincreased hematocrit and an increase in theoxygen-carrying capacity of the blood of the fe-tus. Second, normal fetal metabolism may be al-tered if the supply of oxygen to the fetusbecomes marginal. Generally, the normal fetusreceives about twice as much oxygen acrossthe placenta as needed for normal metabolism.For any fetus, labor increases the likelihood ofdecreased oxygen supply as uterine contrac-tions interfere with blood flow to the intervil-lous space of the placenta. The fetus of thediabetic mother, even if it receives a normaloxygen supply,has, as noted, an accelerated rateof oxidative metabolism. The fetus of the dia-betic mother is therefore at greater risk forhypoxic damage, as illustrated by the presentcase.

When the fetus lacks an adequate oxygensupply, it must satisfy its energy needs throughanaerobic pathways (e.g., glycolysis); however,the glycolytic pathway of glucose utilization,while generating ATP, produces lactic acid (byreducing pyruvate).A fall in the blood pH is in-dicative of anaerobic metabolism. The use ofglycolysis for energy production is an ineffi-cient process that yields only 2 molecules ofATP per molecule of glucose metabolized, com-pared with the approximately 30 ATP mole-cules produced by the combined action ofglycolysis, the citric acid cycle, and oxidativephosphorylation. Indeed, anaerobic glycolysis istoo inefficient to provide the energy the fetusneeds to survive for more than a few hours ordays. The finding of a pH of 6.9 in the infant de-scribed in the present case report was evidenceof significant anaerobic metabolism and im-pending fetal death. It was imperative that thephysicians caring for this patient remove the fe-tus from the hostile intrauterine environment

so that effective oxygenation and aerobic me-tabolism could proceed.

Once one understands the normal metabolicadaptation to extrauterine life, it is easy to seewhy the infant whose mother has diabetes isalso likely to have problems maintaining neo-natal glucose homeostasis. Before birth, infantsof diabetic mothers whose placentas have func-tioned normally will receive a constant over-supply of glucose,which will induce high insulinlevels and low concentrations of the principalcounterregulatory hormones: glucagon and thecatecholamines. Glucose utilization can be in-creased to rates as high as 10 mg/kg per minutein some infants of diabetic mothers. Even withthe fetal blood glucose concentration (becauseof the maternal diabetes) as high as 8.3 mmol/L(representing a glucose pool of 0.45 mmol/L atbirth), this source of glucose will be exhaustedwithin 8 minutes. Therefore, the infant of a dia-betic mother, like a normal infant, must gener-ate endogenous glucose to maintain themetabolism of tissues, such as brain, that areconstantly dependent on this metabolic sub-strate. Such infants have, as mentioned, abun-dant liver glycogen stores.

Over the first hours of life, however, such in-fants persist in having higher than normal in-sulin concentrations and lower than normalglycogen concentrations (Fig. 10-6). In addition,the concentrations of other counterregulatoryhormones, such as catecholamines, remain low.This particular hormonal milieu will seriouslyinhibit glycogenolysis, thereby decreasing therate of glucose release from the liver.

Unfortunately, the endocrine environment ofthe infant whose mother is diabetic also sup-presses the production of glucose from its onlyother major potential source: amino acids. Highinsulin levels in both fetal and neonatal life in-hibit protein breakdown, thereby decreasingthe availability of amino acids from endogenoussources for potential use as gluconeogenic sub-strates. In addition, the marked increase thatnormally occurs in the activities of two impor-tant rate-controlling gluconeogenic enzymes,cytosolic phosphoenolpyruvate carboxykinaseand glucose 6-phosphatase, is suppressed ow-ing to the higher insulin and lower counterreg-ulatory hormone concentrations of infants bornto diabetic mothers. The combined effect of thedecreased capacity for glycogenolysis and glu-coneogenesis in infants of diabetic mothers is amarked decrease in the ability of the newbornto release glucose from the liver into the blood.


Such a decrease in glucose production wouldhave less impact if other metabolic fuels couldbe mobilized and used. As noted, the fetuswhose mother has diabetes has greatly increasedfat stores. In these infants, the amount of fat pres-ent is theoretically sufficient to supply adequatemetabolic substrate for weeks, but the high in-sulin concentrations markedly inhibit lipolysis.As a consequence, tissues like heart and skeletalmuscle, which are otherwise capable of usingfatty acids to satisfy metabolic needs, are com-pelled to continue oxidizing glucose. The infantof a diabetic mother, therefore, in the face of im-paired glucose production, has a higher thannormal rate of glucose consumption. This cir-cumstance precipitates a marked fall in theblood glucose concentration of the neonate.

The cerebral requirement for oxidative sub-strate can be met by only two substrates: glu-cose and ketone bodies. Unfortunately, theendocrine milieu of the infant of the diabeticmother inhibits both lipolysis (i.e., the mobiliza-tion of fatty acid stores) and hepatic ketogene-

sis (formation of ketone bodies). Consequently,the brain is left with no alternative to glucoseas a metabolic substrate, and the use of glucoseto support normal metabolism becomes obliga-tory. Unless steps are taken to supply exogenousglucose or to alter the endocrine environmentso endogenous stores can be mobilized, the in-fant of a diabetic mother will be at great risk forbrain injury.

The present case presentation and discussionof the consequences of maternal hyperglycemiapresume that placental function is sufficientlypreserved to allow maternal hyperglycemia tobe reflected in fetal hyperglycemia. By contrast,some diabetic mothers whose disease is suffi-ciently long-standing or advanced to cause renalor retinal vascular disease will also have vasculardisease that adversely affects the placenta andcompromises its ability to transport substratesto the fetus. In such cases, the infants are af-fected by a chronic undersupply of both oxygenand other critical metabolic substrates, includ-ing glucose. Because the infants have received

116 FUEL METABOLISM AND ENERGETICSP

lasm

a G

luco

se (

mg/

dL)

50

100

0

UV 1

IDM

Control

2 Hours

Pla

sma

Insu

lin (

ng/m

L)

2

4

0

UV 1

IDM

Control

2 Hours

Pla

sma

Glu

cago

n (p

g/m

L)

100

200

0

UV 1

IDM

Control

2 Hours

GlucoseUtilization

GlucoseProduction

Figure 10-6. Plasma glucose, insulin, and glucagon concentrations in infants from dia-betic mothers. UV, umbilical vein. Reprinted with permission from Girard et al. (1992).

subnormal quantities of glucose, they have lowinsulin levels and higher than normal concentra-tions of counterregulatory hormones. Conse-quently, both their glycogen and lipid stores areminimal. Such an undersupply of fuel reservesresults in an infant that is notably smaller thanwould be expected for its gestational age.

Despite the small size of the infant at birth,anobligate need for glucose (∼4 mg/kg/min) stillexists. Thus, an infant born with a subnormalserum glucose concentration of 3.3 mmol/L willhave used up the entire circulating glucosepool within about 8 minutes. With such a rapiddepletion of serum glucose, the infant must relyon glycogenolysis to replenish blood glucose.Although the endocrine environment of suchinfants, specifically one in which there are lowinsulin levels and higher than normal concen-trations of counterregulatory hormones, favorsglycogenolysis, complete depletion of glycogenreserves will occur in about 4 hours, which ismuch faster than normal. These infants mustthen resort to another source of endogenousglucose production, namely, gluconeogenesis.The endocrine environment of such infants isconducive to both mobilization of amino acidsfrom protein breakdown and induction of thehepatic enzymes needed for gluconeogenesis.Nonetheless, gluconeogenesis remains deficientas a result of the lack of free fatty acid fromlipolysis, which diminishes intramitochondrialacetyl-CoA and compromises hepatic gluconeo-genesis.The free fatty acids are not available toprovide metabolic substrate because lipolysis,although favored by the endocrine milieu, islimited by a lack of adipose triglyceride.There-fore, infants born to diabetic mothers, in whomplacental function has been compromised, lackthe endogenous sources of glucose and are alsoat risk for hypoglycemia; however, the etiologyis one of decreased substrate stores rather thanan inappropriate endocrine environment.

THERAPY

Because all the myriad consequences to the in-fant of a diabetic mother arise from maternalhyperglycemia, the therapy of hyperglycemiafor the infant should begin before birth. Carefulmanagement of the maternal diabetes to pre-vent both hypoglycemia and hyperglycemiawill lessen the likelihood to fetal death orneonatal hypoglycemia. In the present case,economic hardship (the husband’s layoff and

loss of health insurance) prevented Mrs. L. fromobtaining adequate prenatal care. The resultantneonatal complications can be traced directlyto this problem.

There are circumstances that preclude the pre-ventive measures outlined above. In such cases,there are two main methods of relieving thehypoglycemia: provision of exogenous substrateor alteration of the infant’s endocrine status.Afterdelivery,assuming the infant is otherwise healthy,the early provision of calories by milk feedingscan provide adequate exogenous sources of glu-cose to prevent hypoglycemia. In some infants,however, either because of other illnesses or be-cause of the severity of the hypoglycemia(Cornblath and Ichord, 2000), it is necessary toprovide intravenous glucose. Although this wasdone in the present case, the provision of glu-cose at only 4 mg/kg/min was inadequate to en-sure a sufficient quantity of glucose for use bythe brain.The initial intravenous therapy shouldprovide at least 6 mg/kg/min.

When hypoglycemia does occur, it is impera-tive for the physician to respond rapidly and ap-propriately.Exogenous glucose delivered directlyinto a vein is the treatment of choice,but cautionmust be exercised because the rapid injectionof quantities of glucose far in excess of require-ments will lead to a sudden increase of bloodglucose to hyperglycemic levels. For infants ofdiabetic mothers, this further stimulates insulinrelease, which then rapidly produces reboundhypoglycemia, as in the present case.The use of10 mL of 25% dextrose produced a blood glu-cose concentration of nearly 22.2 mmol/L, fol-lowed by a rapid decrease. It is more prudentand efficacious to administer 1 to 2 mL/kg of10% dextrose over approximately 5 minutesand then begin a continuous intravenous infu-sion calculated to supply 6 to 8 mL/kg of glu-cose per minute.

In an emergency, when an intravenous sitecannot be rapidly obtained, the endocrineenvironment of the infant can be made moreconducive to glucose release by use of intramus-cular or subcutaneous injection of glucagon(0.3 mg/kg per dose) to produce a short-lived(<30 min) increase in blood glucose while an in-travenous catheter is inserted. Glucagon acts notonly by increasing glycogenolysis but also bystimulating phosphoenolpyruvate carboxyki-nase, which will increase gluconeogenesis.Glucagon may also increase the capacity for fattyacid oxidation, which is also essential for gluco-neogenesis. Infants whose glycogen is depleted


(i.e., growth-retarded neonates) cannot be ex-pected to respond as effectively to glucagon asinfants of diabetic mothers. If possible,milk feed-ing should also be instituted to stimulate the se-cretion of many glucoregulatory gut hormonesand to supply an additional source of glucoseand gluconeogenic precursors.

In a few infants, massive quantities of in-travenous glucose fail to prevent repeatedepisodes of hypoglycemia, and manipulation ofthe neonatal endocrine status is required to min-imize glucose use and maximize glucose produc-tion.The use of a brief course of glucocorticoids,one of the counterregulatory hormones, as inthe present case, can effectively treat the hypo-glycemia. The mechanisms of action are three-fold: (1) glucocorticoids counteract the effect ofinsulin on peripheral muscle glucose uptake,thereby “sparing” glucose for brain use and rais-ing the blood glucose concentration; (2) gluco-corticoids stimulate hepatic glycogenesis andgluconeogenesis; and (3) glucocorticoids in-crease proteolysis in peripheral tissues, therebyincreasing the availability of gluconeogenic sub-strates. The effectiveness of this therapy wasclearly shown in the present case.

It is not surprising that infant L. suffered dif-fuse encephalopathy (brain disorder), a cere-bral infarction, and seizures during the neonatalperiod (Yager, 2002). Both asphyxia and hypo-glycemia are injurious to the brain. The treat-ment for seizures consists of providing normalmetabolic substrates (e.g., glucose) and appro-priate anticonvulsant therapy (phenobarbital),as was done in the present case. The long-termtreatment for the child’s developmental disabili-ties is complex and involves the skills of manymembers of the health care team.

The prognosis of infant L. is uncertain.Asymptomatic hypoglycemia in an infant of adiabetic mother is usually less injurious thanthose episodes associated with central nervoussymptoms. Obviously, in the present case, al-though the child had symptoms of seizures andan abnormal neurological examination, it is notpossible to differentiate these symptoms fromthose that might have resulted from the chronicinadequacy of cerebral oxygenation in utero.Nevertheless, the prognosis for an asphyxiatedchild with an abnormal electroencephalogram,a cerebral infarction, and persistent abnormali-ties on neurological examination is guarded atbest. The bilateral hearing loss and the inabilityto swallow and feed adequately at the time ofdischarge are strong evidence of some degree

of neurological disability in the future. Predic-tions about specific motor and mental poten-tials are impossible to make in the earlyneonatal period.

Generally, the prognosis for infants born todiabetic mothers is good; however, severesymptomatic hypoglycemia can lead to lifelongproblems. Infants whose hypoglycemia reflectsa long-term lack of adequate nutrition and oxy-genation in utero (such as severe growth-retarded infants) are at even greater risk fordevelopmental disabilities. In both cases, theadditional burden of birth asphyxia will in-crease the likelihood of long-term problems.Obviously, the severity of illness in the infantdescribed here and the ominous signs of cere-bral disturbance at the time of discharge bodepoorly for the child’s future development.

QUESTIONS

1. Compare the hormonal changes that oc-cur at birth and their effects on carbohy-drate homeostasis with those that occurin adults during starvation.

2. What was the primary medical cause ofinfant L.’s problems? Why was appropriatetherapy not initiated in the present case?

3. What substance is an important circulat-ing oxidative substrate in fetal life but un-common postnatally? What commonmetabolic substrate of postneonatal life isrelatively less important during fetal life?

4. Can free fatty acids serve as a directsource of glucose (i.e., can fatty acids beconverted into glucose)? Why or why not?

5. Would glucagon or glucocorticoids be aseffective a therapy for hypoglycemia inthe growth-retarded infant of a diabeticmother as in newborns who are large forgestational age?

BIBLIOGRAPHY

Cornblath M, Ichord R: Hypoglycemia in theneonate. Semin Perinatol 24:136–149, 2000.

Cowett RM: The infant of the diabetic mother, inCowett RM (ed): Principles of Perinatal-Neonatal Metabolism. Springer-Verlag, NewYork, 1998, pp. 1105–1129.

Cowett RM: Role of glucoregulatory hormones inhepatic glucose metabolism during the perina-tal period, in Polin RA, Fox WW, Abman SH


(eds): Fetal and Neonatal Physiology. Saun-ders, Philadelphia, 2004, pp. 478–486.

Cowett RM, Farrag HM: Neonatal glucose me-tabolism, in Cowett RM (ed): Principles ofPerinatal-Neonatal Metabolism.Springer-Verlag,New York, 1998, pp. 683–722.

Girard J, Ferre P, Pegorier J-P, et al.: Adaptations ofglucose and fatty acid metabolism during theperinatal period and suckling-weaning transi-tion. Physiol Rev 72:507–562, 1992.

Simmons RA: Cell glucose transport and glucosehandling during fetal and neonatal development,in Polin RA,Fox WW, Abman SH (eds):Fetal andNeonatal Physiology. Saunders, Philadelphia,2004, pp. 487–493.

Srinivasan G, Pildes RS, Cattamanchi G, et al:Plasma glucose values in normal neonates: anew look. J Pediatr 109:114–117, 1986.

Yager JY: Hypoglycemic injury to the immaturebrain. Clin Perinatol 29:651–674, 2002.


Part III

Intermediary Metabolism

CASE REPORTS

Case 1

A 50-year-old female presented passing blackurine after having been unwell for 3 days withfever and headache. There was no previous his-tory of blood disorders, anemia, or similar symp-toms. Past medical history included a diagnosisof multiple sclerosis. Treatment included ste-roids and high-dose vitamins and minerals.These had been given on five separate occa-sions over the preceding 2 weeks and included200 mg vitamin B1,25 mg B2,250 mg B5,100 mgB6, 15 mg B12, 50 g C, 2.5 mg folic acid, 15 mgzinc, 11 mg magnesium, 10 µg molybdenum,100 µg selenium, and 50 µg chromium. In hersocial history, she was of Arabic descent fromEast Africa, currently living in Kuwait. Hernephew was reported to suffer from favism (anacute hemolytic condition occurring after inges-tion of a certain species of bean [Vicia faba] inindividuals with glucose 6-phosphate dehydro-genase [G6PD] deficiency). There was no recenthistory of travel or malaria exposure.

Laboratory tests showed that hemoglobinwas 7.5 g/dL (normal range is 12–15 g/dL), andthe reticulocyte count was 262 × 109/L (normalrange is 50–100 × 109/L). The tests also indi-cated neutrophilia, a white cell count of34 × 109/L (normal range is 4–11 × 109/L), anda platelet count of 328 × 109/L (normal range is150–400 × 109/L). Renal function was normal.Malarial parasite screen and direct Coombs testwere negative. Blood film showed nucleatedred blood cells and anisocytosis with “bite cells”

and irregularly contracted cells consistent withacute oxidant drug-induced hemolysis (Fig. 11-1). G6PD deficiency screen was positive. Onquantitative assay, G6PD measured 2.27 enzymeunits (normal range 5.34–11.34), compatiblewith a diagnosis of severe G6PD deficiency.Management included blood transfusion, intra-venous fluids, folic acid, and iron supplementa-tion.All current medications were stopped.

Over the following days, her hemoglobin con-tinued to drop but stabilized after a further trans-fusion. Her hemoglobinuria (dark rust-coloredurine) gradually improved and the patient madea full recovery. No further hemolytic episodeshave occurred. Family members have beenscreened (two brothers, two sons). The highdose of vitamin C administered to this patientwas thought to be the precipitating factor (seeTable 11-1), but it is possible that the high dosesof the other vitamins may have contributed tothe hemolysis.

Case 2

A 2-day-old baby boy was noted to be jaundiced(yellowing of skin and eyes). His initial bilirubinwas 25 mg/dL (normal range is 2–10 mg/dL). Bythe following day, he was clinically more jaun-diced. Phototherapy was initiated. Preparationwas made for exchange transfusion,but by day 5of age, bilirubin levels began falling and contin-ued to fall over the subsequent few days. Therewas an associated fall in hemoglobin to a mini-mum of 12 g/dL (normal range is 15–21 g/dL),and the reticulocyte count was raised at 20%(normal range 2%–5%). On direct questioning of

11

Glucose 6-Phosphate Dehydrogenase Deficiency and Oxidative Hemolysis

CATHERINE BURTON and RICHARD KACZMARSKI

123

his family history, he had an older brother whohad also suffered with neonatal jaundice, whichhad resolved spontaneously. There was also his-tory of a maternal uncle having had intermittentjaundice and one episode of dark urine when hehad a respiratory tract infection as a child. AG6PD assay was performed and confirmed ery-throcytic G6PD deficiency.

His parents have been counseled regarding thesignificance of G6PD deficiency. They have beenadvised to avoid certain drugs and to seek med-ical attention on development of dark urine. Hisolder brother has been screened and has beenconfirmed to be G6PD deficient.Any future sonsborn to this mother should also be screened.

DIAGNOSIS

G6PD deficiency is the most common enzy-matic disorder of red blood cells and should beconsidered in the differential diagnosis of anycase of hemolytic anemia. It affects 400 millionpeople worldwide with a wide variation inprevalence. It is very rare in the indigenouspopulations of northern Europe but increasesto frequencies of 20% in parts of southern Eu-rope, Africa, and Asia and up to 40% in certainareas of Southeast Asia and the Middle East.

G6PD deficiency is expressed clinically bya spectrum of hemolytic syndromes. Clinical

variation in presentation from asymptomatic toepisodic anemia to chronic hemolysis is due togenetic heterogeneity. The majority of patientsare asymptomatic and initially present with ane-mia due to a precipitating cause, for example,an infectious illness, exposure to certain drugs,or metabolic disturbance.

The diagnosis should be considered in theclinical setting of a patient from an appropriategeographic or ethnic origin who presents withevidence of hemolytic anemia. Clinically, acuteintravascular hemolysis is characterized by hemo-globinuria, jaundice, and symptoms and signs ofanemia. Less-acute presentations may present inthe course of investigation of jaundice. Chronichemolysis may lead to gallstone formation,which may become symptomatic with abdomi-nal pain.

The complete blood count will show a vary-ing degree of anemia and a raised reticulocytecount. The mean cell volume may be raised dueto the reticulocytosis or concurrent folate defi-ciency. As the oxidative denaturation of hemo-globin occurs, hemoglobin collects at one sideof the cell with an adjacent membrane-boundclear zone. This gives the characteristic bite cellappearance on blood smears (Fig. 11-1). Otherblood film appearances include red cell frag-ments, irregularly contracted erythrocytes, andHeinz bodies (seen on supravital stains; repre-sent precipitated denatured hemoglobin). The

124 INTERMEDIARY METABOLISM

Figure 11-1. Peripheral blood features of oxidative hemolysis showing bite cells, irreg-ularly contracted red cells, and nucleated erythrocytes.

biochemical profile will show elevated serumbilirubin levels, mainly unconjugated (indirect)bilirubin, consistent with intravascular hemoly-sis. Serum levels of red cell enzymes such as lac-tate dehydrogenase (LDH) will also be raiseddue to the hemolysis. Renal function may be ab-normal, and some patients develop renal failure,particularly adults who may have previously un-recognized renal damage. This is the result ofthe hemoglobinuria but may be exacerbated bythe intercurrent infection or drugs. Occasion-ally, this will require hemodialysis.

The diagnosis of G6PD deficiency is madeby quantitative or semiquantitative assay. Themethods used are based on the normal functionof the G6PD enzyme in catalyzing the initialstep in the pentose phosphate pathway (PPP).Screening tests depend on the inability of cellsfrom deficient subjects to convert an oxidizedsubstrate to a reduced state. The substrates

used may be the natural one of the enzyme,namely, NADP+, or other substrates linked bysecondary reactions. The resultant reaction canbe demonstrated by fluorescence,color change,or dye deposition methods. There are a numberof screening tests available,using brilliant cresylblue decolorization, methemoglobin reduction,or an ultraviolet spot test.

A simple method for diagnosis of G6PD defi-ciency is the fluorescent spot test. By addinga measured amount of hemolysate to glucose6-phosphate (substrate) and NADP+ (substrate),the rate of NADPH production can be measuredspectrophotometrically by measuring light ab-sorbance of NADPH at 340 nm.Other screeningtests include the estimation of NADPH genera-tion indirectly by measuring transfer of hydro-gen ions from NADPH to an acceptor. In themethemoglobin (metHb) reduction test, meth-ylene blue is used to transfer hydrogen fromNADPH to metHb, thereby promoting its reduc-tion. This technique can detect relative G6PDsufficiency in individual red cells, detecting thecarrier state with 75% accuracy. These tests willdetect a threshold G6PD activity of about 30%of normal and will identify hemizygous defi-cient males and homozygous deficient females.An abnormal screening test should be followedby a quantitation of G6PD activity by spec-trophotometric assay to confirm the diagnosis.

Although in most instances a screening testis adequate to make a diagnosis of G6PD defi-ciency, false-negative results and difficulties indiagnosis are a fairly common problem. False-negative results occur if the most severely defi-cient red cells have been destroyed by hemolysis.Therefore, the remaining cells analyzed have suf-ficient G6PD activity for the diagnosis to bemissed,especially during the reticulocytosis thatfollows acute hemolysis. This is not usually aproblem for male hemizygotes or female ho-mozygotes but can result in a missed diagnosisin female heterozygotes and for both sexes inmilder variants. Females have varying propor-tions of G6PD-deficient and normal red cells,and the reticulocytes of milder variants oftenhave normal or near normal levels of G6PD ac-tivity. To investigate these cases further, familymembers should be studied especially in neo-natal and pediatric cases.Alternatively, reevalua-tion can be repeated a few weeks after thehemolytic episode since red cells of all ages willbe present. If a prompt diagnosis is required,then the oldest remaining cells can be isolatedby differential centrifugation and demonstrated

G6PD Deficiency and Oxidative Hemolysis 125

Table 11-1. Drugs and Chemicals to Be Avoidedin Patients with Glucose 6-phosphateDehydrogenase (G6PD) DeficiencySulfonamides/Sulfones Cytotoxic/Antibacterial

Dapsone Chloramphenicol

Sulfacetamide Cotrimoxazole

Sulfanilamide Nalidixic acid

Sulfapyridine Furazolidone

Sulfasalazine Furmethonol

Salazopyrin Nitrofurantoin

Septrin Nitrofurazone

Sulfamethoxypyrimidine P-Aminosalicylic acid

SulfisoxazoleAntimalarials Cardiovascular Drugs

Chloroquine Procainamide

Hydroxychloroquine Quinidine

Mepacrine (quinacrine) α-Methyldopa

Pamaquine Ascorbic acid

Pentaquine Dimercaprol

Primaquine Mestranol

Quinine Hydralazine

QuinocidAnalgesics/Antipyretics Miscellaneous

Probenicid Vitamin K (water-soluble) analogues

Acetanilid Naphthalene

Amidopyrine (Aminopyrine) Methylene blue

Antipyrine Toluidine blue

Pyramidone

to have low G6PD activity. This technique canalso be used if the patient has been transfused.

A conclusive diagnosis of G6PD deficiencycan be made by DNA analysis. This is of particu-lar use when the most likely variants in a popu-lation are known since polymerase chainreaction amplification of the appropriate regionof the G6PD gene can be performed and themutation identified.


Biology

G6PD deficiency was first reported in 1956(Alving et al., 1956). The importance of G6PDfor red cell integrity was recognized afterAfrican American soldiers taking the antimalar-ial drug primaquine developed acute hemolyticanemia with hemoglobinuria. This led to thediscovery that one of the enzymes required tomaintain adequate intracellular reduced glu-tathione levels was deficient in affected redblood cells.

The majority of glucose 6-phosphate pro-duced by the phosphorylation of glucose byATP is metabolized to lactate and pyruvate inthe glycolytic or Embden-Myerhof pathway. Thedirect nonmitochondrial oxidation of glucose 6-phosphate occurs by the pentose phosphatepathway (PPP), and the G6PD enzyme is the ini-tial enzyme in this alternative route of glucosemetabolism (Fig. 11-2).

The PPP occurs in the cytosol of the cell.It comprises two irreversible oxidative reac-tions followed by a series of reversible sugar-phosphate interconversions. Unlike glycolysisor the citric acid cycle, in which the direction

of reactions is well defined, the interconver-sions reactions of the PPP can occur in differ-ent directions, with the rate and direction ofthe reactions determined by the supply and de-mand of intermediates in the cycle. The path-way provides a major portion of the cell’sNADPH, which functions as a biochemical re-ductant.

The oxidative portion of the PPP consists ofthree reactions that lead to the formation of ribu-lose 5-phosphate, CO2, and two molecules ofNADPH for each molecule of glucose 6-phosphate oxidized.The G6PD enzyme catalyzesan irreversible oxidation of glucose 6-phosphateto 6-phosphogluconolactone. This reaction isspecific for NADP+ as coenzyme. The PPP is reg-ulated primarily at this level with NADPH as thepotent competitive inhibitor of the enzyme. Un-der most metabolic conditions, the ratio ofNADPH to NADP+ is sufficiently high to in-hibit enzyme activity substantially (Fig. 11-3).Increased demand for NADPH decreases the ra-tio of NADPH/NADP+, and increases G6PD activ-ity. Subsequently, 6-phosphogluconolactone ishydrolyzed by 6-phosphogluconolactone hydro-lase (Fig. 11-4). This reaction is irreversible andnot rate limiting. The 6-phosphogluconate pro-duced undergoes oxidative decarboxylation cat-alyzed by 6-phosphogluconate dehydrogenase.This irreversible reaction produces a pentosesugar-phosphate (ribulose 5-phosphate), CO2,and a second molecule of NADPH (Fig.11-5).

G6PD is responsible for maintaining adequatelevels of NADPH inside the cell. NADPH is a re-quired cofactor in many biosynthetic reactionsand also provides electrons to reduce oxidizedglutathione (GSSG) to reduced glutathione (GSH)catalyzed by glutathione reductase (Fig. 11-6).


Figure 11-2. Initial oxidative reactions of the pentose phosphate pathway.

Reduced glutathione acts as a scavenger fordangerous oxidative metabolites in the cell; inparticular, it converts harmful hydrogen perox-ide to water with the help of the enzymeglutathione peroxidase (Yoshida and Beutler,1986).

The malic enzyme also serves to generateNADPH in many cells including hepatocytes.A supply of NADPH is critical for the liver mi-crosomal cytochrome P-450 mono-oxygenasesystem. This is the major pathway for the hy-droxylation of steroids,alcohols,and many drugs.These oxidations also detoxify drugs and for-eign compounds by converting them into solu-ble forms more readily excreted through thekidney. NADPH is also used as a source of elec-trons for the biosynthesis of fatty acids. The sit-uation is unusual in red blood cells as this otherNADPH-producing enzyme is lacking (Scriver etal., 1995). This has a profound effect on the sta-bility of red blood cells since they are especiallysensitive to oxidative stresses as well as havingonly one NADPH-producing enzyme to removethese harmful oxidants (Fig. 11-7).

The accumulation of hydrogen peroxidaseaffects many intracellular processes and resultsin hemolysis. These include the cross-linking ofmembrane proteins; hemoglobin denaturation(manifest as Heinz body formation), which inturn affects the physical properties of the ery-throcyte; and lipid peroxidation, which may af-fect the cell membrane to cause direct hemolysis(Fig. 11-8).The resultant damage leads to a mix-ture of intravascular hemolysis and extravascu-lar hemolysis (by which hemolysis occurs in thereticuloendothelial system). In acute hemolyticepisodes, the clinical picture is of predomi-nantly intravascular hemolysis, while predomi-nantly extravascular hemolysis is seen inpatients with chronic hemolysis.

The G6PD gene consists of 13 exons, whichare the regions of the DNA that code for the en-zyme, and 12 introns, which are intervening se-quences serving no apparent functional purpose(Scriver et al.,1995).The enzyme’s function is de-termined by the sequence and size of the G6PDgene and the mRNA encoded by the gene,whichare 18,500 and 2,269 base pairs in length, re-spectively. The entire genomic sequence of theG6PD gene is described in the original researcharticle by Chen and coworkers (1991).However,the complete three-dimensional structure of theenzyme, which determines the active enzyme’sfunctional properties, has not yet been deter-mined. G6PD, in its active enzyme form, is madeup of either two or four identical subunits, eachsubunit having a molecular mass of about 59 kd(Scriver et al., 1995).

G6PD deficiency is known to have over 400variant alleles or different forms of the samegene (Beutler et al., 1990). The biochemicalcharacteristics of the particular G6PD variantdetermine the extent of enzyme deficiency,


Figure 11-3. Reversible conversion of NADP+ toNADPH.

Figure 11-4. Subsequent oxidative reactions of the pentose phosphate pathway.

which subsequently affects the severity of he-molysis. Mutations can be in the form of pointmutations or deletions that can range from oneto several base pairs as well as replacements inthe DNA (Scriver et al., 1995). Different pop-ulations have different types of mutations, butwithin a specific population,common mutationsare usually shared (e.g., the “Mediterranean” vari-ant in Egypt, the “Japan” variant in Japan)(Scriver et al., 1995). The World Health Organi-zation has classified G6PD-deficient variants intofive classes depending on the level of enzymeactivity (Table 11-2). G6PD A− is the most preva-lent variant,mostly found in Africans and AfricanAmericans. G6PD Mediterranean is the secondmost common variant.

On a molecular level, most individuals withG6PD deficiency have a qualitative abnormalityin the structure of the G6PD enzyme (Scriveret al., 1995). Various models have been pro-posed suggesting possible reasons why an ab-normal enzyme is not fully active.As discussedin Scriver et al. (1995), it was suggested bySharff that the decreased stability of a mutatedenzyme results either from a change in the con-formation of the G6PD molecule or from an in-crease in its susceptibility to proteolyticenzymes, resulting in accelerated decay of the

enzyme during the aging of the erythrocyte.Molecular analysis has demonstrated that eachvariant is associated with a particular biochemi-cal and clinical phenotype, and that mutationscan be found in every exon and are scatteredthroughout the G6PD gene.

Genetics

The gene for the G6PD enzyme is located at theq28 locus on the X chromosome (Pai et al.,1980; Scriver et al., 1995), so the expression ofG6PD deficiency varies between males (XY)and females (XX). All X-linked genetic condi-tions, such as G6PD deficiency, are more likelyto affect males than females. G6PD deficiency isexpressed in males carrying a defective variantgene that produces sufficient enzyme deficiencyto lead to symptoms. In heterozygous females,the mean erythrocyte enzyme activity can benormal, moderately reduced, or severely defi-cient. This is dependent on the degree of ly-onization and the extent of expression of theabnormal G6PD variant.

Lyonization is the random inactivation bymethylation of one of the two X chromosomesin each cell in females, so that the descendentof each cell inherits only one active X chromo-some. This is a random event, and therefore nor-mally a heterozygous female with inactivation ofone X chromosome in each cell via lyonizationhas 50% normal G6PD activity (due to 50% nor-mal erythrocytes) and 50% G6PD-deficient ery-throcytes. Provided there is one good copy ofthe G6PD gene in a female, some normal en-zyme will be produced, and this normal enzymecan then take over the function that the defec-tive enzyme lacks. Excessive lyonization can oc-cur when inactivation of a large proportion ofthe same X chromosome occurs. If this results inthe majority of remaining red cells being G6PD


Figure 11-5. Overview of oxidative reactions ofthe pentose phosphate pathway.

Figure 11-6. Chemical structure of glutathione.

deficient, then the phenotype is comparable tomales and homozygous females. Compound het-erozygosity can also result in G6PD deficiencymanifestation in females when there are two de-fective copies of the gene in the genome.

CLINICAL PRESENTATION

Generally, all of the G6PD variants that occurwith high frequency in various populations(e.g., G6PD A− and G6PD Mediterranean) areasymptomatic and are clinically important only

because of the risk of acute hemolytic anemia.In contrast, symptomatic G6PD variants tend tobe rare in any population. There are four clini-cal presentations of G6PD deficiency: acute he-molytic anemia, neonatal hyperbilirubinemia,congenital nonspherocytic hemolytic anemia,and favism.

Acute Hemolytic Anemia

As illustrated in case 1, acute hemolytic ane-mia occurs when there is destruction of theG6PD-deficient red cells. The fact that the most


Figure 11-7. Consequences of glucose 6-phosphate dehydrogenase (G6PD) deficiency.GSH, reduced glutathione; GSSG, oxidized glutathione.

Figure 11-8. Mechanism of hemolysis in glucose 6-phosphate dehydrogenase deficiency.

prevalent G6PD variants are normally asympto-matic in the steady state means that a level ofenzyme activity of 20% (some studies suggest3%) of normal is probably sufficient for normalred cell function under ordinary conditions. Itcan be demonstrated by isotopic techniquesthat there is slight shortening of the red cellsurvival, but generally there is no evidence ofanemia, and blood film morphology is normal.

Destruction of the G6PD-deficient red cells istriggered by high redox potential drugs (as incase 1), certain infections, and metabolic distur-bances such as acidosis or hyperglycemia. Typi-cally, 2–4 days after drug ingestion there is asudden onset of jaundice, pallor, dark urine, andoften back or abdominal pain. This is associatedwith a rapid fall in hemoglobin. Morphologicalabnormalities such as red cell fragmentation,mi-crospherocytes, and bite cells are present in theblood film. Both hepatic and splenic sequestra-tion (pooling of red blood cells,which no longercontribute to the circulating volume, in these or-gans) can occur.

An appropriate stimulation of erythropoiesisis induced by the anemia.A reticulocytosis is ap-parent within 5 days and is maximal at 7 to 10days after the onset of hemolysis. Even if there iscontinued drug exposure, the acute hemolysisusually resolves after 1 week. This spontaneousrecovery reflects replacement of the older,enzyme-deficient erythrocytes by younger redcells with sufficient G6PD activity to withstandoxidative injury. Provided there is normally func-tioning bone marrow,the continued loss of agingred cells is compensated by the erythropoieticmarrow response.

Neonatal Hyperbilirubinemia

As in case 2, some patients with G6PD variantsexperience hyperbilirubinemia in the neonatalperiod. Neonates with the rare class I variantare particularly at risk, but more commonvariants account for the majority of cases. The

cause of neonatal hyperbilirubinemia is unclearbecause it does not occur in 50% or more ofG6PD-deficient individuals. It is probably due toa combination of increased bilirubin productionsecondary to accelerated red cell breakdownand the immaturity of the neonatal liver, withthe latter factor probably the most significant.

Congenital Nonspherocytic Hemolytic Anemia

Class I G6PD variants have such severe G6PDdeficiency that lifelong hemolysis occurs in theabsence of infection or drug exposure. Mosthave DNA mutations grouped around exon 10of the gene which code for amino acids affect-ing the glucose 6-phosphate or NADP+/NADPHbinding domains. These sites are central to thefunction of G6PD, and it is thought that thefunctional defect is so severe in patients in classI that the red cells are unable to withstand nor-mal stresses within the circulation. Detection ofanemia and jaundice normally occurs in theneonatal period. The degree of hyperbilirubine-mia in these neonates is frequently severeenough to require exchange transfusion to pre-vent kernicterus. Most individuals have chronichemolysis with mild-to-moderate anemia (hemo-globin of 8–10 g/dL) and a persistent reticulocy-tosis. Acute hemolysis can be aggravated byexposure to precipitants such as infection, favabeans, or drugs and chemicals with oxidant po-tential, as described above, resulting in acute hy-perhemolysis.Drugs with mild oxidant potentialthat can be safely given to individuals with classII or III variants can increase hemolysis in class Ivariants.

Generally, these individuals only have mildanemia as increased erythropoiesis compen-sates for the chronic hemolysis. The anemiamay worsen if there is reduced erythropoieticcapacity for any reason, such as hematin defi-ciency, infection, or parvovirus-induced aplasticcrises.


Table 11-2. The World Health Organization Classification of Glucose 6-phosphate DehydrogenaseActivity VariantsClass Level of Enzyme Deficiency Clinical Presentation

I Severe, < 10% Chronic hemolytic anemia

II (e.g., G6PD Mediterranean) Severe, < 10% Intermittent hemolysis

III (e.g., G6PD A−) Moderate, 10%–60% Intermittent hemolysis, usually precipitated

IV None Normal

V Increased activity Normal

In patients who are severely G6PD deficient,on rare occasions an associated abnormality oc-curs in which there is neutrophil dysfunctionsecondary to G6PD deficiency. This defect leadsto impaired neutrophil bactericidal activity,resulting in recurrent infections with catalase-positive organisms.

Favism

Favism usually results from the ingestion offresh, uncooked fava beans, although it can oc-cur with dried, cooked, canned, or frozen beans.It occurs most commonly in males between theages of 1 and 5 years. Clinical symptoms andsigns of acute intravascular hemolysis develop 5to 24 hours after fava bean ingestion. Precedingsymptoms of headache,nausea,back pain,chills,and fever are followed by hemoglobinuria andjaundice. The fall in hemoglobin is normallyacute and severe and can be fatal if blood trans-fusion is not performed.

The glucoside divicine (or its aglyconeisouramil) is the oxidant implicated in favism. Itoverwhelms the low GSH-generating capacity ofthe G6PD-deficient cells and directly compro-mises red cell function. The vast majority of in-dividuals developing favism are G6PD deficient,although only a small percentage of G6PD-deficient individuals are sensitive to the favabean. Also, the same individual’s response canvary. It is unclear why fava bean ingestion is notalways followed by hemolysis in G6PD-deficientindividuals. There are probably a number of fac-tors that are important in contributing to thephenotype. It is likely that hepatic metabolismof fava bean oxidants is important as well as theamount of divicine or isouramil in each favabean and how the fava beans are consumed.The degree of hemolysis is known to correlatewith the quantity of beans consumed and the in-testinal surface area of the individual. Favism ismost commonly associated with the G6PDMediterranean variant and therefore most com-monly occurs in Italy and Greece. AlthoughAfricans and African Americans with G6PD defi-ciency are less susceptible, favism neverthelessdoes occur with the G6PD A− variant.

Precipitants of Oxidative Hemolysis

Infection

Infection is probably the most common precip-itating factor for hemolysis in G6PD-deficient

individuals. The degree of hemolysis is normallymild, but severe intravascular hemolysis has oc-casionally been described. The mechanism ofaccelerated red cell destruction is not entirelyclear. It is likely that phagocytosing macrophagesgenerate oxidants while combating the infec-tion, with resulting red cell damage. Commonlyimplicated infectious agents are Salmonella,Escherichia coli, β-hemolytic streptococci, rick-ettsiae, and viral hepatitis.

Drugs

Compounds known to induce hemolysis inG6PD-deficient individuals are listed in Table11-1 (Avery, 1980; Koda-Kimble, 1978). All ofthese drugs interact with hemoglobin and oxy-gen, resulting in the intracellular formation ofH2O2 and other oxidizing radicals (e.g., O2

−,OH•). These oxidants accumulate in enzyme-deficient red cells with low GSH levels. This re-sults in oxidation of hemoglobin and otherproteins, leading to loss of function and celldeath. Some drugs are dangerous for all G6PD-deficient individuals, whereas some can begiven to individuals with less-severe deficiencyas they only cause modest shortening of thelifespan of their red blood cells.Also, for drugssuch as aspirin previously implicated in causingsevere hemolysis, it is likely that the infectionfor which the drug was given was the trueculprit.

THERAPY

The majority of patients with G6PD deficiencywill be unaware of the problem and go throughlife clinically undiagnosed. For patients whoare diagnosed with G6PD deficiency, their fam-ily members and offspring should be offeredscreening and counseling. Counseling shouldcover an explanation of the genetic back-ground of the condition, ways it can affecthealth, and how to recognize symptoms andsigns of hemolysis and should provide a list ofdrugs to avoid. Particular advice should begiven regarding travel where malaria prophy-laxis is required. The patient should be advisedto carry a Medicalert tag and be sure to adviseany health care professional with whom theyare dealing (doctor, nurse, pharmacist) of theirproblem. Patients should be given this writteninformation in an easily accessible form. TheInternet also provides a number of useful and


patient-friendly sites (e.g., http://rialto.com/g6pd/index.htm).

Treatment of a patient with G6PD deficiencyis dependent on the clinical syndrome of theindividual. All individuals known to be G6PDdeficient should avoid exposure to known pre-cipitants. This includes pregnant and lactatingwomen who are heterozygous for G6PD defi-ciency as in some cases drugs can cross theplacenta and be present in breast milk.

In acute hemolytic episodes, the use of theoffending drug should be stopped and any in-fection treated. Supportive measures, includingfolic acid, are normally sufficient. Occasionally,iron supplements may be required since iron islost in intravascular hemolysis. Transfusion isonly rarely required if there is impaired erythro-poiesis for some reason or if anemia causes car-diovascular compromise. The management ofacute renal failure associated with hemoglobin-uria does not differ from that of acute renalfailure from any other cause. This will requireappropriate management of fluid and elec-trolyte disturbance and addressing any cofac-tors, such as sepsis or drug toxicity.Hemodialysissometimes may be required, but the prognosisfor recovery is good.

The treatment of neonatal jaundice due toG6PD deficiency does not differ from that forother causes. Phototherapy and occasionally ex-change transfusion may be required to preventkernicterus (bilirubin neurotoxicity).Photother-apy results in photo-oxidation, configurationalisomerization, and structural isomerization ofbilirubin. The product of these reactions is lu-mirubin, a compound that can be efficiently ex-creted without hepatic conjugation. Exchangetransfusion rapidly removes serum bilirubin,but the procedure may need to be repeated toremove the residual bilirubin subsequently re-leased from the tissue-bound pool. With ap-propriate management, full recovery can beexpected, but where good medical care is notavailable, severe neurological disability will oc-cur.This therefore remains a major health issuein underdeveloped parts of the world whereG6PD deficiency is prevalent.

Both vitamin E supplements and splenec-tomy have been suggested as potential thera-peutic options, but neither has proven to bebeneficial. Gene therapy is a potential futuretreatment under development but would onlybe clinically justified in the most severe casesaffected by chronic hemolysis.

It is not common practice to screen routinelyfor this disorder as the milder variants are notclinically hazardous, and screening for the rarervariants is impractical. In high-prevalence popu-lations (Africans, Mediterranean) however,there should be a high index of suspicion inpatients presenting with appropriate clinicalsymptoms.

QUESTIONS

1. What are in the abnormalities in thechemical pathways that result in the clini-cal symptoms of G6PD deficiency?

2. What are the causes of false-negative as-says for G6PD deficiency?

3. How might the clinical presentation ofG6PD deficiency vary between individuals?

4. Why does the level of G6PD activityvary with the degree of maturity of theRBC?

5. By what mechanisms is G6PD deficiencyclinically apparent in females?

BIBLIOGRAPHY

Alving A, Carson P, Flanagan C, et al.: Enzymatic de-ficiency in primaquine-sensitive erythrocytes.Science 124:484–485, 1965.

Avery G: Drug Treatment. 2nd ed. ADIS Press,New York, 1980.

Beutler E: G6PD deficiency. Blood 84:3613–3636,1994.

Beutler E, Lisker R, Kuhl W: Molecular biology ofG6PD variants. Biomed Biochim Acta49:S236–S241, 1990.

Chen E, Cheng A, Lee A, et al.: Sequence of humanglucose-6-phosphate dehydrogenase cloned inplasmids and a yeast artificial chromosome. Ge-nomics 10:792–800, 1991.

Koda-Kimble M: Applied Therapeutics for ClinicalPharmacists. 2nd ed.Applied Therapeutics, SanFrancisco, 1978.

Luzzatto L, Gordon-Smith E: Inherited haemolyticanaemias, in Postgraduate Haematology. 4thed. Hoffbrand, AV, Arnold, London, 2001, pp.120–143.

Luzzatto L,Mehta A:Glucose-6-phosphate dehydro-genase deficiency, in Scriver C, Beaudet A, Sly W,Valle D (eds): The Metabolic and MolecularBases of Inherited Disease. 7th ed. McGraw-Hill, New York, 1996, pp. 3367–3398.

Pai G, Sprenkle J, Do T, et al.: Localization of locifor hypoxanthine phophoriboxyltransferase and


http://rialto.com/g6pd/index.htm

http://rialto.com/g6pd/index.htm

glucose-6-phosphate dehydrogenase and bio-chemical evidence of non-random X-chromo-some expression from studies of a humanX-autosome translocation. Proc Natl Acad SciU.S.A. 77:2810–2813, 1980.

Scriver C,Beaudet A,Sly W,et al.: The Metabolic andMolecular Bases of Inherited Disease. 7th ed.McGraw- Hill,New York,1995,pp.3367–3398.

Yoshida A, Beutler E: Glucose-6-phosphate Dehy-drogenase. Academic Press, New York, 1986.


CASE REPORT

A 6-month-old male was brought to the emer-gency department because the parents ob-served that he had “jerking” of his arms andlegs. They said that the child had been sleepingfor longer periods of time over the last few daysand had episodes of vomiting that were not re-lated to the time of feeding.He had been less re-sponsive than usual, and they were concernedbecause his rate of breathing had markedly in-creased this morning. The parents had calledtheir pediatrician, who recommended that theybring the child to see him. While preparing tobring the child to the physician’s office, thechild began jerking and “twitching.” Instead ofbringing their child to the pediatrician, the par-ents called the emergency medical service,which took him directly to the emergency de-partment of the local hospital. By the time theemergency medical team arrived, the child hadstopped seizing but was somnolent.

The child had had a normal and uneventfulprenatal and birth history. He was born after afull-term gestation and weighed 3.3 kg. Until afew days prior to the incident, he had no med-ical problems. His weight, length, and headcircumference had always followed the 40thpercentile on growth charts. He had attainednormal developmental milestones for age.Therewas no known consanguinity between the par-ents and no family history of similar problems,and there was a normal 3-year-old brother.

On physical examination, the child wasafebrile with mild tachypnea and tachycardia.

He was lethargic and exhibited no overt seizureactivity. He also had a fine eczemalike skin rasharound his eyes, mouth, and genital areas andpatchy loss of scalp hair (Fig. 12-1A). Neurologi-cally, he was hypotonic (poor muscle tone)in both the upper and lower extremities withdecreased deep tendon reflexes. His lung, car-diovascular, and abdominal examinations wereunremarkable. The remainder of his physicalexamination was normal.

On further questioning, the parents statedthat they thought the rash was just a diaperrash and had not noticed the redness aroundhis eyes and mouth. They related that just a fewdays prior they had found hair on the child’sbed but thought the hair loss was normal.

The major concern of the emergency de-partment physicians was the lethargy, hypo-tonia, and seizure activity. Initial laboratorystudies revealed that the child had a normalcomplete blood count and smear. Other bloodtests revealed metabolic acidosis with a bicar-bonate concentration of 11 mEq/L (normal is20–25 mEq/L) and an anion gap of 22 mEq/L(normal is < 15 mEq/L). His serum glucose, cal-cium, and magnesium concentrations were nor-mal. To exclude the diagnosis of meningitis, aspinal tap was performed. The cell counts andchemistries of the cerebrospinal fluid were nor-mal. The physicians considered that the childmight have sepsis and administered antibioticsand intravenous fluids. Prior to administrationof antibiotics, blood, urine, and cerebrospinalfluid were sent for bacterial culture.

A urinalysis showed no signs of infection, but

12

Biotinidase Deficiency:A Biotin-Responsive Disorder

BARRY WOLF

134

there was marked ketonuria. Plasma ammoniaconcentration was mildly elevated. Based onthe metabolic acidosis, increased base deficit,hyperammonemia in the presence of thelethargy, hypotonia, and decreased responsive-ness, the possibility of an inherited metabolic dis-order was considered. To screen for a variety ofmetabolic disorders, plasma amino acid concen-trations were tested and found to be normal.Uri-nary organic acid analysis showed increasedconcentrations of lactate, β-hydroxyisovalerate,β-methylcrotonylglycine,and methylcitrate.Thesefindings were consistent with multiple carboxy-lase deficiency (Fig.12-2).Based on these results,the child was given 10 mg of oral biotin (Fig.12-3) and 10 mg per day thereafter.

To determine the type of multiple carboxy-lase deficiency,blood was obtained to determinethe biotin holocarboxylase synthetase activity inleukocytes,and serum was sent to determine thebiotinidase activity.The results of the serum bio-tinidase activity returned first and indicated lessthan 1% of mean normal serum activity, confirm-ing that the child had profound biotinidase defi-ciency (less than 10% of mean normal serumbiotinidase activity). Subsequently, biotin holo-carboxylase synthetase activity was found to benormal. Although many states screen for bio-tinidase deficiency in the newborn period, thischild was born in a state where newborn screen-ing for biotinidase deficiency is not performed.

Over the next few days, the child becamemarkedly more alert and exhibited no seizure

activity.All the bacterial cultures were negativeafter 72 hours, and the antibiotics were discon-tinued. His skin rash disappeared in a week, andhair growth returned to the areas of alopecia(Fig. 12-1B). The child has had no further bio-chemical, neurological, or cutaneous abnormali-ties since the biotin was started.

Because ophthalmologic problems and hear-ing loss occur commonly in symptomatic chil-dren with profound biotinidase deficiency,extensive eye and auditory testing were per-formed. The child was found to have mild opticatrophy and moderate mid- and high-frequencyhearing loss.To ensure normal language devel-opment, he was given hearing aids.

The parents and the sibling were shown tohave about 50% of mean normal serum bio-tinidase activity, consistent with heterozygosity.Mutation analysis of DNA from the child revealedhomozygosity for a common missense mutation.

DIAGNOSIS

Symptoms, such as vomiting, respiratory prob-lems, hypotonia, and seizures, in young childrenoften are characteristic of various serious in-fections, including meningitis and sepsis. Al-though these are usually accompanied by fever,neonates with sepsis may be hypothermic oreuthermic (normal temperature). In the ab-sence of seizures,gastrointestinal obstruction andcardiorespiratory problems are in the differential

Biotinidase Deficiency 135

Figure 12-1. A child with biotinidase deficiency (A) prior to diagnosis and therapy and(B) after several months of biotin treatment.

Leucine

Oxaloacetate

PyruvateAlanine Lactate Propionate

Methylcitrate

3-Hydroxypropionate

Propionyl CoA

Glucose

ValineIsoleucineMethionineThreonineOdd-chain Fatty Acids

3-Methylcrotonyl CoA3-Methylcrotonylglycine

3-Hydroxyisovalerate

3-Methylglutaconyl CoA

Citrate Succinate

TCACycle

Methylmalonyl CoA

Acetyl CoAMalonyl CoAFatty AcidsPCC

ACC

PC

βMCC

Figure 12-2. Metabolic pathways involving the four biotin-dependent carboxylases.The solid rectangular blocks indicate the locations of the enzymes: ACC, acetyl-CoA car-boxylase; βMCC, β-methylcrotonyl-CoA carboxylase; PC, pyruvate carboxylase; PCC,propionyl-CoA carboxylase. Isolated deficiencies of the first three carboxylases (mito-chondrial) have been established (isolated ACC deficiency has not been confirmed).Atleast the activities of the three mitochondrial carboxylases can be secondarily deficientin the untreated multiple carboxylase deficiencies, biotin holocarboxylase synthetasedeficiency and biotinidase deficiency.Lowercase characters indicate metabolites that arefrequently found at elevated concentrations in urine of children with both multiple car-boxylase deficiencies.The isolated deficiencies have elevations of those metabolites di-rectly related to their respective enzyme deficiency.

diagnosis.However,when these symptoms occurtogether with biochemical abnormalities, suchas metabolic acidosis, ketonuria, or hyper-ammonemia,an inborn error of metabolism mustbe suspected. In addition, occasionally there aresymptoms that are more specific for particularinherited metabolic disorders, such as skin rashand alopecia that lead the physician to the cor-rect diagnosis. Plasma amino acid and urinary or-ganic acid analyses are helpful screening tools toinclude or exclude disorders from the largegroup of possible inborn errors of metabolism.

Urinary organic acid analysis is useful for dif-ferentiating isolated carboxylase deficienciesfrom the biotin-responsive multiple carboxy-lase deficiencies. β-Hydroxyisovalerate is themost common urinary metabolite observedin isolated β-methylcrotonyl-CoA carboxylasedeficiency, biotinidase deficiency, biotin holo-carboxylase synthetase deficiency, and acquiredbiotin deficiency. In addition to β-hydroxy-isovalerate, elevated concentrations of urinarylactate, methylcitrate, and β-hydroxypropionateare indicative of multiple carboxylase deficiency.

Children with any of the isolated carboxylasedeficiencies do not improve with biotin supple-mentation, whereas those with multiple car-boxylase deficiency do.A trial of biotin is oftenexpedient and useful in discriminating betweenthe isolated carboxylase deficiencies and themultiple carboxylase deficiencies. Isolated car-boxylase deficiencies can be definitively con-firmed by demonstrating deficient enzymeactivity of one of three mitochondrial carboxy-lases in extracts of peripheral blood leukocytes(prior to biotin therapy) or cultured fibroblasts,whereas the activities of the other two carboxy-lases are normal.

The organic acid analysis in the urine of thischild was consistent with biotin deficiency ormultiple carboxylase deficiency. Biotin defi-ciency usually can be excluded unless there is ahistory of dietary indiscretion, such as consum-ing a diet containing raw eggs or few biotin-containing foods, or there is a history ofprolonged parenteral hyperalimentation with-out biotin supplementation. Low serum biotinconcentrations can be useful in differentiating


d-Biotin

COO−CH2CH2

CH2H2C CH

CH

C

O

HC

HN NH

CH2

S

N-Carboxybiotin Lysyl Residue

EnzymeCH2 CH2 CH2 CHCH2

CH2 CH2 CH2 CH2H2C CH

CH

C−OOC

O

HC

N NH

NH

NH

S

C

O

O

C

Figure 12-3. The structure of d-biotin and N-carboxybiotin covalently attached to an ε-amino group of a lysyl residue of a carboxylase. Biocytin is the biotin similarly attachedto the ε-amino group of a lysine. The dotted line indicates the amide bond linkage.

biotin and biotinidase deficiencies from biotinholocarboxylase synthetase deficiency, butthese determinations are laborious, only per-formed in a few laboratories, and depending onthe methodology, can provide uninformative in-formation. For example, only the methods thatcan distinguish biotin from biocytin or boundbiotin provide useful information. Methods thatmeasure total biotin and biotin plus biotin de-rivatives may not indicate biotin deficiency.

Children with both biotin holocarboxylasesynthetase deficiency and biotinidase defi-ciency often exhibit similar nonspecific clinicalfeatures, such as feeding problems, vomiting,hypotonia, and lethargy. Other symptoms char-acteristic of biotin-responsive multiple carboxyl-ase deficiencies, such as skin rash or alopecia,can lead more rapidly to the correct diagnosis,but also can be seen in children with zinc or es-sential fatty acid deficiencies. Frequent viral,bacterial, or fungal infections, due to immuno-logical dysfunction, may occur in children withmultiple carboxylase deficiency. On the otherhand, an infection may be the stress than pre-cipitates the symptoms of a metabolic disorder.Children with biotin holocarboxylase syn-thetase deficiency may have metabolic acidosis,a large anion gap, and elevated concentrationsof lactate in the serum and urine.An amino acidanalysis may reveal hyperglycinemia,which alsocan occur in other organic acidemias.

Biotinidase deficiency must be differentiatedfrom biotin holocarboxylase synthetase defi-ciency.The majority of patients with biotinidasedeficiency exhibit metabolic lactic acidosis, ke-tosis, and organic aciduria similar to that seen inbiotin holocarboxylase synthetase deficiency.Patients with both disorders have elevated con-centrations of urinary β-hydroxyisovalerate andβ-methylcrotonylglycine. However, some symp-tomatic patients with biotinidase deficiency(about 20%) do not excrete the characteristicorganic acids, even when they are seriously illand metabolically compromised.Therefore, thepresence of normal urinary organic acids, evenwhen the child is symptomatic, does not ex-clude a diagnosis of biotinidase deficiency.Patients with both disorders may have mildhyperammonemia. When neurological or cuta-neous symptoms are suggestive of multiplecarboxylase deficiency, appropriate enzymetesting should be performed.

The age of onset of symptoms may be usefulin discriminating between these two multiplecarboxylase deficiencies. Although children

with biotin holocarboxylase synthetase defi-ciency usually exhibit symptoms before 2months of age and whereas those with bio-tinidase deficiency usually manifest symptomsafter 3 months of age, there are, however, ex-ceptions for both disorders.

Both multiple carboxylase deficiencies arecharacterized by deficient activities of the threemitochondrial carboxylases in peripheral bloodleukocytes prior to biotin treatment. The car-boxylase activities increase to near normal ornormal after treatment with pharmacologicaldoses of biotin. Patients with biotin holocar-boxylase synthetase deficiency have deficientactivities of the three mitochondrial carboxy-lases in fibroblasts incubated in medium withlow biotin concentrations (containing only thebiotin contributed by fetal calf serum added tothe medium for cell growth), whereas patientswith biotinidase deficiency have normal car-boxylase activities under these conditions.The activities of the carboxylases in biotinholocarboxylase synthetase deficiency becomenear normal to normal when cultured inmedium supplemented with high concentra-tions of biotin.

Biotinidase deficiency and biotin holocar-boxylase synthetase deficiency can be defini-tively diagnosed by direct enzymatic assay.Biotinidase activity in plasma or serum is usu-ally determined by using the artificial substrate,biotinyl-p-aminobenzoate. If biotinidase activityis present, then biotin is cleaved, releasing p-aminobenzoate.The p-aminobenzoate then isreacted with reagents that result in the develop-ment of purple color that can be quantitatedcolorimetrically. In the absence of biotinidaseactivity, p-aminobenzoate is not liberated. Bio-tinidase activity in patients with an isolated car-boxylase deficiency or biotin holocarboxylasesynthetase deficiency is normal.

Plasma biotin concentrations may be defi-cient in patients with biotinidase deficiency,butalso can be normal prior to therapy.Again, it isimportant to be certain that the method used todetermine the biotin concentration measuresfree biotin and does not also measure biotin de-rivatives, such as biocytin.

The age of onset of symptoms of childrenwith profound biotinidase deficiency variesfrom several months to 10 years old, with amean age of presentation between 3 and 6months old. The most common neurologicalfeatures of this disorder are seizures, hypotonia,and ataxia. Myoclonic seizures are the most


common type of seizure, but some childrenhave exhibited grand mal and focal types, and afew have been described as infantile seizures.Some children have exhibited breathing abnor-malities, such as hyperventilation, laryngeal stri-dor, and apnea. One child who was initiallydiagnosed as having sudden infant death syn-drome was retrospectively found to have bio-tinidase deficiency.

Older children with biotinidase deficiencyoften exhibit ataxia and developmental delay.Sensorineural hearing loss and eye problems,such as optic atrophy, have been described inmany untreated children. A recent study hasshown that 76% of symptomatic children withprofound biotinidase deficiency have sen-sorineural hearing loss. Frequently, these chil-dren have cutaneous symptoms, including skinrash and alopecia. Several patients have had cel-lular abnormalities, manifested by fungal infec-tion. These immunological aberrations arevaried and may represent the effects of abnor-mal organic acid metabolites or biotin defi-ciency on normal immunological function.Biotin therapy rapidly corrects the immunologi-cal dysfunction. Some children with biotinidasedeficiency only exhibit one or two of these fea-tures, whereas others have many of the neuro-logical and cutaneous findings.

Children who are diagnosed and treatedwith biotin at birth, before developing symp-toms, or who did not experience recurrentepisodes of metabolic compromise will usuallyremain asymptomatic. Severe neurologicalproblems most frequently occur in those chil-dren who have had recurrent symptoms andmetabolic compromise. The irreversible symp-toms include deafness, optic atrophy, and devel-opmental delay.


Biotin, an essential water-soluble B-complex vi-tamin, is the coenzyme for four human carboxy-lases (Fig. 12-2): These include the threemitochondrial enzymes: pyruvate carboxylase,which converts pyruvate to oxaloacetate and isthe initial step of gluconeogenesis; propionyl-CoA carboxylase, which catabolizes severalbranched-chain amino acids and odd-chain fattyacids; and β-methylcrotonyl-CoA carboxylase,which is involved in the catabolism of leucine;and the principally cytosolic enzyme, acetyl-CoA carboxylase, which is responsible for the

first, committed step in the synthesis of fattyacids (Fig. 12-2). Therefore, these enzymes areinvolved in carbohydrate, protein, and lipid me-tabolism.

Biotin is attached to the various apocarboxyl-ases by the enzyme biotin holocarboxylasesynthetase forming holocarboxylases throughthe two partial reactions shown below:

A. Biotin + ATP ←→ Biotinyl-5'-AMP + PPi

B. Biotinyl-5'-AMP + Apocarboxylase → Holocarboxylase + AMP

Net: Biotin + ATP + Apocarboxylase → Holocarboxylase + AMP + PPi

All four carboxylases bind a carbon dioxidemoiety to a nitrogen atom in the heterocyclicportion of biotin and transfer the carbon diox-ide to an acceptor molecule through the follow-ing two partial reactions:

A. Enzyme-Biotin + ATP + HCO3−

←→ Enzyme-Biotin-CO2 + ADP + Pi

B. Enzyme-Biotin-CO2 + Acceptor←→ Enzyme-Biotin + Acceptor-CO2

Net: HCO3− + ATP + Acceptor

←→ Acceptor-CO2 + ADP + Pi

The carboxyl group of biotin is linked by anamide bond to an ε-amino group of a specific ly-sine residue of the apoenzymes (Fig. 12-3).Afterthe holocarboxylases are degraded proteolyti-cally to biocytin (biotinyl-ε-lysine) or small bi-otinyl peptides, biotinidase cleaves the amidebond, releasing lysine or lysyl peptides and freebiotin, which can then be recycled (Fig. 12-4).Biotinidase also plays a role in the processing ofprotein-bound biotin, thereby making the vita-min available to the free biotin pool. Recentstudies suggest that biotinidase also plays a rolein the transfer of biotin from biocytin to spe-cific proteins, such as histones.

Inherited isolated deficiencies of the threemitochondrial biotin-dependent carboxylaseswere described during the 1970s (Fig. 12-2).Children with each of the isolated deficienciesexhibit neurological symptoms during infancyor early childhood associated with metaboliccompromise caused by the accumulation of ab-normal metabolites resulting from the respectiveenzyme block. Each isolated deficiency is dueto a structural abnormality in the respective mi-tochondrial enzyme, whereas the activities of


the other carboxylases are normal. Each disor-der is treated by dietary restrictions, but all failto improve following treatment with pharmaco-logical doses of biotin.

In 1971, a child with biotin-responsive β-methylcrotonylglycinuria was reported.This indi-vidual had metabolic ketoacidosis and elevatedconcentrations of urinary β-methylcrotonic acidand β-methylcrotonylglycine. Several days afterbeing given oral biotin, his symptoms resolved,and the urinary metabolites cleared. He subse-quently was shown to have deficient activities ofall three mitochondrial carboxylases in his pe-ripheral blood leukocytes and skin fibroblasts.This was the first child to be diagnosed withwhat was called multiple carboxylase deficiency.

Additional children with multiple carboxy-lase deficiency were reported. Initially, thesechildren were classified as having either theearly-onset form (referred to as the neonatal orinfantile form) or the late-onset form (referredto as the juvenile form) of multiple carboxylasedeficiency, depending on the age of onset of

symptoms. Most of the children with the early-onset form ultimately were shown to have bi-otin holocarboxylase synthetase deficiency.

Two children with the late-onset form ini-tially were reported as having a defect in intes-tinal transport of biotin. This conclusion wassupported by finding low plasma biotin con-centrations when these children were adminis-tered oral biotin compared to the concentrationsof plasma biotin of unaffected control subject.In 1983, it was demonstrated that the primarybiochemical defect in most patients with late-onset multiple carboxylase deficiency was a de-ficiency of serum biotinidase activity. The twochildren with a putative defect in intestinalbiotin transport both were confirmed to havebiotinidase deficiency. This disparity was recon-ciled by demonstrating that, in both cases, thechildren were biotin depleted at the time thebiotin-loading studies were performed. There-fore, when the children initially were givenbiotin, although the vitamin was transportedinto the blood normally, it was rapidly taken up


ProteolyticDegradation

Dietary Biotin

FreeProtein-Bound

Biotin

Biocyclin

Apocarboxylases(PCC, MCC, PC, ACC)

HolocarboxylaseSynthetase

Holocarboxylases

LipidsFatty AcidSynthesis

CarbohydratesGluconeogenesis

ProteinsAmino AcidCatabolism

Lysine orLysyl-Peptides

HOC

O

O

H

N N

TheBiotinCycle

Biotinidase

Biotinidase O

H

N

N

NH2

HCOOHN

C

O

O

H

N

NH

N

C

O

Figure 12-4. The biotin cycle shows the actions of biotin holocarboxylase synthetase inbiotinylating carboxylases and of biotinidase in cleaving biocytin, thereby recycling biotin.

by the tissues, thereby resulting in low plasmabiotin concentrations. When these childrenwere replete in biotin, their plasma biotin con-centrations increased normally in response tothe biotin. These latter studies indicated thatneither child had a defect in intestinal biotintransport.

Children with profound biotinidase activityhave less than 10% of mean normal serum en-zyme activity. Deficient biotinidase activity hasalso been demonstrated in extracts of leuko-cytes and fibroblasts. At least one patient wasshown to have deficient biotinidase activity inhis liver. More than 300 symptomatic individu-als have been reported with biotinidase defi-ciency. The parents of these children usuallyhave serum enzyme activities intermediate be-tween those of the patients and those of nor-mal individuals.

Biotinidase deficiency may present with vary-ing degrees of metabolic acidosis,ketosis,and hy-perammonemia.The metabolic acidosis is usuallyaccompanied by an increased anion gap andlactate elevations.The ketosis is due to the accu-mulation of abnormal organic acid metabo-lites, such as propionate and lactate, in blood,and β-hydroxypropionate, methylcitrate, β-hydroxyisovaleric acid, β-methylcrotonylglycine,and β-hydroxybutyrate in urine. Hyperammone-mia plays a major role in causing the lethargy,somnolence, and coma that can occur in thedisorder. The hyperammonemia is due to thesecondary inhibition of N-acetylglutamate syn-thetase, which produces N-acetylglutamate, theactivator of carbamyl phosphate synthetase inthe urea cycle.

Individuals with untreated biotinidase defi-ciency develop biotin deficiency because theycannot recycle endogenous biotin. The biotindeficiency subsequently results in the lack ofsubstrate for biotin holocarboxylase synthetase.Without the availability of biotin to be added tothe apocarboxylases, multiple carboxylase defi-ciency occurs, and the abnormal metabolitesaccumulate.

Serum biotinidase activity is not altered by bi-otin deficiency. This was demonstrated by find-ing normal serum biotinidase activity in severalpatients who became biotin deficient while be-ing treated with parenteral hyperalimentationlacking biotin. Biotinidase appears to play an im-portant role in the processing of protein-boundbiotin, either by being secreted into the intes-tinal tract, where it can release biotin from

bound dietary sources that can subsequently beabsorbed, or by cleaving biocytin or biotinylpeptides in the intestinal mucosa or in blood.

Most patients with biotinidase deficiency ex-crete large quantities of biocytin in their urine,but there has been no evidence of accumula-tion of this metabolite in tissues. It remains tobe determined whether biotin therapy is harm-ful because it may increase the concentration ofbiocytin in these children.

Biotinidase activity in cerebrospinal fluid andthe brain is very low. This suggests that thebrain may not recycle biotin effectively and de-pends on biotin transported across the blood-brain barrier. Several symptomatic childrenwho have failed to exhibit peripheral lacticacidosis or organic aciduria have had elevatedlactate or organic acids in their cerebrospinalfluid. This compartmentalization of the bio-chemical abnormalities may explain why theneurological symptoms usually appear beforeother symptoms. Peripheral metabolic keto-acidosis and organic aciduria subsequently occurwith prolonged metabolic compromise.

Most patients with biotinidase activity whohave hearing loss had this problem beforebeginning biotin therapy. Hearing loss is usuallyirreversible, although several young affectedchildren have shown some improvement inhearing with biotin therapy. Some researchershave speculated that the biocytin or biotinylpeptides accumulate and are toxic and even ag-gravate hearing deficits. This remains to beproven, however. Recently, biotinidase has beenlocalized to discrete regions of the brain andcochlea, including the hair cells and spiral gan-glion. Lack of biotinidase in these areas likelyhas a direct role in causing the hearing loss.

Electroencephalographic findings in individ-uals with untreated profound biotinidase defi-ciency have ranged from normal to markedlyabnormal and are usually nonspecific. Demyelin-ation and other cerebral abnormalities havebeen seen in some children with biotinidase de-ficiency. These findings sometimes improve fol-lowing biotin treatment.

Human serum biotinidase has been purifiedto homogeneity and characterized. The cDNAthat encodes for the enzyme and the genomicorganization have been elucidated. The genethat encodes for the serum enzyme has been lo-calized to chromosome 3p25. Over 100 differ-ent mutations that cause biotinidase deficiencyhave been identified.


THERAPY

A child with biotinidase deficiency should beadequately hydrated to facilitate the excretionof the abnormal organic acid metabolites (Fig.12-2). Adequate nutrition is essential. In theacute stage of the disorder, protein may be re-stricted. It is imperative to supply sufficientcalories in the form of parenteral glucose ororal polysaccharides, such as polycose. Severeacidosis may initially require bicarbonate sup-plementation.

The mainstay of therapy in biotinidase defi-ciency is biotin supplementation. To date, allsymptomatic children with biotinidase defi-ciency have improved after treatment with 5 to10 mg of biotin per day. Biotin appears to be re-quired in the free form as opposed to thebound form. This is based on the findings oftwo children who were fed yeast as a form oftherapy. Neither improved because essentiallyall of the biotin in yeast is protein bound, andthese children could not recycle the biotin.These children, however, did improve whentreated with free biotin. Treatment with biotinis essential and sufficient to prevent or resolvethe symptoms. It is not necessary to treat chil-dren with biotinidase deficiency with protein-restricted diets as it is in some of the isolatedcarboxylase deficiencies because with biotintherapy all the carboxylase activities are normal.Symptoms of biotinidase deficiency are prevent-able if patients are diagnosed and treated atbirth or before symptoms occur.

If the child with profound biotinidase defi-ciency is symptomatic, then the biochemicalabnormalities and seizures resolve rapidly afterbiotin treatment. This is usually followed byimprovement of the cutaneous abnormalities.Hair growth returns over a period of weeks tomonths in the children with alopecia. Opticatrophy and hearing loss are usually resistantto therapy, especially if several months haveelapsed between the time of diagnosis and theinitiation of treatment. Some treated childrenhave rapidly regained lost developmental mile-stones, whereas others have continued to showdeficits.

Biotinidase deficiency met most of the crite-ria for inclusion in a neonatal screening pro-gram, such as ease and efficacy of treatment,frequency of the disorder, and the relativelylow cost of testing. Therefore, a simple colori-metric method to determine biotinidase activ-ity using the same soaked filter-paper blood

spots required for other neonatal screeningtests was developed. Usually, this colorimetricenzymatic assay uses the same reagents de-scribed for confirmational testing. Currently,more than 25 countries and 25 states in theUnited States screen for biotinidase deficiencyin the newborn period. The frequency of bio-tinidase deficiency is about 1 in 60,000 new-borns.

Biotinidase activity can be measured in cul-tured amniotic fluid cells and in amniotic fluid.Therefore,prenatal diagnosis of biotinidase defi-ciency is possible. Prenatal diagnosis has beenperformed in two at-risk pregnancies in whichamniocentesis was performed because of ad-vanced maternal age. The fetuses were found tobe unaffected, and this was confirmed afterbirth. In addition to enzyme determination inamniocytes, a fetus was correctly shown to be aheterozygote by molecular mutation analysis inan at-risk pregnancy. Because treatment is so ef-fective in this disorder, some laboratories arenow performing prenatal diagnosis.

Biotinidase deficiency represents a readily di-agnosable inherited metabolic disorder forwhich the symptoms can be prevented or ame-liorated with simple, direct therapy. If a childmust have an inborn error of metabolism, thenbiotinidase deficiency is the disorder to have.

QUESTIONS

1. Although there have been reports of in-herited isolated deficiencies of the threemitochondrial biotin-dependent carboxy-lases, there have not been any confirmedreports of inherited isolated acetyl-CoAcarboxylase deficiency.What is a possibleexplanation for why isolated acetyl-CoAcarboxylase deficiency has not been ob-served?

2. Some children with profound biotinidasedeficiency exhibit symptoms, such asseizures, but do not have metabolicacidemia or elevations of the characteris-tic organic acids in their plasma or blood.What is a possible explanation for this ob-servation?

3. Biotinidase activity in newborn screeningand in confirmational testing is usually de-termined by using the artificial substratebiotinyl-p-aminobenzoate in the colori-metric assay.What are the ramifications ofnot using the natural substrate biocytin to


identify children with biotinidase defi-ciency?

4. Children with biotinidase deficiency can-not cleave biocytin. It has been suggestedthat biocytin may accumulate in tissues.Can you design an experiment to deter-mine if biocytin is toxic or harmful?

5. Biotinidase activity in extracts of brain tis-sue is very low (when activity is expressedon a per-mg of-protein basis) relative tothe activity in plasma, liver,or kidney.Whatmight this finding suggest about the distri-bution of biotinidase in the brain?

6. Children with biotinidase deficiency areusually treated with 5–10 mg of oral bi-otin. We often use the same dose for in-fants as we do for older children. Thisobviously means that the dose of biotinper body weight is decreasing as the childgets older. How might one determine ifthe dose of biotin is adequate or if higherdoses are necessary?

7. Yeast contains large concentrations of bi-otin or biotinyl derivatives. However,when yeast extracts are administered tosymptomatic children with profound bio-tinidase deficiency, there has been no clin-ical improvement. What is a possibleexplanation for this observation?

8. A mouse can be made biotin deficient byfeeding the animal food lacking biotin andsupplemented with the egg white proteinavidin, which strongly binds biotin, for 8weeks. If the animal is then given radiola-

beled biotin, as much as 50% of the biotinlocalizes to the nucleus of hepatocytes.What is a possible role of biotin in thenucleus?

BIBLIOGRAPHY

Cole H, Reynolds TR, Buck GB, et al.: Human serumbiotinidase: cDNA cloning, sequence and charac-terization. J Biol Chem 269:6566–6570, 1994.

Hymes J,Wolf B: Biotinidase and its roles in biotinmetabolism. Clin Chim Acta 255:1–11, 1996.

Wolf B: Disorders of biotin metabolism, in ScriverCR, Beaudet AL, Sly WS,Valle D (eds):The Meta-bolic Basis of Inherited Disease. McGraw-Hill,New York, 1992a, pp. 2083–2103.

Wolf B: Disorders of biotin metabolism: treatableneurological syndromes, in Rosenberg R,Prusiner SB, DiMauro S, Barchi RL, Kunkel LM(eds):The Molecular and Genetic Basis of Neu-rological Disease. Butterworth, Stoneham, MA,1992b, pp. 569–81.

Wolf B: Biotinidase deficiency: new directions andpractical concerns. Curr Treat Options Neurol5:321–328, 2003.

Wolf B, Feldman GL: The biotin-dependent car-boxylase deficiencies. Am J Hum Genet34:699–716, 1982.

Wolf B, Grier RE, Allen RJ, et al.: Biotinidase defi-ciency: the enzymatic defect in late-onset multi-ple carboxylase deficiency. Clin Chim Acta131:273–281, 1983.

Wolf B, Heard GS, Weissbecker KA, et al.: Bio-tinidase deficiency: initial clinical features andrapid diagnosis. Ann Neurol 18:614–617, 1985.


CASE REPORT

An 11-year-old Caucasian male was referred forevaluation of deteriorating school performanceand the recent onset of ataxic (i.e., wide-basedand unsteady) gait. He had been previously di-agnosed with attention deficit disorder and hadbeen treated with Ritalin for the past 3 years.There was no history of trauma or infection. Hecame to attention after his fifth grade teachernoted that his handwriting skills had declinednoticeably from the end of fourth grade, andthat there was a general decline in his cognitiveabilities, including difficulties with mathemati-cal calculations. In addition, his mother re-ported that he was having difficulty reading,had become clumsy, and was displaying behav-ioral changes.

On physical examination, he had slightlyslurred speech, was unable to follow complexcommands, and had right-left disorientation. Hisgait was abnormal with circumduction (i.e., ac-tive circular movement of the leg) on the rightand a steppage gait, right-sided weakness, anda positive Romberg sign (falling forward wheneyes are closed, and feet are close together dueto loss of position sensation).Further neurologi-cal assessment revealed dysgraphia (i.e., inabil-ity to write) and dyscalculia (i.e., inability toperform mathematical calculations). No otherphysical findings were present. Magnetic reso-nance imaging (MRI) of the brain revealeddiffuse abnormal signal in the white matter ofthe posterior hemisphere extending anteriorly(Fig. 13-1).

Based on the clinical history and findings andthe results of the brain MRI, the diagnosis of

X-linked adrenoleukodystrophy (X-ALD) wassuspected. Plasma was obtained for the mea-surement of very long chain fatty acids(VLCFA), which revealed a C26:0 concentrationof 1.32 µg/mL, a C24:0/C22:0 ratio of 1.88, anda C26:0/C22:0 ratio of 0.08 (normal is < 0.02),confirming the biochemical diagnosis of X-ALD.Subsequent formal neuropsychological testingwas obtained and revealed a performance IQ of70. Adrenal function was assessed by cortico-tropin (ACTH) stimulation test and revealed ad-renal insufficiency that was treated withreplacement therapy with hydrocortisone.

The family history was significant for a ma-ternal uncle, who had died at the age of 18years from sepsis and who was reported tohave bronze skin, both of which presumably re-sulted from unrecognized adrenal insufficiency,and for another maternal uncle, who had beendiagnosed with mental illness and was living inan adult residence. In addition, the 70-year-oldmaternal grandmother had been confined to awheelchair for almost 15 years due to weakness,the cause of which had not been previouslyidentified. Biochemical studies to detemineVLCFA ratios confirmed the carrier status of thepatient’s mother and revealed that his only sib-ling, a 9-year-old brother, had the biochemicaldefect.Two of his three maternal aunts also werefound to be carriers of X-ALD, one of whom wasfound to have an affected 1-year-old son.

After the diagnosis of X-ALD was established,the patient was placed on a regimen of 45 ccdaily (qd) of Lorenzo’s oil, 30 cc qd glycerol tri-oleate oil, and dietary restriction of VLCFA,which resulted in normalization of the plasmaVLCFA. His family had decided to pursue bone

13

Adrenoleukodystrophy

MARGARET M. MCGOVERN

144

marrow transplantation (BMT) using marrow de-rived from his father,who was a histocompatible,antigen-identical match. In the post-transplantperiod, he continued to decline neurologically,lost all speech, and became progressively de-mented. Several months later he succumbed toan infection.

DIAGNOSIS

X-ALD is a clinically heterogeneous disorderthat can result in progressive nervous systemmanifestations, adrenal insufficiency, and in-

volvement of the testis. The diagnosis shouldbe considered in boys who present with be-havioral disturbances, vision loss, deterioratinghandwriting, incoordination, and other pro-gressive neurological deficits. A concomitanthistory of attention deficit disorder is fre-quently present. A family history and pedigreeshould be obtained to look for a pattern ofX-linked transmission of disease, although ahigh index of suspicion must be employedsince the phenotype can vary markedly evenamong affected members of the same family.The diagnosis of X-ALD also should be consid-ered in men who present with progressive gait

Adrenoleukodystrophy 145

Figure 13-1. MRI of brain in a patient with X-linked adrenoleukodystrophy. Brain mag-netic resonance image obtained shortly after diagnosis shows diffuse abnormal signal inthe white matter of the posterior hemisphere extending anteriorly.

disorders and abnormalities of sphincter con-trol and sexual dysfunction, which characterizea later onset form of the disease called adreno-myeloneuropathy (AMN). In addition, specificbiochemical testing should be obtained in maleswith primary adrenocortical insufficiency evenif no neurological symptoms are present.

Clinical studies that contribute to the diagno-sis include a thorough neurological examinationand MRI of the brain in patients with neurologi-cal symptoms. Brain MRI is always abnormal inneurologically symptomatic males.The imagingfindings usually include a characteristic patternof symmetrical enhanced T-2 signal in theparieto-occipital region with contrast enhance-ment at the advancing margin.

Definitive diagnosis is achieved by the mea-surement of VLCFA in plasma.The results of thisanalysis are abnormal in nearly all males withX-ALD. The specific analyses that are includedare the concentration of C26:0, the ratio ofC24:0/C22:0, and the ratio of C26:0/C22:0, all ofwhich are elevated in most cases. Increased con-centration of VLCFA in plasma and/or culturedskin fibroblasts also is present in approximately85% of female carriers, although identificationof carriers is now available using molecularmethods which are more reliable. Testing of at-risk family members is recommended and canidentify affected males who may be candidatesfor therapy (see Therapy). In addition, it facili-tates assessment of adrenocortical function inbiochemically affected males, permitting treat-ment of those with adrenal insufficiency. Thistesting is accomplished by ACTH stimulation.


Historical Perspective and Clinical Phenotypes

X-ALD is a disorder of peroxisomal β-oxidation ofVLCFA and is estimated to occur in approxi-mately 1:20,000 males.The first description of anaffected patient was in 1910.This patient was apreviously healthy boy who developed visual dis-turbances at the age of 6 years, followed by dete-rioration of his schoolwork,and then spastic gait.His older brother had died of a similar illness inchildhood. The brain pathology in this child re-vealed diffuse loss of myelin in the cerebral hemi-spheres. Involvement of the adrenal gland wassubsequently reported in later-identified patients,

and the name adrenoleukodystrophy (ALD) wasproposed in 1970.

Studies of the adrenal gland revealed thepresence of lamellar inclusions and led to therecognition that X-ALD involved the abnormalstorage of lipids. Identification of the stored ma-terial as cholesterol esterified with VLCFA led tothe recognition that patients with X-ALD haveimpaired degradation of VLCFA. This discoveryresulted in X-ALD being classified as a peroxiso-mal disorder. It was hypothesized that the de-fect was an enzymatic deficiency of VLCFA-CoAsynthase, the first enzymatic step in the metab-olism of VLCFA, although this theory was laterdisproven when the gene was identified in1993 (see below). Nevertheless, the identifica-tion of the biochemical defect allowed the es-tablishment of a diagnostic assay by quantitationof VLCFA in plasma.

The availability of a diagnostic assay led to theidentification of patients with the biochemicaldefect who had varying phenotypic presenta-tions that have now been delineated as severaldistinct X-ALD phenotypes (Table 13-1). Thesephenotypes range from a severe childhood cere-bral form with progressive demyelination of thebrain, to a milder adult form, adrenomyelo-neuropathy (AMN), that is slowly progressiveand limited to the spinal cord, and an evenmilder form characterized by Addison’s diseaseonly, with no neurodegeneration.The childhoodform,which accounts for 35% of patients,usuallypresents between the ages of 4 and 8 years withbehavioral or learning deficits, often diagnosedas attention deficit disorder or hyperactivity, fol-lowed by neurological symptoms suggestive of amore serious underlying disorder.

The neurological symptoms can include inat-tention, deterioration of handwriting skills, di-minishing school performance, difficulty inunderstanding speech; reading difficulties, clum-siness; and visual disturbances. In some patients,behavioral disturbances predominate and can in-clude aggression.The rate of disease progressionis highly variable, with some patients progress-ing very rapidly and succumbing within severalyears, and others surviving for years, albeit withsignificant deficits.The cerebral phenotype is notonly observed in childhood but may also presentlater in adolescence (adolescent cerebral ALD)or in adulthood (adult cerebral ALD).

AMN, which accounts for 45% of cases, usu-ally presents in the second or third decade withprogressive stiffness and weakness of the legs,


abnormalities of sphincter control, and sexualdysfunction.A subset of these patients also haveabnormal findings on MRI, and in a small num-ber, the brain disease can be progressive, leadingto cognitive and behavior abnormalities.The dis-order also can manifest as Addison’s diseaseonly, which presents with unexplained vomitingand weakness or coma and increased skin pig-mentation resulting from excessive ACTH secre-tion. A major pathophysiological differencebetween the cerebral and noncerebral forms ofthe disease is the presence of an inflammatoryreaction in the cerebral white matter in the for-mer, which resembles multiple sclerosis.

In addition to the manifestations in affectedmales, some carrier females also display diseasesymptoms. In particular, up to 40% of womenhave mild-to-moderate spastic paraparesis (i.e.,weakness and stiffness of the legs) in middleage or later. Although adrenal insufficiency infemales has been reported, it is rare, as is cere-bral involvement.

Peroxisomal Function andPeroxisomal Diseases

X-ALD is a disorder of peroxisomal function.Normally, the peroxisome is the cellular site forβ-oxidation of VLCFA and is also involved in thesynthesis of bile acids, cholesterol, and plas-malogens. β-Oxidation of some long-chain fattyacids also occurs in the peroxisome, althoughmost long-chain fatty acids are metabolized inthe mitochondria. The peroxisome also playsa role in amino acid and purine metabolism.There are a number of genetic disorders associ-ated with defects in the peroxisome. These in-clude defects in genes that encode specificproteins, usually ones with enzymatic activity,and peroxisomal biogenesis defects that affectall of the metabolic pathways of the organelle.The former group includes hyperoxaluria type I

due to alanine:glyoxylate aminotransferase defi-ciency; Refsum’s disease resulting from defi-ciency of phytanoyl-CoA hydroxylase; X-ALDdue to deficiency of a membrane protein as de-scribed below; rhizomelic chondrodysplasiapunctata types II and III due to dihydroxy-acetone phosphate acyltransferase deficiency;and the β-oxidation disorders due to acyl-CoAoxidase, bifunctional protein, and thiolase defi-ciencies.The peroxisomal biogenesis disorders,which result from defects in the peroxisomebiogenesis (PEX) genes, are severe disordersthat include Zellweger syndrome, neonatal ALD,infantile Refsum disease, and rhizomelic chon-drodysplasia punctata type I.

The Biochemical and Molecular Defect in X-linkedAdrenoleukodystrophy

The biochemical abnormality in X-ALD is theaccumulation of saturated VLCFA such as tetra-cosanoic (C24:0) and hexacosanoic acid (C26:0)in the brain, adrenal glands, testes, and plasma.Although the mitochondria are the primary sitefor the oxidation of ingested and stored fats,VLCFA are metabolized in the peroxisome. Themetabolic pathway, which is shown in Figure13-2, leads to the shortening of the carbonchain of the VLCFA by two carbons and the re-lease of acetyl-CoA. This β-oxidation pathwayinvolves consecutive reactions, including dehy-drogenation, hydration of the double bond, de-hydrogenation again, and thiolytic cleavage.Through this four-step mechanism a two-carbonunit is split from the fatty acid in the form of anacetyl-CoA unit, which can then be transportedto the mitochondria and further oxidized in thecitric acid Krebs cycle to produce CO2, H2O,and energy.

The source of the VLCFA substrates metab-olized in the peroxisome is from both dietary


Table 13-1. Phenotypic Presentations of X-linked Adrenoleukodystrophy in Affected MalesPhenotype Age of Onset Clinical Features

Childhood cerebral <10 years Progressive neurological and behavioral deficits with total disability within 3 years

Adolescent cerebral 10–21 years Progressive neurological and behavioral deficits with eventualtotal disability

Adrenomyeloneuropathy Second or third decade Progressive paraparesis, late cerebral involvement in 45%

Addison disease only Any Adrenocortical dysfunction with no neurological disease

Asymptomatic Biochemical defect with no symptoms

intake and endogenous synthesis. The latteroccurs via the microsomal fatty acid elongationsystem. Both dietary and endogenously synthe-sized VLCFA have been implicated in the patho-genesis of X-ALD, which has led to thedevelopment and evaluation of a variety of ther-apeutic interventions that have attempted todecrease exposure to VLFCA from each of thesesources.

The elevation of VLCFA can be detected inmost body tissues and fluids and forms the basisfor a diagnostic assay for the identification of af-fected individuals.The most frequently used testis the measurement of VLCFA in plasma whichhas been shown to be very sensitive. AlthoughVLCFA are increased in some of the other perox-isomal disorders, including Zellweger syndrome,infantile Refsum disease, and neonatal ALD, theclinical presentations of these disorders are verydifferent from X-ALD, and the discrimination of

these patients from those with X-ALD is usuallyapparent from the clinical findings. In femalecarriers of X-ALD, the measurement of VLCFA inthe plasma also reveals elevations in about 85%of obligate carriers.Therefore, the test does havesome false negatives in carriers, and more defini-tive diagnosis can be achieved using molecularmethods as detailed below.

For many years, it was assumed that the in-ability of patients with X-ALD to degrade VLCFAwas due to a primary deficiency of the first en-zyme in this pathway, very long chain fattyacyl-CoA (VLCF-CoA) synthase. Indeed, studiesin cultured skin fibroblasts from patients withX-ALD have revealed that this enzyme activityis decreased when compared to cells fromnormal controls. However, when the X-ALDgene was isolated in 1993, it was discoveredthat the underlying defect is the deficiency ofa peroxisomal membrane protein, called ALD


Figure 13-2. The metabolic pathway for the metabolism of very long chain fatty acids(VLCFA) in the peroxisome.

protein (ALDP).The gene maps to Xq28 and en-codes a protein product that is a member of theadenosine triphosphate binding cassette (ABC)transporter superfamily. There are three othermammalian peroxisomal ABC transporters thatare closely related to ALDP:ALDRP, PMP70, andPMP69.

The precise functions of the peroxisomalABC transporters and the nature of their interac-tion with VLCFA-CoA synthase are not well un-derstood, although their sequence similaritiessuggest that they may have related or overlap-ping functions in peroxisomal metabolism. In-deed, the fact that X-ALD cells, which lack ALDP,maintain residual activity of the synthase en-zyme suggests that the presence of one or moreof the other ABC transporters allows partial ex-pression of the enzymatic function. However,to date the precise nature of the interactionbeetwen ALDP and VLCFA-CoA synthase has notbeen completely elucidated, although recentstudies have shown that peroxisomes fromX-ALD cultured cells contain normal levels ofVLCFA-CoA synthase protein. This finding sug-gests that ALDP is not required for the localiza-tion or stability of the VLCFA-CoA synthase butmay play a role in its activation.

Since the identification of the genetic defectthat results in X-ALD, the underlying moleculardefects in affected patients have been exten-sively studied. Most families have been found tohave private missense mutations, although non-sense and frameshift mutations, splice defects,and deletions also have been identified. Asmight be expected from the fact that differentphenotypes can manifest in the same family, nogenotype-phenotype relationships have beenidentified. Similarly, although 70% of patientshave no demonstrable ALDP, no correlation be-tween disease severity and the level of ex-pressed protein has been established, and theVLFCA levels also do not correlate with pheno-type.Therefore, it has been suggested that mod-ifier genes play an important role in theexpression of the phenotype.

Pathophysiology in X-linkedAdrenoleukodystrophy

Studies of the pathophysiology of the disorderhave focused on the metabolic and pathologiceffects of high levels of VLCFA, although it ispossible that future studies may identify otherroles for ALDP that have yet to be recognized.To date, these investigations have revealed that

VLCFA accumulation has an adverse effect onmembrane structure and function.For example,in cultured adrenocortical cells, the addition ofC26:0 to the media results in increased micro-viscosity of the cell membrane and decreasedsecretion of cortisol after ACTH stimulation.Al-though similar studies have not been carriedout in nerve cells, the effect of VLCFA on neuralcell membranes may also result in the neurolog-ical manifestations of patients with X-ALD.

In addition to the effect of increased VLCFAon membrane and possibly cellular function,the rapid cerebral form of X-ALD is character-ized by an inflammatory response that is be-lieved to contribute to the demyelination thatcharacterizes this phenotype and which is sim-ilar to that seen in multiple sclerosis. These ce-rebral lesions are characterized by breakdownin myelin with sparing of the axons accompa-nied by the accumulation of cholesterol esterin the neurons.A perivascular inflammatory re-sponse with infiltration of T cells, B cells, andmacrophages also is present. Therefore, it is be-lieved that the rapid cerebral disease has an im-munologically-mediated component. It hasbeen suggested that the inflammatory responseoccurs in response to the elevated levels ofVLCFA in lipids, which elicits an inflammatorycascade that may be mediated in part by cyto-kines. Once this cascade begins, it may be moredifficult to intervene in the disease process, andin general therapeutic interventions studied todate have been most effective when initiatedearly. Therefore, prevention of the initiation ofthe immune response is important for improv-ing outcome.

THERAPY

To date, no definitive therapeutic approach hasbeen identified that is applicable to all patientswith X-ALD, although a number of interven-tions have been evaluated as described in thissection. The clinical evaluation of the efficacyof therapeutic approaches for this disorder iscomplicated by the wide spectrum of pheno-typic severity, even within the same family, andthe lack of a biochemical or genetic markerto predict phenotype. For example, it is notknown when a particular patient enters a clini-cal trial what that patient’s precise diseasecourse will be. Therefore, it requires carefullyplanned clinical trials to determine whetherthe intervention is having a beneficial effect,


taking into consideration that not all patientswould go on to have the rapid cerebral form ofthe disease.

In addition to the attempts described below tointerfere with the onset or progression of neu-rological disease, it is also important to carryout endocrine testing on all biochemically iden-tified males to assess their adrenocortical axis.In particular, all patients who have adrenal in-sufficiency require adrenal hormone replace-ment to prevent life-threatening complicationsof insufficiency.

Dietary Therapy

The fact that VLCFA ingested in the diet hasbeen implicated in the pathogenesis of the dis-order led to the development of a diet that re-stricts intake of saturated VLCFA by limitingtotal fat intake to 20 to 30 g per day. However,the use of this low-fat diet alone does not re-duce the plasma VLCFA concentrations anddoes not have an impact on clinical progression,most likely due to the fact that endogenous pro-duction of VLCFA continues to occur and con-tributes to the disease phenotype. Therefore,attempts have been made to disrupt the endoge-nous production of VLCFA by administration ofa 4:1 mixture of glycerol trioleate oil and glyc-eryl trierucate oil, which is commonly refereedto as Lorenzo’s oil, so named after a child af-fected with X-ALD whose parents actively par-ticipated in the development of this therapy.This mixture, which includes erucic acid, an un-saturated 22-carbon fatty acid, acts by compet-ing with saturated fatty acids for the microsomalVLCFA elongation system.

Although Lorenzo’s oil has been found tonormalize VLCFA levels in the plasma of pa-tients with X-ALD, this therapy has not beenshown to improve outcomes in patients who al-ready have neurological degeneration. Clinicaltrials have shown, however, that it may preventneurological disease when provided to bio-chemically identified males who have not yetbegun to manifest any neurological symptomsand who have a normal MRI. Therefore, it hasbeen advocated by some that the administra-tion of Lorenzo’s oil to reduce endogenous pro-duction of VLCFA in combination with dietarytherapy to reduce intake may be useful inasymptomatic boys, although careful follow-upand frequent clinical evaluation is needed tofacilitate early identification of those patients

who progress despite this intervention andwho may become candidates for BMT.

Bone Marrow Transplantation

BMT has been attempted in a number of af-fected boys with X-ALD. The rationale for theuse of BMT in this disorder is that bonemarrow-derived cells are known to be capableof degrading VLCFA. Perhaps more important, itwas expected that some bone marrow-derivedcells would enter the brain since infiltration ofthe brain with perivascular lymphocytes andmacrophages is known to occur. Experiencewith the procedure has shown that it can stabi-lize the disease course in a subset of patients,and several cases of reversal of neurologicaldamage also have been reported in patients inthe early stages of their disease.

Currently, BMT is not recommended for pa-tients who already have advanced disease andwho will most likely continue to deteriorate de-spite the intervention, presumably because theinflammatory cascade has already begun. Simi-larly, it is not recommended for boys with a bio-chemical diagnosis but with no symptoms orMRI changes as it is possible that they maynever develop the rapid cerebral form.For thesepatients, exposing them to the high morbidityand mortality associated with BMT is not justi-fied. However, careful neurological assessment,imaging studies, and neuropsychological testingare needed in asymptomatic patients in order todetect cerebral abnormalities early enough topermit the consideration of this intervention ifevidence of disease progression is found. In gen-eral, BMT for this disease is now carried out pri-marily in patients with the biochemical defectwho also have early evidence of cerebral in-volvement. It is not recommend in AMN.

Lovastatin

Lovastatin, commonly known as an inhibitor ofcholesterol synthesis, is another agent that hasbeen shown to reduce plasma levels of VLCFAand to increase the ability of X-ALD fibroblaststo metabolize VLCFA. The mechanism for thiseffect may involve upregulation of ABCD2,which encodes a protein (ALDR) that can sub-stitute for the function of ALDP. However, clini-cal trials with this agent have demonstratedvariable reactions, and no clear clinical benefithas been demonstrated.


QUESTIONS

1. What is the underlying genetic defect inX-ALD?

2. How does peroxisomal β-oxidation differfrom mitochondrial β-oxidation?

3. How does the concept of genetic hetero-geneity apply to the phenotypic expres-sion of X-ALD?

4. How would you evaluate a patient sus-pected of having X-ALD?

5. What is the most reliable method for iden-tifying carriers of X-linked ALD? Why?

6. Which patients with elevated levels of VL-CFA are the most suitable candidates forBMT?

BIBLIOGRAPHY

Kemp S,Pujol A,Waterham HR,et al.: ABCD1 muta-tions and the X-linked adrenoleukodystrophymutation database: role in diagnosis and clinicalcorrelations. Hum Mutat 18:499–515, 2001.

Moser AB, Kreiter N, Bezman L, et al.: Plasma verylong chain fatty acids in 3,000 peroxisome dis-ease patients and 29,000 controls. Ann Neurol45:100–110, 1999.

Moser HW.: Adrenoleukodystrophy: phenotype,genetics, pathogenesis and therapy. Brain 120:1485–1508, 1997.

Moser HW, Raymond GV, Koehler W, et al.: Evalua-tion of the preventive effect of glyceryl trioleate-trierucate (“Lorenzo’s oil”) therapy in X-linkedadrenoleukodystrophy: results of two concurrenttrials.Adv Exp Med Biol.544:369–387,2003.

Moser HW, Smith KD, Watkins PA, et al.: X-linkedadrenoleukodystrophy, in Scriver CR, BeaudetAL, Sly WS, Valle D (eds): The Metabolic andMolecular Bases of Inherited Disease. 8th ed.McGraw-Hill, New York, 2000, pp. 3257–3301.

Mosser J, Douar AM, Sarde CO, et al.: PutativeX-linked adrenoleukodystrophy gene shares un-expected homology with ABC transporters.Nature 361:726–730, 1993.

Paintlia AS, Gilg AG, Khan M, et al.: Correlation ofvery long chain fatty acids accumulation and in-flammatory disease progression in childhoodALD: implications for potential therapies. Neu-robiol Dis 14:425–439, 2003.

Shapiro E, Krivit W, Lockman L, et al.: Long-termeffect of bone-marrow transplantation forchildhood-onset cerebral X-linkedadrenoleukodystrophy. Lancet 356:713–718,2000.

van Geel BM, Bezman L, Loes DJ, et al.: Evolution ofphenotypes in adult male patients with X-linkedadrenoleukodystrophy.Ann Neurol 49:186–194,2001.

Whitcomb RW,Linehan WM,Knazek RA.:Effects oflong-chain, saturated fatty acids on membranemicroviscosity and adrenocorticotropin respon-siveness of human adrenocortical cells in vitro.J Clin Invest 81:185–188, 1988.


CASE REPORT

A 10-year-old girl from China was evaluated atthe University of Pennsylvania for homozygousfamilial hypercholesterolemia (FH) and possi-ble participation in an experimental gene ther-apy protocol for her condition. Cutaneousxanthomas (lesions characterized by abnormalaccumulation of lipids in macrophages) werefirst observed at 3 months of age. However, theorigins were not fully investigated until the ageof 3 years, when she was found to have a totalcholesterol level of 1150 mg/dL (desirable lev-els are < 200 mg/dL) with normal triglyceridelevels, and the diagnosis of homozygous FH wasmade. Both of her parents’ total cholesterol lev-els were in the range of 300–400 mg/dL, consis-tent with the diagnosis of heterozygous FH. Apositive family history for premature coronaryartery disease was present on both sides.

The proband remained in apparent goodhealth until she was 9 years old,when she devel-oped progressive symptoms of exertional chesttightness and shortness of breath. Her physicalexamination at 10 years of age revealed numer-ous cutaneous and tendon xanthomas (overlyingher elbows, wrists, knees, and buttocks and onthe Achilles’ tendons) as well as a systolic ejec-tion murmur radiating to the neck and bilateralcarotid bruits (abnormal sound heard during aus-cultation that suggests the presence of block-age).A carotid Doppler showed the presence ofa stenosis (blockage) of approximately 50% inboth common carotid arteries. She underwent athallium exercise stress test that was negative; an

exercise echocardiography that showed infer-oseptal and inferobasal hypokinesis (decreasedcontractile function of the myocardium) duringexercise that resolved at rest; and a cardiaccatheterization that showed a 90–99% stenosisof the right coronary artery’s ostium and a 40%stenosis of left anterior descending artery, in ad-dition to mild aortic regurgitation.

Functional and genetic studies were per-formed to confirm the diagnosis of homozygousFH. Low-density lipoprotein (LDL) receptor ac-tivity, measured in an in vitro functional assaythat used fibroblasts isolated from the patient’sskin, was less than 2% of the normal value(functionally confirming the diagnosis of ho-mozygous FH). LDL metabolism studies showedthat she removed LDL from her blood at amarkedly slower rate than normal, consistentwith lack of LDL receptors. Finally, geneticanalysis confirmed that she was homozygousfor an LDL receptor gene mutation in exon 10(W462X) that created a premature stop codonat amino acid 462; this mutant truncated LDLreceptor has been shown to be nonfunctional.

She underwent coronary artery bypass graft-ing of the right coronary artery. Her exertionalchest tightness and shortness of breath werenoted to be entirely resolved following surgery.Two months after surgery, she was readmittedat the Hospital of the University of Pennsylvaniafor a research protocol of ex vivo liver-directedgene therapy. This was an experimental proto-col designed to genetically correct liver cellstaken from the patient with the normal LDL re-ceptor gene and then readminister these liver

14

Low-Density Lipoprotein Receptors and Familial Hypercholesterolemia

MARINA CUCHEL and DANIEL J. RADER

152

cells to the patient (Raper et al., 1996). The firststage of the protocol consisted of the resectionof the left lateral segment of the patient’s liverand the isolation of hepatocytes (liver cells) toestablish primary hepatocyte cultures. The cul-tures were then infected with a recombinantretrovirus that contained the human LDL recep-tor gene. One day later, the genetically modifiedhepatocytes were harvested and reinfused intothe patient’s liver via a portal venous catheter.The patient tolerated the procedure well andleft the hospital several days later.

Three months later, her cholesterol levelswere about 20% lower than prior to the genetherapy protocol. Repeat LDL metabolism stud-ies showed that she cleared LDL from her bloodsignificantly faster than prior to the gene ther-apy, consistent with some liver expression ofthe LDL receptor. A liver biopsy was done anddemonstrated clusters of liver cells that wereexpressing the normal LDL receptor gene.

The patient was followed yearly. Because hercholesterol level remained very high, she wastreated with a low-fat, low-cholesterol diet andthe maximal dose of a statin (the most commondrug class for treating high cholesterol) withmodest effects.Although LDL apheresis (a phys-ical method of purging the blood of choles-terol) is strongly recommended in patients withhomozygous FH, it was not an available optionwhere she lived. Her cardiovascular conditiongot progressively worse, and she experiencedmyocardial infarction at the age of 14 years. Herrisk of developing recurrent atheroscleroticvascular disease remains high because of hervery elevated cholesterol.

DIAGNOSIS

FH is a well-characterized genetic disordercaused by mutations in the gene encoding theLDL receptor, the receptor responsible for theuptake and subsequent removal of LDL fromcirculation. As a consequence, LDL levels inplasma are increased. FH is autosomal codomi-nant: Heterozygotes have elevated LDL choles-terol levels (250–500 mg/dL), and homozygoteshave even higher levels (500–1200 mg/dL). Itsfrequency is 1 in 500 subjects in the heterozy-gous state and 1 in 1 million in the homozygousstate. In certain populations, such as AshkenaziJews,Afrikaners in South Africa, French Canadi-ans, and Christian Lebanese, the frequency issubstantially higher due to a “founder effect”

(i.e., to the fact that these population groupsoriginated from a small number of individualscarrying LDL mutations at higher frequencythan that of the general population).

Heterozygous FH patients have LDL choles-terol levels that are two- to threefold higherthan normal subjects, tendon xanthomas (accu-mulation of cholesterol in tendons, causingthem to be thickened and irregular, especiallyin the backs of the hands and the Achilles’ ten-dons) in adulthood, and coronary atherosclero-sis that becomes clinically manifest usually afterage 30 years in men and 40 years in women. Onthe other hand, patients with homozygous FHhave LDL levels that are more than three timesthe normal values; tendon, tuberous, and cuta-neous xanthomas in childhood; and prematureatherosclerosis with symptomatic coronary ar-tery and other atherosclerotic vascular diseasefrequently before the age of 20 years. Interest-ingly, as in the case presented, the early athero-sclerosis manifestation in homozygous FH isoften due to the formation of atheroma alongthe aortic root, aortic valve, and coronary ostia,which result in the presence of systolic mur-mur as well as a pressure gradient across theaortic valve, narrowing of the root, and obstruc-tive ostial coronary disease. This often requiressurgical intervention that may include coronarybypass and/or valve replacement.

There is a substantial interindividual variationin LDL cholesterol levels among patients withFH. Generally, LDL cholesterol levels are in-versely related to the residual LDL receptor activ-ity, as measured in the in vitro assay that usesskin fibroblasts. Patients with homozygous FHare classically divided into two groups based onthe fibroblast LDL receptor activity.Patients withless than 2% activity, as the patient describedin the case report, are classified as receptor-negative. Patients with 2%–20% LDL receptoractivity are classified as receptor-defective. Thenatural history of the disease is much more se-vere in receptor-negative patients,who, if left un-treated, rarely survive beyond the second decadeof life. Receptor-defective patients, in contrast,have less-severe hypercholesterolemia and amore delayed onset of coronary artery diseaseand mortality.

The diagnosis of FH is made on clinicalgrounds. The diagnosis of homozygous FH isusually made on the basis of markedly elevatedLDL cholesterol levels, normal triglyceride level,presence of xanthomas, and family history of hy-percholesterolemia and/or premature coronary

LDP Receptors and Familial Hypercholesterolemia 153

artery disease. The differential diagnosis of het-erozygous FH includes other less-frequent mono-genic causes of hypercholesterolemia, such asfamilial defective apolipoprotein B-100 (apoB-100;due to mutations in apoB, the major proteinin LDL and the ligand for binding to the LDLreceptor), autosomal recessive hypercholes-terolemia (due to mutations in the ARH gene en-coding a novel adaptor protein important forthe internalization of the LDL receptor), auto-somal dominant hypercholesterolemia (due tomutations in PCSK9, a gene of unknown func-tion), and sitosterolemia (due to mutations inABCG5 or ABCG8 that are responsible for excre-tion of sterols from the body) (Rader et al.,2003).


The landmark studies by Brown and Goldstein(see Goldstein et al., 2001 for review) that ledto the discovery of the genetic cause of FH andits metabolic and clinical consequences are fas-cinating and of historical importance becausethey were the first to elucidate the mechanismsof receptor-mediated endocytosis. Brown andGoldstein received the Noble Price in Medicinein 1985 for their work on the LDL receptor andmolecular basis of FH.

Cells need cholesterol to survive. Cells caneither synthesize cholesterol de novo or obtainit from exogenous sources, such as lipoproteinspresent in the circulation (see Table 14-1 andRader and Hobbs, 2005, for review). All of thelipoprotein classes contain phospholipids, es-terified and unesterified cholesterol, and triglyc-erides to varying degrees. LDL are the mostabundant lipoproteins in humans. They are

composed of a core of neutral lipids, mainlycholesteryl ester, surrounded by a shell of phos-pholipids and held together by a protein calledapoB. Their precursors are very low densitylipoproteins (VLDL), triglyceride-rich lipopro-teins that are synthesized and secreted by theliver. Once in circulation,VLDL undergo lipoly-sis of the triglyceride by two enzymes calledlipoprotein lipase and hepatic lipase and areconverted to their remnants, intermediate-densitylipoproteins (IDL), and finally to LDL.

The LDL receptor is a cell-surface glycopro-tein able to bind both apoB, the sole protein ofLDL,and apoE,an apolipoprotein that is presentin multiple copies on other lipoproteins, includ-ing VLDL and IDL, but not LDL. ApoE has atwofold higher affinity than apoB for the recep-tor, and in general IDL are removed from the cir-culation more efficiently than LDL. In addition,while its precursors may be metabolized viaother pathways, LDL removal from circulationdepends heavily on functional LDL receptors.

LDL receptors are synthesized in the endo-plasmic reticulum (ER), undergo glycosylationin the Golgi, and are then transported to theplasma membrane, where they localize incoated pits, specialized regions of the cell mem-brane capable of invaginating and forming coatedvesicles (Fig. 14-1). Once at the cell surface, theLDL receptor can bind to the apoB on LDL, trig-gering invagination of the coated pit and inter-nalization of the LDL receptor/LDL complex in acoated vesicle. Coated vesicles fuse with largervesicles called endosomes. In the acidic environ-ment of the endosomes, the LDL receptor un-dergoes conformational changes that cause thedissociation of bound LDL. The receptor is thenrecycled to the membrane surface, while the


Table 14-1. Major Lipoprotein ClassesDensity Electrophoretic Apolipoproteins(g/mL)* Size (nm)† Mobility‡ Major Other

Chylomicrons <0.95 75–1200 Origin B-48 A-I,A-IV, C-I, C-II, C-III

VLDL <1.006 30–80 Pre-beta B-100 E,A-I,A-II,A-V, C-I, C-II, C-III

IDL 1.006–1.019 25–35 Slow pre-beta B-100 E, C-I, C-II, C-III

LDL 1.019–1.063 18–25 Beta B-100

HDL 1.063–1.21 5–12 Alpha A-I A-II,A-IV, E, C-I, C-II, C-III

Lp(a) 1.050–1.120 25 Pre-beta B-100 Apo(a)

VLDL, very low density lipoprotein; IDL, intermediate-density lipoprotein; LDL, low-density lipoprotein; HDL, high-density lipoprotein;apo, apolipoprotein; C, cholesterol;TG, triglycerides; PL, phospholipids.

*The density of the particle is determined by ultracentrifugation.

†The size of the particle is measured using gel electrophoresis.

‡ The electrophoretic mobility of the particle on agarose gel electrophores reflects the size and surface charge of the particle, with betathe position of LDL and alpha the position of HDL.

dissociated LDL is transferred into lysosomes,where its lipid and protein components are hy-drolyzed.

Numerous mutations (>1000) at the LDL re-ceptor locus have been described as causingFH. Interindividual phenotypic variation amongpatients with either heterozygous or homozy-gous FH is at least partly dependent on theresidual activity of the LDL receptor. This inturn depends on the site of mutation within theLDL receptor gene. The gene contains 18 exonsand 17 introns and encodes a protein of 860amino acids. The protein contains multiple do-mains ( ligand-binding domain, EGF or epider-mal growth factor homology domain, O-linked

sugar domain, membrane-spanning domain, andcytoplasmic domain), all of which are impor-tant for its function. Numerous mutations havebeen reported in each of the exons.

LDL receptor mutations can be divided intofive different functional classes as shown inFigure 14-1. Class 1 mutations (null alleles) donot produce immunoprecipitable protein. Class2 mutations produce proteins that are not(completely, 2A; partially, 2B) transported to theGolgi due primarily to defective folding of theprotein, which is then targeted for degradation.Class 3 mutations produce proteins that are nor-mally transported to the plasma membrane butdo not bind to LDL normally. Class 4 mutations


Figure 14-1. The biochemical role of the low-density lipoprotein (LDL) and the classesof mutations that affect its function. The LDL receptors are synthesized in the endoplas-mic reticulum (ER), undergo glycosylation in the Golgi, and are transported to theplasma membrane, where they localize in coated pits. Once at the cell surface, the LDLreceptors bind to the apolipoprotein B on low-density lipoprotein (LDL), triggering in-vagination of the coated pit and internalization of the LDL receptor/LDL complex in acoated vesicle. Coated vesicles fuse with larger vesicles called endosomes. In the endo-somes, the LDL receptor undergoes conformational changes that cause the dissociationof bound LDL. The receptor is then recycled to the membrane surface,while the dissoci-ated LDL is transferred into lysosomes, where its lipid and protein components are hy-drolyzed. The free cholesterol obtained from the degradation of LDL diffuses into thecytoplasm and plays an important role in regulating intracellular cholesterol homeosta-sis through several mechanisms, including downregulation of hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase activity and LDL receptor expression. There are fiveclasses of LDL receptor mutations, indicated by circled numbers in the figure:class 1 mu-tations are null alleles. Classes 2–5 affect different stages of processing, ligand binding,and internalization of the LDL receptor. See text for details.

produce proteins that are able to bind LDL butdo not cluster on the coated pits and fail to in-ternalize. Class 5 mutations produce proteinsthat are not able to release the LDL in the endo-some and cannot be recycled. Class 1 and class2A are usually associated with less than 2% LDLreceptor activity in cultured fibroblasts, whileother classes are associated with some degreeof activity.As discussed earlier, subjects with ho-mozygous FH who are receptor-negative (lessthan 2% activity) tend to have higher choles-terol levels and more severe clinical manifesta-tion than patients who are receptor-defective.

The importance of functional LDL receptorsfor the removal of LDL from circulation in sub-jects with FH has been confirmed studying invivo the clearance rate of radiolabeled LDL in-jected intravenously, as was done for the patientdescribed in the case report before and aftergene therapy. The removal time of LDL in sub-jects with heterozygous FH is approximatelytwice as long and in those with homozygousFH more than three times as long compared tothat of normal controls.

Free cholesterol obtained from the degrada-tion of LDL diffuses into the cytoplasm andplays an important role in the feedback mecha-nism that regulates intracellular cholesterolconcentration through several mechanisms.Among these are the suppression of the activityof hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase (the rate-limiting enzyme incholesterol synthesis) and the downregulationof LDL receptor synthesis to prevent excesscholesterol accumulation. In the absence of anexogenous source of cholesterol, both HMG-CoA reductase activity and LDL receptor syn-thesis are upregulated and hydrolysis of esterifiedcholesterol is increased, so that ultimately theneeded amount of intracellular cholesterol ismaintained.

Members of a family of nuclear transcriptionfactors called sterol regulatory element-bindingproteins (SREBP) are responsible for the regula-tion of these cholesterol feedback mechanisms.SREBP are able to activate a number of genesencoding for proteins involved in the home-ostasis of cholesterol and other lipids, includingthe LDL receptor gene itself.

The promoter region of the LDL receptorgene contains several regulatory elements thatcontrol its expression. One in particular, sterolregulatory element 1, is the binding site forSREBP. Under baseline conditions, SREBP areinactive proteins bound to the endoplasmic

reticulum (ER). In the ER, SREBP bind to an-other protein called SREBP-cleavage activatingprotein (SCAP). SCAP is a sensor of sterols.When the cholesterol content in the cell is low,the SCAP-SREBP complex is able to move to theGolgi,where the SREBP are sequentially cleavedby two proteases, S1P and S2P, thus releasingthe NH2 terminal of the SREBP to the cyto-plasm. Once in the cytoplasm, SREBP translocateto the nucleus, where they activate transcriptionby binding to the SRE-1 in the promoter regionsof the LDL receptor gene, the HMG-CoA reduc-tase gene, and other sterol-regulated genes, thusgenerating more cellular cholesterol in a homeo-static response.

Conversely, if intracellular sterols are in ex-cess, then SCAP undergoes conformationalchanges, and the complex SCAP-SREBP can nolonger migrate to the Golgi. The net resultingeffect is that the NH2 terminal of SREBP is notcleaved. Under these circumstances, the SREBPdo not translocate to the nucleus and do not ac-tivate transcription of the target genes HMG-CoA reductase and the LDL receptor.

This finely regulated mechanism is physio-logically very important as it is this mechanismthat regulates the response to a high-fat diet aswell as the response to therapeutic interven-tion with cholesterol-lowering drugs.A diet richin saturated fat and cholesterol is associatedwith an increased flux of dietary cholesteroland fats from the intestine to the liver. Thiscauses an increase in intracellular cholesterolconcentration and the downregulation of LDLreceptor gene expression in the liver, mediatedby the SREBP-regulated feedback just de-scribed. The net effect is that LDL cholesterol isremoved more slowly from circulation and ac-cumulates in the plasma, causing an increase inits levels.

An opposite effect is at the basis of the up-regulation of LDL receptors in response totreatments with bile acid sequestrants, intes-tinal cholesterol absorption inhibitors, andHMG-CoA reductase inhibitors. The first classof drugs inhibits the intestinal reabsorption ofbile acids, thus promoting increased conver-sion of cholesterol to bile acids in the liver. Theincreased demand for cholesterol results in ac-tivation of the SREBP system and upregulationof LDL receptor synthesis (as well as choles-terol synthesis via upregulation of HMG-CoAreductase). Similarly, inhibition of intestinalcholesterol absorption with ezetimibe resultsin a reduction in the hepatic cholesterol pool


and compensatory upregulation of the LDL re-ceptor. Finally, upregulation of the LDL recep-tors is also obtained with HMG-CoA reductaseinhibitors (statins) that decrease intracellularcholesterol synthesis by inhibiting the rate-limiting enzyme in the cholesterol synthesispathway.

THERAPY

The primary goal of therapy is the control ofthe hypercholesterolemia and prevention ofatherosclerotic cardiovascular disease. Patientswith heterozygous FH can usually be success-fully treated with medications to lower the LDLcholesterol to acceptable levels (Table 14-2).They are generally responsive to treatment withstatins, alone or in combination with otherdrugs, such as bile acid sequestrants (such ascholestyramine) or cholesterol absorption in-hibitors (such as ezetimibe) that act additively toupregulate the expression of the functioningLDL receptor as described in the “BiochemicalPerspectives” section. In a few cases, a more ag-gressive treatment with LDL apheresis (discussedin this section) may have to be considered in or-der to reach acceptable LDL cholesterol levels.

In contrast, individuals with homozygous FHare largely unresponsive to conventional drugtherapy and have limited treatment options. Inparticular, the response to statins is practicallynil in receptor-negative patients, while it maybe present to a small degree in receptor-defective patients. Surgical interventions, suchas portacaval shunt and ileal bypass, have beenused in the past in an attempt to decrease circu-lating cholesterol levels; however, they resultedin partial and transient LDL cholesterol lower-ing at best. The replacement of the patient liverwith a healthy donor liver (orthotopic liver

transplantation) has been demonstrated to sub-stantially reduce LDL cholesterol levels in pa-tients with homozygous FH, but it is not thetreatment of choice due to the disadvantagesand risks that are associated with this invasiveapproach and the lifelong need for immunosup-pressive therapy.

Currently, the preferred treatment option forhomozygous FH is LDL apheresis.This is a pro-cedure in which apoB-containing lipoproteins(VLDL, IDL, and LDL) are selectively removedfrom the plasma when passed extracorporeallyover a column designed to bind apoB. HDL andother plasma protein concentrations are notaffected by this selective method. This proce-dure has been shown to result in regression ofxanthomas and may reduce the rate of progres-sion of atherosclerotic plaques. However, be-cause rapid reaccumulation of LDL occurs, ithas to be repeated frequently (every 1–2 weeks),and many patients with homozygous FH remainat high risk for atherosclerotic cardiovasculardisease. Furthermore, this is a procedure that isavailable only in few specialized centers, andmany patients with homozygous FH, like theone described in the case report, do not haveaccess to it.

For these reasons, there is a great medicalneed for more effective treatments. Among fu-ture treatment options, liver-directed gene ther-apy is certainly a possibility. Gene therapy is anintervention with the goal of modifying the ge-netic material of a target organ’s cells, and it isconsidered a theoretically attractive approachfor the treatment of rare monogenic diseasessuch as homozygous FH. In the ex vivo ap-proach, cells are genetically modified ex vivoand then returned to the patient. In the in vivoapproach, cells are modified in vivo via the ad-ministration of vectors containing the gene thattarget the desired tissue. The young patient


Table 14–2. Low-Density Lipoprotein-Cholesterol Lowering TreatmentsFrequently Used in Patients with Familial HypercholesterolemiaTreatment Major Mechanism Major Effect

Statins ↓ C synthesis ↑ LDLR activity

Bile acid sequestrants ↓ Bile acids reabsorption ↑ LDLR activity

Cholesterol absorption inhibitors ↓ Intestinal C absorption ↑ LDLR activity

Nicotinic acid ↓ VLDL production ↓ VLDL synthesis

Low-fat, low-C diet ↓ Dietary C and fats ↑ LDLR activity

LDL apheresis Remove LDL ↓ LDL

C, cholesterol; LDL, low-density lipoprotein; LDLR, LDL receptor; VLDL, very low density lipoprotein.

described in the case report above was one ofthe five patients who participated in a pilotstudy to test the feasibility of ex vivo genetransfer for homozygous FH (Grossman et al.,1995). Current efforts focus on delivering theLDL receptor gene in vivo using viral vectorsthat have hepatic specificity.

Viral vectors are preferentially used in genetherapy because their higher efficiency intransducing cells in vivo as compared withother vectors such as liposomes or the ad-ministration of DNA only. In the lipoproteinresearch field, adenoviruses have been exten-sively used as vectors in animal studies be-cause of their ability to efficiently infect theliver and to result in high levels of hepatictransgene expression. However, their adminis-tration is associated with cellular immune re-sponses, which limits the duration of the geneexpression. Recent efforts have been focusedon the use of adeno-associated viruses (AAV).Adeno-associated viruses are parvoviruses thatdepend on other viruses (such as adenovirus)for replication. Importantly, they have notbeen associated with toxicity or inflammatoryresponse, and stable long-term gene expres-sion has been achieved in animal studies.Although homozygous FH remains an excel-lent candidate for the development of liver-directed gene therapy, more work must bedone on the development of safe, liver-specificvectors that will provide long-term expressionof the LDL receptor.

Alternative pharmacological approaches arealso being explored. Preliminary results in ani-mal models of FH using an inhibitor of the mi-crosomal transfer protein, a key enzyme in theassembly and secretion of apoB-containinglipoproteins by the liver (see chapter 27 onabetalipoproteinemia for details) showed a sig-nificant decrease in LDL cholesterol levels andsuggests that this may be a more realistic thera-peutic approach in the near future.

QUESTIONS

1. What are the clinical characteristics ofboth heterozygous and homozygous FH?

2. What is the genetic cause of FH?

3. Once LDL is bound to the LDL receptor,what is the intracellular fate of this com-plex?

4. What is the effect of dietary cholesteroland saturated fats on the expression of theLDL receptor by the liver?

5. What is the mechanism of action of com-monly used lipid-lowering drugs, such asstatins and ezetimibe?

6. What are SREBP and SCAP? What role dothey play in the regulation of cholesterolhomeostasis?

7. What are the therapeutic options for thetreatment of both heterozygous and ho-mozygous forms of FH?

BIBLIOGRAPHY

Brown MS, Anderson RG, Goldstein JL: Recyclingreceptors: the round-trip itinerary of migrantmembrane proteins. Cell 32:663–667, 1983.

Brown MS, Goldstein JL: A receptor-mediated path-way for cholesterol homeostasis. Science 232:34–47, 1986.

Goldstein JL, Hobbs HH, Brown MS: Familial hyper-cholesterolaemia, in Scriver CR, Beaudet AL, SlyWS,Valle D (eds): The Metabolic and MolecularBases of Inherited Disease. vol. 2. McGraw-Hill,New York, 2001, pp. 2863–2913.

Grossman M, Rader DJ, Muller DWM, et al:A pilotstudy of ex vivo gene therapy for homozygousfamilial hypercholesterolaemia. Nat Med 1:1148–1154, 1995.

Horton JD, Goldstein JL, Brown MS: SREBPs: activa-tors of the complete program of cholesterol andfatty acid synthesis in the liver. J Clin Invest10:1125–31, 2002.

Naoumova RP, Thompson GR, Soutar AK: Currentmanagement of severe homozygous hypercholes-terolaemias.Curr Opin Lipidol 15:413–422,2004.

Rader DJ,Cohen J,Hobbs HH:Monogenic hypercho-lesterolemia: new insights in pathogenesis andtreatment. J Clin Invest 111:1795–1803,2003.

Rader DJ, Hobbs HH: Disorders of lipoprotein me-tabolism, in Braunwald E, Fauci AS, Kasper DL,Hauser SL, Longo DL, Jameson JL (eds): Harri-son’s Principles of Internal Medicine. 16th ed.McGraw-Hill, New York, 2005, pp. 2286–98.

Raper SE, Grossman M, Rader DJ, et al: Safety andfeasibility of liver-directed ex vivo gene therapyfor homozygous familial hypercholesterolemia.Ann Surg 223:116–116, 1996.


CASE REPORTS

Case 1

T. L. was a 5-year-old patient who underwent atonsillectomy in 1960. His enlarged tonsils wereyellowish-gray,with many septa running throughmultiple abnormally lobulated regions. Onpathological examination, many foam cells(macrophages with lipid deposits) were found inand among the lymph follicles. This histiocyticappearance of the foam cells led to a tentative di-agnosis of either Hand-Schüller-Christian (a multi-focal eosinophilic granuloma) or Niemann-Pick(a lysosomal lipid storage) disease. Thinking thatthe tonsillar honeycomblike appearance was un-usual for either of these diseases, the attendingphysician, Dr.Thomas Edmonds, referred T. L. forfurther analyses to the NIH and thereby initiateda series of investigations that led to the discoveryof Tangier disease by Dr. Donald S. Fredricksonand colleagues in 1961.

A physical examination found T. L. to be nor-mal except for moderate hepatosplenomegaly(liver enlargement) and scattered enlargedlymph nodes. The diagnosis of Hand-Schüller-Christian disease was quickly eliminated whenliver and lymph node biopsies lacked the histio-cytic granuloma manifestation of the disease.A neurological examination of T. L. was unre-markable, a finding that is extremely rare inNiemann-Pick disease for a 5-year-old patient.Histochemical analysis to characterize the lipidsin the biopsies was negative with the Smith-Dietrich stain, considered to be specific for the

unsaturated phospholipids normally associatedwith Niemann-Pick tissues but was positive forthe Schultz test, suggesting high levels of choles-terol or a related sterol. In addition, this lipid de-posit was not membrane-bound. Subsequentchemical analysis of the lipid confirmed an ex-traordinarily high cholesterol ester contentbut normal levels of phospholipids, includingsphingomyelin, thus definitively dismissing thepossibility of Niemann-Pick disease. The investi-gators involved in the case suspected a novellipid accumulation disorder and named it Tang-ier disease after the locality where T. L. resided.

Tangier Island is a geographically isolated is-land in the Chesapeake Bay of Virginia. It had in1961 about 900 inhabitants, a large number ofwhom descended from a few original inhabi-tants who settled there in the late 1600s. Sincesuch a genetically insulated locale provides anopportune setting for manifestation of a raregenetic disorder, a throat examination of morethan a tenth of the island inhabitants was car-ried out after the diagnosis of T. L. All werenormal except for E. L., the patient’s sister.Although E. L. was healthy without signs ofhepatosplenomegaly or lymphadenopathy (ab-normal enlargement of lymph nodes), she, likeT. L., had enormous lobulated tonsils (Fig. 15-1)and no neurological abnormalities.

When the lipid content of the tonsils from E.L. was compared to a nonsymptomatic control,the main difference was a 50-fold increase incholesterol esters. This finding prompted the in-vestigators to focus their research on cholesterolmetabolism in the probands.As all results of tests

15

Tangier Disease: A Disorder in the ReverseCholesterol Transport Pathway

LIEN B. LAI,VIJAYAPRASAD GOPICHANDRAN, and VENKAT GOPALAN

159

for cholesterol synthesis were normal, plasmalipid profiling was subsequently explored.Whilethe plasma triglyceride level was sixfold higherthan that of controls, cholesterol and phospho-lipids levels were lower. Further analyses re-vealed that the high-density lipoproteins (HDL)were reduced to almost undetectable levels,butthe low-density lipoproteins (LDL) appeared tobe unaffected. The mother of T. L. and E. L. alsohad low levels of HDL. A partial pedigree ofthese children indicated consanguinity in theirgrandparents’ (both maternal and paternal) andparents’ generations.

Case 2

A 32-year-old man experienced weakness andburning pain in his hands for the 2 months.Theweakness started insidiously and had been grad-ually progressing. As a professional pianist, hefound that the mobility of his fingers wasgreatly limited since the onset of the problems.The burning pain started a few days after theweakness and was spontaneous with no spe-cific aggravating or relieving factors.

His medical history was unremarkable ex-cept for a recurrent sore throat during child-

hood that resulted in tonsillectomy at age 12years. He was not diabetic or hypertensive andhad no history of exposure to toxic metals.

Physical examination found him to be moder-ately built and nourished.Both hands showed di-minished power of the flexors and the extensorsto 2/5 on the right and 3/5 on the left.Wasting ofthe small muscles of the thenar eminence wasalso noted in both hands. Reflexes of the bra-chioradialis, biceps, and triceps were diminishedto 1/4 in the right and 2/4 in the left arm.Therewas loss of pain and temperature sensation overthe distribution of median and radial nerves.Allother muscle groups and sensations were nor-mal. Cranial nerves were normal.Abdominal pal-pation revealed hepatosplenomegaly, which wassubsequently confirmed by ultrasound. Oral ex-amination found yellowish plaques on the poste-rior pharyngeal wall. Eyes, ears, and nose werenormal.

Differential diagnoses of peripheral neuropa-thy were entertained. Laboratory tests revealedthat serum parameters for electrolytes and pro-teins were all within the normal range. Urineporphyrinogen and porphobilinogen levels werenormal.Tests were negative for serum rheuma-toid factor and antinuclear antibodies, the latterused in detection of connective tissue diseasessuch as systemic lupus erythematosus andpolyarteritis nodosa that could present withfeatures of peripheral neuropathy. Nerve con-duction studies of the radial, ulnar, and mediannerves revealed delayed conduction.Biopsies ofthe ulnar and radial nerves showed loss ofnerve fibers and sudanophilic (indicating lipid)deposits in the Schwann cells of the neurons.Similarly, the yellowish plaques of the pharynxshowed abundant macrophages filled with su-danophilic material. These deposits were notmembrane-bound.

Subsequently, serum lipid and lipoproteinprofiles were obtained: 70 mg/dL total choles-terol (normal is 130–200 mg/dL), 1 mg/dL HDLcholesterol (optimal is ≥ 60 mg/dL), 180 mg/dLtriglycerides (normal is 100–150 mg/dL), andless than 5 mg/dL apolipoprotein A-I (apoA-I;normal is ∼ 140 mg/dL).Cholesterol efflux frompatient skin fibroblasts to apoA-I, the main pro-tein component of HDL, was reduced to 30% ofnormal. These results indicated Tangier disease,the definitive diagnosis of which was madewhen the sequencing of the ATP-binding cas-sette transporter A1 (ABCA1) gene revealed anonsense mutation within exon 12.


Figure 15-1. The enlarged and lobulated tonsilsof E. L. Dr. Donald Fredrickson annotated his expe-rience on encountering the sight: “When I sawthose gigantic bright orange organs, I had thesame feeling one gets at looking at a mountainthat has never been climbed” (http://profiles.nlm.nih.gov/FF/B/B/L/G/_/ffbblg.txt). Photographwas reproduced from http://profiles.nlm.nih.gov.

http://profiles.nlm.nih.gov/FF/B/B/L/G/_/ffbblg.txt

http://profiles.nlm.nih.gov/FF/B/B/L/G/_/ffbblg.txt

http://profiles.nlm.nih.gov

DIAGNOSIS

Tangier disease is an extremely rare genetic dis-order with fewer than 100 cases reported sincethe index cases in 1961. Clinical signs andsymptoms are variable and may depend on theage of the patients. For instance, signs of neu-ropathy were not observed in T. L. and E. L., incontrast to our 32-year-old patient. Assmannand colleagues (2001) have succinctly compiledthe reported findings (Table 15-1),most of whichare discussed here.

Of all the physical signs of Tangier disease(Table 15-1), perhaps the most remarkable fea-ture is enlarged and lobulated tonsils with a dis-tinctive yellow-orange color. Patients withtonsils removed are often found to have had ahistory of recurrent tonsillitis or oropharyngealobstruction. Another physical diagnostic char-acteristic of the disease is the appearance of 1-to 2-mm discrete orange-brown speckles on therectal mucosa. Similar speckles have also beenreported in the colon, stomach, and intestine.These patients may be prone to frequent stools,intermittent diarrhea, and abdominal pain. Inthe eye, the most frequent finding is slightcorneal opacification.Although about one thirdof Tangier patients have hepatomegaly, this signmight be transient as tests of liver function areoften normal, and foam cells are found only oc-casionally. Many patients with splenomegaly(spleen enlargement) also have mild hemolytic

anemia, stomatocytosis (erythrocytes with deeppitting on the cell membrane due to an alteredmembrane cholesterol:phospholipids ratio)and thrombocytopenia (low platelet count),with 30,000–120,000 platelets/mL (normal is150,000–400,000/mL).

Lymphadenopathy is most often not clinicallymanifested; however, bright yellow plaques anda cholesterol ester content 100-fold higher thannormal have been documented for bothnormal-size and enlarged lymph nodes in Tang-ier patients. Biopsies of bone marrow and theaffected tissues have revealed many foam cellsthat are smaller than those observed in lipidstorage diseases. In addition, these cells containsudanophilic deposits which are not membrane-bound, as is the case for lysosomal storagediseases. Foam cells have also been found inotherwise normal skin, ureters, renal pelvises,tunica albuginea (white fibrous capsule) of tes-ticles, mitral and tricuspid valves, and aorta,coronary, and pulmonary arteries.

About 30% of adult Tangier patients presentwith peripheral neuropathy, the onset ofwhich usually takes place after 10 years of age.These neuropathic symptoms may be subtle orovert, transient or permanent, and includeweakness, increased sweating, diplopia (dou-ble vision), ptosis (abnormally drooping eye-lids), ocular muscle palsies, and diminished orabsent deep-tendon reflexes. There are twoprototypes of neuropathy found in Tangier dis-

Tangier Disease 161

Table 15-1. Frequencies of Clinical Symptoms and Findings in Tangier DiseaseNumber of Cases

Symptoms Present Absent Not Examined/Not Reported

Yellowish tonsils or 44 6 18pharyngeal plaques

Peripheral neuropathy 36 20 12

Splenomegaly 33 17 18

Abnormal rectal mucosa 28 4 36

Hepatomegaly 19 24 25

Corneal opacities 15 28 25

Coronary heart disease 15 33 20

Thrombocytopenia 14 1 44

Anemia or stomatocytes 9 6 53

Lymphadenopathy 6 26 36

Unexplained diarrhea 5 1 62

These data were obtained from the case reports of 66 Tangier patients whose ages ranged from 2 to 72years (Assman et al., 2001). Reproduced with permission from McGraw-Hill Companies, Inc.©2001.

ease, a syringomyelialike syndrome and a mul-tiple mononeuropathy. Of the two, the formeris much more debilitating.

Syringomyelialike syndrome can appear earlyand be progressive, with loss of pain and tem-perature sensation, paresthesia (abnormal skinsensation with no physical cause), and muscleatrophy beginning in the face and distal parts ofthe upper limbs.Eventually,dysesthesia (impairedsensitivity to touch) may progress to the trunkand lower limbs.These symptoms are caused byaxonal degeneration of small myelinated andunmyelinated fibers. Multiple mononeuropathyoften presents with transient, relapsing, or nosymptomatic complications. When manifested,impairments are either mononeuropathic orasymmetrically polyneuropathic and may in-volve cranial nerves. Morphologically, this typeof neuropathy is characterized by de- and re-myelination of peripheral nerves without ax-onal deficit. Atypical lipid accumulation inSchwann cells may be the underlying cause ofthese symptoms.

Profiling of plasma lipoproteins and serumlipids can often aid in the diagnosis of Tangierdisease. There are four classes of lipoproteins:(1) chylomicrons, which transport dietary cho-lesterol and triglycerides from the intestines tothe tissues; (2) very low-density lipoproteins(VLDL) and (3) low-density lipoproteins (LDL),both of which transport de novo synthesizedcholesterol and triglyceride from the liver tothe tissues; and (4) high-density lipoproteins(HDL), which mediate reverse cholesterol trans-port, a process in which excess cholesterolfrom peripheral tissues is transported to theliver.

In all Tangier patients, plasma HDL level isnearly undetectable (optimal is ≥60 mg/dL).Analysis of the major apolipoproteins on HDL bynephelometry (a technique that quantitates spe-cific antigens by analyzing increases in turbiditycaused by antigen-antibody complex formation)often reveals a marked decrease in apoA-I(1%–3% normal) and apoA-II (5%–10% normal).LDL levels are also reduced (∼40% of normal) inTangier patients.Analysis of serum lipids usuallyshows that the total cholesterol averages about78 mg/dL (normal is 130–200 mg/dL). Triglyc-eride levels are usually higher than normal butvary significantly depending on diet.

Cultivated skin fibroblasts from Tangierpatients fail to efflux cholesterol, phosphatidyl-choline, or sphingomyelin to exogenous lipid-free apolipoproteins or artificial amphipathic

polypeptides. Since 1999, definitive diagnosisof Tangier disease can be achieved by sequenc-ing the ABCA1 gene (Lapicka-Bodzioch et al.,2001), which encodes the ATP-binding cas-sette transporter responsible for the efflux ofcholesterol and phospholipids from cells.

Several diseases have overlapping symptomswith Tangier disease, and differential diagnosismust be considered. First, very low plasma levelsof HDL and apoA-I are also consistent with famil-ial dyslipidemias,such as apoA-I and lecithin:cho-lesterol acyltransferase (LCAT) deficiencies andobstructive liver disease. Increased levels ofproapoA-I (the precursor) relative to that of themature apoA-I, a distinctive feature of Tangier dis-ease, can be demonstrated by isoelectric focus-ing and be used to eliminate the possibility of afamilial dyslipidemia. The hypercholesterolemicsymptom found in obstructive liver disease dis-tinguishes it from Tangier disease. Second, thepresence of foam cells may indicate a lipid stor-age disorder (e.g., Niemann-Pick disease), butthese disorders are not usually associated withthe tonsillar abnormalities and low HDL levels. Inaddition, the absence of membrane surroundingthe lipid deposits in foam cells is a clear indica-tion of Tangier disease and not a storage disorder.Finally, low plasma cholesterol levels are also ob-served in abetalipoproteinemia, but patientswith Tangier disease frequently also have ele-vated triglyceride and much lower HDL levels.


Almost four decades after the index cases ofTangier disease were reported, the gene respon-sible for the disease was identified as ABCA1 byseveral groups (Bodzioch et al., 1999; Brooks-Wilson et al., 1999; Remaley et al., 1999; Rust etal., 1999).ABCA1 belongs to the ABC superfam-ily of ATP-binding cassette transporters, whichare responsible for the translocation across cel-lular and intracellular membranes of varioussubstances, including peptides, metabolites,drugs, and lipids. As one might expect fromtheir ubiquitous presence and high degree ofconservation, mutations in a number of thesetransporters have been incriminated in severalhuman diseases, such as cystic fibrosis, maculardegeneration, and immunodeficiency syn-dromes. We discuss next the role of ABCA1 inlipid metabolism and how mutations in thistransporter result in Tangier disease.


Reverse Cholesterol Transport andABCA1-Mediated HDL Formation

ABCA1 mediates the first step in the energy-dependent efflux of cholesterol from the cell toform HDL for reverse cholesterol transport (Fig.15-2). While all tissues in the body can synthe-size cholesterol, only the liver and steroidogenictissues can metabolize it. Surplus cholesterol incells of the peripheral tissues is transported tothe liver for either redistribution to other cellsor for excretion either as free cholesterol or as abile salt after conversion in the liver. Therefore,this reverse cholesterol transport system plays apivotal role in cholesterol homeostasis withHDL as one of the key players.

Due to dynamic remodeling within differentHDL subfractions and between HDL and otherlipoproteins, HDL is a heterogeneous class oflipoproteins that can be generalized by a highdensity (>1.063 g/mL) and a small size (5- to 17-nm diameter). Differences are characterized byshape, density, size, charge, and antigenicity,

which are influenced by the lipid andapolipoprotein (e.g.,A-I,A-II, C-II, E) content.

The formation of discoidal, nascent HDL par-ticles commences with the two-step lipidationof apoA-I by ABCA1. These steps consist of theefflux of phospholipid followed by that of cho-lesterol and involve different mechanistic ac-tions of ABCA1, the details of which arecurrently unknown. The function of ABCA1 wasshown recently to include modification of lateendocytic membranes to bring together poolsof cellular cholesterol and those of endocy-tosed apoA-I (Neufeld et al., 2004). ABCA1present on the cell surface stimulates apoA-Iuptake into endosomes.These endosomes con-taining ABCA1 and apoA-I then travel to late en-docytic compartments where ABCA1 appearsto mediate the conversion of late endocyticpools of cholesterol to pools that associate withapoA-I. The lipidated apoA-I-harboring late en-docytic vesicles may then move to the cell sur-face where their contents are released as nascentHDL particles. ABCA1-mediated lipidation of

Tangier Disease 163

Figure 15-2. A simplified schematic of cholesterol transport.Cholesterol travels to non-hepatic cells, such as the macrophage, via VLDL and LDL particles, while excess choles-terol is shuttled to the liver via HDL particles. Note that ABCA1 mediates nascent HDLformation by translocating cellular cholesterol and phospholipids to apolipoprotein A-I(apoA-I) in an active,energy-dependent reaction.CETP,cholesteryl ester transfer protein;LCAT, lecithin:cholesterol acyltransferase; LDLR, low-density lipoprotein receptor; SR-B1,scavenger receptor B1.

apoA-I could also take place at the plasma mem-brane and in early endocytic vesicles.

The nascent HDL particles change shape andcomposition as they acquire additional freecholesterol by passive cellular diffusion of freecholesterol from cell membranes or from otherplasma lipoproteins.HDL surface-localized LCATprogressively converts the free cholesterol onthe surface of the particles to cholesterol ester,which occupies the core of the lipoprotein par-ticle. This process converts the shape of theHDL particles from discoidal to spherical. Thelipid unloading of HDL in the liver follows atleast two pathways. In the first route, the cho-lesterol ester transfer protein (CETP) mediatescholesterol ester transfer from HDL to VLDLand LDL in exchange for triglyceride; LDL inturn are taken up by the liver via the LDL recep-tor. In the second route, HDL binds to the scav-enger receptor B1, and cholesterol esters areselectively taken into the liver cells without in-ternalization of HDL proteins (Fig. 15-2).

In Tangier disease in which ABCA1 is defec-tive, formation of nascent HDL particles is com-promised, leading to the observed clinicalfindings. First, cellular levels of cholesterol arehigh in the absence of active reverse choles-terol transport. Second, lipophilic compoundssuch as vitamin E, retinyl esters (yellow), andcarotenoids (orange) also accumulate inside thecell, perhaps due to their decreased efflux re-sulting from the ABCA1 mutation, thus account-ing for the yellow-orange appearance of tissueplaques in Tangier disease. Third, failure to as-semble functional HDL particles leads to rapidrenal filtration and catabolism of apoA-I, thusaccounting for the decreased level of HDL inplasma.Any residual HDL-like particles in Tang-ier plasma are small and lipid poor. Last, plasmatriglyceride levels are often high (two- to three-fold greater than normal) in adult patients. Theelevated accumulation of triglyceride in VLDLand chylomicrons appears to result from thefailure to activate lipoprotein lipase due to theabsence of its activator apoC-II, which is nor-mally transferred from HDL to VLDL.

ABCA1

The ABCA1 gene has been localized to 9q31onthe long arm of chromosome 9. It spans 149 kband has 50 exons. The first intron has beenshown to act as an alternative promoter.ABCA1is ubiquitously expressed, with notable expres-sion in liver, brain, and steroidogenic tissues.Furthermore, it is expressed in confluent but

not growing fibroblast cell culture, indicatingthe balance between cholesterol efflux and in-tracellular requirement.Consistent with this ob-servation, overloading macrophages withcholesterol increased ABCA1 transcription dra-matically. This induction is mediated by thetranscription activators liver X receptor (LXR)and retinoid X receptor (RXR), which act as ob-ligate heterodimers in the presence of eitheroxysterols or retinoic acid (Costet et al., 2000).LXR/RXR binds to the DR4 elements in theABCA1 promoter and first intron.Hence, the ac-cumulating cholesterol must be converted toan oxysterol to elicit ABCA1 induction.

ABCA1 is a 2261-amino acid integral mem-brane protein consisting of two transmembranedomains, each with six transmembrane-spanninghelices, connected by a regulatory domain richin Ser and Thr residues.It also has two nucleotide-binding folds (NBFs) where ATP is bound and hy-drolyzed, thus generating the energy needed forABCA1 action. Each NBF, also known as an ATP-binding cassette, contains a signature motif (orC motif ) inserted between two conserved re-gions, the Walker A and Walker B motifs. Morethan 50 mutations have been identified through-out the gene; these include various missense,nonsense, insertional, deletional, and even in-tronic mutations that can affect splicing (Sin-garaja et al., 2003).The genetic defect in theoriginal Tangier disease kindred (case report 1)was determined to be a homozygous dinu-cleotide deletion (3283–3284TC) in exon 22,resulting in a frameshift at amino acid 1055 anda premature stop codon after amino acid 1084(Remaley et al., 1999).

Although definitive answers are just forth-coming as to why these various mutations inABCA1 cause a reduction in its ability to medi-ate cholesterol efflux, possibilities include mis-localization that causes a reduction in thefraction of ABCA1 that reaches the plasma mem-brane; altered NBFs that fail to bind and utilizeATP, thereby compromising active transport; andweaker interactions between ABCA1 and apoA-Ithat reduce lipidation of the latter. Establishingstructure–function relationships in ABCA1 willbe important in further elucidating the bio-chemical basis for Tangier disease.

ABCA1 Zygosity and Coronary Artery Disease

Tangier disease is inherited clinically as an auto-somal recessive trait but biochemically as anautosomal codominant trait. This is because


clinical phenotypes (e.g., hyperplastic tonsils)appear only in homozygotes, but biochemicalphenotypes (e.g., reduced ABCA1 activity andcholesterol efflux) are also present in heterozy-gotes. This dichotomy can be appreciated bybriefly examining the pathophysiology of the dis-ease.

The main activity of tissue macrophages is toengulf apoptotic and necrotic cells.As the effluxof cholesterol extracted from the phagocytosedcell membranes is ABCA1-dependent, loss offunction of both ABCA1 alleles (in homozy-gotes) leads to tissue hyperplasia. As the het-erozygotes do not have symptoms of lipid-loadedtissue macrophages characteristic of Tangier dis-ease, one functional ABCA1 allele may suffice toprevent cholesterol deposition in macrophages.

While the homozygotes display the variousphenotypes typical for Tangier patients, the het-erozygotes are associated with another disorderknown as familial HDL deficiency,which is asso-ciated with an increased risk of coronary arterydisease (CAD). Despite the well-documented in-verse correlation between HDL levels and pre-mature CAD, homozygotes with virtually nocirculating HDL do not have a markedly in-creased risk of CAD compared to heterozy-gotes, who have half-normal levels of HDL. Theunderlying reason for this counterintuitive find-ing may be the low LDL levels (40% of normal)found in the homozygotes (Oram, 2000).

Why do low LDL levels confer antiathero-genic benefits? In arterial macrophages, most ofthe cholesterol comes from ingestion of oxi-dized LDL generated by reactive oxygen speciesin the bloodstream. To prevent endothelial in-jury, the oxidized LDL is engulfed by arterialmacrophages. The combined assault of an im-paired cholesterol efflux activity and low serumHDL levels (as in both homozygotes and het-erozygotes) results in the transformation ofthese LDL-engorged arterial macrophages intofoam cells due to their failure to egress their cel-lular stores of cholesterol. Under continued ox-idative stress, these foam cells necrose and causeinflammation, stimulating further macrophagerecruitment to the site and proliferation ofsmooth muscle and neointima, leading to athero-sclerotic plaque formation (Frishman, 1998).Therefore, low serum LDL levels are consideredto be atheroprotective.

This discussion highlights the interplaybetween pro- and antiatherogenic factors in dic-tating the overall risk of CAD in heterozygotesand homozygotes. However, why are LDL levelspreferentially decreased in the Tangier homozy-

gotes? Since a significant amount of cholesterolin LDL originates from HDL via the action ofCETP, a decrease in HDL would be expected tolower LDL levels. Moreover, alterations in theHDL-mediated return of cholesterol to the livermight enhance LDL receptor activity andthereby facilitate a more rapid clearance ofplasma LDL. The fact that Tangier patients dodevelop CAD despite their strikingly low LDLlevels suggests that other proatherogenic ab-normalities in lipid metabolism (such as the ele-vated levels of triglycerides, which can alsocause arterial hardening) might partially offsetthe benefits of lower LDL levels.These observa-tions shed light on why ABCA1 mutations areassociated with only a moderately increasedrisk of atherosclerosis, albeit not as high as thatobserved in other lipidemias, such as familialhypercholesterolemia.

THERAPY

Although there is no specific treatment forTangier disease, symptom relief and alleviationof complications can improve the quality of lifefor the patients. Families with Tangier patientsmay also benefit from genetic counseling.

Surgical removal of tissues may be necessaryin some cases to alleviate symptoms, such asoropharyngeal obstruction (tonsillectomy) andprogressive anemia and thrombocytopenia(splenectomy).

Supplements frequently prescribed for pa-tients with neuropathy, such as omega-3 fattyacids, antioxidants, and vitamin E, have notbeen effective in preventing the progression ofneurological symptoms in patients with Tangierdisease. Therefore, pain-relieving medicationsand physiotherapy are prescribed for these pa-tients. Decompression surgery and nerve resec-tion procedures may be used to alleviateintractable pain. The strength of the musclescan be improved by exercise, electrical stimula-tion, and physiotherapy.

Drugs used to increase HDL levels (fibrates,nicotinic acid, and statins) in otherwise nor-mal people do not have the same effect in pa-tients with Tangier disease. Therefore, it isnecessary to identify and treat other risk fac-tors associated with CAD. Exercise, weight re-duction, dietary cholesterol and saturated fatreduction, and smoking cessation are the firstline in management of low HDL cholesterol.Dietary management with low fat intake isbeneficial in reducing the risk for CAD, as well

Tangier Disease 165

as in preventing hepato- and splenomegaly.Al-though ineffective in increasing plasma HDLlevels, fibrates have been useful in loweringtriglyceride levels in patients with Tangier dis-ease. In addition, drugs that lower HDL choles-terol, such as β-blockers, benzodiazepines, andandrogens, should not be taken (if possible).

An exciting and intensively investigated areafor therapeutic intervention in Tangier diseaseis the use of drugs to increase the expression ofABCA1 at the transcriptional level. Even overex-pression of a mutant version that possesses lessthan wild-type activity could confer some bene-fit. However, the success of such approacheswill be predicated on the ability to elicit tran-scriptional activation of ABCA1 specifically.

The prescient remarks of Fredrickson in1961 that investigations into Tangier diseasewill shed light on poorly understood lipid andfat metabolic pathways have now been borneout.An unanticipated bonus has been the iden-tification of a molecular target for pharmacolog-ical manipulation might help lower the risk ofCAD in the normal population.

QUESTIONS

1. In general, do you expect biochemicallydominant phenotypes for inherited reces-sive disorders? Using the reasoning foryour answer, explain why patients withTangier disease may have a lower risk ofCAD than heterozygotes.

2. What would be the serum lipid andlipoprotein profile of a person homozy-gous or heterozygous for a mutation inthe ABCA1 gene?

3. Based on your understanding of the roleof ABCA1 in the formation of HDL parti-cles, how would you design a therapeuticmodality to overcome a weakened interac-tion between ABCA1 and apoA-I that re-sults from a mutation in ABCA1 gene?

4. If an ABCA1 mutation does not com-pletely eliminate the cholesterol efflux ac-tivity, then what approach would you taketo enhance the level of expression of thismutant protein to alleviate the pheno-typic consequences?

5. How would you modulate the reversecholesterol transport pathway to increasecholesterol efflux from cells in normal in-dividuals?

Acknowledgment: We thank Dr. Alan Remaley, NIH, forcritical reading of the manuscript and providing valuablesuggestions for improvement.

BIBLIOGRAPHY

Assmann G, von Eckardstein A, Brewer JB Jr.: Famil-ial analphalipoproteinemia: Tangier disease, inScriver CR, Beaudet AL, Sly WS, et al. (eds): TheMetabolic and Molecular Bases of InheritedDisease. 8th ed. McGraw-Hill, New York, 2001,pp. 2937–2960.

Bodzioch M, Orsó E, Klucken J, et al.:The gene en-coding ATP-binding cassette transporter 1 is mu-tated in Tangier disease. Nat Genet 22:347–351,1999.

Brooks-Wilson A, Marcil M, Clee SM, et al.: Muta-tions in ABC1 in Tangier disease and familialhigh-density lipoprotein deficiency. Nat Genet22:336–345, 1999.

Costet P, Luo Y, Wang N, et al.: Sterol-dependenttransactivation of the ABC1 promoter by theliver X receptor/retinoid X receptor. J BiolChem 275:28240–28245, 2000.

Fredrickson DS,Altrocchi PH,Avioli LV, et al.: Tang-ier disease—combined clinical staff conferenceat the National Institute of Health. Ann InternMed 55:1016–1031, 1961.

Frishman, WH: Biologic markers as predictors ofcardiovascular disease.Am J Med 104A:18S–27S,1998.

Lapicka-Bodzioch K, Bodzioch M, Krull M, et al.:Homogeneous assay based on 52 primer sets toscan for mutations of the ABCA1 gene and itsapplication in genetic analysis of a new patientwith familial high-density lipoprotein deficiencysyndrome. Biochim Biophys Acta 1536:42–48,2001.

Neufeld EB, Stonik JA, Demosky SJ, et al.: TheABCA1 transporter modulates late endocytic traf-ficking. J Biol Chem 279:15571–15578,2004.

Oram JF:Tangier disease and ABCA1. Biochim Bio-phys Acta 1529:321–330, 2000.

Remaley AT, Rust S, Rosier M, et al.: Human ATP-binding cassette transporter 1 (ABCA1): ge-nomic organization and identification of thegenetic defect in the original Tangier diseasekindred. Proc Natl Acad Sci U.S.A. 96:12685–12690, 1999.

Rust S, Rosier M, Funke H, et al.: Tangier diseaseis caused by mutations in the gene encodingATP-binding cassette transporter 1. Nat Genet22:352–355, 1999.

Singaraja RR, Brunham LR,Visscher H, et al.: Effluxand atherosclerosis—the clinical and biochemi-cal impact of variations in the ABCA1 gene.Arterioscler Thromb Vasc Biol 23:1322–1332,2003.


CASE REPORT

The patient was a 34-year-old woman of Ashke-nazic Jewish ancestry.When she was four yearsold (September 1974), her parents noticed thatshe had frequent nosebleeds and bruised easily.In addition, she tired with little exertion andlooked pale. They took her to the emergency de-partment, and on admission to the hospital,a bone marrow biopsy revealed the presenceof Gaucher cells as seen in Figure 16-1. She wasdiagnosed as having Gaucher disease, an inbornerror of metabolism affecting β-glucosidase (alsoknown as acid β-glucosidase or glucocerebrosi-dase).White blood cell (WBC) β-glucosidase as-says were performed on the patient and herimmediate family members. Her parents andonly sibling had slightly reduced levels of WBC β-glucosidase, whereas the patient’s β-glucosidaseactivity was only 9% of the control mean level,thus confirming the diagnosis of Gaucher dis-ease on a biochemical-enzymatic basis. In addi-tion, she was found to have an elevated level ofserum acid phosphatase (ACP) activity, a charac-teristic of most patients with Gaucher disease.

At age 15 years, her spleen had greatly ex-tended into the pelvis and right lower quadrantof the abdomen, and she subsequently under-went splenectomy to alleviate thrombocyto-penia ( low platelet count) aggravated by thissplenomegaly. The pathologist reported thather spleen weighed 2070 g (normal range is100–170 g), which is over 15 times normal size.Further microscopic examination of the spleenand of biopsies from the liver and adjacentlymph nodes confirmed the presence of Gaucher

cells. Two months following the surgery, herhematological parameters were restored to nor-mal levels: Her platelet count rose from 80 × 109

to 291 × 109/L (normal range is 140 × 109 to440 × 109/L), and her hemoglobin had increasedfrom 12.5 to 15.0 g/dL (normal range is13.5–16.0 g/dL).

The patient felt well after her surgery untilseveral months later, when she developed painin her left hip. She began to limp to reduce herpain and resorted to the use of crutches to min-imize the pressure placed on her left hip. Evenwith the use of crutches, however, she contin-ued to have pain when walking. Prolonged sit-ting also caused her discomfort, particularlywhen rising. Radiological examination (Fig. 16-2)revealed the classical picture of Gaucher dis-ease: osteonecrosis (death and degradation ofbone tissue) of the left hip, with significant col-lapse of the head of the femur. In May 1990, sheelected to undergo hip replacement surgery.The left femoral head was resected, and a totalhip replacement device and acetabular implantwere inserted.

Beginning in June 1991, the patient was ad-ministered a newly approved intravenous drug,Ceredase (the commercial name of a placental-derived glucocerebrosidase preparation). Shewas given a comprehensive evaluation at thebeginning of the treatment and was monitoredevery 3 months thereafter. The initial assess-ment included a full blood analysis (hemoglo-bin, platelet count, biochemical markers) inaddition to visceral, skeletal, and pulmonaryevaluations. At her first evaluation after begin-ning therapy, her hematological parameters had

16

Gaucher Disease: A Sphingolipidosis

WILLIAM C. HINES and ROBERT H. GLEW

167

improved somewhat, but they had not nor-malized. Nevertheless, she reported increasedstamina and a sense of well-being followingCeredase administration.

The patient experienced several bone crises(severe bone pain caused by ischemia fromswelling and decreased blood supply) duringthe first year of treatment, but the duration andintensity were less severe than prior to enzymereplacement. Following the first year of treat-ment, a progressive reduction in generalizedbone pain was noted.With subsequent attemptsto titrate down the costly Ceredase dosage, thepatient reported a gradual return of bone painand lethargic feeling. These symptoms de-creased after her dosage was returned to thehigher levels. As shown in Table 16-1, over thecourse of 4 years of treatment, there was an im-provement in the patient’s hematological in-dices, including significant increases in plateletcount and size. Furthermore, there was a no-table reduction of liver volume. In 1995, a re-combinant form of glucocerbrosidase,Cerezyme,was approved by the Food and Drug Adminis-tration. In 1996, the patient was switched tothis less-expensive form and has been receiv-ing it ever since. She remains stable to this day

and has not reported any additional compli-cations.

DIAGNOSIS

Patients suffering from Gaucher disease areplaced in one of three clinical subtypes (Table16-2). The most common form, type I, originallydescribed by Gaucher, is also termed chronicnonneuropathic or adult form Gaucherdisease—even though the onset of symptomsin two thirds of patients occurs during child-hood. This form of the disease usually follows achronic progressive course characterized by he-patosplenomegaly, anemia, thrombocytopenia,and erosion of the long bones.The range of clin-ical severity is great; some patients suffer in-tensely before their third decade of life,whereasothers remain undiagnosed into their sixth orseventh decade of life, hence the adult-onsetdesignation.

The other, and much rarer, clinical subtypes,type II (acute neuropathic or, less commonly,malignant or infantile) and type III (chronic neu-ropathic or, less commonly, juvenile or subacuteneuropathic), are distinguished from type I, as


Figure 16-1. Wright-stained Gaucher cells obtained from bone marrow showing thetypical wrinkled appearance of the cytoplasm (magnification 175×).

Figure 16-2. (A) X-ray of the patient’s normal-appearing hip taken in 1982 and (B) acollapsed and necrotic left femoral head in 1986.

169

their names indicate, by the presence of neuro-logical involvement. These neurological formsare distinguished from each other by their clini-cal severity. Type II Gaucher disease is the mostdrastic form, with symptoms occurring earlierand death usually following the third year of life,whereas type III Gaucher disease usually ap-pears later in childhood and displays a slowercourse of progression.

Gaucher disease is the most common of thelysosomal storage diseases. It is an autosomal re-cessively inherited disorder that is prevalentwithin all populations; however, it is mostprevalent in members of the Ashkenazic Jewishpopulation, occurring at a rate of 1 in 450 indi-viduals (the heterozygote “carrier” frequency isabout 4%). In fact, it is the most frequent gene-tic disorder within this population. Conse-quently, descendants of this lineage account forthe majority of cases within the United States,accounting for nearly two thirds of all Gaucherdisease cases.

The suspicion of Gaucher disease often oc-curs when a relative of the patient suffers thedisease; however, diagnosis is not always thisstraightforward. In the absence of any suspicions

due to bloodlines, it may be difficult to identifypatients with Gaucher disease because the dis-ease shares signs and symptoms with other dis-orders. Therefore, physicians from severaldifferent specialties (pediatrics, hematology,pathology, radiology, orthopedics, neurology)should be prepared to suspect Gaucher diseaseon observing a patient with one or more of thefollowing: reduced platelet count, abnormalx-ray, frequent nosebleeds, fatigue, bone pain,frequent fractures, enlarged abdomen, cognitiveproblems,myoclonic seizures,or disturbances ingait (characterized by a particular manner ofmoving the feet or walking—a common distur-bance indicating neurological involvement).

Bone marrow biopsies are commonly per-formed to investigate the origin of cytopenias(reduced blood cell counts), and the diagnosisof Gaucher disease is often made or confirmedby a pathologist, who identifies the highlycharacteristic “Gaucher cells” within thestained aspirate of bone marrow. Gaucher cells,which are pathognomonic for Gaucher disease,are mononuclear cells of the macrophage-reticuloendothelial system (Figs.16-1 and 16-3).They occur in abundance in sites of this sys-tem, namely, the spleen, liver, and bone marrow.They occur elsewhere in reduced numbers,usually in the lymph nodes and lungs.With pa-tients having the neurological forms of Gaucherdisease, these unique storage cells are alsofound in the brain, particularly in the perivascu-lar areas. When cellular aspirates are treatedwith the Wright stain (a commonly used stainfor blood and bone marrow that consists of amixture of basic [methylene blue derivates] andacidic [eosin] dyes), Gaucher cells appear largewith an abundant cytoplasm resembling wrin-kled tissue paper. The nucleus is usually ovaland positioned eccentrically. The wrinkled, stri-ated appearance of the cytoplasm is due tothe accumulation of the glucocerebroside-rich


Table 16-1. Therapeutic Response to EnzymeReplacement Therapy

December October Therapeutic Response 1991 1995

Liver volume (cc) 3151 2271

RBC count (106/mm3) 3.87 4.67

Hemoglobin (g/dL) 12.4 13.9

Hematocrit (%) 36.8 42.6

MCV (fL) 95 91

Platelet count (103/mm3) 133 326

MPV (fL) 10.3 93.0

MCV, mean corpuscular volume; MPV, mean platelet volume;RBC, red blood all.

Table 16-2. Clinical Features of the Gaucher PhenotypesClinical Features Type I Type II Type III

Onset Child/adult Infancy Juvenile

Hepatosplenomegaly + + +Hypersplenism + + +Bone crises/fractures + − +Neurodegenerative course − +++ ++Life expectancy Variable <2 years Third to fourth decade

Ethnic predilection Ashkenazic Panethnic (Norrbottnian) Swedish

storage material referred to as Gaucher deposits(Fig. 16-1). This finding, although indicative ofGaucher disease, can occur in other diseases,such as thalassemia or chronic granulocyticleukemia.Complicating the issue, the bone mar-row aspiration procedure may not provide asample (a “dry” tap) or contain enough Gauchercells needed to make a diagnosis. Therefore, theβ-glucosidase enzymatic assay, specific only toGaucher disease, should always be performedto make a definitive diagnosis.

Although the β-glucosidase assay can be per-formed on any tissue, the tissues of choice areperipheral blood leukocytes or cultured fibrob-lasts grown from a punch biopsy of the skin.Antenatal diagnosis can be made using extractsof cultured amniotic cells obtained from amnio-centesis. There are several different β-glucosidesubstrates that can be used in the assay: the nat-ural substrate glucocerebroside or structurallysimilar, artificial substrates.

In the case of the glucocerebrosidase-dependent assay, the two products, glucose andceramide, are separated based on their differingsolubilities to assess the degree of enzymaticactivity. It is convenient to use [3H]glucocerebro-side labeled in the glucose moiety ([3H]glucose-ceramide) because it is soluble in chloroform and

insoluble in water, whereas the radiolabeledproduct of the reaction, [3H]glucose, is watersoluble:

[3H]Glucocerebroside + H2O→ [3H]Glucose + Ceramide

After incubating the substrate and tissueextract at 37°C in the appropriate buffer (0.2 Msodium acetate, pH 5.5) and in the presenceof activator lipid (e.g., sodium taurocholate,2% w/v), the reaction is terminated by extractionof the incubation medium with CHCl3-methanol(2:1 v/v); the top layer (the water phase) contain-ing [3H]glucose is counted in a liquid scintilla-tion spectrometer to give a measure of theextent of reaction.Extracts from tissues from per-sons with Gaucher disease will usually generateless than 10% as much [3H]glucose as controls.

Since radiolabeled molecules are hazardousand difficult to dispose, most researchers andclinical laboratories prefer to use nonisotopicmethods of labeling molecules. For this reason,investigators have established β-glucosidaseassay conditions that allow the estimation ofglucocerebrosidase content of human tissueswhen nonphysiological β-glucosides serve asthe substrate. The artificial substrate of choice,

Gaucher Disease: A Sphingolipidosis 171

Figure 16-3. Splenic Gaucher cells, light micrograph.

and in widest use today, is 4-methylumbelliferyl-β-D-glucopyranoside (MUGlc). MUGlc is non-fluorescent, but one of the products of thereaction, 4-methylumbelliferone, is intenselyfluorescent when the solution is brought to analkaline pH (Fig. 16-4).Thus, the extent of thereaction may be determined by measuringthe quantity of fluorescence generated by aparticular tissue extract and comparing it tothe fluorescence of standard solutions of 4-methylumbelliferone.

Complicating matters, however, is the pres-ence of a second β-glucosidase in most humantissues, particularly leukocytes. This second β-glucosidase, which is not associated withGaucher disease, does not cleave glucocerebro-side and is not deficient in most cases ofGaucher disease; however, it does cleave the ar-tificial substrate MUGlc, which can lead to mis-leading results when this artificial substrate isused in the diagnostic assay. Fortunately, it wasfound that the inclusion of bile salts, such assodium taurodeoxycholate, in the assay media(used to keep the substrate soluble) inhibitedthe activity of the second β-glucosidase, whilestimulating the activity of lysosomal glucocere-brosidase. Thus, one can use the nonphysiologi-cal β-glucoside (MUGlc) within the assay andremain confident that the activity is from thelysosomal, and Gaucher-specific, β-glucosidase

and not from the coexisting nonspecific iso-enzyme.

Although not currently used for the diagno-sis, there are other enzymes (serum acid phos-phatase, chitinase) that are of significant clinicalvalue due to their ability to signal abnormalphysiological changes, and these are used tomonitor the progression of Gaucher disease.These cell-specific enzymes, whose activitiesthat are usually low in the serum,are commonlyelevated in disorders that cause bone degra-dation or macrophage destruction. The firstserological enzymatic marker identified as cor-related with Gaucher disease was serum acidphosphatase (ACP). The elevation of ACP activ-ity occurs in most, but not all, cases of Gaucherdisease. Hence, there is the possibility that a pa-tient may have a normal serum ACP level.There-fore, the diagnosis of Gaucher disease cannotbe ruled out just because of normal serumACP activity. Nevertheless, the finding of anelevated serum ACP value along with otherclinical symptoms should heighten suspicion ofGaucher disease (95% of patients with Gaucherdisease exhibit an elevated serum ACP value).One confounding problem regarding the diag-nostic utility of the ACP test is its lack of speci-ficity. The sera of patients with other lysosomalstorage diseases, such as Niemann-Pick disease,may also exhibit elevations of ACP. This is whyACP is not used to make a diagnosis; however,once the diagnosis of Gaucher disease has beenmade, results of this enzymatic test are highlyuseful as they closely reflect the onset and pro-gression of common symptoms and sequelae.

Another enzyme,chitotriosidase (also knownas chitinase) has also been found to be ele-vated in the plasma of patients with Gaucherdisease. This enzyme, which is synthesized inmacrophages and Gaucher cells, is believed tobe involved in the degradation of chitin wallsof microorganisms. It is a stable protein and isreadily analyzed; however, a drawback is thehigh frequency of a null allele within the generalpopulation that leads to the complete absenceof activity in 3%–5% of individuals. Nevertheless,because plasma chitotriosidase activity is soclosely associated with glucocerebroside storageand clinical manifestations, it has recently be-come an important clinical marker for Gaucherdisease.

A new marker, a chemokine named CCL18/PARC, has recently been identified to be in-creased within the serum of patients withGaucher disease, much like that of chitotriosi-


Figure 16-4. The β-glucosidase reaction illustrat-ing use of the artificial substrate MUGlc (4-methylumbelliferyl-β-D-glucopyranoside).

dase.However,unlike chitotriosidase,null allelesare not believed to exist in the general popula-tion; thus, this new marker may take the place ofchitotriosidase or provide a mechanism to moni-tor the clinical manifestations within individualswho have chitotriosidase null alleles.

It should be emphasized that, although theenzymatic assays have proved 100% effective indiagnosing Gaucher disease, the β-glucosidaseassay is not completely reliable in identifyingGaucher disease heterozygotes. Experience hasshown that when using the β-glucosidase assay,about 1 of every 10 obligate heterozygotes willbe incorrectly assigned to the control category.Thus, identification of heterozygotes must relyon other methods; in this case, molecular meth-ods are the preferred and logical choice. Molec-ular diagnosis of Gaucher disease relies on theidentification of the mutation(s) within the glu-cocerebrosidase gene. There are several differ-ent molecular strategies for identifying patients,and the most common methods include com-plete gene sequencing, restriction site analysis,or detection based on polymerase chain reac-tion (PCR). These analyses are easily performedon DNA isolated from blood leukocytes or buc-cal (cheek) cells. Each method has its limita-tions, but when used in combination, they cancorrectly identify carriers of β-glucosidase mu-tations in nearly all cases.


Gaucher disease is the most common of thesphingolipidoses, a group of related genetic dis-eases that result from the impaired lysosomalenzyme function. Figure 16-5 shows the rela-tionships among the lysosomal enzymes thatconstitute the pathway of sphingolipid catabo-lism. The substrates and products that comprisethe pathway are linked by a series of irre-versible reactions that remove a sugar moiety, asulfate residue, or a fatty acid; all of these reac-tions are hydrolytic (use water for cleavage),and the product of one reaction becomes thesubstrate for the next enzyme in the sequence.In the case of sugar residues, the oligosaccha-ride domains are disassembled by removal ofmonosaccharides one at a time from the nonre-ducing end of the sugar chain (the reducing endof sugar describes the anomeric carbon contain-ing the free carbonyl group; the nonreducingend does not have a carbonyl group, (e.g.,carbon-4 of glucose)). All reaction sequences

converge on ceramide, which is degraded inturn by an amidase called ceramidase.

Interestingly, all of the enzymes that com-prise the sphingolipid-catabolic pathway areglycoproteins and may be found in differentplaces within the organelle. For example,glucocerebrosidase is firmly associated withthe lysosomal membrane, whereas others,like hexoseaminidase A, exist largely in solu-ble form in the lysosomal matrix. A commonproperty of the lysosomal hydrolases, how-ever, is their expression of maximum activityat a relatively acidic pH (i.e., pH 4.0–5.5),hence the term acid hydrolase. This is not un-expected because ATP-driven proton pumpssustain an acidic milieu (pH 5.2) within thelysosome.

All tissues of the patient under discussion, likeeach patient with Gaucher disease, are markedlydeficient in lysosomal glucocerebrosidase activ-ity. Glucocerebrosidase catalyzes the reactionshown in Figure 16-6, and insufficient activityresults in the accumulation of substrate withincells, such as macrophages, where glucocere-broside precursors are being catabolized. Mostof the glucocerebroside stored in the liver,spleen, and bone marrow is derived from thecatabolism of membranes of white blood cells(WBC) and red blood cells (RBC) by themacrophages of the reticuloendothelial system.

In Gaucher disease, as the oligosaccharidechains of the sphingolipids are catabolized, me-tabolism is aborted at the level of glucocerebro-side.This lipid molecule self-associates to formbilayers when it accumulates and will subse-quently stack to form membranous sheets thatcreate the histological tissue paper appearancewithin cells. The insolubility of glucocerebro-side in water or bodily fluids is problematic be-cause the lipid cannot be readily excreted inthe urine, and only a relatively low rate of ex-cretion occurs by way of the hepatic-biliaryroute. Thus, glucocerebroside molecules aggre-gate and accumulate in the lysosomes of cells(Fig. 16-7) within spleen and liver, giving rise tosplenomegaly and hepatomegaly. The enlargedspleen and liver in turn destroy RBC,WBC andplatelets prematurely, thereby contributing tothe anemia, leukopenia, and thrombocytopeniathat are common in patients with Gaucher dis-ease.

The storage cells in the bone marrow have amyelophthisic effect; that is, they crowd out thenormal hematopoietic tissues. As a result ofGaucher cell accumulation in the bone marrow,


Gaucher patients often suffer from skeletalinvolvement; this includes frequent fracturesand severe bone and joint pain, particularly inthe femur, and is accompanied by swelling andtenderness of the surrounding tissue. Theseacute attacks, or “bone crises,” are a major im-pairment for patients with Gaucher disease.Avascular necrosis and collapse of the femoralhead occur in many patients and are often se-vere enough to require hip replacement sur-gery, as exemplified by the subject of the casereport. Frequently seen in younger patients un-dergoing rapid bone growth is the “Erlenmeyerflask” deformity that arises from expansion ofthe cortex of the distal femur.

The patient described in the present reporthas the type I form of Gaucher disease and is

spared central nervous system (CNS) involve-ment. However, in infants with the type II formof the disease, the brain contains Gaucher cells,and CNS involvement is extensive.This form ischaracterized by strabismus (lack of parallelismof the visual axes of the eyes), muscle hyper-tonicity (spasticity), retroflexion of the head,and failure to thrive. These infants usually diebefore the age of 2 years. Patients with type IIIGaucher disease usually live longer than type IIpatients, but they also suffer from neurologicalinvolvement. Gaucher cells are also present inneuronal and nonneuronal tissues from type IIIpatients, whose condition, although similar tothat of type II patients, is usually less severe.

The accumulation of storage lipid withinlysosome-rich cells of the reticuloendothelial


Figure 16-5. The pathway of sphingolipid catabolism.Diseases that result from specificenzyme deficiencies are as follows: (1) GM, gangliosidosis; (2) GM2 gangliosidosis (Tay-Sachs disease); (3) sialidosis; (4) Fabry disease; (5) Gaucher disease; (6) Niemann-Pick dis-ease; (7) Krabbe disease; (8) metachromatic leukodystrophy; (9) Farber disease.Cer, Ceramide; Glc, glucose; Gal, galactose; GalNAc, N-acetylgalactosamine; NANA,N-acetylneuraminic acid.

system produces other effects, including therupture of glucocerebroside-laden cells. Afterrupturing, the cellular contents of these cellsenter the lymphatic system and eventually thebloodstream,causing elevated enzyme activities(ACP, chitinase) that are common in patientswith Gaucher disease.

Glucocerebrosidase has a subunit molecularweight of about 67,000, and these inactive sub-units have a tendency to aggregate into catalyt-

ically active dimers, tetramers, and even largerforms. The enzyme is firmly embedded in thelysosomal membrane, and solubilization re-quires extraction with detergents such as Tri-ton X-100 or bile salts (e.g., sodium cholate).Subsequent separation of enzyme protein anddetergents using a butanol extraction step notonly renders glucocerebrosidase soluble inaqueous media, but also removes endogenous,natural lipid activators that are found within


Figure 16-6. The reaction catalyzed by glucocerebrosidase.

Figure 16-7. An electron micrograph of the typical lysosomal inclusions of a splenicGaucher cell.

the lysosomal membrane. Thus, the in vitrodemonstration of activity after such a proce-dure requires inclusion of some phospholipidor bile salt in the assay medium.The enzyme’spH optimum is in the range of 5.0–5.8, andthe specific activity of the enzyme depends onthe nature of the lipid activator used in theassay.

Glucocerebrosidase is particularly respon-sive to acidic lipids and an 11,000 molecularweight heat-stable glycoprotein, referred to his-torically as heat-stable factor (HSF), and morerecently as sphingolipid activator protein (SAP).The acidic membrane lipids that are most effec-tive in reconstituting glucocerebrosidase arephospholipids such as phosphatidylserine andphosphatidylinositol and gangliosides such asGM1. These acidic lipids activate glucocere-brosidase by increasing the Vmax and decreas-ing the Km, and the activation, at least in diluteaqueous buffers, occurs with conversion of theenzyme from a low to a high molecular weightform. Regarding substrate specificity, glucocere-brosidase has a rather strict specificity for β-D-glucosides; it will cleave the glucose moietyfrom glucocerebroside as well as from non-physiological substrates like MUGlc and p-nitrophenyl-β-D-glucopyranoside but has littleor no activity toward β-D-galactosides or β-D-xylosides.

In terms of molecular genetics, the humanglucocerebrosidase gene is located within agene-rich region within band q21 of the longarm of chromosome 1. It consists of 7,620 basepairs that reside within 11 exons separated by10 introns (Fig. 16-8). A consensus sequencerepresenting an inducible promoter occursimmediately upstream of two independentstart codons. The 5.6-kb primary transcript isprocessed and translated into a polypeptidecontaining 497 amino acids which has an

approximate molecular mass of 67,000 d.Withinthe primary structure of the enzyme, thereare five consensus sequences for N-linkedoligosaccharide chain glycosylation sites (Asn-X-Ser or Asn-X-Thr), four of which are occupiedby complex-type oligosaccharide chains con-taining mannose, fucose, galactose, N-acetylglu-cosamine, and sialic acid (Fig. 16-9). The activesite of glucocerebrosidase is localized to the car-boxyl terminal region of the enzyme and is en-coded by exons IX and X.

Notably, there exists a glucocerebrosidasepseudogene located 16 kb downstream fromthe functional gene (a pseudogene is a regionof DNA that has sequence similarity to a knowngene but is not functional). It shares a high ex-onic sequence homology (96%) to the activegene but does contain many sizable intronicand exonic deletions as well as several non-sense and missense point mutations. Thepseudogene is actively transcribed from a weakpromoter but is not translated or processed fur-ther. Several types of mutations have arisenfrom the homologous recombination betweenthis pseudogene and the glucocerebrosidasegene.

To date, over 250 different mutations havebeen found within the glucocerebrosidasegene, comprising both private (unique) andpublic (common) mutations. Five mutations(N370S, L444P, R436C, 84insG, and IVS2+1G→A) are most commonly detected but havevarying frequencies among differing popula-tions. The first three mutations representchanges that have occurred in the coding re-gion of the gene that cause a substitution ofone amino acid for another. The 84insG muta-tion denotes a frameshift/nonsense mutation, asingle base insertion of guanine at the 84thbase pair within the coding region that createsa premature translational termination codon.


Figure 16-8. The glucocerebrosidase gene and sites of the most common mutations.The glucocerebrosidase pseudogene is located 16 kbp downstream. Over 250 differentmutations have been identified. The most common mutations shown here are describedin the text.

The IVS2+1G→A mutation is a G-to-A changewithin the splice-donor site of intron 2, result-ing in the deletion of the second exon from themRNA.

The most commonly occurring mutationwithin most populations is N370S, whereas theothers are found with variable frequenciesamong the different populations. For example, arecent study found that the 84insG mutationoccurred only in Ashkenazi Jewish patientswith Gaucher disease, whereas these same pa-tients did not have any of the pseudogene-recombinant alleles that were found inmembers of the general population.

Gaucher disease is autosomal recessive, re-sulting from the inheritance of two copies of amutated gene. Homoallelic disease is the re-sult of inheriting precisely the same mutationfrom both parents. Heteroallelic, or complex,defects arise when an alteration in one allelecoexists with a different mutation (or muta-tions) in the corresponding allele. Thus, nu-merous combinations of two of the 250+mutations could possibly exist within a partic-ular individual.Although the precise mutationscan be identified within Gaucher patients, thisknowledge cannot, with confidence, be usedto predict the severity of the disease within aparticular patient. However, some associationsdo exist; for example, the presence of at least asingle copy of the N370S allele within a patientis largely associated with the nonneuropathic(type I) variant of the disease. Although somegeneralizations like this can be made, there isgreat clinical heterogeneity between patients

with Gaucher disease, even when they haveidentical mutations.

THERAPY

Successful implementation of enzyme replace-ment therapy was initiated in 1990, and hasgreatly improved the quality of life for manypersons suffering from Gaucher disease. Beforethat time, the therapy, like that of the othersphingolipidoses, remained largely supportive,aimed at lessening the impact of the variousmanifestations of the disease process. As thesplenomegaly associated with Gaucher diseaseoften results in anemia and thrombocytopenia,the latter placing the individual at risk forbleeding diathesis, splenectomy was eventu-ally performed on almost all patients.Althoughsome physicians contend that postponingsplenectomy as long as possible reduces theage of onset and rate of development of boneand liver involvement, there is little evidence tosupport this position. Surgical intervention—for example, the implantation of a hip prosthe-sis in the woman discussed in this chapter—usually results in decreased bone pain and in-creased mobility. Two years after her left hipwas reconstructed, the patient’s artificial hipwas functioning well, and she was free of painat the site.

In 1966, just a year or two after the enzy-matic basis of Gaucher disease was identified,Brady and coworkers proposed a novel pro-tocol for treating the disease, namely, the


Figure 16-9. Schematic representation of the complex N-linked oligosaccharidescontained in glucocerebrosidase.

intravenous infusion of exogenous glucocere-brosidase (as discussed in Brady and Barranger,1983). Eight years later, the earliest attempts atenzyme supplementation were undertaken us-ing native glucocerebrosidase purified from hu-man placenta. Despite the modest reductions inplasma and hepatic glucocerebroside levelsthat were achieved with a single infusion of glu-cocerebrosidase, the outcome of the initialtreatment was disappointing because the in-fused enzyme did not effectively penetrate intothe affected cells (i.e., macrophages) where theoffending glycolipid is stored. The principal ob-stacle was the rapid clearance of the infused en-zyme from the circulation via selective uptakeby hepatocytes. Because glucocerebroside ac-cumulates in the reticuloendothelial system ofpatients with Gaucher disease, a strategy wasneeded to target the placental enzyme specifi-cally to macrophages.At about the same time,lectinlike receptors on the plasma membraneof macrophages were identified that selec-tively bind glycoproteins with oligosaccharidechains that terminate in α-linked mannoseresidues. It was subsequently shown thatmacrophages use a receptor-mediated endocy-totic mechanism to internalize glycoproteinsthat contain exposed mannose residues at thenonreducing ends of their oligosaccharidechains. Hepatocytes, in contrast, express sur-face receptors that bind galactose-terminatedglycoproteins.

Native placental glucocerebrosidase containsfour oligosaccharide chains, all of which termi-nate with sialic acid (Fig. 16-9). To optimizeglucocerebrosidase uptake by macrophages,the enzyme must be treated sequentiallywith neuraminidase, β-galactosidase, and β-N-acetylglucosaminidase, a process that results inthe exposure of mannose residues.

In the early 1990s,macrophage-targeted gluco-cerebrosidase supplementation was proven tobe a highly effective and safe therapy for mostcases of Gaucher disease. Several accounts de-scribe a gradual resolution of the clinical fea-tures of the disease, an overall decrease in lipidaccumulation in organs and the circulation, anda dramatic improvement in the quality of life ofthe patients. No adverse effects following treat-ment were reported. As seen in Table 16-1, thepatient in the present report experienced an im-provement in hematological parameters and asignificant reduction in liver volume after en-zyme replacement therapy.

To treat more severe forms of the disease, amore aggressive regimen is advised. In mostcases, therapy results in a significant decreasein splenic volume and in the level of circulatingglucocerebroside. Generalized improvement inhematologic variables follows clearance ofGaucher cell infiltrates from the bone marrow.In a few subjects, there is slow and partial re-covery from skeletal complications. Patientsgain weight, grow taller, and experience in-creased vigor, satiety, decreased bone pain, andrelief from chronic fatigue. Normalization ofhematological parameters, including elevationsin hemoglobin concentration and plateletcount, an increase in the survival time of autolo-gous labeled RBC, and decreased percentage ofreticulocytes, occurs after 3–4 months of en-zyme supplementation. Clearance of Gaucherinfiltrates from the bone marrow and as muchas a 60% reduction of splenic volume is thoughtto underlie the basis of such changes. Analysisof bone marrow aspirates of patients receivingenzyme supplementation therapy reveals quali-tative changes in Gaucher cells, specifically a re-duction in size and a decrease in lipid content.Amelioration of hematological variables is dosedependent; discontinuing therapy causes thehematological indices to revert to pretreatmentvalues.

The cost of treating a single patient withglucocerebrosidase is several hundred thou-sand dollars per year, raising several concernsregarding compensation since insurance plansmay refuse to authorize coverage. Such prohibi-tive cost often places a severe financial burdenon patients in this country and makes enzymereplacement therapy in less-developed nationsan impossibility.

One year of therapy for a patient infusedwith Ceredase, the placental-derived glucocere-brosidase drug, requires more than 22,000placentas for production. To meet demandfor therapy, Genzyme, the manufacturer ofCeredase, built a processing plant to extractglucocerebrosidase from donated placentas,and at one point, 35% of all unwanted placen-tas from the United States were passingthrough this French plant. Even with this large-scale processing, it was evident that not allpatients with Gaucher disease would be ableto receive enzymatic therapy. Fortunately,due to advancements in molecular biology,researchers were able to clone the glucocere-brosidase coding region and produce a recom-


binant form of the enzyme. This recombinantform could be altered for macrophage uptake,just as the placental-derived enzyme, and itdemonstrated equivalent clinical efficacy.This recombinant form was approved by theFood and Drug Administration in 1995 andis produced by Genzyme under the nameCerezyme.

A formidable challenge remains in targetingglucocerebrosidase to the involved cells in theCNS of those afflicted with type 2 disease. Intra-ventricular infusion or temporary alteration ofthe blood-brain barrier have been suggested; atthis time, no effective delivery strategy has yetbeen formulated for the neuronopathicGaucher phenotype.

QUESTIONS

1. What are some of the ethical questionsthat would arise if (a) one attempted toimplement a screening program to detectcarriers of Gaucher disease that employeda biochemical test with a diagnosticspecificity of 0.95? (b) a woman was en-couraged to undergo amniocentesis to de-termine if the fetus was affected?

2. How would you go about determining thenature of the phospholipid in the lyso-somal membrane that is the true in vivoactivator of glucocerebrosidase?

3. What is the cause of elevation in serumacid phosphatase (ACP) activity in Gaucherdisease?

4. What is the major source of the accumu-lated glucocerebroside in Gaucher dis-ease? In what form is the accumulatedlipid stored in the lysosome?

5. How would you assess whether gluco-cerebrosidase replacement therapy waseffective in a teenager with Gaucher dis-ease?

6. What consequences of immunoglobulinproduction against glucocerebrosidasemight occur in a patient receiving placen-tal enzyme replacement therapy?

7. What is the cause of the pancytopenia as-sociated with Gaucher disease?

8. Assume you have discovered a new muta-tion at base position 276 in the glucocere-brosidase gene. How would you use thisinformation to design an assay based onpolymerase chain reaction for the pres-

ence of this mutation in the parents andsiblings of the index case?

BIBLIOGRAPHY

Baldellou A, Andria G, Campbell PE, et al.: Paedi-atric non-neuronopathic Gaucher disease: rec-ommendations for treatment and monitoring.Eur J Pediatr 163:67–75, 2004.

Barton NW, Brady RO, Dambrosia JM, et al.:Replacement therapy for inherited enzymedeficiency—macrophage-targeted glucocere-brosidase for Gaucher’s disease. N Engl J Med324:1464–1470, 1991.

Basu A, Glew RH: Characterization of the phospho-lipid requirement of a rat liver β-glucosidase.Biochem J 224:515–524, 1984.

Basu A, Glew RH, Daniels LB, et al.: Activators ofspleen glucocerebroside from controls and pa-tients with various forms of Gaucher’s disease. JBiol Chem 259:1714–1719, 1984.

Beutler E.: Gaucher disease as a paradigm of cur-rent issues regarding single gene mutations ofhumans. Proc Natl Acad Sci U.S.A. 90:5384–5390, 1993.

Boot RG,Verhoek M, de Fost M, et al.: Marked ele-vation of the chemokine CCL18/PARC inGaucher disease: a novel surrogate marker forassessing therapeutic intervention. Blood103:33–39, 2004.

Brady RO, Barranger JA: Glucosylceramide lipido-sis: Gaucher’s disease, in Stanbury JB, Wyngaar-den JB, Frederickson DS, et al. (eds): TheMetabolic Basis of Inherited Disease. McGraw-Hill, New York, 1983, pp. 842–56

Brady RO, Barton NW, Grabowski GA:The role ofneurogenetics in Gaucher disease. Arch Neurol50:1212–1242, 1993.

Brady RO, Furbish FS: Enzyme replacement ther-apy: specific targeting of exogenous enzymes tostorage cells, in Martonosi AN (ed): Membranesand Transport.Vol. 2. Plenum Press, New York,1982, pp. 587–92

Brady RO, Murray GJ, Oliver KL, et al.: Managementof neutralizing antibody to Ceredase in a patientwith type 3 Gaucher disease. Pediatrics 100:E11, 1997.

Cox TM.: Future perspectives for glycolipid re-search in medicine. Philos Trans R Soc Lond BBiol Sci 358:967–973, 2003.

Daniels LB, Glew RH: β-Glucosidase assays in thediagnosis of Gaucher’s disease. Clin Chem28:569–577, 1982.

Elstein D,Abrahamov A,Hadas-Halpern I,et al.:Rec-ommendations for diagnosis, evaluation, andmonitoring of patients with Gaucher disease.Arch Intern Med 159:1254–1255, 1999.


Furbish FS, Steer CJ, et al.: Uptake and distribu-tion of placental glucocerebrosidase in rathepatic cells and effects of sequential deglyco-sylation. Biochim Biophys Acta 673:425–434,1981.

Gauchers Association Web site.Available at: http://www.gaucher.org.uk. Accessed 03/12/06.

Genezyme Web site. Available at: http://www.genzyme.com. Accessed 03/12/06.

Glew RH, Basu A, Prence EM, et al.: Biology ofdisease—lysosomal storage diseases. Lab Invest53:250–269, 1985.

Glew RH, Daniels LB, Clark LS, et al.: Enzymic dif-ferentiation of neurologic and nonneurologicforms of Gaucher’s disease. J Neuropathol ExpNeurol 41:630–641, 1982.

Grabowski GA: Gaucher disease: gene frequenciesand genotype/phenotype correlations. GenetTest 1:5–12, 1997.

Grabowski GA, Andria G, Baldellou A, et al.:Pediatric non-neuronopathic Gaucher disease:presentation, diagnosis and assessment. Con-sensus statements. Eur J Pediatr 163:58–66,2004.

Grabowski GA, Barton NW, et al.: Enzyme therapyin type 1 Gaucher disease: comparative efficacyof mannose-terminated glucocerebrosidasefrom natural and recombinant sources. AnnIntern Med 122:33–39, 1995.

Hill SC, Parker CC, et al.: MRI of multipleplatyspondyly in Gaucher disease. Arch Neurol50:1212–1224, 1993.

National Gaucher Foundation Web site. Availableat: http://www.gaucherdisease.org. Accessed03/12/06.

Parker RI, Barton NW, et al.: Hematologic improve-ment in a patient with Gaucher disease on longterm-enzyme replacement therapy: evidence fordecreased splenic sequestration and improvedred blood cell survival. Am J Hematol 38:130–137.

Sidransky E, Bottler A, Stubblefield B, et al.: DNAmutational analysis of type 1 and type 3 Gaucherpatients: how well do mutations predict pheno-type? Hum Mutat 3:25–28, 1994.

Zhao H, Grabowski GA: Gaucher disease: perspec-tives on a prototype lysosomal disease. Cell MolLife Sci 59:694–707, 2002.


http://www.gaucher.org.uk

http://www.gaucher.org.uk

http://www.genzyme.com

http://www.genzyme.com

http://www.gaucherdisease.org

CASE HISTORY

The patient (Fig. 17-1) was a 38-week-gestationproduct of a 22-year-old G,P1Abo (gravida 1,first pregnancy; paradelivery, first delivery;abortion, 0) woman and her 31-year-old first-cousin partner. Pregnancy was complicated byoligohydramnios (deficiency in the amount ofamniotic fluid); an amniocentesis revealed anormal male karyotype. The infant was deliv-ered by cesarean section because of maternalpreeclampsia (development of hypertensiondue to pregnancy). Apgar scores were 8/9,weight was 2200 g, and length was 18.5 inches.On newborn examination, microcephaly (abroad nasal bridge with anteverted nostrils),micrognathia (smallness of the jaws, especiallythe underjaw), hypospadias (a defect in thewall of the urethra), and undescended testeswere seen.A tentative diagnosis of Smith-Lemli-Opitz syndrome was postulated.

By 9 months of age, marked developmentaldelay was noted. Repeat evaluation revealed bi-lateral inguinal hernias, dysmorphic featureswith synophrys (growing together of the eye-brows), and thick upper lip with gum hypertro-phy; limitation of motion at the knees, hips,elbows, and wrists; camptodactyly (bendingof fingers); rib flaring; hepatosplenomegaly;corneal clouding; and hypotonia in addition tothose findings reported at birth.A lysosomal en-zyme screen indicated elevated levels of a num-ber of lysosomal enzymes in plasma and urineand sialyloligosaccharides in the urine. Thesefindings and the physical examination are con-sistent with a diagnosis of I-cell disease.

DIAGNOSIS

Postnatal

I-cell disease (mucolipidosis II, ML-II) is a rareautosomal recessive lysosomal storage disorderfirst described in 1966 by DeMars and Leroywhile studying a group of suspected Hurler pa-tient biopsies. The histopathologic hallmark ofI-cell disease is the presence of characteristicphase-dense, cytoplasmic, spherical (approxi-mately 0.5–1.0 µm in diameter) vesicles (i.e., in-clusions) (Fig. 17-2) surrounding the nucleusand juxtanuclear Golgi apparatus within themesenchyme-derived tissues throughout thebody. These cells were so distinctive com-pared with cells cultured from other patientswith similar clinical symptoms that the appella-tion I cell was coined. Ultrastructurally, the in-clusions are bound by a single membrane andcontain fibrogranular and membranous lamellarmaterial. Subsequently, I-cell disease was classi-fied as a mucolipidosis (designated ML-II bySpranger and Wiedemann in 1970),characterizedby intracellular (i.e., lysosomal) accumulation ofacid mucopolysaccharides, sphingolipids, or gly-colipids in visceral and mesenchymal cells.

I-cell disease is suspected clinically by thephenotype and is confirmed biochemically (cf.“Biochemical Perspectives” section). The infantwith I-cell disease is usually small for gestationalage and is clinically differentiated from havingHurler’s syndrome by earlier onset of signsand symptoms, the absence of excessive mu-copolysacchariduria, short stature, and the rap-idly progressive course, leading to death usuallyby age 4 years.

17

I-Cell Disease (Mucolipidosis II)

JAMES CHAMBERS

181

The disease progresses with the child develop-ing coarse facial features with puffy eyelids,prominent epicanthal folds, flat nasal bridge, an-teverted nostrils, macroglossia (enlargement ofthe tongue), and severe skeletal abnormalities.Craniofacial abnormalities (i.e., microencephalyand metopic craniosynostosis), restricted jointmovement with generalized hypotonia, congeni-tal hip dislocation, hernias, and bilateral talipesequinovarus (clubfoot) may be observed in theneonatal period. Gingival hypertrophy is morestriking than in Hurler syndrome. Hard, firm sub-cutaneous nodules may represent accumulationof storage material. In addition,severe gingival hy-perplasia is also observed in some presentations.

By 6 months of age, psychomotor retar-dation is usually obvious. Joint immobilityprogresses with development of claw-handdeformities and kyphoscoliosis. Hepatomegalyis prominent, but splenomegaly is minimal.Corneal haziness may be present but is subtle,and corneal opacities due to the accumulationof storage material are not as striking as inHurler syndrome. Examination of peripheralblood smears reveals the presence of abnormalinclusions in cells such as lymphocytes, and in-creased lysosomal enzyme activity in wholeblood as well as cultured fibroblasts is confir-matory of I-cell disease.

Radiographically, I-cell disease is mani-fested in two successive stages of infancy, one

observed in the neonatal period and the otherat about 1 year of age.When detected at birth,peculiar radiographic changes of the bones re-flect a more severe disturbance in bone devel-opment and growth, and strongly resemblerickets or osteomalacia (softening and bendingof the bones) as evidenced by the presence ofrachiticlike (related to rickets) lesions andchanges similar to those observed in hyper-parathyroidism. The radiographic abnormali-ties found in older children appear to be morenonspecific and are typical of Hurler-like signsof dysostosis multiplex (roentgenographic ab-normalities indicating bone changes at multi-ple sites).

Biochemically, I-cell disease is characterizedby excessive secretion of newly synthesizedlysosomal enzymes into body fluids and con-comitant loss of respective intracellular activi-ties in fibroblasts. Shown in Table 17-1 arerepresentative lysosomal enzyme activity levelsin serum from patients with I-cell disease andthose with the closely related disorder pseudo-Hurler polydystrophy, indicating significantly in-creased levels of lysosomal enzyme activity.Germane to the biochemical diagnosis is thecharacteristic pattern of lysosomal enzyme defi-ciency in cultured fibroblasts, that is,an increasein the ratio of extracellular to intracellular en-zyme activity (Table 17-2). It is interesting tonote that not all lysosomal (i.e. intracellular)


Figure 17-1. A 9-month-old Hispanic male infant with I-cell disease. From the Divisionof Medical Genetics, Children’s Hospital, Los Angeles.

enzyme activities are decreased in I-cell fibrob-lasts, suggesting alternative lysosomal routing;acid phosphatase and β-glucosidase are the ex-ceptions. Major increases in glycosylaspara-ginase activity in I-cell patient serum and urineand the markedly reduced activity in cultured

fibroblasts can also be used as a diagnostic aidfor detecting I-cell disease (Ylikangas andMononen, 1998). It is important to note thatincreased levels of lysosome enzymes in serumcannot distinguish between ML-II and ML-III(pseudo-Hurler polydystrophy), which usually

I-Cell Disease (Mucolipidosis II) 183

Figure 17-2. Microscopic examination of cultured I-cell skin fibroblasts. (A) Phase con-trast microscopy. (B) Electron microscopy showing the inclusions. Photographs cour-tesy of Dr. Robert DeMars.

must be differentiated on the basis of clinicalevents.

Degradation of most lysosomal substratesoccurs by the stepwise activity of a series ofhydrolases, with each step requiring the actionof the previous hydrolase to modify the sub-strate, so that the substrate can be further de-graded by the next enzyme in the pathway.If one step in the process fails, then furtherdegradation ceases, and the partially degradedsubstrate accumulates. Thus, lysosomal accumu-lation of partially degraded substrate in manycases give rise to a metabolic “traffic jam”affect-ing the architecture and function of cells, tis-sues, and organs.

Analysis of I-cell patient glycosphingolipids re-veals elevations in trihexosylceramide (GL-3)and the ganglioside GM3 (Fig. 17-3) but notthe massive accumulation associated with theglycosphingolipidoses. In the spleen, the onlyapparent lipid abnormality is an increase in thelevel of GM3. The brain has normal levels of gan-gliosides, apart from a small increase in GM3 and


Table 17-1. Lysosomal Enzyme Activities inSerum from Patients with I-Cell Disease andPseudo-Hurler Polydystrophy

Pseudo-Enzyme Controls I-Cell Hurler

β-Galactosidase* 16.3 30.6 ND

α-Galactosidase* 4.6 14.5 ND

α-Mannosidase* 31.4 1116 3976

β-Glucuronidase* 14.2 163 740

α-Fucosidase* 235 5184 4389

Arylsulphatase A* 14.2 1664 ND

α-Glucosaminidase* 15.8 1531 76

β-Glucosaminidase* 621 5040 4194

Glycosylasparaginase† 13.4 142 ND

β-Galactosidase† 58.9 926.5 ND

ND, nondetermined.

*Data from Poenaru et al. (1988). Specific activity is expressed asnmol/h/ml serum.†Data from Ylikangas and Mononen (1998). Specific activity isexpressed as mU/L.

Table 17-2. Intra- and Extracellular Levels of Various Lysosomal Enzymes of Cultivated I-Cell Disease FibroblastsEnzyme Control I-Cell Disease

Intracellular activityα-Fucosidase* 72 2.2β-Galactosidase* 584 6.7α-Mannosidase* 98 9.3Neuraminidase* 39 0.5Acid Phosphatase* 2,954 2,620β-Glucosidase* 162 148Phospholipase† 2.8 0.45Palmitoyl-protein thioesterase‡ Detected, Undetected,

immunoblot immunoblotGlycosylasparaginase§ 70 6.0β-Galactosidase§ 21,800 890

Extracellular activityα-Fucosidase* 0.8 2.7β-Glucuronidase* 41 345Hexosaminidase A* 97 878α-Mannosidase* 2.3 24β-Glucosidase* 0.2 0.2Phospholipase† 0.0 0.58Palmitoyl-protein thioesterase‡ Undetected, Detected,

immunoblot immunoblot

*Data from Ben-Yoseph et al. (1986). Specific activity of intracellular enzymes is expressed as nmol/h/mg protein.Activities detected in the medium (extracellular) were those secreted by 1 mg cell protein in 24 hours.†Data from Jansen et al. (1999). Intracellular and extracellular specific activities are expressed as nmol/h/mg fibroblastprotein and nmol/h/ml medium, respectively.

‡Data from Verkruyse et al. (1997). Intracellular palmitoyl-protein thioesterase (PPT) was determined using whole celllysates and anti-PPT antibody. PPT protein in the medium (extracellular) was detected following partial purification toenrich for PPT prior to immunoblotting with anti-PPT antibody.§Data from Ylikangas and Monenen (1998). Specific activity is expressed as pmol/min/mg protein.

lower than normal concentrations of sulfatideGL-1bS and cerebroside GL-lb (Fig.17-3).

Due to the extensive mesenchymal tissue in-volvement in I-cell disease, the observation oflarge amounts of unsulfated chondroitin in thecartilage matrix and compensatory decrease ofchondroitin 4-sulfate has been viewed as sug-gesting possible abnormalities in the posttransla-tional assembly and subsequent modification ofglycosaminoglycans. However, I-cell fibroblastshave been shown to (1) internalize cell-surfaceproteoglycans; (2) remove glycosaminoglycanchains from the proteoglycan core protein;and (3) partially degrade heparan sulfate glycos-aminoglycan chains in identical fashion as thatobserved for normal fibroblasts (Brauker andWang, 1987).Also suspect is the degradation ofcollagen (type I) in I-cell disease fibroblasts as aresult of the greatly reduced levels of cathepsinB, which possesses potent collagenolytic activ-ity; however, the rate of intracellular degrada-tion of proline analogues by I-cell diseasefibroblasts is comparable to rates of degradationobserved for normal fibroblasts.

Many tissues from patients with I-cell diseaseexhibit normal levels of “intracellular” lysosomalenzymatic activities (e.g., brain, liver, kidney,and spleen). In the liver, lysosomal enzyme lev-els are normal except for diminished β-galactosidase and elevated β-hexosaminidase,β-xylosidase,and α-galactosidase. This fairly spe-

cific decrease in liver activity of β-galactosidasecontrasts with the diminution of almost all lyso-somal enzymes in cultured fibroblasts.

Prenatal

Prenatal diagnosis of I-cell disease has beenbased on greatly reduced phosphotransferaseactivity (cf.“Biochemical Perspectives” section)and abnormal intracellular-extracellular distri-bution of lysosomal enzymes in cultured amni-otic fluid cells (Table 17-3).As indicated in Table17-3, amniotic fluid cells secrete large amountsof lysosomal enzymes into the extracellularmedium. Decreased levels of lysosomal en-zymes in chorionic villi obtained by biopsyhave also been observed in I-cell disease; how-ever, the characteristic secondary effect (i.e., in-creased levels of lysosomal enzymes in theextracellular compartment) is only partially ex-pressed or not expressed at all in chorionicvilli, suggesting an alternative mechanism forthe transport of lysosomal proteins. Although

I-Cell Disease (Mucolipidosis II) 185

Figure 17-3. Important glycolipids affected inI-cell disease. The dotted lines represent the re-spective β-linked glycosides. Ceramide, 2-N-acylsphingosine; NANA, N-acetylneuraminic acid.GL-1bs is shown sulfated in the 3 position of galac-tose (SO3H → 3).

Table 17-3. Prenatal Diagnosis: Levels of GlcNac-PO4 Transferase and Various Lysosomal Enzymesin Amniotic Fluid and Cultivated Amniotic FluidCellsEnzyme Control I-Cell Disease

Amniotic fluid cellsGlcNac-PO4 Transferase 0.68 0.05

Intracellular activityα-Fucosidase 478 22.3β-Galactosidase 745 27.3β-Glucosidase 136 278β-Glucuronidase 244 17.4Hexosaminidase A 1345 119α-Mannosidase 190 14.7β-Mannosidase 462 31.1

Extracellular activityα-Fucosidase 53.4 285β-Galactosidase 14.6 69.0Hexosaminidase A 242 1843α-Mannosidase 5.6 44.4Amniotic fluidHexosaminidase 155 1620β-Glucuronidase 15.6 135α-Fucosidase 18.0 140α-Mannosidase 7.2 76α-Galactosidase 0.78 2.4

Data from Parvathy et al. (1989) and Besley et al. (1990).Intracellular activities are expressed as nmol substrate cleaved/h/mg cell protein. Extracellular activities are expressed as theactivity secreted into the medium/24 h/mg cell protein. Activitiesin amniotic fluid are expressed as nmol/h/mg protein.

the diagnosis is made more reliably using cellsobtained by amniocentesis, direct examinationof the phosphotransferase activity of chorionicvilli has been effectively used to diagnose I-celldisease prenatally. This approach affords a morerapid diagnosis than the 2 to 4 weeks requiredfor diagnosis by amniocentesis. In addition, elec-tron microscopic examination of chorionic vil-lus tissue has been used for rapid prenataldiagnosis of I-cell disease (Carey et al., 1999).Although not the focus of this discussion, pre-natal diagnosis of the closely related disorderexhibiting the same primary biochemical de-fect as I-cell disease (i.e., pseudo-Hurler polydy-strophy) was accomplished using moleculartools targeting the γ-subunit of the phospho-transferase enzyme (Falik-Zaccae et al., 2003).


Initial Observations

Historically, the important observation made byHickman and Neufeld (1972) that I-cell fibrob-lasts are capable of endocytosing (the processby which cells take up macromolecules) lysoso-mal enzymes secreted by normal cells but thatnormal cells are incapable of internalizing theenzymes secreted by I-cell fibroblasts suggestedthat some kind of recognition marker for inter-nalization of lysosomal enzymes was absent inI-cell fibroblasts. This recognition marker hassubsequently been identified as mannose 6-phosphate.

It is widely believed that I-cell disease (ML-II)and pseudo-Hurler polydystrophy (ML-III) arevariants of the same disorder because fibrob-lasts from both these disorders are deficient inphosphotransferase activity. At the molecularlevel, the primary and most significant differ-ence is that in I-cell disease the deficiency isessentially total, whereas in pseudo-Hurlerpolydystrophy there appears to be significantresidual activity (approximately 10% of controlvalues). The molecular basis for these two dis-eases may, however, be distinct, as evidenced bygenetic complementation studies demonstrat-ing complementation of ML-II by some ML-III fi-broblasts (Mueller et al., 1983). Although twoclinical presentations of I-cell disease have beendescribed (i.e., the neonate and 6- to 12-month-old patient), correlation of the degree of defi-ciency of phosphotransferase activity withclinical severity in the two I-cell disease presen-tations remains controversial.

Molecular Theme

The Marking Event

The I-cell patient has proved invaluable for elu-cidation of the complex nature of intracellularpackaging and sorting of lysosomal enzymes.The physiological importance of this signal-mediated pathway is evident in that fibroblastsfrom patients with I-cell disease and pseudo-Hurler polydystrophy secrete rather than targetmost of their lysosomal enzymes. Thus, the mo-lecular theme of I-cell disease is that of “faulty”lysosomal targeting, the inability to transport(i.e., sort) lysosomal enzymes from their site ofsynthesis to the lysosome.

The biosynthetic pathways responsible forboth membrane and secretory glycoproteinshave been localized to specific lumenal re-gions of the Golgi and endoplasmic reticulum(for an excellent review see Kornfeld and Ko-rnfeld, 1985). As shown in Figure 17-4, the“core”oligosaccharide (three glucose, nine man-nose,and two N-acetylglucosamine residues) ispreassembled in an “activated” form and istransferred en bloc from a lipid-linked (dolichyl-pyrophosphate) carrier to distinct asparagineresidues of the nascent polypeptide while theprotein is being synthesized on membrane-bound polysomes. This transfer usually takesplace before the folding of the polypeptidechain. Once glycosylation of the protein has oc-curred, processing of the oligosaccharide is ini-tiated by trimming reactions that remove threeglucose residues and one mannose unit. Theproteins then move by vesicular transport tothe Golgi, where they undergo a variety of addi-tional posttranslational modifications and aresorted for targeting to the proper destination,that is, to the lysosome, secretory granules, orplasma membrane.

The trafficking of lysosomal hydrolases tothe lysosome in higher eukaryotes dependson the specific modification of asparagine-linked oligosaccharides to contain a mannose6-phosphate recognition marker. This man-6-P-dependent sorting system requires the con-certed action of two key enzymes (Fig. 17-5).The initial and determining step in the gen-eration of the mannose 6-phosphate recogni-tion marker is catalyzed by the enzymeUDP-N-acetylglucosamine:lysosomal-enzyme N-acetylglucosaminyl-1-phosphotransferase(EC 2.7.8.17), commonly referred to as phos-photransferase.

The primary biochemical defect in I-cell dis-


ease, as well as in pseudo-Hurler polydystro-phy, is in the phosphotransferase enzyme. Thismetabolic error creates a secondary deficiencyof most lysosomal enzymes, as described in thissection. The reaction catalyzed by the phos-photransferase enzyme (cf. reaction 1, Fig.17-5)results in formation of the phosphodiester N-

I-Cell Disease (Mucolipidosis-II) 187

Figure 17-4. Schematic pathway of lysosomal enzyme targeting to lysosomes. Lyso-somal enzymes and secretory proteins are synthesized in the rough ER (RER) and glyco-sylated by the transfer of a preformed oligosaccharide from dolichol-P-P-oligosaccharide(Dol). In the RER, the signal peptides ( ) are excised. The proteins are translocatedto the Golgi, where the oligosaccharides of secretory proteins are processed tocomplex-type units, and the oligosaccharides of lysosomal enzymes are phosphorylated.Most of the lysosomal enzymes bind to mannose 6-phosphate receptors (MPRs) ( )and are translocated to an acidified prelysosomal compartment, where the ligand disso-ciates. The receptors recycle back to the Golgi or to the cell surface, and the enzymesare packaged into lysosomes, where cleavage of their propieces is completed ( ).The Pi may also be cleaved from the mannose residues. A small number of the lysosomalenzymes fail to bind to the receptors and are secreted along with secretory proteins.These enzymes may bind to surface MPRs in coated pits ( ) and be internalizedinto the prelysosomal compartment. ( ), N-Acetylglucosamine; ( ), mannose; ( ) glu-cose; ( ), galactose; ( ) sialic acid. Reprinted with permission from Kornfeld(1987). Copyright 1987, FASEB J.

acetylglucosamine-l-phosphate-6-mannose inter-mediate. A second “uncovering” enzyme (i.e.,N-acetylglucosamine-l-phosphodiester-α-N-acetyl-glucosaminidase or EC 3.1.4.45) cleaves the ter-minal α-N-acetylglucosamine residue to uncoverthe mannose 6-phosphate group that serves asthe required “recognition marker” component

responsible for high-affinity binding to mannose6-phosphate receptors (MPRs) in the Golgi.

The MPRs and their bound ligands (i.e., lyso-somal enzymes) are packaged into clathrin-coated vesicles that bud from the trans-Golginetwork (TGN) and subsequently fuse with anacidic endosomal compartment, where the lowpH promotes dissociation of receptor and lig-and. Thus, these receptors mediate delivery ofthe “targeted” enzymes to endosomal compart-ments, where they are discharged and subse-quently packaged into lysosomes (for reviews,see Kornfeld, 1986, 1992; von Figura and Hasi-lik, 1986). In I-cell disease, lysosomal enzymesare not modified by addition of mannose 6-phosphate.Thus, they are not segregated by theMPRs into the appropriate transport vesicles inthe TGN and instead are carried to the cell sur-face and secreted.

The Enzyme

Bao and coworkers (1996) have purified bovinephosphotransferase as a 54-kd complex com-posed of three different subunits (α2, β2, and γ2).Bovine phosphotransferase is the product of twogenes, one encoding the α- and β-subunits, thesecond coding the γ-subunit. The multisubunitnature of the enzyme gives much insight into therelationship between subunits, ML-II comple-mentation groups, and the mucolipidoses (ML-IIand ML-III). The gene coding for the human γ-subunit has been localized to chromosome 16p.Examination of human γ-subunit in ML-III pa-tients indicated a single nucleotide insertion dif-ference in the phosphotransferase subunit.

Although both I-cell disease and pseudo-Hurler polydystrophy patients are deficient inphosphotransferase activity, they exhibit differ-ent patterns of intracellular and extracellular

lysosomal enzyme activities in cultured skin fi-broblasts with respect to electrophoretic mobil-ity, lectin-binding properties, and responsivenessto sucrose loading. Complementation studies offibroblasts from I-cell disease and pseudo-Hurlerpolydystrophy indicate three genetic groups (A,B, and C) exhibiting altered phosphotransferaseactivity. In these studies,fibroblasts from individ-ual patients are fused with cells from other pa-tients. Variants have been categorized on thebasis of phosphotransferase activity toward α-methylmannoside as acceptor in the phospho-transferase assay as well as decreased secretionof lysosomal enzymes into the extracellularcompartment.

Complementation group A is the largest andincludes all I-cell disease and many pseudo-Hurler polydystrophy patients. The defects inpatients from group A are thought to be allelesat the same locus and are characterized by achange in the catalytic portion of the phospho-transferase that renders it unable to use the arti-ficial acceptor substrate (α-methymannoside).In patients of complementation group A, thephosphotransferase enzyme appears to besmaller than normal, suggesting the absence ofa catalytically important enzyme component.

Groups B and C consist entirely of the less-common pseudo-Hurler polydystrophy (ML-III)variants. Group B is comprised of patients withdefective phosphotransferase catalytic function.The enzyme from patients with complementa-tion group B phosphotransferase deficiencyappears to be larger than normal, suggestingabnormal aggregation of the enzyme. Group Cincludes patients exhibiting a phosphotrans-ferase with normal catalytic function toward α-methylmannoside but one that fails to recognizeendogenous lysosomal enzymes as acceptorsubstrate. The mutation in group C is thought


Figure 17-5. Synthesis of the mannose 6-phosphate recognition marker. R representsthe high-mannose oligosaccharide of newly synthesized lysosomal enzymes. Reaction 1is catalyzed by UDP-N-acetylglucosamine:lysosomal enzyme N-acetylglucosaminyl-1-phosphotransferase. Reaction 2 is catalyzed by N-acetylglucosamine-1-phosphodiester-N-acetylglucosaminidase.

to affect a gene coding a noncatalytic phospho-transferase site or a component that is involvedin the recognition of, or specific binding to,lysosomal enzyme precursors. Work by Raas-Rothschild and coworkers (2004) strongly sug-gests that, despite the existence of the threecomplementation groups, the γ-subunit of thephosphotransferase gene plays a major role inML-III.

The Phosphotransferase Assay

As shown in Figure 17-5, the phosphotrans-ferase transfers N-acetylglucosamine-l-phosphateresidues en bloc to the C-6 oxygen of particularmannose residues in “high-mannose” oligosac-charide units of lysosomal enzymes. This re-action occurs in two stages: transfer ofN-acetylglucosamine-l−phosphate to a mannoseresidue on the a-1,6 branch of a high-mannoseoligosaccharide and addition of a second N-acetylglucosamine-l−phosphate to the a-1,3branch.The two stages are thought to occur ina pre-Golgi compartment and the cis Golgi, re-spectively. A number of assay procedures havebeen described for the phosphotransferaseusing 32P-, 3H-, and 14C-labeled UDP-GlcNAc (ra-dioactivity is located in the GIcNAc moiety) asthe Ν-acetylglucosamine-1-phosphate donorand a variety of acceptor substrates (e.g., β-hexosaminidase, glycopeptides, and α-methyl-mannoside).

Generally, phosphotransferase is assayed us-ing α-methylmannoside as acceptor.After incu-bation, the reaction is terminated, and thephosphodiester product of the transfer reac-tion, (i.e., [14C or 3H]GlcNac-α-l-phospho-6-mannose-α-l-methyl) is chromatographicallyseparated from the reaction components, andthe radioactivity is determined.The general re-action scheme is

UDP-[3H/14C]GlcNAc + α-CH3-Man ↔ UMP+ α-CH3-Man-(6←P)-GlcNAc[3H/14C]

In vivo, the CH3 group in the reaction scheme isrepresentative of the rest of the high-mannoseoligosaccharide intermediate.

Phosphotransferase Protein Target Recognition

Efficient targeting of enzymes to lysosomes re-quires the phosphotransferase to recognize and

bind with high affinity to a protein determinantthat is common to lysosomal enzymes but absentfrom nonlysosomal glycoproteins. Because lyso-somal enzymes do not share linear amino acidsequences, the phosphotransferase must specifi-cally interact with a conformation-dependentprotein determinant that is expressed in 40 to50 different lysosomal hydrolases while display-ing low affinity for hundreds of other glycopro-teins that contain identical high-mannose-typeoligosaccharides.

The high-affinity interaction required for effi-cient phosphorylation in vivo and in vitro ap-pears to be mediated primarily throughprotein–protein structural interactions (Langet al., 1984). This is based on the fact that iso-lated high-mannose-type oligosaccharide andglycopeptides are extremely poor phospho-transferase acceptor substrates. Furthermore,the phosphorylation of intact lysosomal en-zymes is markedly inhibited by the inclusion ofdeglycosylated lysosomal enzymes in reactionmixtures, supporting the hypothesis that thephosphotransferase recognizes a protein domainthat is common to all lysosomal enzymes butabsent in nonlysosomal glycoproteins.

Much insight into this problem has beenderived from studies of cathepsin D, a bilobedlysosomal aspartylprotease that contains oneasparagine-linked oligosaccharide per lobe. Us-ing chimeric proteins containing either aminoor carboxyl lobe sequences of cathepsin Dsubstituted into a glycosylated form of the homo-logous secretory protein pepsinogen, the mini-mal elements of a recognition domain requiredfor phosphorylation involving two specific ly-sine residues have been identified that arethought to be shared among lysosomal enzymesand recognized by the phosphotransferase.

Mutation of these two residues results in70% inhibition of mannose phosphorylation(Sandholzer et al., 2000). A phosphotransferaserecognition domain located on either lobe ofthe cathepsin/glycopepsinogen chimeric mole-cule is sufficient to allow phosphorylation ofoligosaccharides on both lobes. Interestingly,the carboxyl lobe oligosaccharides of cathepsinD acquire two phosphates, whereas the aminolobe oligosaccharides acquire only one phos-phate. Molecular modeling of the cathepsin Dsequences inclusive of the lysosomal enzymerecognition domain indicates that these residuescome together in three-dimensional space toform a surface patch on the carboxyl lobe ofthe molecule.


Although much of the focus of I-cell diseasecenters around lysosomal enzyme carbohydrateprocessing, it is important to remember thatlysosomal enzymes are synthesized as pre-proenzymes with amino terminal extensionsthat require further proteolytic processing. Inaddition to removal of the signal peptide inthe ER, the newly synthesized protein may besubject to additional limited proteolysis duringtransport (Fig. 17-4). This proteolytic pro-cessing appears to be initiated in the prelyso-somal compartments and is completed after theenzymes arrive in the lysosomes. In some cases,there are further internal cleavages of the pep-tide as well as carboxyl-terminal processing.The biological significance of this processing isnot well understood but may play an early rolein the sorting process (e.g., maintenance of thecorrect folding and conformation of the nas-cent protein that subsequently becomes a sub-strate for the phosphotransferase).

Mannose 6-Phosphate Receptor: An Intracellular Lysosomal Enzyme Receptor

Newly synthesized lysosomal enzymes contain-ing mannose 6-P residues are recognized bytwo distinct mannose 6-phosphate receptors(MPRs) with molecular masses of 46,000 and300,000 d. Both mediate the transport of newlysynthesized soluble lysosomal enzymes fromthe trans-Golgi network (TGN) to a prelyso-somal compartment that is part of the endocyticroute. There the receptor-ligand complexes aredissociated, and hydrolases are subsequentlypackaged into lysosomes.The smaller receptorrequires divalent cations for its binding activityand is referred to as cation dependent (i.e., CD-MPR). In contrast, the larger MPR is cation inde-pendent (i.e., CI-MPR). Both MPR are type Itrans-membrane glycoproteins. The two recep-tors recycle regardless of whether they areloaded with ligands or not between the TGN,the endosomal compartment, and the cell sur-face. MPR are never detected in the lysosomes.Through increasingly better-understood path-ways (reviewed by Eskelinen et al., 2003), re-spective lysosomal substrates and the enzymesresponsible for their breakdown are brought to-gether in the endosomal compartment.

In humans and in many other species, thelarger receptor is identical to the receptor forthe insulinlike growth factor II (IGF-II) and thusis referred to as the mannose-6-P/insulin growth

factor II (man-6-P/IGF-II) receptor. The MPRsare distributed over several cellular compart-ments. A single pool of MPRs cycles constitu-tively between the Golgi, endosomes, and theplasma membrane. At the plasma membrane,the man-6-P/IGF-II receptor binds and mediatesendocytosis of extracellular ligands, whereasthe CD-MPR appears not to function in endocy-tosis under physiological conditions despiterapid internalization from the cell surface. Thedistribution of MPRs in fibroblasts from someI-cell patients is abnormal, with the receptorsfound almost exclusively in Golgi cisternae andin coated vesicles located near the cisternae.Thus, defective receptors, as well as functionalreceptors wrongly distributed, represent possi-bly yet additional mechanisms that can result inthe I-cell phenotype.

Whereas both types of MPRs bind mannose6-P with the same affinity (7–8 × 10−6 M), theman-6-P/IGF-II receptor binds diphosphory-lated oligosaccharides with a significantly higheraffinity than does the CD-MPR (2 × 10−9 M vs.2 × 10−7 M, respectively). Because oligosaccha-rides with two phosphomonoesters bind to theMPRs with an affinity similar to that observedfor lysosomal enzymes, the high-affinity bindingof lysosomal enzymes can be explained by atwo-site model in which two phosphomannosylresidues on the lysosomal enzyme interact withthe receptor. Individual phosphomannosylresidues located on different oligosaccharidesappear to interact with the receptor withhigher affinity than do two phosphomannosylresidues present on the same oligosaccharide.

How Does Sorting Occur?

The signal for the rapid internalization of MPR300 at the plasma membrane has been localizedto the pentapeptide KYSKV (residues 24–29 inits 163-residue cytoplasmic tail). MPR 46 ap-pears to contain in its cytoplasmic tail of 67residues several distinct signals for rapid inter-nalization and binding to clathrin adaptors.Both receptors have a leucine-based sorting sig-nal close to their C terminus, which is criticalfor their ability to sort lysosomal enzymes (Den-zer et al., 1997).

The question of why two MPR are present inmammalian cells remains unanswered. Analysisof cell lines made deficient in MPR by targeteddisruption of the respective MPR gene frommice as well as from tumors has shown that eachof the two MPRs contributes to the targeting of


newly synthesized lysosomal enzymes and thatthe complement of lysosomal enzymes trans-ported by either receptor is distinct but largelyoverlapping. However, neither type of MPR ap-pears to be sufficient for targeting of lysosomalenzymes to lysosomes. This has been demon-strated by expression of either MPR 46 or MPR300 in cells that lack endogenous MPRs (Kasperet al., 1996).

Does defective lysosomal catabolism in I-celldisease somehow feed back to affect the ex-pression of lysosomal proteins and their re-ceptors? Compared with control fibroblasts,twofold increases in man-6-P/IGF-II receptorshave been observed for fibroblasts from pa-tients with I-cell disease. This increase in recep-tor concentration stems from an increased rateof synthesis, not from differences of receptorstability. Interestingly, when they are exposedto insulinlike growth factors I and II or tumor-promoting phorbol esters, I-cell fibroblasts re-spond differently from control fibroblasts.Theseobservations indicate multiple regulatory sitesin the man-6-P/IGF-II receptor pathway.

Lysosomes have more complex functionsthan simply being the end point of a degrada-tive pathway. It is now emerging that the limit-ing membranes around these organelles andtheir associated proteins (LAMP-1, LAMP-2, andlysosomal integral membrane protein 2/lysoso-mal membrane glycoprotein 85) exhibit very in-teresting functions. A multidomain family ofproteins implicated in protein trafficking ofmannose 6-phosphate receptors and associatedcargo from the TGN to endosomes (the GGA orGolgi localizing, γ-adaptin ear homology do-main, ARF binding) have now been shown toparticipate in lysosomal enzyme sorting via theCD-MPR (Doray et al., 2002), and specific GGAinteract with an adaptor protein (AP-1) to pack-age MPR cooperatively into coated vesicles atthe TGN (Bai et al., 2004). Thus, these interac-tions may be physiologically important in en-suring proper transfer of the MPR into coatedvesicles at the TGN.

The discovery that MPR 300 binds and inter-nalizes IGF-II, which does not contain mannose6-phosphate residues, raised the fascinatingpossibility that this receptor may function intwo diverse biological processes: protein traf-ficking and transmembrane signaling. The re-sults of several studies have shown that IGF-IImediates transmembrane signaling through anindependent receptor. These responses includestimulation of glycogen synthesis, amino acid

uptake, cell proliferation, Na+/H+ exchange, in-ositol triphosphate production, and Ca2+ influx.

In light of the increased number of man-6-P/IGF-II receptors in I-cell fibroblasts, the aboveinteractions of IGF could have far-reaching ef-fects. For example, I-cell disease has not beentypically associated with abnormalities in phos-phorous/calcium metabolism. The extensiveskeletal deformities could involve impairmentof mechanisms of orderly calcium deposition.Rather than resulting from a primary disorderof calcium metabolism, it is possible that thebone lesions in I-cell disease are secondary toaltered lysosomal processing events in the kid-ney or liver.

Terman and coworkers (2002) recently re-ported that although I cells can undergo apop-tosis, their response to different proapoptoticagonists is substantially delayed, likely due tofunctional inefficiency of multiple lysosomalhydrolases resulting from their nonlysosomalcompartmentalization. This finding supportsthe emerging notion that lysosomal enzymesare important inducers/mediators of apoptosisand suggests that I cells may be a suitable modelfor exploring this process.

THERAPY

The notion that lysosomal storage diseasescould be treated by replacing the defective en-zyme with its normal counterpart was first sug-gested by de Duve 40 years ago. Only recentlyhas this approach become available for treat-ment of three human lysosomal storage dis-eases: Gaucher disease, Fabry disease, andmucopolysaccharidosis (MPS) type I (Bartonet al., 1991; Desnick, 2001; Kakkis et al., 2001).Although promising, this approach has yet to betried with I-cell disease, and except for sympto-matic therapies, I-cell disease remains untreat-able. Thus, the clinical course of I-cell disease ischronic, progressive, and fatal.

Because lysosomal hydrolases are synthe-sized in the rough ER and cotranslationallymodified in the rough ER lumen, a major con-sideration in I-cell disease enzyme replacementtherapy is one involving correct insertion of vi-able phosphotransferase enzyme in the Golgi.Alternatively, therapy could provide multiplenormal enzymes to the lysosomes of abnormalcells.

One case of I-cell disease has been reportedto exhibit biochemical and clinical improvement


following bone marrow transplantation (Yam-aguchi et al., 1989). The precise mechanism forthis improvement remains obscure because lyso-somal storage diseases are not specific to bloodcells. However, macrophages are important ther-apeutic targets for many lysosomal storage dis-eases. These phagocytic cells have a complicatedendocytosis mechanism involving several differ-ent receptors (mannose, mannose 6-phosphate,asialoglycoprotein, and various scavenger recep-tors) for internalization of specific lysosomal en-zymes. One plausible explanation is that cellsderived from hematopoietic progenitors of thedonor (i.e., circulating leukocytes and tissuemacrophages) can donate lysosomal enzymes tothe deficient cells in all tissues of the host eitherthrough secretion or direct cell–cell interactions.

Considerable progress in gene replacementtherapy has been made using retroviral vectorscarrying lysosomal enzyme-encoding cDNAssince the initial work of Wolfe and coworkers(1990) correcting deficient human and caninefibroblasts. Several new animal models have re-cently been characterized (Ellinwood et al.,2004). A viral vector gaining increased interestfor use in lysosomal storage is the adeno-associated viral vector,which exhibits low toxic-ity and supports long-term transgene expressionin mice as well as large animals. Importantly,most of the cDNAs that encode lysosomal en-zymes are relatively small.

A clever alternative to the above enzyme-replacement approach is that of LeBowitz andcoworkers (2004) entailing the use of a peptide-based targeting system for delivery of enzymesto lysosomes in a murine MPS type VII model.This strategy depends on the interaction of afragment of IGF-II with the IGF-II binding siteon the bifunctional, IGF-II cation-independentmannose 6-phosphate receptor. A chimericprotein containing a portion of mature humanIGF-II fused to the C terminus of human β-glucuronidase containing the catalytic site wastaken up by MPS VII fibroblasts in a mannose 6-phosphate-independent manner, and its uptakewas inhibited by the addition of IGF-II. Thetagged enzyme was delivered effectively to clini-cally significant tissues in the MPS type VIImouse model and was effective in reversing thestorage pathology. Although both mannose-6-phosphate receptors are involved in lysosomaltargeting, this approach is ideal for I-cell disease,with the major consideration the generation of afamily of chimeric proteins rather than just oneto replace only one deficient enzymic activity.

QUESTIONS

1. In certain cell types (e.g., fibroblasts)from patients with I-cell disease, lyso-somal enzymes are secreted into the extra-cellular milieu rather than targeted tolysosomes. In contrast, other cells (such ashepatocytes, Kupffer cells, and leuko-cytes) contain nearly normal levels of lyso-somal enzymes, even though these cellsare also deficient in phosphotransferaseactivity. What do these observations sug-gest about the targeting of lysosomal en-zymes in these different kinds of cells?

2. You are interested in devising a methodfor prenatal diagnosis of I-cell disease. Us-ing the enzymes listed in Table 17-3,which of these enzyme activities wouldbe most useful in this endeavor and why?

3. Acid sphingomyelinase is a lysosomalenzyme that catalyzes the breakdown ofsphingomyelin to ceramide and phosphoryl-choline. A deficiency of this enzyme leadsto lysosomal accumulation of sphingo-myelin in patients with Niemann-Pick dis-ease. Recent data indicate that correctintracellular targeting of acid sphingo-myelinase to lysosomes is dependent onthe mannose 6-phosphate-mediated path-way. Does this imply that the I-cell patientwill present with Niemann-Pick symptoms?Can I-cell disease be viewed as a constella-tion of many lysosomal storage diseases?

4. In this chapter, we indicate that the meta-bolic error (i.e., deficiency of phospho-transferase activity in I-cell disease) givesrise to a secondary phenotype of general-ized diminished lysosomal enzyme activ-ity. What other metabolic defects in themannose 6-phosphate-mediated uptakesystem could result in such a phenotype,and how would you confirm the defect?

Acknowledgments: I thank Dr. Robert DeMars for pro-viding the original photographs first describing the I celland Dr. Robert H. Glew for his helpful comments andsuggestions.

BIBLIOGRAPHY

Bai H, Doray B, Kornfeld S: GGA1 interacts withthe adaptor protein AP-1 through a WNSF se-quence in its hinge region. J Biol Chem279:17411–17417, 2004.


Bao M, Elmendorf BJ, Booth JL, et al.: Bovine UDP-N-acetylglucosamine:lysosomal-enzyme N-acetyl-glucosamine-1-N-phosphotransferase. J Biol Chem271:31446–31451, 1996.

Barton NW, Brady RO, Dambrosia JM, et al.: Re-placement therapy for inherited enzymedeficiency—macrophage-targeted glucocere-brosidase for Gaucher’s disease. N Engl J Med324:1464–1470, 1991.

Ben-Yoseph Y, Pack BA, Mitchell DA, et al.: Charac-terization of the mutant N-acetylglucosaminyl-phosphotransferase in I-cell disease andpseudo-Hurler polydystrophy: complementa-tion analysis and kinetic studies. Enzyme 35:106–116, 1986.

Besley GTN, Broadhead DM, Nevin NC, et al.: Pre-natal diagnosis of mucolipidosis II by early am-niocentesis. Lancet 335:1164–1165, 1990.

Brauker JH, Wang JL: Nonlysosomal processing ofcell-surface heparan sulfate proteoglycans. JBiol Chem 262:13093–13101, 1987.

Carey WF, Jaunzems A, Richardson M, et al.: Prena-tal diagnosis of mucolipidosis II-electron mi-croscopy and biochemical evaluation. PrenatDiagn 19:252–256, 1999.

Demars R, Leroy JG:The remarkable cells culturedfrom a human with Hurler’s syndrome: an ap-proach to visual selection for in vitro geneticstudies. In Vitro 2:107–118, 1966.

Denzer K,Weber B, Hille-Rehfeld A, et al.: Identifi-cation of three internalization sequences in thecytoplasmic tail of the 46 kd mannose 6-phosphate receptor. Biochem J 326:497–505,1997.

Desnick RJ: Enzyme replacement and beyond. J In-herit Metab Dis 24:251–265, 2001.

Doray B, Bruns K, Ghosh P, et al.: Interaction of thecation-dependent mannose 6-phosphate recep-tor with GGA A proteins. J Biol Chem277:18477–18482, 2002.

Ellinwood NM,Vite CH, Haskins ME: Gene therapyfor lysosomal storage diseases: the lessons andpromise of animal models. J Gene Med 6:481–506, 2004.

Eskelinen EL, Tanaka Y, Saftig P: At the acidicedge: emerging functions for lysosomal mem-brane proteins. Trends Cell Biol 13:137–145,2003.

Falik-Zaccai TC, Zeigler M, Bargal R, et al.: Mucol-ipidosis III type C: first-trimester biochemicaland molecular prenatal diagnosis. Prenat Di-agn 23:211–214, 2003.

Hickman S, Neufeld EF: A hypothesis for I-celldisease: defective hydrolases that do not enterlysosomes. Biochem Biophys Res Commun 49:992–999, 1972.

Jansen SM, Groener JEM, Poorthuis VJHM: Lysoso-mal phospholipase activity is decreased inmucolipidosis II and III fibroblasts. BiochimBiophys Acta 1436:363–369, 1999.

Kakkis ED, Muenzer J, Tiller GE, et al.: Enzyme-replacement therapy in mucopolysaccharidosisI. N Engl J Med 18:182–188, 2001.

Kasper D, Dittmer F, von Figura K, et al.: Neithertype of mannose 6-phosphate receptor is suffi-cient for targeting of lysosomal enzymes alongintracellular routes. J Cell Biol 134:615–623,1996.

Kornfeld S: Trafficking of lysosomal enzymes innormal and disease states. J Clin Invest 77:1–6,1986.

Kornfeld S: Trafficking of lysosomal enzymes.FASEB J 1:462–468, 1987.

Kornfeld S: Structure and function of the man-nose 6-phosphate/insulin-like growth factor IIreceptor. Annu Rev. Biochem. 61: 307–330,1992.

Kornfeld R, Kornfeld S: Assembly of asparagine-linked o1igosaccharides. Annu Rev Biochem54:631–664, 1985.

Lang L, Reitman M,Tang J, et al.: Lysosomal enzymephosphorylation. Recognition of a protein-dependent determinant allows specific phospho-rylation of oligosaccharides present on lysosomalenzymes. J Biol Chem 259:14663–14671,1984.

LeBowitz JH,Grubb JH,Maga JA,et al.:Glycosylation-independent targeting enhances enzyme deliv-ery to lysosomes and decreases storage inmucopolysaccharidosis type VII mice. Proc NatlAcad Sci U.S.A. 101:3083–3088, 2004.

Leroy JG, Ho MW, MacBrinn MC, et al.: I-cell dis-ease: biochemical studies. Pediatr Res 6:752–757, 1972.

Mononen IT, Kaartinen VM, Williams JC: A fluoro-metric assay for glycosylasparaginase activityand detection of aspartylglycosaminuria. AnalBiochem 208:372–374, 1993.

Mueller OT, Honey NK, Little LE: Mucolipidosis IIand III. The genetic relationships between twodisorders of lysosomal enzyme biosynthesis. JClin Invest 72:1016–1023, 1983.

Parvathy MR, Mitchell DA, Ben-Yoseph Y: Prenataldiagnosis of I-cell disease in the first and secondtrimesters. Am J Med Sci 297:361–364, 1989.

Poenaru L, Castelnau L,Tome F, et al.:A variant ofmucolipidosis II. Clinical, biochemical andpathological investigations. Eur J Pediatr147:321–327, 1988.

Raas-Rothschild A, Bargal R, Goldman O, et al.:Genomic organization of the UDP-N-acetyl-glucosamine-1-phosphotransferase γ subunit(GNPTAG) and its mutations in mucolipidosisIII. J Med Genet 41:e52, 2004.

Sandholzer U, von Figura K, Pohlmann R: Functionand properties of chimeric MPR 46-MPR 300mannose-6-phosphate receptors. J Biol Chem275:14132–14138, 2000.

Spranger JW, Wiedemann HR: The genetic mucol-ipidoses. Diagnosis and differential diagnosis.Humangenetik 9:113–139, 1970.


Terman A, Neuzil J, Kagedal K, et al.: Decreasedapoptotic response of inclusion-cell disease fi-broblasts: a consequence of lysosomal enzymemissorting? Exp Cell Res 274:9–15,2002.

Verkruyse LA, Natowicz MR, Hofmann SL:Palmitoyl-protein thioesterase deficiency in fi-broblasts of individuals with infantile neuronalceroid lipofuscinosis and I-cell disease. BiochimBiophys Acta 1361:1–5, 1997.

von Figura K, Hasilik A: Lysosomal enzymes mindtheir receptors.Annu Rev Biochem 55:167–193,1986.

Wolfe JH, Schuchman EH, Stramm LE, et al.:Restoration of normal lysosomal function in mu-copolysaccharidosis type VII by retroviralvector-mediated gene transfer. Proc Natl AcadSci U.S.A. 87:2877–2881, 1990.

Yamaguchi K,Hayasaka S,Hara S,et al.: Improvementof tear lysosomal enzyme levels after treatmentwith bone marrow transplantation in a patientwith I-cell disease.Ophthal Res 21:226–229,1989.

Ylikangas PK, Mononen IT: Glycosylasparaginaseas a marker enzyme in the detection of I-cell dis-ease. Clin Chem 44:2543–2544, 1998.


CASE REPORT

A male infant was born at term following an un-complicated pregnancy to a healthy womanpregnant for the first time.The delivery was un-complicated, and growth was appropriate forgestational age. Initially, he was breastfed every3 h.On the second day of life,he developed mildrespiratory distress and fed poorly. Antibioticswere started, but no evidence for an infectionwas found. On day 3, he became increasingly un-responsive. Arterial blood gases revealed a nor-mal PO2 with a respiratory alkalosis (pH 7.56,PCO2 17, HCO3

− 15 µmol/L (normal 21–28µmol/L)). Blood glucose and electrolytes werenormal.He became comatose and required artifi-cial ventilation. Repeat tests showed persistentrespiratory alkalosis, and the plasma ammoniumwas 1250 µmol/L (normal 5–30). Serum elec-trolytes, plasma glucose, and lactate and a toxi-cology screen were normal.Plasma was obtainedfor quantitative amino acid analysis and quantita-tive carnitine and for an acylcarnitine profile;urine was obtained for orotic acid measurementand organic acid profiling.

All enteral feeds were stopped.The baby wasgiven intravenous glucose, L-arginine, sodiumbenzoate, and sodium phenylacetate. Hemodial-ysis was initiated. At this time, there were nospontaneous respirations, there was no re-sponse to painful stimuli, and brainstem reflexeswere absent. The plasma amino acid results re-vealed a glutamine level of 1500 µmol/L (nor-mal 254–823), and citrulline was undetectable(normal 10–34 µmol/L). Quantitative carnitine,plasma acylcarnitine,and urine organic acid pro-files were normal.The urine orotic acid concen-

tration was normal at 2 µmol/L. Over the next12 h, the ammonium concentration decreased to100 µmol/L; however, the baby remained coma-tose until the fifth day of life, when he began toregain brainstem reflexes, spontaneous respira-tions, and response to stimuli. He was treatedwith a low-protein diet and oral medications.

Blood leukocyte DNA was used for mutationanalysis of the carbamoyl phosphate synthetaseI (CPSI) gene, and he was found to carry twoseparate, previously described, disease-causingmutations. Subsequently, the boy suffered globaldevelopmental delay, reduced head growth, anda seizure disorder. He had several hospital admis-sions for episodic hyperammonemia associatedwith intercurrent viral illnesses. In a subsequentpregnancy, the parents had prenatal diagnosis bymolecular analysis of chorionic villus.The fetuswas not affected, and a healthy baby was born.

DIAGNOSIS

The clinical syndrome of acute neonatal hyper-ammonemic encephalopathy described in thecase report represents the classical presenta-tion of a patient with a urea cycle disorder(UCD). It is important to note that this neonatalcourse represents only the most common andsevere presentation of a UCD. This holds truefor all the diseases listed in Table 18-1, with theexceptions of arginase (ARG-1) deficiency,which results in progressive spasticity of thelower limbs, and of the mitochondrial mem-brane transporters citrin and ornithine trans-porter 1 (ORNT-1). Deficiency of citrin resultsin adult-onset encephalopathy; deficiency of

18

Inborn Errors of Urea Synthesis

PRANESH CHAKRABORTY and MICHAEL T. GERAGHTY

195

Table 18-1. Enzymes, Transporters, and Biochemical Features of Inborn Errors of Urea SynthesisDisease Gene Chromosome Tissue Plasma Urine

Enzymes

N-Acetylglutamate NAGS deficiency NAGS 17q21.31 Liver Citrulline ↓ Orotic acid ↓synthetase (NAGS) Arginine ↓

Carbamoyl Phosphate CPS-1 deificiency CPS1 2q35 Liver Citrulline ↓synthase 1 (CPS) Arginine ↓ Orotic acid ↓

Ornithine transcarbamoylase OTC deficiency OTC Xp21.1 Liver Citrulline ↓ Orotic acid ↑↑(OTC) Arginine ↓

Arginino succinate Citrullinemia type 1 AS 9q34 Liver Skin Citrulline ↑↑↑ Orotic acid ↑synthetase (AS) Arginine ↓ Homocitrulline ↑

Arginino succinate aciduria AL 7cen-q11.2 Liver, skin, RBC Citrulline ↑ Orotic acid ↑lyase (AL) Arinine ↓

ASA, arginino succinic acid ↑↑↑

Arginase Argininemia ARG1 6q23 Liver, RBC Citrulline ↑Arginine ↑↑↑ASA, arginino succinic acid ↑ Orotic acid ↑

Transporters

Ornithine transporter HHH disease ORNT1 13q14 Liver, skin Citrulline ↑ Orotic acid ↑Ornithine ↑ Homocitrulline ↑↑

Aspartate-glutamate Citrullinemia type 2 CITRIN 7q21.3 Liver Citrulline ↑transporter Arginine normal- ↑

ASA ↑ Orotic acid ↑HHH, hyperornithinemia, hyperammonemia, homocitrullinuria; RBC, red blood cells.

ORNT-1 presents with recurrent hyperam-monemia and protein aversion in infancy andchildhood. For the remainder, variant clinicalsyndromes presenting at almost any age havebeen described.

Several classes of inborn errors of metabo-lism in addition to inborn errors of urea synthe-sis can cause neonatal hyperammonemia.Theseinclude organic acidurias, fatty acid oxidationdefects, amino acidopathies, and mitochondrialrespiratory chain disorders. All of these disor-ders have a number of features in common. La-bor and delivery tend to be normal, and thereare no predisposing risk factors.Clinical featurespresent after 24 h of life and are progressive.They are inherited, and thus a family history ofpreviously affected children or neonatal deathsmay be present.While most are inherited in anautosomally recessive manner, ornithine tran-scarbamoylase (OTC) deficiency is X linked, anda family history of affected males in the mater-nal pedigree is not uncommon.

In the majority of cases, a UCD can be distin-guished from other inborn errors of metabo-lism by routinely available clinical chemistrytests such as blood gases, acid/base balance,plasma glucose, ammonium, or lactate. Ureaproduction, and hence serum urea nitrogen, isdecreased in UCDs. Respiratory alkalosis hasfew causes and is an important diagnostic clueof hyperammonemia that should trigger mea-surement of plasma ammonium.

Alkalosis is a consequence of hyperventilationcaused by increased intracerebral ammonium orglutamine accumulation. Hyperammonemia as-sociated with an increased anion gap metabolicacidosis strongly suggests an organic acidemia(e.g., propionic or methylmalonic acidemia),while concomitant hypoketotic hypoglycemia issuggestive of a defect in fatty acid oxidation(e.g., very long chain acyl-CoA dehydrogenasedeficiency). Analysis of urine organic acids,plasma carnitine levels, and a plasma acylcarni-tine profile will distinguish these groups of disor-ders and often establishes a definitive diagnosis.As ammonium incorporation into urea occursprimarily in the liver, acute or chronic liver fail-ure may result in hyperammonemia.

Several preanalytical factors can cause a falseelevation of the measured plasma ammoniumconcentration.These include prolonged tourni-quet application, hemolysis, specimen trans-portation at room temperature, delay in plasmaseparation, and even ambient ammonium fromcleaning agents or cigarette smoke. These fac-

tors seldom cause measured ammonium con-centrations exceeding 100 µmol/L.

Plasma Amino Acids

Quantitation of plasma amino acids, especiallyof citrulline, is the first step in determining theprecise enzyme or transport protein defect inpatients with a UCD. If the defect involves N-acetyglutamate synthetase (NAGS), CPSI, orOTC, then plasma citrulline concentration willbe low. Marked hypercitrullinemia (>2000µmol/L) is seen in argininosuccinate synthetase(AS) deficiency,while moderate increases (>200µmol/L, normal undetectable) are found inargininosuccinate lyase (AL) and citrin deficien-cies. In AL deficiency, the presence of argini-nosuccinic acid and its anhydrides furtherdistinguishes this disorder.

Consideration of other plasma amino acidsalso informs the diagnosis of inborn errors ofurea synthesis. The plasma concentrations ofglutamine and alanine are often elevated in par-allel with or prior to the ammonium concentra-tion as they act as a nitrogen buffer. Plasmaarginine concentrations are low since the onlysynthetic route for arginine in humans is viathe urea cycle. In contrast, the arginine con-centration is elevated in ARG-1 deficiency. Hy-perornithinemia and homocitrullinuria are thecharacteristic features of the hyperammone-mia, hyperornithinemia, and homocitrullinuria(HHH) syndrome caused by a defect in the or-nithine transporter (ORNT-1).

Urine Orotic Acid

Orotic acid is an intermediate in pyrimidinesynthesis. It is synthesized from the transcar-bamylation of aspartic acid and subsequentintramolecular condensation. Any defect inureagenesis causing accumulation of intracel-lular carbamoyl phosphate provides substratefor orotic acid synthesis.Therefore, a defect ofOTC, or any defect distal to this step, can causeorotic aciduria.The detection of elevated oroticacid in the urine is most useful in differentiatingbetween patients with OTC deficiency and ei-ther CPSI- or NAGS-deficient patients in whomorotic aciduria is not present.

Enzymology

Definitive diagnosis of a UCD can be made byspecific enzyme assays. A deficiency of OTC,

Inborn Errors of Urea Synthesis 197

CPSI, or NAGS is established by liver biopsy asthese enzymes are expressed only in liver. Dueto random X inactivation, women with OTC de-ficiency may have normal enzyme activity.Thediagnosis in these women must be establishedby interpretation of the analytes describedabove or by mutation analysis. AS is assayed inskin fibroblasts as well as liver;AL is assayed in anumber of additional tissues, including red andwhite blood cells.ARG-1 activity is measured inliver or red blood cells.

Mutation Analysis

The genes for the six enzymes and two trans-porters involved in the urea cycle have beenidentified, and disease-causing mutations havebeen reported in each. Most mutations are pri-vate,with no common mutation accounting for alarge proportion of affected individuals. Muta-tion testing is therefore complicated by the needto screen for mutations in the entire coding re-gion and, in some cases, for larger gene re-arrangements or mutations in regulatory regions.


In contrast to the storage of glucose in glyco-gen, and two-carbon skeleton storage in fattyacids, animals do not possess a mechanism tostore nitrogen. Amino acids are primarily usedfor the synthesis of protein and other nitroge-nous compounds such as purine nucleotidesand heme. Excess amino acids are deaminated,and their carbon skeletons are used directly asan energy source or for gluconeogenesis or ke-togenesis.All animals therefore require a mecha-nism to excrete excess nitrogen.

Several evolutionary solutions exist to ac-complish this. From a teleological viewpoint,there are two main reasons for the develop-ment of nonammonium nitrogen-containingexcretion products. First, ammonium is poorlysoluble and is therefore associated with a largeobligate water loss. Second, ammonium ishighly toxic to vertebrate central nervous sys-tems. Humans, like other mammals, excreteurea and are therefore called ureotelic organ-isms.This word derives from the Greek ouronmeaning “urine” and telos meaning “end.” Incontrast, birds and reptiles are primarily uri-cotelic (excreting uric acid), and marine ani-mals are primarily ammoniatelic (excretingammonium).

Urea is synthesized via the urea cycle (Fig.18-1). In 1932, Krebs and Henseleit publisheddata demonstrating that ornithine stimulatesthe synthesis of urea without stoichiometricconsumption of this intermediate. This appar-ent catalytic function was determined to be theresult of the cyclic nature of the pathway.Thiswas a revolutionary idea since metabolic path-ways were conceptualized as purely linearprior to the publication of these observations.In the following sections, we discuss the bio-chemical processes involved in urea formation.

Nitrogen Sources for Urea

Nitrogen incorporated into urea in the liveroriginates from both intrahepatic and extrahep-atic sources. Urea contains two nitrogen atoms;one is derived from ammonium,while the otheris derived from aspartic acid. In liver hepato-cytes, ammonium is generated by the oxidativedeamination of glutamic acid, catalyzed by glu-tamate dehydrogenase,

Glutamate + NAD+ ↔ α-Ketoglutarate + NH3

+ NADH + H+

as well as from deamination of a number ofother amino acids. Alternatively, the α-nitrogenfrom glutamate or alanine may be incorporatedinto urea via aspartate following transaminationof oxaloacetate.

The extrahepatic organs involved in ureasynthesis are muscle, small bowel, and kidney.Muscle is an important source of alanine thatfeeds both the hepatic urea cycle as well as he-patic gluconeogenesis. Once in the hepatocyte,the α-nitrogen of alanine may be transferred toα-ketoglutarate to produce glutamate used forammoniagenesis or formation of aspartate.Theother product of alanine transamination is pyru-vate, which is used in gluconeogenesis. Muscleis also a net producer of glutamine, which isused in the small intestine to produce ammo-nium, alanine, and citrulline. These urea cycleprecursors are efficiently transported to theliver via the portal circulation, where ammo-nium and alanine are metabolized as described.While citrulline may be used in the hepaticurea cycle, a substantial amount is metabolizedto arginine by the kidney prior to completionof the urea cycle in the liver.There is also a re-nal glutaminase activity, which may supply am-monium for the hepatic urea cycle.


Figure 18-1. Urea synthesis in the liver.Enzymes and transporters are in bold.N-Acetylglutamate (NAG),an allosteric activator of carbamoylphosphate synthetase (CPSI) is formed by NAG synthase (NAGS). Organic acids are competitive inhibitors of this reaction. CP, carbamoylphosphate; OTC, ornithine transcarbamoylase; AS, argininosuccinate synthetase; AL, argininosuccinate lyase; ARG-1, arginase; ORNT-1,ornithine-citrulline antiporter. Nitrogen atoms are represented as *.

Carbamoyl Phosphate Synthetase I

CPSI catalyzes the formation of carbamoylphosphate from bicarbonate, ammonium, andtwo adenosine triphosphate molecules (Fig.18-1).This first step of the urea cycle occurs inthe mitochondrial matrix and assimilates thefirst of the two nitrogen atoms that will even-tually be found in urea. While two ATP mole-cules are hydrolyzed, there is formation of alower energy bond in carbamoyl phosphate.CPSI is a homodimer that accounts for 15–30%of the total protein mass in liver mitochondria.N-Acetylglutamate (NAG) is an essential al-losteric activator of CPSI activity, and magne-sium ions are also required for its activity.

A second, cytosolic CPS activity (CPSII) oc-curs in mammals as part of the CAD trifunc-tional protein that catalyzes the first three stepsof pyrimidine synthesis (CPSII, asparate tran-scarbamoylase, and dihydroorotase).The activi-ties of these three enzymes—CPSII, aspartatetranscarbamoylase, and dihydroorotase—resultin the production of orotic acid from ammo-nium, bicarbonate, and ATP. CPSII has no role inureagenesis, but orotic aciduria results fromhepatocellular accumulation of carbamyl phos-phate and helps distinguish CPSI deficiency fromother UCDs. Defects in CPSI classically presentwith neonatal acute hyperammonemic en-cephalopathy. The plasma citrulline and urineorotic acid concentrations are both low.A defini-tive diagnosis can be established by enzyme assayof biopsied liver tissue or by mutation analysis.

N-Acetylglutamate Synthase

NAG is an essential allosteric activator of CPSIactivity that is produced by intramitochondrialcondensation of acetyl-CoA and glutamate cat-alyzed by NAGS. NAG plays an important role inthe regulation of the urea cycle. Its productionis upregulated by arginine; its transport into thecytosolic compartment for degradation is inhib-ited by glucagon. The clinical and laboratorypresentation of NAGS deficiency is indistin-guishable from CPSI deficiency. Diagnosis canbe confirmed by NAGS mutation analysis.

Ornithine Transcarbamoylase

The second intramitochondrial step of theurea cycle is the reversible condensation ofcarbamoyl phosphate and ornithine, catalyzedby OTC. The high-energy phosphate bond incarbamoyl phosphate is cleaved during this

reaction, and equilibrium favors the forward re-action to produce citrulline. OTC is a ho-motrimer of 35-kd polypeptides. It has a highspecific activity and is found outside the liverin small intestine and to some extent in kidney.

OTC deficiency is the most common urea cy-cle defect.As it is X linked,affected boys typicallyhave severe disease with neonatal presentationas described in this chapter. The disease inwomen who carry an OTC mutation on one Xchromosome ranges from severe early-onset dis-ease to complete absence of symptoms. Further-more,affected women may decompensate in thecontext of a metabolic stress such as an infectionor following parturition. OTC-deficient patientshave low plasma citrulline and high urine oroticacid. Confirmation of the diagnosis requires mu-tation analysis or a liver biopsy for enzymology.The carrier status of women is most accuratelydetermined by mutation analysis.

Ornithine-Citrulline Transporter

Citrulline is exchanged for ornithine across theinner mitochondrial membrane by ORNT-1. Or-nithine is produced in the cytosol as the finalstep in the urea cycle and must be returned tothe mitochondrial matrix for transcarbamoyla-tion by OTC. A second ornithine-citrulline an-tiporter (ORNT-2) is also expressed in the livermitochondria and may attenuate the severity ofdisease in patients with HHH (Hyperammonemia,Hyperornithinemia, Homocitrullinuria) diseasedue to ORNT-1 deficiency.This disorder typicallymanifests later in life with intermittent hyperam-monemic encephalopathy and protein aversion.Intramitochondrial ornithine deficiency causesboth hyperammonemia and hyperornithinemiadue to a lack of substrate for OTC. Homoc-itrullinuria occurs due to the use of lysine byOTC as an alternate substrate. The diagnosis isconfirmed by mutation analysis.

Argininosuccinate Synthetase

Once citrulline is in the cytosol, argininosuc-cinic acid is formed by condensation of cit-rulline with aspartate.This is where the secondnitrogen atom enters the cycle. Argininosucci-nate synthetase, a homotetramer of a 46-kdpolypeptide catalyzes the reversible reactionaccompanied by hydrolysis of ATP to AMP andpyrophosphate. The subsequent hydrolysis ofpyrophosphate shifts the equilibrium to theright and results in the consumption of twohigh-energy phosphate bonds.


AS activity is found in liver,kidney, skin fibrob-lasts, and some areas of the brain. Patients withcitrullinemia classically present with neonatalhyperammonemic encephalopathy. The diseaseis easily distinguishable by the very high concen-tration of citrulline in plasma. Urine orotic acidis usually elevated.The major source of circulat-ing citrulline is extrahepatic. Observations ofpatients with OTC or AS deficiency who haveundergone liver transplantation support thisfinding. Citrulline levels in such patients do notnormalize; they remain low in the OTC-deficientpatients and do not entirely normalize in AS-deficient patients.

Argininosuccinate Lyase

AL is found in a wide variety of tissues, includingliver, kidney, skin fibroblasts, red blood cells, andparts of the brain.It cleaves argininosuccinic acidto form arginine and fumarate.The formation ofarginine via the urea cycle is the reason that it isa nonessential dietary amino acid.The cytosolicfumarate hydratase and malate dehydrogenaseactivities act on fumarate to produce oxaloac-etate. Oxaloacetate may be transaminated toform aspartate, which is then used again in theurea cycle or in cytosolic gluconeogenesis. Pa-tients with AL deficiency or argininosuccinicaciduria usually develop early-onset hyperam-monemic encephalopathy. Plasma citrulline ismoderately elevated. Argininosuccinic acid andits two anhydrides are detected in the plasmaamino acid profile.The anhydrides may coelutewith the neutral amino acids isoleucine orleucine or with the dipeptide homocystine inmany ion exchange chromatography methodsused for amino acid quantitation. Patients oftendevelop progressive liver dysfunction and devel-opmental delays in spite of treatment.

Arginase

The final step of the urea cycle is the cleavageof arginine to release urea and regenerateornithine. Ornithine then reenters the mito-chondria via the ORNT-1 ornithine-citrullineantiporter. ARG-1 is a cytosolic homotrimericenzyme of 35-kd monomers that is expressedin liver and red blood cells.A second mitochon-drial arginase (ARG-2) most likely plays a rolein nitric oxide synthesis and is most abundantin brain, kidney, and prostate.ARG-1 deficiencyis unique among the urea cycle deficiencies aspatients do not present with hyperammonemiaand encephalopathy but rather develop pro-gressive spasticity of the lower limbs. Biochem-

ically, the plasma arginine concentration ismarkedly elevated.

Aspartate-Glutamate Transporter(Citrin)

Citrin is an aspartate-glutamate antiporter thathas a role both in the urea cycle and in themalate aspartate shuttle. It is necessary for thetransport of aspartate produced in the mito-chondria into the cytosol, where it is used byAS. Its role in the malate-aspartate shuttle is totransport cytosolic NADH reducing equivalentsinto the mitochondria, where they are used inoxidative phosphorylation. Defects in citrincause citrullinemia type II. Patients manifestlater-onset intermittent hyperammonemic en-cephalopathy as in HHH syndrome.

THERAPY

Treatment for patients with a UCD can be di-vided into two parts: acute management of hy-perammonemic encephalopathy and long-termcontrol to prevent further episodes of hyperam-monemia while maintaining normal growth anddevelopment.

Acute Management of Hyperammonemic Encephalopathy

Ammonia is very toxic to the CNS, and elevatedlevels over a period of time result in irreversiblebrain damage and poor outcome.The treatmentof choice for acute hyperammonemic coma ishemodialysis (or peritoneal dialysis if hemodial-ysis is not available).While awaiting hemodialy-sis, the following measures should be initiatedimmediately. Protein (exogenous nitrogen) in-take must be stopped and adequate caloriesprovided in the form of glucose and fat to pre-vent further catabolism (endogenous nitrogen).

Intravenous arginine should be providedsince it becomes an essential amino acid inUCDs,and it promotes the formation of citrulline(in AS deficiency) or argininosuccinic acid (in ALdeficiency), both of which are excreted in theurine, thus providing some waste nitrogen excre-tion. In addition, intravenous sodium benzoateand sodium phenylacetate can be given. Theseprovide an alternate mechanism of waste nitro-gen disposal (Fig. 18-2). Once the plasma ammo-nium has been normalized, these drugs can becontinued while refeeding (intravenous or en-teral) is reintroduced. Oral sodium benzoate andsodium phenylacetate (or alternatively sodium


phenylbutyrate) is introduced,but careful match-ing of nitrogen intake versus waste disposal bythese medications must be observed.

Long-Term Management of Patientswith Urea Cycle Disorders

Long-term measures to reduce ammonium lev-els rely on dietary protein restriction. However,this approach alone is not sufficient.Two majoradvances in treatment have contributed togreater survival and improved outcome. First,the recognition that arginine is an essentialamino acid in patients with a UCD, and that de-ficiency leads to endogenous protein break-down and negative nitrogen balance. Second, itwas recognized that one could provide an alter-nate pathway for waste nitrogen disposal by theuse of sodium benzoate and sodium phenylac-etate (Fig. 18-2). Sodium benzoate is conjugatedwith glycine (a non-essential amino acid con-taining a single nitrogen atom) to form hippuricacid. Sodium phenylacetate is conjugated withglutamine (a nonessential amino acid with twonitrogen atoms) to form phenylacetylglutamine.Both hippuric acid and phenylacetylglutamineare excreted in the urine.

The combination of protein reduction and al-ternate waste nitrogen establishes an alternatehomeostasis. From the physician’s point of view,it remains very difficult to balance the ongoingdynamic of nitrogen turnover using a static treat-ment of prescribed diet and fixed-dose medi-cations. Changing growth rates, illness, andinadvertent protein intake all contribute to re-current hyperammonemia.Constant surveillanceof growth (e.g., height, weight, head circumfer-ence) and of adequate nutrition (blood concen-trations of hemoglobin, prealbumin, vitamins,and trace elements) and monitoring of diseasecontrol (e.g., plasma concentrations of ammo-nium, glutamine, alanine, and other amino acids)are required throughout life.

Treatment of CPS, OTC, or NAGS deficiencygenerally requires daily restriction of dietarynitrogen to 0.7g/kg/day milk protein and0.7 g/kg/day essential amino acids. The lattercontains roughly 12% nitrogen as opposed toapproximately 16% in milk protein. Nitrogenrequirements are normally increased in thevery young, during accelerated growth phases(infancy and puberty), and after acute illnessand surgery. Citrulline supplementation at170 mg/kg/day is given. Citrulline has two ad-


Figure 18-2. Nitrogen (*) incorporation into benzoate and phenylacetate as pheny-lacetylglutamine (PAG) and hippuric acid (HA) provide an alternate path to urea forwaste nitrogen excretion.

vantages over arginine: It is deficient in both ofthese disorders, and it contains less nitrogen (3atoms) than arginine (4 atoms), thus minimiz-ing intake of nitrogen. Sodium benzoate(250 mg/kg/day) and sodium phenylactetate(250 mg/kg/day) are recommended. More re-cently, sodium phenylbutyrate (500 mg/kg/dayor 12 g/m2/day), which has a less-offensiveodor, has been used. Sodium phenylbutyrate re-quires one round of fatty acid oxidation to pro-duce phenylacetyl-CoA, which is thenconjugated with glutamine.A structural analogof NAG, carbamoylglutamate has been shownto bind and activate CPSI, and several patientswith NAGS deficiency have been reported torespond to this compound.

Treatment of AS deficiency is similar to thatfor CPS and OTC deficiencies. Protein intake isrestricted to 1.5–1.75 g/kg/day, and sodiumbenzoate and sodium phenylacetate (or sodiumphenylbutyrate) are provided. Arginine supple-mentation is required not only for protein syn-thesis but also to stimulate waste nitrogenexcretion in the form of citrulline.Unfortunately,citrulline formation allows incorporation of onlya single ammonium nitrogen atom, and its renalclearance is only 25% that of the glomerular fil-tration rate. Nevertheless, citrulline does pro-vide an alternative, albeit reduced, form ofwaste nitrogen excretion. Increasing attentionis being paid to the long-term metabolic effectsof chronic hypercitrullinemia and hyperarginine-mia as a result of therapy.

In AL deficiency, argininosuccinic acid is amuch better waste nitrogen product than cit-rulline. It contains two waste nitrogen atoms(i.e., one from carbamoyl phosphate and onefrom aspartate), and its renal clearance is equalto that of the glomerular filtration rate. Thus,arginine supplementation alone allows suffi-cient waste nitrogen excretion to prevent hy-perammonemia. Moderate protein restriction(1.5–1.75 g/kg/day) is also recommended. Asin AS deficiency, the long-term metabolic effectsof increased citrulline and argininosuccinicacid in this disorder are unclear. Despite ade-quate treatment (i.e., prevention of recurrenthyperammonemia), many patients have devel-opmental delays and unexplained liver disease.

ARG-1 deficiency is treated by arginine re-striction to about 300 mg/day. Patients thus re-quire strict limitation of protein supplementedwith essential amino acids.The effectiveness ofsuch treatment is unclear. Patients with ORNT-1or citrin deficiency are treated with low-proteindiets. The use of sodium benzoate or sodium

phenylbutyrate to provide an alternate waste ni-trogen disposal pathway may be beneficial inthese disorders.

QUESTIONS

1. List several metabolic fates of the aminoacid arginine in the human. What is thedifference between essential and non-essential amino acids? How is this relevantto the treatment of urea cycle defects?

2. Carbamoyl phosphate is a substrate fortwo separate enzymes. What are the en-zymes involved and describe the meta-bolic fates of their respective products?

3. A patient with OTC deficiency has recur-rent hyperammonemic events over a 3-month period despite reduced proteinintake and increased doses of sodium ben-zoate and sodium phenylacetate.What arethe factors that need to be considered as acause of the hyperammonemia in this in-dividual? (Consider the dynamic of nitro-gen balance over this period.) What is thelikely treatment?

4. Describe how recent advances in mo-lecular genetics have contributed to ourknowledge of the urea cycle. List some ap-plications of these advances to medicalpractice.

5. How does the location of the gene thatencodes OTC on the X chromosome con-tribute to greater clinical variation in pa-tients with deficient enzyme activity? Listreasons why the diagnosis of OTC defi-ciency is difficult in women.

BIBLIOGRAPHY

Brusilow SW: Treatment of urea cycle disorders, inDesnick RJ (ed):Treatment of Genetic Diseases.Churchill Livingstone, New York, 1991, pp.79–94.

Brusilow SW, Horwich AL: Urea cycle enzymes, inScriver C,Beaudet A,Sly W,et al. (eds): The Meta-bolic and Molecular Bases of Inherited Dis-ease. 8th ed. McGraw-Hill, New York, 2001, pp.1909–1963.

Brusilow SW, Maestri NE: Urea cycle disorders:diagnosis, pathophysiology and therapy. AdvPediatr 43:127–170, 1996.

McCabe RB, Bachmann C, Batshaw, et al (eds):New developments in urea cycle disorders. MolGenet Metab 81(suppl 1):S4–S91, 2004.


CASE REPORT

Penny Urick is a 22-year-old G1P0 (one preg-nancy, no births) woman who was seen at thefamily practice clinic for routine prenatal careduring her first pregnancy. Mrs. Urick describedherself as “generally healthy.”At the time of hervisit, she stated that she took no medicationsand had no known medical allergies; however,her medical record revealed a history ofphenylketonuria (PKU). When further ques-tioned about her experience with PKU Mrs.Urick stated that:

I had it as a child, but I am fine now.They putme on this diet, which I had to stay on until Iwas 18 years old, when I was told that I didn’tneed it anymore. I haven’t had any problems atall since I stopped it.

On further questioning, it became clear thatMrs. Urick was concerned that her baby mightinherit the disorder, and that if so, the babywould need to be on the PKU diet. Mrs. Urickdescribed her experience with PKU manage-ment as follows:

It [the diet] was terrible! You can’t eat anythingeveryone else can. I couldn’t eat meat or fish orcheese or eggs.While everyone else was eatingthese things, I had to eat potatoes and cereal.You get pretty sick of potatoes and cereal everyday! I had to constantly watch everything I ate,exchanging one thing for another,depending onthe amount of phenylalanine the various foodshad. I also had to take a daily supplement whichwas not exactly tasty. I met with a dietician con-

stantly,which my parents said was pretty expen-sive. Plus, I had to get my blood tested all thetime. That’s why I stopped the diet when Iturned 18. My doctor said he thought it wouldbe fine if I tried stopping it. I haven’t had anytrouble being on regular food, and I don’t wantmy child to have to go through all of that.

During the clinical interview Mrs. Urick ex-pressed an interest in knowing more aboutchanges in treatment should her baby havePKU. She was especially curious about tetrahy-drobiopterin (BH4) therapy, which she learnedabout while searching the Web, where shefound several sites indicating that BH4 therapymay be helpful in some patients with PKU.

DIAGNOSIS

The standard diagnostic test for PKU is theGuthrie bacterial inhibition assay.The test usesthe growth of a strain of bacteria on a speciallyprepared agar plate as a sign for the presence ofhigh levels of phenylalanine (PA), phenylpyru-vate, or phenyllactate. The compound β-2-thienylalanine will inhibit the growth of thebacterium Bacillus subtilis (ATCC 6051) onminimal culture media. If PA,phenylpyruvate,orphenyllactate is added to the medium, thengrowth is restored. Such compounds will bepresent in excess in the blood or urine of pa-tients with PKU. If a suitably prepared sampleof blood or urine is applied to the seeded agarplate, then the growth of the bacteria in thetest will be a positive indicator for PKU in thepatient.

19

Phenylketonuria

WILLIAM L. ANDERSON and STEVEN M. MITCHELL

204

Despite the fact that widespread newbornscreening for PKU has been in place since the1960s,accurate data regarding the incidence andprevalence of PKU is sparse,and it becomes diffi-cult to identify the best epidemiological evi-dence or clinical experience to use as the basisfor a recommendation for Mrs.Urick or her baby.It is clear from existing data that Mrs. Urick’s PAlevels ought to be controlled as tightly as possi-ble during her pregnancy. However, the ques-tions of how closely dietary restriction should becontrolled in nonpregnant adults and whichother treatments are most effective and availableare currently under debate.This chapter presentsthe basic science and current best clinical evi-dence regarding the most current potential carefor Mrs.Urick and her baby.

EPIDEMIOLOGY

Some of the most convincing data regarding theepidemiology of PKU comes from the 1994Newborn Screening Report of the Council of Re-gional Networks for Genetic Services. However,these data are limited, and there was a lack ofuniformity of criteria used by individual statesto screen for PKU and non-PKU hyperphenylala-ninemia (HPA). In addition, there was consider-able variability in the actual data reported. Inthe United States, the total incidence of pheny-lalanine hydroxylase (PAH) deficiency of varyingdegrees is about 1 in 10,000 births, and theprevalence of the heterozygous condition isroughly 2%.The incidence of frank PKU (serumPA levels >1 mM) is 1 in 13,500 to 19,000 births.There is wide variation regarding the reportingof the incidence of non-PKU HPA; however, cur-rent composite estimates suggest an incidenceof 1 in 48,000 among newborns.HPA is less com-mon in blacks, Asians, and Hispanics, with ahigher incidence in white and Native Americans(NIH Consensus Statement,2000).

CLINICAL MANIFESTATIONS

The clinical course of PKU is progressive and ex-hibits considerable variability among individuals.Mental retardation is the hallmark of the dis-ease in untreated patients, but other neurologi-cal manifestations are often present (Table 19-1).Newborns initially appear normal but fail to at-tain early developmental milestones, includingfailure to roll by 4 months, failure to sit by 6

months, failure to walk by 1 year, and delayedlanguage development.These delays are usuallynoticeable by 6 months. All patients with un-treated PKU will be mentally retarded, whichcannot be accurately diagnosed until after age 6years but can be strongly suspected based onpoor language skills and late motor milestones.Mental dysfunction worsens during early child-hood,with increasing dietary consumption of PAand continuing myelination during brain devel-opment. Commonly, patients eventually develophyperactivity, abnormalities of gait, disturbancesin posture and stance, seizures, and severe men-tal retardation, including autisticlike syndrome.Once brain maturation is complete, the condi-tion is thought by some physicians to stabilize,while other investigators have documented neu-rological difficulties in adults who do not strictlyadhere to a PA-restricted diet and recommendthat all affected individuals remain on the dietfor their lifetime (Koch, 2002; Longo, 1998; NIHConsensus Statement Online, 2000; Poustie,2004).

The precise cause of brain damage in patientswith PKU is unclear and likely multifactorial(Table 19-2) but most probably reflects the accu-mulation of aromatic metabolites of PA in neuraltissues.Due to accumulation of phenylacetate,pa-tients often have what is described as a “mousy”odor to their skin, hair, and urine. Hypopigmenta-tion results from decreased availability of tyrosineand competitive inhibition of tyrosinase by anoverabundance of PA (Longo, 2001; NIH Consen-sus Statement Online,2000;Scriver,2000).


The biochemical basis of Ms. Urick’s problem isan inability to dispose of excess dietary PA. Mildforms of the condition lead to HPA,and in severecases of PKU, mental retardation results from

Phenylketonuria 205

Table 19-1. Clinical Manifestations ofHyperphenylalanemia

Mental retardation

Cerebral and basal ganglion dysfunction

Rigidity, chorea, spasms, and hypotonia

Hyperactivity

Seizures

Gait abnormalities

Unstable temperature regulation

“Mousy”odor of skin, hair, and urine

accumulation of PA in the blood and tissues.Normal serum PA levels are less than 0.24 mM,and in an individual with PKU, this level in-creases above 0.9 mM. Normally, PA is not de-tected in the urine; however, an individual withPKU can excrete as much as 2 g of PA per dayin the urine. Non-PKU HPA results from a par-tial deficiency of PAH, with serum PA levels inthe range 0.24–0.9 mM. Both conditions are au-tosomal recessive disorders that are the resultof different mutations to the PAH enzyme; con-sequently, severity and clinical presentationvaries considerably between individuals.

As illustrated in Figure 19-1, there are twodifferent competing pathways available forthe catabolism of PA. One pathway involvestransamination of PA to phenylpyruvate, fol-lowed by the phenylpyruvate dehydrogenasereaction to form phenylacetyl-CoA. The prod-ucts of these reactions (phenylpyruvate,phenylacetyl-CoA) and the excretion productsphenylacetate and phenyllactate cannot befurther metabolized and consequently accu-mulate in the blood, urine, and tissues.

A second competitive pathway for the dis-posal of PA requires the initial conversion of PAinto tyrosine. This reaction is catalyzed by theenzyme PAH (phenylalanine-4-monooxygenase;EC 1.14.16.1). The resulting tyrosine moleculecan then be catabolized into fumarate and ace-toacetate. Both products are nontoxic and canbe further catabolized in the citric acid cycle. InMrs. Urick and the majority of individuals suf-fering from HPA and PKU, there is a defect inthe PAH enzyme system (NIH Consensus State-

ment Online, 2000). In the absence of an effec-tive PAH, the alternate transaminase pathway isthe only available route for the disposal of PA.

Phenylalanine Hydroxylase

A schematic representation of the phenylala-nine hydroxylation reaction is shown in Figure19-2. PAH reduces and cleaves molecular oxy-gen into two hydroxyl groups. One of the hy-droxyl functional groups is found in the reactionproduct tyrosine, while the other oxygen atomis used to modify the reaction cofactor BH4,cre-ating the pterin 4α-carbinolamine.

PAH, a nonheme iron-containing enzyme, is amember of a larger BH4-dependent amino acidhydroxylase family. In addition to PAH, the en-zyme family includes tyrosine hydroxylase andtryptophan hydroxylase. The enzymes in thisfamily participate in critical metabolic steps andare tissue specific. PAH catabolizes excess di-etary PA and synthesizes tyrosine. In adrenal andnervous tissue, tyrosine hydroxylase catalyzesthe initial steps in the synthesis of dihydrox-yphenylalanine. In the brain, tryptophan isconverted to 5-hydroxytryptophan as the firststep of serotonin synthesis. Consequently, theseenzymes are highly regulated not only by theirexpression in different tissues but also byreversible phosphorylation of a critical serineresidue found in regulatory domains of thethree enzymes. Since all three enzymes arephosphorylated and dephosphorylated by dif-ferent kinases and phosphatases in response tothe need for the different synthetic products, itis not unexpected that the exact regulatory sig-nal for each member of the enzyme family isunique.

Although regulation is unique for each mem-ber of the aromatic amino acid hydroxylasefamily, the catalytic mechanism and cofactor re-quirements for members of the family are iden-tical. During the reactions of all three enzymes,the dioxygen molecule is cleaved and incorpo-rated as a hydroxyl group into both the aro-matic amino acid and BH4. Each enzyme in thefamily displays its own unique substrate speci-ficity profile. Two interesting questions aboutthis enzyme family relate to the actual hydroxy-lation mechanism and how enzyme activity isaltered by changes in BH4 levels. Problems inany one of these hydroxylation systems canarise from either an inadequate supply of theBH4 cofactor or a defect in the enzyme or its ex-pression.


Table 19-2. Potential Causes of Impaired BrainDevelopment in PKU

Increased competitive inhibition of transport of otheramino acids required for protein synthesis in thebrain

Increased oxidative stress

Increased production of metabolites of phenylalaninethat inhibit synthesis of a variety of substancesrequired for normal brain growth

Inhibition of N-methyl-D-aspartate receptors, which areinvolved in memory and learning

Competitive inhibition of transport of other amino acidsrequired for protein synthesis

Impaired polyribosome formation or stabilization

Reduced synthesis/increased degradation of myelin

Decreased formation of norepinephrine and serotonin

Altered myelin structure and function

Glu

NADH + H+

NAD+

HO CHCCH2

OO

O C CCH2

OO NAD+

CoASH

H3N CHCCH2

OO

H3N CHCCH2

OO

OH

BH4

O2

phenylalanine hydroxylase

NADH + H+

CO2 CCH2

OS-CoA

pterin 4α-carbinolamine

transaminase dehydrogenase

dehydrogenase

Continued metabolismto fumarate and

acetoacetate

PLPα-KG

FADTPP

Lipoamide

PHENYLALANINE

TYROSINE

PHENYLPYRUVATE

PHENYLLACTATE

PHENYLACETYL-CoA

Tyrosine derived products

Figure 19-1. Pathways for the metabolic disposal of phenylalanine.There are two com-petitive pathways for the disposal of phenylalanine.One pathway involves a transaminaseenzyme phenylpyruvate, while the first step in the second pathway requires phenylala-nine to be initially converted to tyrosine. Continued metabolism of the phenylpyruvateproduced by the first pathway leads to products that cannot be further metabolized,while tyrosine can be converted into citric acid cycle intermediates.Glu,glutamate;αKG;CoASH, coenzyme A; BH4, tetrahydrobiopterin; TPP, thiamine pyrophosphate.

207

Biopterin Synthesis and Recycling

Mrs. Urick’s questions to her physician con-cerned the possible effectiveness of elevatingBH4 levels as a therapeutic intervention for this

condition,which raises the possibility of BH4 de-ficiency as a cause of her disease.As discussed,since BH4 is consumed in the hydroxylation re-action,without adequate BH4 it is possible that aBH4 deficiency could be a causative factor in


HN

N

O

H2N NH

HN

CH3

OHHON

N

O

H2N NH

NCH3

OHOH

HO

H3N CHCCH2

OO

H3N CHCCH2

OO

OH

H3N CHCCH2

OO

HN

TRYPTOPHAN

O2

H3N CHCCH2

OO

OH

H3N CHCCH2

OO

OHOH

DOPA

H3N CHCCH2

OO

HNOH

HYDROXYTRYPTOPHAN

PHENYLALANINE TYROSINE

TYROSINE


tyrosine hydroxylase

tryptophan hydroxylase

TETRAHYDROBIOPTERIN( BH4 )

PTERIN-4α-CARBINOLAMINE

O2

BH4

O2

BH4

pterin-4α-carbinolamine

pterin-4α-carbinolamine

Figure 19-2. Aromatic amino acid hydroxylase reaction.Aromatic amino acids are hy-droxylated by a common mechanism catalyzed by a family of hydroxylases.The enzymefamily consists of phenylalanine hydroxylase, tyrosine hydroxylase, and tryptophan hy-droxylase. In addition to substrate, all three enzymes require molecular oxygen and thecofactor tetrahydrobiopterin.Tetrahydrobiopterin is consumed in this reaction and con-verted into pterin 4α-carbinolamine. DOPA, dihydroxyphenylalanine.

this disease. BH4 can be either synthesized denovo from GTP or salvaged from the pterin-4α-carbinolamine product of the hydroxylase reac-tion. Both synthetic and recycling pathways forBH4 are illustrated in Figure 19-3. Consideringthat these metabolic pathways include a num-ber of different steps, there are several possiblesources for an error in BH4 production, and pa-tients with most of these possible errors havebeen identified.

The specific difficulty in answering Mrs.Urick’s question is illustrated by Figure 19-4,which compares the frequency of PKU casesknown to be caused by a BH4-related problemwith reports on effectiveness of BH4 supplemen-tation therapy.Although defects in biopterin syn-thesis and the BH4 recycling system account forless than 2% of all PKU cases (NIH ConsensusStatement Online,2000), there is a growing bodyof data that suggests that a significantly higherpercentage of PKU and HPA cases will respondfavorably (decreased serum PA levels) to BH4

therapy. In one study, 41% of 41 random patientshad PAH mutations that responded to BH4 ther-apy (Trefz and Blau, 2003), while in a retrospec-tive study of 1919 patients with mild PKU andHPA,greater than 60% responded to BH4 therapy(Bernegger and Blau,2002).

Both of these observations have led to sug-gestions that, in all newly diagnosed cases ofPKU and HPA, serum levels of PA should bemonitored following a BH4 challenge to deter-mine if the serum PA levels decrease in responseto the BH4 challenge. It has been suggested thatthis BH4 challenge test be a routine part of pa-tient management (Trefz and Blau, 2003). If thedefect in the PAH enzyme resides in the BH4

binding site and results in a lower affinity forthe cofactor, then one can easily justify a mech-anistic rationale for the proposed BH4 therapyin more than 2% of patients.

Enzyme Structure

The active PAH enzyme is a tetramer of identicalpolypeptides. Each polypeptide subunit is com-posed of three different domains: an amino ter-minal regulatory domain, the iron-containingcatalytic domain, and a carboxyl terminal poly-merization domain.To date, the crystal structureof the intact enzyme has yet to be elucidated;however, the protein database contains severalhigh-resolution human and rat PAH structuresthat were first truncated by removal of eitherthe regulatory or the polymerization domain,

thereby allowing the structure of the intact pro-tein to be determined by combining differentcrystal structures. A schematic structure of themonomeric regulatory and catalytic domains ofthe rat enzyme is illustrated in Figure 19-5.

Fitzpatrick (2003) reported that when sev-eral different crystal structures are superim-posed on a single PAH monomer, the differentmolecules demonstrate a remarkable conserva-tion of three-dimensional shape in the catalyticdomain for all members of the enzyme family.Moreover, the catalytic domains exhibit about80% amino acid sequence homology. As wouldbe expected by both their different tissue ex-pression and their biological functions, struc-tural differences between the regulatorydomains are significantly greater.

Erlandsen and Stevens (1999) used these dif-ferent overlapping structures to create a com-posite model of the intact enzyme. The ferriciron atom, located in a deep pit of the catalyticdomain, is coordinated by interactions with his-tidine residues, glutamic acid, water molecules,and the BH4 cofactor. Prior to catalysis, the fer-ric iron must be reduced to the ferrous state. Ithas been assumed that BH4 is the physiologicalreducing agent; however, this has yet to be con-firmed. Descriptions of the iron coordinationare variable and consist of reports of octahedralcoordination and distorted pentahedral coordi-nation, depending on which crystal structure isanalyzed, suggesting structural alterations arepossible on crystallization. More important, thisobservation complicates attempts at using crys-tal structure to propose or test a catalytic mech-anism for the enzyme because unique contactresidues and bond angles vary between crystalstructures. Moreover, until recently, crystalstructures of enzyme-substrate complexes havenot been available.

The location of the BH4 binding site and theamino acid contact to BH4 are very relevant toMrs. Urick’s questions about the effectivenessof dietary BH4 supplementation as a therapeu-tic intervention. This site has been located inseveral different structures and provides a sec-ond possible explanation for the large numberof putatively BH4-responsive individuals. Pre-sumably, if the cofactor binding site representsa mutational “hot spot,” then mutations toresidues in the BH4 binding site could lowerthe affinity for the cofactor, thus requiring anincreased concentration of the cofactor to sat-urate the binding site, resulting in a higher-than-expected frequency of BH4-responsive

Phenylketonuria 209

HN

N

N

O

H2N N

O

OH

H HH HOH

O P OO

OPO

OO P

OOH

O

HN

N

O

H2N NH

HN

CH3

OHHO

HN

N

O

H2N NH

HN

CH3

OO

N

N

O

H2N NH

NCH3

OHHOOH

N

N

O

H2N NH

NCH3

OHHO

HN

N

O

H2N

O

NH

NCH2

OHHO

P OO

OPO

OO P

OOH

O

HCOOH

2 NADPH

Phenylalanine2 NADP+

TyrosineO2

NADP+

NADPH

H2O

7,8-dihydroneopterin triphosphate

GTP cyclohydrolase

6-pyruvoyl-tetrahydropterin

6-pyruvoyl-tetrahydropterinsynthase

tetrahydrobiopterin

reductase

pterin 4α-carbinolamine


dihydrobiopterin

dehydratase

dihydrobiopterinreductase

Pi + PPi

GTP

Figure 19-3. De novo biopterin synthesis and recycling. Tetrahydrobiopterin can besynthesized de novo from GTP or resynthesized from the pterin 4α-carbinolamine.

210

enzymes. Mutations of the PAH enzyme havebeen intensively studied, and currently 400 dif-ferent mutations have been documented. Byidentifying single-nucleotide mutations andmapping the polymorphisms, the amino acidchange involved in the mutation, and the loca-tion of the mutation, one should be able topredict whether a particular patient might beresponsive to BH4 therapy. The database onthe mutations to PAH is extensive and rapidlygrowing.This database includes not only natu-ral mutations to the enzyme but also the re-sults of in vitro expression of laboratorymanipulations to the PAH gene. An in-depthdiscussion of these modifications is beyondthe scope of this chapter. Moreover, thesedata have been summarized (Pey et al., 2003;Teigen et al., 1999). One of the most up-to-datelistings of PAH mutations can be obtainedfrom the PAH Web site (http://www.pahdb.mcgill.ca).

The problem with a genetic approach to in-vestigating BH4 responsiveness is that (1) thereis no evidence that the BH4-binding site repre-sents a mutational hot spot, and (2) mutationsoutside the known BH4-binding site also havebeen shown to produce a BH4-responsive PAH(Bernegger and Blau, 2002; Steinfeld et al.,2001).The dissociation between the location ofthe mutation in the x-ray crystal structure and

BH4 responsiveness has led Dr. Blau to create afunctional database for the PAH molecule inwhich individuals can contribute to the func-tional mapping of genetic mutations (http://www.bh4.org).

Catalytic Mechanism

The mechanism by which elevated BH4 levelsenhance the catalytic effectiveness of a mutatedPAH, even when the mutation does not involvethe BH4-binding site, is unclear. The answer tothis problem will likely be found in a fuller un-derstanding of the enzyme mechanism itself.Understanding the catalytic mechanism of thePAH enzyme and how the dioxygen molecule iscleaved have proven to be difficult experimen-tal challenges. Kinetic investigations indicatedthat a ternary complex of enzyme, biopterin,and oxygen must be formed prior to the bind-ing of PA. In the active site of the enzyme, theoxygen molecule replaces one of the watermolecules and is coordinated between the ironatom and the BH4 cofactor. This complex, il-lustrated in Figure 19-6, effectively preventssubstrate access to the oxygen molecule andsuggests that a reactive oxygen molecule mustbe formed prior to substrate binding.Two dif-ferent species have been suggested as the ac-tive oxygen molecule. One possibility is an

Phenylketonuria 211

Figure 19-4. Differences between the frequency of phenylketonuria (PKU) and hyper-phenylalaninemia (HPA) caused by problems in tetrahydrobiopterin (BH4) metabolismand reports of positive therapeutic responses to BH4 therapy. (A) Frequency of PKU andHPA cases documented to be caused by defects in BH4 metabolism. ,Clinical cases docu-mented to be caused by defects in BH4; , clinical cases presumable due to a defect in thephenylalanine hydroxylase enzyme. (B) Frequency of PKU and HPA cases that have beenreported to respond positively to BH4 therapy. , Positive response to BH4 therapy; , noresponse to BH4 therapy.Different reports in the literature varied with respect to the num-bers of individuals responding to BH4 therapy. The differences in reported numbers ofBH4 responders are indicated by the boxes with cross-hatching.

http://www.bh4.org

http://www.bh4.org

http://www.pahdb.mcgill.ca

http://www.pahdb.mcgill.ca

active Fe(IV)-oxygen complex (Bugg, 2001;Klinman, 2001; Solomon et al., 2000); the othersuggestion requires the biopterin cofactor to beconverted into a 4-hydroperoxide (Eberlein etal, 1984). Evidence supporting both hypothesesexist, but an in-depth evaluation of that evi-dence is beyond the scope of the present dis-cussion. These data have been reviewed byseveral investigators (Bassan et al., 2003; Fitz-patrick, 2003), who concurred that that theFe(IV)-oxygen intermediate is the most likelyoxidizing species.

The difficulty in obtaining crystal structuresfrom active substrate enzyme complexes hassignificantly hampered the ability to confirm thehypotheses about the mechanism of PAH ac-tion.This problem was solved by Andersen andcoworkers (2002) using 3-(2-thienyl)-L-alanine asa substrate analogue. These authors reportedthat, during catalysis, the enzyme undergoesmajor structural changes as part of the activa-tion of the dioxygen molecule. They also re-ported that a tyrosine residue (Tyr 138) thoughtto be exposed on the surface of the enzyme ismoved to the interior and may play an impor-tant, but as-yet-unrecognized, role in the activeenzyme-substrate complex.

These new findings also raise questions aboutthe assignment of residues to roles in catalysisand suggest that the mutation databases should

be reevaluated in the light of these dramatic con-formational changes.The possibility of a changein global enzyme conformation playing a rolein the catalytic mechanism also provides anexplanation for the observations related to thesuccessful use of high levels of BH4 in patienttreatment (Erlandsen and Stevens, 2001).Theseinvestigators argued that a catalytic mecha-nism involving major structural changes couldbe dramatically influenced by cofactor levels.Mutations that cause slow or delayed foldingof the enzyme would be cleared and degradedby the proteosome system. However, if high lev-els of BH4 cofactor were present during thefolding, then the interaction with enzyme mightstabilize the enzyme or enhance proper fold-ing, which might explain the observation thata BH4 effect is found in more than just a fewpatients.

In the past two years the number of reportedBH4 responsive cases of HPA and classical PKUhas increased dramatically, suggesting that theBH4 response rate may be far greater than ini-tially observed. Moreover, BH4 responses haveeven been reported in classical PKU (Matalonet al., 2005). In response to this changing view


Figure 19-6. Hypothetical active site complex ofphenylalanine hydroxylase. The nonheme ironatom of phenylalanine hydroxylase is coordinatedwith protein amino acid residues (top of the fig-ure) and molecular oxygen. The BH4 cofactor isalso coordinated with molecular oxygen (bottomof the figure). It has been suggested that this com-plex decomposes to either an iron-oxygen com-plex or a biopterin-4 hydroperoxide complexprior to interaction with the amino acid sub-strate.

Figure 19-5. Phenylalanine hydroxylase struc-ture. Ribbon structure of the catalytic and regula-tory domain of the rat phenylalanine hydroxylaseenzyme (PDB access number 1PHZ) (Kobe et al.,1999). The catalytic domain is on the left and anonheme iron atom is present in the center of thedomain. The regulatory domain is shown on theleft.

about PKU, in addition to the holding the “FirstInternational Conference on the Role of BH4

and Phenylketonuria Treatment” in 2005, thejournal, Molecular Genetics and Metabolism,devoted an entire supplement to “New De-velopments in Phenylketonuria and Tetrahy-drobiopterin Research” (Kock et al., 2005).Although the mechanism of the BH4 effect re-mains elusive, evidence is accumulating that inaddition to an elevated Km for BH4 by thephenylalanine hydroxylase enzyme, BH4: in-creases enzyme stability, protects against pro-teolytic degradation, has chaperon-like activity,alters regulation of BH4 synthesis, induces PAHexpression and the stability of its mRNA (Er-landsen et al., 2004; Blau and Erlandsen, 2004).The altered view about the potential role ofBH4 therapy in PHA and classic PKU has led tothe suggestion that all cases of HPA and PKUbe evaluated using the BH4 loading test (Fiegeet al., 2005).To date the only contraindicationfor BH4 therapy appears to be the cost of thecofactor.

THERAPY

Testing

Unlike most other genetic diseases, PKU isunique in that, once diagnosed, it can be suc-cessfully treated. Moreover, while medical inter-vention is not inexpensive, it is far morecost-effective than treating the manifestations ofthe disease itself (NIH Consensus Statement,2000). For these reasons, screening for HPA ispublic policy in the United States and manyother countries. In patients with HPA,plasma PAlevels are typically normal at birth but quicklyrise after the initiation of protein feedings.

Screening must take place within 3 weeks af-ter birth to prevent mental retardation.Blood PAlevels are measured using the Guthrie bacterialinhibition assay. Elevated values are then con-firmed using quantitative amino acid analysis. Ifscreening reveals HPA (plasma levels persistent-ly greater than 0.125 mM), then investigation ofBH4 homeostasis must be initiated to rule out di-hydropteridine reductase deficiency. Deficiencyof dihydropteridine reductase is confirmed bydirect enzyme measurement in dried blood sam-ples (Dhondt, 2002). Dihydrobiopteridine defi-ciency and subsequent block in BH4 synthesiscan also be detected in utero via assays on cul-tured amniocytes.

If BH4 levels are normal, then HPA due to aproblem with PAH becomes established by ex-clusion. About half of these patients will havePKU, with PA levels above 1 mM, and half willhave non-PKU HPA, with PA levels between0.125 and 1.0 mM.While not widespread as yet,DNA analysis is becoming more common andclinically relevant (NIH Consensus StatementOnline, 2000; Scriver, 2000).

Treatment

As discussed in the preceding section, of all ge-netic diseases, screening for PKU is importantbecause the manifestations of the disease canbe prevented with proper medical interven-tion. Currently, the primary treatment of pa-tients with HPA is dietary restriction of PA,which was first introduced by Bickel in 1953.Normal PA intake is more than 1000 mg per day,but in patients with PKU, only 250–500 mg/dayPA can be tolerated.There are a variety of semi-synthetic diets currently available that are de-void of or low in PA.

It is important to note that, while PA intakemust be significantly reduced in affected individ-uals,PA is still an essential amino acid and cannotbe completely eliminated from the diet.As such,growth rate and PA levels need to be monitoredregularly. Unfortunately, as clearly identified byMrs. Urick, these diets are often expensive anduniversally unpalatable. As such, compliance isfrequently a clinical issue (Bodamer, 2003;Scriver, 2000). In this case, the physician is leftwith several questions: Was it appropriate forMrs. Urick to discontinue her diet? What are theimplications, if any, for her unborn baby? and Willshe or her baby be a candidate for BH4 therapy?

Unfortunately, few clinical studies have inves-tigated the efficacy and timing of specific di-etary intervention with patients who have PKU.In fact, there are no randomized clinical trialscomparing treatment from diagnosis versus notreatment at all. On the other hand, evidencefrom nonrandomized studies overwhelminglydemonstrated that implementation of dietaryrestriction of PA early in life can significantlydecrease the mental retardation and neurologi-cal impairments associated with PKU.While theneed for intervention via a PA-restrictive diet isclear, some questions still remain regarding ex-actly how early to begin treatment, which di-etary supplements should be added to PArestriction, what degree of metabolic control

Phenylketonuria 213

should be achieved, and when, if ever, the dietshould be discontinued (Koch, 2002; NIH Con-sensus Online, 2000; Poustie and Rutherford,2004).

Current evidence suggests dietary restrictionshould begin as soon as possible, preferablywithin the first week of life (NIH ConsensusStatement Online, 2000), but certainly withinthe first 20 days after birth (Poustie and Ruther-ford, 2004).There is considerable evidence thatearlier initiation of dietary restriction directlycorresponds to higher IQ levels, and elevatedPA levels even in the first 2 weeks of life may af-fect development of the visual system (NIHConsensus Statement Online, 2000).

Questions arise when discussing the neces-sary degree of control of plasma PA levels.While results from a variety of studies varywidely and there is no consensus what the opti-mal plasma PA level should be, current best evi-dence suggests that tighter metabolic control(low PA levels) correlates directly with im-provements in IQ, motor skills, attention span,behavioral skills, and cognitive ability. In mostof these areas, effects of lower PA levels havebeneficial effects throughout the life of the pa-tient (NIH Consensus Statement Online, 2000).

As discussed, PA restriction diets are poorlypalatable, difficult to follow, and expensive, thusmaking compliance difficult.These diets also re-quire regular support from specialists and mayresult in inappropriate eating behaviors. In ad-dition, proper nutrition can be difficult to main-tain. Adding to the problem, physicians havebeen giving patients inconsistent advice on theappropriate length of time the diet should bemaintained (Poustie and Rutherford, 2004).

As such, and as illustrated by Mrs. Urick’s ex-perience, reduced compliance often begins inmidchildhood. By late adolescence, a significantnumber of affected individuals have discontin-ued the diet. While screening and interventionfor patients with PKU have been widely accept-able and established as public policy for over40 years, there is considerable uncertaintyabout when, if ever, the diet should be discon-tinued and what other dietary supplementsshould be included. It seems clear that discon-tinuing the diet prior to 6 years of age has anegative effect on performance on IQ tests(NIH Consensus 2000; Poustie and Rutherford,2004). However, it is less clear that discontinua-tion at later ages has adverse consequencesand, if so, to what degree.

There are some data suggesting that continu-

ing the diet as long as possible may be benefi-cial to long-term neurological health. First, up-per motor neuron disturbances have beendocumented in older children and adults whohave returned to a normal diet. In addition, ele-vated PA levels in these individuals can lead towhite matter changes in the brain, as demon-strated by magnetic resonance imaging (Poustieand Rutherford, 2004). Adults who discontinuePA restriction do not usually exhibit poorer per-formance on IQ tests,but they do not do as wellon measures of attention span or speed of pro-cessing. Surveys of physicians and evidencefrom several studies suggested that, if possible,the PA restrictive diet should be maintainedthroughout life for best neurological outcomein the individual. However, there is some evi-dence that metabolic control may be able to berelaxed later in life (NIH Consensus StatementOnline, 2000).

Supplementation for patients who are re-stricting PA intake varies with the particulardiet that patient is receiving. Many patients re-quire vitamins and other supplements alongwith other essential amino acids that may bedeficient in the restricted diets (Poustie andRutherford, 2004). Several studies have investi-gated tyrosine supplementation. Given that in-dividuals with PKU suffer from both an excessof PA and a deficiency of tyrosine, one wouldexpect that tyrosine supplementation might behelpful in addressing the manifestations of thisdisease. However, there have yet to be shownany significant differences between patientstaking tyrosine supplements and those who arenot, other than an increase in plasma tyrosine.Due to the small number of individuals lookedat in these studies, a definitive answer to tyro-sine supplementation cannot be reached withcertainty as yet, and a large multicenter random-ized clinical trial would be necessary to do so.

EPILOGUE

Mrs.Urick immediately began following a restric-tive diet and successfully maintained plasma PAlevels below 0.9 mM. She reported that the dietwas hard to follow, but it was well worth it tohave a healthy child. She was also very inter-ested in exploring BH4 therapy. While she wasunable to enroll in a clinical evaluation of BH4

treatment during her pregnancy, she main-tained good plasma PA levels throughout gesta-tion, carried the baby to term, and delivered a


healthy baby girl, who was screened for PKUand found not to have the disease. After thebirth of her baby, Mrs. Urick decided to try tocontinue the restrictive diet “at least for awhile.” She also expressed interest in findingout more about BH4 supplementation and otherpotential treatments.

QUESTIONS

1. How would you justify the observationthat that two different patients possessingthe same mutation in the PAH enzymemay exhibit very different phenotypeswith respect to PA metabolism?

2. How can you explain the observation thatin some individuals with a mutation to thePAH enzyme outside the BH4-binding siteshow a decrease in serum PA levels whengiven dietary BH4?

3. Although mutations to dihydrobiopterinreductase and PAH are equally likely to befound in the population, how can you ac-count for the observation that mutationsto dihydrobiopterin reductase accountsfor only about 2% of patients with PKUand HPA?

4. What would you propose as the least-expensive and most rapid method toidentify the existence of a defect in BH4

metabolism as the cause of a patient’sPKU or HPA?

5. Based on metabolic pathways, what phe-notypic differences would you expect toobserve between individuals sufferingfrom either a defect in BH4 metabolism ora defect in the PAH enzyme?

6. Mrs. Urick was initially counseled that asan adult she did not need to follow thestrict PKU diet. How do you respond tothis advice? Based on PA metabolism, howdo you justify your response?

7. Based in biochemical and physiologicalmechanisms, how can you explain im-paired brain development as a conse-quence of uncontrolled PKU?

BIBLIOGRAPHY

Andersen OA, Flatmark T, Hough E: Crystal struc-ture of the ternary complex of the catalytic do-main of human phenylalanine hydroxylase withtetrahydrobiopterin and 3-(2-thienyl)-L-alanine,and its implications for the mechanism of catal-

ysis and substrate activation. J Mol Biol 320:1095–1108, 2002.

Bassan A, Blomberg MRA, Siegbahn PEM: Mecha-nism of dioxygen cleavage in tetrahydrobiopterin-dependent amino acid hydroxylases. Chem Eur J9:106–115,2003.

Bernegger C, Blau N: High frequency of tetrahy-drobiopterin-responsiveness among hyperpheny-lalaninemias: a study of 1919 patients observedfrom 1988 to 2002. Mol Genet Metab 77:304–313, 2002.

Bickel H, Gerrard J, Hickmans EM: Influence ofphenylalanine intake on phenylketonuria.Lancet 265:812–813, 1953.

Blau N, Erlandsen H:The metabolic and molecularbases of tetrahydrobiopterin-responsive pheny-lalanine hydroxylase deficiency. Mol. Genet.Metab. 82:101–111, 2004.

Bugg TDH: Oxygenases: mechanisms and struc-tural motifs for O2 activation. Curr Opin ChemBiol 5:550–555, 2001.

Dhondt JL: Screening of tetrahydrobiopterin defi-ciency among hyperphenylalanenic patients.Ann Biol Clin Paris 60:165–71, 2002.

Eberlein GA, Bruice TC, Lazarus RA, et al.:The in-terconversion of the 5,6,7,8-tetrahydro-, 6,7,8-dihydro, and radical forms of 6,6,7,7-tetramethyldihydropterin. A model for thebiopterin center of aromatic amino acid mixedfunction oxidases. J Am Chem Soc 106:7916–7924, 1984.

Erlandsen H, Pey AL, Gámez A, et al.: Correction ofkinetic and stability defects by tetrahydro-biopterin in phenylketonuria patients with cer-tain phenylalanine hydroxylase mutations.Proc. Natl. Acad. Sci. (USA) 101:16903–16908,2004.

Erlandsen H, Stevens RC: The structural basis ofphenylketonuria. Mol Genet Metab 68:103–125, 1999.

Erlandsen H, Stevens RC: A structural hypothesisfor BH4 responsiveness in patients with mildforms of hyperphenylalaninaemia. J InheritMetab Dis 24:213–230, 2001.

Fiege B, Bonafé L, Ballhausen D, et al.: Extendedtetrahydrobiopterin loading test in the diagno-sis of cofactor-responsive phenylketonuria: Apilot study. Mol Genet Metab 86:s91–s95,2005.

Fitzpatrick PF: Mechanism of aromatic amino acidhydroxylation. Biochemistry 42:14083–14091,2003.

Klinman JP: Life as aerobes: are there simple rulesfor activation of dioxygen by enzymes? J BiolInorg Chem 6:1–13, 2001.

Kobe B, Jennings IG, House CM, et al.: Structuralbasis of autoregulation of phenylalanine hydrox-ylase. Nat Struct Biol 6:442, 1999.

Koch R, Burton B, Hoganson B, et al.: Phenylke-tonuria in adulthood: a collaborative study. J In-herit Metab Dis 25:333–346, 2002.

Phenylketonuria 215

Kock R,Matalon R,Stevens RC (eds):New develop-ments in phenylketonuria and tetrahydro-biopterin research. Mol Genet Metab V86(Supplement 1):1–156, 2005.

Longo N:“Inherited Disorders of Amino Acid Me-tabolism and Storage.” Harrison’s Principles ofInternal Medicine. Eugene Braunwald (ed).15th Edition. New York, NY. The McGraw-HillCompanies, Inc. 2001. Chap 352: 2301–2309.

Matalon R, Michals-Matalon K, Koch R, et al.: Re-sponse of patients with phenylketonuria in theU.S. to tetrahydrobiopterin. Mol Genet Metab86:s17–s21, 2005.

NIH Consensus Statement Online: Phenylke-tonuria: screening and management. October16–18; 17:1–27, 2000. http://consensus.nih.gov/2000/2000phenylketonuria113html.htm.Accessed 4/5/06.

Pey AL,Desviat LR,Gámez A,et al.:Phenylketonuria:genotype-phenotype correlations based on ex-pression analysis of structural and functional mu-tations in PAH.Hum Mutat 21:370–378,2003.

Poustie VJ, Rutherford P: Dietary interventions forphenylketonuria, Cochrane Review, in: TheCochrane Library, Issue 1. Wiley, Chichester,

UK, 2004. http://www.mrw.interscience.wiley.com/cochrane/clsysrev/articles/cd001304/frame.html.Accessed 4/5/06.

Scriver CR: “The Hyperphenylalaninemias andAlkoptonuria.” Cecil Textbook of Medicine. Vol.1. Ed. Lee Goldman and J. Claude Bennette. 21stedition. Philadelphia, PA, W.B. Saunders Com-pany, 2000. Chap 209: 1108–1110.

Solomon EI, Brunold TC, Davis MI, et al.: Geomet-ric and electronic structure/function correla-tions in non-heme iron enzymes. Chem Rev100:235–349, 2000.

Steinfeld R, Kohlschutter A, Zschocke J, et al.:Tetrahydrobiopterin-responsiveness associatedwith common phenylalanine hydroxylase muta-tions distant from the tetrahydrobiopterin bind-ing site. J Inherit Metab Dis 24:29, 2001.

Teigen K, Frøystein NÅ, Martínez A:The structuralbasis of the recognition of phenylalanine andpterin cofactors by phenylalanine hydroxylase:implications for the catalytic mechanism. J MolBiol 294:807–823, 1999.

Trefz FK, Blau N: Potential role of tetrahydro-biopterin in the treatment of maternal phenylke-tonuria.Pediatrics 112:1566–1569, 2003.


http://consensus.nih.gov/2000/2000phenylketonuria113html.htm

http://consensus.nih.gov/2000/2000phenylketonuria113html.htm

http://www.mrw.interscience.wiley.com/cochrane/clsysrev/articles/cd001304/frame.html



CASE REPORT

D.E.C. was a normal female infant (weight 6 lb4 oz, length 20 inches) born via spontaneousvaginal delivery to an 18-year-old primigravida(first pregnancy). The pregnancy was un-planned and complicated only by maternal useof antidepressants (Zoloft and Depakote) dur-ing the first months of gestation. When D.E.C.was 1 month of age, she appeared to havemacrocephaly and mild hypotonia; a cranial sec-tor scan was performed and was normal. Shewas initially breast-fed but was a poor feeder,had gastroesophageal reflux, and gained weightslowly (Fig. 20-1). She sweated profusely whensleeping, and at 4 months of age had hypotoniaand poor head control.

She was initially hospitalized at 4 months ofage during a family vacation. An initial viral ill-ness with low-grade fever and vomiting led tolethargy within 12 h.When she developed gener-alized seizures, she was taken to the emergencyroom and admitted. She experienced respiratorydecompensation, was intubated, and requiredventilator support for 72 h. Laboratory studiesdemonstrated marked metabolic acidosis withhypoglycemia (45 mg/dL; normal >80 mg/dL)and hyperammonemia (plasma ammonium>1100 µmol/L;normal 50–200 µmol/L).Her liverenzymes were markedly elevated. Her urinecontained a number of organic acids, including3-hydroxyisovaleric acid, 3-methylglutaconicacid, and 3-hydroxy-3-methylglutaric acid, all by-products of leucine catabolism. Plasma aminoacid values also suggested a problem inbranched-chain amino acid metabolism. Studieswith skin fibroblast cultures confirmed a defi-

ciency of 3-hydroxy-3-methylglutaryl (HMG)-CoA-lyase deficiency with zero enzyme activity.

Although leucine is an essential amino acid,it was necessary to restrict the diet of this pa-tient to about half of the leucine intake of nor-mal healthy infants. She was therefore placedon a low-leucine formula consisting of threecomponents: Similac with iron (a common in-fant formula), I-Valex 1 (a medical food that is aleucine-free, carnitine-enriched formula), andPolycose powder (glucose polymers; Ross Labo-ratories) to provide additional carbohydratecalories. The initial formula was designed toprovide her with age-appropriate infant nutri-tion of 150 mL/kg, 130 Kcal/kg, 2 g protein/kgwhile limiting leucine to 121 mg/kg, the mini-mum requirement to support normal growth.After a 10-day hospitalization, she was dis-charged taking the formula mixture, a multivita-min, and L-carnitine supplement. Due to poororal intake, she required supplemental nasogas-tric tube feeding.

Physical exam at 5 months of age revealed afrail-appearing small girl with nasogastric tubein place (Fig. 20-2). Her weight was significantlybelow the fifth centile, and neurological examrevealed truncal hypotonia and focal increaseddeep tendon reflexes.The neurological findingssuggested brain injury secondary to her initialmetabolic insult. She laughed and smiled buthad poor head control and could not lift herhead from prone.

By 10 months of age, she had required sev-eral hospitalizations for viral illnesses with milddehydration, hypoglycemia, and mild acidosis.She also had several brief episodes of right-sided focal seizures lasting as long as 10 s.As a

20

HMG-CoA Lyase Deficiency

VIRGINIA K. PROUD and MIRIAM D. ROSENTHAL

217

Figure 20-1. Growth chart, weight for length, for D.E.C. Note poor weight for length at 5months before placement of gastrostomy tube (first arrow) and increased weight for lengthby 2 years 6 months,when she was admitted for metabolic studies and feeding therapies.

218

result of her chronic organic acidemia, reflux,frequent hospitalizations, hypotonia, and devel-opmental delay, she developed oral aversionand poor feeding.There was constant concernabout hypoglycemia, and at one clinic visit shehad a blood sugar of 37 mg/dL (normal >80 mg/dL).A permanent gastrostomy tube (G-tube) wasplaced at 6 months of age, and a Nissen fun-doplication was performed to treat gastroe-sophageal reflux. After the G-tube placement,metabolic balance was achieved more easily,with no further hospitalizations for dehydra-tion, acidosis, or hypoglycemia.

By 2.5 years of age, her weight was greaterthan the 95th centile, but she was taking verylittle food by mouth (Fig. 20-1).Although it wasnecessary to decrease her caloric intake to al-low her to lose weight and to stimulate her ap-petite, there was concern that she mightdevelop hypoglycemia. She was therefore ad-mitted to the special feeding unit at the Chil-dren’s Hospital to monitor her metabolic statuswhile her gastrostomy-tube intake was de-creased and a team of feeding specialists, nutri-tionists, and occupational therapists worked

with her to improve her oral intake. Duringthe 6-week hospitalization, a fasting study con-firmed that she was stable for up to 14 h with-out hypoglycemia; this probably reflectedimproved availability and adequacy of glycogenstores with liver growth and maturation. Withcontinued feeding therapy, all of her caloric in-take was by mouth by 4 years of age, and by 6years of age approximately 75% of her calorieswere from foods other than the medical for-mula.Her appetite remained poor,however, andmealtimes often extended over an hour.

At 7 years of age, she had a completely nor-mal physical examination. Her weight was22.1 kg (60th centile), height was 123.4 cm(60th centile), and head circumference was51 cm (40th centile). The neurological examwas normal without focal findings. She did wellin first grade and led a normal life in all areaswith the exception of her diet. Although hereating was more mechanical than natural, herappetite had improved, and she was eating a va-riety of foods. She still could not eat meat, eggs,or other high-protein foods, and her family muststill measure and weigh all foods to determine

HMG-CoA LYASE DEFICIENCY 219

Figure 20-2. Photograph of D.E.C. at 5 months. Note nasogastric feeding tube andlistless expression.

leucine content. Her mealtimes had decreasedto only 30 min. She drank her medical foodmixed with juice and must adhere to a consis-tent eating schedule to avoid fasting.The med-ical team continues to monitor her biochemicalparameters by analyzing plasma amino acidsand modifies her diet to optimize growth.

DIAGNOSIS

Organic acidurias comprise a unique group ofdisorders of amino acid and fatty acid metabo-lism characterized by a predisposition formetabolic decompensation. Several of these dis-orders, including methylmalonic aciduria, pro-pionic aciduria, and isovaleric aciduria, canhave similar clinical presentations. However,qualitative examination of urinary organic acidsdemonstrates a unique profile for each disease.In children with HMG-CoA lyase deficiency, theurine organic acids include 3-hydroxyisovalericacid, 3-methylglutaric acid, and 3-hydroxy-3-methylglutaric acid. In addition, HMG-CoA lyasedeficiency is unique among organic acidurias inthat hypoglycemia is not associated with in-creased urinary levels of the so-called ketonebodies, acetoacetate and β-hydroxybutyrate.

The clinical presentation of HMG-CoA lyasedeficiency depends on the amount of residualenzyme activity as well as the specific illness ormetabolic stress that precipitates the clinicalproblems. There may be acute neonatal meta-bolic acidosis with hypoglycemia, hypotonia,and seizures, or there may be a more indolentcourse with several months of hypotonia,devel-opmental delay, and failure to thrive. Older in-fants, between 6 months and 2 years, maypresent with profound metabolic acidosis, lac-tic acidosis, hypoglycemia, and hyperammone-mia.The rapid-onset metabolic decompensationoften results from an acute viral illness, espe-cially if associated with fever, vomiting and di-arrhea, and dehydration, but can also beprecipitated by a fever secondary to routine im-munizations.The clinical presentation often in-cludes hepatomegaly and shock. If there is achange in neurological status, especially withlethargy, then diagnostic studies should includeblood gas to determine degree of metabolic aci-dosis and compensatory respiratory alkalosis aswell as plasma levels of glucose, lactic acid, am-monia, amino acids, carnitine (total and free),and urinary organic acids. In older children,MRI of the brain may also show changes in the

white matter caused by poor myelinization,which with recurrent episodes of metabolic in-sult can progress to cortical atrophy, seizures,and neurological deterioration.

Many states have expanded newborn meta-bolic screening programs and are using tandemmass spectrometry to identify as many as 60amino acid and fatty acid oxidation defects, in-cluding phenylketonuria,maple syrup urine dis-ease, and homocystinuria, from the analysis ofone small blood sample. An elevated plasmalevel of leucine is indicative of an inborn errorin leucine catabolism. Further diagnostic testsare then required to characterize the specificenzyme deficiency. More widespread use of thistechnology will permit identification of manyinfants and children before acute metabolic de-compensation occurs, thus preserving optimalneurological and cognitive function.


HMG-CoA lyase deficiency was first reportedby Faull and co-workers (1976). The patient, a7-month-old infant, presented with diarrhea,vomiting, dehydration, metabolic acidosis, hy-poglycemia, and organic aciduria.Analysis of uri-nary metabolites by gas-liquid chromatography/mass spectrophotometry identified high concen-trations of 3-hydroxylisovaleric,3-methylglutaric,3-methylglutaconic, and 3-hydroxyl-3-methylglu-taric acids.This profile differed from that seen inpreviously described diseases related to leucinecatabolism (including maple syrup urine dis-ease) and suggested a deficiency in the cleavageof HMG-CoA to acetoacetate and acetyl-CoA.

HMG-CoA lyase is normally present in the mi-tochondrial matrix.To understand the complex-ity of the metabolic problems of a patient withHMG-CoA lyase deficiency, it is necessary toconsider the role of this enzyme in two verydistinct metabolic pathways: catabolism ofleucine and ketogenesis.

Catabolism of Leucine

Leucine is a branched chain-amino acid that isessential or required in the diet. Mitochondrialcatabolism of excess leucine occurs by thepathway shown in Figure 20-3. The initialtransamination step (removal of the aminogroup) is followed by a decarboxylation reac-tion to produce isovaleric acid. It is this decar-boxylation of the α-keto analogs of the three


branched-chain amino acids leucine, isoleucine,and valine that is deficient in maple syrup urinedisease. Catabolism of isovaleric acid ultimatelyproduces HMG-CoA, which, as discussed, iscleaved by HMG-CoA lyase to produce acetoac-etate and acetyl-CoA.Subsequent metabolism ofacetoacetate involves an acetoacetyl-CoA inter-mediate that is cleaved to form two more mole-cules of acetyl-CoA.

Individuals who are deficient in HMG-CoAlyase are unable to complete the metabolism ofleucine. The increased urinary excretion of 3-hydroxy-3-methylglutaric acids is the primarybiochemical criterion that distinguishes thisparticular enzymatic defect from other defectsin enzymes of leucine catabolism that also re-sult in metabolic acidosis and abnormal or-ganic aciduria.There is also substantial urinaryexcretion of intermediates of leucine catabo-lism, such as 3-methylglutaconic acid, and theirmetabolites, including 3-hydroxy-isovaleric acidproduced from isovaleric acid.

Ketogenesis

Ketogenesis or “ketone body” formation oc-curs in the liver and, to some extent, renal cor-tex during the fasted state and is an obligatorypathway for energy metabolism during fast-ing. Under these conditions, the hormoneglucagon stimulates hepatocytes to maintainblood sugar levels both by mobilizing glucosefrom liver glycogen stores and by gluconeo-genesis. The major substrates for gluconeo-genesis are amino acids such as alanine andglutamine. As shown in Figure 20-4, oxaloac-etate is a key intermediate in gluconeogenesis.Oxaloacetate also plays a key role in the tricar-boxylic acid (TCA) cycle, in which it is com-bined with the entering acetyl-CoA to formcitrate. When gluconeogenesis occurs, diver-sion of oxaloacetate to glucose synthesis de-pletes the TCA cycle of this key intermediate,resulting in shutdown of the oxidation of acetyl-CoA to CO2 and the ATP generation linked to re-ducing equivalents generated through the TCAcycle.

Utilization of amino acids for gluconeogene-sis requires initial removal of their aminogroups, usually by transamination linked tosubsequent glutamate dehydrogenase activity.The resultant NH4

+ is detoxified by urea syn-thesis.

Both gluconeogenesis and urea synthesis re-quire net energy input as ATP. In the fasting


Figure 20-3. Pathway for the catabolism ofleucine. The last step in the pathway is catalyzedby 3-hydroxy-3methyl-CoA lyase, which is deficientin 3-hydroxy-3-methylglutaryl-CoA lyase deficiency.

state, the liver relies on β-oxidation of fattyacids for synthesis of ATP.The reducing equiva-lents (NADH and FADH2) generated duringeach cycle of β-oxidation can flow directly intothe electron transport chain, independent ofthe TCA cycle, and generate ATP via oxidativephosphorylation. There is, however, one prob-lem:What do liver cells do with the acetyl-CoAgenerated by β-oxidation of fatty acids? The so-lution is the synthesis of so-called ketone bod-ies. This process (shown in Fig.20-5) essentiallyconverts two molecules of acetyl-CoA to onefour-carbon acetoacetate, with release of thetwo molecules of reduced coenzyme A (CoA-SH). HMG-CoA, the six-carbon intermediate inthis pathway, is formed by successive condensa-tions of three acetyl-CoA molecules. HMG-CoAlyase then cleaves the HMG-CoA to releaseacetyl-CoA from acetoacetate.

Ketogenesis is thus an alternate pathway inwhich the partially oxidized intermediates offatty acid oxidation are exported from the liver

as acetoacetate (and β-hydroxybutyrate). Theacetoacetate can then be further oxidized bytissues such as muscle that are not gluco-neogenic and that retain an active TCA cycle.

An individual who is deficient in HMG-CoAlyase cannot synthesize acetoacetate from HMG-CoA; in the absence of a pool of free CoA-SHmolecules, β-oxidation of fatty acids cannotcontinue in the liver. This often causes he-patomegaly (liver enlargement) with diffuseaccumulation of lipid droplets in swollen hepa-tocytes. Lack of sufficient ATP generation impairsurea synthesis and results in hyperammone-mia. Most important, in the absence of ongoingATP generation, gluconeogenesis cannot occur.Stores of liver glycogen, usually more limited inneonates and young children, are thus depletedmore rapidly than normal.The resultant fastinghypoglycemia often precipitates a metaboliccrisis.

For most patients with HMG-CoA lyase defi-ciency, strict dietary control can maintain


Figure 20-4. Biochemical pathways for gluconeogenesis in the liver. Alanine, a majorgluconeogenic substrate, is used to synthesize oxaloacetate. The carbon skeletons ofglutamine and other glucogenic amino acids feed into the TCA cycle as α-ketoglutarate,succinyl-CoA, fumarate, or oxaloacetate and thus also provide oxaloacetate. Conversionof oxaloacetate to phosphoenolpyruvate and ultimately to glucose limits the availabilityof oxaloacetate for citrate synthesis and thus greatly diminishes flux through the initialsteps of the TCA cycle (dashed lines). Concurrent β-oxidation of fatty acids providesreducing equivalents (NADH and FADH2) for oxidative phosphorylation but results inaccumulation of acetyl-CoA.

imize the need for gluconeogenesis, particu-larly in neonates and young children, whoseliver glycogen stores are often less adequate tomeet long-term needs.As described, metabolicdecompensation often occurs during illnesses,such as viral infections, when decreased di-etary intake and fever result in dehydrationand endogenous protein catabolism. In normalindividuals, the endogenous protein catabo-lism provides amino acid substrates for synthesisof immunoglobulins and acute phase proteins aswell as for gluconeogenesis to meet the in-creased needs of the stress response. However,individuals with HMG-CoA lyase deficiencycannot tolerate the increased mobilization ofleucine from muscle proteins and will quicklydecompensate and develop profound meta-bolic acidosis.

Leucine is a ketogenic amino acid with a car-bon skeleton that cannot serve as a substrate forgluconeogenesis. Normally, much of the leucine(and other branched-chain amino acids) mobi-lized during fasting is actually not released fromthe muscle. Instead, the amino groups are re-moved by transamination and exported from themuscle as alanine or glutamine; the resultantbranched α-keto acyl-CoA molecules are metabo-lized within the muscle for energy. In individualswho are deficient in HMG-CoA lyase, increasedleucine mobilization during the catabolic stateinduced by illness results in increased accumula-tion of 3-hydroxy-3-methylglutaric acid andother organic acids.

Role of Carnitine

Carnitine deficiency complicates HMG-CoAlyase deficiency and other inborn errors of me-tabolism, which results in organic acidemia.L-Carnitine or β-hydroxy-γ-trimethylammoniumbutyrate is a carrier molecule that transportslong-chain fatty acids across the inner mito-chondrial membrane for subsequent β-oxi-dation. L-Carnitine also facilitates removal oftoxic metabolic intermediates or xenobioticsvia urinary excretion of their acyl carnitinederivatives. Indeed, individuals with HMG-CoAlyase deficiency have been shown to excrete3-methylgluatarylcarnitine (Roe et al., 1986).In the absence of ketogenesis, the formationof the acyl carnitine derivative of 3-hydroxy-3-methylglutarate from HMG-CoA also servesto regenerate free CoA in the mitochondriaand permits continued β-oxidation of fattyacids.


Figure 20-5. Biochemical pathway for ketogene-sis.Condensation of three molecules of acetyl-CoAresults in synthesis of 3-hydroxy-3-methyl-glutaryl(HMG)-CoA and release of free CoA-SH, which isthen available for continued β-oxidation. Defi-ciency of HMG-CoA lyase results in the inability tocleave the HMG-CoA to form acetoacetate (the ini-tial “ketone body”) and its reduction product β-hydroxybutyrate.

homeostasis and prevent metabolic decompen-sation while providing adequate leucine intakefor growth. A regimen of frequent feeding andsufficient carbohydrate intake is utilized to min-

Individuals with HMG-CoA lyase deficiencyare particularly susceptible to carnitine defi-ciency. With restriction of red meats and dairyproducts, dietary carnitine intake is quite low.Carnitine is also synthesized endogenouslyfrom the modified, methylated lysine resides ofvarious proteins; free trimethyllysine is releasedwhen the protein is degraded.Since the therapyfor patients with HMG-CoA lyase deficiencymust minimize their endogenous protein catab-olism, they also have limited availability oftrimethyllysine for carnitine synthesis.

THERAPY

Nutritional therapy for HMG-CoA lyase defi-ciency has two major goals. First, the prescribeddiet aims to provide enough total protein andcalories to achieve normal growth and maintainmetabolic balance in the context of a leucine-restricted diet. Equally important, the nutritionaltherapy focuses on preventing excess catabo-lism, acidosis, and hypoglycemia, especially dur-ing times of acute illness. For these patients, itis particularly important to avoid fasting at anytime.

Monitoring plasma amino acids and growthon a regular basis in a metabolic clinic is nec-essary to ensure adequate leucine and totalprotein intake. Leucine tolerance varies widelyfrom patient to patient. Furthermore, theleucine needs of each patient will change as aresult of growth spurts, age, illness, and ade-quacy of total protein and energy intakes.Leucine deficiency can cause poor weightgain, hair loss, skin rash, loss of appetite, and ir-ritability. Excess leucine intake, on the otherhand, will lead to the accumulation of organicacids, metabolic acidosis, and potentially toacute decompensation. Plasma samples shouldbe drawn 2–4 h postprandially (after a meal)and the diet adjusted to maintain leucine lev-els at the low end of normal (50–100 µmol/L).For the younger child, routine infant formula isused to supply the prescribed amount ofleucine in the form of natural protein frommilk.The medical food ( I Valex 1) provides ad-ditional protein in the form of individualamino acids, but it is totally deficient inleucine. It also provides additional fluid, fat,carbohydrate, vitamins, and minerals. If con-sumed as prescribed, then the final formula,which is a combination of natural protein and

medical food, will meet the nutritional needsof the growing child.

Manipulation of the intake from infancy,however, frequently results in oral aversion, ageneral refusal to engage in normal eating be-havior. The child may exhibit poor oral-motorskills, decreased appetite, food refusal, and sen-sory problems, especially with textures. Causesof oral aversion include lack of eating experi-ence, especially when a gastrostomy tube isnecessary to facilitate feeding of special for-mula, anorexia secondary to organic acids inthe blood, and loss of the normal hunger cyclein individuals whose diet is regulated to avoidfasting. Lack of adequate oral intake can resulteither in excessive dependence on continuedgastrostomy tube feeding or in failure to thrive.Ultimately, failure to thrive leads to impairedbrain growth and developmental delay or evenmetabolic encephalopathy with loss of motorand cognitive skills. Oral aversion is a very com-plex disorder and must be treated in a multidis-ciplinary way with parents, occupational andspeech therapists, the nutritionist, and meta-bolic geneticist working together. Sometimes,medication is indicated to stimulate the child’sappetite and optimize eating behavior. Often,months or years of feeding therapy and behav-ior modification are necessary to teach thechild to eat normally.

Complications occur in spite of optimumnutritional therapy, especially if the child ac-quires a viral illness, ear infection, or gastroen-teritis and cannot maintain the necessaryformula intake. It is important that he or she beseen early to optimize hydration and maintaincaloric and protein intake. Maintaining bio-chemical homeostasis requires a fine balancein these patients, especially during the infantand toddler years when their ability to main-tain homeostasis during illness is much lessthan that of older children. Levels of plasmaglucose, electrolytes, pH, and HCO3

− togetherwith the specific gravity of urine are impor-tant indicators of the extent of metabolic de-compensation and the need for intravenousfluids with dextrose and HCO3

−. Unlike chil-dren with other inborn errors of metabolism,individuals with HMG-CoA lyase deficiency willnot accumulate either plasma or urinary ke-tones, and ketosis therefore cannot be used asan indicator of dehydration or catabolism. If thechild is vomiting or cannot take food by mouth,then gastrostomy tube feeding or intravenous


hyperalimentation must be started within 24–48 h to avoid further catabolism.

In summary, HMG-CoA lyase deficiency is aunique inborn error of metabolism with pro-found effects on both amino acid catabolismand metabolic homeostasis in the fasted state.Management of these patients is difficult andrequires constant attention to daily nutritionand timely intervention during acute illness.Fortunately, nutritional therapy treatment thatprovides a diet adequate for growth but withlimited intake of leucine and prevents fastingand hypoglycemia enables individuals withHMG-CoA lyase deficiency to live normal ac-tive lives.

QUESTIONS

1. Why does deficiency of HMG-CoA lyaseresult in the excretion of a large numberof organic acids in the urine?

2. Why do individuals with HMG-CoA lyasedeficiency have difficulty fasting?

3. Why might infection precipitate meta-bolic decompensation in an individualwith HMG-CoA lyase deficiency?

4. What is the rationale for providing supple-mental carnitine to individuals with HMG-CoA lyase deficiency?

5. Would you expect an individual with HMG-CoA lyase deficiency to have continued re-liance on medical foods as an adult? Why?

Acknowledgments: We wish to thank genetic coun-selor Heather Creswick, metabolic nutritionist MelodyPersinger-Yeargin, and administrative assistants ChristinaMcCalla and Trina Moore for their contributions to thischapter and preparation of the manuscript.

BIBLIOGRAPHY

Acosta PB,Yannicelli S: Disorders of leucine catab-olism, in Acosta PB,Yannicelli S (eds): NutritionSupport Protocols; the Ross Metabolic FormulaSystem. 4th ed.Abbott Laboratories, Columbus,OH, 2001, pp. 103–122.

Elsas LJ II,Acosta PB: Nutritional support in inher-ited metabolic disease, in Shils ME, Olson JA,Shike M, and Ross AC (eds): Modern Nutritionin Health and Disease. 9th ed. LippincottWilliams and Wilkins, Baltimore, MD, 1999, pp.1003–1056.

Faull K, Bolton P, Halpern B, et al.: Patient with de-fect in leucine metabolism. New Engl J Med294:1013, 1976.

Mitchell GA, Fukao T: Inborn errors of ketonebody metabolism, in Scriver CR, Sly WS, et al.(eds): The Metabolic and Molecular Bases ofInherited Disease. 8th ed. McGraw-Hill, NewYork, 2001, pp. 2327–2356.

Nyhan WL, Barshop BA, Ozand PT: 3-Hydroxy-3-methylglutaryl CoA lyase deficiency, in NyhanWL, Barshop BA, Ozand PT (eds): Atlas of Meta-bolic Diseases, 2nd edition. Chapman and HallMedical, 2005, pp. 253–258.

Roe CR, Millington DS, Maltby DA: Identificationof 3-methylglutarylcarnitine: a new diagnosticmetabolite of 3-hydroxy-3-methylglutaryl-coen-zyme A lyase. J Clin Invest 77:1391–1394, 1986.


CASE REPORT

A 35-year-old man presented to his family physi-cian after he noted a 20-min episode of slurredspeech and weakness of his left arm. He hadhad a similar episode, lasting 10 min, 3 days ear-lier. He had never smoked cigarettes, and hehad no history of high blood pressure,high cho-lesterol, diabetes, stroke, or neurological prob-lems. He did have a history of blood clots in hisright leg (deep vein thrombosis) and lungs (pul-monary embolism) at age 15 years. From age 15until age 20 he took the anticoagulant drugwarfarin, which inhibits blood clotting byblocking the synthesis of vitamin K-dependentcoagulation factors.

At age 26, he first noted pain and swellingof his right leg. These symptoms worsenedprogressively over the next 2 years, becomingso severe that he was no longer able to jog orplay basketball. At age 28, ultrasonography ofhis right lower extremity revealed chronicocclusion (blockage) of his right femoralvein with accompanying venous insufficiency(leaky valves in the veins of his leg). He thenunderwent catheter-directed thrombolysis,a procedure in which the drug urokinase,which dissolves blood clots, is infused intothe femoral vein. Following the thrombolysisprocedure, a metal stent was placed in thefemoral vein to help prevent its reocclusion.He again took warfarin along with clopido-grel, a drug that inhibits blood clotting bypreventing aggregation of platelets. Exceptfor some intermittent swelling of his right an-kle after exercise, he did well for the next 7years.

At the time he noticed the episodes of slurredspeech and left-sided weakness, he was takingno medications except for warfarin and clopi-dogrel. His diet consisted mainly of meats andfried potatoes, with very few fruits and vegeta-bles. He admitted to drinking up to a 12-pack ofbeer on the weekends. His family history wasnegative for stroke or blood clots, but his fatherhad suffered a fatal myocardial infarction (heartattack) at age 44.

On physical examination, his neurologicalfunction was completely normal, and hisspeech was intact. Magnetic resonance imag-ing of his brain was normal, but ultrasonog-raphy of his neck revealed a 50% stenosis(narrowing) of the right internal carotid artery.Laboratory valuation indicated that the con-centration of total homocysteine in his bloodplasma was elevated to 38 µmol/L (the normalrange is 5–12 µmol/L). The level of vitamin B12

in his serum was within the normal range(312 pmol/L), but the level of folate in hisserum was slightly below normal (3.4 ng/L;normal level >4.1 ng/L). His serum creatinine,an indicator of kidney function, also was nor-mal (0.8 mg/dL). Genetic testing revealed thathe was homozygous for the C677→T polymor-phism of the methylene tetrahydrofolate reduc-tase (MTHFR) gene.

He was diagnosed with moderate hyperho-mocysteinemia and treated with 2.0 mg folicacid daily. He continued to take warfarin andclopidogrel. After 1 month, his plasma total ho-mocysteine level had decreased to 9 µmol/L. Hehad no further episodes of weakness or otherneurological symptoms during the ensuing 12months.

21

Hyperhomocysteinemia

ANGELA M. DEVLIN and STEVEN R. LENTZ

226

DIAGNOSIS

Hyperhomocysteinemia is defined as an ele-vated level of homocysteine in the blood. It isdiagnosed by measuring the concentration oftotal homocysteine (tHcy) in a sample ofplasma or serum. The laboratory measure-ment of tHcy quantifies the total amount ofhomocysteine and all of its disulfide deriva-tives. The normal reference range for plasmatHcy is defined as the 2.5th-to-97.5th per-centile interval for presumed healthy individ-uals. This range is typically 5–10 µmol/L, butthe upper limit may vary between 10 and15 µmol/L depending on age, gender, ethnicgroup, and dietary intake of folate. Levels tendto increase with age and are higher in menthan women.

Mild hyperhomocysteinemia is defined as aplasma tHcy concentration of 10–30 µmol/L;moderate hyperhomocysteinemia is classifiedas 30–100 µmol/L.A very severe form of hyper-homocysteinemia,which produces plasma tHcyconcentrations greater than 100 µmol/L, can becaused by one of several inborn errors of me-thionine metabolism. Patients with these disor-ders also have high levels of tHcy in their urine,a condition known as homocystinuria.

Because hyperhomocysteinemia may be as-sociated with a variety of different nutritionaland genetic factors, additional laboratory test-ing is often useful in confirming the diagnosisand defining the cause of hyperhomocysteine-mia.Adjunctive tests may include measurementof blood levels of folate (vitamin B9), cobalamin(vitamin B12), or methionine and genetic testingfor mutations or polymorphisms in genes, suchas MTFHR, that are involved in homocysteinemetabolism.


Homocysteine, first described by Butz and Vin-cent du Vigneaud in 1932, is a sulfur-containingamino acid that is structurally and metabolicallylinked to the dietary amino acids methionineand cysteine (Fig. 21-1). Homocysteine is foundin both prokaryotic and eukaryotic organismsand is a sensitive marker of vitamin deficiencyand kidney function.

Homocysteine is formed in the cytoplasmduring the intracellular metabolism of methion-ine.Within the methionine cycle (Fig. 21-2), me-thionine is converted to S-adenosylmethionine

(SAM) by methionine adenosyltransferase. SAMserves as a methyl donor for a variety of methylacceptors, including DNA,protein,neurotransmit-ters, and phospholipids.S-Adenosylhomocysteine(SAH) is produced following methyl donation bySAM, and homocysteine is formed through theliberation of adenosine from SAH by the en-zyme SAH hydrolase. Unlike methionine andcysteine, homocysteine is not incorporated intopolypeptide chains during protein synthesis. In-stead, homocysteine has one of two metabolicfates: transsulfuration or remethylation to me-thionine.

The transsulfuration pathway involves con-version of homocysteine to cysteine by the se-quential action of two pyridoxal phosphate(vitamin B6)-dependent enzymes, cystathionine-β-synthase (CBS) and cystathionine γ-lyase (Fig.21-2). Transsulfuration of homocysteine occurspredominantly in the liver, kidney, and gastroin-testinal tract. Deficiency of CBS, first describedby Carson and Neill in 1962, is inherited in anautosomal recessive pattern. It causes homo-cystinuria accompanied by severe elevations inblood homocysteine (>100 µM) and methion-ine (>60 µM). Homocystinuria due to defi-ciency of CBS occurs at a frequency of about 1in 300,000 worldwide but is more common insome populations such as Ireland, where thefrequency is 1 in 65,000. Clinical features in-clude blood clots, heart disease, skeletal defor-mities, mental retardation, abnormalities of theocular lens, and fatty infiltration of the liver. Sev-eral different genetic defects in the CBS genehave been found to account for loss of CBS ac-tivity.

Cysteine is considered a nonessential nu-trient because it can be synthesized from me-thionine via the transsulfuration pathway(Figs. 21-1 and 21-2). Production of cysteine ismetabolically important because it serves as asource of sulfur for incorporation into pro-teins and detoxification reactions. A lack ofcysteine needed for incorporation into thestructural protein collagen may be responsi-ble for the musculoskeletal abnormalitiesseen in patients with CBS deficiency. A majormetabolic use of cysteine is in the productionof glutathionine (γ-glutamylcysteinylglycine),an important antioxidant. Another importantpathway for cysteine metabolism is its oxida-tion to cysteinesulfinate, which serves as a pre-cursor for taurine, an amino acid that stabilizescell membranes in the brain.

Remethylation of homocysteine to methionine

Hyperhomocysteinemia 227

is primarily accomplished by the cobalamin-dependent enzyme methionine synthase (MS),which uses 5-methyltetrahydrofolate (5-MTHF)as the methyl donor for the remethylation re-action. 5-MTHF is produced by the folate cycleenzyme methylenetetrahydrofolate reductase(MTHFR), which catalyzes the reduction of5,10-methylenetetrahydrofolate to 5-MTHF. Inaddition to serving as the precursor to 5-MTHF,5,10-methylenetetrahydrofolate also providesmethyl groups for the synthesis of thymidy-late, a pyrimidine required for the synthesis ofDNA.

Deficiency of MTHFR can produce severe hy-perhomocysteinemia and homocystinuria butis distinguishable from CBS deficiency in thatblood levels of methionine are not elevated.The first case of homocystinuria caused by defi-ciency of MTHFR was described by Mudd et al.in 1972. The case was a patient who presentedwith homocystinuria without elevation ofblood levels of methionine. Since then, several

genetic defects in the MTHFR gene resulting insevere deficiency of enzyme activity and homo-cystinuria have been described.

A common thermolabile variant of MTHFR,first described by Frosst et al. in 1995, resultsfrom a C→T substitution at nucleotide 677 of theMTHFR gene. Homozygosity for the C677→Tvariant produces deficient MTHFR activity andis associated with mild or moderate hyperho-mocysteinemia. Individuals homozygous forthe C677→T variant have an increased risk ofcardiovascular disease, especially under condi-tions of low dietary folate intake (Klerk et al.,2002). The frequency of the 677TT genotyperanges from 10% to 20% in Caucasian popula-tions in North America and Europe and hasbeen shown to be slightly higher in MexicanAmerican populations and lower in AfricanAmerican populations.

During the remethylation of homocysteineto methionine, the methyl group from 5-MTHFis transferred to cobalamin, which serves as an


Figure 21-1. Structural and metabolic relationships between methionine, homocys-teine, and cysteine. CBS, cystathionine b-synthase; CTH, cystathionine γ-lyase; MAT, me-thionine adenosyltransferase; MS, methionine synthase; 5-MTHF, 5-methyltetrahydrofolate;MTs, methyl transferases; PLP, pyridoxal phosphate; SAH, S-adenosylhomocysteine; SAHH,SAH hydrolase;THF, tetrahydrofolate.

intermediate methyl carrier (Fig. 21-3). Themethyl group is subsequently transferred tohomocysteine, generating methionine. Theother product of the reaction is tetrahydrofo-late, which is then remethylated to 5,10-methylenetetrahydrofolate by the enzymeserine hydroxymethyltransferase, completingthe folate cycle (Fig. 21-2). Over time, the cat-alytic activity of MS decays due to oxidation ofits cofactor, cobalamin. Regeneration of activeMS requires the reductive methylation ofcobalamin via methionine synthase reductase(MTRR), which uses SAM as the methyl donor(Fig. 21-3).

Deficiency of MTRR can produce homo-cystinuria accompanied by megaloblastic ane-mia (a type of anemia characterized by thepresence in the blood of large, immature redblood cells). This condition is also known ascobalamin-responsive homocystinuria and isclassified as the cbl E complementation type of

cobalamin deficiency. Defects in the intestinalabsorption, metabolism, or transport of cobal-amin can also result in hyperhomocysteinemiawith megaloblastic anemia (see Chapter 28 onVitamin B12 deficiency).

An alternative reaction for the remethylationof homocysteine to methionine can be accom-plished by betaine homocysteine methyltrans-ferase, which uses betaine instead of 5-MTHFas the methyl donor. Unlike the MS reaction,which is believed to be ubiquitously present inall tissues, the betaine homocysteine methyl-transferase reaction occurs only in the liverand kidney. Betaine is not a required nutrientsince the liver can synthesize betaine fromcholine. Betaine supplements, however, havebeen shown to lower plasma total homocys-teine concentrations successfully in subjectswith deficient homocysteine remethylation dueto defects in MTHFR or MTRR and in thosewith deficient CBS activity.


Figure 21-2. Metabolism of homocysteine. BHMT, betaine:homocysteine methyl-transferase; CBS, cystathionine β-synthase; Cob, cobalamin; CTH, cystathionine γ-lyase;DHF, dihydrofolate; DMG, dimethylglycine; FAD, flavin adenine dinucleotide; MAT,methionine adenosyltransferase; 5-MTHF, 5-methyltetrahydrofolate; 5,10-MTHF, 5,10-methylenetetrahydrofolate; MTHFR, methylenetetrahydrofolate reductase; MS, methio-nine synthase; MTRR, methionine synthase reductase; MTs, methyl transferases; PLP,pyridoxal phosphate; SAH, S-adenosylhomocysteine; SAHH, SAH hydrolase; SAM, S-adenosylmethionine; SHMT, serine hydroxymethyltransferase; THF, tetrahydrofolate; Zn,zinc.

Blood Homocysteine

The fraction of intracellular homocysteine thatdoes not undergo transsulfuration or remethy-lation is secreted into the extracellular spaceand ultimately finds its way into the blood. Onemajor source of blood homocysteine is theliver, but some homocysteine is secreted intothe blood by endothelial cells, circulatingblood cells, and other tissues. Only about 2% ofhomocysteine in blood remains in its reduced,thiol form. The remainder circulates as a vari-ety of different oxidation adducts, which in-clude the disulfide, homocystine, as well ashomocysteine-cysteine mixed disulfide and sev-eral protein-bound disulfides.About 70% of totalhomocysteine in blood is bound to the proteinalbumin through a disulfide linkage.When bloodhomocysteine is measured in the clinical labo-ratory, a reducing agent is added to the sample

to convert all of the disulfide forms to homocys-teine, which is then detected using either chro-matographic or immunologic methods. Theresult is reported as tHcy.

Measurement of blood tHcy is usually per-formed for one of three reasons: (1) to screenfor inborn errors of methionine metabolism; (2)as an adjunctive test for cobalamin deficiency;(3) to aid in the prediction of cardiovascularrisk. Hyperhomocysteinemia, defined as an ele-vated level of tHcy in blood, can be caused bydietary factors such as a deficiency of B vita-mins,genetic abnormalities of enzymes involvedin homocysteine metabolism,or kidney disease.All of the major metabolic pathways involved inhomocysteine metabolism (the methionine cy-cle, the transsulfuration pathway, and the folatecycle) are active in the kidney. It is not known,however, whether elevation of plasma tHcy inpatients with kidney disease is caused by de-creased elimination of homocysteine in thekidneys or by an effect of kidney disease on ho-mocysteine metabolism in other tissues. Addi-tional factors that also influence plasma levelsof tHcy include diabetes, age, sex, lifestyle, andthyroid disease (Table 21-1).

Most cases of hyperhomocysteinemia areclassified as mild (a plasma tHcy concentrationof 10–30 µmol/L) or moderate (a plasma tHcyconcentration of 30–100 µmol/L). Mild or mod-erate hyperhomocysteinemia can be caused bynutrient deficiencies (Table 21-2), geneticpolymorphisms in genes encoding enzymes re-quired for homocysteine or methionine metab-olism, impaired kidney function, or excessivealcohol intake. Some drugs, such as nitrous ox-ide, antiepileptic drugs, and certain drugs usedto treat asthma or high cholesterol, can pro-duce mild hyperhomocysteinemia by alteringthe metabolism of homocysteine, folate, orcobalamin. The most severe form of hyperho-mocysteinemia, which produces plasma tHcyconcentrations greater than 100 µmol/L, isfound in patients with homocystinuria causedby severe deficiency of CBS, MTHFR, or MTRR.

Hyperhomocysteinemia and Vascular Disease

It has been recognized for over 30 years thatindividuals with homocystinuria due to CBSdeficiency have a very high incidence of car-diovascular disease. When untreated, affectedpatients have about a 50% chance of develop-ing a myocardial infarction, stroke, or serious


Figure 21-3. The methionine synthase reaction.Methionine synthase catalyzes the remethylationof homocysteine to methionine. In the first half re-action (1), a methyl group is transferred from 5-methyl tetrahydrofolate (5-MTHF) to the reducedform of cobalamin [Cob(I)], generating methyl-cobalamin [Methyl-Cob(III)] and tetrahydrofolate(THF). During the second half reaction (2), themethyl group is transferred from methylcobalaminto homocysteine, generating methionine. Duringthe catalytic reaction,Cob(I) occasionally becomesoxidized,producing an inactive form of cobalamin,cob(II)alamin [Cob(II)]. The enzyme methioninesynthase reductase (MTRR) then reactivates Cob(II)through reductive methylation, producing methyl-Cob(III). SAM,S-adenosylmethionine; SAH,S-adeno-sylhomocysteine.

blood clot before the age of 30. Because themetabolic defect in CBS deficiency producesmarked elevation of blood levels of both homo-cysteine and methionine, it was initially notknown whether the high risk of vascular dis-ease in these patients is related to elevation ofhomocysteine, methionine, or another factor.

In 1969, McCully made an important obser-vation when he noticed severe vascularpathology caused by defects in homocysteineremethylation in children with homocystinuria(McCully, 1969). Like patients with CBS defi-ciency, these children had markedly elevatedlevels of tHcy in their blood and urine, but theirblood levels of methionine were lower than

normal. It was McCully’s observation that vascu-lar disease is a characteristic feature of homo-cystinuria caused by either transsulfuration orremethylation defects; this led him to postulatethat homocysteine itself may be a causativeagent in cardiovascular disease.

Initial attempts to confirm McCully’s hypothe-sis met with mixed results in the 1970s and1980s. Progress in the field was hampered by alack of reliable laboratory methods to measurecirculating levels of plasma tHcy accurately.Theintroduction of newer, automated methods fordetermination of plasma tHcy levels in the 1990sled to an exponential increase in published re-search studies investigating the pathological ef-fects of hyperhomocysteinemia. Elevations inplasma tHcy have now been proven to be associ-ated with increased risk of developing coronaryheart disease, stroke, and blood clots.

Like cholesterol, tHcy appears to be a gradedrisk factor, and even mild hyperhomocysteine-mia confers an increased risk of cardiovascularevents.A meta-analysis concluded that a 25% el-evation in plasma tHcy (about 3 µmol/L) ispredictive of about a 10% increased risk of myo-cardial infarction and a 20% increased risk ofstroke (Homocysteine Studies Collaboration,2002).

Despite a large amount of epidemiologicalevidence supporting the hypothesis that hyper-homocysteinemia is a risk factor for cardio-vascular disease, the molecular mechanismsunderlying the vascular pathogenic effects of ho-mocysteine are still not fully understood. Homo-cysteine is a highly reactive amino acid that canundergo redox reactions, producing oxidants


Table 21-1. Factors Influencing Plasma Levels of Total Homocysteine

Factor

Increase Diabetes

Increasing age

Male sex

Postmenopause

Smoking

Alcohol

Coffee

Hypothyroidism

Renal disease

Decrease Pregnancy

Hyperthyroidism

Exercise

Vitamin intake (Table 21-2)

Table 21-2. Nutrients Required for Homocysteine MetabolismNutrient Function

Folate (vitamin B9) Methionine cycle: methyl donor in the remethylation of homocysteine to methionine viamethionine synthase

Cobalamin (vitamin B12) Methionine cycle: intermediate methyl carrier in the remethylation of homocysteine tomethionine cofactor for methionine synthase

Pyridoxine (vitamin B6) Transsulfuration pathway: cofactor for cystathionine-β-synthase and cystathionine γ-lyase

Folate cycle: methylation of tetrahydrofolate cofactor for serine hydroxymethyltransferase

Riboflavin (vitamin B2) Folate cycle: reduction of 5, 10-methyltetrahydrofolate cofactor for methylene-tetrahydrofolate reductase

Betaine* Methionine cycle: methyl donor in the remethylation of homocysteine to methionine via betaine hydroxymethyltransferase

Choline* Methionine cycle: required for the formation of betaine

Zinc Methionine cycle: remethylation of homocysteine to methionine cofactor for methionine synthase

*Not considered to be essential nutrients in healthy adults.

such as hydrogen peroxide. Homocysteine alsocan participate in disulfide exchange andaminoacylation reactions, resulting in the chem-ical modification of proteins. Studies usingmouse models of CBS deficiency and MTHFRdeficiency have confirmed that mild hyperho-mocysteinemia produces oxidative stress andvascular inflammation, resulting in dysfunctionof the endothelium of blood vessels (Faraci andLentz, 2004).

One major mechanism for endothelial dys-function in hyperhomocysteinemic animalsappears to be mediated by the oxidative inacti-vation of nitric oxide, a diffusible gas thatrelaxes blood vessels and prevents blood clot-ting. Impairment of NO-mediated dilation ofblood vessels also has been observed in hu-man subjects with chronic or acute hyperho-mocysteinemia (Lentz, 2001).

Another mechanism contributing to the vas-cular pathogenic effects of homocysteine maybe a diminished capacity for methylation reac-tions. Hyperhomocysteinemia is often accom-panied by a decreased intracellular SAM/SAHratio, which can inhibit the methylation of DNAand other methyl acceptors (Fig. 21-2). DNA hy-pomethylation, with resulting changes in geneexpression, has been observed in white bloodcells isolated from human subjects with hyper-homocysteinemia (Ingrosso et al., 2003).

Birth Defects and PregnancyComplications

Hyperhomocysteinemia is associated with anincreased risk of birth defects resulting fromincomplete closure of the neural tube duringearly embryogenesis. The molecular mecha-nism responsible for this manifestation of hy-perhomocysteinemia is not well understood,but the risk of a neural tube defect can be de-creased by treatment with folic acid prior toand during pregnancy. Neural tube closure oc-curs very early in gestation,during a time whenmany women may be unaware of their preg-nancy. To help prevent neural tube defects,therefore, the Food and Drug Administrationmandated in 1998 that all grain products in theUnited States be fortified with folic acid. Inaddition, the American College of Obstetri-cians and Gynecologists recommends that allwomen of childbearing age take folic acid sup-plements for the prevention of neural tubedefects.

Hyperhomocysteinemia also has been impli-cated as a risk factor for other complications ofpregnancy, including intrauterine growth retar-dation, eclampsia, birth defects other than neu-ral tube defects, and premature separation ofthe placenta from the wall of the uterus. It isnot yet known, however, whether treatmentwith folic acid or other homocysteine-loweringtherapies will protect pregnant women fromthese complications.

Neurological Disease

Regulation of homocysteine metabolism ap-pears to be especially important in the centralnervous system, presumably because of thecritical role of methyl transfer reactions in theproduction of neurotransmitters and othermethylated products. It has been known for de-cades that mental retardation is a feature of thegenetic diseases, such as CBS deficiency, thatcause severe hyperhomocysteinemia and ho-mocystinuria. Impaired cognitive function isalso seen in pernicious anemia, which causeshyperhomocysteinemia due to deficiency ofcobalamin (see Chapter 28).Hyperhomocysteine-mia also may be linked to depression,schizophre-nia, multiple sclerosis, and Alzheimer’s disease.The molecular mechanisms underlying theseclinical associations have not yet been delin-eated.

THERAPY

Homocysteine-lowering therapy is lifesaving forpatients with severe hyperhomocysteinemiadue to CBS deficiency. Approximately 50% ofpatients respond to treatment with pharmaco-logical doses (100 to 800 mg daily) of pyridox-ine (a form of vitamin B6). Adjunctive therapymay include betaine (2 to 6 g daily), folic acid(5 to 10 mg daily),or cobalamin (0.05 to 1.0 mgdaily), or methionine restriction. Long-termhomocysteine-lowering therapy substantiallydecreases cardiovascular risk in these patients.

Dietary supplementation with B vitamins isalso highly effective in lowering homocysteinein most individuals with mild or moderate hy-perhomocysteinemia.A meta-analysis of 12 ran-domized trials performed prior to folic acidfortification concluded that treatment withfolic acid (0.5 to 5 mg daily) decreased homo-cysteine levels by 25%, and that the addition of


cobalamin (0.5 mg daily) produced a further 7%decrease in homocysteine levels (Brattstromet al., 1998).The addition of pyridoxine did notprovide any additional lowering of homocys-teine levels. In the current era of folic acid forti-fication, the degree of homocysteine loweringexpected from treatment with folic acid may beattenuated. Nevertheless, it is still possible toachieve a normalization of plasma homocys-teine in most individuals with mild or moderatehyperhomocysteinemia, although patients withrenal failure may be resistant.

It is less clear, however, whether homocys-teine-lowering treatment actually improvesclinical outcomes in patients with mild or mod-erate hyperhomocysteinemia. Several random-ized clinical trials of homocysteine-loweringtherapy are now being conducted to addressthis question.These trials are designed to testthe effectiveness of folic acid or cobalamin inpreventing vascular events (such as myocardialinfarction, stroke, or blood clots) in patientswith known coronary heart disease or priorstroke. Until the results of these trials becomeavailable, homocysteine-lowering therapy willcontinue to be used mainly in selected patients(such as those with a history of premature car-diovascular disease, stroke, or blood clots orthose felt to be at high risk due to the presenceof other risk factors such as the MTHFR 677TTgenotype).

When treatment is undertaken, the goal is tomaintain plasma tHcy within the low-normalreference range for the laboratory performingthe measurement. An American Heart Associa-tion science advisory statement has suggestedthat a homocysteine level of below 10 µmol/Lis a reasonable target for subjects with high car-diovascular risk (Malinow et al., 1999).

QUESTIONS

1. How does a diet deficient in folate affecthomocysteine and methionine metabo-lism?

2. What are the metabolic and clinical con-sequences of CBS deficiency?

3. What patient populations are at risk forhyperhomocysteinemia?

4. What is the predominant form of homo-cysteine in the blood, and what methodsare employed to measure levels of bloodhomocysteine?

5. What potential mechanisms have beensuggested to account for the vascularpathogenic effects of homocysteine?

6. Why is the U.S. food supply fortified withfolic acid?

BIBLIOGRAPHY

Brattstrom L, Landgren F, Israelsson B, et al.: Lower-ing blood homocysteine with folic acid basedsupplements—meta-analysis of randomised tri-als. BMJ 316:894–898, 1998.

Butz LW, du Vigneaud V.:The formation of a homo-logue of cystine by the decomposition of me-thionine with sulfuric acid. J Biol Chem99:135–142, 1932.

Carson NA, Neill DW.: Metabolic abnormalities de-tected in a survey of mentally backward individ-uals in Northern Ireland. Arch Dis Child37:505–513, 1962.

Faraci FM, Lentz SR.: Hyperhomocysteinemia, ox-idative stress, and cerebral vascular dysfunction.Stroke 35:345–347, 2004.

Frosst P, Blom HJ, Milos R, et al.: A candidate gene-tic risk factor for vascular disease: a commonmutation in methylenetetrahydrofolate reduc-tase. Nat Genet 10:111–113, 1995.

Homocysteine Studies Collaboration: Homocys-teine and risk of ischemic heart disease andstroke: a meta-analysis. JAMA 288:2015–2022,2002.

Ingrosso D, Cimmino A, Perna AF, et al.: Folate treat-ment and unbalanced methylation and changesof allelic expression induced by hyperhomocys-teinaemia in patients with uraemia. Lancet361:1693–1699, 2003.

Klerk M, Verhoef P, Clarke R, et al.: MTHFR677C→T polymorphism and risk of coronaryheart disease: a meta-analysis. JAMA 288:2023–2031, 2002.

Lentz SR: Homocysteine and cardiovascular physi-ology, in Carmel R, Jacobsen DW (eds), Homo-cysteine in Health and Disease. CambridgeUniversity Press, Cambridge, UK, 2001, pp.441–450.

Malinow MR, Bostom AG, Krauss RM: Homo-cyst(e)ine, diet, and cardiovascular diseases—astatement for healthcare professionals from theNutrition Committee, American Heart Associa-tion. Circulation 99:178–182, 1999.

McCully KS: Vascular pathology of homocysteine-mia: implications for the pathogenesis of arte-riosclerosis. Am J Pathol 56:111–128, 1969.

Mudd SH, Uhlendorf BW, Freeman JM, et al.:Homocystinuria associated with decreasedmethylenetetrahydrofolate reductase activity.Biochem Biophys Res Commun 46:905–912,1972.


CASE REPORT

A newborn male was the product of a 37-weekgestation to a primagravida female whose bloodtype was O and Rh positive. During the preg-nancy, the tests for syphilis and hepatitis B werenegative. The pregnancy was uncomplicated,and the mother received excellent care. She de-nies alcohol consumption and does not smoke.

The labor was prolonged, but the deliverywas spontaneous and vaginal.The child was vig-orous, active, and breathing spontaneously. Priorto delivery, the mother decided to breastfeedand the child was allowed to suckle immedi-ately following birth.

The initial physical exam revealed a birthweight of 2775 g, a birth length of 48 cm, and ahead circumference of 33 cm, which are withinnormal limits for a near-term male child.

The child was vigorous and appropriatefor gestational age. No bruises, skin rashes, orcyanosis were observed. The head, eyes, ears,nose, and throat were normal. He was normo-cephalic with appropriate molding of the cra-nial sutures from the delivery. The child hadexcellent sucking and swallowing coordination.The chest shape and movement were symmetri-cal with good air entry. No abnormal breathsounds were heard.The heart had a regular rateand rhythm. A slight holosystolic murmur washeard along the left sternal border and wasthought to be normal. The abdomen was softand nontender with no organomegally present.The umbilicus was clamped. Two arteries andone vein were visible. The genitourinary examwas normal, with a normal penis and bilaterallydescended testes. His extremities demonstrated

a full range of motion. No deformities werefound. Initial laboratory screening values wereas follows: hematocrit 48%; blood sugar 55 mg/dL; total serum bilirubin 1.0 mg/dL (all withinnormal range).The child was given the standardeye care and vitamin K injection. He was al-lowed to room with the mother, and the motherwas encouraged to allow the child to nurse atleast every 2 h.

The following morning, nearly 24 h later, thechange-of-shift nurse noted that the child ap-peared to be jaundiced, with the whites of theeyes appearing yellow and the skin yellow tothe level of the chest. Following the nurseryprotocol, repeat laboratory screening valueswere obtained. These tests showed the follow-ing values: hematocrit 45%; total serum biliru-bin 7 mg/dL; cord blood type A; and positiveCoombs antibody test. The child’s pediatricianwas informed of the results.

DIAGNOSIS

Hyperbilirubinemia in newborn infants is verycommon and accounts for most of the labora-tory testing done in the neonatal period. Thegreatest challenge to managing neonates withhyperbilirubinemia is the short hospital stay ofmothers and infants.The length of stay is as lit-tle as 24 h in some cases, which does not allowadequate time for the normal bilirubin rise tooccur.This requires the physician to make a di-agnosis quickly and accurately but sometimeswith only limited laboratory data.

Diagnosis of the cause of hyperbilirubinemia ismade by (1) evaluating the total serum bilirubin

22

Neonatal Hyperbilirubinemia

JEFFREY C. FAHL and DAVID L.VANDERJAGT

234

and the hematocrit or the hemoglobin; (2) ex-amining a blood smear for signs of hemolysis;(3) comparing the mother’s and infant’s bloodtype; and (4) obtaining a Coombs test.The dif-ferential diagnosis is broad but can be quicklyrefined based on simple laboratory tests donereliably by all hospital laboratories.This discus-sion focuses on the causes of unconjugated orindirect hyperbilirubinemia since conjugatedhyperbilirubinemia, which should raise clinicalalarms, has a very different differential diagno-sis.The causes of neonatal jaundice are listed inTable 22-1.

The gestational age of the infant is a majorfactor in the development of neonatal hyper-bilirubinemia.The more premature the infant is,the lower the level of expression of the en-zymes necessary for synthesis of conjugatedbilirubin (discussed in the section on HepaticMetabolism of Bilirubin) and the more likelythe child is to develop jaundice. Babies are notroutinely screened for the cause of jaundice un-til the condition manifests itself.Testing wouldbe instituted early if there were a sibling whohad experienced prolonged jaundice, or if themother is blood type O or is Rh negative. Allmothers who have good prenatal care aretested for blood type and Rh antibodies. Thisalerts the physician to potential problems andallows the physician to anticipate the mostcommon forms of jaundice, namely,ABO incom-patibilities.

These incompatibilities occur because ofsmall transfusions of maternal blood into the in-fant at the time of delivery.The naturally occur-ring antibodies that define blood groups thenmay initiate hemolysis. For example, a mother

with blood type O has antibodies to A and Bblood types. If the child has either an A or Bblood type, then the small maternal blood trans-fusion may lead to hemolysis of the child’sblood. However, if the child is blood type O,then no reaction will occur. Even if the motheris blood type O, there is a low probability ofjaundice developing.A higher risk occurs if thechild is Coombs positive, which demonstratesantibody formation to the child’s blood type.(The direct Coombs test detects antibodieson the erythrocytes using Coombs antiglobulinreagent. The indirect Coombs test detectsantierythrocyte antibodies in serum).

Rh-negative mothers have an advantagenow that RhoGAM, an antisera against Rh anti-body formation, is available. RhoGAM is an im-munoglobulin G globulin that, when given toan Rh-negative mother, prevents antibody for-mation to Rh-positive cells from the infant.Thisexposure occurs during pregnancy, particularlyat the time of delivery. If given to the motherwithin 72 h of delivery, then RhoGAM effec-tively prevents the mother from developing an-tibodies to Rh-positive blood and preventshydrops fetalis, which would often lead to ma-jor neonatal problems prior to 1965.

Although ABO incompatibility and Rh-anti-body formation are the most common causes ofhemolysis, other causes need to be considered,such as cephalohematoma formation duringdelivery with resultant increase in biliru-bin production as the hematoma is resorbed;hereditary spherocytosis, which is a red cellmembrane defect that results in prematurebreakdown of the red cells; and glucose 6-phosphate dehydrogenase deficiency, which isinvolved in maintaining adequate reduced glu-tathione levels in the red cell. Infection in theneonatal period is uncommon but still must beconsidered as a cause of jaundice. In particu-lar, infections of the urinary tract lead fre-quently to jaundice as a preliminary symptom.The increase in infection associated withinstrumentation in the premature infant is al-ways a concern.

Infants are slow to begin nursing, often donot drink enough fluids during the first fewdays of life, and lose weight. Dehydration canoccur, and it is a major effort to encouragemothers to feed or nurse their child during theneonatal time period.Dehydration concentratesthe blood volume and increases the bilirubinconcentration.

Breast-fed infants may have problems because

Neonatal Hyperbilirubinemia 235

Table 22-1. Conditions That May ProduceNeonatal Jaundice

Prematurity

ABO incompatibilities

Rh-negative mothers with prior exposure to Rh-positiveinfants

Hemolysis

Infection

Dehydration

Breast-feeding

Hypothryoidism

Maternal drug use

Drugs given to infants

Familial nonhemolytic hyperbilirubinemia

milk flow doesn’t occur immediately afterbirth. Frequent suckling induces milk flow,but the presence of pregnandiol in the breastmilk inhibits expression of bilirubin-UDP-glucuronyltransferase (discussed in the sectionon Hepatic Metabolism of Bilirubin), leadingto the accumulation of indirect hyperbiliru-binemia.

Hypothyroidism is sufficiently common thatall infants are screened for this problem priorto discharge from the nursery. Most infantsare asymptomatic because thyroxine diffusesacross the placental membrane.When hypothy-roidism does occur, it results in lethargy, slowmovement, hoarse cry, macroglossia, umbilicalhernia, large fontanels, hypotonia, dry skin, hy-pothermia, and prolonged unconjugated hyper-bilirubinemia. These are late findings, so earlydetection and treatment will improve long-termoutcome and avoid cretinism, which includesextremely poor developmental and physicalgrowth.

One needs to keep in mind that the use ofdrugs by the mother will sometimes lead toimpairment of the activity of bilirubin-UDP-glucuronyltransferase.Phenothiazines are an ex-ample of this kind of interaction. The use ofdrugs in the neonatal intensive care unit alsocan contribute to hyperbilirubinemia. Usually,the medications that compete for binding siteson albumin are the culprits in this case (see sec-tion on Bilburin Transport).An example of thistype of interaction is the use of furosimide,which is a diuretic used to decrease fluid reten-tion and improve cardiac function and renaloutput.

Finally, there are hereditary causes of non-conjugated, nonhemolytic hyperbilirubine-mias.These are Crigler-Najjar types 1 and 2 andGilbert’s syndrome (discussed in the section onGenetic Diseases of Bilirubin Metabolism).


Bilirubin Metabolism

Heme Oxygenase

About 75% of bilirubin is derived from degrada-tion of the heme obtained from hemoglobinthat is ingested by phagocytic cells in the de-struction of senescent erythrocytes, especiallyin spleen and liver. Most of the remainingbilirubin is derived from turnover of otherheme-containing proteins, including myoglobin,

catalase, and cytochromes. Some hemoglobin isreleased into the circulation when senescenterythrocytes are destroyed.This hemoglobin iscomplexed as oxyhemoglobin dimers to hapto-globins, which are a family of α2-globulins syn-thesized in the liver and secreted into thecirculation. The haptoglobin-hemoglobin com-plexes are removed from the circulation byphagocytes, including Kupfer cells in the liver.Any free heme that might be released is com-plexed by the globulin hemopexin as well as byserum albumin and transported to the liver.

The heme that is dissociated from globin inthe phagocytic process is Fe+3-heme (ferrihemeor ferriprotoporphyrin IX).This form of heme isthe substrate for the microsomal heme oxyge-nases, which are mixed-function oxidases thatutilize molecular oxygen and NADPH, similar tothe family of cytochrome P450s (Fig. 22-1).Theoxidation of heme, catalyzed by heme oxyge-nase, releases carbon monoxide, iron, andbiliverdin IXα.The Fe+3 iron in heme is reducedby microsomal NADPH-dependent cytochromeP450 reductase in this process and is releasedas Fe+2.The carbon monoxide is obtained fromthe α-methene carbon of the cyclic tetrapyrrolering of heme.

Two genetically distinct heme oxygenases(HOs) have been characterized; HO-1 is an in-ducible form, whereas HO-2 is constitutive. HO-2 is abundant in selective tissues such as brainand the cardiovascular system. HO-1 is highlyinducible in many tissues in response to a vari-ety of stresses, including infection, exposure toxenobiotics, hypoxia, and proinflammatory cy-tokines. The recognition that heme oxygenaseis expressed in many different tissues led tostudies of the role of heme degradation apartfrom the pathway for formation and eliminationof bilirubin derived from destruction of senes-cent erythrocytes.

Biology of Carbon Monoxide

It is well recognized that low levels of carbonmonoxide (CO) derived from the heme oxyge-nase reaction in certain tissues play an impor-tant role in the physiological response to stress.CO participates in several signaling pathwaysin which CO exerts vasoregulatory, anti-in-flammatory, antiapoptotic and antiproliferativeeffects. Evidence is accumulating that CO playsan especially important role in neural signaling.Thus, the pathway of heme degradation (Fig.22-1) should be viewed in the context of both


induction of HO-1 and production of CO as partof the cellular response to stress and as a majordetoxification/elimination pathway to rid thebody of excess bilirubin.

Biliverdin Reductase

The linear tetrapyrrole biliverdin IXα that is aproduct of the heme oxygenase reaction is thesubstrate for biliverdin reductase, which cat-alyzes the NADPH-dependent reduction ofbiliverdin IXα at the γ−methene bridge to formbilirubin IXα (Fig. 22-1). Many of the pathologi-cal problems associated with excess produc-tion or diminished elimination of bilirubin are

the result of the low solubility of bilirubin IXα.Other isomers of bilirubin are more solublethan the IXα isomer;however, the IXα isomer isthe main product of heme degradation in hu-mans. Bilirubin IXα is much less soluble thanbiliverdin IXα. This difference in solubility be-tween bilirubin IXα and biliverdin IXα is notobvious from a comparison of their structures(Fig.22-1).However,bilirubin IXα exists in solu-tion as an intramolecularly H-bonded molecule,as shown in Figure 22-2. This intramolecularH-bonding markedly diminishes the solubilityof this bilirubin isomer.Thus, much of the sub-sequent metabolism of bilirubin involving trans-port to the liver, conjugation with glucuronic


Figure 22-1. Production of bilirubin (BR). The degradation of Fe+3-heme by molecularoxygen and NADPH, catalyzed by microsomal heme oxygenase, produces biliverdin, CO,and Fe+2. Subsequent reduction of biliverdin by NADPH, catalyzed by biliverdin reduc-tase, produces bilirubin. Bilirubin that is produced in phagocytes from degradation ofsenescent erythrocytes is transported to liver for conjugation with glucuronic acid, cat-alyzed by bilirubin-UDP-glucuronyltransferase. In some cells, the bilirubin is used as anantioxidant, where it recycles through the biliverdin reductase reaction.

acid, and elimination through the biliary systemand intestine is designed to deal with the lowsolubility of bilirubin and the problems that canarise if bilirubin is not eliminated from thebody.

The widespread expression of biliverdin re-ductase, which parallels the widespread ex-pression of heme oxygenase, suggests thatbilirubin may play a special role in particulartissues. Studies suggested that bilirubin is an ex-cellent lipid-soluble antioxidant.Bilirubin reactswith reactive oxygen species (ROS) that oxi-dize bilirubin back to biliverdin.The biliverdinthen is reduced again to bilirubin.This repeatedcycling results in the net use of NADPH to pro-tect cells from ROS (Fig. 22-1).This system maybe especially important in neurons.

Bilirubin Transport

The bilirubin that is produced in phagocyticcells from degradation of hemoglobin repre-sents the majority of the bilirubin that is pro-duced and must be eliminated. This initiallyrequires transport of bilirubin from the phago-cytic cells to the liver. Normally, bilirubin is se-creted from phagocytic cells and complexedwith albumin for transport to the liver. It is es-sential that bilirubin is transported through thecirculation bound to albumin. The toxicity of

unbound bilirubin, especially that associatedwith neonatal hyperbilirubinemia, occurs whenthe binding capacity of serum albumin is ex-ceeded. Unbound bilirubin is readily depositedin skin and other tissue due to its affinity formembrane phospholipids. Unbound bilirubinreadily enters the brain, where it can producebilirubin encephalopathy (kernicterus). Albu-min has a high-affinity binding site for bilirubin.A serum albumin concentration of 3 g/dL is suf-ficient to complex approximately 25 mg/dLbilirubin.The movement of serum through theliver sinusoids allows bilirubin to be transferredfrom albumin into liver hepatocytes for furthermetabolism (Fig.22-3).This is a carrier-mediatedprocess.

Hepatic Metabolism of Bilirubin

Within hepatocytes, bilirubin forms a complexwith an abundant cytosolic family of enzymes,the glutathione S-transferases, that have bilirubin-binding sites and function to transport bilirubinto the surface of the endoplasmic reticulum,where one or two glucuronic acid moietiesare added to the propionic acid side chains ofbilirubin (Fig. 22-1).This reaction, catalyzed bybilirubin-UDP-glucuronyltransferase, involvestransfer of glucuronic acid from UDP-glucuronideto bilirubin.The effect of attachment of one or


Figure 22-2. Photobilirubins. Light-induced isomerization of bilirubin at the 4- and15-double bonds produces photoisomers that are water soluble. In addition, bilirubincan be photochemically cyclized to water-soluble lumirubin.

two glucuronic acids to bilirubin is to increaseits water solubility and to prepare bilirubinfor transfer across the bile cannaliculus (Fig.22-3).

Bilirubin that contains one or two glu-curonic acids is referred to as conjugatedbilirubin. Clinically, conjugated bilirubin iscalled direct-acting bilirubin, and unconjugatedbilirubin is called indirect-acting bilirubin.Thisis related to a colorimetric reaction called thevan den Bergh reaction that is sometimes usedto quantify the two forms of bilirubin, which isimportant in differential diagnosis of thecauses of hyperbilirubinemia. In the van denBergh reaction, conjugated bilirubin reacts di-rectly in an azo-dye-coupling reaction. On theother hand, unconjugated bilirubin must firstbe treated with alcohol to alter some intramole-cular H-bonds that prevent the dye-coupling re-action. It is therefore referred to as indirectbilirubin.

Bilirubin-UDP-glucuronyltransferase is onemember of a family of glucuronyl transferasesthat participate in the metabolism of xenobi-otics (foreign compounds) by increasing their

water solubilities.These enzymes are located onthe ER. It is important developmentally that thegene for bilirubin-UDP-glucuronyltransferasenot be activated early in fetal development,when it would be undesirable to produce con-jugated bilirubin and secrete it into the biliarysystem.Throughout most of fetal development,bilirubin produced by the fetus is transferredto the maternal circulation. The activity ofbilirubin-UDP-glucuronyltransferase increasesjust before birth and continues to increase afterbirth, giving the newborn the ability to conju-gate and excrete the bilirubin it produces. Inpremature infants, the activity of bilirubin-UDP-glucuronyltransferase is often low, leading to im-paired conjugation and secretion of bilirubin.Therefore, the hyperbilirubinemia associatedwith premature births is mainly unconjugatedbilirubin.

The transfer of conjugated bilirubin in thehepatocyte across the bile canaliculus depositsthe conjugated bilirubin along with other he-patic secretions into the biliary system fortransport to the small intestine. This transportprocess is catalyzed by one or more members


Figure 22-3. Transport and hepatic metabolism of bilirubin. Bilirubin that is producedin phagocytes is transported to liver as an albumin-bilirubin complex. Uptake into thehepatocytes takes place in liver sinusoids. Within the hepatocyte, bilirubin is trans-ported to the endoplasmic reticulum (microsomes) bound to glutathione S-transferase(GST).Bilirubin is made water soluble by addition of one or two glucuronic acid moietiesobtained from UPD-glucuronic acid, catalyzed by bilirubin-UDP-glucuronyltransferase.The product, conjugated bilirubin, is transported across the bile canalicular membranefor secretion into the biliary system, with subsequent movement into the intestines.

of the ATP-requiring transporters that belong tothe ABC family (ATP binding cassette) of trans-porters.

Metabolism of Bilirubin in theIntestine

Conjugated bilirubin that is released into thebiliary system is concentrated, along with otherhepatic and biliary secretions such as bile salts,lecithin and bicarbonate, in the gallbladder andis delivered to the small intestine with the othercontents of the gallbladder when the gallblad-der contracts. Conjugated bilirubin is absorbedpoorly in the intestine. In the lower intestineand colon, bacterial β-glucuronidases removeglucuronic acid to form unconjugated biliru-bin.Further metabolism of bilirubin,by bacteria,reduces bilirubin to two colorless compounds,urobilinogen and stercobilinogen. A smallamount of urobilinogen is absorbed and entersinto the enterohepatic circulation.A minor frac-tion of the absorbed urobilinogen remains inthe circulation, where it is ultimately excretedby the kidney, partly as the oxidized, coloredcompound urobilin, which imparts the charac-teristic yellow color of urine.The stercobilino-gen is excreted in stool mainly as the oxidizedform, stercobilin,which imparts the characteris-tic color of stool.

Genetic Diseases of BilirubinMetabolism

A number of genetic diseases are associatedwith defects in one or more of the enzymes in-volved in the metabolism of bilirubin.Three ofthese diseases result in elevated levels of un-conjugated bilirubin, and two of these result inelevated levels of conjugated bilirubin.All threeof the diseases that produce elevated levels ofunconjugated bilirubin are related to defects inthe level of expression or in the inherent activ-ity of bilirubin-UDP-glucuronyltransferase (alsocalled UGT1A1).The mildest and most commonof these disease is Gilbert syndrome, which ispresent in about 10% of the Caucasian popula-tion.

Gilbert syndrome is the result of the pres-ence of an additional TA in the TATAA box inthe proximal promoter of the UGT1A1 gene;this insertion reduces transcription, resulting inbilirubin-UDP-glucuronyltransferase activity thatis about 20% of normal in homozygous individ-uals. This condition is generally benign, with

affected individuals exhibiting mild elevations inunconjugated bilirubin. However, in some indi-viduals with other underlying genetic conditions,such as glucose 6-phosphate dehydrogenase de-ficiency, there may be much higher levels of un-conjugated bilirubin. In addition, some patientswith Gilbert syndrome have a problem with he-patic uptake of bilirubin.

The other two diseases related to UGT1A1are Crigler-Najjar (CN) syndrome types I and II.Many different mutations in UGT1A1 have beenidentified. CN type I patients express a mutatedprotein that is essentially devoid of activity.These patients therefore are unable to conju-gate bilirubin. Until recently, this condition waslethal in childhood; however, now these pa-tients can be treated with phototherapy (dis-cussed in the next section on Photobilirubin)and liver transplantation. Patients with CN typeII express a mutated protein that retains someactivity. These patients generally respond tophenobarbital, which increases the transcrip-tion of UGT1A1. Both CN types I and II are re-cessive disorders and rare.

The two genetic diseases associated with ele-vated levels of conjugated bilirubin are rare andbenign diseases that involve altered expressionor activity of the ATP-dependent pump thattransports conjugated bilirubin across the bilecanaliculus. Dubin-Johnson syndrome results ina chronic elevation of conjugated bilirubin andis also characterized by liver histology thatshows lysosomal accumulation of a black pig-ment. Rotor’s syndrome is similar in many re-spects to Dubin-Johnson syndrome exceptthat liver histology is normal.The exact differ-ences that distinguish these diseases are not yetknown.

Photobilirubin

For many years,phototherapy has been the stan-dard treatment of neonatal hyperbilirubinemia.The effectiveness of this form of therapy isbased on the ability of photons of the appropri-ate wavelength to convert the intramolecularlyH-bonded bilirubin IXα with its low solubilityinto photoisomers of bilirubin in which thenormal Z,Z stereochemistry at the 4- and 15-positions is changed, resulting in photoisomersthat are more water soluble. In addition, biliru-bin can be converted into the cyclic lumirubin,which is soluble and can be excreted in theurine (Fig. 22-2). Photoproducts can also beexcreted through the liver pathway without


requiring conjugation.Thus, phototherapy usesphotochemical energy to drive the excretion ofbilirubin by way of its photoproducts.Blue light(approximately 450 nm), longer wavelengthgreen light, or more commonly, fluorescentwhite light have been used for phototherapy.

TREATMENT

Treatment begins with prevention and an aware-ness of risk factors, summarized in Table 22-2.Breast-feeding is now the preferred method ofnourishment for infants, and usually frequentnursing is enough to overcome the hydrationproblems associated with hyperbilirubinemia.However, the combination of low breast milkvolume during the first 3 days of life and result-ant dehydration affect the level of bilirubin. In-terestingly, the addition of water to the feedingof the infant does little to reduce the bilirubinconcentration, but the addition of formula ishelpful. Knowing the blood type and Rh statusof the mother is important and is essential forgood prenatal care. In the absence of prenatalcare, assessment of these values at the time ofbirth becomes critical. The use of RhoGAM inRh-negative mothers is another preventive mea-sure that has improved the outcome of at-riskinfants.

Prevention alone is often not enough. Chil-dren still become jaundiced even after carefulassessment.When this occurs, it is important toassess the risk of developing significant hyper-bilirubinemia. The main concern is that thechild will develop acute bilirubin encephalopa-thy, which is caused by the toxicity of bilirubinon the basal ganglia and other brain stem nu-clei. There are early, middle, and late stages ofacute bilirubin encephalopathy (Table 22-3).The term kernicterus is applied to chronic

bilirubin encephalopathy (Table 22-4). Ker-nicterus occurs most often in infants who havedeveloped the late form of acute bilirubin en-cephalopathy. However, since it can occur inchildren with few if any of the signs of acutebilirubin encephalopathy, early diagnosis andtreatment are imperative.

Unfortunately, visual assessment has not beencorrelated with actual bilirubin levels.Therefore,laboratory measurement of the actual total serumbilirubin levels is mandatory.The timing of thesemeasurements becomes important because ofthe bilirubin load and level of maturation ofbilirubin-UDP-glucuronyltransferase.

Acceptable serum bilirubin levels for agehave been known for many years, and a nomo-gram for age versus bilirubin level is availableto help the clinician determine severity anddefine the level of intervention. Most childrenrequire careful monitoring only during theirhospital stay. However, outpatient monitoringmay be needed depending on where they fallon the nomogram. By knowing the age and thebilirubin level, the clinician can predict therise in bilirubin that will occur for that child.Furthermore, these data also predict whichchildren will require phototherapy or possiblyexchange transfusion.

The criteria for use of phototherapy are basedon risk factors and a rising bilirubin level that ex-ceeds the rate predicted by the nomogram orlevels between 15 and 20 mg/dL The child is at


Table 22-2. Risk Factors for Neonatal Jaundice

Jaundice in the first 24 hours

Blood group incompatibility

Gestational age less than 38 weeks

Sibling received phototherapy

Cephalohematoma or significant bruising

Exclusive breast-feeding, especially if not going well

Infant of a diabetic mother

Maternal age greater than 25

Male gender

East Asian race

Table 22-3. Stages of Acute BilirubinEncephalopathyEarly Middle Late (Irreversible)

Lethargy Moderate stupor Retrocollis-opisthotonos

Hypotonia Irritability Shrill cry

Poor suckling Hypertonia Refusal to feedApneaFeverDeep stupor to comaSeizuresPossible death

Table 22-4. Description of Chronic BilirubinEncephalopathy (Kernicterus)

Severe athetoid cerebral palsy

Auditory dysfunction

Dental-enamel dysplasia

Paralysis of upward gaze

Intellectual handicaps

very high risk if bilirubin levels are greater than20 mg/dL for a sustained period of time or ifthe child is not responding to phototherapy, inwhich case exchange transfusion may be con-sidered. Exchange transfusion replaces the in-fant’s blood volume with blood of the same orcompatible type that has a lower bilirubin level.This allows for the equilibration of tissue biliru-bin and the removal of antibodies that may leadto hemolysis.The exchange transfusion is donein combination with phototherapy.

Using this approach, the incidence of acutebilirubin toxicity has markedly decreased, andthe occurrence of kernicterus is now extremelyrare.

BIBLIOGRAPHY

American Academy of Pediatrics Clinical PracticeGuideline: Management of hyperbilirubinemiain the newborn infant 35 or more weeks of ges-tation. Pediatrics 114:297–316, 2004.

Bhutani VK, Johnson L, Sevieri EM: Predictive abilityof a predischarge hour-specific serum bilirubin

for subsequent significant hyperbilirubinemia inhealthy term and near-term newborns.Pediatrics103:6–14,1999.

Bosma PJ: Inherited disorders of bilirubin metabo-lism. J Hepatol 38:107–117, 2003.

Dennery PA, Seidman DS, Stevenson DK: Neonatalhyperbilirubinemia.N Engl J Med 344:581–590,2001.

Freda VJ,Gorman JG,Pollack W,et al.:Prevention ofRh hemolytic disease—ten-years’ clinical experi-ence with Rh immune globulin. N Engl J Med292:1014–1016, 1975.

Ip S, Chung M, Kulig J, et al.: An evidence-basedreview of important issues concerning neona-tal hyperbilirubinemia. Pediatrics 114:130–153,2004.

Kaplan M, Hammerman C, Maisels MJ: Bilirubin ge-netics for the nongeneticist: hereditary defectsof neonatal bilirubin conjugation. Pediatrics111:886–893, 2003.

McDonagh AF: Turning green to gold. NatureStruct Biol 8:198–200, 2001.

Ryter SW, Otterbein LE: Carbon monoxide in biol-ogy and medicine. BioEssays 26:270–280, 2004.

Shapiro SM: Bilirubin toxicity in the developingnervous system. Ped Neurol 29:410–421, 2003.


Part IV

Digestion, Absorption, andNutritional Biochemistry

CASE REPORT

T.F. is a 14-year-old girl who came into the officefor a well-care examination accompanied byher mother. She had a history of asthma andused albuterol (a bronchodilator) as needed.She was concerned that her weight was toohigh, and that she was often teased at school byother children. Both T.F. and her mother havebeen reading about weight loss and exercise;they were ready to make a change to decreaseT.F.’s weight.

The patient lived with her mother and hersister. Her father died in a car accident 2 yearsprior to the visit. Her grandmother had type 2diabetes and high cholesterol. T.F. ate fast foodmore than three times a week, drank three 8-ounce cans of nondiet soda each day, andwatched about 4 h of television per day. Sherarely exercised to break a sweat and did notparticipate in any sports activities. T.F. was170 cm tall (90%), weighed 91 kg (>90%), andhad a blood pressure of 110/70 mm Hg (normalfor her age and height is <125/80 mm Hg).

Her body mass index (BMI) was 31.5 kg/m2,which is more than 97% of that for her age. Herthyroid was not enlarged, the cardiac examina-tion was normal, and the lung examination wasnormal without any wheezing (a high-pitched,flutelike noise with expiration often seen withasthma). She had a dark, thick skin rash in thefolds on her neck called acanthosis nigricans.

Laboratory studies indicated that her fast-ing serum glucose level was 85 mg/dL (nor-mal is less than l00). Her serum lipid profilewas within the normal range: total cholesterol168 mg/dL; HDL cholesterol 43 mg/dL; LDL

cholesterol 105 mg/dL;VLDL cholesterol 20 mg/dL; triglycerides 101 mg/dL. She also had nor-mal serum levels of the enzymes AST (aspar-tate transaminase),ALT (alanine transaminase),and GGT (γ-glutamyl transpeptidase), which areserum markers of liver function.

DIAGNOSIS

Obesity is a common medical condition seendaily by most physicians. BMI is a useful clinicaltool to evaluate a patient’s size compared withothers of the same age.This measure takes intoaccount the patient’s height as well as weight.BMI is calculated with either of the followingtwo formulas depending on which measure-ment system is used:

BMI = Weight (kilograms)

Height (meters) × Height (meters)

BMI = Weight (pounds)× 703

Height (inches) × Height (inches)

T.F.’s BMI can be calculated as 31.5 (91 ÷ 1.7 ÷1.7); a BMI calculator is provided by the NIH athttp://nhlbisupport.com/bmi/bmicalc.htm.

Adults having a BMI between 25.0 and29.9 kg/m2 are considered to be overweight; in-dividuals with a BMI greater than or equal to30 kg/m2 are classified as obese (Obesity Educa-tion Initiative, 1998).

The BMI value can be plotted on a referenceBMI curve to determine a child’s or adolescent’s

23

Obesity: A Growing Problem

MIRIAM D. ROSENTHAL and LAWRENCE M. PASQUINELLI

245

http://nhlbisupport.com/bmi/bmicalc.htm

Figure 23-1. T.F.’s body mass index growth curve over time. Plotted on the referencechart of the National Center for Health Statistics,Centers for Disease Control; available athttp://www.cdc.gov/nchs/howto/howto.htm.

246

http://www.cdc.gov/nchs/howto/howto.htm

BMI percentile as a function of age (Fig. 23-1).Children and adolescents with a body mass in-dex percentile of 85% or greater and less than95% are considered “at risk of overweight”;those with a BMI of 95% or more are consid-ered “overweight.” Plotting BMI at the time ofmedical examinations can be helpful in moni-toring growth trends over time. Indeed, plottingBMI for T.F. clearly indicates the progressive de-velopment of first overweight and then obesityover a period of many years.

Several methods are available to evaluate apatient’s actual body composition rather thantotal body mass. Skin-fold measurement maybe of value in evaluating subcutaneous adipos-ity (adipose tissue accumulation); proper tech-nique is required for reliable results. Otheranthropomorphic measurements such as bio-electrical impedance, dual-energy x-ray absorp-tiometry, and total body water immersion arealso available.These last techniques are often ofvalue in research studies, but it is clinically im-practical to use them routinely (Elberg et al.,2004).

For adults, measurement of waist circumfer-ence is an easy and accurate indicator of upperbody or abdominal adiposity.A waist circumfer-ence greater than 102 cm (40 inches) for menand greater than 88 cm (35 inches) for womenis indicative of increased risk of cardiovasculardisease. Comparable standards are not availablefor children and teens.

Typically, overweight children are tall fortheir age when compared with their peers ofthe same sex and age. If a child is short and over-weight, then further studies should be done toexclude a genetically determined problem suchas Prader-Willi or Bardet-Biedl syndrome.

Obesity or excess adiposity was relativelyrare in earlier centuries. Indeed, an ample bellywas often seen as a sign of affluence and pros-perity. In the last two decades, however, adultobesity has reached epidemic proportions inthe United States and other developed coun-tries. This is strikingly evident from data com-piled by the Centers for Disease Control andPrevention (CDC), which estimates obesityrates by state. Data from 1990 indicate thatall states had obesity rates (BMI > 30 kg/m2

or ∼ 30 lb overweight) of less than 15%. By con-trast, in 2002, every state had an obesity preva-lence rate of at least 15%–19%; 29 states hadrates of 20%–24%; and the rates in three stateswere over 25%. Graphic representation ofthese data, updated annually, is available at

http://www.cdc.gov/nccdphp/dnpa/obesity/trend/maps/index.htm.

There has also been a marked increase inchildhood obesity in recent years. The above-referenced CDC Web site also includes data oncross-sectional studies on the weight of chil-dren. Data from 1963 to 1970 showed that 4%of 6- to 11-year-olds and 5% of 12- to 19-year-olds were overweight. Data from 1999 to 2000have shown an increased incidence in thisnumber: 15% for 6- to 11-year-olds and 15% for12- to 19-year-olds.

The chance that an obese child will becomean obese adult is estimated to increase from ap-proximately 20% if overweight at age 4 years toapproximately 80% if overweight during adoles-cence (Guo and Chumlea, 1999). Patients whoare overweight are at a higher risk for type 2 di-abetes, hypertension, and dyslipidemia.

Blood pressure should be measured onoverweight patients as part of each physicalexamination. Based on her height and age,T.F.’sblood pressure should be less than 120/80 mgHg; her blood pressure (100/70 mm Hg) wasacceptable. Blood pressure curves are avail-able that list acceptable blood pressure for pa-tients based on their age and height percentile(http://www.nhlbi.nih.gov/health/prof/heart/hbp/hbp_ped.htm).

Because her BMI is greater than 95%, it putsher at a higher risk for the development of bothtype 2 diabetes and hypercholesterolemia, fast-ing plasma glucose and lipid levels were ob-tained as part of T.F.’s physical examination. Inher case, all of the serum levels were withinnormal limits. If her fasting glucose had beenhigher than 100 mg/dL, then considerationwould have been made for doing a glucose tol-erance test to determine how her body man-ages sugar.A fasting blood glucose level greaterthan 126 mg/dL is diagnostic for type 2 dia-betes and requires treatment. Similarly, if T.F.’sfasting cholesterol had exceeded 170 mg/dL orher triglycerides were greater than 200 mg/dL,then further dietary investigation, testing, nutri-tional counseling,and possible pharmacologicalintervention would need to be considered.

T.F. had acanthosis nigricans, a velvety skinrash that mimics the appearance of poorlywashed skin and is commonly seen in the neckfolds, axilla, or under the breasts.Although thiscondition may be associated with type 2 dia-betes, BMI is a better predicator of risk for thedevelopment of type 2 diabetes than the find-ing of acanthosis nigricans.

Obesity 247

http://www.cdc.gov/nccdphp/dnpa/obesity/trend/maps/index.htm

http://www.cdc.gov/nccdphp/dnpa/obesity/trend/maps/index.htm

http://www.nhlbi.nih.gov/health/prof/heart/hbp/hbp_ped.htm

http://www.nhlbi.nih.gov/health/prof/heart/hbp/hbp_ped.htm

Nonalcoholic steatohepatitis (accumulationof triacylglycerol in the liver) or NASH is begin-ning to be identified more often in overweightindividuals.This condition may progress to cir-rhosis (fibrosis and loss of normal lobularstructure) of the liver. In the case of T.F., mea-surement of serum levels of AST, ALT. andGGT were used to rule out abnormal liver func-tion.

Patients who are overweight often sufferfrom depression and poor self-image.Evaluationand screening for these conditions are an es-sential aspect of caring for patients who areoverweight or obese. In addition, although manypeople are concerned that their increasedweight is related to the function of their thyroidgland, it is the cause of obesity in fewer than 1%of patients.


Individuals who are overweight or obese prima-rily have excess adiposity or excess storage oftriacylglycerols. The most common cause forpatients being overweight is that the individualconsumes more calories than are consumed forbasic metabolic needs and energy expenditure.For each additional 3500 kcal consumed thatare not utilized as fuel, 1 pound of fat is accu-mulated in the body.

The human organism is well adapted to inter-mittent availability of food.When we eat, excessenergy is stored for future use during intervalsof scarcity. Indeed, this fed-fasted cycle is a dailyphenomenon, with mobilization of stored fuelused to sustain the individual and maintain crit-ical concentrations of blood sugar during thenight.

Most of the body’s energy reserves are in theform of fat or triacylglycerol. Estimates are thatthe reference 70 kg (154 lb) adult male has15 kg of stored fat compared to a mere 0.2 kgof stored carbohydrate, primarily glycogen. Ifthat same individual were to gain 70 kg inweight and become quite obese, then 65 kg ofthe increased weight would be triacylglycerol,with the rest comprised of increased fluid and amodest increase in total body protein.

The human diet provides energy primarilyfrom a mixture of carbohydrates and fats, withprotein accounting for perhaps 12%–30% of to-tal energy intake.As noted above, the potentialfor storing carbohydrates is limited. Further-more, there are no protein stores in the body

per se, although some muscle proteins are read-ily mobilized during fasting to provide aminoacid precursors for gluconeogenesis. Instead,the dietary carbohydrates and protein-derivedamino acids not required for the immediatemetabolic and biosynthetic needs of the bodyare converted to triacylglycerol for storage(Fig. 23-2). Similarly, excess dietary fatty acidsare esterified into triacylglycerols and stored inadipose tissue.

In the postprandial or fasted state, the bodyrelies on stored triacylglycerols to provide mostof its energy needs. As seen in Figure 23-3, theadipocyte triacylglycerols are hydrolyzed byhormone-sensitive lipase, and the released freefatty acids are made available to muscle, liver,heart, and other organs. Some organs, notablybrain and red blood cells, cannot utilize freefatty acids as an energy source and are thereforedependent on a continued supply of glucose.The pancreatic hormone glucagon stimulatesthe release of a modest amount of glucose (upto 70 g) from liver glycogen. Muscle glycogenis unavailable for this purpose. (Muscle cellslack glucose 6-phosphatase, and any glucose-phosphate obtained from glycogenolysis is re-tained for utilization within the muscle ratherthan released into the blood.) The liver glyco-gen stores are, however, insufficient to meetglucose requirements during an overnight fast.Fortunately,when glucagon stimulates glycogenmobilization, it also stimulates gluconeogenesisand thus the synthesis of glucose from glyceroland amino acid precursors (Fig 23-3).

Food energy is usually measured in caloriesor kilocalories, with 1 kcal representing theamount of heat that is required to raise the tem-perature of 1 kg of water 1°C.The caloric yieldof different energy sources is shown in Table23-1.The physiological energy yield of protein(4 kcal/g) is less than the 5.7 kcal/g that is ob-served in a bomb calorimeter; the differencerepresents the energy required for the synthesisof urea.Note that fats provide more calories pergram than either protein or carbohydrate. Con-sumption of alcoholic beverages can also pro-vide significant energy, even if they containrelatively few carbohydrates. Dietary ethanol isoxidized to acetyl-CoA, which, depending onthe availability of other energy sources, is eitherfurther oxidized or utilized for synthesis of fattyacids.

There are two main components of dailyenergy expenditure: basal or resting energyexpenditure and energy expended in exercise.

248 DIGESTION, ABSORPTION, AND NUTRITIONAL BIOCHEMISTRY

For adults, the resting energy expenditure is esti-mated at about 1800 kcal/day for males andabout 1400 kcal/day for females (RecommendedDietary Allowances, 1989). At all ages, restingenergy expenditure is closely correlated withlean body mass (amount of actual metabolic tis-sue); individuals who increase their musclemass through physical activity actually have ahigher energy expenditure at rest.

The other major component of daily energyexpenditure is due to physical activity. Formost individuals in the United States, physicalactivity accounts for 15%–30% of the totaldaily energy. Most people can sustain meta-bolic rates 8–10 times their resting value(8–10 METS or multiples of the resting meta-bolic rate) with reasonably strenuous activi-

ties such as running, biking, or fast walking.When estimating energy expenditure due tophysical activity, one must consider both theincreased metabolic rate per minute and thelength of the activity. Thus, individuals whorun for 30 minutes per day before going towork at a desk job are still primarily seden-tary. Unfortunately, more and more childrenalso lead essentially sedentary lifestyles. Mus-cle work, independent of sustained physicalexercise, can also be a source of significantenergy expenditure.When 24-h energy expen-diture was studied in a respiratory chamber,much of the differences between individualswere the result of variability in the degree ofspontaneous physical activity (i.e.,“fidgeting”)(Ravussin et al., 1986).

Obesity 249

Figure 23-2. Metabolism in the fed state.An adequate supply of carbohydrate providesglucose to replenish glycogen stores. Dietary protein provides amino acids for proteinsynthesis. Dietary carbohydrates, fats, and proteins can all be metabolized to generateATP. (For clarity, ATP generation during β-oxidation of fatty acids and substrate-levelphosphorylation during glycolysis is not depicted.) Excess dietary carbohydrates andamino acids are converted to fatty acids and, along with excess dietary fatty acids, storedas triacylglycerols. DHAP, dihydroxyacetone phosphate.

The accumulation of excess adipose tissuetriacylglycerol, which results in development ofan overweight status and subsequent obesity,is a gradual process resulting from a long-term,often small imbalance in energy intake andenergy expenditure.A modest energy excess in-take of 50 kcal/day will add up to 18,250 kcalper year, or approximately 5 lb of additionalstored fat (3500 kcal/lb). Similarly, substantialweight loss requires negative energy balanceover a prolonged time period. Based on the3500 kcal/lb of fat, one can reasonably expectthat a deficit of 500 kcal/day will result in aweight loss of about 1 lb/week. Consulting theRDA tables (Recommended Dietary Allowances,1989), one finds the average energy a teenagersuch as T.F. expends per day is 2200 kcal. If shewere to fast completely (which is not medicallyrecommended), then she would not depletemore than about 4 pounds of fat per week. In-deed, much of the initial rapid weight loss on

popularized “diets” often reflects depletion ofmuscle protein for gluconeogenesis or loss offluids.

The goal of a healthy weight reduction pro-gram is the gradual depletion of excess triacyl-glycerol reserves.This occurs when, over time,the replenishment of triacylglycerol stores af-ter meals is less than the utilization of thosestores between meals. Exercise helps this pro-gram in several ways. First, exercise increasesfuel utilization, with skeletal muscle utilizinga mixture of muscle glycogen, plasma glucose,and plasma free fatty acids. The ratio of glu-cose to fatty acids oxidized depends on the in-tensity and duration of the exercise. Since thesubsequent meal is then utilized to replenishthe glycogen stores as well for triacylglycerolsynthesis, the important consideration is thetotal amount of energy expended rather thanthe specific fuel mixture. Second, as noted, reg-ular exercise increases the lean muscle mass


Figure 23-3. Mobilization of fuels during an overnight fast. Adipocyte triacylglycerolsare mobilized to provide the major fuel for the body. The released free fatty acids are uti-lized by muscle and most other tissues. Blood sugar supplies to the brain are maintainedby the liver, which releases glucose synthesized by gluconeogenesis as well as from itsglycogen stores. Amino acids from catabolism of muscle proteins and glycerol fromadipocyte triacylglycerol provide the substrates for gluconeogenesis. Under these condi-tions, liver utilizes β-oxidation of free fatty acids as a means of generating ATP. The result-ant two-carbon moieties from acetyl-CoA are exported as ketone bodies (acetoacetateand β-hydroxybutyrate) and provide an additional fuel supply for muscle. During pro-longed fasting, the ketone bodies also become a significant fuel source for the brain, thussparing glucose.

and thus the resting energy expenditure aswell.

The medical problems of obesity are severeand multiple. In addition to the obvious in-creased physical burden to organs such as theheart and joints, obesity is associated with sig-nificant metabolic changes. Studies have demon-strated that adipose tissue is not a passivereservoir for energy storage as triacylglycerolbut is instead an active endocrine organ (Ker-shaw and Flier,2004).Adipocytes produce a vari-ety of bioactive peptides called adipocytokines,which affect metabolism on both the local andsystemic level. For example, increased circulat-ing levels of one such adipocytokine, inter-leukin 6, are associated with impaired glucosetolerance and development of cardiovasculardisease. By contrast, adiponectin, a secondadipocyte-derived hormone, appears to have aninverse relationship with development of in-sulin resistance and inflammatory states.

Excess adiposity, particularly the abdominalobesity associated with increased waist circum-ference, is associated with insulin resistance,hy-pertension, and proinflammatory states. Theprevalence of this complex of comorbidities as-sociated with obesity, now referred to as themetabolic syndrome, is reaching epidemic pro-portions in the United States (Grundy et al.,2004; Roth et al., 2002). Indeed, increased ab-dominal adiposity is one of a cluster of factorsthat are used in the diagnosis of metabolic syn-drome.Abdominal tissue in the trunk occurs inseveral compartments, including subcutaneousand intraperitoneal or visceral fat.Visceral fat inparticular appears to contribute to perturbedfuel metabolism by at least two mechanisms.First, hormones and free fatty acids releasedfrom visceral fat are released into the portal cir-culation and impact directly on metabolism ofthe liver. Second, the visceral adipose depotproduces a different spectrum of adipocytokinesthan that produced by subcutaneous fat (Ker-shaw and Flier, 2004).

Adipocytes are also a significant source ofsteroid hormones, including cortisol, andro-

gens, and estrogens. Visceral adipocytes appearto be more active than subcutaneous adipocytesin production of glucocorticoids. Excess pro-duction of sex steroids by adipocytes has alsobeen linked to problems of polycystic ovarydisease, infertility, and dysfunctional uterinebleeding.

Modification of the quantity of energy intakeis clearly a primary focus for any approach toprevention or treatment of excess adiposity(Obesity Education Initiative, 1998). Althoughexcess caloric intake, whether carbohydrate,fat,protein,or even ethanol, is readily convertedto triacylglycerol for storage, earlier therapiesoften focused on decreasing total fat in the diet.This was in part because of the high caloricdensity of fat and the increased caloric contentof fried foods and in part because of the estab-lished linkage between high saturated fat andcholesterol intake and hypercholesterolemia.With the awareness of the prevalence of hyper-cholesterolemia and dyslipidemia in overweightindividuals,many medical experts use the Amer-ican Heart Association’s guidelines for a pru-dent diet (limit saturated fat to less than 10% ofthe patient’s total calories, limit total caloriesfrom fat to no more than 30%, and limit choles-terol intake to less than 300 mg/day) for theiroverweight patients (Willett, 2002).

It has become increasingly clear, however,that low fat does not necessarily mean lowcalorie. Indeed, the increased caloric intake ofthe American public in the last decade hascome primarily from carbohydrates. T.F.’s in-take of sugared soda beverages is typical ofthis situation. Furthermore, epidemiologicaldata have strengthened the link between obe-sity and type 2 diabetes, including the in-creased prevalence of this “adult-onset”diseaseat younger and younger ages. This has led toconcern about overall carbohydrate intake,particularly simple sugars and refined carbohy-drate foods such as white flour, which have ahigh glycemic index (or impact on postpran-dial blood glucose) and thus provoke a high in-sulin response.While both extreme low-fat andlow-carbohydrate advocates have received a lotof popular press, current studies support amore nuanced approach that distinguishes be-tween vegetable oils and saturated animal fatsand between whole grain carbohydrate sourcesand processed sugars (Willett and Stampfer,2003).

Metabolic research on obesity has also soughtto elucidate hormonal and neuroendocrine

Obesity 251

Table 23-1. Approximate Energy Values ofMetabolic Fuels

Carbohydrate 4 kcal/g

Protein 4 kcal/g*

Fat 9 kcal/g

Ethanol 7 kcal/g

*Net energy yield in human metabolism.

mechanisms of appetite control. Leptin, a circu-lating protein produced by adipocytes, con-tributes to regulation of energy homeostasisthrough effects on both the hypothalamus andon peripheral tissues. Studies have indicated,however, that leptin’s primary role is to serve asa metabolic signal of energy sufficiency ratherthan excess (Kershaw and Flier, 2004). Indeed,common forms of obesity are neither leptindeficient nor responsive to exogenous leptin.Another molecule that has attracted much re-search interest is ghrelin, a recently discoveredorexigenic (appetite-stimulating) gastric hor-mone; its production is induced by lack of foodin the stomach. There is, however, no com-pelling evidence that mutations in ghrelin con-tribute to human obesity (Cowley and Grove,2004). Furthermore, studies with mice deficientin either ghrelin or its receptor suggest thatghrelin is not essential for normal growth or ap-petite (Lang, 2004). Further studies on leptin,ghrelin, and other proteins that contribute tometabolic signaling may, however, lead to phar-macological products useful in the treatment ofwasting syndromes or morbid obesity.

THERAPY

For most overweight and obese individuals, theprimary therapeutic approach involves chang-ing the balance between energy intake and uti-lization. Efforts need to be made to encouragepatients both to decrease the amount of calo-ries consumed and to increase energy expendi-ture. Programs that incorporate exercise withdietary choice intervention have been moresuccessful in helping patients to lose weightand then to maintain their weight loss (Epsteinet al., 1985). Furthermore, identification of thepatient’s and family’s motivation for change isan important step in treatment. Obesity is afamily-centered problem, and all members ofthe family need to be willing to work to helpthe affected family members. Without this sup-port, long-term success is difficult.

T.F. and her mother were motivated to makebehavioral changes to help T.F. lose weight. Set-ting goals for behavioral change at each visit isimportant.The goals should be decided by thepatient with the family and should be goals thatare realistic and achievable. Once the old goalshave been met, new goals should be developed.Each visit is an opportunity to meet and to re-

view progress; frequent follow-up in an individ-ual or group setting greatly enhances successfulweight loss.

At the visit,T.F. agreed to limit her televisionwatching to less than 2 h a day. For children,limiting the amount of television watching toless than 2 h a day has been associated withweight loss (Robinson,2001).She also agreed tobegin walking in her neighborhood three timesa week for 30 min. She would drink only dietsoda and try to drink more water. Rather thanlimiting her eating out at restaurants, she wouldseek to become aware of portion sizes and tryto avoid high-fat, high-sugar meals. She mightconsider packing lunch for school if the avail-able choices are unsuitable. She and her motheragreed to come back for a follow-up visit in 2weeks. Her mother agreed to make an effort toincrease the availability of foods and snackswith lower caloric density (particularly vegeta-bles and fruits) available at home but to leaveactual food choices to T.F.

A realistic weight loss goal should be approx-imately 0.5 to 1 pound per week, which meansthat it will require time to reach the patient’spersonal goal. Weight loss, though important,should not be the sole measure of success. Im-provements in how clothing fits or how the pa-tient feels about herself should also berecognized and praised. For younger pediatricpatients who are still growing, maintaining thechild at a constant weight may allow the childto “grow into” a weight that is appropriate forthe child’s height.

Effective behavior modification programsalso require a long-term approach. It takes timefor individuals to acquire appropriate nutri-tional information and to develop sustainableprograms for physical activity. For this reason,T.F. and her mother were referred to a weightmanagement program provided by the localchildren’s hospital.The 7-week program was de-signed to provide additional education, skills,and tools to assist families as they help theirchildren adopt a healthier lifestyle.The weekly2-h group sessions, separate for individuals of8–11 and 12–18 years of age, incorporate physi-cal activity as well as education and peer sup-port. Each patient is also evaluated by anutritionist, licensed clinical social worker, anda physical therapist.

Unlike T.F., some obese patients have severemedical complications related to their obesity,such as uncontrolled diabetes,severe depression,


or sleep apnea. Under these circumstances,more aggressive measures may be necessary.One such measure is a medically supervisedvery low calorie, ketogenic diet (high protein,very low carbohydrate) for rapid weight loss.Use of a ketogenic diet in children requires in-tensive family dedication and may require inpa-tient hospitalization to meet the necessarydietary goals.

A pharmacological approach can also beconsidered (Therle and Aronne, 2003). Cur-rently, there are few medications available orapproved by the Food and Drug Administra-tion for weight loss in children and adoles-cents. Orlistat, which inhibits pancreatic lipaseto decrease fat absorption, and sibutramine, anorepinephrine-, serotonin-, and dopamine-reuptake inhibitor, have been used with someoverweight pediatric patients. Current opin-ion recommends that such medications beused as part of a multicenter research proto-col under the supervision of a physician quali-fied to monitor patients on these medicationsproperly. When patients have life-threateningcomplications resulting from their obesity orwhen all other therapies have been ex-hausted, there may be a role for bariatric sur-gery, which prevents the patient from eatinglarge quantities of food at a single meal. Thesurgical procedures are not without risk orpotential for complications, including death.Surgical intervention should therefore be theoption of last resort, especially in adolescentpatients.

QUESTIONS

1. Trying to stay awake to study, a second-year medical student added three 12-ozcans of caffeinated soda a day (∼40 gsugar/can) to her diet. She did not other-wise change her diet or physical activity.How much weight would she be ex-pected to gain during the next 6 months?

2. After 6 months, the student described inquestion 1 now wants to lose the weightshe has gained. She decides to switch todiet sodas and starts to walk 3 miles fivetimes per week. She looks up exercisephysiology studies and determines thatshe will burn about 75 kcal/mile walked.How long will it take to lose the addedweight?

3. What are the advantages and disadvan-tages (if any) to utilizing fat rather thancarbohydrate as one’s primary dietaryfuel?

4. What carbohydrate-containing foods wouldbe an important part of a balanced diet fora teenager such as T.F.?

5. What suggestions could you make to ateenager to decrease the caloric density ofher diet?

6. What social and economic obstacles mightan individual such as T.F. face in trying tocontrol her body weight?

BIBLIOGRAPHY

Cowley MA, Grove KL: Ghrelin—satisfying ahunger for the mechanism. Endocrinology 145:2604–2606, 2004.

Elberg J, McDuffie JR, Sebring NG, et al.: Compari-son of methods to assess change in children’sbody composition. Am J Clin Nutr 80:64–69,2004.

Epstein LH, Wing RR, Penner BC, Kress MJ: Ef-fect of diet and controlled exercise on weightloss in obese children. J Pediatr 107:358–361,1985.

Grundy SM, Brewer HB Jr, Cleeman JI, et al.: Defini-tion of metabolic syndrome: report of the Na-tional Heart, Lung, and Blood Institute/AmericanHeart Association conference on scientific issuesrelated to definition. Circulation 109:433–430,2004.

Guo SS, Chumlea WC:Tracking of body mass indexin children in relation to overweight in adult-hood. Am J Clin Nutr 70(suppl):145S–148S,1999.

Kershaw EE, Flier JS: Adipose tissue as an en-docrine organ. J Clin Endocrinol Metab 89:2548–2556, 2004.

Lang L: Studies question role of ghrelin in appetiteregulation. Gastroenterology 127:374–375,2004.

Obesity Education Initiative: Clinical Guidelineson the Identification, Evaluation, and Treat-ment of Overweight and Obesity in Adults:TheEvidence Report. 1998. NIH Publication 98-4083.Available at:www.ncbi.nlm.nih.gov/books.

Ravussin E, Lillioja S,Anderson TE, et al.: Determi-nants of 24-h energy expenditure in man. Meth-ods and results using a respiratory chamber.J Clin Invest 76:1568–1578, 1986.

National Research Council: Recommended Di-etary Allowances. 10th ed. National ResearchCouncil, National Academy Press, Washington,D.C., 1989.

Obesity 253

www.ncbi.nlm.nih.gov/books

Robinson TN: Television viewing and childhoodobesity. Pediatr Clin North Am 48:1017–1025,2001.

Roth JL,Mobathan S,Clohisy M:The metabolic syn-drome: where are we and where do we go?Nutr Rev 60:335–341, 2002.

Thearle M,Aronne LJ: Obesity and pharmacologic

therapy. Endocrinol Metab Clin North Am32:1005–1024, 2003.

Willett WC: Balancing life-style and genomics re-search for disease prevention. Science 296:695–698, 2002.

Willett WC, Stampfer MJ: Rebuilding the food pyra-mid. Sci Am 288:64–71, 2003.


CASE REPORTS

Case 1: Kwashiorkor

A 22-month-old male child was seen in a pedi-atric clinic in the United States (Carvalho et al.,2001). He had been suffering from intermittentskin eczema since 2 weeks of age. After wean-ing at 13 months of age, he was given cow’smilk, and several episodes of vomiting were ob-served soon after. Since his parents suspectedmilk intolerance and were apprehensive aboutthe cow’s milk causing a possible worsening ofthe eczema, he was immediately switched to alactose-free diet that was primarily a rice milkbeverage.This rice drink was extremely low inprotein but (somewhat) adequate in caloriccontent. Although well educated, the parentsmistakenly assumed that the beverage, becauseof its fortified contents and high cost, was a su-perior food product for their child even thoughthe container stated that the rice milk beveragewas not meant for use as an infant formula.Despite normal growth during infancy, poorweight gain was obvious during the secondyear, when the child was on a daily intake of1.5 L of the rice drink, which is equivalent to3 g protein and 790 calories/day (norm for ages1–3, 16 g protein and 1300 calories/day). In theweeks before admission, he was irritable, wasless active, and displayed an increase in skin le-sions.

At the time of admission, the child weighed10.8 kg (10th percentile), and height was81 cm (5th percentile). His physical examina-tion revealed generalized edema; alternatinghyper- and hypopigmented patches in the skin;

thin, sparse scalp hair; abdominal distension;and irritability. Laboratory tests of the child’sserum revealed 1.0 g/dL albumin (normal is3.5–5.5 g/dL), less than 0.5 mg/dL urea nitrogen(normal is 2–20 mg/dL), and 2.2 mg/dL phos-phorus (optimal is 4.5 mg/dL). Hematologicalanalysis showed normocytic anemia withmarked anisocytosis (variation in cell size).Thewhite blood cell (WBC) count of 11,500cells/mm3 was normal.

A diagnosis of kwashiorkor was made aftera detailed workup to rule out other causesof hypoproteinemia such as proteinuria (lossof proteins in urine) and protein-losing en-teropathy (loss of serum proteins through thegut). As there was severe anorexia, gradualrefeeding was started using a nasogastrictube.A soy-based, protein-rich formula was ini-tiated at the rate of 80 mL/kg/day and gradu-ally increased to 120 mL/kg/day. Potassium,phosphorus, zinc, folate, and other vitaminswere supplemented.The child also received a10-day course of oral cotrimoxazole (a broad-spectrum antibiotic) for his otitis media(acute inflammation of the middle ear) witheffusion. He was subsequently switched to amilk-based nutritional supplement since soy-beans are relatively deficient in methionine,an essential dietary amino acid. Within 3weeks of treatment, rising serum albumin lev-els and resolution of the edema were ob-served. After 2 months, his body weight of11.5 kg placed him in the 20th percentile.The1-year follow-up revealed normal developmentafter having been maintained on a milk-basedpediatric nutritional supplement containingmicronutrients.

24

Protein-Energy Malnutrition

VIJAYAPRASAD GOPICHANDRAN, LIEN B. LAI, and VENKAT GOPALAN

255

Case 2: Marasmus

A 10-month-old male child was examined dur-ing a health-screening camp for women andchildren in a developing country. The mothercomplained that her child was always hungryand crying for food and was having loose stoolsfor the past 1 month.The striking feature aboutthis irritable and constantly crying child wasthe emaciated appearance, with gross wastingof all muscle groups. The skin was hanging inloose folds and lacked the normal shine andtexture.

The child’s height was 70 cm, and weightwas 5 kg, which were in the 10th and belowthe 5th percentile, respectively, of normalheight and weight for his age. There was noedema. The midarm circumference (MAC) was10 cm, and the triceps skinfold thickness (TSF)was 5 mm. The midarm muscle circumference(MAMC), defined as (MAC, cm) − [π (TSF,mm)/10], was calculated be to 8.5.All these an-thropometric values, when compared to stan-dard tables for age group and sex, were belowthe 5th percentile and indicated wasting, lowfat, and low protein reserve.

On further evaluation in a referral center,the child’s laboratory test results were as fol-lows: 60 mg/dL fasting blood glucose (normalis 70–100 mg/dL); 174 mg/dL 2-h postprandialblood glucose (normal is 120 mg/dL); 1.8 g/dLserum albumin (normal is 3.5–5.5 g/dL);8 mg/dLserum prealbumin or transthyretin (normal is16–40 mg/dL); 0.6 mmol/day urine creatinine(normal is 1–1.3 mmol/day). The urinarycreatinine-to-height index (CHI) was 46% as cal-culated by the formula CHI (%) = [(Actual24–hr creatinine excretion)/(Expected 24–hrcreatinine excretion for height)] × 100. TheWBC count of 3,500 cells/mm3 was low (normalis 4,500–10,000 cells/mm3).The T-cell count at300 cells/mm3 was significantly below the ex-pected value (i.e., 8%–22% of the normal WBCcount). A routine stool examination revealedplenty of fat globules, pus cells, and red bloodcells (10–12/hpf (high-power field)).

The child was diagnosed as suffering frommalnutrition of the marasmus type. Cotrimoxa-zole was given twice daily to treat the intes-tinal infection. Nutritional refeeding wasinitiated with a liquid gruel containing 1 g eggprotein/kg ideal body weight, wheat flourequivalent to 1000 calories/day, and plenty ofvegetables and fruits to replace the vitaminsand minerals. Vitamins A and D were given as

oral supplements.A follow-up analysis 6 monthslater revealed a nearly complete recovery asreflected in the anthropometric measurementsand biochemical tests.

DIAGNOSIS

Signs and Symptoms

The signs and symptoms of protein-energymalnutrition (PEM) depend on various factors,including the duration of the nutritional inade-quacy, age at onset, and frequency/types of con-comitant infections. Figure 24-1 shows childrendiagnosed with kwashiorkor and marasmus andoutlines some of the diagnostic features dis-cussed in this section (Scrimshaw and Behar,1961).

A hallmark of kwashiorkor (absent in maras-mus) is the rotund appearance caused by gener-alized edema in the hands, feet, and abdomen(Fig. 24-1; Hendrickse, 1991). While low bodyweight for height is always found in childrenwith marasmus, fluid retention and edema fre-quently lead to increased weight in kwashior-kor and tend to mask growth failure. Childrenwith kwashiorkor usually have lower serum al-bumin values than those diagnosed with maras-mus. “Flaky paint” dermatitis, discolored andbrittle hair devoid of natural sheen and curl, en-larged fatty liver, growth retardation, and apathyare features associated with kwashiorkor. Theeasy peeling of the epidermis results in infec-tions of the underlying tissues. Periodic varia-tions between poor and acceptable nutritionalstatus lead to the multicolored hair that lookslike a “striped flag” with alternating depig-mented and pigmented portions. According toHendrickse (1991), the mental state and behav-ior of those with kwashiorkor are distinguishedby extreme irritability and the absence of asmile.

Children with marasmus are likely to havebeen weaned early and look extremely thindue to loss of muscle mass and body fat.The dis-ease of hunger is portrayed most poignantly inthe skeletal spectre of marasmic children. Thehead appears disproportionately large, withbones and joints jutting out prominently. Thefacial fat (i.e., the buccal pad) is usually re-tained even when subcutaneous fat elsewhereis being consumed rapidly.These facial featureslend the children an elderly appearance. Mus-cle wasting, weight loss, modest irritability,


frequent infections, an absence of edema, and agrossly shrunken appearance typify marasmus.

Since children suffering from PEM exhibitgrowth retardation, lower scores on develop-mental measures, and weak cognitive perfor-mance, it is imperative for the diagnosis to bemade as early as possible to remedy nutritionalinadequacies and ensure appropriate develop-mental progress. In addition, even though PEMis usually associated with developing countriesor those who are at war or facing famine, it hasbeen found elsewhere. In many instances, foodfads and fallacies in developed countries lead toPEM, and many of these cases may remain ei-ther undetected or unreported as there is littleclinical suspicion of PEM.

General Nutritional Assessment

The tools for nutritional assessment includemedical history and screening aides,physical ex-amination and anthropometric measurements,biochemical assessment, and tests of immunefunction.A general health assessment and med-ical history are required to rule out causes of sec-ondary malnutrition such as poor oral health,chronic illness, disease, and medication. Malnu-trition is influenced by lifestyle, which includesalcohol usage in adults, food preference, eatinghabits, social interactions, and economic status.Various screening tools, such as the DETER-MINE checklist (White et al., 1991), are avail-able to assess the risk of malnutrition.

Protein-Energy Malnutrition 257

Figure 24-1. Typical features found in children with kwashiorkor and marasmus.Photographs reprinted with permission from Scrimshaw and Behar (1961).© 1961, AAAS.

The physical examination in children con-centrates on anthropometric measurementssuch as height, weight, and calculation of thepercentile of the height and weight for age(Gurney and Jelliffe, 1973). Signs of associatedvitamin and mineral deficiencies may also bedetected. The TSF and MAMC values indicatethe body fat and protein reserve, respectively.In adults and the elderly, anthropometrics in-clude weight, height, and body mass index,which is calculated as (Weight in kg)/(Heightin m)2.Taken together, these patient data, whencompared to the average normal values, canhelp draw inferences of stunting, wasting, orboth.

PEM is frequently associated with changes inserum markers such as protein, lipid, and nitro-gen (Hendrickse, 1991). Decreased serum albu-min levels have been used as a standard PEMindicator. Since albumin has a long half-life of18 days in the plasma and a large pool size thatdecreases in various illnesses (such as hepaticinsufficiency and nephrosis), it is not a specificmarker for recent nutritional inadequacies. Pre-albumin (or transthyretin), by virtue of itshigher catabolic rate (half-life = 1.9 days) andsmall pool size, is a very sensitive and fairly spe-cific indicator of acute malnutrition and short-term responses to nutritional replacements inpatients of all ages (Bernstein and Ingenbleek,2002).

While low serum cholesterol levels havebeen observed in malnourished patients, largelyas a result of decreased synthesis of lipopro-teins in the liver, hypocholesterolemia occurslater in the course of malnutrition and is there-fore not useful as a screening test. PEM usuallyresults in low serum urea nitrogen (BUN), uri-nary urea, and total nitrogen.Estimation of 24–hurine creatinine excretion is also a valuablebiochemical index of muscle mass (when thereis no impairment in renal function). The uri-nary CHI is correlated to lean body mass andanthropometric measurements. In edematouspatients, for whom the extracellular fluids con-tribute to body weight and spuriously highbody mass index values, the decreased CHIvalues are especially useful in diagnosing mal-nutrition.

Yet another indication of PEM is usually ob-tained from an analysis of the immunodefi-ciency of the patient. An atrophied thymus inmalnourished patients results in a lower T-cellcount. In addition, macrophages from PEM indi-viduals have decreased interleukin 1 activity, for

which a sensitive (albeit expensive) assay isavailable.


The word “marasmus” is derived from theGreek word marasmos, which means towaste away. Marasmus represents the severearrest in growth that accompanies the sus-tained deprivation of nutrients and is the mostcommon form of malnutrition. Pursuant toCicely Williams’ documentation of kwashior-kor (which means “sickness of weaning con-comitant with birth of the next child” in the Galanguage) in Ghana in the 1930s, marasmusand kwashiorkor were considered distinct forseveral years until the realization that bothare associated with malnutrition. Since proteinsupplementation alone did not significantly re-duce the unfortunately high occurrence ofmalnutrition-associated deaths in different partsof the world, the term protein-energy or -caloriemalnutrition (PEM or PCM) was coined to em-phasize factors other than protein in the diet(Golden, 2002). PEM now encompasses threeclinical phenotypes: marasmus, marasmic-kwashiorkor, and kwashiorkor.

The finding of kwashiorkor in parts of theworld where starchy, watery gruels derivedfrom cassava, maize, plantain, and yam are usedas infant foods (immediately after weaning) ledto the view that kwashiorkor results from high-carbohydrate diets, rich in calories but poor inprotein. It is becoming increasingly evident thatfactors other than nutritional deficiency mightdictate the onset and progression of kwashior-kor (Golden, 2002; see Etiology section). In con-trast, marasmus is clearly associated with a dietdeficient in calories but balanced in nutrientcomposition.

Metabolic Alterations and Causative Hormonal Changes in Protein-Energy Malnutrition

When food intake decreases, the utilization offat and protein reserves in the body enables var-ious essential metabolic processes to continueduring the nutritional inadequacy. In the earlystage of fasting or starvation, glucose require-ments of the brain and nervous system are ful-filled by mobilization of glycogen in the liver.This short-term adaptation lasts only a day untilglycogen stores are exhausted. Gluconeogenesis


is then initiated in the liver, using as precursorscarbon backbones obtained from deaminationof amino acids, which in turn were obtainedfrom the breakdown of muscle proteins.

Adaptation to decreased food intake alsoactivates a hormone-sensitive lipase, which hy-drolyzes triglyceride reserves in fat cells to gen-erate fatty acids and glycerol. The fatty acidsprovide substrate for the citric acid cycle andfurnish the carbon backbone for the synthesisof ketone bodies in liver.Once their levels stabi-lize, ketone bodies can serve, in part, as an alter-native fuel in place of glucose for the brain, andmuscle protein is no longer degraded as exten-sively to furnish precursors for gluconeogene-sis. Since unremitting muscle protein loss(particularly in respiratory tissue) will be fatal,the utilization of ketone bodies by the brainhelps to spare lean body mass until the subcuta-neous fat tissue in the body is depleted. De-creased resting energy expenditure also servesto minimize protein turnover in PEM (Jackson,2003).

Although the proportionate shortage of allmacronutrients in marasmus ensures home-ostasis to a large extent, there is a severe meta-bolic derangement in kwashiorkor. Since thediet in kwashiorkor is deficient in protein, it isnot surprising that decreased de novo proteinsynthesis grossly affects the functioning ofvarious organs. In the liver, the failure to syn-thesize transport proteins results in the depriva-tion of lipids, vitamins, and other micronutrientsto peripheral tissues. Deficiency in key meta-bolic enzymes also impairs several hepaticprocesses, including fatty acid oxidation, bilesalt formation, and detoxification of xenobi-otics. Moreover, the absence of pancreatic di-gestive enzymes leads to poor digestion, andmalabsorption of fat and fat-soluble vitaminsmanifests as loose stools (steatorrhea). Thesedrastic changes do not occur in marasmus (ex-cept after prolonged starvation) since theamino acid pool derived in part from musclecatabolism and in part from the modest di-etary intake provides precursors for synthesisof key hepatic proteins such as albumin andlipoproteins.

The hormonal fluxes that orchestrate metab-olism among various organs and tissues alsovary between these two forms of PEM and con-tribute to their respective clinical phenotypes(Torun and Chew, 1999). In the case of maras-mus, the overall decreased food intake (i.e., re-duced concentration of blood glucose and free

amino acids) causes a decline in insulin levelswith a concomitant increase in catabolic hor-mones such as glucagon, epinephrine, and cor-tisol. Glucagon activates glycogenolysis andhelps increase blood glucose concentration. Ep-inephrine activates lipolysis in adipocytes tosupply fatty acids for oxidation in other tissuesand glycerol for gluconeogenesis. In the liver,fatty acid oxidation provides acetyl-CoA as asubstrate for the synthesis of ketone bodies.Cortisol triggers proteolysis, specifically of mus-cle protein, to generate amino acids for use bythe liver as precursors for gluconeogenesis andfor de novo synthesis of serum proteins. Highcortisol levels also cause diminished sensitivityof the tissues to insulin as well as a decreasein glucose utilization by directly inhibiting glu-cose transport.

The carbohydrate-rich, protein-deficient dietof kwashiorkor causes, at least initially, an in-crease in the anabolic hormone insulin, an un-desirable situation in a malnourished state.Despite an energy shortage, insulin inhibitslipolysis and promotes lipogenesis (fatty acidsynthesis) using the excess carbohydrate inthe diet as precursors. The increase in insulinis accompanied by dampening of the adrener-gic response that prevents fat and proteinbreakdown and by a decreased mobilization ofmuch needed precursors for de novo serumprotein synthesis.Although initially stimulated,insulin secretion decreases after prolongedprotein deficiency in kwashiorkor due to ex-haustion and atrophy of the pancreatic betacells. Glucose intolerance is therefore observedin kwashiorkor but not in marasmus. The de-creased insulin results in palliative endocrinechanges (such as increased glucocorticoid lev-els) that stimulate increased lipolysis and mus-cle protein catabolism. However, musclewasting with little or no dietary intake of es-sential amino acids cannot entirely amelioratedecreased hepatic protein synthesis in kwashi-orkor.

Both forms of PEM are associated with hy-percortisolemia.The level of cortisol in kwashi-orkor is lower, however, than in marasmus,likely due to decreased adrenocortical functioncaused by low protein intake (and not adrenalfailure). If a sufficiently high level of cortisol isnot maintained, then adequate muscle proteinis not mobilized to sustain hepatic protein syn-thesis. Indeed, hypoproteinemia, evident by thedecreased serum albumin and transferrin levels,is more acute in kwashiorkor than marasmus.


The reduced levels of serum lipoproteins neededfor lipid transport also contribute to the “fattyliver”usually observed in kwashiorkor.

Pathophysiology

In conserving energy, cellular processes arecompromised, aggravating the effects of malnu-trition. Decreased rate of cellular replicationcauses atrophy of the intestinal mucosal cellsand leads to both bacterial colonization of thegut and poor bioavailability of nutrients.WhenATP-requiring reactions are downregulated,processes such as the transport of nutrientsacross the intestinal lumen or pumping of ionsacross the cell membrane are perturbed signifi-cantly. In the absence of an active Na+/K+

ATPase pump, the concentration gradient favorspotassium efflux and sodium influx as borneout by the increased intracellular sodium anddecreased potassium that are observed in PEM.Membrane leakiness due to alteration of lipidcomposition may also underlie this change. Infact, recently designed rehydration regimens forPEM patients provide less sodium and morepotassium on account of this imbalance (seeTherapy section). Economizing energy also in-vokes oxidative stress. Due to reduced synthe-sis of new red blood cells (RBC) and ferritinduring PEM, the iron released from degradedRBC is not sequestered.The resulting free ironpromotes the dangerous generation of freeradicals.

During the last four decades, the interde-pendence between malnutrition, decreased im-munity, and infection has become increasinglyclear (Jolly and Fernandes, 2000). Since PEMwas initially attributed entirely to dietary insuf-ficiencies, it was thought that enrichment ofthe diet was the key. It soon became clear, how-ever, that food consideration alone could notadequately explain all facets of malnutrition.In 1959, Scrimshaw and coworkers proposedthat malnutrition makes a person susceptibleto infections, which in turn perpetuates themalnutrition state. In fact, the high mortalityobserved in PEM has been attributed to the in-creased susceptibility to infections arising fromimmunodeficiency. These observations haveled to a two-prong therapeutic modality forchildren with malnutrition: stem the infectionwhile enhancing the nutritional status. Ad-vances in immunology have now facilitated acomprehensive understanding of the defensemechanisms of the body and placed decreased

immunity in the vicious triangle concomitantwith malnutrition.

Malnutrition causes a breakdown in theseemingly invincible array of defense mecha-nisms and renders the host highly susceptibleto frequent infections and even malignancies.There is a dramatic atrophy of the thymus anda decrease in the number of functionally ma-ture and differentiated T cells (Jolly and Fernan-des, 2000). Moreover, lower cytokine activityresults in reduced proliferation of T cells. Al-though circulating B-lymphocyte levels areusually normal, response to vaccine antigensis low in PEM, as might be expected from thereduced antigen presentation by T cells. Thebreakdown in mucosal immunity, as reflectedin the decreased secretory immunoglobulin Alevels, might account for the increased risk ofrespiratory and gastrointestinal tract infectionsusually observed in PEM.There is also a reduc-tion in the components of the complementsystem in PEM and as a consequence the abil-ity of macrophages to phagocytose variouspathogens is impaired.

How does infection exacerbate malnutri-tion? Infection is a catabolic state accompaniedby various metabolic alterations. Calorigenesis(increased caloric usage), required for pro-ducing fever, utilizes the already limiting storesof glucose, fats, and amino acids. Infectionleads to increased inflammation and repair,which demand further use of reserves andcauses nutrient losses. The liver uses valuableamino acids from muscle breakdown to syn-thesize acute-phase reactants, such as the tu-mor necrosis factor, instead of serum proteins.The acute-phase reactants also lead to cachexia(general wasting and decrease in body mass).Albeit counterintuitive, malnourished indi-viduals also lose appetite for food, allowingthe cycle of infection and malnutrition tobe self-perpetuating. Halting cytokine-inducedcachexia merits consideration as a future phar-macological approach to circumvent thisproblem.

The synergy between illness and malnutri-tion is clear. Malnutrition does not merely ar-rest development but is life threatening forchildren in many developing countries. In fact,recent studies examining the risks of dying fromunderweight and malnutrition concluded thata large number of deaths from pneumonia,diarrhea, malaria, and measles could be pre-vented by eradication of undernutrition amongchildren.


Etiology

PEM arises due to a negative energy balance,that is, a combined intake of proteins and calo-ries less than that required for body expendi-ture. Inadequate food intake in PEM may havemultiple causes, ranging from secondary malnu-trition to sociopolitical problems (Fig. 24-2).

There has been considerable debate over thepathogenesis of kwashiorkor and marasmus(Golden, 2002).The edema unique to kwashior-kor was initially attributed to the low plasmaprotein content and the resulting decrease inplasma oncotic pressure. Results from variousstudies in the last few decades challenged thisnotion and suggested consideration of a multi-factorial basis for the etiology of kwashiorkor(reviewed by Golden, 2002).Yet, the concept ofan acute protein deficiency with normal caloricintake as the underlying basis survived due tothe absence of a substitute paradigm.

In 1968, Gopalan (who does not share an-tecedents with an author of this chapter) pro-

posed a new theory of “dysadaptation” and wasprescient in drawing early attention to the pos-sible role of nondietary factors in the etiologyof kwashiorkor. Finding no significant differ-ence in the dietary patterns between kwashi-orkor and marasmic children in India, heconcluded that only children with kwashiorkorsuffer from a form of dysadaptation to the unfa-vorable environment, and that amino acidswere being shunted from synthesis of visceralproteins to synthesis of acute-phase reactants.The resulting decrease then causes reducedplasma oncotic pressure and edema.Other chal-lenging findings include observations such asthe inability to induce kwashiorkor in animalsfed low-protein diets, the cure of children withkwashiorkor by a low-protein diet and withcomplete resolution of edema in the absence ofan increase in serum albumin levels, and breast-fed infants suffering from kwashiorkor despitehealthy mothers.

The observation that glutathione, a free-radicalscavenger, was reduced in the erythrocytes of


Disasters:personal,natural,man-made,war

Congenital defects:prematurity,metabolic errors,anatomical gastrointestinal defects,mental retardation

Increased diet needs(individual variation):growth, injury,pregnancy, illness,location, work

Maldistribution of wealth,underdevelopment,improvidence

Lack of dietary adaptation:emotional,somatic

Diarrhea,dysentery,parasites,toxins

Other diseases:infections,allergies

Bacterial changes,enzyme changes,intestinal atony,intestinal atrophy

Physical,biochemicalbalancepresentation

QualityQuantityTiming

Lack of education,poor sanitation

Anorexia

Gastrointestinal diseases(origin doubtful):sprue, colitis, etc.

Food habits,traditions

IgnorancePoverty Carelessness

Inadequate food intake

Malnutrition

Contaminated food

Malabsorption

Maternal deprivation,psychological disorders

Figure 24-2. Multiple etiologies of protein-energy malnutrition. Reprinted fromWilliams (1986) with permission from Elsevier.

children with kwashiorkor but not marasmusled Golden and Ramdath (1987) to propose anew pathogenesis model in which oxidativestress was ascribed a central role. They postu-lated an overproduction of reactive oxygenspecies and inadequate scavenging of these freeradicals in kwashiorkor. The sophisticated an-tioxidant system designed to counter the ad-verse effects of free radicals was hypothesizedto fail in its objective largely due to the unavail-ability of protein, micronutrients, and vitamins.In light of this model, it is of interest to examinethe oxidative stress in the two forms of PEM.

Although glutathione is specifically decreasedin kwashiorkor, blood levels of selenium-dependent glutathione peroxidase (a scavengerof peroxides) and vitamins A, C, and E (all mem-bers of the antioxidant machinery) are lowerin both kwashiorkor and marasmus (Ashouret al., 1999). Why then are marasmic children,also deficient in some antioxidants, spared theoxidative stress? Does a weakened antioxidantdefense manifest as a serious threat only in thepresence of pro-oxidant activities of the type en-countered in kwashiorkor? What is a possibletrigger for the increase in free radicals, and howmight this account for some of the phenotypicalterations in kwashiorkor?

Unlike marasmic or well-nourished children,those with kwashiorkor have low transferrin lev-els and detectable free iron in the plasma (Siveet al., 1996). Uncomplexed iron is extremelytoxic due to its ability to generate free radicalsby the Haber-Weiss and Fenton reactions.The on-slaught of opportunistic infections in the mal-nourished also elicits production of free radicalsand accentuates the oxidative stress.

Free radicals through lipid peroxidation cancause membrane damage, induce electrolyteimbalance and edema. Indeed, children withkwashiorkor display low levels of polyunsatu-rated fatty acids (e.g., linoleic acid) in the ery-throcyte membrane compared to the marasmicchildren, presumably due to increased lipid per-oxidation (Leichsenring et al., 1995). Interest-ingly, cysteinyl leukotrienes, which can causeedema by altering capillary permeability, arealso enhanced in those with kwashiorkor butnot marasmic children (Mayatepek et al., 1993).

Studies indicated that cysteine supplementa-tion is beneficial in restoring glutathione levelsin children with severe edematous malnutrition(Badaloo et al., 2002). If the reduced rate of glu-tathione synthesis in kwashiorkor is convinc-ingly attributed to a shortage of protein in the

diet, then the new findings on oxidative stresscould indeed be gratifyingly integrated with thelong-held belief that kwashiorkor is at least at-tributable in part to a protein deficit.

Despite the advances enumerated in this sec-tion, the conundrum posed by Gopalan (1968)regarding whether the same diet could lead toboth kwashiorkor and marasmus remains. Prof-itable future investigations might entail an exam-ination of the genetic contributions to oxidativestress and the consequent phenotypic sequelaewhile on an inadequate diet.

THERAPY

The high mortality rate observed with severemalnutrition has been attributed to errors ofmanagement precipitated by the lack of an inte-grated treatment for the various problems.A pri-mary and unfortunately frequent mistake hasbeen the injudicious approach of correctingweight loss without managing illness and re-pairing the highly damaged cellular machinery.The advances in our understanding of PEM to-gether with the documented failures from priorapproaches have led the World Health Organi-zation (WHO) to issue a manual for managingsevere malnutrition. The principles of clinicalmanagement elaborated by WHO consist of 10steps that encompass an initial treatment phaseaimed to resolve life-threatening situations anda subsequent rehabilitation phase that focuseson reestablishing good nutritional status andcatching up lost growth (Table 24-1).

Resolving Life-Threatening Situations: The Initial Treatment Phase

PEM is often associated with two specific sets ofphysiological problems: (1) infections,hypother-mia, and hypoglycemia; and (2) fluid and elec-trolyte derangements ( Jackson,2003).Treatmentof infection plays an important role in PEM man-agement. Broad-spectrum antibiotics like cotri-moxazole should be used for addressing gut,respiratory, or urinary infections. In addition toimmunization for measles,antihelminthic tablets(e.g., mebendazole) for deworming the gut andantimalarial tablets (e.g., chloroquine) for com-bating malarial parasites are prescribed in thetropical countries. Loss of body mass con-tributes to reduced thermal insulation. Themalnourished patient presenting with a body


temperature less than 35.5°C is in imminentdanger and needs to be immediately warmed.The depletion of glucose that results either fromcombating infections or increasing body tem-perature causes hypoglycemia and reduces thesupply of glucose to the brain.A multiprong ap-proach of combating infections, reducing theheat loss and supplying glucose by intra-venous/oral means is absolutely essential to res-cue the patient from the medical emergencycaused by severe malnutrition.

The signs of malnutrition can occasionallymask the clinical signs of hypovolemia (de-creased volume of circulating blood). For exam-ple, the laxity of skin, which is a good marker ofhypovolemia, cannot be relied on in the case ofthe marasmic child, who has an inelastic skin.Valuable indicators of hypovolemia include ex-aggerated thirst, dry mouth, decreased urinaryoutput, tachycardia (abnormally rapid heart-beat) with a thready pulse, hypotension withorthostasis (dizziness caused only on standing),and declining levels of consciousness. Rehydra-tion should be initiated with a modified versionof the original WHO-recommended oral rehy-dration salt (ORS) solution that has an osmolar-ity of 300 mOsm/L and contains 125 mMglucose, 45 mM sodium, 40 mM potassium,70 mM chloride, 7 mM citrate, 3 mM magne-sium, 0.3 mM zinc, and 0.045 mM copper. Forinfants, breast-feeding should be continued dur-ing rehydration. The fluid replacement shouldcontinue at the rate of 70–100 mL ORS/kg bodyweight until a urinary output of 200 mL/day in

children (and 500 mL/day in adults) is reached.If the patient has incessant vomiting and is un-able to tolerate ORS, then nasogastric fluid re-placement should be considered.

Reestablishing Nutritional Status:The Rehabilitation Phase

The atrophied gut mucosa and the pancreaticinsufficiency will contribute to an intoleranceof rapid and vigorous refeeding regiments. Infact, it is well documented that malabsorptioncaused the high mortality among prisoners ofwar who were starved for long durations andsubjected to rapid refeeding on release. Nutri-tional replacement is best done orally and intwo stages: the therapy stage, which starts withlow doses of foods to restore vital cellular func-tions, and the rehabilitation stage,which aims atcatching up and regaining nutritional adequacy.

Formula diets F-75 and F-100 are comprisedof defined amounts of dried skimmed milk,sugar, cereal flour, oil, minerals, and vitamins de-signed both for the initial (F-75, 75 kcal/100 mL) and the rehabilitation (F-100, 100 kcal/100 mL) stages. While the initial stage shouldentail caloric intake of 80–100 kcal/kg bodyweight/day, the rehabilitation stage should be-gin with about 130 kcal and gradually increaseto 150–200 kcal/kg body weight/day. The re-turn of a normal appetite signals the clearanceof infections and the end of the initial stageand permits the switch to rehabilitation. Judi-cious and careful refeeding is essential for ensur-


Table 24-1. Time Frame for Management of a Child with Severe MalnutritionInitial Treatment Rehabilitation Follow-up

Activity Days 1–2 Days 3–7 Weeks 2–6 Weeks 7–26

Treat or prevent:

Hypoglycemia ––––––––––→Hypothermia ––––––––––→Dehydration ––––––––––→

Correct electrolyte imbalance –––––––––––––––––––––––––––––––––––––––––→Treat infection –––––––––––––––––––––––→Correct micronutrient deficiencies ←−− Without iron −−→ ←−− With iron −−→

––––––––––––––––––––––––––––––––––––––––––→

Begin feeding –––––––––––––––––––––––→Increase feeding to recover lost weight

(“catch-up growth”) –––––––––––––––––––––––––––––––––––→Stimulate emotional and sensorial

development –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––→Prepare for discharge –––––––––––––→Reproduced with permission from World Health Organization (1999).

ing a speedy and complete recovery.Food shouldbe given in small amounts and frequently; therecommended intakes can be found in theWHO manual.

If formula diets are not used, then the type ofprotein used is important due to considerationsof biological value and digestibility. Milk, animalproteins, egg, certain legumes, and soy productsare appropriate, rich sources. If lactose intoler-ance is encountered, then milk should be re-placed with soy-based feeds or an alternative.Fish and vegetable oils are good sources of fats asthey provide generous amounts of essential fattyacids and long-chain unsaturated fatty acids.

In addition to proteins and calories, the pa-tients with PEM must be replenished with re-spect to vitamins and micronutrients especiallyto enhance antioxidant status. Due to heavy vi-tamin A losses that occur during infections andPEM, vitamin A supplements are necessary toprevent blindness. Selenium, zinc, manganese,and cobalt are important micronutrients defi-cient in patients with PEM and therefore needto be replaced.These micronutrients play a vitalrole in the function of several enzymes (e.g.,selenium in glutathione peroxidase). Iron mustbe excluded from the diet in the initial stagesince early administration increases the risk offree-radical production and infection.

Psychosocial Support

Children with PEM frequently do not realizetheir full mental and developmental potentialand display mental apathy, lethargy, and psy-chomotor retardation. Developmental and psy-chosocial rehabilitation are needed togetherwith restoration of nutritional status. Modestphysical activity that helps in rebuilding leanmass and stimulatory activities that facilitate re-covery of brain function should be pursued af-ter the patient has been rescued from the initialemergency phase.A caring and supportive envi-ronment is absolutely necessary for ensuringthe emotional well-being and expediting recov-ery to normalcy.

QUESTIONS

1. Is it necessary to identify and differentiatebetween the two types of PEM while for-mulating a plan of treatment? Explainyour reasoning.

2. Draw a comparison between energy me-tabolism in a child with PEM and anotherfed with excess carbohydrates and pro-teins.

3. For the adverse factors complicating mal-nutrition such as oxidative stress discussedin the chapter,explain what genetic factorswould further exacerbate the problem.

4. Postulate biochemical mechanisms for thedeterioration of health that occurs afterrapid refeeding of a child with PEM.

5. If a child is fed a diet that is poor in pro-teins but rich in carbohydrate and antioxi-dants, would the edema associated withkwashiorkor be prevented?

6. Do you expect hibernating animals toshow signs of PEM under any circum-stances?

BIBLIOGRAPHY

Ashour MN, Salem SI, El-Gadban HM, et al.: Antiox-idant status in children with protein-energy mal-nutrition living in Cairo, Egypt. Eur J Clin Nutr53:669–673, 1999.

Badaloo A, Reid M, Forrester T, et al.: Cysteine sup-plementation improves the erythrocyte glu-tathione synthesis rate in children with severeedematous malnutrition. Am J Clin Nutr 76:646–652, 2002.

Bernstein LH, Ingenbleek Y: Transthyretin: its re-sponse to malnutrition and stress injury. Clinicalusefulness and economic implications. ClinChem Lab Med 40:1344–1348, 2002.

Carvalho NF, Kenney RD, Carrington PH, et al.: Se-vere nutritional deficiencies in toddlers resultingfrom health food milk alternatives. Pediatrics107:E46, 2001.

Golden MHN:The development of concepts of nu-trition. J Nutr 132:2117S–2122S, 2002.

Golden MHN, Ramdath D: Free radicals in thepathogenesis of Kwashiorkor. Proc Nutr Soc46:53–68, 1987.

Gopalan C: Kwashiorkor and marasmus: evolutionand distinguishing features, in McCance RA,Widdowson EM (eds): Calorie Deficiencies andProtein Deficiencies. Little, Brown and Com-pany, Boston, 1968, pp. 49–58.

Gurney JM, Jelliffe DB:Arm anthropometry in nu-tritional assessment: normograms for rapid cal-culation of muscle circumference and crosssectional muscle and fat mass. Am J Clin Nutr26:912–915, 1973.

Hendrickse RG: Protein-energy malnutrition, inHendrickse RG, Barr DGD, Mathhews,TS (eds):Paediatrics in the Tropics. Blackwell Scientific,Oxford, UK, 1991, pp. 119–131.


Jackson AA: Severe malnutrition, in Warrell DA,Cox TM, Firth JD, et al. (eds):Oxford Textbook ofMedicine. 4th edition Oxford University Press,Oxford, UK, pp. 1054–1061, 2003.

Jolly CA, Fernandes G: Protein-energy malnutritionand infectious disease, in Gershwin ME,GermanJB, Keen CL (eds): Nutrition and Immunology:Principles and Practice. Humana Press,Totowa,NJ, 2000, pp. 195–202.

Leichsenring M, Sutterlin N, Less S, et al.: Polyun-saturated fatty acids in erythrocyte and plasmalipids of children with severe protein-energymalnutrition. Acta Paediatr 84:516–520, 1995.

Mayatepek E, Becker K, Gana L, et al.: Leukotrienesin the pathophysiology of kwashiorkor. Lancet342:958–960, 1993.

Scrimshaw NS, Behar M: Protein malnutrition inyoung children. Science 133:2039–2047, 1961.

Scrimshaw NS, Taylor CE, Gordon JE: Interactionsof nutrition and infection. Am J Med Sci 237:367–372, 1959.

Sive AA, Dempster WS, Malan H, et al.: Plasma freeiron: a possible cause of oedema in kwashior-kor. Arch Dis Child 76:54–56, 1996.

Torun B, Chew F: Protein-energy malnutrition, inShils ME (ed): Modern Nutrition in Health andDisease. 9th ed.Lippincott,Williams and Wilkins,Baltimore, MD, 1999, pp. 963–988.

White JV, Ham RJ, Lipschitz DA, et al.: Consensus ofthe Nutrition Screening Initiative: Risk factorsand indicators of poor nutritional status in olderAmericans. Journal of the American DiabeticAssociation 91:783–787, 1991.

Williams SR: Nutritional deficiency diseases, inWilliams SR (ed): Nutrition and Diet Therapy.5th ed. Elsevier, Philadelphia, 1985, pp. 327–350.

World Health Organization: Management of Se-vere Malnutrition: A Manual for Physiciansand Senior Health Workers. WHO, Geneva,Switzerland, 1999.


CASE REPORT

A 6-year-old African American boy presentedwith a 6-month history of nausea,abdominal dis-comfort, and diarrhea with three to four waterystools a day associated with passage of muchgas. These complaints began soon after hestarted first grade when he was required todrink 2 cups of milk with his school lunch.Thesymptoms occurred after lunch, and he feltmuch better after having a bowel movement.Noblood was noted in the stool. He was mostlysymptom free on weekends, when he rarelydrinks milk. There has been no fever or noweight loss,and his appetite has remained good.

In general, the boy has been healthy sincebirth at full term. He was breast-fed for 6months, at which time he was weaned to regu-lar infant formula, and started taking solids inthe form of rice cereal. He began eating tablefood at 1 year of age.Neither parent drinks muchmilk. He is the only child of both parents, andthere are no other family health concerns.

The patient was brought to the clinic forevaluation directly from school and soon afterlunch. Physical examination revealed a pleasantand healthy looking young boy. His growth anddevelopment were appropriate, with both hisheight and weight in the 50th percentile on thegrowth chart. He had a normal temperature,pulse and respiratory rates, and blood pressure.The only positive findings on systematic exami-nation were in the abdomen, which was mildlydistended with increased bowel sounds andtympanicity on percussion (the abdomen is res-onant when tapped with fingers or an instru-ment because of distension of the abdomen

with air or gas).There was no tenderness. Nei-ther the liver nor the spleen was palpable, andno other mass was palpated.There was mild peri-anal excoriation (destruction and loss of thesurface of the skin around the anus), and rectalexamination revealed loose stool in the rectum.The stool was negative on testing for occultblood (blood present in such small quantitiesthat it can only be detected by chemical test-ing). Further examination of a stool specimenin the laboratory showed no red or white bloodcells, Rotazyme stool testing for rotavirus wasnegative, and no bacterial pathogen was identi-fied on stool culture.The stool pH was 5.5, andClinitest (for detecting reducing sugar in stool)was positive. Blood cell count and serum elec-trolytes were normal.

As this patient’s history indicated that hisproblems began soon after he started taking 2cups of milk with lunch at school and his physi-cal examination and stool examination findingssuggested he had carbohydrate malabsorption,a diagnosis of lactose intolerance was made. Itwas recommended that his milk intake at lunchbe decreased to half a cup (about 4 oz). Hissymptoms improved considerably and almostimmediately. His milk intake was then graduallyincreased such that he was ultimately able totake up to 8 oz of milk without much complaint,especially when the milk was taken with otherfoods.

DIAGNOSIS

It is necessary to confirm the clinical impres-sion of lactose intolerance by specific tests

25

Lactose Intolerance

MARCY P. OSGOOD and ABIODUN O. JOHNSON

266

based on measurement of products of lactosedigestion or fermentation. Such tests demon-strate lactose maldigestion and malabsorptionand may be direct or indirect. In the direct test,lactase activity in a small intestinal mucosalspecimen is determined by enzymatic assay.Thesmall intestinal specimen is obtained by biopsy.However, because this test is invasive, it has noplace in routine clinical management; it is usednow essentially in research situations.The indi-rect tests include testing for stool acidity andpresence of lactose, a reducing sugar; measure-ment of blood glucose levels in the lactose tol-erance test; and the determination of breathhydrogen levels.

In the lactose tolerance test, an oral dose of2 g lactose per kilogram body weight (maxi-mum dose of 50 g in adults) is administered tothe patient after an overnight fast and afterdrawing venous blood for fasting blood sugarbaseline. Serial venous blood samples are thentaken every half hour for 2 h for measurementof blood glucose levels. The test is positive ifthe blood glucose level increases less than20 mg/dL above the fasting level, indicating thelack of hydrolysis of lactose to glucose andgalactose and therefore little or no transport ofglucose into the bloodstream from the small in-testine. Also, symptoms of lactose intolerancemay occur during the test, but this is not neces-sary to make the diagnosis.

False-positive and false-negative results occurin up to 20% of subjects due to variable gastricemptying times and individual differences inglucose metabolism. It is sometimes necessaryto demonstrate that the lactose tolerance test istruly positive because of failure to hydrolyzelactose and not due to a mucosal absorptive dis-order. This is done by repeating the test afterthe patient drinks an aqueous solution contain-ing half of the carbohydrate dose as glucoseand the other half as galactose.The blood sugarlevel should now rise at least 20 mg per dLabove the fasting level.

The hydrogen (H2) breath test is generallyconsidered the diagnostic method of choice be-cause it is sensitive, noninvasive, and relativelyinexpensive. Normally, humans exhale very lit-tle H2 gas as H2 is not a product of human me-tabolism. However, H2 and other gases areproduced when undigested lactose in the colonis fermented by bacteria. The H2 is absorbedand carried through the bloodstream to thelungs, where it is exhaled and can be detected.Raised H2 levels in exhaled air indicate insuffi-

cient digestion of lactose in the small intestineor the presence of unusually large numbers ofH2-producing bacteria in the colon. After anovernight fast, a baseline sample of exhaledbreath is collected in a balloonlike collectionbag. The patient then drinks an aqueous solu-tion of lactose at a dose of 2 g/kg body weight(maximum dose 50 g). Serial 15-min breathsamples are then collected, and the H2 level ineach collection bag is determined by gas chro-matography. A rise in breath H2 concentrationgreater than 20 ppm over the baseline level sug-gests hypolactasia. The hydrogen breath test ispositive in about 90% of patients with lactosemalabsorption. Gastrointestinal symptoms anda positive H2 breath test have been shown to bestrongly correlated (Hermans et al., 1997).

A false-negative test can occur in the absenceof colonic bacteria (often due to recent use ofantibiotics or a high colonic enema), absence ofany H2-producing colonic bacteria,or overabun-dance of methane-producing bacteria in thegut. Sleep deprivation, exercise, use of aspirin,gum, or mouthwash, or smoking may increasebreath H2 secretion independent of lactose loador hypolactasia. Again, a stool pH equal to orless than 5.6 and the presence of reducingsugar in the stool by Clinitest are important an-cillary diagnostic findings needed for diagnosis.

Our patient manifested most of the aboveclinical and laboratory findings, which im-proved when the amount of milk (and lactose)he ingested was reduced. He had tolerated bothbreast milk and cow milk-based formula in in-fancy and developed symptoms only when hewas induced to drink relatively large amountsof regular milk at one sitting.Thus, he had thevery common primary, hereditary, late-onsettype of lactose intolerance. This type is evenmore pronounced in adults, resulting in thecommon nomenclature, “adult-type” lactose in-tolerance, which is really a misnomer since thiscondition can start anytime after the weaningperiod. Our patient was gradually able to toler-ate about 1 cup of milk a day with his meal.

Because the symptoms of lactose intoleranceare nonspecific, lactose intolerance needs to bedifferentiated from other common disorderscausing diarrhea and abdominal pain or discom-fort. These differential diagnoses include milkprotein allergy, gastroenteritis, colitis, and irrita-ble bowel syndrome. Allergy or sensitivity toone or more of cow’s milk proteins characteris-tically occurs in formula-fed infants, and al-though no age is exempt, it is rare in the older

Lactose Intolerance 267

child and in adults. Gastroenteritis is commonlycaused by toxins or by infections by viruses orbacteria. In infectious gastroenteritis, the stoolcontains white blood cells and may containgross or occult blood,especially if it is caused byan invasive bacterial pathogen such as Shigella,Salmonella, or enteroinvasive Escherichia coli.These pathogens can be cultured from thestool. In colitis, the stools are loose, mucousy,and bloody, and the characteristic intermittentcrampy lower abdominal pain is often relievedby defecation. Irritable bowel syndrome is char-acterized by the alternation of diarrhea andconstipation.

The diagnosis of lactose intolerance shouldbe strongly suspected ahead of any of theseother differential diagnoses when a patient pres-ents with the characteristic symptoms enunci-ated above following ingestion of lactose-containing drinks or foods such as milk, parti-cularly if the patient belongs to an at-risk popu-lation.


General Biochemistry of Lactose as a Dietary Sugar

Carbohydrates in various forms make up a sig-nificant portion of the human diet in mostparts of the world. Dietary carbohydrates rangefrom the simple polyhydroxy aldehydes andketones such as glucose, galactose, and fruc-tose, to the large polysaccharides of glucose,starch, and glycogen.Also important are the di-etary disaccharides, which include lactose (β-D-galactopyranosyl-(1→4)- α-D-glucopyranose),the main carbohydrate in mammalian milk(Fig. 25-1); sucrose (α-D-glucopyranosyl-(1→2)-β-D-fructofuranoside) from sugar beet and sugarcane;maltose (α-D-glucopyranosyl-(1→4)- α-glu-copyranose) from corn syrup and naturallyoccurring in several fruits; and, less com-monly, trehalose (α-D-glucopyranosyl-(1→1)-α-D-glucopyranoside) from mushrooms, insects,and some seaweeds.

Sugars capable of reducing ferric or cupricion are called reducing sugars; lactose, whichhas a free anomeric carbon (not involved in aglycosidic bond), is a reducing sugar, as is malt-ose. Sucrose and trehalose, both of which haveno free anomeric carbon, are nonreducing sug-ars. Reduced digestion, transport, or absorptionof any of the dietary carbohydrates can cause

mild-to-severe symptoms of carbohydrate mal-absorption.

Lactose is only found in high quantities inmammalian milk, which is an essential foodfor suckling mammals. Lactose milk concentra-tion varies from almost undetectable in themilks of some marine mammals to as much as7 g/100 mL in mature human milk. Once in-gested, lactose is digested to its componentmonosaccharides by the enzyme lactase, whichis located in the brush border (microvilli) of thesmall intestine mucosal cell (enterocyte).Trans-port of the released glucose and galactose oc-curs through sugar transporters on the smallintestine brush border (Fig. 25-2).

Forms of Lactose Intolerance

Lactase deficiency leading to lactose intoler-ance may be congenital, primary (hereditary,delayed onset, adult), or acquired. Congenitallactase deficiency is a very rare condition innewborn babies, and it is inherited by an auto-somal recessive mechanism. It is characterizedby total absence of intestinal lactase activity(alactasia). Diarrhea and gaseous abdominal dis-tension occur soon after the newborn is fedbreast milk or milk-based infant formula. Be-cause lactase activity is vital in all baby mam-mals in order to digest the lactose in milk, aninfant’s inability to utilize its primary food (milk)and the dehydration and discomfort caused bythe lactose malabsorption can lead to a severeillness that can be fatal if not diagnosedpromptly and if lactose is not excluded fromfurther feedings. Cases of congenital alactasiahave been reported mostly in Finland, and it isone of the “Finnish recessive disorders” (Jarvelaet al., 1998).

Primary, hereditary delayed-onset, or adultlactose intolerance (also referred to as lactasenonpersistence) is the most familiar form ofthis condition. Lactase nonpersistence is actu-ally much more common worldwide than adult


Lactose

CH2OHO

OH

H

OHH

HO

H

OH

H

CH2OHO

H

H

OH

OHH

H

OH

H

Figure 25-1. Lactose.

lactase persistence. Developmental expressionof lactase is tightly controlled. Intestinal lactaseactivity rises relatively late in the fetus, reachingoptimal levels at term; thus, the preterm new-born has reduced lactase levels, and the morepremature the baby is, the greater the risk of re-duction in lactase level. Intestinal lactase activ-ity remains at peak levels from birth in thefull-term baby through infancy and early child-hood, when the diet is mostly milk and lactose-containing milk products. Lactase activity fallsgradually with reduction in milk intake and theintroduction of non-lactose-containing weaningfoods. Intestinal lactase levels become quitelow after the child is well established on cerealsand by the time weaning is complete.The abil-ity to digest lactose becomes compromisedthereafter, and this type of lactose intolerance iswell established by 3–7 years of age dependingon the weaning practices of the particular pop-ulation (Scrimshaw and Murray, 1988). As lac-tase is not an inducible enzyme (Gilat et al.,1972), the fall in lactase activity with weaningin most of the world’s population is ultimatelygenetic.Thus, this state of lactase insufficiency(hypolactasia) is normal in all suckling mam-mals, including humans. However, populationsof northern European descent and some no-madic tribes in Africa and Asia have retained theability to digest lactose past the weaning period(Saavedra and Perman, 1989; Scrimshaw andMurray, 1988;Thomas et al., 1990).

Finally, lactose intolerance may be secondaryto disorders that damage the normal formation

and functioning of the small intestinal mu-cosa.These may be due to acute diarrheal dis-eases (e.g., enteritis), parasitic infestations(e.g., giardiasis), enteropathies (e.g., celiac dis-ease), chronic inflammatory bowel disease(e.g., Crohn’s disease), ionizing radiation, an-timetabolite medications, and intestinal surgeryor resection.

Physiological Effects of LactaseInsufficiency

Carbohydrates that reach the large intestineprovide a metabolic resource to other nonmam-malian organisms, specifically certain types ofcolonic bacteria; to these, lactose is a fer-mentable substrate. Microbial fermentation oc-curs when appropriate substrates are availableand oxygen is excluded from the environment(i.e., under anaerobic conditions). Metabolicend products of fermentation vary according tothe specific organisms present but can includeshort-chain acids and alcohols and variousgases.The presence of fermentation products inthe colon does not necessarily lead to discom-fort. However, in the absence of sufficient, orsufficiently active, lactase in the small intes-tine, significant amounts of unhydrolyzed lac-tose enter the large intestine.The disaccharideosmotically draws fluid into the bowel lumen,increasing the volume and fluidity of the gas-trointestinal contents. The accumulation ofwater in the lumen produces watery stools, re-sulting in fluid and electrolyte loss from the gas-trointestinal tract and hemoconcentration ofthe blood. In addition, bacterial hydrolysis ofthe lactose in the large intestine splits the disac-charide into its component monosaccharides,which cannot be absorbed by the colonic mu-cosa, thereby further increasing the osmoticload. Bacterial fermentation of the glucose andgalactose leads to the evolution of carbon diox-ide, hydrogen, and methane gases.The osmoticimbalance and increase in concentration offermentation end products results in the gas-trointestinal symptoms of lactose intolerance:excessive gas, bloating, increased bowel sounds(borborygmi), abdominal pain, and acidic diar-rheal stools containing reducing sugar. How-ever, not all individuals with lactaseinsufficiency (lactose maldigesters) have symp-toms of lactose intolerance; and not all lactose-intolerant people are lactose maldigesters(Carrocio et al., 1998; Johnson et al., 1993a)(Figure 25-3).


Lactose

CH2OHO

OH

H

OHH

HO

H

OH

H

CH2OHO

H

H

OH

OHH

H

OH

H

Galactose Glucose

LACTASE

CH2OHO

H

H

OHH

HO

HOH

OH

H

CH2OHO

H

H

OH+

OHOH H

H

OH

H

Figure 25-2. Hydrolysis of lactose to componentmonosaccharides by the enzyme lactase.

Lactase

Lactase (lactase phlorizin hydrolase [3.2.1.23])has two activities: β-galactosidase activity,which catalyzes the hydrolysis of lactose, and aβ-glucosidase activity, which is responsible forhydrolyzing phlorizin, a disaccharide found insome seaweeds and in the roots and bark ofplants of the family Rosaceae.

In the small intestine mucosal cells (entero-cytes), lactase is synthesized as a propolypep-tide that is proteolytically cleaved inside thecell.The mature enzyme has an apparent molec-ular weight of approximately 160,000. Lactasehas a carboxy-terminal, membrane-spanning do-main that acts to anchor the enzyme, but thebulk of the enzyme projects into the lumen ofthe intestine. The protein dimerizes on thebrush border membrane of the small intestineto form the active enzyme. Like all brush bor-der hydrolases and transporters, lactase is gly-cosylated. There is evidence of both N- andO-glycosylation, and there are person-to-person differences in these terminal glycosyla-tions related to the ABH blood group antigens(Grand et al., 2003; OMIM; Swallow et al.,2001) (Fig. 25-4).

Primary Hypolactasia

Primary hypolactasia results in up to 75% of di-etary lactose passing unhydrolyzed through thesmall intestine into the colon, where it is rap-idly metabolized by colonic bacteria. The gas-

trointestinal symptoms that result vary from pa-tient to patient and are related to the quantityof lactose ingested, the relative hypolactasia,and the patient’s pain or discomfort tolerancelevel.They occur between half an hour and 2 hafter ingestion of lactose, depending in part onwhether lactose is taken by itself (e.g., in anaqueous solution) or as part of a meal withother nutrients (such as fat) ingested at thesame time. The amount of lactose required toproduce symptoms also varies, ranging from 4.5to 18 g (3–12 oz milk). Severity of symptomscan be directly related to the osmotic pressurecaused by the presence of unhydrolyzed lac-tose in the colon.

Ingestion of larger amounts of the disaccha-ride, faster gastric emptying time, and faster in-testinal transit time all contribute to moresevere symptoms. In general, since an increasein the fat content of food slows the rate of gas-tric emptying, lactose intolerance symptomswould be expected to decrease with ingestionof higher-fat milk. However, symptoms reportedby lactose maldigesters were not different afteringestion of fat-free compared to full-fat milk(Vesa et al., 1997).

A common misconception is that lactosemalabsorption and lactose intolerance symp-toms are caused by an allergy to milk or otherdairy products. Milk allergy is an immunologi-cally mediated reaction to milk proteins thatmay involve single systems (e.g., the gastroin-testinal tract, skin, respiratory tract) or multiple


Glucose

Lactase

Normal Stools Diarrhea, Flatulence

Lactose

Normal Lactose Digestion Lactose Intolerance

Lactose

SmallBowel

Colon

Bacteria Water

Irritation

IncreasedMotility

Fermentation

Lactic AcidH2

CH1CO2

Galactose

Figure 25-3. Genesis of symptoms in lactose malabsorption.

systems (systemic anaphylaxis). Also, milk pro-tein allergy may be an immediate or delayedtype hypersensitivity reaction. A recent studyshowed that children who are hypersensitive tocow’s milk (that is, are allergic to milk proteins)are tolerant of lactose even from bovinesources, clearly delineating the two conditions(Fiocchi et al., 2003).

Bacterial Fermentation of Lactose in the Colon

The catabolism of any organic energy source(carbohydrates, lipids, amino acids) involves itsoxidation: the transfer of electrons from theenergy source to an electron acceptor. Theelectrons released during fuel oxidation aretransferred to NAD+ (or other electron carriers)to form the reduced NADH. But, NADH must bereoxidized to NAD+ for the catabolic pathwayto continue operating, either through the pro-cess of cellular respiration, using oxygen (or ni-trate or sulfate in some bacteria) as a terminalelectron acceptor or through the process of fer-mentation.

In anaerobic environments like the colon,

where oxygen is not available as a final electronacceptor, fermentative bacteria transfer elec-trons from NADH to some organic compoundas the mechanism to regenerate NAD+.The re-dox reactions occur in the cytosol of the mi-croorganism, and ATP is produced throughsubstrate-level phosphorylation. There are anumber of variations of microbial fermenta-tion, but all need to have an acceptor for theelectrons produced during the oxidations. So,fermentations are characterized by excretionof relatively large quantities of reduced or-ganic compounds, including alcohols (such asethanol) and organic acids (such as lactic,acetic, propionic, and butyric acids).

For example, lactic acid bacteria are a largeand heterogeneous group of gram-positive bac-teria whose sole or major fermentation prod-uct from carbohydrates is lactate. Lactobacilliand bifidobacteria are two groups of lactic acidbacteria normally found in the human colon. Inhomofermentative lactate fermentation, glu-cose is oxidized to pyruvate via the 10 steps ofglycolysis. The NADH produced in the oxida-tion step of glyceraldehyde 3-phosphate to 1,3-bisphosphoglycerate is used to reduce pyruvate


Figure 25-4. Model of the molecular forms of lactase during synthesis and processingin the human villus enterocyte. MVM, microvillus membrane.

to lactate; NAD+ is regenerated concurrently.Those lactic acid bacteria that utilize what isknown as heterofermentative lactate fermenta-tion generate lactate using the decarboxylationand transferase reactions of the pentose phos-phate pathway. Lactate is ultimately createdfrom the reduction of pyruvate to regenerateNAD+, and ethanol is also produced in otherNADH-using, NAD+-regenerating reactions.

Frequently, hydrogen gas is also a fermen-tation end product; in some bacterial species,protons can be used as electron acceptors byenzymes known as hydrogenases, which cat-alyze reactions that essentially reduce protonsto hydrogen gas: NADH + H+ → H2 + NAD+. Hy-drogen gas is therefore a product only of partic-ular bacterial fermentative processes and not ofhuman metabolism.

The excreted end-products of fermenta-tions, including hydrogen gas, can be used byother anaerobic bacteria, and what is knownas an anaerobic food chain develops. At the“bottom” of the food chain are the archaealmethanogens, different species of which con-vert hydrogen gas, carbon dioxide, or acetate toCH4. The process of methanogenesis from H2

and CO2 occurs as follows:

4H2 + HCO3– + H+ → CH4 + 3 H2O

Methanogenesis from acetate occurs as follows:

CH3COO– + H2O → CH4 + HCO3–

Therefore, the colonic hydrolysis of lactoseto glucose and galactose, the fermentation ofthese monosaccharides to various organicacids and alcohols and H2, and the further con-version of these end products to CO2, and CH4

require the combined action of several types ofbacteria.

Because of this sequential catabolism, thepresence or absence of certain types of mi-croorganisms in the mixed populations of thecolon can have a significant effect on the symp-toms experienced by lactase-nonpersistentpatients and even on the accuracy of measure-ment of symptoms. Billions of microorganismsare present in each gram of feces, and these in-clude organisms from many different bacterialfamilies.The exact balance of individual speciesof bacteria within the total flora can changerapidly according to diet, colonic conditionssuch as pH, and interactions between the vari-

ous guilds that range from syntrophic to com-petitive.

Three examples illustrate how the microbialecology of the small and large intestine can in-fluence the symptoms of lactose intolerance.First, the absence of H2-producing bacterialguilds, which would result in less H2 gas pro-duced, or the presence of large numbers ofmethanogens, which would result in more H2

gas consumed, can alter the net amount of gasin the colon, with or without the presence oflactose as a fermentable substrate.Methane pro-duction consumes 4 mol of H2 to reduce 1 molof CO2 to CH4 (Pelletier et al., 2001; Verniaet al., 2003).A second example of how the mi-crobial ecology of the gut can influence thesymptoms of lactose intolerance has to do withthe presence of lactic acid bacteria. Lactic acidbacteria that are ingested with food (e.g.,Lacto-bacillus bulgaricus and Streptococcus ther-mophilus in yogurt) may contribute their ownlactase and thereby increase lactose hydrolysisand improve lactose absorption in the small in-testine. Ingestion of yogurt by lactase-nonper-sistent patients has in fact been shown toreduce H2 production beyond that expectedsimply because of the reduced lactose contentof yogurt (25%–50% less than in the sameamount of milk) due to its fermented state.Such a beneficial effect of a bacterial food sup-plement is called a probiotic effect (Hove et al.,1999; Pelletier et al., 2001). Finally, exposing thecolon to increasing amounts of lactose overtime can change the numbers and types of bac-teria present, so that the colonic flora becomesbetter suited to processing undigested lactoseentering from the small intestine (Hertzler andSavaiano, 1996; Johnson et al., 1993b; Pribila etal., 2000).

Prevalence

As mentioned earlier, primary late-onset heredi-tary (adult) hypolactasia is developmentally the“normal” state. The level of expression of lac-tase is controlled by a genetically determinedregulatory polymorphism that shows large dif-ferences in allele frequencies in human popula-tions and results in the widely varying prevalenceof hypolactasia among ethnic backgrounds. Es-timates range from 2%–5% hypolactasia in per-sons of northern European descent to nearly100% hypolactasia in adult Asians and NativeAmericans of North America. African Ameri-cans and Ashkenazi Jews have hypolactasia


prevalences of 60%–80% and Latinos 50%–80%.In North America alone, it is estimated thatabout 50 million people are lactose intolerant.

While lactase nonpersistence is the morecommon phenotype worldwide, lactase persis-tence tends to be the most frequent in popula-tions where fresh milk historically has been asignificant part of the adult diet, specifically innorthern Europeans and in some pastoral no-madic tribes of Africa and Asia. The very highfrequency of the lactase persistence allele inthese populations has resulted from positiveselection over time for a mutation that pro-vided a selective advantage by opening upthe enzymatic opportunity to benefit from thehigh nutritional content of milk. Human cul-tural evolution influenced this molecularevolution.About 40,000 years ago, humans do-mesticated milk-producing animals, and about9,000–10,000 years ago the pastoralist age be-gan. The pastoralists were nomads who trav-eled with their herds from one pasture toanother. They continued to hunt and gatherother foods but complemented their diet withthe milk of their animals, often in the form ofcheese and fermented milks. The presence oflactose in the diet after weaning was a nutri-tional windfall, and those with lactase persis-tence who could take advantage of it benefitedand passed on the opportunity to their de-scendents. Recent population genetics evi-dence suggests that the persistence of lactaseactivity in certain populations represents oneof the strongest signals of recent positive se-lection so far documented in the humangenome (Bersaglieri et al., 2004).

It appears that the allele for lactase persis-tence has coevolved with certain cattle milkprotein genes. Distribution of the allele for hu-man lactase persistence is most frequent innorth central Europe, and this closely matchesthe area of high allelic richness and genetic dis-tinctiveness of cattle milk proteins. It is specu-lated that the advantages conferred by milkconsumption led to the keeping of larger herdsof cattle and artificial selection within thoseherds for increased yields and altered milk pro-tein composition. Geographic co-presence ofcertain cattle milk protein genes with the hu-man lactase persistence allele represents a rareexample of cultural-genetic coevolution be-tween humans and another species. Previously,instances of coevolution were documented forhuman genes and genes of parasites (e.g., Plas-modium); this was the first non-disease-related

example, emphasizing the importance of ani-mal domestication in shaping human genomesas well as societies (Beja-Pereira et al., 2003).

Molecular Basis of Lactase Persistence/NonpersistencePolymorphism

Lactase is coded for by the gene designated LCTon chromosome 2q21. It is made up of 17 ex-ons, about 49-kb long, and the mRNA tran-scribed from it is a little more than 6 kb afterprocessing. For many years, a sort of molecularmystery surrounded the lactase persisstence/nonpersistence phenomenon; although it hasbeen assumed that regulation of gene transcrip-tion was the mechanism of control, the genethat encodes lactase did not appear to be differ-ent in people with these differing conditions.But, studies done in 2002 on nine extendedFinnish families found variations in the DNA lo-cated outside the lactase gene itself that werecorrelated to the differences in ability to digestlactose (Ennatah et al., 2002).

This and subsequent work (Kuokkanen et al.,2003; Troelsen et al., 2003) showed that thephenotypic polymorphism of lactase persis-tence or nonpersistence into adult life is associ-ated with two single nucleotide polymorphisms(SNPs) that occur upstream of the LCT gene.One of these SNPs, a T/C polymorphism at posi-tion −13910, is perfectly associated with lactasepersistence/nonpersistence. All individuals inthe group studied who had low lactase levels(nonpersistent) were homozygous for C at thissite; those who were lactase persistent had C/Tor T/T genotypes. The other SNP in the up-stream region is at position −22018; this is aG/A polymorphism. Its association with thecondition is not perfect but very high (97%).G/G was the genotype in those patients withlactase nonpersistence and G/A or A/A in thosewith the persistent phenotypes.The genotypescorrelated with the level of the disaccharidaseactivities in intestinal biopsies.

The −13910 T/C SNP is located in a strongenhancer region. Functional studies using lu-ciferase reporter constructs showed that theseregions of DNA in both lactase-persistent and -nonpersistent people have enhancer activity,but the −13910 (T) variant enhances the lac-tase gene promoter approximately four timesmore than the −13910 (C) variant. Control overthe LCT gene promoter regulates the transcrip-tion of the lactase gene and thus the level of


lactase produced. Functional evidence forthe nonpersistent −13910 (C) and persistent −13910 (T) allele was also obtained by relativequantitation of the expressed lactase alleles;several times higher expression of the lactasemRNA in the intestinal mucosa is found in indi-viduals with the T allele compared to thatfound in individuals with the C allele, confirm-ing regulation of the lactase gene at the tran-scriptional level.

The SNPs identified and a role for differentialactivity of an enhancer in lactase persis-tence/nonpersistence do not, however, explainhow those individuals with the genotype lead-ing to adult lactase nonpersistence were ableto produce sufficient lactase in infancy. Thereare developmental questions that remain to beanswered. Congenital lactase deficiency hasbeen localized to a separate region, also onchromosome 2q21 but not within the LCT gene( Jarvela et al., 1998).

Interestingly, the lactase-persistent phenotypefound in some pastoral groups in sub-SaharanAfrica is not correlated with the −13910 (T)genotype as it is in northern European popula-tions.This implies that the −13910 SNP is not acausative mutation or not the only causativemutation influencing lactase persistence.Thereis speculation that other variants may be in-volved in this region of the DNA and are yet tobe identified (Bersaglieri et al., 2004; Mulcare etal., 2004; OMIM).

Health Consequences of LactoseIntolerance

The gastrointestinal manifestations of primaryhereditary delayed-onset hypolactasia,while un-comfortable, are not life threatening and can becompletely eliminated by dietary manipulation.Is lactose intolerance a serious health issue orjust a nuisance? In certain circumstances, specif-ically at the extremes of life (infancy and oldage) lactose maldigestion may have grave healthconsequences.

When substantial carbohydrate maldigestionoccurs in infants, it can lead to diarrhea.This ismost often seen when the infant has experi-enced some other insult that has damaged thesmall intestine enterocytes, producing a second-ary hypolactasia.This is more so when the infanthas a rotavirus infection, and rotavirus is themost important cause of gastroenteritis in in-fancy. Rotavirus infects only mature enterocytes,

in which it induces apoptosis with consequentloss of lactase, as well as resulting in defectivemucosal absorption (Boshuizen et al., 2003).Thesmall size, relatively large body surface area, andsmall circulating blood volume of infants placethem at greater risk of significant dehydration,hypotension, and metabolic acidosis from severediarrhea. Although these can be life-threateningcomplications, they are quite uncommon com-plications of isolated lactose intolerance.

Old age represents the other extreme of lifeduring which lactose intolerance can presentdangerous consequences. A severe bout oflactose-induced osmotic diarrhea in a frail el-derly person may interfere with already precari-ous glucose regulation. Lactose maldigestionand malabsorption result in reduced energy(glucose) intake, which can lead to hypo-glycemia.This in turn may produce dizziness, apotential cause of a fall. Because the frail elderlyperson often has brittle bones and fragile vas-cular walls, a fall can result in bone fracturesor head injuries with intracranial bleeding.Theelderly also commonly have hemodynamicinstability. Significant diarrhea from lactosemaldigestion in this population may thereforealso result in hypovolemia and circulatory col-lapse or reduced arterial perfusion and is-chemia (Solomons, 2002).

Connection Between LactoseIntolerance and Other Diseases

Studies from Sardinia and Finland have raisedthe possibility of an association between lac-tase persistence genotypes, lactase activity, anddiabetic mellitus. Even though the findings ofthese studies are presently inconclusive, thesestudies are important because Sardinia and Fin-land represent areas with populations with thehighest incidence of type I diabetes in theworld, and Finland has the highest annual con-sumption of milk products, with an average of220 kg milk products/capita (Enattah et al.,2004; Meloni et al., 2001).Also, the genetic pre-disposition for adult lactose intolerance signifi-cantly affects calcium intake, bone density, andlikelihood of fractures in postmenopausalwomen. Whether lactose intolerance in youngadults prevents the achievement of adequatepeak bone mass and predisposes to severe os-teoporosis remains controversial (Di Stephanoet al., 2002; Kudlacek et al., 2002; Obermayer-Pietsch et al., 2004; Prentice, 2004).


Public Health and Cultural Concerns

The public health and political/cultural aspectsof lactose intolerance may be more compli-cated than the simple biochemistry of the con-dition. For example, the U.S.-based consumeradvocate group Physicians Committee for Re-sponsible Medicine has challenged the U.S. De-partment of Agriculture in a legal action overthe issue of the Food Pyramid and its recom-mendation for a range of daily servings of dairyfoods and for their inclusion of these foods ingovernment-sponsored feeding programs suchas school lunches. They argue that the AfricanAmerican population (and other minorities)tend to have genetic lactase nonpersistenceand that milk-based beverages can produce in-tolerance symptoms (Solomons, 2002). Sincelactase nonpersistence is known to be the bio-logical norm, why do most nutrition programsin the United States still treat it as a disease?Many schools require a physician’s note beforethey will provide an alternate beverage.The na-tional School Lunch Program reimburses forcow’s milk in a lunch but not for soy or ricemilk (Barnard, 2003).

On the other hand, some diseases for whichAfrican Americans are at greater risk, such ashypertension and stroke, may be made worseby a low intake of calcium. Average intake ofcalcium in African American, Hispanic, andAsian populations are at the threshold(600–700 mg/day) below which bone loss andhypertension can result. Though many mem-bers of these groups are lactase nonpersistent,intolerance symptoms can be reduced to ac-ceptable levels with commonsense dietarypractices that still allow sufficient intake ofdairy products for health. Partial reduction ofnational health disparities between ethnicgroups may be possible by overcoming the “bar-rier of lactose intolerance” (Jarvis and Miller,2002).

THERAPY

Lactose intolerance can be readily managed byensuring that the amount of lactose ingested isrestricted to the amount the individual can tol-erate, which is related to the individual’s levelof residual intestinal lactase activity. Lactosemaldigesters can determine their individualthreshold for the occurrence of lactose intoler-

ance symptoms and adjust their lactose intakeaccordingly (Johnson et al., 1993b). Indeed,most adult lactose maldigesters know from ex-perience how much milk (and lactose) they canhandle, and consuming smaller servings of milkat a time improves tolerance. Total eliminationof milk and dairy products from the diet is un-necessary and is nutritionally unwise becausesuch a management excludes intake of impor-tant and essential nutrients, including high-quality protein, calcium, vitamins A and D, andriboflavin.

Tolerance to lactose can be improved byslowly increasing intake of lactose-containingdairy products. As lactase is not an inducibleenzyme, this adaptation is likely to be due toqualitative or quantitative changes in colonicbacterial flora (Johnson et al., 1993b). Also,consuming lactose-containing foods as part of ameal improves tolerance to lactose. In additionto the amount of lactose, the type of dairy foodconsumed influences symptoms of lactose in-tolerance in lactose maldigesters (Scrimshawand Murray, 1988).Also, chocolate milk appearsto be better tolerated than unflavored milk bylactose maldigesters (Lee and Hardy, 1989).Yo-gurt and buttermilk are fermented milk prod-ucts that may be better tolerated than milk inindividuals with lactose intolerance. The en-hanced efficiency of lactose digestion fromyogurt is attributed to the activity of a β-lactosidase (lactase) enzyme contained in bac-teria (S. thermophilus, L. bulgaricus) present inyogurt. The bacterial lactase survives passagethrough the stomach and is released in thesmall intestine, where it aids the digestion ofthe yogurt’s lactose (Martini et al., 1991). Kefiris a fermented milk beverage that contains dif-ferent cultures than yogurt; ingestion of thisbeverage was shown to lower levels of breathhydrogen and the perceived severity of flatu-lence (Hertzler and Clancy, 2003).

If specific treatment is considered neces-sary, supplemental lactase can be provided asreplacement therapy. This is derived from spe-cific species of yeast or fungus and is availableas caplets, tablets, and drops. Whole milk pre-treated with lactase is marketed in some statesas Lactaid and can be used by lactosemaldigesters.The availability of the replacementenzyme has made unnecessary the use of thelactose-free diet, which can be nutritionally un-sound, as discussed. Also, since lactose is com-mon in many foods and drinks and is used in


compounding some drugs, compliance with alactose-free diet is likely to be difficult even ifadvisable.

QUESTIONS

1. Explain the pathophysiological basis ofthe common manifestations (symptoms)of lactose intolerance.

2. All patients with lactase insufficiencyhave symptoms of lactose intolerance.Dis-cuss why this statement is false.

3. Describe the biochemistry behind the fol-lowing diagnostic tests: stool pH of 5.5;positive stool test for presence of reduc-ing sugar (what other sugars besides lac-tose would give a positive test?).

4. How are H2, CO2, and methane producedin the human gut? Why is expired CO2 notused as a measure of bacterial fermenta-tion?

5. Why is an overnight fast required beforethe lactose tolerance test or the breath hy-drogen test is administered?

6. What is the point (to the fermenting bac-teria) of producing alcohols, acids, and H2

gas?7. From your general understanding of how

biomolecules interact with each other,how could change of a single nucleotidealter transcriptional activity? Describepossible bonds/interactions that could bealtered by C→T or G→A changes in aDNA sequence.

8. Many lactose-intolerant patients can, overtime, increase the amount of lactose in-gested without an increase of symptoms.What is the mechanism of this “toler-ance”? Is there a limit to its extent?

9. Discuss the clinical consequences of lac-tose intolerance.

BIBLIOGRAPHY

Barnard ND:The milk debate goes on and on andon! Pediatrics 112:448, 2003.

Beja-Pereira A, Luikart G, England PR, et al.: Gene-culture coevolution between cattle milk proteingenes and human lactase genes. Nat Genet35:311–313, 2003.

Bersaglieri T, Sabeti PC, Patterson N, et al.: Geneticsignatures of strong recent positive selection of

the lactase gene. Am J Hum Genet 74:1111–1120, 2004.

Boshuizen JA, Reimerink JH, Korteland-van MakeAM, et al.: Changes in small intestinal homeosta-sis, morphology, and gene expression during ro-tavirus infection of infant mice. J Virol 77:13005–13016, 2003.

Carrocio A, Montalto G, Cavera G, et al.: Lactoseintolerance and self-reported milk intolerance:relationship with lactose maldigestion and nu-trient intake. J Am Coll Nutr 17:631–636, 1998.

Di Stephano M, Veneto G, Malservisi S, et al.: Lac-tose malabsorption and intolerance and peakbone mass. Gastroenterology 122:1793–1799,2002.

Enattah NS, Forsblom C, Rasinpera H, et al., TheFinnDiane Study Group: The genetic variant oflactase persistence C (−13910) T as a risk factorfor type I and II diabetes in the Finnish popula-tion. Eur J Clin Nutr 1–4, 2004.

Enattah NS, Sahi T, Savilahti E, et al.: Identificationof a variant associated with adult-type hypolac-tasia. Nat Genet 30:233–237, 2002.

Fiocchi A, Restani P, Gualtiero L, et al.: Clinical tol-erance to lactose in children with cow’s milk al-lergy. Pediatrics 112:359–362, 2003.

Gilat T, Russo S, Gelman-Malachi E, et al.: Lactase inman: a nonadaptable enzyme. Gastroenterology62:1125–1127, 1972.

Grand RJ, Montgomery RK, Chitkara DK, et al.:Changing genes; losing lactase.Gut 52:617–619,2003.

Hermans MM, Brummer RJ, Ruijgers AM, et al.: Therelationship between lactose tolerance test re-sults and symptoms of lactose intolerance. Am JGastroenterol 92:98–104, 1997.

Hertzler SR, Clancy SM: Kefir improves lactose di-gestion and tolerance in adults with lactosemaldigestion. J Am Diet Assoc 103:582–587,2003.

Hertzler SR, Savaiano DA: Colonic adaptation todaily lactose feeding in lactose maldigesters re-duces lactose intolerance. Am J Clin Nutr64:232–236, 1996.

Hove H, Norgaard H, Mortensen PB: Lactic acidbacteria and the human gastrointestinal tract.Eur J Clin Nutr 53:339–350, 1999.

Jarvela I, Enattah NS, Kokkonen J, et al.: Assign-ment of the locus for congenital lactase defi-ciency to 2q21, in the vicinity but separate fromthe lactase-phlorizin hydrolase gene. Am J HumGenet 63:1078–1085, 1998.

Jarvis JK,Miller GD: Overcoming the barrier of lac-tose intolerance to reduce health disparities. JNatl Med Assoc 94:55–66, 2002.

Johnson AO, Semenya JG, Buchowski MS, et al.:Correlation of lactose maldigestion, lactose in-tolerance, and milk intolerance. Am J Clin Nutr57:399–401, 1993a.


Johnson AO, Semenya JG, Buchowski MS, et al.:Adaptation of lactose maldigesters to continuedmilk intakes. Am J Clin Nutr 58:879–881,1993b.

Kudlacek S, Freudenthaler O,Weissboeck H, et al.:Lactose intolerance: a risk factor for reducedbone mineral density and vertebral fractures? JGastroenterol 37:1014–1019, 2002.

Kuokkanen M, Enattah NS, Oksanen A, et al.: Tran-scriptional regulation of the lactase-phlorizinhydrolase gene by polymorphisms associatedwith adult-type hypolactasia. Gut 52:647–652,2003.

Lee CM, Hardy CM: Cocoa feeding and human lac-tose intolerance. Am J Clin Nutr 49:840–844,1989.

Martini MC, Kukielka D, Savaiano DA: Lactose di-gestion from yogurt: influence of a meal and ad-ditional lactose. Am J Clin Nutr 53:1253–1258,1991.

Meloni GF, Colombo C, La Vecchia C, et al.: Highprevalence of lactose absorbers in NorthernSardinian patients with type 1 and type 2 dia-betes mellitus. Am J Clin Nutr. 73:582–585,2001.

Mulcare CA,Weale ME, Jones AL, et al.: The T alleleof a single-nucleotide polymorphism 13.9 kbupstream of the lactase gene (LCT)(C-13.9kbT)does not predict or cause the lactase-persis-tence phenotype in Africans. Am J Hum Genet74:1102–1110, 2004.

Obermayer-Pietsch BM, Bonelli CM, Walter DE,et al.: Genetic predisposition for adult lactoseintolerance and relation to diet, bone densityand bone fractures. J Bone Miner Res 19:42–47,2004.

OMIM Online Mendelian Inheritance in Man.www.ncbi.nlm.nih.gov/entrez.

Pelletier X, Laure-Boussuge S, Donazzolo Y: Hydro-gen excretion upon ingestion of dairy productsin lactose-intolerant male subjects: importanceof the live flora. Eur J Clin Nutr 55:509–512,2001.

Prentice A: Diet, nutrition and the prevention of os-teoporosis.Public Health Nutr 7:227–243,2004.

Pribila BA, Hertler SR, Martin BR et al.: Improvedlactose digestion and intolerance amongAfrican-American adolescent girls fed a dairy-rich diet. J Am Diet Assoc. 100:524–528, 2000.

Saavedra JM, Perman JA: Current concepts in lac-tose malabsorption and intolerance. Annu RevNutr 9:475–502, 1989.

Scrimshaw NS, Murray EB: The acceptability ofmilk and milk products in populations with ahigh prevalence of lactose intolerance. Am JClin Nutr 48:1083–1159, 1988.

Solomons NW:Fermentation, fermented foods,andlactose intolerance. Eur J Clin Nutr 56:S50–55,2002.

Swagerty DL,Walling AD,Klein RM:Lactose intoler-ance. Am Family Phys 65:1845–1850, 2002.

Swallow DM, Poulter M, Hollox EJ: Intolerance tolactose and other dietary sugars. Drug MetabDispos 29:513–516, 2001.

Thomas S, Walker-Smith JA, Senewiratne B, et al.:Age dependency of the lactase persistence andlactase restriction phenotypes among childrenin Sri Lanka and Britain. J Trop Pediatr.36:80–85,1990.

Troelsen JT, Olsen J, Moller J, et al.: An upstreampolymorphism associated with lactase persis-tence has increased enhancer activity.Gastroen-terology 125:1686–1694, 2003.

Vernia P, Di Camillo M, Marinaro V, et al.: Effect ofpredominant methanogenic flora on the out-come of lactose breath test in irritable bowelsyndrome patients. Eur J Clin Nutr57:1116–1119, 2003.

Vesa TH, Lember M, Korpela R: Milk fat does not af-fect the symptoms of lactose intolerance. Eur JClin Nutr 51:633–636, 1997.

Vesa TH, Marteau P. Korpela R: Lactose intolerance.J Am Coll Nutr 19:165S–175S, 2000.

White D: The Physiology and Biochemistry ofProkaryotes. Oxford University Press, NewYork, 1995.


www.ncbi.nlm.nih.gov/entrez

CASE REPORT

A 45-year-old male patient presented with diar-rhea and weight loss (12 kg during the past year;his current body weight was 63 kg, height was181 cm, and Body Mass Index (BMI) was19.2 kg/m2 [normal is 19.2–24.9]). He reportedthat his stools were bulky, loose, and odorous,particularly following large meals rich in fat.Dur-ing the last 15 years, the patient had sufferedfrom recurrent attacks of abdominal pain. Mostepisodes were self-limited and subsided after 1–3days. In time, the patient found that avoidingfood intake during acute attacks was helpful. Inassociation with one attack, a peptic ulcer wassuspected, and the patient was treated unsuc-cessfully with an acid-suppressing drug (possiblya proton pump inhibitor), the name of whichhe did not remember; upper gastrointestinal en-doscopy and ultrasound were not performed.

The patient had observed a gradual decreaseover time of both frequency and intensity ofacute attacks and therefore did not pursue asystematic diagnostic workup. On further in-quiry, the patient admitted to substantial, con-tinuous alcohol consumption over the last 20years of about 2–3 L of beer per day (on week-ends and “special occasions”even more) as wellas hard liquor. He did, however, abstain from al-cohol during acute attacks.Thus, his mean dailyconsumption can be estimated as about 120 gpure alcohol. Moreover, the patient was a heavysmoker, smoking 20–30 cigarettes per day fornearly 30 years.

Physical examination indicated poor nutri-

tional status (BMI 19.2 kg/m2) and reducedgeneral condition but otherwise normal re-sults. Routine laboratory examinations showedmacrocytosis (mean corpuscular volume oferythrocytes increased to 105 fL; normal is82–101 fL) and elevated γ-glutamyltransferase(152 U/L; normal is 5–39 U/L) compatible withongoing alcohol abuse and liver damage. Fast-ing blood glucose concentrations and serumconcentrations of amylase and lipase were nor-mal.Abdominal ultrasound showed a fatty liver,nonhomogeneous pancreatic tissue with evi-dence of calcifications, and a widened andirregular pancreatic duct (Fig. 26-1).A plain ab-dominal radiogram confirmed pancreatic calci-fications. Endoscopic ultrasound revealed noevidence of a pancreatic tumor but confirmedstructural changes as observed on abdominalultrasound.

Quantitative fecal fat excretion was markedlyincreased (28 g/day; normal is <7 g/day), andfecal elastase 1 concentration was reduced to24 µg/g (lower level of normal is 200 µg/g).Thepatient was diagnosed with severe pancreaticexocrine (but not endocrine) insufficiencydue to alcoholic chronic calcifying pancreati-tis and advised to stop alcohol and cigaretteconsumption.

The patient also was advised to change his di-etary habits by avoiding large meals and eatingseveral smaller or medium meals.He was treatedwith pancreatin, a medication containing a mix-ture of porcine pancreatic enzymes, includinglipase, amylase, and proteases, with the dosagedetermined by units of lipase activity (Layer and

26

Pancreatic Insufficiency Secondary to Chronic Pancreatitis

PETER LAYER and JUTTA KELLER

278

Keller, 2003).The initial dosage (25,000 U lipasewith each small meal or snack and 50,000 unitswith a medium meal) resulted in partial im-provement. Symptoms virtually disappearedwith subsequent doubling of enzyme dosage. Inparticular, stools became normal, and the pa-tient gained 4 kg of weight within 3 months.

DIAGNOSIS

Pancreatic exocrine insufficiency with nutrientmaldigestion as described in the case reportis both a classical complication and a definingleading symptom of chronic pancreatitis.There-fore, taking a careful history provides the keyfor the diagnosis.

In the majority of patients, pancreatic ex-

ocrine insufficiency develops as a late compli-cation, mostly in the second decade followingfirst onset of symptoms. Its typical manifesta-tion is insidious, with gradual progression.Thus, it is necessary to consider this diagnosisin patients in whom weight loss, usually associ-ated with steatorrhea (fat in the stool, per-ceived by the patient as bulky, loose, odorousstools) develops following a long-standing his-tory of abdominal pain and/or known chronicpancreatitis. As a further clue, pancreatic ex-ocrine insufficiency may be associated with di-abetes mellitus, in which there is destructionof endocrine as well as exocrine pancreaticparenchyma. Diabetes mellitus is a late conse-quence in more than 60% of patients with pan-creatic exocrine insufficiency.

It is important to keep in mind, however, thatnot all patients with chronic pancreatitis de-velop clinical pancreatic exocrine insufficiency;approximately 25% of patients still have suffi-cient exocrine function after 25 years of dis-ease. On the other hand, it is also important tonote that 10%–15% of patients with chronicpancreatitis have primary painless disease; inthese patients,pancreatic exocrine insufficiencymay be the first (and possibly only) clinical man-ifestation.Thus, the absence of pain or a historyof pancreatitis does not exclude the diagnosis(DiMagno et al., 1993).

Irreversible destructive morphological changes(such as ductal alterations, calcifications, scar-ring, pseudocysts, etc.) are essential features ofchronic pancreatitis and are therefore usually, al-beit not invariably, detectable in pancreatic ex-ocrine insufficiency by imaging methods, thatis, morphological investigations. These includeplain abdominal x-ray, sonography, computedtomography (CT), endosonography, endoscopicretrograde cholangiopancreatography (ERCP),and magnetic resonance imaging (MRI)/mag-netic resonance cholangiopancreatography(MRCP). (ERCP involves endoscopic injection ofa radioopaque solution and radiological imagingof the pancreatic duct system. Magnetic reso-nance cholangiopancreatography uses specificresonance properties of compartmentalized fluidwithin the pancreatic duct system for visualiza-tion.) It has to be kept in mind,however, that im-aging findings may provide evidence for chronicpancreatitis but do not prove pancreatic ex-ocrine insufficiency (with the possible future ex-ception of functional MRI techniques). Hence, inorder to diagnose pancreatic exocrine insuffi-ciency, secretory function needs to be assessed.

Pancreatic Insufficiency 279

Figure 26-1. Abdominal ultrasound showingpathological alterations typical of chronic pancre-atitis, i.e. a nonhomogenous pancreatic tissue withcalcifications (A; several calcifications are markedwith an arrow) and a widened and irregular pan-creatic duct (B; diameter of pancreatic duct D1 is8.2 mm, upper limit of normal is 3 mm). CourtesyC. Pachmann, MD.

Pancreatic function tests are therefore indi-cated if and when one or more of the followingaspects need be clarified: Is a symptom or signcaused by pancreatic exocrine insufficiency?Has pancreatic exocrine insufficiency devel-oped in the course of chronic pancreatitis?Does a patient require enzyme supplementa-tion treatment?

To be of value in routine clinical practice,any pancreatic function test is required to benot only precise and reliable, but also uncom-plicated and inexpensive. However, due to thepeculiarities of pancreatic anatomy and physi-ology, no currently available test meets all thesecriteria. Indeed, despite astounding develop-ments in virtually every other diagnostic field(including pancreatic imaging), pancreatic ex-ocrine function is estimated today by usingvery much the same principles and even meth-ods as three decades ago.

In principle, pancreatic function can be testedby quantifying exocrine secretory output directlyby duodenal intubation, juice aspiration,and mea-surement of enzymes under defined stimulatoryconditions, usually by intravenous application ofcholecystokinin and/or secretin. ERCP-guided di-rect cannulation and collection of pure pancre-atic juice is, however, not a routine proceduredue to its substantial inherent risk of iatrogenicpancreatitis. Pancreatic enzymes can also bequantified indirectly by measuring a representa-tive enzyme in a stool sample.

The other approach is indirect, namely, tomeasure not enzymes as such, but their physio-logical biochemical efficacy. This can beachieved by oral administration of a compositetest substance, which is hydrolyzed (“digested”)into its components exclusively by a specificpancreatic enzyme and subsequently absorbedand eliminated by the renal or the respiratorysystem. Ideally, the urinary excretion/respiratoryexhalation of a metabolite of the test substanceis proportional to its hydrolysis in the small in-testine, which in turn is directly dependent onthe quantity of pancreatic enzymes present.


Physiology of Human PancreaticEnzyme Secretion

Depending on nutrient intake, the human pan-creas secretes about 3 L of alkaline juice per day.Pancreatic juice contains hydrolytic enzymes

(or their precursors, see below) for digestion ofcomplex carbohydrates, proteins, and lipids,with the most important ones α-amylase,trypsin, chymotrypsin, lipase, and colipase. Inaddition, elastases, carboxypeptidases, phos-pholipase A2, carboxylester lipase, and ribo-and deoxyribonucleases are also secreted.Some of the enzymes are present in more thanone form (e.g., cationic and anionic trypsin-ogen). The principal inorganic components ofpancreatic juice are water, potassium, chloride,and bicarbonate.The high bicarbonate contentof pancreatic juice is responsible for its alka-line pH. This serves to protect enzymes fromacidic denaturation and increases enzymatic ac-tivity within the intestinal lumen (pH optimawithin the alkaline range). α-Amylase and lipaseare secreted in active form, whereas proteasesenter the duodenum as inactive zymogens(Rinderknecht, 1993). Within the duodenum,trypsinogen is partly converted to trypsin by theduodenal enzyme enterokinase. Subsequently,trypsin has an autocatalytic effect and also acti-vates the other proteases (Fig. 26-2).

Under physiological circumstances, two dis-tinctive secretory patterns are observed: Dur-ing the fasting state, the healthy humanpancreas secretes low amounts of pancreaticjuice; pancreatic secretion is cyclical and inte-grated with upper gastrointestinal motility, gas-tric acid secretion, and entry of bile into theduodenum (Keller and Layer, 2002). One cycletakes 90–120 min and consists of a period of se-cretory and motor quiescence (phase I) fol-lowed by irregular secretion and motility (phaseII, about 80% of time).Maximal interdigestive se-cretion rates (about 50% of maximum) occur atthe end of phase II and just before onset ofphase III (motility), in which the motility in theduodenum is characterized by regular and ab-orally propagated contractions at maximum fre-quency. This motor pattern has a strongpropulsive effect and is thought to clean thesmall intestine of residual nutrients and endoge-nous substances.The clearing effect is enhancedby the high volumes of pancreatic juice and bilesecreted just prior to it. Pancreatic secretion de-clines during duodenal phase III.A short phaseof irregular motor activity (phase IV) may followphase III before onset of the next phase I (Di-Magno and Layer, 1993).

The digestive pattern is initiated by stimula-tion of pancreatic exocrine secretion due toimagination, sight, smell, and taste of food (i.e.,the so-called cephalic phase of pancreatic


exocrine secretion). Distention of the stomach(gastric phase) may also increase pancreatic se-cretion via vagal reflexes. However, the mostimportant stimulatory mechanism of prandialpancreatic exocrine secretion is nutrient expo-sure of the duodenal mucosa. There is experi-mental evidence that digestive products ofenzymatic hydrolysis (such as free fatty acids)rather than intact macromolecules (such astriglycerides) evoke neurohormonal stimula-tion of the physiological enzyme response to ameal. The most important stimulatory hor-mones are cholecystokinin and secretin. Bothare peptide hormones produced and releasedby the enterochromaffin cells of the proximalsmall intestinal mucosa. Cholecystokinin is thestrongest physiological stimulator of enzymesecretion. By contrast, secretion of bicarbonateis stimulated by secretin, which is releasedmainly in response to duodenal acid exposure(Owyang and Logsdon, 2004). Following a regu-lar meal, the digestive pattern will prevail forabout 4–7 h.Thus,with three or more meals perday, the interdigestive pattern will usually onlyoccur during nighttime (DiMagno and Layer,1993).

Sufficient absorption of macronutrients tomaintain energy supply requires prior hydroly-sis by pancreatic enzymes. Thus, complete ab-sence of pancreatic enzymes (e.g., due topancreatectomy) is, if untreated, incompatiblewith life.Normally,however, in healthy individu-als the cumulative postprandial pancreatic en-zyme response exceeds by far (10- to 15-fold)the quantity required to prevent overt maldiges-tion. In particular, lipid digestion is fast and

highly efficient under physiological circum-stances. Hydrolysis of dietary triacylglycerolsdepends on an intricate interplay among pan-creatic lipase, colipase, and bile salts. Pancreaticlipase binds to the oil-water interface of oildroplets and hydrolyzes triacylglycerol to twofatty acid molecules and a monoacylglycerol,with the fatty acid esterified at position 2; bothbile salts and colipase are important to achieveits full lipolytic activity: Bile salts enhance theemulsification of triacylglycerols and form mi-celles, which remove fatty acids and monoacyl-glycerol from the oil-water interface. In thepresence of bile salts, colipase forms a complexwith lipase that acts to anchor the lipase and al-lows its action in a more hydrophilic environ-ment.After incorporation into micelles, the fattyacids, and monoacylglycerol are absorbed, in-corporated into chylomicrons, and transportedto the liver.

In healthy humans, up to 80% of lipids maybe absorbed during duodenal transit (i.e., be-fore reaching the ligament of Treitz (Holtmannet al., 1997)). In exocrine pancreatic insuffi-ciency, on the other hand, 7%–10% of normalprandial secretory rates may be enough to pre-vent elevated fecal fat excretion. Still, even inhealthy human subjects, digestion and absorp-tion of nutrients after a regular meal is in-complete, and substantial nutrient quantitiesregularly escape intestinal absorption and passthe ileocecal junction. Nutrient exposure ofthese intestinal sites inhibits upper gastroin-testinal secretory and motor functions (Layeret al., 1990;Read et al., 1984).This mechanism iscalled ileal brake. In addition, malabsorbed


Enterokinase

TAP

Trypsinogen

TrypsinogenChymotrypsinogenProelastaseProcarboxypeptidase A/BProcolipaseProphospholipase A2

Trypsin

ActiveEnzymes

Activationpeptides

Figure 26-2. The pancreatic enzyme cascade. Pancreatic proteases enter the intestinallumen as inactive zymogens.Within the duodenum, a specific enzyme of the duodenalmucosa, enterokinase, activates trypsinogen by releasing the trypsinogen activation pep-tide (TAP). Subsequently, active trypsin activates the other zymogens and acts autocat-alytically.

nutrients serve as exogenous energy supply ofthe colonic microflora.

After having been secreted intraluminally,enzymatic activity decreases during small intes-tinal transit. The rate of inactivation differs be-tween enzymes because they have differentstability against degradation (Layer et al., 1986).For example, about 60% of duodenal proteaseactivities reach the jejunum and between 20%and 30% reach the ileum. It should be notedthat trypsin’s hydrolytic activity survives betterthan its immunoreactivity, which suggests thatcomplete structural integrity may not be re-quired for its enzymatic action. Pancreatic amy-lase is rather stable because it is not easilyproteolyzed and consequently has been foundto have a high duodeno-ileal survival rate. Bycontrast, lipase is inactivated very rapidly, par-ticularly in the absence of triglycerides, andonly a small proportion may reach the distalsmall bowel. Chymotrypsin appears to be ofparticular importance for destroying lipase ac-tivity.

Pancreatic Exocrine Insufficiency in Chronic Pancreatitis

Nutrient maldigestion due to pancreatic ex-ocrine insufficiency is a classical complicationof chronic pancreatitis. In most cases, it occurs

late in life because of the enormous functionalreserve of the gland described above. In alco-holic pancreatitis, its manifestation usually de-velops within the second decade of clinicaldisease, but it may also appear more rapidly.

Steatorrhea, the clinical result of insufficientintraluminal lipid hydrolysis, is the most im-portant digestive malfunction in pancreatic ex-ocrine insufficiency. As a rule, concomitantmalabsorption of the lipid-soluble vitamins A,D, E, and K must be suspected in these pa-tients. Naturally, potential differential diag-noses have to be considered in patients whopresent with steatorrhea (Table 26-1).The piv-otal role of fat malabsorption in chronicpancreatitis is due to several interacting mech-anisms:

1. In progressive chronic pancreatitis, acinarsynthesis and release of lipase decreaseearlier and more markedly compared withproteases.

2. Pancreatic bicarbonate secretion is alsodiminished in exocrine pancreatic insuffi-ciency. There may thus be insufficientintraduodenal buffer protection for en-zymes against denaturation by gastric acidemptied with postprandial chyme. Indeed,intraluminal pH may decrease below 4.0,which results in irreversible destruction oflipase.


Table 26-1. Clinically Important Differential Diagnoses of SteatorrheaPathophysiology Disease

Pancreatic exocrine insufficiencyDecreased intraluminal fat digestion Chronic pancreatitis

Cystic fibrosis

Disturbed secretion of bile acidsImpaired fat solubilization, decreased Primary biliary cirrhosis

formation of micelles Primary sclerosing cholangitis

MalabsorptionReduced absorptive capacity of intestinal Celiac disease

mucosa due to inflammation Morbus WhippleCrohn’s diseaseParasitosis

Nutrient unavailability

Abdominal surgery, radiation therapyLoss of absorptive surface Short-bowel syndromeBypass Enteroenteric fistualNutrient unavailability, deconjugation Bacterial overgrowth

of bile acidsDisturbed solubilization of lipids, Loss of bile acids

decreased formation of micelles

3. During small intestinal transit, lipase isproteolyzed more rapidly than other en-zymes.

4. Extrapancreatic lipolytic mechanisms (suchas gastric lipase) play only a minor roleand cannot compensate for the loss ofpancreatic lipase.

Protein and starch digestion, on the otherhand, have potent nonpancreatic compensa-tory mechanisms. Due to the compensatoryaction of salivary amylase and brush borderoligosaccharidases, a substantial proportion ofstarch digestion can be achieved without pan-creatic amylase. Similarly, protein denaturationand hydrolysis is initiated by gastric proteolyticactivity (acid and pepsin) and continued by in-testinal brush border peptidases, and is thuspartly maintained even in the absence of pan-creatic proteolytic activity.

The coefficient of fat absorption is defined asthe amount of fat absorbed as a percentage ofthe ingested amount. This coefficient normallyexceeds 93% and is used (rather than crude fe-cal fat excretion) to indicate efficacy of luminalfat digestion following different dietary lipid in-takes. By contrast, fecal carbohydrate measure-ments do not fully reflect the extent of starchmalabsorption because carbohydrates are me-tabolized by the intracolonic microbial flora.Since intracolonic metabolism of carbohydrates

produces H2, starch malabsorption can be mea-sured by determining hydrogen breath concen-trations.

In addition to steatorrhea and nutritionaldeficiencies, patients with pancreatic exocrineinsufficiency also develop symptoms such aspostprandial pain, cramps, bloating, and disten-tion.These are caused by profound alterations ofupper gastrointestinal secretory and motorfunctions in response to increased nutrient de-livery to the distal small intestine, particularlythe ileum. In the first 5–10 years of chronic pan-creatitis, overt malabsorption is usually neitherdetected nor a major clinical problem, althoughenzyme output may decrease by 60%–90%. Still,there is evidence that, even in the early stages ofchronic pancreatitis, the site of maximal nutri-ent digestion and absorption is shifted from theduodenum to the more distal small intestine.

Moreover, recent studies suggest that gastricemptying and small intestinal transit aremarkedly accelerated in patients with untreatedpancreatic insufficiency (Fig. 26-3) (Layer et al.,1997).The available time for digestion and ab-sorption is therefore markedly decreased inthese patients, which may contribute further tomalabsorption. Disturbance of both digestionand gastrointestinal transit are corrected byenzyme replacement, suggesting that malab-sorption is not only a consequence but also acause of abnormal motor function. In turn, this


Figure 26-3. Patients with severe pancreatic exocrine insufficiency (n = 14,mean ± SE)show marked alterations of gastrointestinal functions with shortened small bowel transittime, 90% gastric emptying time, and duration of the fed pattern. These alterations arecorrected by enzyme replacement therapy (*P < .05 vs. controls, # P < .05 vs. placebo)(Layer et al., 1997).

pathophysiological mechanism explains the al-leviating effects on pain and dyspepsia of pan-creatic enzyme supplementation in a subset ofpatients.

Direct Pancreatic Function Tests

The gold standard for measuring pancreatic ex-ocrine function is direct measurement of pan-creatic enzyme and bicarbonate output inresponse to a specific and defined hormonalstimulus such as cholecystokinin (or an ana-logue such as cerulein) alone or in combinationwith secretin (secretin-cerulein [SC] test). Be-cause sensitivity and specificity of this test bothexceed 90%, it is the only method apt to diag-nose mild-to-moderate pancreatic insufficiencyreliably. Unfortunately, the test is invasive, de-manding, costly, and not standardized across ex-pert centers. It requires placement of adouble-lumen nasoduodenal tube for continu-ous elimination of gastric juice and completecollection of duodenal juice in ice at 10- to 15-min intervals during intravenous infusion of se-cretin and cerulein. Analyses of the samplesinclude measurements of volume, bicarbonate,trypsin, chymotrypsin, lipase, and amylase out-puts.

The Lundh test also includes intestinal intu-bation and direct measurement of enzyme out-put in duodenal juice but uses a standardizedtest meal as a pancreatic stimulus. Because thistest requires release of physiological regulatorymediators from the duodenal mucosa, it is lessspecific than the SC test and may render false-positive results in intestinal diseases such asceliac sprue.

Although all pancreatic enzymes are inacti-vated during intestinal transit, fecal outputs ofseveral enzymes correlate with pancreatic en-zyme secretion. Fecal chymotrypsin activity,which is comparatively stable in the lumen aswell as in extracorporal fecal samples, can bemeasured by a commercially available photo-metric test kit. When performed on three con-secutive days, this test detects severe pancreaticexocrine insufficiency, but sensitivity and speci-ficity are low in mild-to-moderate cases. In addi-tion, the test does not differentiate betweenporcine and human chymotrypsin, so that pan-creatin supplements need to be discontinued 5days prior to the test. For this reason, however,the test is able to monitor a patient’s compli-ance in severe pancreatic insufficiency appar-ently refractory to enzyme treatment. Patients

who still suffer from steatorrhea although theytake the pancreatin as prescribed will have highfecal chymotrypsin activities, whereas thosewho fail to ingest their medication will havelow chymotrypsin activities.

Fecal elastase 1 concentration can be mea-sured by an ELISA test kit using an antibodyspecific for the human enzyme; pancreatin sup-plements do not interfere with this pancreaticfunction test and need not be discontinued.Al-though measurement of fecal elastase 1 excre-tion appears to be somewhat more sensitivethan fecal chymotrypsin, its specificity and posi-tive predictive value are similarly low, and false-positive results can be expected in patients withintestinal diseases. Conversely, mild-to-moderatestages of pancreatic exocrine insufficiency can-not be diagnosed reliably.

Small quantities of pancreatic enzymes arereleased from the pancreas into the blood-stream even physiologically and are detectableas low serum activities and/or concentrations oflipase,amylase, trypsinogen,and chymotrypsino-gen, respectively. Since progressive destructionof the organ occurs in chronic pancreatitis,this should theoretically be reflected by de-creased serum enzymes, but so far these testsare not clinically useful because of their lowaccuracy.

Indirect Pancreatic Function Tests

Measurement of stool weight and quantitativefecal fat excretion on three consecutive daysduring a balanced diet are common screeningtests for both pancreatic insufficiency andother pathologies that result in malabsorption.However, these tests are insensitive and non-specific for pancreatic malfunction: Steator-rhea occurs only after loss of more than 90% ofexocrine parenchyma, and other causes of mal-absorption (e.g., celiac sprue or Crohn’s dis-ease) may also induce abnormal fecal fatexcretion of more than 7 g/day or more than5 g/100 g.

Orally administered fluorescein is readily ab-sorbed in the small intestine.By contrast fluores-cein coupled to dilaurate cannot be absorbedunless the composite molecule is cleaved intra-duodenally by the pancreatic cholesterol es-terase to form lauric acid and (absorbable)fluorescein. After its absorption, fluorescein ispartly glucuronidated in the liver and then ex-creted in urine, predominantly as fluoresceindiglucuronide. Thus, in pancreatic-insufficient


patients, decreased secretion of cholesterol es-terase results in incomplete cleavage and de-creased generation and absorption of freefluorescein (Fig. 26-4). The Pancreolauryl® testinvolves ingestion of a standard breakfast to-gether with 0.5 mmol of fluorescein dilaurate.Urine is collected for 10 h postprandially fordetermination of fluorescein excretion. After awashout period of at least 24 h, the protocol isrepeated with free fluorescein instead of fluo-rescein dilaurate; this procedure serves to con-trol for individual differences in postdigestiveprocessing of free fluorescein. Results are ex-pressed as the ratio of fluorescein excreted onthe test and the control days in percentage(normal is > 30%, abnormal is < 20%, equivocalis 20%–30%).

The test is valid for diagnosing severe pancre-atic exocrine insufficiency but has the sensitivitylimitations described above with respect to de-tecting mild impairment of pancreatic function.Conversely, despite the inclusion of the controltest, false-positive results may occur in patientswith intestinal and biliary diseases, as well as fol-lowing gastric resection or Y-en-Roux proce-dures. In the latter patients, intraluminal lack ofpancreatic enzymes (despite normal secretory

capacity of the pancreas) is caused predomi-nantly by postprandial asynchrony of pancreaticsecretory output and chyme delivery and subse-quent impaired mixing of enzymes and nutrientmacromolecule substrates, as well as reduced re-lease of stimulatory mediators. However, the testmay be useful in such patients because it reflectsthe efficacy of postprandial intraluminal nutrientdigestion as well as the effects of enzyme ther-apy.

The test principle of the NBT-PABA (N-benzoyl-L-tyrosyl-P-aminobenzoic acid, ben-tiromide) test very much resembles thePancreolauryl® test (Fig. 26-4).Again, the unsplitmolecule NBT-PABA cannot be absorbed, andabsorption of its metabolite PABA depends onprior hydrolysis by chymotrypsin within the in-testinal lumen. Subsequently, PABA undergoesconjugation in the liver and is excreted in theurine. In healthy subjects, at least 50% of the ad-ministered dose is excreted in the urine during6 h postprandially.

A noninvasive and simple H2 breath test hasbeen suggested for measurement of pancreaticfunction based on the principle that undigestedstarch will pass into the colon, be metabolizedby the colonic flora, and thus lead to an in-crease in breath H2 exhalation. However, nu-merous mechanisms for false-positive (e.g.,bacterial overgrowth) and -negative (e.g., insuf-ficient H2 production by colonic bacteria) limitits validity.

Several breath tests using 13C-labeled lipidshave been developed for measurement of pan-creatic exocrine function. They are based onthe common principle that intestinal triglyc-eride absorption requires prior hydrolysis bypancreatic lipase to produce free fatty acids andmonoacylglycerol.These metabolites are incor-porated into micelles, absorbed, and trans-ported to the liver. Further degradation byhepatic enzymes and β-oxidation results in for-mation of 13CO2, which is absorbed into thebloodstream, transported to the lungs, and ex-haled (Fig.26-5).Thus, an increase in 13CO2 con-centration in breath correlates with intestinallipid digestion as a marker of pancreatic ex-ocrine function. Measurements are performedby mass spectrometry or isotope-selective in-frared spectrometry. Varying test modificationshave been developed using 13C-tripalmitin,13C-triolein,13C-Hiolein® (a mixture of physiolog-ical long-chain triglycerides), 13C-trioctanoin, orso-called 13C-mixed triglycerides [1,3-distearyl,2(carboxyl-13C)octanoyl glycerol] as substrates.


Pancreaticenzyme

Photometry

SubstrateSubstrate•Marker

Marker

Serum

Urine

Figure 26-4. Principle of the Pancreolauryl® andthe NBT-PABA (N-benzoyl-L-tyrosyl-P-aminobenzoicacid, bentiromide) test. The composite moleculeconsisting of a substrate and a marker molecule can-not be absorbed but can be cleaved intraduodenallyby pancreatic enzymes (cholesterol esterase andchymotrypsin, respectively), leading to release ofan absorbable marker substance.Following absorp-tion and hepatic conjugation, the marker is ex-creted in urine. In pancreatic-insufficient patients,decreased secretion of pancreatic enzymes resultsin incomplete cleavage of the composite molecule.This results in decreased absorption and subse-quent excretion of the marker, which can be mea-sured photometrically.

The breath test using 13C-mixed triglyceridesis best established because it offers several ad-vantages over other substrates. Octanoic acid asa medium-chain fatty acid is metabolized fasterand therefore allows a shorter test durationthan substrates with long-chain fatty acids.Since a normal diet contains only low amountsof octanoic acid, the marker is not diluted byunmarked substrate. Furthermore, maximal andcumulative 13C-exhalation are higher with13C-mixed triglycerides than with long-chaintriglycerides because incorporation of carbonatoms from octanoate into the fatty tissue islower (Vantrappen et al., 1989). Most currentprotocols report sensitivities of about 90% for

severe but less than 60% for mild pancreatic ex-ocrine insufficiency, which is similar to otherindirect pancreatic function tests.

Differential Use of Pancreatic Function Tests

Direct tests of secretory function such as fecalchymotrypsin and elastase 1 are the tests of firstchoice if the main diagnostic goal consists ofnoninvasive confirmation of chronic pancreati-tis. Indirect tests may be preferred, however, ifthe main goal is to verify maldigestion (whichneeds not be due to loss of pancreatic secre-tory capacity) or to optimize enzyme treat-ment. For patients for whom noninvasivedirect or indirect tests are negative or equivo-cal and diagnosis or exclusion of pancreatic ex-ocrine insufficiency appears relevant, theinvasive secretin-cerulein (SC) test should beconsidered.

THERAPY

Basis of Enzyme Treatment

To restore nutrient digestion in exocrine pancre-atic insufficiency, sufficient enzymatic activitymust be administered into the duodenal lumensimultaneously with meal substrates. Intraluminallipid digestion in postprandial chyme requireslipase activity of at least 40–60 IU/mL through-out the digestive period, which translates into25,000 to 40,000 IU intraduodenal lipase for di-gestion of a regular meal. Because plain enzymepreparations undergo rapid lipase inactivationdue to acid and proteolytic destruction, it isnecessary to administer up to 10-fold more li-pase orally to achieve these quantities withinthe duodenum.

Pharmaceutical inhibition of gastric acid out-put (proton pump inhibitor, H2 receptorblocker) may have beneficial effects when com-bined with an unprotected pancreatin prepara-tion, not only by protection of enzymes duringtheir gastric passage,but also by increasing duo-denal pH and thereby improving enzymaticaction. (Note that, in chronic pancreatitis, duo-denal pH is lower due to impairment of pancre-atic bicarbonate secretion, see above).

Coating of pancreatin preparations withpolymers that resist acidic pH but allow disinte-gration and release of enzymes at pH 5.5 to 6.0can be used to protect exogenous enzymesfrom premature acidic denaturation. However,


2

3

13C-MTG

Hydrolysis�-Oxidation

2-(1-13C) Octanoyl-glycerol +

2 Stearic acid

1,3-Distearyl-2-(1-13C) octanoylglycerol

1

13CO2

Figure 26-5. Principle of the 13C-mixed tri-glyceride breath test.Absorption of 13C-mixed tri-glycerides requires prior hydrolysis by pancreaticlipase (1), which leads to production of free fattyacids (stearic acid) and monoacylglycerol [2-(1-13C)octanoylglycerol].These metabolites are incor-porated into micelles, absorbed,and transported tothe liver (2). Further degradation by hepatic en-zymes and β-oxidation results in formation of13CO2, which is absorbed into the bloodstream,transported to the lung, and exhaled (3).Thus, ex-halation of 13CO2 reflects intestinal lipolysis and isa marker of pancreatic exocrine function.

solid particles that cannot be reduced to piecesless than 2 mm by gastric contractions are notemptied prandially. Instead, they are retained inthe stomach until the return of the interdiges-tive motor pattern. Consequently, acid-resistantcoating of a capsule containing pancreatin pro-tects enzymes from acidic denaturation butleads to dissociation between duodenal deliv-ery of enzymes and nutrients (Keller and Layer,2003). Thus, modern pancreatin preparationsconsist of acid-resistant coated pancreatin mi-crospheres with diameters not exceeding2 mm.These mix intragastrically with the mealand are emptied intact into the duodenum si-multaneously with the meal. In the duodenum,increasing pH causes release and hydrolytic ac-tion of enzymes.

Although the small, coated pancreatin prepa-rations have been shown to be superior com-pared with unprotected pancreatin extracts, itshould be noted that enzyme release requiresseveral minutes after exposure to the intestinalmilieu. This may further delay digestive actionand shift the site of maximal absorption moredistally; this pathophysiological mechanism isexacerbated by accelerated gastrointestinal

transit characteristic for pancreatic exocrine in-sufficiency. For this reason, the physicochemi-cal properties of the microsphere coating arecrucial for the efficacy of enzyme therapy.

Practical Enzyme Supplementation

Enzymes for Steatorrhea

At present, the majority of patients with ex-ocrine pancreatic insufficiency are treated withpH-sensitive pancreatin microspheres takenwith each meal. To reduce steatorrhea to lessthan 15 g of fat per day, a minimal dose of25,000–40,000 IU of lipase per meal are re-quired (Fig. 26-6) (Keller and Layer, 2003). It isrecommended that, for small meals or snacks, 1tablet/capsule (containing 10,000–25,000 IU oflipase) should be swallowed with the start ofthe meal, and that for larger meals, a seconddose should be taken during the meal. In manycases, these doses must be doubled (case re-port).

Efficacy of treatment is checked by clinicalcriteria, mainly by monitoring body weight andconsistency of feces. In treatment-refractory


Figure 26-6. Therapeutic algorithm in pancreatic exocrine insufficiency.

cases, enzyme dosage should be furtherincreased, and the patient should be advised todistribute nutrient intake over five or six smallermeals. If the steatorrhea still does not respond,then the compliance of the patient should bechecked by fecal chymotrypsin measurement;low activities suggest an insufficient intake of en-zymes.

A dose-dependent risk of developing stenoticfibrosing colonopathy (colonic strictures andmarked submucosal fibrosis) has been observedin patients with cystic fibrosis taking ultrahighdoses of pancreatin.Although the mechanism re-mains controversial and may be specific for cys-tic fibrosis,and there is no evidence that patientswith chronic pancreatitis also have an increasedrisk, we therefore do not generally recommenddosages of more than 75,000–100,000 IU of li-pase per meal. In refractory cases, pathophysio-logical and therapeutic alternatives should beconsidered.

Following gastric or intestinal resections,bacterial overgrowth or mucosal infectionssuch as Giardia lamblia as well as other intes-tinal absorption disorders may further compro-mise absorption. Patients with acceleratedgastric emptying due to gastric resections or gas-troenterostomies should be supplemented withpancreatin granule or powder preparations.Similarly, patients lacking gastric acid secretion,including those continuously receiving acidblockers, can be treated successfully with con-ventional unprotected pancreatin preparations.If patients are treated with enzyme supplemen-tation, then use of medium-chain triglyceridesas a dietary adjunct does not further enhancelipid absorption.

Enzymes for Pain Treatment

There is evidence that, in humans, a luminal,protease-mediated, negative feedback systemmay be operative under certain circumstances(Slaff et al.,1984),but it is controversial whether(and rather unlikely that) this mechanism con-tributes to the pathogenesis of pain in patientswith chronic pancreatitis. Several controlledtherapeutic trials in patients with chronic pan-creatitis have yielded conflicting results. More-over, experimental data suggest that hormonally-induced inhibition of pancreatic secretion aloneis ineffective in painful pancreatitis. It is morelikely that amelioration of pain following en-zyme administration originates from correctionof disturbed motor function, such as ileal brake

mechanisms (inhibition of upper gastrointesti-nal functions by ileal nutrient exposure), in-duced by maldigested nutrients.

QUESTIONS

1. How is digestive pancreatic exocrine se-cretion induced?

2. Why is steatorrhea the most important di-gestive malfunction in pancreatic exocrineinsufficiency in chronic pancreatitis?

3. What proportion of normal pancreatic ex-ocrine secretion is needed to preventsteatorrhea?

4. Which pancreatic function tests are indi-rect tests?

5. Which hormonal mediators stimulate pan-creatic exocrine secretion and are usedfor direct pancreatic function testing?

6. Why are pancreatin microspheres withacid-resistant coating usually more effec-tive in reducing steatorrhea than identicallipase doses provided as either plain pan-creatin powder or monolithic capsules ortablets with acid-resistant coating?

Acknowledgmens: Our studies cited in this article weresupported by the German Research Foundation (DFG,grants La 483/5) and the Anna-Lorz-Foundation.

BIBLIOGRAPHY

Chowdhury RS, Forsmark CE: Review article: pan-creatic function testing. Aliment PharmacolTher 17:733–750, 2003.

DiMagno EP, Clain CE, Layer P: Chronic pancreati-tis, in Go VL (ed): The Pancreas: Biology, Patho-biology and Disease. Raven Press, New York,1993, pp. 665–706.

DiMagno EP, Layer P: Human exocrine pancreaticenzyme secretion, in Go VL (ed): The Pancreas:Biology, Pathobiology and Diseases. RavenPress, New York, 1993, pp. 275–300.

Holtmann G, Kelly DG, et al.: Survival of humanpancreatic enzymes during small bowel transit:effect of nutrients,bile acids, and enzymes.Am JPhysiol 273:G553–G558, 1997.

Keller J, Layer P: Circadian pancreatic enzyme pat-tern and relationship between secretory andmotor activity in fasting humans. J Appl Physiol93:592–600, 2002.

Keller J, Layer P: Pancreatic enzyme supplementa-tion therapy. Curr Treat Options Gastroenterol6:369–374, 2003.


Layer P, Go VL, DiMagno EP: Fate of pancreatic en-zymes during small intestinal aboral transit inhumans. Am J Physiol 251:G475–G480, 1986.

Layer P, Keller J: How to make use of pancreaticfunction tests, in Buechler M (ed): Chronic Pan-creatitis: Novel Concepts in Biology and Ther-apy. Blackwell Science, Oxford, UK, 2002, pp.233–242.

Layer P, Keller J: Lipase supplementation therapy:standards, alternatives, and perspectives. Pan-creas 26:1–7, 2003.

Layer P, Peschel S, et al.: Human pancreatic secre-tion and intestinal motility: effects of ileal nutri-ent perfusion. Am J Physiol 258:G196–G201,1990.

Layer P, von der Ohe MR, Holst JJ, et al.: Alteredpostprandial motility in chronic pancreatitis:role of malabsorption. Gastroenterology 112:1624–1634, 1997.

Owyang C, Logsdon CD: New insights into neuro-hormonal regulation of pancreatic secretion.Gastroenterology 127:957–969, 2004.

Read NW, McFarlane A, Kinsman RI, et al.: Effect ofinfusion of nutrient solutions into the ileum ongastrointestinal transit and plasma levels of neu-rotensin and enteroglucagon. Gastroenterology86:274–280, 1984.

Rinderknecht H: Pancreatic secretory enzymes, inGo VL (ed): The Pancreas: Biology, Pathobiol-ogy and Disease. Raven Press, New York, 1993,pp. 219–251.

Slaff J, Jacobson D, Tillman CR, et al.: Protease-specific suppression of pancreatic exocrine se-cretion. Gastroenterology 87:44–52, 1984.

Vantrappen GR, Rutgeerts PJ, et al.: Mixed triglyc-eride breath test:a noninvasive test of pancreaticlipase activity in the duodenum. Gastroenterol-ogy 96:1126–1134, 1989.


CASE REPORT

A 28-year-old man who immigrated to theUnited States a few years ago saw his primarycare physician for mild recurrent abdominaldiscomfort and diarrhea. Past medical historywas benign except for what the patient re-ferred to as “occasional abdominal distress.”Further discussion revealed this to include diar-rhea, flatus, and abdominal pains. He describedthe stool during these episodes to be greasy,pale, and foul smelling. Family history was unre-markable. Both of his parents, as well as threesiblings, had no history of abdominal symp-toms. The patient had been diagnosed withceliac disease and treated with a gluten-freediet at a previous visit. Even though he wascompliant with dietary instructions, symptomscontinued for the next year.The patient notedthat he had experienced these symptoms peri-odically throughout his life, but that they hadworsened since his arrival in the United States.He had not traveled since his immigration andhad no pets or recent animal exposures, andthere was no association between his abdomi-nal discomfort and specific foods. He deniedpast or present use of drugs or alcohol and tookno medications. Concerned that celiac diseasemay have been a misdiagnosis, a physical exam-ination, blood tests, and a diagnostic procedurewere performed.

Physical inspection revealed a healthy adultmale, 1.8 m in height and weighing 76.5 kg. Hewas hemodynamically stable. Noticeable bruiseswere observed on his thighs.The patient attrib-uted them to bumping into furniture,“not an un-common event,” as he sometimes left the lights

off on entering the house to avoid disturbing hiswife. Further examination revealed a mild pe-ripheral neuropathy with numbness, decreasedproprioception, and dulled response to painfulas well as vibratory stimuli in the extremities.Deep tendon reflexes were also reduced.Musclestrength was preserved bilaterally. He had apositive Romberg sign (loss of balance whilestanding eyes closed) and slight ataxic gait. Fun-doscopic examination revealed pigmentedretinopathy.

Blood tests were performed and had thefollowing results: hematocrit 33% (normal is38%–45%); reticulocyte count 2.0% (normal is0.5%–1.5%); mean corpuscular volume 90 (nor-mal). A decreased erythrocyte sedimentationrate and a prothombrin time (PT) of 15 s (nor-mal 11–13 s) were noted. A complete bloodcell count; values for total protein and albumin,ALT, and AST, fasting glucose, Hb-A1c, as well asamylase and lipase were within normal range.Plasma levels of vitamins A and K were belownormal, and there was near absence of vitaminE. Dysmorphic red blood cells (RBC) with mul-tiple thorny projections (acanthocytes) werepresent on a blood smear. All other laboratorytests and findings were within normal limits.

Lipid analysis revealed low plasma choles-terol.A fasting lipid profile showed total choles-terol was 1 mmol/L (normal is 3.5–5.0 mmol/L).On fractionation, HDL was found to be normalwhile LDL was undetectable. Postprandial elec-trophoresis of plasma lipoproteins demon-strated the presence of HDL, but absence ofchylomicrons, VLDL, and LDL. Plasma levels ofapolipoprotein B (apoB) were well below thesensitivity of the test used for their measurement

27

Abetalipoproteinemia

PAUL RAVA and M. MAHMOOD HUSSAIN

290

and determined to be absent, while those ofapoA1 were only moderately reduced.

Examination of a stool sample revealedsteatorrhea (foul-smelling stool with high fatcontent). Stool was negative for fecal occultblood, leukocytes, ova, parasites, and cultures.An endoscopic procedure was performed.Macroscopically, the small intestine was atypi-cally white/yellow in appearance. Microscopyof a biopsy specimen taken from the jejunumshowed normal villous architecture,high fat con-tent in epithelial cells visible on staining withOil-Red O, and no evidence of inflammation, ul-ceration, or parasites. On reviewing the findings,a diagnosis of abetalipoproteinemia (ABL) wasmade,and the patient was treated accordingly.

DIAGNOSIS

Abetalipoproteinemia (ABL) is a rare, autosomalrecessive disease first described by Bassen andKornsweig in 1950. It is characterized by the ab-sence of plasma apoB lipoproteins, fat-solublevitamin deficiencies (A, E, and K), and the pres-ence of acanthocytosis (Table 27-1). Other signsinclude fat malabsorption presenting as steator-rhea, flatus, abdominal discomfort, and progres-sive ataxic neuropathy. The key diagnosticfeature is an extremely low plasma total choles-terol and absence of all apoB lipoproteins (chy-lomicrons,VLDL, and LDL).

Three disorders of lipoprotein metabolismshare these characteristics: familial hypobetal-ipoproteinemia, chylomicron retention disease,and ABL (Table 27-2).The presence or absenceof specific plasma apoB lipoproteins, as well astheir mode of inheritance, can be useful whenattempting to differentiate between these disor-ders. Symptoms associated with familial hypo-betalipoproteinemia are usually milder than forthe other two and are inherited as dominanttraits, that is, symptoms are observed in at leastone parent of an affected offspring. Chylo-micron retention disease is an autosomal reces-sive disorder with a severe phenotype commonlypresenting soon after birth. Plasma lipoproteinanalysis from affected individuals shows a spe-cific absence of chylomicrons (apoB48) butnormal amounts of VLDL and LDL (apoB100). Inour patient, evidence of recessive inheritanceand absence of all apoB-containing lipoproteinsimplicates ABL as the most likely diagnosis.

Malabsorption, an early nonspecific clinicalsign of ABL, is commonly encountered in many

diseases of the gastrointestinal tract. It can beloosely defined as the intestine’s inability to ab-sorb lumenal contents (i.e., sugars, nutrients,fat, water, and/or carbohydrates). Presentingsymptoms of malabsorption are abdominal dis-comfort and diarrhea. Some major causes ofacute and chronic malabsorption are listed inTable 27-3.

In ABL, fat malabsorption is responsible forthe diarrhea and flatus. Symptoms are promi-nent in affected newborns with the severitycoupled to their lipid-rich diets. Chronic diar-rhea results in nutrient wasting and, in somecases, failure to thrive. Since fat malabsorptionis responsible for the initial gastrointestinalsymptoms, it is not uncommon for affected indi-viduals to restrict dietary fat independent ofmedical advice, alleviating the frequency of di-arrhea and flatus. Occasional exacerbations stilloccur,but not serious enough to provoke a clin-ical visit. As a result, affected individuals oftendevelop normally, unaware of their disease, un-til a gross secondary neuropathy becomessymptomatic toward the third decade of life oreven earlier.

The patient described in the case report is atextbook example of this. His history suggeststhe presence of chronic malabsorption sincebirth and a self-discipline resulting in alleviationof gastrointestinal symptoms before his move

Abetalipoproteinemia 291

Table 27-1. AbetalipoproteinemiaSigns and symptoms associated with

abetalipoproteinemiaMalabsorption syndrome (steatorrhea, diarrhea,

flatus, failure to thrive)Spinocerebellar disease

Ataxia, positive Romberg signDecreased proprioception and vibratory sensesLoss of deep tendon reflexes

Night blindness

Laboratory and other diagnostic findingsLow plasma cholesterol levelsAbsence of plasma apoB lipoproteins (chylomicrons,

VLDL, and LDL)Fat-soluble vitamin (A, E, and K) deficienciesPigmented retinopathyAcanthocytosisNormocytic anemia with compensatory

reticulocytosisDecreased erythrocyte sedimentation rateHepatic steatosisGelee blanche intestine by endoscopic evaluationHistological presentations: normal villi, absence of

inflammation, accumulation of neutral lipids

to the United States. The likely reason for thereappearance in severity of diarrhea and overalldiscomfort was the introduction to a high-fatWestern diet.This clearly demonstrates the im-portance of fat restriction in the treatment ofABL symptoms.

ABL is frequently misdiagnosed due to a mul-titude of nonspecific and unrelated symptoms,as well as rarity of the disease.An initial diagno-sis of celiac disease, a more common malab-sorptive disorder, is therefore frequently made.Celiac disease also presents as abdominal dis-comfort and diarrhea with frequent remissionsand exacerbations. Its etiology is unknown, butthere is a clear association with gliadin, a con-stituent of most grains, as well as an undefinedimmunological component. Symptoms of celiacdisease often abate when the individual isplaced on a gluten-free diet, whereas those as-sociated with ABL do not, as this patient clearlydemonstrated.

Endoscopic analysis and intestinal biopsiesare useful to rule out other diseases of the intes-tine and to confirm ABL. In ABL, the intestinal lu-men has a “gelee blanche” or white frothyappearance from massive accumulation oflipids within the mucosa, which persists evenin the setting of a low-fat diet. On microscopicevaluation of a biopsy sample taken from thisregion, inflammation commonly observed inother malabsorption syndromes is usually ab-

sent, and villi have normal appearance. Inspec-tion of individual cells shows distension withcytoplasmic lipid droplets that stain positive forneutral lipids (triglycerides and cholesterolesters). Variable-size droplets, which may ormay not be surrounded by membranes, havebeen noted by electron microscopy (Berriot-Varoqueaux et al., 2000).

Neurological problems are also diagnostic fea-tures of ABL and are consequences of prolonged


Table 27-2. Prominent Genetic Disorders with Decreased or Absent Plasma apoB LipoproteinsGene/Mode Defective Symptoms/ Plasma

Disease of Inheritance Assembly Step Diagnostic Features Lipoproteins

Abetalipoproteinemia mttp/ Biosynthesis of apoB Fat malabsorption, Absence of all (ABL) recessive lipoproteins particle fat-soluble vitamin apoB lipoproteins

deficiencies, spinocerebellar (chylomicrons,disease, retinitis pigmentosa, VLDL, and LDL)acanthocytosis, galee blanche intestine,fatty liver

CMRD/Anderson sar1b/ Budding of vesicles Similar to ABL, absence of Absence ofdisease recessive containing nascent fatty liver, failure to thrive chylomicrons;VLDL

lipoproteins from in infancy, no acanthocytosis and LDL presentER of intestinal cells

CMRD-MSS sar1b/ Same as CMRD Same as CMRD; also Similar to CMRDrecessive congenital cataracts, mental

deficiency, and cerebellaratrophy

Hypobetalipo- apob/ Synthesis of apoB Presents only in a Low to absent apoBproteinemia dominant peptide homozygous or complex lipoproteins

heterozygous state with symptoms similar to ABL

CMRD, chylomicron retention disease; CMRD-MSS, CMRD with Marinesco Sjogren syndrome.

Table 27-3. Causes of DiarrheaAcute

Bacterial, parasitic, and viral infectionsFood poisoningMedications (antibiotics, chemotherapy, laxatives,

nonsteroidal anti-inflammatory drugs etc.)Ingestion of large amount of nonabsorbable sugarIntestinal ischemia

ChronicIrritable bowel syndromeInflammatory bowel disease (i.e., Crohn’s disease)Chronic infectionRadiation enteritisMedicationsLocal cancer/lymphomaPrevious surgery (i.e., gastrectomy)Endocrine causes (diabetes, hypo- and

hyperthyroidism, etc.)Malabsorption syndromes (i.e., celiac disease,

abetalipoproteinemia, etc.)

vitamin E deficiency.The function of vitamin Ein nervous tissue is largely undefined, but gen-eral consensus is that it plays an important an-tioxidant role preventing membrane lipidoxidation, especially of myelin. One of the firstsymptoms to present is the loss of deep tendonreflexes. Proprioception, vibratory senses, anddevelopment of an ataxic gait soon follow. ARomberg sign is usually present in older indi-viduals as long-term untreated disease severelyaffects the dorsal columns. Paraethesias, in astocking-and-glove distribution, may also be ob-served. Autopsy results from patients with ABLhave revealed demyelination within the centraland peripheral nervous systems with loss oflarge myelinated fibers (Sorbrevilla and Good-man, 1964). Electromyography reveals a slowconduction velocity in patients, again suggest-ing demyelination in ABL.

The spinocerebellar degenerative symptomsobserved in ABL are often confused withFriedreich ataxia. Friedreich ataxia is also anautosomal recessive disease, but involves ex-pansion of trinucleotide (GAA) repeats withinthe frataxin gene. Like ABL, presenting symp-toms include progressive ataxia and loss of vi-bratory senses beginning around the seconddecade of life owing to ordered regression oflarge myelinated axons in the peripheral aswell as central nervous systems. Friedreichataxia, however, does not show malabsorptionsymptoms.

Plasma levels of vitamin A, and to a lesser ex-tent vitamin K, are also below normal in ABL.Vi-tamin A deficiency presents as progressivenight blindness.This could explain the bruiseson the patient’s thighs resulting from decreasedvisual acuity in the dark, and thus leading to ex-cessive bumping into furniture. Individuals withchronic low levels may also complain of diffi-culty in driving at night. Pigmented retinopathyis also associated with vitamin A deficiency butis not specific to ABL.

Low levels of vitamin K cause hemostatic ab-normalities and hemorrhage. A prolonged pro-thrombrin time (PT) is a measurable evidenceof this. PT is commonly elevated in liver dis-eases, anticoagulant therapies, inborn or ac-quired deficiencies in the clotting pathways,and severe vitamin K deficiency. In ABL, vitaminK deficiency symptoms are either generallymild or absent. Nonetheless, spontaneous gas-trointestinal bleeding can occur in some indi-viduals.

Other nonspecific findings in ABL include

acanthocytosis and associated abnormalities re-sulting from it. Acanthocytosis is the appear-ance of dysmorphic, “starry-shaped” RBC on asmear.These peculiar looking cells are believedto result from decreased cholesterol contentand abnormal cholesterol:phospholipid ratio incell membranes. Acanthocytes are fragile andprone to increased destruction. A mild normo-cytic anemia (hematocrit < 38% and a normalmean corpuscular volume) with compensatedreticulocytosis is indicative of this destruction.Acanthocytes also contribute to the decreasederythrocyte sedimentation rates observed inABL. Normal RBC tend to aggregate under invitro conditions, forming clusters referred to asrouleaux and sediment more quickly. Acantho-cytes decrease the structured formation ofrouleaux, in effect slowing the rate of erythro-cyte sedimentation.


Lipoprotein Assembly andMicrosomal Triglyceride Transfer Protein

Apolipoprotein B (apoB) is an essential struc-tural component of triglyceride-rich lipopro-teins. These apoB lipoproteins are responsiblefor plasma transport of lipids and fat-soluble vi-tamins, including triglycerides, phospholipids,and cholesterol esters, as well as vitamins A, E,and K (Hussain et al., 1996; Kane et al., 1995).A single apob gene is inherited from both par-ents; however, two tissue-specific, liver(apoB100) and intestine (apoB48), forms of theprotein exist (Fig. 27-1). ApoB100 (100%, full-length single polypeptide of 4536 amino acids),expressed mainly in hepatocytes, is the essen-tial structural component of VLDL, IDL, andLDL. In contrast, differentiated intestinal cellsproduce apoB48, a vital element of chylomi-crons. Production of apoB48 is accomplishedthrough the activity of an intestinal cell-specificnuclear cytosine deaminase, apobec-1.Apobec-1deaminates deoxycytosine to deoxyuracil at po-sition 6666 of the apoB mRNA, resulting in thegeneration of a stop codon (UAA) in place ofthe glutamine codon (CAA) present in apoB100mRNA.The stop codon terminates apoB mRNAtranslation in the intestine, producing a stableand functional single polypeptide of 2152amino acids, apoB48.

Translation of apoB mRNA initiates on free ri-bosomes in the cytoplasm. On production of an


295

Figure 27-1. Schematic diagram of apoB lipoprotein assembly. In the nucleus, the apoB gene is transcribed into a 15-kb mRNA. In the intestine, thismRNA undergoes posttranscriptional editing, resulting in deamination of cytosine6666 to uracil, converting a glutamine codon into a stop codon. In theliver,no such editing occurs. The apoB mRNA is transported from the nucleus to cytosol,where its N-terminal signal sequence is translated.On the bind-ing of the signal sequence to the signal recognition particle, the translation of the mRNA is stalled, and the complex is translocated to the ER. After bind-ing to the ER membrane, translation of the apoB mRNA resumes.Complete translation of apoB mRNA results in the synthesis of apoB48 and apoB100 inintestine and liver, respectively.The newly translated apoB can either be used for lipoprotein assembly (1) or be degraded (2) by endoplasmic reticulum-associated degradation (ERAD) involving proteosomes. Microsomal triglyceride transfer protein (MTP) plays a critical role in the lipoprotein assembly(1) by different mechanisms: (A) MTP can transfer lipids either from membrane to the nascent apoB, forming a primordial lipoprotein. (B) MTP cantransfer lipids from lipid droplets to nascent apoB. (C) MTP can import triglycerides into the ER lumen that can be used for lipoprotein assembly or lipiddroplet formation. (D) MTP can interact with the nascent apoB and help in early folding and lipidation of the molecule leading to the synthesis of a pri-mordial lipoprotein particle. (E) MTP may facilitate the fusion of lipid droplets and the primordial lipoprotein resulting in the synthesis of a nascentlipoprotein. Incompletely assembled apoB lipoproteins are degraded in the ER (3) by mechanisms other than ERAD. PDI, protein disulfide isomerase.

N-terminal ER signal sequence, the complex isescorted to the rough ER by cytoplasmic chap-erones, where protein synthesis continues.Proper folding, disulfide bond formation, andacquisition of lipids (phospholipids, triglyc-erides, and cholesterol) accompany cotransla-tional translocation of the growing peptide intothe lumen. Defects in any of these events leadto degradation of the nascent apoB polypep-tide. Since apoB mRNA is constitutively pro-duced, degradation of the nascent peptidesignificantly influences lipoprotein assemblyand secretion. ApoB lipoprotein assembly be-gins with an initial highly dense “primordiallipoprotein” and culminates with a larger “nas-cent lipoprotein” particle (Hussain et al., 2003).Quality control processes monitoring apoBlipoprotein assembly are extremely efficient.Secretion is only allowed to occur upon apoBobtaining an appropriately folded structure andsufficient lipid content.

In ABL, an early step in apoB lipoprotein as-sembly shared by intestinal and liver cells isdefective. The net result is near absence of allplasma apoB lipoproteins.ApoB synthesis froma mRNA transcript occurs, but its successful as-sembly into the mature lipoprotein particledoes not. The inability to assemble apoB intolipoproteins was shown to be due to a defect inthe mttp gene in affected individuals (Wetterauet al., 1992). Its translational product is an 894-amino acid, 97-kd, polypeptide that exists in theER complexed with a 55-kd protein disulfideisomerase which is believed to maintain solubil-ity, physiologic activity, and ER retention of the97-kd peptide. The heterodimeric complex ofthe 97-kd and 55-kd subunits is referred to asmicrosomal triglyceride transfer protein (MTP)(Wetterau et al., 1992).

Several functions of MTP have been identi-fied; all have been implicated in coordinatingsuccessful lipoprotein assembly (Fig. 27-1). MTPtransfers lipids between vesicles in vitro, andthis activity is likely to be its major function.MTPcan pick up lipids from membrane (step A) orvesicles and droplets (step B) and transfer themto the nascent apoB. In addition, the lipid trans-fer activity of MTP has been implicated in the ac-cretion of neutral lipids from the cytosol into theER lumen (step C). Compounds that inhibit invitro transfer activity of MTP decrease apoB se-cretion by cells, indicating that this activity is es-sential for apoB lipoprotein secretion.Apart fromtransferring lipids,MTP has been shown to inter-act physically with apoB (step D). This activity

also appears important since inhibition of MTP-apoB binding results in decreased apoB secre-tion. Last, MTP can also associate with lipiddroplets in the ER and may stabilize these parti-cles. It may also help in the fusion (step E) oflipid droplets and primordial lipoproteins, result-ing in the formation of nascent, triglyceride-richlipoproteins (Hussain et al., 2003).

Biochemical Explanation for Symptoms

Chronic malabsorption does not fully explainthe different extents of fat-soluble vitamin defi-ciencies associated with ABL. More specifi-cally, why are plasma vitamin E levels moreseverely affected than those of vitamins A orK? The answer for this can be traced to apoBlipoprotein biosynthesis and catabolism (Fig.27-2). Just as observed for lipids, hydrophobic,fat-soluble vitamins require apoB lipoproteinsas vehicles for plasma transport. The relianceof each fat-soluble vitamin on apoB lipopro-teins varies, and this variable dependency is di-rectly related to the severity of symptomsobserved in ABL.

Vitamin E, like neutral lipids, requires apoBlipoproteins at every stage of its transport (Fig.27-2). Dietary vitamin E becomes emulsified inmicelles produced during the digestive phaseof lipid absorption and permeates the intestinalepithelium, similar to fatty acids and choles-terol. Uptake of vitamin E by enterocytes ap-pears to be concentration dependent. Withinintestinal cells, vitamin E is packaged into chy-lomicrons and secreted into lymph. Duringblood circulation of chylomicrons, some vita-min E may be released to the tissues as a conse-quence of partial lipolysis of these particles byendothelial cell-anchored lipoprotein lipase.The rest remains associated with chylomicronremnants.Remnant particles are mainly endocy-tosed by the liver and degraded, resulting in therelease of fat-soluble vitamins.

In hepatocytes,vitamin E can take two routes.A fraction of it is packaged as VLDL and reentersthe circulation, while excess is excreted in thebile. Plasma lipolysis of the VLDL particle againresults in release not only of lipids, but also of vi-tamin E, with the remainder left with the LDLparticles.This fraction can be further distributedto tissues via LDL receptor-mediated endocytosisor transferred between lipoproteins, mainly toHDL, by plasma lipid transfer proteins.Thus, mo-bilization of vitamin E from intestinal and liver


cells is critically dependent on apoB lipoproteinassembly and secretion.

Uptake of vitamin A by intestinal cells is car-rier mediated at low concentrations. If higheramounts are present, then it enters intestinalcells via diffusion. Like vitamin E, it also exploitschylomicrons as a vehicle to exit the enterocytes

and reaches the liver as a component of theremnant particles. In the liver, the metabolismsof vitamin A and vitamin E diverge. Here, unlikevitamin E and triglycerides, vitamin A is eitherstored in Ito cells or secreted bound to retinol-binding protein for plasma transport and fur-ther tissue distribution.


Figure 27-2. An overview of lipoprotein metabolism. Intestine and liver are two majororgans that play crucial roles in transporting dietary and endogenous lipids. Intestine(dashed lines): Dietary triglycerides, phospholipids, and cholesterol esters are hy-drolyzed in the lumen of the intestine and are absorbed by enterocytes. In the intestinalcells, fat is resynthesized and assembled into particles referred to as chylomicrons. Dur-ing lipoprotein assembly, fat-soluble vitamins are also incorporated into these particles.Assembly of these particles requires apoB48 and microsomal triglyceride transfer pro-tein (MTP). In the circulation, chylomicorns are hydrolyzed by endothelial cell-boundlipoprotein lipase (LPL), and adjacent cells take up the released free fatty acids. Thelipoproteins left after hydrolysis, called chylomicron (chylo) remnants, contain signifi-cant amounts of dietary lipids and fat-soluble vitamins.These particles acquire apoE (E)from plasma, which is a high affinity ligand for LDL receptors (LDL-R) and LDL receptor-related protein (LRP). Recognition of apoE by these receptors results in the internaliza-tion and degradation of lipoprotein particles.Liver (solid lines): In hepatocytes, lipids areresynthesized and packaged into lipoproteins called very low density lipoproteins(VLDL).The VLDL contain apoB100, lipids, and vitamin E. In the circulation, these parti-cles are also hydrolyzed by the endothelial cell-bound lipoprotein lipase-generatingintermediate-density lipoproteins (IDL) and low-density lipoproteins (LDL). IDL acquireapoE, similar to chylomicron remnants, and are cleared rapidly from plasma by receptor-mediated endocytosis. LDL do not acquire apoE; however, apoB100 present in these par-ticles is recognized by LDL receptors expressed in liver and other peripheral tissues.Since apoB100 has a lower affinity for the receptor than does apoE, LDL are removedslowly with a plasma half-life of about 24 h.

The ultimate dependence of vitamin E fortransport by liver and intestinal cells via apoBlipoproteins may explain the severe vitamin Edeficiency observed in ABL. In contrast, apoBlipoprotein assembly is required only for themobilization of dietary vitamin A by the intes-tinal cells; liver distributes the endogenousvitamin A to other tissues by retinol-bindingprotein.Unlike vitamins E and A,dietary vitaminK may only partially depend on apoB lipopro-teins for its transport across the intestinal ep-ithelial cells and tissue targeting and mayexplain why bleeding diathesis is rarely ob-served in ABL.

THERAPY

Presently, ABL is not a curable disease, but theassociated morbidity and premature death arepreventable. Thus, early diagnosis and treat-ment are essential to avoid secondary neuro-logical and ophthalmic pathologies and toprovide an opportunity of a normal life foraffected individuals. Initiation of treatment, af-ter delayed diagnosis and symptomatic devel-opment, can prevent further progression ofsymptoms.

To treat the malabsorption and subsequentdiarrhea, lipid-poor diets (<5 g/day) should beimplemented with a restriction of triglyceridescontaining long-chain fatty acids. Medium-chainfatty acids rely on other protein carriers besidesapoB (i.e., albumin) for plasma transport, mak-ing them an ideal lipid substitute. However,long-term supplementation should be cau-tioned as associated hepatic fibrosis could oc-cur. Diets should also contain increased proteinand carbohydrate content to compensate forcaloric loss from fat restriction.

Most of the neurological pathologies arepreventable even in the setting of chronic mal-absorption associated with ABL. Lifelong sup-plementation is required to treat vitamindeficiencies, especially that of vitamin E. Oralroutes are effective. However, large doses arerequired to overcome the inefficiency of alter-nate transport pathways (e.g., albumin andHDL). Prescription of large doses of vitamin E(100–300 mg/kg/day) results in only a moder-ate increase in plasma levels (<50% of the nor-mal range), yet this appears sufficient toprevent development and progression of theassociated symptoms.

Vitamin A should be supplemented at 5000

IU/day by the oral route with routine bloodwork to monitor plasma levels.Toxicity due toexcess vitamin A is not common but is poten-tially fatal. Acute toxicity presents as vertigo,diplopia, seizures, and exfoliative dermatitis.Chronic toxicity manifests very differently (i.e.,dry skin, alopecia, amenorrhea, symptoms ofliver fibrosis, etc.). It is imperative to monitorplasma vitamin A levels regularly, especially inthe setting of the large doses required to treatABL.

Since vitamin K deficiency varies amongpatients, not all will require supplementation.When it is needed, the recommended oral doseis 1–2 mg/day. Toxicity from dietary vitamin Khas not been described in adults, but a par-enteral dose may cause hemolytic anemia in in-fants; therefore, caution should be taken whenadministered to such patients.

In summary,ABL is a rare, often-misdiagnoseddisease, presenting initially as severe fat malab-sorption and later with predominantly neuro-logical sequelae.The symptoms and findings arenonspecific except for the complete absence ofplasma apoB lipoproteins and below normalplasma total cholesterol. Before widespread useof cholesterol screening, this treatable diseasewas easily overlooked or misdiagnosed. Treat-ment includes dietary restriction of long-chainfatty acids as well as lifelong supplementationof vitamins A, E, and K.

QUESTIONS

1. Why do abetalipoproteinemia (ABL), chy-lomicron retention disease, and homozy-gous hypobetalipoproteinemia present sosimilarly? How would one differentiate be-tween them?

2. Acanthocytosis is associated with ABL, butnot with chylomicron retention disease.Explain.

3. Hypobetalipoproteinemia presents clini-cally only in the homozygous state. Whydoes it sometimes present similar to ABLwith absent plasma apoB lipoproteins andat other times as a much milder pheno-type?

4. Of the fat-soluble vitamins (A, D,E, and K),only vitamin D deficiency is not associ-ated with ABL and is not a required sup-plement.Why?

5. Why has the use of MTP lipid transferactivity inhibitors been unsuccessful as


a treatment for hypercholesterolemia?(Hint: think of possible adverse affects invivo.)

6. A nonfunctional MTP gene/protein causesABL.What are the possible affects on lipidlevels and cardiovascular pathology if thisprotein is overexpressed?

Acknowledgment: National Institutes of Health grantsDK46900 and HL64272 partially supported this work.

BIBLIOGRAPHY

Bassen FA, Kornzweig AL: Malformation of the ery-throcytes in a case of atypical retinitis pigmen-tosa. Blood 5:381–387, 1950.

Berriot-Varoqueaux N, Aggerbeck LP, et al.: Therole of the microsomal triglyceride transfer pro-tein in abetalipoproteinemia. Annu Rev Nut.20:663–697, 2000.

Chan L, Chang HJ, Nakamuta M, et al.: Apobec-1and apolipoprotein B mRNA editing. BiochimBiophys Acta 1345:11–26, 1997.

Hussain MM,Kancha RK,Zhou Z,et al.:Chylomicronassembly and catabolism: role of apolipoproteinsand receptors. Biochim Biophys Acta 1300:151–170,1996.

Hussain MM, Shi J, Dreizen P: Microsomal triglyc-eride transfer protein and its role in apoB-lipoprotein assembly. J Lipid Res 44:22–32,2003.

Kane JP, Havel RJ: Disorders of the biogenesis andsecretion of lipoproteins containing the Bapolipoproteins, in Scriver CR, Beaudet AL, SlyWS,Valle D (eds): The Metabolic and MolecularBasis of Inherited Diseases. McGraw-Hill, NewYork, 1995, pp. 1853–1886.

Sorbrevilla LA, Goodman ML: Demyelinating CNSdisease, macular atrophy and acanthocytosis(Bassen-Kornzweig syndrome). Am J Med 37:821–832, 1964.

Wetterau JR, Aggerbeck LP, Bouma M, et al.: Ab-sence of microsomal triglyceride transfer pro-tein in individuals with abetalipoproteinemia.Science 258:999–1001, 1992.


CASE REPORT

The patient, an 84-year-old male resident of aveteran’s home, was admitted to the hospitalfollowing a fall in which he had sustained afracture of his left radius and several lacera-tions. Physical examination on admission re-vealed a well-nourished elderly male in no greatdistress (body mass index (BMI), 22.5 kg/m2).He was moderately confused but with prompt-ing was oriented to time, place, and person.There was no obvious pallor, cyanosis, or jaun-dice, and his skin and mucous membranes wereunremarkable apart from the injuries he sus-tained in the fall. His right lower extremity wasswollen and tender, and he walked cautiouslyand with a pronounced limp.

His blood pressure was 145/85 mm Hg (nor-mal is 120/80 mm Hg). His pulse rate was 80beats per minute (normal is 75 beats perminute), and his respiratory rate was 16 perminute (normal is 12–14 per minute). His tem-perature was 98.2°F (normal is 98.4°F). Exceptfor a soft systolic ejection murmur in the neckand in the left parasternal area, his cardiovascu-lar, respiratory, abdominal, and genitourinarysystems were all normal. The neurological ex-amination was normal except for impaired cog-nitive function, a positive Romberg sign, andequivocal plantar responses. (A patient is con-sidered to have a positive Romberg sign when,in a standing position, the patient is more un-steady with eyes closed than with eyes open.This indicates a loss of proprioceptive control.)There were no signs of head injury.

The director of the veteran’s home statedthat the patient’s memory had been gradually

failing over the past 2 years. His balance hadalso become mildly impaired, which resulted inthe fall and associated injuries. His medical his-tory revealed that the patient had been a heavysmoker, but he stopped smoking 20 years ago.It also revealed that he had suffered from nu-merous recurrences of a duodenal ulcer thatwas managed with diet, antacids, and H2 recep-tor antagonist drugs. However, he has had norecurrences of ulcer symptoms since age 70years.

Laboratory investigations yielded the follow-ing results: hemoglobin 148 g/L (normal is140–180 g/L); hematocrit 45% (normal 42% is52%); mean corpuscular volume 98 fL (normalis 80–100 fL); red cell distribution width 14%(normal is 11%–13 %); reticulocytes 1% (nor-mal is 0.5%–1.5%); white blood cell count7.8 × 103/mm3 (normal is 4–8 × 103/mm3)with a normal differential count; platelets230 × 103/mm3 (normal is 140–460 × 103/mm3);prothrombin time 12 s (normal or control is11.9–14.8 s). The following analytes were allwithin the normal range: blood urea nitrogen(BUN), creatinine, electrolytes, calcium, phos-phorus, magnesium, glucose,TSH, total protein,albumin, lactate dehydrogenase (LDH), andliver function markers, including alkaline phos-phatase. Serum iron and folate concentrationswere also within the normal range; serumcobalamin (vitamin B12) was 230 pg/mL, at thelower end of the normal range (normal is200–850 pg/mL). Holotranscobalamin II (holo-TCII) was undetectable. The low normalserum vitamin B12 and undetectable holo-TCIIindicated deficient intake of vitamin B12. How-ever, serum 25-hydroxyvitamin D, vitamin C,

28

Vitamin B12 Deficiency

DOROTHY J.VANDERJAGT and DENIS M. McCARTHY

300

carotene, and trace minerals were all withinnormal limits.

Chest x-ray showed mild unfolding of theaortic arch and scattered calcifications in thearea of the aortic valve.There was no change incardiac size. To exclude the presence of sub-dural hematoma, a CT of the head was obtainedbut showed no focal abnormalities. However,there was mild diffuse bilateral cortical atrophy,especially in the frontal areas.

On the day of admission, the patient had de-veloped a deep venous thrombosis in his rightcalf, a site not involved in the injury. In investi-gating the underlying cause of the deep venousthrombosis, serum homocysteine was mea-sured and found to be 17.4 µmol/L (normalis < 14 µmol/L).To distinguish between folic acidand vitamin B12 deficiencies, a serum methyl-malonic acid (MMA) assay was performed; ityielded a result of 0.59 µmol/L MMA (normalis < 0.30 µmol/L). This confirmed the presenceof vitamin B12 deficiency, despite a serum B12

concentration that was within the normal range.

DIAGNOSIS

Most often, the classical vitamin B12-deficient pa-tient is an elderly subject who presents withinsidious onset of nonspecific symptoms of ane-mia (e.g., lack of energy, easy fatigability, deterio-ration in work ability). If the anemia is severe,then cardiovascular symptoms such as dyspneaon exertion may also be evident. Depending onthe etiology of the condition, other signs of mal-nutrition or some other underlying disease (e.g.,Crohn’s disease) may also be present.At this latestage of the disease (now rarely seen), neurologi-cal symptoms predominate, notably paresthesiassuch as “tingling,”“pins and needles,”“numbness,”and other sensory loss manifesting as a “glove-and-stocking” distribution involving fingers andtoes.These symptoms usually start in the lowerlimbs. Physical examination reveals marked pal-lor, with a pale lemon yellowish tint to the skin,conjunctivae, and mucous membranes. Thetongue is painful, atrophic, and pale, but onlyrarely ulcerated. Neurologically, there may beloss of vibration sense, diminished reflexes (es-pecially in the ankles), and a positive Rombergsign or other evidence of involvement of theposterior or lateral columns of the spinal cord.

In pernicious anemia (PA), the urine containsexcess urobilinogen, and gastric analysis revealshistamine-fast achlorhydria. Laboratory examina-

tion reveals a low serum B12 (usually under200 µg/mL) and the presence of a severe macro-cytic anemia in the peripheral blood, with ahigh mean corpuscular volume (MCV, usually110–125 fL) and a blood smear showing aniso-cytosis, poikilocytosis, a few nucleated red cells,normoblasts, rarely even megaloblasts, and fewor no reticulocytes.There is also leukopenia in-volving granulocytes, and the polymorphs pres-ent are mature and hypersegmented and show3–7 nuclei per cell; platelets are moderately re-duced (Fig. 28-1).When the cause of the vitaminB12 deficiency is classical PA, the patient’s serummay also contain antibodies directed againstparietal cells or intrinsic factor (IF). Such pa-tients may also show serological or clinical fea-tures of a variety of other autoimmune diseasessuch as type 1 diabetes, hypothyroidism, pri-mary biliary cirrhosis, hypoadrenalism, or myas-thenia gravis.

Over the past 50 years, the presentation ofvitamin B12 deficiency has changed consider-ably, at least in developed countries, wheremegaloblastic anemia has become less com-mon and is usually detected and treated beforethese relatively late hematological and neuro-logical features develop. It is widely acceptedthat neurological and psychiatric features candevelop in the absence of anemia (Lindenbaumet al., 1988), especially in patients receiving fo-late supplements, and can be accompanied byserum vitamin B12 concentrations within nor-mal limits, albeit at the low end of the normalrange. At the other end of the socioeconomicscale, in areas where malnutrition is wide-spread, there is mounting concern about theneed for vitamin B12 supplements, at least invulnerable groups (e.g.,young pregnant women,infants, and the elderly).As recently as 1992, anoutbreak of myeloneuropathy in Cuba due todietary vitamin B12 deficiency involved morethan 50,000 patients (Roman, 1994).

In North America and the United Kingdom,there is growing concern that many elderly sub-jects with mild or no anemia and normal serumB12 concentrations are suffering from preventa-ble and correctable vitamin B12 deficiency dis-ease.Today, in economically advanced countries,as many as 75% of vitamin B12-deficient patientspresent with various neurological symptoms,including somatic and autonomic neuropathies,myelopathies, cortical atrophy, and dementia(Carranza, 2002). In addition, vitamin B12-depleted subjects may develop postoperativemyeloneuropathy when exposed to anesthesia

Vitamin B12 Deficiency 301

with nitrous oxide (Metz, 1992). In the absenceof screening tests, many individuals institutional-ized for senile dementia or Alzheimer disease mayactually have treatable vitamin B12 deficiency(Carranza, 2002).The reversibility of such neuro-psychiatric illnesses following vitamin B12

therapy is still under debate; it is clear that re-versibility deteriorates with delay in institutingtherapy.Because screening is not readily availableand testing is expensive, most geriatricians in theUnited States today simply prescribe monthly or3-monthly injections or oral dietary supplementsof vitamin B12 for all patients over age 65 years.

Vitamin B12 status is usually determined bymeasuring vitamin B12 concentrations in serumby competitive protein-binding radioimmunoas-say. In such an assay, the vitamin B12 in the pa-tient’s serum and the unlabeled B12 in the

standard compete with the radioactive tracer(57Co) for binding sites on purified intrinsicfactor (binding protein). Because of the relation-ship between folate and vitamin B12 metabolism,folate and vitamin B12 should both be assayed si-multaneously in the patient’s serum using ra-dioactive tracers and binding proteins specificfor each vitamin. Normal values for serum vita-min B12 are 200–835 pg/mL (148–616 pmol/L).However, low or normal serum vitamin B12 con-centrations may be present in clinically appar-ent vitamin B12 deficiency.

Direct assays for vitamin B12, such as thecompetitive protein-binding immunoassays, de-tect all forms of vitamin B12 in the serum, in-cluding physiologically inactive analogues. Thevitamin B12 bound to transcobalamin II (TCII)has been shown to be the physiologically


Figure 28-1. Characteristic features of vitamin B12 or folate deficiency in the peripheralblood include larger-than-normal mature red blood cells (oval macrocytes with an ele-vated mean corpuscular volume) and large hypersegmented neutrophils (1 neutrophilwith six nuclear lobes or more than 5% neutrophils with five nuclear lobes).These char-acteristics are associated with megaloblastic changes in the bone marrow and are the re-sult of fewer cell divisions due to impaired DNA synthesis.

relevant pool of vitamin B12, and a decrease inholo-TCII concentrations is one of the earlysigns of deficient intake of vitamin B12 (Her-zlich and Herbert, 1988). Methods have beendeveloped to measure holotranscobalamin(Nexo et al., 2002). (Holotranscobalamin refersto transcobalamin that has B12 bound to it.)The expected concentration range of holo-TCIIin healthy adults is between 40 and150 pmol/L.

Indirect indicators of vitamin B12 deficiencyinclude measurements of the metabolites ho-mocysteine and methylmalonic acid (MMA) inserum and MMA in urine (see the BiochemicalPerspectives section). Whereas the serum ho-mocysteine concentration increases during fo-late or vitamin B12 deficiencies, the serum andurine MMA concentrations increase only in vi-tamin B12 deficiency. Therefore, MMA determi-nations can be used to differentiate vitamin B12

deficiency from folate deficiency. The normalconcentration of MMA in serum ranges from0.08 to 0.28 µmol/L. MMA is quantified usinggas-liquid chromatography and mass spectro-metry. Elevated concentrations of MMA andhomocysteine in serum may precede the de-velopment of hematological abnormalities andreductions in serum vitamin B12 concentra-tions. One should be aware that other condi-tions, including renal in sufficiency and inbornerrors of metabolism, can also result in ele-vated serum levels of MMA.


A severe anemia now known to be due to vita-min B12 deficiency was first described in 1855by the English physician Thomas Addison.Shortly thereafter, it was discovered that thisanemia was linked to severe macroscopic andmicroscopic atrophy of the gastric mucosa. Be-cause of its usually fatal outcome, the conditionwas named pernicious anemia (PA) in 1872 bythe German physician Anton Biermer. Addison-ian pernicious anemia was the subject of thefirst definitive description of the disease byGardner and Osler in 1877, who pioneered theuse of microscopy to describe the unique ap-pearance of strange, large, nucleated red bloodcells (macro-ovalocytes) in the peripheralblood. These patients had mild jaundice (alemon-yellow color to their skin), severe pe-ripheral neuropathy, and progression to deathwithin 1 year from the time of onset of neurol-

gical complications and anemia. In 1880,Ehrlich coined the descriptor megaloblastic forthe appearance of the circulating blood cells;however, another half century passed before itwas discovered that feeding patients largeamounts of animal liver was curative.

This observation led to the discovery that adietary compound (extrinsic factor) was ab-sorbed only after combination with a proteinsecreted by the normal stomach (intrinsic fac-tor [IF]), and that the IF was missing from thesecretions of the atrophic stomach found in pa-tients with PA.The extrinsic factor, later namedvitamin B12, was obtained in crystalline form in1948, and its structure was defined by x-raycrystallography by Dorothy Hodgkins, an ac-complishment for which she received the No-bel Prize for Chemistry in 1964.

Over the next 20 years, it was learned thatthe gastric mucosal and glandular atrophy wereprincipally due to an autoimmune gastritis ac-companied by circulating antibodies directedagainst parietal cells, IF, and other antigens.These antibodies damaged the patient’s mucosaand abolished the secretion of hydrochloricacid and IF, both of which were believed tooriginate from parietal cells in humans.Thus, insome definitions, the use of the term perni-cious anemia is restricted to those cases of vi-tamin B12 deficiency anemia caused by absentsecretion of IF as a result of autoimmune gas-tropathy.

Other causes of gastric atrophy, such asthose due to Helicobacter pylori,AIDS, or radia-tion injury, can lead to a similar outcome butfrom different pathogenic mechanisms. There-fore, vitamin B12 deficiency, resulting in neuro-logical,psychiatric,metabolic,and hematologicaldisorders, can arise from any one of the manycauses listed in Table 28-1. For this reason, theterm pernicious anemia (PA) is used here to de-scribe only the classical disease that is associ-ated with IF deficiency due to autoimmunegastritis.

Some authors use the term adult onset per-nicious anemia to distinguish this conditionfrom rare disorder subdivisions of “perniciousanemia” due to congenital defects in IF secre-tion or structure or to various types of entero-cyte cobalamin malabsorption. In all othersituations, the term vitamin B12 deficiency isused, and an associated anemia, if consequenton it, is called megaloblastic anemia, bearingin mind that identical appearances of the pe-ripheral blood and the bone marrow may be


observed in patients deficient in folic acid, asdescribed below.

The prevalence and impact of vitamin B12

deficiency vary in different groups, races, andgeographic areas (Stabler and Allen, 2004). InNorth America, true autoimmune PA is uncom-mon, far less common than vitamin B12 defi-ciency as a whole, especially in the elderlypopulation. Older studies, mainly from Scandi-navia, indicated a prevalence of PA in the popu-lation of 0.1%–0.2%, but a study of people over60 years of age in Los Angeles found 2.3% ofsubjects affected, of whom most (1.9%) wereundiagnosed (Carmel, 1996).The prevalence ofundiagnosed PA was 2.7% in women and 1.4%in men; the rates were 4.3% in black women,4.0% in white women, but lower in women ofother races.These figures refer to patients whohad characteristic antibody profiles and whosedisorder corrected with administration of IF.Vitamin B12 deficiency per se was not assessed(Carmel, 1996). Other studies in subjects over

60 years of age also suggest a prevalence of PAof 1%–3% (Stopeck, 2000).

However, there has been a growing aware-ness that gastric atrophy from H. pylori infec-tion, particularly when there is long-term use ofproton pump inhibitors (PPIs) in infected pa-tients and the organism has not been elimi-nated, are causing much of the problem.Reports suggested that about 15% of over 40million elderly people living in the UnitedStates have metabolic or clinical evidence of vi-tamin B12 deficiency, and the number of elderlyat risk is projected to grow to 65 million by theyear 2050 (Dharmarajan and Norkus, 2005), anumber much higher than the currently esti-mated 800,000 cases of PA in the United States(Carmel, 1996).

Among the remaining cases of B12 deficiency,food-cobalamin malabsorption due to achlorhy-dria or hypochlorhydria caused by H.pylori gas-tritis, even in the absence of atrophy or thewidespread use of proton pump inhibitors or


Table 28-1. Differential Diagnosis of the Causes of Vitamin B12 Deficiency1. Dietary deficiency of vitamin B12

Vegetarian diets, especially in vegansProlonged severe malnutrition

2. Gastric disorders involving intrinsic factorAutoimmune Addisonian pernicious anemia (major)Severe gastric atrophy due to Helicobacter pylori infectionCongenital abnormalities of intrinsic factorTotal gastrectomy

3. Inadequate gastric release of food-bound vitamin B12

Atrophic gastritis with achlorhydria/deficient proteases (H.pylori [major],AIDS and other causes)Partial gastrectomy, gastroenterostomySustained hypochlorhydria due to drugs (proton pump inhibitors, high dose H2-receptor antagonists)*

4. Impaired degradation of R-proteins in intestinePancreatic enzyme deficiency*Zollinger-Ellison syndrome*

5. Competition for vitamin B12 with hostBacterial overgrowth syndromes in various diseases (diabetes,* scleroderma, jejunal diverticulosis, blind loops)Tape worms (Diphyllobothrium latum)

6. Absent or diseased ileal mucosaResection of at least 3 feet of terminal ileum for any reasonileal bypass operations, ileal loop bladdersDiseases of ileal mucosa (Crohn’s or Bechet’s disease,TB, actinomycosis, lymphoma, ulcerative ileo-jejunitis [due to

sprue], radiation enteritis, HIV enteropathy, Yersinia enteritis)

7. Impaired translocation of vitamin B12 or vitamin B12–F complexAbsent ileal IF receptorTranscobalamin II deficiencyImerslund-Grasbeck syndromeMetformin therapy,* possibly colchicine*

*Although these have been linked to malabsorption of B12 (e.g., cause abnormal Schilling tests), they have not been associated withmegaloblastic anemia or neuropsychiatric diseases. Intrinsic factor is normally secreted in amounts greatly in excess of normalrequirements, so that so that a 90% reduction in intrinsic factor secretion may be of little or no consequence.

IF, intrinsic factor.

H2 receptor antagonists for various acid pepticdisorders, accounts for much of the clinical andsubclinical deficiency. Such individuals usuallyhave sufficient IF to absorb low doses of aque-ous vitamin B12.Malabsorption of vitamin B12 insubjects with H. pylori gastritis reverses witheradication of the infection without therapeuticadministration of B12 (Avcu et al., 2001; Kaptanet al., 2000), as does the malabsorption associ-ated with abnormal bacterial colonization ofthe stomach or small intestine. Nevertheless,there are numerous current investigations ex-amining whether H. pylori, through mecha-nisms of molecular mimicry or through theperpetuation of autoimmune gastritis, may alsoplay some part in the genesis of classical PA.

Lacking large, prospective national studiesthat screen for inadequate vitamin B12 intakewith serum holo-TCII assays and for subclinicalor clinical disease using either urinary or serumassays of MMA, the true magnitude of vitaminB12 deficiency remains unknown. It is clear thatscreening based on serum vitamin B12 measure-ment is inadequate to the task of detecting neuro-logical and psychiatric disease, and unknownnumbers of cases with reversible neuropathy,cognitive impairment, and dementia are beingleft untreated or undetected until it is too latefor successful therapy. Screening defined popu-lations for vasculotoxic and neurotoxic concen-trations of homocysteine in serum raises anumber of additional issues, some of which alsoinvolve deficiencies of folic acid.Vitamin B12 de-ficiency in the U.S.population is further compli-cated by U.S. Food and Drug Administrationregulations that mandate folate fortification ofthe diet and, by preventing megaloblastic ane-mia, mask or exacerbate B12 deficiency statesother than anemia. Similar issues arise in othercountries where malnutrition is common, af-fecting particularly pregnant women and chil-dren; where meat is scarce; and where there arelarge populations of vegetarians or subjectsconsuming parasite-infested fish.

The term vitamin B12 (cobalamin) is usedto describe a family of compounds that containa planar corrin ring with a single cobalt atom atthe center. The structure of vitamin B12 isshown in Figure 28-2. The corrin ring system,which resembles the porphyrin rings of heme,is composed of four reduced pyrrole ringslinked by three methylene bridges and a singlesaturated bond between rings 3 and 4. Thecobalt atom is coordinated to each of the nitro-gen atoms of the four pyrrole rings, a dimethyl-

benzimidazole group, and a sixth ligand, whichmay be a cyanide, hydroxyl, methyl, or adenosylmoiety.

The vitamin B12 that occurs in nature is pro-duced almost entirely by bacterial synthesis inanimals but not in humans (Battersby, 1994).The richest dietary sources of vitamin B12 areorgan meats, such as liver and kidney. Lesseramounts are present in shellfish, chicken, fish,muscle meats, and dairy products (the principalsource in lacto-vegetarians). Plants contain novitamin B12 unless they are contaminated bybacteria, and foods that contain microorgan-isms often provide the only source of vitaminB12 for strict vegetarians, such as the vegans ofsouthern India.

The ligand attached to the cobalt atom deter-mines the activity of vitamin B12 in human enzy-matic reactions.The two active coenzyme formsare methyl-cobalamin and 5'-adenosylcobalamin,the primary form of vitamin B12 in tissues.Cyanocobalamin, the therapeutic form of vita-min B12 contained in vitamin supplements, isproduced by the cleavage of the unstable link


Figure 28-2. The structure of vitamin B12 (cobal-amin).Vitamin B12 is composed of a planar corrinring containing a cobalt atom at the center. Thecorrin ring system is composed of four pyrrolerings and is similar to porphyrin in heme. Thecobalt atom is coordinated to the nitrogens ofeach pyrrole ring and to a dimethylbenzimidazolegroup.The R group attached to the sixth coordina-tion site of the cobalt atom in vitamin B12 can be a–CN, –OH, –CH3, or adenosyl group.

between the 5'-deoxyadenosyl group and thecobalt ion in adenosylcobalamin that occursduring its isolation from liver. The adenosylgroup is replaced by a molecule of cyanide thatis leached from the charcoal columns used in thepurification procedure. Cyanocobalamin, whichis not physiologically active, is readily hydrolyzedin tissues to hydroxycobalamin. Hydroxy-cobalamin is converted in the body to adenosyl-cobalamin by successive reactions involving twoflavoprotein-dependent reductases, which re-duce Co3+ sequentially to Co1+. Co1+, a powerfulnucleophile, then attacks the 5'-carbon of ATP,ex-pelling a triphosphate anion and resulting in thesynthesis of 5'-adenosylcobalamin.

In the diet, vitamin B12 is bound to proteins.Although some release of protein-bound vita-min B12 begins in the mouth, most of the re-lease occurs in the stomach on exposure offood to gastric acid (HCl) and the proteolyticenzyme pepsin. For this reason, either hypo-chlorhydria (abnormally low concentration ofHCl in gastric fluid) or achlorhydria (the ab-sence of HCl in gastric fluid) may decrease theavailability of dietary vitamin B12 for absorptionby preventing the activation of pepsinogen topepsin, the principal enzyme responsible forproteolysis in the stomach. Achlorhydric pa-tients with adequate production of IF may havelow normal or subnormal serum B12 concentra-tions because of failure to liberate B12 bound tofood.

Vitamin B12 is absorbed primarily by an ac-tive transport process in the ileum and by sim-ple diffusion at high doses, although with lowerefficiency. The receptor-mediated intestinal ab-sorption of vitamin B12 and delivery of the vita-min to tissues requires a series of vitamin B12

carrier proteins (Fig. 28-3).These include R pro-teins (also referred to as haptocorrins), IF, andthe transcobalamins (TCI, II, III). The two vita-min B12-binding proteins present in the stom-ach are IF and R proteins (Neale, 1990).The Rproteins consist of a family of glycoproteinsthat can be fractionated on the basis of theirsialic acid content (sialic acid is a nine-carbonacidic sugar).They were designated R proteinson the basis of their more rapid migration rela-tive to IF during electrophoresis.

R proteins are synthesized primarily ingranulocytes but are also present in saliva,plasma, bile, tears, and milk. The affinity of Rproteins for vitamin B12 at pH 2.0 is about 50times greater than that of IF. At low pH, the Rprotein-B12 complex is resistant to proteolytic

degradation by pepsin.Thus,most vitamin B12 isbound to R proteins during transit through thestomach.

Intrinsic factor (IF) is a 60,000-d glyco-protein that is synthesized and secreted by gas-tric parietal cells in humans. It is highly specificfor vitamin B12 and does not bind other non-functional cobalamin analogues ingested infood or excreted in the bile. IF secretion is stim-ulated by several endogenous agents, includinginsulin, gastrin, histamine, and acetylcholine.The average concentration of IF in gastric fluidis 1 µg/mL, an amount approximately 50 timesin excess of that estimated to be needed forphysiological purposes.

When the gastric contents reach the duode-num and the pH of the chyme is raised to neu-trality, the haptocorrins (R proteins) becomesusceptible to degradation by pancreatic prote-olytic enzymes (e.g., trypsin), causing vitaminB12 to be released (Fig. 28-3). The free vitaminB12 is then immediately bound to IF. In the pres-ence of pancreatic disease, the lack of prote-olytic enzymes to degrade the B12-R proteincomplex can result in malabsorption of vitaminB12. In the duodenum, gastric contents aremixed with bile, which contains approximately0.5 nmol/L cobalamins secreted by the liver. In-trinsic factor is thus available in duodenal con-tents to bind not only recently ingested vitaminB12 but also any physiologically active cobal-amin excreted in bile. In this way, about 90% ofthe physiologically active, nontoxic metabolitesof vitamin B12 in bile are reabsorbed.Thus, theenterohepatic recirculation provides for effi-cient conservation of vitamin B12 and, togetherwith the fact that large amounts are stored inthe liver, accounts for the fact that vitamin B12

deficiency does not occur rapidly even in theprolonged absence of vitamin B12 intake. Ane-mia in most cases takes 5 to 7 years to developafter the onset of reduced vitamin B12 intake orabsorption. An exception to this generalizationoccurs with individuals who are infested withthe fish tapeworm Diphyllobothrium latum,which absorbs and uses large amounts of vita-min B12. In tapeworm infestation, anemia candevelop in less than 3 years.

IF has two binding sites, one for vitamin B12

and one for the ileal receptor cubulin,which rec-ognizes the IF-vitamin B12 complex. Cubulin issynthesized by epithelial cells lining the terminalileum and other absorptive tissues. It is co-expressed with megalin,a multiligand transmem-brane endocytic receptor. The vitamin B12-IF


complex (ligand) binds to cubulin (receptor),and the ligand-receptor complex is internalizedvia clathrin-coated pits and coated vesicles.Thevitamin B12-IF ligand is dissociated in the acidicvesicles, and cubulin and megalin return to themembrane for reuse.After dissociation from thecubulin-megalin complex,the vitamin B12-IF com-plex is transferred to lysosomes, where IF is de-graded and vitamin B12 is released. After it isdissociated from the IF, vitamin B12 is transferredto TCII for transport to tissues.

TCII is a 50,000-d β-globulin synthesized inthe ileum and the liver. It is the transport pro-tein for recently absorbed vitamin B12. About20% of the plasma vitamin B12 is bound to TCII;the remainder of the vitamin B12 in plasma istransported by TCI and TCIII, primarily in theform of 5-methylcobalamin. Normal serum con-tains 0.7–1.5 nmol/L of TCII, which is capableof binding 600–1300 ng of vitamin B12; how-ever, only about 15% of its binding capacity isused at any one time. Unlike TCI and TCIII, the

other serum proteins that transport vitaminB12,TCII is not a glycoprotein.The specific func-tions of TCI and TCIII have not been deter-mined; however, they are believed to transportpotentially toxic vitamin B12 analogues to theliver for excretion in the bile (Fig. 28-4).

The absence of TCI and TCIII is clinically be-nign; however, the prolonged absence of TCIIresults in the typical hematological, neurologi-cal, and metabolic symptoms that characterizevitamin B12 deficiency. Infants who have a gene-tic defect resulting in a nonfunctional form ofTCII show evidence of vitamin B12 deficiencywithin a few months following birth.

The TCII-vitamin B12 complex is recognizedby specific high-affinity plasma membrane re-ceptors present on the cells that utilize vita-min B12. After internalization, the complex istransferred to lysosomes, where TCII is rapidlydegraded, and the hydroxycobalamin is re-leased into the cytosol and methylated. Thehydroxycobalamin that enters the mitochondria


Figure 28-3. The absorption of vitamin B12 in humans.The sequence of transitions be-tween protein-bound dietary vitamin B12 and circulating B12 bound to TCII. R, R pro-teins, haptocorrins; IF, intrinsic factor; TCII, transcobalamin II.

is converted to the adenosyl form by the en-zyme adenosyl transferase. Vitamin B12 presentin plasma in excess of the binding capacity ofthe transport proteins is excreted in the urine.Exogenous radiolabeled vitamin B12 can beused to determine glomerular filtration rate inhumans.

In the early stages of reduced vitamin B12

absorption, the amount of vitamin B12 boundto TCII decreases rapidly (10–21 days) withoutany decrease in the total serum B12 level.Thetotal vitamin B12 concentration begins to de-cline only when the saturation of TCII falls be-low 5%. Reduced saturation of TCII is thus oneof the earliest indicators of reduced intake ofvitamin B12 and is detectable in advance ofclinical disease. This makes holo-TCII parti-cularly valuable for use as a screening test insusceptible populations. Elevated serum andurine concentrations of homocysteine andMMA occur somewhat later and representearly evidence of cellular dysfunction that oc-curs when body (liver) stores have beengreatly depleted.

The total body pool of vitamin B12 in humansranges from 2 to 5 mg, most of which is storedin the liver. Daily losses of vitamin B12 are usu-

ally 1–3 µg or approximately 0.1% of totalbody stores. These losses occur primarilythrough excretion of vitamin B12 in bile andthrough sloughing of gastrointestinal epithe-lial cells.Although a dietary intake of 1 µg/dayis probably sufficient to meet the needs ofmost adults, the recommended dietary al-lowance is 2 µg/day to allow for individualvariability and to ensure maintenance of bodystores (Herbert, 1987). The average Americandiet contains 5–30 µg/day of vitamin B12. Reg-ularly scheduled vitamin B12 injections are gen-erally preferred by clinicians since measurementof IF and TCII are expensive and are not readilyavailable.

The total serum B12 concentration reflectsbody stores of B12 only when liver concentrationsfall below 0.6 µg/g, wet weight (normal is0.6–1.5 µg/g, wet weight). Moderately low serumvitamin B12 concentration (110–148 pmol/L) maynot be specific for vitamin B12 deficiency; how-ever, serum vitamin B12 concentrations below74 pmol/L are almost always associated with clin-ical manifestations of B12 deficiency,except in ve-gans or strict vegetarians.

Vitamin B12 is required by only two enzymesin human metabolism: methionine synthetaseand L-methylmalonyl-CoA mutase. Methioninesynthetase has an absolute requirement formethylcobalamin and catalyzes the conversionof homocysteine to methionine (Fig. 28-5). 5-Methyltetrahydrofolate is converted to tetrahy-drofolate (THF) in this reaction. This vitaminB12-catalyzed reaction is the only means bywhich THF can be regenerated from 5-methyltetrahydrofolate in humans.Therefore, invitamin B12 deficiency, folic acid can become“trapped” in the 5-methyltetrahydrofolate form,and THF is then unavailable for conversion toother coenzyme forms required for purine,pyrimidine, and amino acid synthesis (Fig. 28-6).All folate-dependent reactions are impaired invitamin B12 deficiency, resulting in indistin-guishable hematological abnormalities in bothfolate and vitamin B12 deficiencies.

The catabolism of certain amino acids (e.g.,valine, isoleucine, methionine) and odd-chainfatty acids (17:0) produces propionyl-CoA.Propionyl-CoA enters the TCA (citric acid) cyclefollowing conversion to succinyl-CoA, as shownin Fig.28-7.Propionyl-CoA is first carboxylated toproduce D-methylmalonyl-CoA, which in turn isthen racemized to L-methylmalonyl-CoA. In an in-tramolecular rearrangement reaction catalyzedby L-methylmalonyl-CoA mutase, a vitamin B12-


Figure 28-4. The enterohepatic circulation of vi-tamin B12. TCII (transcobalamin II) is the transportprotein that carries newly absorbed dietary vita-min B12 from the intestine to tissues. Approxi-mately 20% of circulating vitamin B12 istransported by TCII; the remainder of vitamin B12

is bound to TCI.THF, tetrahydrofolate.

requiring enzyme, methylmalonyl-CoA, is con-verted to succinyl-CoA. In vitamin B12 deficiency,cellular levels of both propionyl-CoA andmethylmalonyl-CoA increase. Propionyl-CoA canthen substitute for acetyl-CoA in fatty acid syn-thesis, leading to the production of fatty acidscomprised of odd numbers of carbon atoms.Similarly, methylmalonyl-CoA can substitute formalonyl-CoA in fatty acid synthesis, resulting inthe synthesis of branched-chain fatty acids.

Myelination is essential to the normal func-tion of sensory neurons. Because the synthesisof normal myelin is dependent on the availabil-ity of specific fatty acids, the inclusion of abnor-mal fatty acids (e.g., odd-chain, branched-chainfatty acids) in myelin may alter neural functionor cause premature demyelination.This hypoth-esis has been put forth to explain the neurolog-ical impairment observed in vitamin B12

deficiency (Shevell and Rosenblatt, 1992).Methionine, through subsequent conversion

to S-adenosylmethionine (see Fig. 28-4), isalso required for the synthesis of choline and

choline-containing phospholipids and for themethylation of myelin basic protein, a majorprotein component of myelin. Although themechanisms responsible for the developmentof neurological impairment resulting from vita-min B12 deficiency have not been precisely de-fined, defective methionine metabolism mayalso contribute to neurological complicationsin vitamin B12-deficient patients.

THERAPY

In the setting of established B12 deficiency,treatment is usually by parenteral injection,with 1000 µg given daily for the first 5 days.Thereafter, cyanocobalamin (1000 µg) is givenmonthly by intramuscular or deep subcuta-neous injection.While orally administered drugis not recommended in most patients who havea B12 deficiency due to malnutrition or achlorhy-dria (failure to release food-bound B12), 1000 µgoral vitamin B12 can be administered daily. A


Figure 28-5. The reaction catalyzed by methionine synthase, a vitamin B12-requiringenzyme. In this reaction, homocystine is converted to methionine, with the simultane-ous production of tetrahydrofolate (THF) from 5-methyltetrahydrofolate.Methionine canthen be converted to S-adenosylmethionine (SAM), the universal methyl-group donor.

transnasal vitamin B12 preparation is available(Nascobal) that can be used by the patient in adose of 500 µg/week, thereby obviating theneed for injection.The transnasal drug is moreexpensive but avoids indirect costs such as nee-dles, syringes, patient and nursing time, travel,and office expenses that are part of injectiontherapy and greatly dwarf the cost of the in-jected drug. Although most patients could betaught self-injection, this form of therapy re-mains little used.

For prevention of disease in the elderly, thepregnant, or other susceptible groups, nationalfortification of food with vitamin B12 appearssensible and inexpensive but at present is notused and, in the absence of population screen-ing, is unlikely to be mandated by governmentaledict. In general terms, the hematological mani-festations of vitamin B12 deficiency are rapidlyand fully correctable, although deficiencies ofother micronutrients such as iron, folic acid,pyridoxine, copper, or vitamin C may be un-masked in the process and may limit the bonemarrow’s response until they are also cor-rected.

Therapy of patients with megaloblastic ane-mia should always start with vitamin B12 re-placement and never with folic acid; giving

folate before adequate vitamin B12 replacementhas occurred is dangerous, not only because itmay mask vitamin B12 deficiency,but also,moreimportantly, because it may promote utilizationof vitamin B12 by bone marrow, leading to criti-cal deterioration of vital neurological func-tions. The likelihood that the neurologicalfunctions will recover, even slowly, depends onthe duration of the vitamin B12 deficiency be-fore therapy: If less than 1 year, the prognosis isgood; if over 2 years, it is poor; and if only 1–2years, it may be partial or incomplete. Whilethe treatment may appear simple, for many rea-sons compliance is often poor, needed therapyis discontinued, and the patient ends up withinadequate vitamin B12 replacement and dete-riorating disease, an unfortunate outcome.

QUESTIONS

1. What metabolites in serum and urinecan be used to differentiate between afolate deficiency and a vitamin B12 defi-ciency?

2. How does a vitamin B12 deficiency causea decrease in the hematocrit and hemo-globin concentration?


Figure 28-6. The “folate trap.”

3. How can pancreatic insufficiency con-tribute to the development of vitamin B12

deficiency despite the presence of ade-quate intake of vitamin B12 and normal se-cretion of intrinsic factor?

4. How can a deficiency of intrinsic factordevelop?

5. What are the metabolic and clinical con-sequences of using folic acid to treatmacrocytic anemia caused by a vitaminB12 deficiency?

6. What are the most common causes of vi-tamin B12 deficiency in elderly nursinghome residents?


Figure 28-7. The metabolism of branched-chain amino acids and odd-chain fatty acids viapropionyl-CoA. Propionyl-CoA is converted to D-methylmalonyl-CoA by propionyl-CoAcarboxylase. D,L-Methylmalonyl-CoA racemase catalyzes the conversion of D-methylmalonyl-CoA to L-methylmalonyl-CoA. L-methyl malonyl-CoA mutase, an adenosylcobalamin-requiring enzyme, converts L-methylmalonyl-CoA to succinyl-CoA.TCA cycle is citric acidcycle or Kreb’s cycle.

BIBLIOGRAPHY

Avcu N,Avcu F, Beyan C, et al.:The relationship be-tween gastric-oral Helocobacter pylori and oralhygeine in patients with vitamin B12-defciencyanemia. Oral Surg Oral Med Oral Pathol OralRadiol Endod 92:166–169, 2001.

Battersby AR: How nature builds the pigments oflife: the conquest of vitamin B12. Science 264:1551–1557,1994.

Carmel R: Prevalence of undiagnosed perniciousanemia in the elderly. Arch Int Med 156:1097–1100, 1996.

Carranza E:The neurological manifestations of vi-tamin B12 deficiency, in Herbert V (ed): VitaminB12 Deficiency. Royal Society of Medicine,Round Table Series 66, RSM Press, London, pp.21–26, 2002.

Dharmarajan TS, Pais W, Norkus EP: Does anemiamatter? Anemia, morbidity, and mortality inolder adults: Need for greater recognition. Geri-atrics 60:22–27, 2005.

Dharmarajan TS, Ugalino JT, Kanagala M, et al.:Vita-min B12 status in hospitalized elderly fromnursing homes and the community. J Am MedDir Assoc 1:21–24, 2000.

Herbert V: Recommended dietary intake (RDI) ofvitamin B12 in humans. Am J Clin Nutr 45:671–678, 1987.

Herzlich B, Herbert V: Depletion of serum holo-transcobalamin II. An early sign of negative

vitamin B12 balance. Lab Invest 58:332–337,1988.

Kaptan K, Beyan C, Ural AU, et al: Helicobacterpylori—is it a novel causative agent in vitaminB12 deficiency? Arch Int Med 160:1349–1353,2000.

Lindenbaum J, Healton EB, Savage DG, et al.:Neuropsychiatric disorder caused by cobal-amin deficiency in the absence of anemia ormacrocytosis. N Engl J Med 318:1720–1728,1988.

Metz J: Cobalamin deficiency and the pathogene-sis of nervous system disease. Annu Rev Nutr12:59–79, 1992.

Neale G: B12 binding proteins.Gut 31:59–63, 1990.Nexo E, Christensen A-L, Hvas A-M, et al.: Quantifi-

cation of holo-transcobalamin, a marker of vita-min B12 deficiency. Clin Chem 48:561–562,2002.

Roman GC:An epidemic in Cuba of optic neuropa-thy, sensorineural deafness, peripheral sensoryneuropathy and dorsolateral myeloneuropathy.J Neurol Sci 127:11–28, 1994.

Shevell MI, Rosenblatt DS:The neurology of cobal-amin. Can J Neurol Sci 19:472–486, 1992.

Stabler SP, Allen RH: Vitamin B12 deficiency as aworldwide problem. Annu Rev Nutr 24:299–326, 2004.

Stopeck A: Links between Helicobacter pylori in-fection, cobalamin deficiency, and perniciousanemia. Arch Int Med 160:1229–1230, 2000.


CASE REPORTS

Case 1

A boy aged 3 years and 8 months was admitted toa local hospital in Thailand with the followingsymptoms: high fever, difficulty breathing, and di-arrhea with approximately 10 episodes per daylasting 1 week.The child showed eye symptoms(see Fig. 29-1), including Bitot’s spot and cornealxerosis, that were consistent with the early stagesof xerophthalmia, the clinical syndrome associ-ated with vitamin A deficiency. (The clinically de-fined stages of xerophthalmia are provided inTable 29-1.) Night blindness was also identifiedthrough interview of the parents using the localterm for night blindness. (For cultures in whichvitamin A deficiency has been common, there isoften a specific term in the local language fornight blindness.) Concomitant with vitamin A de-ficiency, the child was diagnosed as experiencingprotein-energy malnutrition type 2 (moderate),pneumonia, and chronic diarrhea.The boy spent14 days in the hospital and was treated with vita-min A capsules (200,000 IU/day) on days 1, 2, 3,and 14 as well as with standard antibiotics for theinfections. Following treatment with vitamin A,the child’s eye symptoms disappeared, and his vi-sion returned to normal.He continued to receivevitamin A capsules providing 200,000 IU every3–6 months throughout his childhood.

Case 2

A 3-year-old boy was admitted to a hospitalwith a high temperature, vomiting, and diar-

29

Vitamin A Deficiency in Children

NUTTAPORN WONGSIRIROJ, EMORN WASANTWISUT,and WILLIAM S. BLANER

313

rhea that had persisted for over 2 weeks. Onexamination of his eyes, he showed keratoma-lacia (see Table 29-1) in both eyes. In addition,he refused to eat and to drink milk. One weekbefore coming to the hospital, the boy had redeyes and eye pain and cried frequently. His par-ents indicated that, at the onset of symptoms,the boy had experienced night blindness andhad become inactive. He subsequently devel-oped a high fever and refused his normal foodand milk. On evaluation, the child was diag-nosed as experiencing protein-energy malnu-trition type 3 (severe), keratomalacia, diarrhea,and sepsis. The boy was treated with an intra-muscular injection of water-miscible vitamin A(200,000 IU) and standard antibiotics. He con-tinued to receive vitamin A capsules (200,000IU) every 3–6 months throughout his child-hood. Owing to this bout with vitamin A defi-ciency, the boy developed a corneal scar inboth eyes (see Fig. 29-2) that will remainthroughout his life.

DIAGNOSIS

Vitamin A deficiency remains a major publichealth problem in the developing world. Datacompiled in 2002 by the World Health Organiza-tion (WHO) suggest that worldwide as many as21% of children in the age range 0.5 to 5 yearssuffer from vitamin A deficiency. Extrapolationsfrom these data indicate that as many as 140 mil-lion children under the age of 5 years and morethan 7 million pregnant women experience vita-min A deficiency every year (West, 2002).

Vitamin A deficiency in preschool children isnow rare in Thailand due to public health pro-grams that have been implemented since theearly 1970s. In rare instances, it still can be foundin remote areas of northern Thailand (Was-antwisut, 2002). From 1995 to 1999, there were39 cases of malnutrition with vitamin A defi-ciency reported. (It should be noted that vitaminA deficiency is usually associated with other nu-tritional deficiencies, especially protein-energymalnutrition.) Two of these cases are presentedin this chapter.

The eradication of vitamin A deficiency inThailand did not arise from a government pro-gram focused specifically on vitamin A defi-ciency but rather from a comprehensive nationalprogram to promote primary health care in localcommunities. Vitamin A supplements were ad-ministered only to children who showed eyesymptoms but not to all children deemed to beat risk of vitamin A deficiency.The program was

aimed at improving the quality of health careavailable in communities, food availability andquality, and community awareness of the symp-toms of malnutrition and associated illness. Edu-cational programs were put into place to trainand motivate community members to identifychildren experiencing malnutrition (especiallyprotein-energy malnutrition and micronutrientdeficiencies) and/or infection. In addition to amarked reduction in the observed rates of vita-min A deficiency in children, the program wasalso successful in bringing about the virtual elim-ination of protein-energy malnutrition and brokethe downwardly reciprocating relationship be-tween malnutrition and infection. Hence, thesuccess of the program in Thailand is creditedprimarily to improved public health measures,including educational programs and empower-ment of local communities to recognize and acteffectively to prevent and treat malnutrition andits associated infections in children


Figure 29-1. Two views of eye symptoms for a boy aged 3 years and 8 months with vita-min A deficiency (case study 1).Bitot’s spot and corneal xerosis are present in these views.Following treatment with vitamin A for 14 days, the eye symptoms returned to normal.

Table 29-1. Clinical Classification of XerophthalmiaXerophthalmia Classification Symptom

XN Night blindness Cannot see properly in dim light

X1A Conjunctival xerosis Dryish, sandylike change in the conjunctiva

X1B Bitot’s spot Foamy and cheeselike patches on the whites of the eyes

X2 Corneal xerosis Drying of the cornea, scaly appearance

X3A Corneal ulceration/keratomalacia Formation of holes on the cornea (involving less than a third of corneal surface)

X3B Corneal ulceration keratomalacia Cornea becomes cloudy and soft (involving a third or more of the corneal area)

XS Corneal scar Scar on the cornea

XF Xerophthalmia fundus Structural damage to the rods in retina, followed by degenerationof the both rods and cones

From Sommer and West (1996).

Table 29-2. Classification of Vitamin A Status bySerum Retinol LevelsSerum Retinol Levelµg/dL µM (µmole/L)* Vitamin A Status

≥30 ≥1.05 Normal

20–30 0.7–1.05 At risk

10–20 0.35–0.69 Moderate

<10 <0.35 SevereFrom sommer and West (1966).

*1 µM equals 28.6 µg/dL.

Serum retinol concentrations are used as ameasured index for assessing vitamin A status.Routinely, serum retinol levels in the range of0.35–0.69 µM (10–20 µg/dL) define moderate vi-tamin A deficiency, whereas levels below0.35 µM (<10 µg/dL) indicate severe vitamin Adeficiency (see Table 29-2). The clinical symp-toms of xerophthalmia, the name of the syn-drome associated with vitamin A deficiency, aremost evident in the eye and progress initiallyfrom night blindness (stage XN) to the develop-ment of Bitot’s spot (stage X1B), corneal ulcera-tion and keratomalacia (stage X3), and formationof a corneal scar (stage XS) and ultimately resultin xerophthalmia fundus (stage XF).The full clin-ical classifications of the stages of xerophthalmiaare listed in Table 29-1. The early stages of xe-rophthalmia are readily reversible on administra-tion of vitamin A to the patient; however, thelater stages commencing with the formation of acorneal scar are irreversible.

Subclinical vitamin A deficiency results in anincreased risk of morbidity and mortality arisingprimarily from impaired immunity (Sommer andWest, 1996). Serum levels of retinol in the rangeof 0.70–1.05 µM (20–30 µg retinol/dL serum)should be taken to indicate an at-risk populationfor subclinical vitamin A deficiency.


Vitamin A is an essential fat-soluble compoundacquired from the diet.The parent form of vita-min A is all-trans-retinol.Vitamin A is needed tomaintain normal vision, normal reproduction(including spermatogenesis, conception, andplacenta formation), and normal cell differentia-tion (including bone remodeling, maintenanceof differentiated epithelial linings and skin, em-

bryogenesis, and immune system function) (Gu-das et al., 1994). Because vitamin A is importantto so many essential processes in the body, vita-min A deficiency results in severe consequencesfor the health of human beings.The body is ableto store substantial quantities of vitamin A, pri-marily in liver, and this stored vitamin A is usedto maintain (defend) blood vitamin A levels.Adults normally acquire enough vitamin A overtheir lifetime to have stores that provide protec-tion from developing vitamin A deficiency intimes of dietary insufficiency. Since tissue vita-min A stores are relatively low at birth, youngchildren, especially those born to mothers withpoor vitamin A reserves, are vulnerable to devel-oping vitamin A deficiency.

When vitamin A stores are adequate, the liversecretes retinol bound to retinol-binding pro-tein (RBP) into the circulation to provide tis-sues with a constant supply of vitamin A. In thecirculation, the retinol-RBP complex is foundbound to another circulating protein of hepaticorigin, transthyretin (TTR). TTR also binds thy-roid hormone and consequently plays a role inthe transport of both vitamin A and thyroid hor-mone. The molecular size of the retinol-RBPcomplex is quite small, and the formation of the

Vitamin A Deficiency in Children 315

Figure 29-2. Residual corneal scarring following administration of vitamin A therapy toa 3-year-old who had experienced vitamin A deficiency (case study 2).

retinol-RBP-TTR complex serves to prevent re-nal filtration of retinol-RBP. When hepaticstores of vitamin A become depleted, retinol-RBP is no longer secreted from the liver, andconsequently blood levels of retinol decline.The retinol bound to RBP is measured diagnos-tically (see Diagnosis section) to assess the vita-min A status of a patient. Low blood retinollevels indicate insufficient vitamin A stores.However, normal blood retinol levels are not agood indicator for assessing liver stores sincethe liver secretes retinol-RBP until the storesare completely exhausted.

Chemistry, Biochemistry andNutrition of Vitamin A

The term retinoid refers to any naturally occur-ring or synthetic compound bearing a struc-

tural resemblance to all-trans-retinol with orwithout the biological activity of vitamin A(O’Byrne and Blaner, 2004). The structures ofsome important vitamin A species as well as themost active provitamin A carotenoid, all-trans-β-carotene, are shown in Figure 29-3.

Structurally, vitamin A and many syntheticretinoids consist of a β-ionone ring, a polyunsat-urated polyene chain, and a polar end group.The polar end group can exist in several oxida-tion states, as retinol, retinal, or retinoic acid.Retinol and retinyl esters are the most abundantvitamin A forms found in the body (Blaner andOlson, 1994). Retinol can be esterified withlong-chain fatty acids (mainly palmitate, oleate,and stearate) to form retinyl esters, which arethe body’s storage form of vitamin A. Retinolalso can undergo oxidation to retinal,which canbe oxidized further to retinoic acid. The active


(I) all-trans β-Carotene

(II) all-trans Retinol (IV) 11-cis Retinal(III) Retinyl Ester (R = Acyl Chain)

OH

O

O R

OH

O

O

(VII) 13-cis Retinoic Acid(VI) 9-cis Retinoic Acid(V) all-trans Retinoic Acid

O OHO OH

Figure 29-3. Chemical structures of important vitamin A species and the provitamin Acarotenoid β-carotene. All-trans-β-carotene (I) is the most important provitamin Acarotenoid,which can be converted to all-trans-retinal and then all-trans-retinol (II),whichby definition is vitamin A.All-trans-retinol can be esterified with long-chain fatty acids toform retinyl ester (III), the storage form of vitamin A in the body.The active form of vitaminA in vision is 11-cis-retinal (IV).The transcriptionally active forms of vitamin A are all-trans-retinoic acid (V) and 9-cis-retinoic acid (VI). 13-cis-Retinoic acid (VII) has poor transcrip-tional regulatory activity but is used clinically as isotretinoin to treat skin diseases.

form of vitamin A in vision is 11-cis-retinal (seemore details in the Vision section) (Wald, 1968).Outside of vision, vitamin A acts by regulatinggene transcription (Chambon, 1994; Mangels-dorf et al., 1994).The preponderance of the bio-logical activity of vitamin A arises from effectson gene expression. The transcriptional regula-tory actions of vitamin A are mediated by all-trans-retinoic acid and 9-cis-retinoic acid (seesection on Transcriptional Regulation). A largenumber of synthetic retinoids have been gener-ated and used in clinical trials as potential treat-ments for many different diseases, especiallycancers and skin diseases.

All vitamin A present in the body must be ac-quired from the diet (O’Byrne and Blaner,2004;Vogel et al., 1999).Two forms of vitamin Aare available in the diet, preformed vitamin Aand provitamin A carotenoids. The preformedvitamin A consists primarily of retinol andretinyl ester that have been consumed in ani-mal food products. Provitamin A carotenoidslike β-carotene are consumed in dark greenand yellow fruits and vegetables (mangoes, pa-paya, carrots, dark green leafy vegetables, andred palm oil are some rich dietary sources ofprovitamin A carotenoids). Dietary provitaminA carotenoids are converted in the intestineand other tissues to vitamin A by the actions ofthe enzyme β-carotene-15,15'-monooxygenase(also called carotene cleavage enzyme). Oncedietary provitamin A carotenoid has been con-verted in the body to vitamin A, it is metaboli-cally and functionally indistinguishable fromvitamin A originating in the diet as preformedvitamin A.

The vitamin A content of foods is often givenin terms of the international unit (IU). One IUof vitamin A is defined as 0.3 µg of all-trans-retinol. The term retinol equivalent (RE) isused to convert all sources of vitamin A andcarotenoids in the diet to a single unit. One REis by definition 1 µg of all-trans retinol, 12 µg ofβ-carotene, or 24 µg of other (mixed) provita-min A carotenoids. The recommended dietaryallowance for vitamin A ranges from 375 µgRE/day for infants to 1,000 RE/day for adults.

Excessive intake of vitamin A (hyper-vitaminosis A), like too little intake, can result inadverse health consequences. Approximately60%–80% of vitamin A is stored in the liver inhepatic stellate cells (also called Ito cells andfat-storing cells). Retinyl esters are the mainstorage form of vitamin A in the liver and arefound in lipid droplets present in the hepatic

stellate cells (Fig 29-4). Both the size and num-ber of these lipid droplets increase in responseto vitamin A intake. Initially, hypervitaminosis Apromotes cheilitis (dryness and cracking of thelips) as well as dryness of nasal mucosa, eyes,and skin; hair loss; and nail fragility. Later clini-cal indicators of vitamin A toxicity include bonepain, nausea, vomiting, and hepatomegaly. Somecases of excessive vitamin A intake have re-sulted in irreversible liver damage.

Biochemical Actions of Vitamin A

Vision

The form of vitamin A required in vision is 11-cis-retinal. In the retina, the initiating event ofvision occurs when a photon of light excites amolecule of 11-cis-retinal that is covalentlybound to the protein opsin (11-cis-retinalbound to opsin is known as rhodopsin or thevisual pigment) (Saari, 1999). As shown in Fig-ure 29-5, when rhodopsin is struck by a photonof light, the 11-cis-retinal undergoes photoiso-merization to all-trans-retinal, thus initiating asignal transduction cascade that results in trans-mission of the visual impulse to the brain.Thisaction of vitamin A involves a large number ofvery well characterized enzymatic and nonen-zymatic processes that arise ultimately as the re-sult of the photoisomerization of 11-cis-retinalto all-trans-retinal. These vitamin A-dependentprocesses are known as the visual cycle of vita-min A and are unique to the eye and vision(Saari, 1999).

Transcriptional Regulation

Outside of vision, the physiological actions ofvitamin A arise through the regulation of vita-min A-dependent gene expression (Chambon,1994; Mangelsdorf et al., 1994). More than 500different genes may be responsive to vitamin A(Balmer and Blomhoff, 2002). The genes re-ported to be responsive to vitamin A are quitediverse and include ones encoding transcrip-tion factors, peptide hormones, growth factors,growth factor receptors, protein kinases andphosphatases, and enzymes involved in inter-mediary metabolism and the regulation of thismetabolism.

Vitamin A acts to regulate gene expressionthrough interactions with its specific ligand-dependent transcription factors, which aresometimes referred to as nuclear receptors(Chambon, 1994; Chawla et al., 2001; Mangels-


dorf et al., 1994).The vitamin A-dependent tran-scription factors consist of three retinoic acidreceptors (RARs) and three retinoid X recep-tors (RXRs). The RARs and RXRs are membersof the large steroid/thyroid/retinoid superfam-ily of ligand-dependent transcription factorsthat regulate hormone- and nutrient-dependentgene expression.Thus, vitamin A functions bio-chemically within the body in a manner that isvery similar to steroid hormones and thyroidhormone.

The form of vitamin A that is needed for reg-ulating vitamin A-dependent gene expression isretinoic acid.All-trans-retinoic acid can interactwith each of the three RARs (RARα, RARβ, andRARγ) and each of the three RXRs (RXRα,RXRβ, and RXRγ). 9-cis-Retinoic acid binds andeffectively transactivates only the RXRs. TheRARs and the RXRs recognize well-defined cis-acting response elements, termed retinoic acidresponse elements (RAREs) and retinoid X re-sponse elements (RXREs) that are present inthe promoter regions of responsive genes.TheRARs and RXRs bind to RAREs and/or RXREs asdimers, either homodimers or heterodimers.

When bound to a response element, the vita-min A-dependent nuclear receptors are able tointeract with coactivators and corepressorspresent in the nucleus either to stimulate or torepress transcriptional activity. Interactionswith coactivators or corepressors will result re-spectively in transcriptional activation or re-pression. Thus, retinoic acid availability as aligand for binding to RARs and/or RXRs and theability of these receptors to bind alternativelyto either coactivators or corepressors providesa mechanism through which genes can be ei-ther activated or repressed by retinoic acid. It isthis gene regulatory mechanism that accountsfor the great majority of vitamin A action in thebody.

Vitamin A has a very important role in regu-lating the actions of other hormone- andnutrient-dependent genes because the RXRscan interact with other nuclear receptors toform heterodimers that are able to bind andregulate hormone and nutrient responsiveness.The RXRs are able to heterodimerize with thevitamin D receptor, the thyroid hormone re-ceptors, the peroxisome proliferator activated


Figure 29-4. Stellate cell lipid droplets present in a biopsy obtained from the liver of apatient experiencing vitamin A toxicity. Image obtained from electron microscopy of abiopsied human liver showing the characteristic lipid droplets (LD) found in hepaticstellate cells (SC).These lipid droplets are highly enriched in vitamin A, and the size andnumber of lipid droplets is influenced by dietary vitamin A intake and nutritional status.In this image of a human stellate cell, the nucleus (N) is compressed by the surroundinglipid droplets, and very little cell cytoplasm within the stellate cell can be seen in thisview.Adjoining the stellate cell are two hepatocytes (H).

Figure 29-5. The visual cycle of vitamin A is central to vision. In the retina, light stimulates the conversion of 11-cis-retinal, part of rhodopsin (Rho), to all-trans-retinal and activates rhodopsin (Rho*).This initiates the first step of the signal transduction cascade that results in the transmission of the visual signal tothe brain.The visual cycle involves biochemical and metabolic events in both the photoreceptors (rods) and the retinal pigment epithelium.

receptors, as well as a number of otherhormone-/nutrient-dependent transcription fac-tors (Chawla et al., 2001).Through this interac-tion, vitamin A and the RXRs can regulate abroad spectrum of hormonally or nutrient-responsive genes.These receptors and their lig-ands are illustrated in Figure 29-6.

The Biochemical Basis for the Clinical Manifestations of Vitamin A Deficiency

The initial effects of vitamin A deficiency onchild health are subclinical and manifested asan increased risk of morbidity and mortality


Receptor (R) Ligand (L)

Retinoid X Receptor (RXR)

Retinoic Acid Receptor (RAR)

Thyroid Hormone Receptor (TR)

Vitamin D Receptor (VDR)

Peroxisome ProliferatorActivated Receptor (PPAR)

9–cis–retinoic acid

all–trans–retinoic acid

Triiodothyronine (T3)

1,25–(OH)2 –Vitamin D

Fatty Acid Metabolite

Figure 29-6. Gene transcription is regulated by retinoic acid.All-trans-retinoic acid and9-cis-retinoic acid are ligands for retinoic acid receptors (RARs) and retinoid X receptors(RXRs), respectively.The RXRs can form heterodimers with RARs and with the thyroidhormone receptors (TRs), the vitamin D receptor (VDR), and the peroxisomeproliferator-activated receptors (PPARs) and a number of other hormone- and nutrient-responsive transcription factors to moderate gene transcription.Because of the ability ofRXR to form heterodimers with other nuclear receptors, vitamin A has abroad effect onmany hormonally and nutrient-responsive genes.

arising primarily from gastrointestinal and res-piratory infections. This results from a dimin-ished immune response and disruption of theintestinal barrier.Vitamin A is needed to ensurenormal cell proliferation and differentiation ofboth white blood cells and the intestinal epithe-lium.Thus, retinoic acid and the transcriptionalregulatory actions of the RARs and RXRs areneeded to ensure optimal immune cell actionand maintenance of an intact intestinal barrier.

The first clinical sign of vitamin A deficiencyis night blindness.This arises because too little11-cis-retinal is produced in the retina to sup-port normal night vision and the visual cycle.As the effects of xerophthalmia progressfrom night blindness through Bitot’s spot andcorneal xerosis to keratomalacia, the clinical ef-fects of vitamin A deficiency result from impair-ments of vitamin A-dependent gene expression.Corneal xerosis is thought to arise from inap-propriate expression of collagenase genes thatare normally repressed by retinoic acid. In theabsence of retinoic acid, these genes becomeactivated and collagenases are produced. Thisleads to digestion of connective tissue neededto maintain the structural integrity of the eye.Ultimately, if vitamin A deficiency is not recog-nized and treated it will result in death owingto impaired immunity and infection.

THERAPY

Patients presenting with any stage of xeroph-thalmia should immediately be given high-dosevitamin A supplements according to WHO treat-ment dosing guidelines (Table 29-3). This doseshould be repeated the next day and 1–4 weekslater (WHO/UNICEF/IVACG, 1997).

Because of the potential teratogenic effectsof high-dose vitamin A,caution must be taken inthe treatment of severe vitamin A deficiencyamong pregnant women as well as women ofreproductive age. Women of reproductive ageshould be treated with 200,000 IU only whenthey display active corneal xerophthalmia. For

milder eye signs (night blindness or Bitot’sspot), women should be given 5,000–10,000 IUper day but less than 25,000 IU per week for 4weeks or longer.

The preferred form and route of administra-tion is retinyl palmitate in oil given orally.The full dosing schedule should be given for allstages of xerophthalmia, aside for pregnantwomen and women of reproductive age. Incases of severe diarrhea or vomiting,which mayprevent ingestion and absorption, an intra-muscular injection of water-miscible vitamin A(at the same doses) may be administered.A re-cent field trial in Ethiopia (Biesalski et al., 1999)indicated that an aerosol of retinyl palmitate forinhalation resulted in well-absorbed vitamin Athrough the respiratory tract and may in the fu-ture become the preferred treatment option forpatients with severe diarrhea or vomiting.

QUESTIONS

1. A measure of an individual child’s serumretinol level may not be a good indicatorof the vitamin A status of the child.Why isthis?

2. What protocols should be followed inproviding vitamin A to a child who is sus-pected of being either subclinically orclinically vitamin A deficient?

3. Although vitamin A must be acquiredfrom the diet, why might it be thought ofas being hormonelike?

4. Clinical signs of vitamin A deficiency (xe-rophthalmia) are often diagnosed to-gether with infections such as pneumoniaand diarrhea.Why?

5. What would be the recommended treat-ment for pregnant women with nightblindness? Explain the basis for such a rec-ommendation.

6. For a preventive measure, high-dose vita-min A is administered every 4–6 months.What is the biochemical basis behind thisperiodic regime?

BIBLIOGRAPHY

Balmer JE, Blomhoff R.: Gene expression regula-tion by retinoic acid. J Lipid Res 43:1773–1808,2002.

Biesalski H, Reifen R, Furst P, et al.: Retinylpalmitate supplementation by inhalation of an


Table 29-3. Vitamin A Supplement According to WHO GuidelinesAge Treatment Dose (IU)

<6 months of age 50,000

6–12 months of age 100,000

>12 months of age 200,000

From WHO/UNICEF/IVAGG (1997).

aerosol improves vitamin A status of preschoolchildren in Gondar (Ethiopia). Br J Nutr 82:179–182, 1999.

Blaner WS, Olson JA: Retinol and retinoic acid me-tabolism, in Sporn MB,Roberts AB,Goodman DS(eds): The Retinoids: Biology, Chemistry, andMedicine. 2nd ed. Raven Press, New York, 1994,pp. 229–256.

Chambon P: The retinoid signaling pathway: mo-lecular and genetic analysis. Cell Biol. 5:115–125, 1994.

Chawla A, Repa JM, Evans RM, et al.: Nuclear recep-tors and lipid physiology: opening the X-files.Science 294:1866–1870, 2001.

Gudas LJ, Sporn MB, Roberts AB: Cellular biologyand biochemistry of the retinoids, in Sporn MB,Roberts AB, Goodman DS (eds): The Retinoids,Biology, Chemistry and Medicine. 2nd ed.Raven Press, New York, 1994, pp. 443–520.

Mangelsdorf DJ, Umesono K, Evans RM: Theretinoid receptors, in Sporn MB, Roberts AB,Goodman DS (eds): The Retinoids, Biology,Chemistry and Medicine. 2nd ed. Raven Press,New York, 1994, pp. 319–350.

O’Byrne SM, Blaner WS: Introduction to retinoids,in Packer L, Kraemer K, Obermueller U, Sies H(eds): Carotenoids and Retinoids. AOCS Press,Champaign, IL, 2005, pp. 1–22.

Saari JC: Retinoids in mammalian vision, in Nau H,Blaner WS (eds): The Handbook of Experimen-

tal Pharmacology, Retinoids: The Biochemicaland Molecular Basis of Vitamin A andRetinoid Action. Vol. 139. Springer-Verlag, Hei-delberg, 1999, pp. 563–588.

Sommer A, West K: Vitamin A Deficiency: HealthSurvival and Vision. Oxford University Press,New York, 1996.

Vogel S, Gamble MV, Blaner WS: Retinoid uptake,metabolism and transport, in Nau H, Blaner WS(eds): The Handbook of Experimental Phar-macology, Retinoids: The Biochemical andMolecular Basis of Vitamin A and RetinoidAction. Springer-Verlag, Heidelberg, 1999, pp.31–96.

Wald G: Molecular basis of visual excitation. Na-ture 219:800–807, 1968.

Wasantwisut E:A combined approach to vitamin Adeficiency in Thailand. Sight Life Newsl 3:63–67, 2002.

West KP Jr: Extent of vitamin A deficiency amongpreschool children and women of reproductiveage. J Nutr 132:2857S–2866S, 2002.

WHO: The World Health Report 2002: ReducingRisks, Promoting Healthy Life.World Health Or-ganization, Geneva, 2002.

WHO, UNICEF, IVACG Task Force: Vitamin A Sup-plements. A Guide to Their Use in the Treat-ment and Prevention of Vitamin A Deficiencyand Xerophthalmia. 2nd ed. World Health Or-ganization, Geneva, 1997.


CASE REPORT

The patient was a 3-year-old female who livedwith her family in a rural village in northernNigeria that was situated 40 miles distant fromthe nearest urban center. She was the fifth ofeight children. Her parents were subsistencefarmers who grew millet and maize that, to-gether with a small vegetable garden, providedmost of the family’s diet. Because of their highcost, dairy products were rarely consumed bythe parents or children.

The patient was brought by her mother tothe local community health clinic because thechild had a bacterial infection on her knee.Theinfection was successfully treated with a 10-daycourse of antibiotics.During that clinic visit, thephysician noted that the patient’s legs were se-verely bowed (genu varum) (Fig. 30-1) and ad-vised the mother that it would be beneficial forthe child to receive a more thorough examina-tion at the teaching hospital in the nearby city.She also noticed that the patient’s 6-year-oldbrother had a pronounced “K-leg” (rotation ofknees toward one another, genu valgum). How-ever, the youngest of the eight children, a 1.5-year-old girl who accompanied the mother andwho was still being breast-fed, exhibited no evi-dence of outstanding skeletal problems.

The physician at the rural clinic madearrangements to transport the child and hermother to the teaching hospital in the nearbycity. At the hospital, she was given a completeexamination by a physician in the Departmentof Family Medicine, and blood and urine sam-ples were obtained. The weight of the patient(14.8 kg) was above the 50th percentile for fe-

males in her age range according to WorldHealth Organization (WHO) standards (zscore = +0.66) (WHO, 1983). A z score is amethod used to compare an individual’s owndata to a reference population. The z score iscalculated as the difference between the indi-vidual’s value and the corresponding mean ofthe reference population at the same age andgender divided by the standard deviation forthe corresponding reference population. Thepatient’s standing height was 80 cm, which cor-responds to a z score of −2.1.The patient wasotherwise in good health.The mother reportedthat the child cried when she was required towalk for even a short distance, and that she car-ried her daughter a good part of the time.

Stereotypical signs of rickets were evident inthe patient, including bowed legs and the pres-ence of a “rachitic rosary”(palpable beadlike en-largement of the costochondral junctions orribs). X-rays of the long bones of the patient re-vealed widening of the growth plates and meta-physeal cupping and fraying.The metaphysis isa growth area of bone located between the dia-physis (the shaft of long bone) and the epiph-ysis, which is the section of bone beyond thegrowth plate. During growth, endochondral os-sification, the process by which cartilage istransformed into bone, occurs at the growthplate of the long bones.This results in lengthen-ing of bone.There was severe osteomalacia, asevidenced by increases in the volume, surfacearea, and thickness of the osteoid, and a re-duced calcification front.

The results of the laboratory analyses aresummarized in Table 30-1. The serum total cal-cium concentration was low (7.9 mg/dL; normal

30

Calcium-Deficiency Rickets

DOROTHY J.VANDERJAGT and ROBERT H. GLEW

323

is 8.8–10.8 mg/dL), as was the phosphorusconcentration (4.4 g/dL; normal is 4.5–5.5 g/dL).Serum albumin was 41 g/L (normal is 38–54 g/L).The intact parathyroid hormone (PTH) concen-tration (283 pmol/L;normal is 1.05–6.84 pmol/L)

and serum alkaline phosphatase activity (1500U/L;normal is 160–320 U/L) were both elevated.Urinary calcium excretion was below the nor-mal range (normal is 50–100 mg/24 h), but uri-nary phosphate excretion was increased(1.9 g/24 h;normal is 0.4–1.3 g/24 h).Analysis ofa serum sample that had been sent to a labora-tory in the United States revealed that the child’sserum 25-hydroxyvitamin D level was normalbut at the lower end of the reference range(46 nmol/L; normal is 35–150 nmol/L); how-ever, her serum 1,25-dihydroxyvitamin D levelwas above the upper limit of the referencerange for children (250 pmol/L; normal is42–169 pmol/L).

The laboratory also measured biochemicalmarkers of bone metabolism. The urinary ex-cretion of deoxypyrolidone, a marker of boneresorption, was elevated (128 nmol/mmol crea-tinine; normal is 31–110 nmol/mmol creati-nine). Bone-specific alkaline phosphatase, amarker of bone synthesis, was also elevated at400 U/L (normal is 70–139 U/L for girls be-tween 3 and 8 years of age) (Tsai et al., 1999).The concentration of another marker of boneresorption, NTx (N-teleopeptide), in the pa-tient’s serum was elevated (110 nmol BCE/L;normal is 20–68 nmol BCE/L [bone collagenequivalents]).

DIAGNOSIS

The clinical presentation and laboratory datafor this patient are consistent with the bone


Figure 30-1. A 3-year-old child in Nigeria with ev-idence of rickets, specifically genu varum.

Table 30-1. Summary of the Laboratory Analyses for the PatientPatient After Therapy

Parameter Before Therapy (3 months) Normal Range

Total calcium (mg/dL) 7.9 9.2 8.8–10.8

Phosphorus (mg/dL) 4.4 5.0 4.5–5.5

Albumin (g/L) 41 43 38–54

iPTH (pmol/L) 283 8 1.05–6.84

Alkaline phosphatase activity (U/L) 1500 300 160–320

Urinary calcium (mg/24 h) n.d. 45 50–100

Urinary phosphorus (mg/24 h) 2.0 0.8 0.4–1.3

25-Hydroxyvitamin D (nmol/L) 46 52 35–150

1.25-Dihydroxyvitamin D (pmol/L) 250 165 42–169

Urinary deoxypyrolidone (nmol/mmol creatinine) 128 60 31–110

Bone-specific alkaline phosphatase (U/L) 400 145 70–139

Serum NTx (BCE/L) 110 67 20–68

n.d., not detectable; PTH, intact parathyroid hormone; BCE, bone collagen equivalents; NTx, N-teleopeptide.

disorder known as rickets (Wharton andBishop, 2003). Nutritional rickets can be causedby a deficiency of calcium (calcipenic rickets )or phosphate (phosphopenic rickets). Calcium-deficiency results from an inadequate calciumintake or a deficiency of vitamin D, the mainregulator of calcium homeostasis.Vitamin D oc-curs as vitamin D3 (cholecalciferol), which isproduced endogenously in the skin, or as vita-min D2 (ergocalciferol),which is synthesized bythe irradiation of ergosterol from yeast andplants (Holick, 2004). When vitamin D is ex-pressed without a subscript, it is referring to acombination of the two forms. Other nonnutri-tional causes of rickets in children are related togenetic defects in the kidney enzyme 1α-hydroxylase, which converts 25-hydroxyvitaminD to 1,25-dihydroxyvitamin D (type 1 vitaminD-deficient rickets) or a mutation in the vitaminD receptor (type 2 vitamin D-resistant rickets).

The principal method of diagnosis of ricketsis radiographic (Jergas and Genant, 2002). Cup-ping and fraying of the metaphysis and longi-tudinal widening of the growth plates arehallmarks of rickets and are common to vitaminD-deficiency and calcium-deficiency rickets.Bone consists of two phases: an osteoid proteinphase composed primarily of collagen pro-duced by osteoblasts on which the mineralphase (calcium and phosphorus) is depositedas a crystal (hydroxyapatite). Rickets is charac-terized by an increase in demineralized osteoid.In adults, in whom the growth plate has closed,decreased mineralization of the osteoid is re-ferred to as osteomalacia. Proper mineralizationof the osteoid matrix in the growth plate re-quires adequate calcium and phosphate to al-low crystal formation. If inadequate calcium isprovided in the diet of a growing child, then theosteoid will not be sufficiently mineralized, re-sulting in soft bones that can bend due to theweight of the child.This is the basis of the bow-leg or knock-knee seen in rickets.

The site and type of bone deformity seen inrickets depend on the age of the child. In asmall infant, deformities of the forearms and an-terior bowing of the distal tibias are more com-mon. Clinical features such as craniotabes(areas of thinning and softening in the bones ofthe skull), hypotonia, and tetany are commonin vitamin D-deficiency rickets, which occursmore frequently in infants 1 year old or younger.These features may be absent in calcium-deficiency rickets, which usually presents afterthe age of 1 year or after the child has been

weaned from the breast and begins consumingfood that is low in calcium (Thacher, 2003).Anexaggeration of the normal bowing of the legs(genu varum) may be more common in a tod-dler, whereas an older child may present withK-leg (genu valgum) or windswept deformity(valgus deformity of one leg and a varus defor-mity of the other).

In rickets and adult osteomalacia, the ratio ofunmineralized to mineralized bone is increased.This is in contrast to osteoporosis, where theratio of unmineralized to mineralized bone isnormal. Osteoporosis is characterized by a de-crease in bone mass caused by an imbalance inbone formation relative to reabsorption but isnot associated with changes in the histologicalappearance of bone.

The biochemical features of calcium-deficiency and vitamin D deficiency are verysimilar. Both disorders result in a low-to-normalserum calcium concentration, an elevated PTHlevel, a decreased or normal phosphorus con-centration, and increased alkaline phos-phatase activity. The serum concentration of25-hydroxyvitamin D is normal or slightly de-creased in calcium-deficiency rickets but ismarkedly decreased in vitamin D deficiency. Onthe other hand, the serum concentration of1,25-dihydroxyvitamin D is greatly elevated incalcium-deficiency rickets but is normal oreven slightly decreased in vitamin D-deficiencyrickets.

The dietary history and laboratory findingsfor this patient pointed to calcium-deficiencyas the cause of her rickets: a low serum cal-cium within the normal range; secondaryhyperparathyroidism with an elevated serumconcentration of PTH; and a normal serumconcentration of 25-hydroxyvitamin with an el-evated serum concentration of 1,25-dihydroxy-vitamin D, the active hormone form of vitaminD.The patient’s calcium-deficiency was due tothe fact that she did not have access to milk be-cause of the expense and the lack of adequatestorage facilities for fresh dairy products in thehome.

Therapy for this patient consisted of a dailysupplement of elemental calcium (1000 mg) inthe form of calcium carbonate. Because of theelevated concentration of 1,25-dihydroxyvi-tamin D and the normal concentration of 25-hydroxyvitamin D,vitamin D supplements werenot prescribed for this child.The subject of thiscase report responded well to calcium supple-mentation.After 3 months of supplementation,

Calcium-Deficiency Rickets 325

there was radiographic evidence of healing, andthe serum calcium concentration rose to withinnormal limits. She no longer had evidence ofpain when walking.The mother was advised tocontinue giving the child calcium supplementsfor another 3 months and then bring the childback to the hospital 6 months later for an exam-ination. After 6 months of calcium therapy, thepatient’s serum 1,25-dihydroxyvitamin D de-creased to 165 pmol/L, which is in the normalrange (Table 30-1). The concentrations of thetwo markers of bone turnover were also mea-sured and found to be within normal limits:Theserum bone-specific alkaline phosphatase was145 U/L, and the serum NTx was 67 BCE/L.Theolder brother was also diagnosed with calcium-deficiency rickets, and he also received calciumsupplements.

Because the younger child was ready to beweaned from the breast, the mother was givennutritional counseling by a dietitian in the Di-etetics Department of the teaching hospital toprevent the younger child from also developingcalcium-deficiency rickets.The mother was alsoprovided information regarding how to in-crease calcium in the diet of the entire familyusing locally available foods such as dried fishand baobab leaf, which contain high amountsof calcium.


Calcium accounts for 2%–4% of the gross bodyweight in humans.Approximately 99% of the to-tal body calcium is in bones and teeth. Calcium,in the form of hydroxyapatite [Ca10(PO4)6

(OH)2], accounts for approximately 40% of thebone mass and, in combination with the pro-tein components of bone, contributes to bonestrength.Type 1 collagen is the most abundantextracellular protein present in bone. In addi-tion to its role as a structural component inbone, calcium serves as an intracellular signal-ing molecule and is required for the normalneural excitation and relaxation of muscles. In-sufficient ionized calcium in the extracellularfluid (ECF) results in tetany (continuous excita-tion of muscle due to hypocalcemia). Approxi-mately 1 g of calcium is contained in theplasma and ECF bathing the cells, and about6–8 g are found in the tissues, primarily in stor-age vesicles (e.g., sarcoplasmic reticulum).Serum calcium is maintained within a narrowconcentration range (8.8–10.8 mg/dL) and is in

equilibrium with the ECF calcium, which is onthe order of 5 mg/dL. The calcium concentra-tion of the ECF is less than that of plasma dueto the lower concentration of proteins in theECF. It is the calcium concentration in the ECFthat is regulated by the vitamin D/parathyroidendocrine system.

The main sources of dietary calcium are milkand dairy products. In the United States, it is es-timated that 73% of calcium is obtained frommilk products, 9% from fruit and vegetables, 5%from grains, and about 12% from all othersources combined. The optimal intake of di-etary calcium depends on age,gender, and phys-iological status and is summarized in Table 30-2.

Calcium is absorbed in the intestine by twodistinct mechanisms, an active process that isvitamin D dependent and another that is vita-min D independent.When the dietary intake ofcalcium is low, the intestinal uptake of calciumoccurs by an active transport process that is vi-tamin D dependent.Active transport is most ef-ficient in the duodenum and proximal jejunum,areas of the intestine that have a pH close to6.0 and where calbindin, a calcium transportprotein, is present. However, the amount of cal-cium absorbed in the ileum may be greater thanthat absorbed in the duodenum and jejunum


Table 30-2. Adequate Intakes* of Calcium andVitamin D at Different Ages

Calcium Vitamin D†

Age (mg/day) (µg/day)

0–6 month 210 5

7–12 month 270 5

1–3 year 500 5

4–8 year 800 5

9–18 year 1300 5

19–50 year 1000 5

51–70 year 1200 10

>70 year 1200 15

Pregnancy

≤ 18 year 1300 5

19–50 year 1000 5

Lactation

≤ 18 year 1300 5

19–50 year 1000 5From Food and Nutrition Board (2000).

*Adequate intake is believed to cover the needs of all individualsin the group; lack of data prevent the ability to specify withconfidence the percentage of individuals covered by this intake.

†In the absence of adequate exposure to sunlight; 1 µg vitaminD = 40 IU vitamin D.

due to the longer transit time of food in thisportion of the intestine. Calbindin, which is reg-ulated by 1,25-dihydroxyvitamin D, binds twomolecules of calcium on the surface of the in-testinal cell. The calbindin-calcium complex isinternalized via endocytosis into a lysosome, af-ter which calcium is released in the acidic envi-ronment of the lysosome and calbindin isrecycled to the cell surface.Calcium then leavesthe cell via the basolateral membrane of the en-terocyte and enters the blood.

When the luminal content is high, such aswhen a calcium supplement or food high in cal-cium is ingested, calcium is absorbed mainly bya nonsaturable paracellular process that is vita-min D independent. Although theoreticallyparacellular transport is bidirectional, the pre-dominant direction of calcium flux is from theintestinal lumen to the blood.The rate of trans-port depends on the calcium load and the tight-ness of the intracellular junctions in theenterocyte. Water may carry calcium throughthe intracellular junctions by solvent drag.Citrate, when present in the intestinal lumen si-multaneously with calcium, is thought to en-hance the passive transport of calcium.

Both the active and passive modes of cal-cium transport are increased during pregnancyand lactation. This is probably due to the in-crease in calbindin and serum PTH and 1,25-dihydroxyvitamin D concentrations that occurduring normal pregnancy. Intestinal calcium ab-sorption is also dependent on age, with a 0.2%per year decline in absorption efficiency start-ing in midlife.The fractional absorption of cal-cium depends on the form and dietary source.Absorption rates are 29% for the calcium incow’s milk, 35% for calcium citrate, 27% for cal-cium carbonate, and 25% for tricalcium phos-phate.Other factors that limit the bioavailabilityof calcium in the intestine are oxalates and phy-tates, which are found in high quantities in veg-etarian diets and which chelate calcium.

Vitamin D is a fat-soluble vitamin; its mainfunction is the maintenance of normal plasmalevels of calcium and phosphorus (Zittermann,2003). Vitamin D3 is synthesized in the skinfrom 7-dehydrocholesterol on exposure to sun-light (290–315 nm) (Fig. 30-2). The UV lightcauses a rearrangement of the 5,7 diene bondsin the B ring of 7-dehydrocholesterol, resultingin a break in the B ring to form previtamin D3.


Figure 30-2. Structures of vitamin D2 and vitamin D3 and their precursors.

Previtamin D3 is thermodynamically unstableand rearranges its double bonds to form themore thermodynamically stable vitamin D3. Af-ter synthesis, vitamin D3 is transported to theliver by a vitamin D-binding protein, which isan α-globulin.

The efficiency of vitamin D3 synthesis de-pends on the melanin content of skin and age.Melanin, a skin pigment, reduces the efficiencyof vitamin D3 synthesis in the skin because itabsorbs UV light in the same region as 7-dehydrocholesterol. Although increased skinpigmentation can decrease the amount of vita-min D3 synthesized, melanin may also protectthe vitamin D which is synthesized from beingdestroyed as a result of sunlight exposure. Be-cause vitamin D3 is also converted to inactiveproducts such as lumisterol, tachysterol, andother sterols when exposed to UV light, vitaminD3 production in the skin plateaus after 30 minof sunlight exposure (Hollick, 1994).Therefore,extended sunlight exposure does not result inthe production of toxic amounts of vitamin D3.Approximately 80% of the required vitamin D3

can be provided by endogenous synthesis;how-ever, vitamin D3 synthesis in the skin may be re-duced by as much as 75% by the age of 70 years(Holick et al., 1989).

In the absence of inadequate endogenoussynthesis, vitamin D must be obtained from di-etary sources or from supplements. Few foodscontain vitamin D except for the flesh of fattyfish (salmon, mackerel, sardines), fish liver oils,and eggs from hens fed feed enriched with vita-min D. In the United States, all commerciallyproduced milk is fortified with vitamin D2 at alevel of 400 IU/L (1 IU = 0.025 µg of vitaminD3). Therefore, in the United States (and othereconomically advanced countries) most dietaryvitamin D is obtained from milk and other vita-min D2-fortified foods. Both vitamin D2 and vita-min D3 are converted at the same rate to25-hydroxyvitamin D by a hydroxylase in theliver and are equally active as a prohormone.Because dietary uptake of vitamin D is depend-ent on normal fat absorption, conditions inwhich fat malabsorption is present can result invitamin D deficiency. Because breast milk con-tains little vitamin D, vitamin D deficiency canoccur in infants who are solely breastfed, arenot exposed to adequate sunlight, and are notreceiving vitamin D supplements.The adequateintake of vitamin D for children is 5 µg/day(200 IU/day) (Table 30-2).

Although vitamin D3 toxicity cannot be

caused by extensive exposure to sunlight, ex-cessive dietary intake of vitamin D can result indeleterious concentrations of vitamin D. Vita-min D toxicity is characterized by hypercal-cemia (elevated serum calcium concentration)and hyperphosphatemia (elevated serum phos-phate concentration).The end result of hypervi-taminosis D is calcinosis (calcification of softtissues) of the kidney, heart, and lungs. Calcifica-tion of the tympanic membrane of the ear canresult in deafness. The tolerable upper limit ofvitamin D intake is 1000 IU/day for infants and2000 IU/day for children and adults.

Vitamin D that is taken up by the liver is con-verted to 25-hydroxyvitamin D by a microsomalhydroxylase (Fig. 30-3). 25-Hydroxyvitamin D isthe main circulating form of vitamin D in theserum and the best indicator of vitamin D sta-tus. Normal serum levels are 14–60 ng/mL(35–150 nmol/L). When serum calcium con-centrations decline, 25-hydroxyvitamin D isconverted to 1,25-dihydroxyvitmin D by 1α-hydroxylase, a mixed-function oxidase that islocated in the inner mitochondrial membranein kidney tissue and whose expression is regu-lated by parathyroid hormone (PTH). Themain function of 1,25-dihydroxyvitamin D isto increase the intestinal absorption of dietarycalcium and phosphorus. When serum con-centrations of calcium and phosphorus are nor-mal or when large doses of vitamin D areadministered, 25-hydroxyvitamin D is metabo-lized to 24,25-dihydroxyvitamin D in the renal


Figure 30-3. Metabolism of vitamin D.

cortex. Normal levels of 24,25-dihydroxyvitaminD in the circulation are between 1 and 4 ng/mLbut are not detectable in patients with vitamin Ddeficiency.

Tissues contain two types of receptors for1,25-dihydroxyvitamin D: a classic steroid hor-mone nuclear receptor and a putative membranereceptor. 1,25-Dihydroxyvitamin D interacts withthe nuclear receptor to form a receptor-ligandcomplex (Fig. 30-4).This complex then interactswith other nuclear proteins, such as the retinoicacid receptor (RXR) to form a functional tran-scription complex. The main effect of this tran-scription complex is to alter the amount ofmRNAs coding for selected proteins such as cal-bindin, the calcium transport protein in the intes-tine, and the vitamin D receptor. In concert withPTH, 1,25-dihydroxyvitamin D acts to mobilizecalcium from bone.As a consequence, serum cal-cium and phosphate homeostasis is maintainedby a combination of 1,25-dihydroxyvitamin Dstimulation of intestinal absorption and boneturnover.

Calcium homeostasis is maintained by a regu-latory system comprised of vitamin D, PTH, andvarious feedback mechanisms (Fig. 30-5). Theconcentration of calcium in the ECF is moni-tored by a calcium receptor (CaR) present inthe parathyroid gland and renal tubular cells.The CaR is a 120-kd G-protein-coupled receptorthat is responsive to the concentration of extra-cellular calcium. At normal or elevated serumcalcium concentrations, the receptor is inac-tive, and PTH release from the parathyroidgland is inhibited. Conversely, reducing the ECFionized calcium concentration below the setpoint for CaR activation removes this inhibitionand triggers PTH release. If the serum calciumconcentration falls below the set point for thereceptor, CaR undergoes a conformationalchange that removes the inhibitory pathwaythat triggers PTH release.

Chief cells in the parathyroid gland synthe-size, store, and secrete PTH. PTH is synthesizedas a pre-pro PTH precursor. The pre- and pro-segments are cleaved enzymatically during


Figure 30-4. Hormone action of 1,25-dihydroxyvitamin D. VDR, vitamin D receptor;RXR, retinoic acid receptor;VDRE, vitamin D response element.

intracellular synthesis and processing in theGolgi apparatus. PTH is then secreted, stored, ordegraded.The mature intact PTH molecule thatis secreted is a 9.5-kd, 84-amino acid peptidewith a half-life of less than 5 min. Intact PTH ismetabolized by the liver and kidneys to gener-ate the metabolically active 2.5-kd fragmentthat contains the amino terminal portion of theintact peptide, and the inactive carboxy termi-nal fragment, which is excreted in urine. IntactPTH accounts for only 5%–25% of the total cir-culating PTH.

PTH interacts with receptors located on theplasma membrane of target cells such as boneand kidney,where it increases the cAMP concen-tration, resulting in an increase in intracellularcalcium. This increase in intracellular calciuminitiates a cascade of intracellular events medi-ated by phospholipase C acting on phos-phatidylinositol bisphosphate. Phospholipase Ccatalyzes the hydrolysis of phosphatidylinositolbisphosphate to diacylglycerol and inosi-toltrisphosphate.The primary role of PTH is toincrease bone turnover to increase blood cal-cium. In bone, the osteoblast is the primary tar-get of PTH. PTH regulates gene expression in

the osteoblast, promoting synthesis of matrixproteins required for new bone formation andproteins associated with matrix degradation andturnover.

Chronic exposure to PTH increases boneresorption by altering the number and activityof osteoblasts (bone-forming cells) and osteo-clasts (bone-resorbing) cells.The effect of PTHon osteoclasts is indirect and is exertedthrough the release of local mediators pro-duced by the osteoblast or released from thebone matrix. PTH decreases collagen synthesisin osteoblasts and increases osteclastic boneresorption, resulting in a net increase in cal-cium and phosphate release from bone intothe ECF. The serum calcium concentration isrestored to normal, but hypophosphatemiapersists, and results in impaired minneraliza-tion of bone. In the absence of disease, an in-crease in serum calcium reduces PTH secretionthrough a negative-feedback loop to maintainhomoeostasis.

PTH also activates renal 1α-hydroxylase, in-creasing the amount of the active form of vita-min D (1,25-dihydroxyvitamin D), which inturn enhances intestinal calcium absorption. In


Figure 30-5. Regulation of calcium homeostasis by the combined action of 1,25-dihydroxyvitamin D and parathyroid hormone (PTH). ECF, extracellular fluid.

the kidney, PTH increases calcium reabsorptionin the distal tubules and decreases reabsorptionof phosphate in the proximal tubule, therebypromoting phosphaturia.Approximately 85% ofrenal phosphorus reabsorption occurs in theproximal tubule. PTH synthesis and secretionare controlled not only by the extracellular cal-cium concentration but also by negative feed-back inhibition by 1,25 dihydroxy-vitamin D.

Rickets can also be the result of hypo-phosphatemia. Because a dietary deficiency ofphosphorus is very rare, hypophosphatemia isusually the result of excess phosphate excretionin the urine due to one of several rare inheriteddisorders such as X-linked hypophosphatemicrickets or autosomal dominant hypophos-phatemic rickets.These disorders share commonfeatures, including renal phosphate wasting, low1,25-dihydroxyvitamin D concentrations, shortstature,bone pain, rickets, and osteomalacia.Mu-tations in the product of the PHEX gene (phos-phate-regulating gene with homologies toendopeptidase on the X chromosome) havebeen identified as the cause of X-linked hypo-phosphatemic rickets.These mutations result inan excess of the hormonelike phosphatoninprotein, leading to urinary phosphate leakageand ultimately hypophosphatemia (Schiavi andKumar, 2004).

Bone is a dynamic tissue in which both bonemodeling and remodeling occur. Bone model-ing occurs during skeletal growth and results inan increase in bone mass. Remodeling occursafter skeletal growth has ceased and is the re-sponse to stress on the skeleton and changes in

diet (e.g., low calcium intake).Bone remodelingis also necessary for repair of minor fracturesthat occur in bones over time. It has been esti-mated that, at any one time, approximately 4%of bone is undergoing remodeling. Remodelingis initiated by the activation of osteoclasts by var-ious cytokines that are produced by osteoblasts.Prolonged calcium deprivation results in in-creased recruitment of osteoclasts which cat-alyze the release of calcium phosphate andpeptides from the bone matrix, includingcollagen-derived hydroxyproline and the pyridi-noline collagen cross-links discussed in the nextparagraph. Prolonged exposure of osteoblasts tothe increased PTH level associated with calcium-deficiency eventually leads to increases in theserum concentrations of biochemical markers ofbone synthesis, such as alkaline phosphatase andosteocalcin.

The principal protein in bone is type 1 colla-gen, a triple helix comprised of two identicalα1-chains and one α2-chain. Collagen is synthe-sized as precursor type 1 procollagen, contain-ing both N- and C-terminal extensions (Fig.30-6). Procollagen undergoes posttranslationalprocessing reactions that include hydroxylationof proline and lysine residues, glycosylation,and formation of interchain disulfide bonds.Af-ter it is secreted, procollagen is converted tocollagen by extracellular processing involvingenzymatic cleavage of the N- and C-propep-tides (Fig. 30-7). Three amino acid side chainsreact to form a trivalent amino acid structurethat contains a pyridinium ring. Deoxypyridino-line is formed from two hydroxylysyl residues


Figure 30-6. Schematic diagram of type 1 collagen. Vertical arrows indicate cleavagesites between the mature collagen molecule, the N-propeptide, and the C-propeptide.Reproduced with permission from Rossert J and de Crombrugghe B (2002).

and one lysyl residue, whereas pyridinoline isformed when three hydroxylysyl moieties com-bine to form a cross-link. Deoxypyridinoline oc-curs in bone, dentine, ligaments, and aorta,whereas pyridinoline is more widespread inhard connective tissues that also contain largeamounts of soft cartilage. Since these cross-linksare formed in mature collagen and are not fur-ther metabolized after bone resorption, theycan be used as markers of bone turnover.

Markers of bone resorption can be measuredin serum or urine, whereas bone formationmarkers, such as bone-specific alkaline phos-phatase, are usually measured in serum. Mea-surement of bone markers allows for real-timeassessment of bone resorption or formation andcan be used to monitor therapy (Ravn et al.,2003). The pyridonolines (deoxypyridinolineand pyridinoline) and the N- and C-teleopep-tides are the most frequently measured markersof bone resorption. Pyridinoline and teleopep-tides (NTx and CTx) are increased in individu-als with metabolic bone diseases associated

with increased bone resorption, such as osteo-malacia and primary hyperparathyroidism, andthey are decreased in individuals with hypo-parathyroidism.

Alkaline phosphatases from liver, bone, andkidney are members of an isozyme family. Tissue-specific isoforms are produced by tissue-specificposttranslational processing. Because osteoblastsare the source of bone-specific alkaline phos-phatase (BSAP) and serum levels of this iso-enzyme reflect osteoblast activity, BSAP can beused as a marker of bone formation. It has a rela-tively long half-life (1–2 days) and is unaffectedby diurnal variation. Reference intervals foradults are 7–30 U/L and are dependent on ageand gender. Children have much higher levels ofBSAP activity, especially during growth spurts.

THERAPY

The type of therapy for rickets depends on thespecific cause of the disease. For patients with


Pyrroliccross-links

PYL

Cancellousbone

N

CH2

CH2

CH2CH

Collagen fibrils

Mineral

DPD

Pyridiniumcross-links

OH NH2

CH2 CH COOH

COOH

CH2

NH2

CH2 CH COOH

CH2

CHNH2

DPL

N

CH2 CH2

CH2

CH2

N

HO

NH2

CH2 CH COOH

COOH

CH2

NH2

CH2 CH COOH

CH2

CHNH2

N+

CH2 CH2

CH2

CH

CH2

NH2

CH2 CH COOH

COOH

CH2

NH2

CH2 CH COOH

NH2

HO

PYD

N+

CH2 CH

CH2

CH

CH2

NH2CH

CH2 CH COOH

COOH

CH2

NH2

CH2 CH COOH

NH2

HO

HO

N+

OH

Figure 30-7. Stabilization of bone matrix by pyridinium and pyrroline cross-links. PYD,pyridinoline cross-link; DPD, deoxypyridinoline cross-link. Reproduced with permissionfrom Rossert J and de Crombrugghe B (2002).

pure vitamin D deficiency due to lack of expo-sure to sunlight or inadequate dietary intake,treatment with oral preparations of vitamin Dor natural sources rich in the vitamin, such asfish oil, can be used. Conventional treatmentconsists of the administration of 2000–5000IU/day of vitamin D until the serum alkalinephosphatase levels returns to normal or radiol-ogy shows signs of bone healing. This usuallytakes from 2 to 10 weeks. Thereafter, mainte-nance is achieved with 400 IU/day vitamin D or800–1200 IU/day if the patient is also takinganticonvulsants. Assuming laboratory facilitiesare available, one can monitor serum levels ofcalcium, phosphate, and alkaline phosphatase.Alternatively, but less commonly, the 25-hydroxyform of vitamin D (20–50 µg/day) or 1,25-dihydroxivitamin D (0.5–1 µg/day) can begiven. For type 1 vitamin D-deficiency rickets,the treatment consists of 1,25-dihydroxyvitaminD. Patients with type 2 vitamin D-resistant rick-ets, are treated with large doses of calcium sup-plements.

In West Africa and many other regions of sub-Saharan Africa where lack of adequate dietarycalcium is the main cause of rickets, the goal ofimmediate therapy is to provide the rachiticchild with 1000–1500 mg of calcium per dayfor 6 months to promote bone healing.This canbe accomplished through calcium replacementtherapy using daily calcium carbonate supple-ments or three to four glasses of milk per day,depending on which is more economically fea-sible. Each 250 mL of milk provides 300 mg cal-cium.The effectiveness of the calcium therapyshould be monitored weekly, as described inthe Diagnosis section.

The major emphasis should be the preventionof rickets by providing adequate calcium and vi-tamin D in the diet, particularly if exposure tosunlight is restricted. One teaspoon (4 mL) ofcod-liver oil provides 360 IU of vitamin D. In de-veloping countries where dairy products areprohibitively expensive, local foods such as driedfish containing small soft bones can add to thecalcium intake (Larsen et al., 2000).

QUESTIONS

1. Assume that you are a public health offi-cial in northern Nigeria. What actionsmight you take at the population level toreduce the risk and incidence of calcium-deficiency rickets in the region?

2. How would you distinguish if a child’srickets was caused by inadequate dietarycalcium versus a deficiency of vitamin D?

3. Assume that a 30-year-old man who hadbeen in good health previously developsan ectopic PTH-producing tumor.What arethe biochemical and pathological changesyou would expect to find in this patient?

4. What are the consequences of a geneticdeficiency of 1α-hydroxylase on the bonestatus of an individual, and what treatmentwould be appropriate for a patient withthis disorder?

5. How would a mutation that inactivatesthe parathyroid calcium receptor (CaR) af-fect calcium homeostasis?

6. How can biochemical markers be used toassess the therapeutic efficacy of drugs totreat osteoporosis?

BIBLIOGRAPHY

Food and Nutrition Board: Dietary Reference In-takes: Applications in Dietary Assessment. In-stitute of Medicine, National Academy Press,Washington, DC, 2000.

Hollick MF:Vitamin D: importance in the preventionof cancers, type 1 diabetes, heart disease, and os-teoporosis.Am J Clin Nutr 79:362–371,2004.

Holick MF, Matsuoka LY,Wortman J:Age, vitamin Dand solar ultraviolet. Lancet 2:1104–1105, 1989.

Holick MF: McCollum Award Lecture, Vitamin D:new horizons for the 21st century. Am J CinNutr 60:619–630, 1994.

Jergas M, Genant HK: Radiologic imaging of meta-bolic bone disease, in Coe FL, Favus MJ (eds):Disorders of Bone and Mineral Metabolism.Lippincott Williams and Williams, Philadelphia,2002, pp. 428–447.

Larsen T, Thilsted SH, Kongsback K, et al.: Wholesmall fish as a rich calcium source. Br J Nutr83:191–196, 2000.

Ravn P,Thompson DE, Ross PD, et al.: Biochemicalmarkers for prediction of 4-year response inbone mass during bisphosphonate treatmentfor prevention of postmenopausal osteoporosis.Bone 33:150–158, 2003.

Rossert J, de Crombrugghe B: Type 1 collagen:structure, synthesis, and regulation, in BilezikianJP, Raisz LG, Rodan GA (eds): Principles of BoneBiology. Academic Press, San Diego, CA, 2002,pp. 189–210.

Schiavi SC, Kumar R:The phosphatonin pathway:new insights in phosphate homeostasis. KidneyInt 65:1–14, 2004.

Thacher T: Calcium-deficiency rickets. Endocr Dev6:105–125, 2003.


Tsai K-S, Jang M-H, Hsu S H-J, et al.: Bone alkalinephosphatase isoenzyme and carboxy-terminalpropeptide of type 1 procollagen in healthy Chi-nese girls and boys.Clin Chem 45:136–138,1999.

Wharton B, Bishop N: Rickets. Lancet 362:1389–1400, 2003.

World Health Organization: Measuring Changesin Nutritional Status. World Health Organiza-tion, Geneva, Switzerland, 1983.

Zittermann A: Vitamin D in preventive medicine:are we ignoring the evidence? Br J Nutr 89:552–572, 2003.


CASE PRESENTATION

A 46-year-old Caucasian man was referred tothe gastroenterology clinic for evaluation of ab-normal liver chemistries. The patient had ini-tially presented to his primary care physician 1month earlier complaining of new-onset fa-tigue, impotence, and diminished libido. He de-nied any abdominal pain, confusion, or changein skin color. He did note a recent 10-poundweight loss.

The patient had recently immigrated to theUnited States from Scotland. He claimed onlysocial use of alcohol and denied any illicit druguse. He worked at a bank and denied exposureto toxic materials.

His primary care physician obtained the fol-lowing serum tests, which indicated liver dam-age: alanine aminotransferase 103 U/L (reference10–55 U/L); aspartate aminotransferase 967 U/L(reference 10–55 U/L); alkaline phosphataselevel 125 U/L (reference 45–115 U/L); and totalbilirubin 0.8 mg/dL (reference 0–1.0 mg/dL). Inaddition, the serum glucose level was 711 mg/dL(reference 30–100 mg/dL). Decreased levels offollicle-stimulating and luteinizing hormoneswere also noted.

On further testing, the patient displayedthe biochemical signs of iron overload. Hehad a serum iron of 197 mg/dL (reference30–360 µg/dL), a total iron binding capacity of202 µg/dL (reference 228–428 µg/dL), and aferritin level of 4890 ng/mL (reference 30–300 ng/mL). His serum transferrin saturationwas calculated to be 97.5% (reference 20%–50%).

A liver biopsy was performed and revealed

31

Hereditary Hemochromatosis

SCOTT A. FINK and RAYMOND T. CHUNG

335

advanced-stage cirrhosis and hemosiderosiswith marked iron deposition in hepatocytes,Kupffer cells, and bile ducts. Hepatic iron con-centration was determined to be 37,880 µg/gdry weight (reference 200–2400 µg/g dryweight).

A diagnosis of hemochromatosis was estab-lished, and the patient was tested for HFE genemutations. He was homozygous for the C282Ymutation.

The patient was begun on a phlebotomy pro-gram, a therapy in which whole blood is takenfrom the patient intravenously, to remove theexcess iron from his blood, eventually leadingto iron equilibrium. His libido improved, he hasachieved better glycemic control, and he nor-malized his liver chemistries.

Given the diagnosis of hereditary hemochro-matosis (HH), his two siblings were tested forHFE mutations. Both were heterozygous for theC282Y mutation and showed no evidence ofiron overload.

DIAGNOSIS

Abnormally high systemic iron levels can leadto cirrhosis of the liver, diabetes mellitus, andheart failure.Although a number of disease pro-cesses can lead to iron overload, this chapter fo-cuses on hereditary hemochromatosis, theprototypical disease of iron overload.

The most common Mendelian genetic disor-der in Caucasians, HH is estimated to occurwith a prevalence of approximately 1 in 200.Among Caucasians, cases are concentrated inthose of northern European origin, specifically

individuals of Nordic or Celtic ancestry (Tavill,2001).

Iron stores are maintained in a delicate bal-ance,controlled primarily at the level of absorp-tion. It is hypothesized that an increase in theiron regulatory “set point” promotes excessivedietary iron absorption in individuals with HHdespite already elevated iron stores (Parkkila,2001).

Patients with HH who have developed ironoverload can present with fatigue, malaise, ab-dominal pain, arthralgias, and impotence. How-ever, the majority of patients are asymptomaticat the time that serum indices of iron overloadare first seen (Bacon, 2001).

Once iron overload has been present formany years, organ damage occurs. Patientsdevelop arthritis, cirrhosis, congestive heart fail-ure, increased skin pigmentation, cardiac ar-rhythmias, and diabetes as a result of pancreaticislet cell infiltration. Physical examination ofpatients with mild iron overload is generallybenign.Those with full-blown hemochromato-sis often exhibit hepatomegaly; stigmata ofchronic liver disease, including characteristicskin lesions; splenomegaly; and in patients withcardiomegaly, findings associated with conges-tive heart failure. In addition,patients whose en-docrine system is affected may have testicularatrophy or signs of hypogonadism and hypothy-roidism (Bacon, 2001).

The diagnosis of HH is established based onserum transferrin saturation (TS), defined asserum iron divided by total iron binding capac-ity (TIBC). Since serum iron and ferritin levelslack specificity for diagnosis when used alone,measurement of fasting TS is currently recom-mended as a first screen to detect iron over-load.TS is the best indirect biochemical markerof iron stores. A fasting TS of greater than 45%will detect over 98% of all cases of phenotypichemochromatosis (Tavill, 2001).

Once serum TS is determined to be greaterthan 45% and the serum ferritin elevated, apolymerase chain reaction (PCR) based genetest for HH is recommended. Two genotypicprofiles are consistent with the diagnosis ofHH: homozygosity for the C282Y mutation orcompound heterozygosity with a C282Y/H63Dgenotype. If homozygosity for the C282Y muta-tion is detected, then PCR testing should be of-fered to first-degree relatives of the proband(Tavill, 2001).

If liver disease is suggested either biochemi-cally or clinically, then the liver should be biop-

sied to establish the presence of advanced fi-brosis or cirrhosis, to rule out iron overload inthe liver when serum markers of iron overloadare equivocal, or to investigate other causes ofliver disease (Tavill, 2001).


The French pathologist Trousseau first de-scribed a patient with hemochromatosis in1865. Four years later, the German patholo-gist Von Recklinghausen (1889) coined thephrase hemochromatosis when describing apatient with pigmentation (“chrom”) thoughtto be caused by a factor in the bloodstream(“hemo”).

Not until 1996 did Feder and colleaguesidentify a novel major histocompatibility com-plex (MHC) class I-like gene in which homozy-gosity for specific mutations were found in 83%of patients with hemochromatosis.

In 1998, Zhou and colleagues used an HFEknockout mouse model to elegantly bring to-gether the pathophysiological and molecular as-pects of the disease. Even when fed a standarddiet, the knockout mice showed abnormallyhigh transferrin saturations and excessive irondeposition in the liver and passed these traitson in an autosomally recessive manner.

A perspective on normal iron metabolism isnecessary for understanding the biochemistryof HH. The healthy adult body has a total ironcontent of 3–5 g of iron, two thirds of which isincorporated in erythrocytes and precursors intheir lineage (Pietrangelo, 2002). Only 1–2 mgof iron, 10% of the total ingested from the diet,is absorbed daily (Parkkila et al., 2001). Eachday, 1–2 mg of iron leaves the body throughprocesses such as menstruation and sloughingof skin (Pietrangelo, 2002).

Approximately 20 mg of iron is requireddaily for erythropoiesis. The iron requirementsare supplied by recycled iron and from residualbody iron stores. One major store is in hepato-cytes where 0.5–1 g of iron is bound to special-ized proteins such as ferritin and hemosiderin.Ferritin is a large, 440-kd cellular storage pro-tein for iron. Measurement of plasma ferritin isseen as a reflection of the cellular ferritinstores, which in turn reflects cellular ironstores. Body stores of iron are maintained by re-cycling: Senescent erythrocytes are ingested bymacrophages, and their iron is taken up byserum transferrin. The iron is delivered to


the bone marrow, where it once again will beincorporated into erythrocytes (Pietrangelo,2002).

Iron levels are tightly regulated through con-trol of dietary absorption of iron. The duode-num and upper jejunum are the only areas ofthe body where this occurs. Since nonhemeiron forms insoluble complexes when ingested,it must first be converted into soluble com-plexes.This is accomplished on the apical sur-face of duodenal villus enterocytes by duodenalferric reductase, which converts insolubleduodenal ferric (Fe3+) iron into soluble and ab-sorbable ferrous (Fe2+) iron. Iron is then trans-ported across the membrane to the cytoplasmthrough a transporter known as the divalentmetal transporter 1 (DMT-1), a proton sym-porter (Harrison and Bacon, 2003).

Once absorbed, iron becomes part of the cel-lular iron pool, either stored as ferritin or trans-ported across the basolateral membrane of theenterocyte into the circulation by an iron trans-porter called ferroportin 1. Hephaestin, a baso-lateral membrane ferroxidase, oxidizes theferrous iron back to its ferric form, thus com-pleting the absorption process (Harrison andBacon, 2003).

In the bloodstream, ferric iron binds tightlyto circulating plasma transferrin (TF) to formdiferric transferrin (FeTF). Absorption of ironinto erythrocytes depends on basolateral mem-brane receptor-mediated endocytosis of FeTFby transferrin receptor 1 (TfR 1). FeTF binds toTfR 1 on the surface of erythroid precursors.These complexes invaginate in pits on the cellsurface to form endosomes. Proton pumpswithin the endosomes lower pH to promote therelease of iron into the cytoplasm from transfer-rin.Once the cycle is completed,TF and TfR 1 arerecycled back to the cell surface. TF and TfR 1play similar roles in iron absorption at the baso-lateral membrane of crypt enterocytes (Parkilla etal.,2001;Pietrangelo,2002).

Iron transport across the intestinal cell oc-curs at both the apical and basolateral inter-faces. Figure 31-1 highlights the importance ofthis polarity in both iron transport into the celland the sensing of iron stores in both the villusand crypt enterocytic apical and basolateral in-terfaces.The apical membrane is specialized totransport heme and ferrous iron into the cellthrough three major pathways. The first is viaDMT-1, which transports ferrous iron and diva-lent metal ions into the enterocyte. Iron canalso be absorbed as the intact heme moiety,

which after easily passing through the brushborder of the apical membrane intact, is brokendown by heme oxygenase into elemental iron.Finally, intestinal mucins and other proteinssuch as mobilferrin, integrins, and ferric reduc-tase can transport iron directly across the apicalmembrane (Parkkila et al., 2001).

The need for iron by erythrocytes dictatesthe direction of body iron stores and is thebody’s iron supply priority. Duodenal epithelialcells must be kept programmed to respond tothese requirements and to the status of thebody’s iron deposits (Pietrangelo, 2002).The de-fect in the molecular iron absorptive machineryin HH is thought to be localized within the in-terplay between the enterocyte and body ironstores. The “stores regulator” hypothesis de-scribes a pathway that would facilitate a slowaccumulation of dietary iron and would pre-vent iron overload after iron stores are deemedadequate. Potentially involving TF-bound iron,serum ferritin, and serum TF, the stores regula-tor is thought to be impaired in individualswith HH. It has been shown, for example, thatafter phlebotomy in patients with HH, there isan increase in intestinal absorption of iron thatpersists even after adequate iron stores are re-plenished.This implies that the defect rests notwith abnormally functioning absorptive ma-chinery, but with abnormal feedback from thebody’s stores regulator, which either is unableto recognize the replenishment of adequateiron stores or cannot communicate this infor-mation to the small intestine (Parkkila et al.,2001).

The HFE gene, mutations of which are re-sponsible for HH, codes for a novel MHC classI-like protein that requires interaction with β2-microglobulin for normal presentation to thecell surface (Bacon, 2001). Located on the shortarm of chromosome 6, it has been detectedby immunohistochemistry in small-intestinalcryptal enterocytes (Parkilla et al., 2001).

The protein encoded by the HFE gene is a343-amino acid protein consisting of a 22-amino acid signal peptide, a large extracellulardomain, a single transmembrane domain, and ashort cytoplasmic tail (Fig. 31-2).As Figure 31-2shows, the extracellular domain includes threeloops (α1, α2, and α3) with intramolecular disul-fide bonds within the second and third loops, astructure similar to other MHC class I proteins(Feder et al., 1996).

The two most common mutations, C282Yand H63D, are in the extracellular domain.The

Hereditary Hemochromatosis 337

C282Y mutation represents a change from cys-teine to tyrosine at amino acid 282, while theH63D mutation is a substitution of aspartate forhistidine at amino acid 63 (Feder et al., 1996). Ithas been estimated that 83%–100% of patients

with HH are homozygous for C282Y (Baconand Briton, 2003).

Ten to fifteen percent of Caucasians of Eu-ropean ancestry are heterozygous for theC282Y mutation. It has been suggested that


Recyclingendosomes

ApicalBasolateral

HFE Transferrinreceptor

Crypt cell

Apotransferrin

Ferritin

β2-microglobulin

Iron

DMT1

Ferroportin 1

Hephaestin

Villus cell

B

A

Figure 31-1. A schematic model of HFE regulation of iron transport in duodenal ente-rocytes. A and B correspond to villus and cryptal enterocytes, respectively. As noted,HFE lies at the center of regulation of iron absorption through its role in sensing bodyiron stores in the villus enterocyte. It communicates this information to the crypt ente-rocyte indirectly through regulation of development of ferroportin and DMT-1.Reprinted with permission from Parkkila et al. (2001). © 2001,American Gastroentero-logical Association.

the mutation originated in Celts and spread tonorthern Europe, perhaps imparting an advan-tage at a time when dietary iron availabilitywas either poor or impaired by parasitic infec-tion (Bacon and Britton, 2003).

The HFE protein interacts with TfR 1, a pro-tein known to be involved in iron metabo-lism, at the HFE-β2 microglobulin complex(Parkilla 2001). By abolishing a disulfide bondin the α3 loop of HFE’s transmembrane do-main, the C282Y mutation interferes with thisinteraction (Feder et al., 1996). HFE binds TfR1 with an affinity similar to that of transferrin(Pietrangelo, 2002). While binding to TfR 1 isnot required for targeting of HFE to the basolat-eral membrane, it is required for HFE to betransported to transferrin-positive endosomesfor regulation of intracellular iron homeostasis(see Fig. 31-1). The biological effect of HFE onTfR may be exerted in the endosomal compart-ment, where iron is released from the TfR-TFcomplex (Parkkila et al., 2001).

As the primary site of iron absorption, theduodenum shows a distinctive pattern of HFEexpression: HFE is highly expressed in cryptbut not villus duodenal cells (Parkilla et al.,2001). As the absorption of ferrous iron fromthe diet is principally mediated by DMT-1 at thevillus tip, a connection between abnormal HFEand overexpression of DMT-1 has been hypoth-esized. While the DMT-1 protein is expressed

primarily in villus cells, mRNA expression forDMT1 begins in crypt cells. Mutated HFE maylead to decreased uptake of plasma iron bycrypt cells, thus diminishing the intracellulariron pool.This might in turn result in increasedexpression of mRNA for DMT-1 and ferroportin1 in villus enterocytes that mature from cryptenterocytes. This hypothesis would link dys-functional HFE in crypt cells to overabsorptionof iron in villus cells (Bacon, 2001).

An intriguing question has been what role, ifany, the liver plays in sensing and regulatingiron absorption.As the liver also expresses theHFE protein and is a major iron-storing organ, adefective signaling mechanism related to abnor-mal HFE has been postulated as a cause of ironoverabsorption. For many years, however, therewas little understanding of possible mecha-nisms for signaling between the liver and thesmall intestine.

Recent studies have characterized a 25-aminoacid, 2- to 3-kd circulating peptide of hepaticorigin called hepcidin. Hepcidin is thought tohave anti-inflammatory properties perhaps re-lated to its ability to downregulate iron stores.It has been shown that hepatic expression ofhepcidin mRNA is significantly lower in pa-tients with hemochromatosis when comparedwith controls. Hepcidin mRNA expressionhas also been shown to be decreased in HFEknockout mice. These data suggest that


Figure 31-2. Model of HFE protein. The HFE protein shares structural similarities toother MHC class I proteins.The locations of the two most common mutations are noted.Reprinted with permission Feder et al. (1996).

circulating hepcidin produced in the liver mayplay a role in inhibiting iron absorption in thesmall bowel, and that this effect is suppressedin patients with HFE mutations (Bridle et al.,2003).

Without a mechanism for its excretion, ironaccumulates in vital organs (Pietrangelo, 2002).Because the liver binds both circulating non-transferrin and transferrin-bound iron, the liveris at particular risk for iron overload. Excessiron causes damage to hepatocytes primarilythrough induction of oxidative stress (Parkillaet al., 2001).

During a cell’s normal life cycle under aero-bic conditions, some of the consumed oxygenis reduced to highly reactive molecules calledreactive oxygen species (ROS).Transition metalions such as iron, with their frequently un-paired electrons, act as excellent catalysts forthe creation of ROS. The body’s inability tomodulate “free iron” availability creates an envi-ronment prone to the formation of ROS andfree-radical induced cellular damage in theevent of iron overload. The “classical” reactionbetween Fe3+ and superoxide (O2

•−) is known asthe Haber-Weiss reaction:

Fe3+ + O2•− → Fe2+ + O2

H2O2 + Fe2+ → OH• + OH− + Fe3+

Figure 31-3 shows the common biochemicalreactions responsible for the initiation of freeradicals (Pietrangelo, 2002). ROS are highly het-erogeneous in terms of half-life and reactivityagainst cellular targets. Common ROS includeH2O2, singlet molecular oxygen (1O2), hydroxyl

(OH•), superoxide (O2•−), alkoxyl (RO•),peroxyl

(ROO•), and nitric oxide radicals (NO•). ROSplay important physiological roles in normalsignal transduction pathways, mitochondrialrespiration, and transcriptional factor activity(Pietrangelo, 1998).

These free radicals are extremely reactiveand capable of attacking cell constituents.Polyunsaturated fatty acids found in mem-brane phospholipids, proteins, and nucleicacids are all vulnerable targets (Pietrangelo,1998). Cellular antioxidant defenses exist tobreakdown ROS. Owing to these defenses, asevere iron burden is necessary to cause dam-age (Pietrangelo, 2002).

Fibrosis of the liver is caused by the exces-sive accumulation of extracellular matrix(ECM) components. These include interstitialcollagens, noncollagenous glycoproteins suchas fibronectin, laminin, undulin, entactin, vitro-nectin, and proteoglycans. They normally pro-vide cohesiveness for cells, promote normaltissue architecture, and play roles in normal cellfunction and differentiation. Chaotic expansionof the ECM leads to pathological disruption ofthe histological architecture of the liver and re-placement of hepatocytes and bile ductuleswith ECM. Cytokines, growth hormones, andother biological peptides promote expansion ofthe ECM (Pietrangelo, 1998).

During the fibrogenic process, Kupffer cellsand invading monocytes stimulate fibrogenesisthrough release of soluble factors such as trans-forming growth factor β (TGF-β),platelet-derivedgrowth factor (PDGF), and chemokines, whichactivate a subpopulation of cells called hepaticstellate cells as shown in Figure 31-4. Once acti-vated, hepatic stellate cells undergo a myo-fibroblastlike transformation. Iron can directlypromote activation of hepatic stellate cells byaccumulating in monocytes and Kupffer cells(Pietrangelo, 1996).

In addition to its effects on the promotion ofROS formation and fibrogenesis, iron is directlytoxic to hepatocytes and causes hepatocellularnecrosis (sideronecrosis). Iron also acts as a co-factor in the promotion of fibrogenesis byother hepatotoxins such as alcohol and viruses(Pietrangelo, 1998).

Iron damages hepatocellular organelles.Mito-chondria exposed to excessive iron show in-creased fragility, increased volume, increasedpH, decreased fluidity, and increased lipid per-oxidation (Britton, 1996). Lysosomes exposedto iron overload show increased fragility and


Figure 31-3. Free radicals resulting from the re-action of ferric iron with hydrogen peroxide canprogress to damage cell membranes through lipidperoxidation (Britton, 1996). LH = polyunsatu-rated fatty acid; R = free radical; LOO = lipid per-oxyl radical; LOOH = lipid peroxide.

release of hydrolytic enzymes directly into thecytoplasm, further promoting cellular damage.

THERAPY

Hereditary hemochromatosis evolves clinicallyin a series of stages. In the first stage, from 0 to20 years of age, there is clinically insignificantiron accumulation of less than 5 g of parenchy-mal iron. After 20–40 years of life, approxi-mately 10–20 g of iron accumulates, and ironoverload is evident in the bloodstream. Diseaseis still not evident clinically. Finally, after ap-proximately 40 years of iron accumulation,more than 20 g of iron accumulates, and the pa-tient progresses to the stage of iron overloadwith organ damage. In its final stages, HH canlead to decompensated cirrhosis,hepatocellularcarcinoma, diabetes mellitus, and cardiomyopa-thy (Tavill, 2001).

Once iron overload is established,current rec-ommendations are to proceed with phlebotomy.Ideally, phlebotomy should be initiated before

the onset of clinically significant disease. Oncesymptoms set in, the malaise, fatigue, skin pig-mentation, and abdominal pain can be reversedwith phlebotomy. However, hemochromatosis-associated arthropathy, hypogonadism, andcirrhosis are less responsive to therapeuticphlebotomy.

As with other end-organ damage, diabetesmellitus can be prevented if phlebotomy is initi-ated prior to pancreatic damage. It is importantto note that, while phlebotomy may reduce in-sulin requirements,hemochromatosis-associateddiabetes with onset prior to initiation of phle-botomy will likely continue (Tavill, 2001). Dia-betes mellitus remains a major cause of death inpatients with noncirrhotic HH, occurring seventimes more frequently in patients with HH whencompared with normal controls (Niederau et al.,1985).

Routine therapeutic phlebotomy, with a goalof the removal of 500 mL of whole blood (ap-proximately 200–250 mg of iron) weekly or bi-weekly, should be continued until iron-limitederythropoiesis develops.This is marked by the


Figure 31-4. Hepatic stellate cell activation. Iron-induced injury to hepatocytes can acti-vate Kupffer cells through the release of soluble growth factors that contribute to scarringof the liver through activation of hepatic stellate cells (HSC). Reprinted with permissionfrom Pietrangelo (1998).© 1998,European Association for the Study of the Liver.

failure of hemoglobin and hematocrit to re-cover before the next phlebotomy. Transferrinsaturation and ferritin levels are monitored peri-odically to predict the return to normal ironstores (Tavill, 2001).

QUESTIONS

1. What organs are most commonly affectedin patients with HH?

2. What are the genetics of hereditary ofHH?

3. How does the HFE protein normally con-tribute to maintenance of iron stores atnormal levels, and how could mutationsaffect this?

4. Describe the mechanisms by which ironmay by toxic to the liver.

5. What are ROS, and how does iron func-tion as a cofactor in their development?

6. How is HH treated?

BIBLIOGRAPHY

Bacon BR: Hemochromatosis: diagnosis and man-agement. Gastroenterology 120:718–725, 2001.

Bacon BR, Britton RS: Hemochromatosis and otheriron storage disorders, in Schiff ER, Sorrell MF,Maddrey WC,et al. (eds):Schiff ’s Diseases of theLiver. Philadelphia, Lippincott Williams andWilkins, 2003, pp. 1187–1205.

Bridle KR, Frazer DM, Wilkins SJ, et al.: Disruptedhepcidin regulation in HFE-associated he-

mochromatosis and the liver as a regulator ofiron homeostasis. Lancet 361:669–673, 2003.

Britton RS. Metal-induced hepatotoxicity. SemLiver Dis 16:3–12, 1996.

Feder JN, Gnirke A,Thomas W, et al.:A novel MHCclass-1-like gene is mutated in patients withhereditary haemochromatosis. Nat Genet13:399–408, 1996.

Harrison SA,Bacon BR:Hereditary hemochromato-sis: update for 2003. J Hepatol 38:S14–S23,2003.

Niederau, Fischer R, Sonnenberg A, et al.: Survivaland causes of death in cirrhotic and in noncir-rhotic patients with primary hemochromatosisN Engl J Med 313:1256–1262, 1985.

Parkkila S, Niemela O, Britton RS, et al.: Molecularaspects of iron absorption and HFE expression.Gastroenterology 121:1489–1496, 2001.

Pietrangelo A: Metals, oxidative stress, and hepaticfibrogenesis. Sem Liver Dis 16:13–30, 1996.

Pietrangelo A: Iron, oxidative stress, and liver fibro-genesis. J Hepatol 28:8–13, 1998.

Pietrangelo A: Physiology of iron transport and thehemochromatosis gene. Am J Physiol Gastroin-test Liver Physiol 282:G403–G414, 2002.

Pietrangelo A: Non-HFE hemochromatosis. SeminLiver Dis 25:450–460, 2005.

Tavill AS:Diagnosis and management of hemochro-matosis. Hepatology 33:1321–1328, 2001.

Trousseau A:Glycosurie,diabete sucre, in Cliniquemedicale de l’Hotel–Dieu de Paris. Vol. 2, 2nded. Balliere, Paris, 1865, p. 663.

Von Recklinghausen FD: Uber hamochromatose.Tageblatt der Versammlung Deutsch. Natur-forsch Arzte Heidelberg 324–325, 1889.

Zhou XY, Tomatsu S, Fleming RE, et al.: HFE geneknockout produces mouse model of hereditaryhemochromatosis. Proc Natl Acad Sci USA95:2492–2497, 1998.


Part V

Endocrinology and Integrationof Metabolism

Diabetes mellitus is a group of metabolic dis-eases characterized by hyperglycemia.This canresult from defects in insulin secretion, defectsin insulin action, or both. Because glucose isa chemically reactive molecule, the chronichyperglycemia of diabetes is associated withlong-term damage, dysfunction, and, ultimately,failure of various organs. These include theeyes, the kidneys, the nervous system, and thecardiovascular system. Several pathogenic pro-cesses are involved in the development of dia-betes.These range from autoimmune destructionof the β-cells of the pancreas with consequentinsulin deficiency (type 1 diabetes, about 10%of cases) to poorly characterized abnormalitiesthat result in resistance to insulin action com-bined with inadequate insulin secretion (type 2diabetes, about 90% of cases). Regardless ofmechanism, the ultimate basis of known abnor-malities in carbohydrate, fat, and protein metab-olism in diabetes is the deficient action ofinsulin on target tissues.

CASE REPORT

Initial Presentation

A 19-year-old girl was brought to the emer-gency department by her parents who reportedthat she had been vomiting and feeling weakfor 24 h. The patient complained of feelinglethargic and fatigued for a few weeks. She hadlost weight despite having a good appetite. De-spite drinking large volumes of water, she con-tinued to feel thirsty all the time. She alsocomplained of an increased frequency of urina-

tion during the day and at night. Her parentssaid that she had been having a mild fever and asore throat for 2–3 days, and that the day beforeshe had started vomiting and had not been ableto get out of bed by herself.

Her past medical history was unremark-able, and she was not currently taking anymedications. She had regular menstrual peri-ods and denied smoking, alcohol use, or recre-ational drug use.The patient’s father had type2 diabetes, and he was concerned that someof her symptoms were similar to what he ex-perienced at the time his diabetes was diag-nosed.

Physical Examination

The patient was a young girl lying in bed,breathing rapidly. Her pulse was 108 beatsper minute (normal is less than 80), and herblood pressure was 94/56 mm Hg supine and72/48 mm Hg sitting (normal is less than140/90 with no decrease on postural adjust-ment). Her temperature was 97.4°F (normal is98.6°F), and she was breathing at a rate of 34breaths per minute (normal is less than 20). Herreported height was 5 ft 6 inches, and shepresently weighed 110 pounds, but her usualweight was 125 pounds. She had poor skin tur-gor and dry mucous membranes. She had afruity odor on her breath. There was no thyro-megaly (enlargement of the thyroid gland), andthe cardiopulmonary examination was unre-markable.Her abdomen was diffusely tender,butpelvic examination was normal. She respondedto questions appropriately, although she wasslow to answer. There were no neurological

32

Type 1 Diabetes Mellitus

SRINIVAS PANJA, ARUNA CHELLIAH, and MARK R. BURGE

345

deficits. Results of the laboratory evaluation areshown in Table 32-1.

DIAGNOSIS

The patient was suffering from newly diag-nosed type 1 diabetes and one of its commonmetabolic decompensations, diabetic ketoacid-osis (DKA).

The diagnosis of type 1 diabetes is made by acombination of typical clinical features and lab-oratory tests. The American Diabetes Associa-tion (ADA) and World Health Organizationdiagnostic criteria for diabetes mellitus areshown in Table 32-2. The diagnosis of diabetesmellitus can be made on the basis of classicsymptoms, including polyuria (frequent urina-tion), polydipsia (frequent water drinking), andweight loss in addition to an unequivocal eleva-tion of fasting or postprandial plasma glucoseconcentrations. The vast majority of patientswith newly presenting type 1 diabetes willmeet these criteria (as in our illustrative case).In the absence of these classical findings (i.e.,when random plasma glucose values are lessthan 11.1 mmol/L) or in the absence of classicsymptoms, the most useful diagnostic tests are a

fasting plasma glucose or an oral glucose toler-ance test (OGTT) (ADA, 2004a).

Oral Glucose Tolerance Test

For the OGTT, subjects are given a syrupy bev-erage containing 75 g of glucose after an 8- to14-h fast. Plasma glucose is sampled prior to theglucose load and again 2 h after the glucoseload. Table 32-3 gives the guidelines for inter-preting OGTT results.

Glycated Hemoglobin andHemoglobin A1C

Glycated hemoglobin is formed continuously inerythrocytes as the product of a nonenzymaticreaction between hemoglobin (HgA) and glu-cose, first forming the labile Schiff base (or pre-HbA1C), then the more stable Amadori product(Fig. 32-1A). HbA1C specifically refers to theAmadori product of the N-terminal valine of eachβ-chain of HbA with glucose.The basic chemicalstructure of an advanced glycation end product,as shown in Figure 32-1B, demonstrates how

346 ENDOCRINOLOGY AND INTEGRATION OF METABOLISM

Table 32-1. Laboratory and RadiologicalEvaluation of Case Study*Capillary blood glucose in the emergency department:

“High”

Arterial blood gas:

pH = 7.15 (7.35–7.45)

pO2 = 84 mm Hg (83–108)

pCO2 = 24 mm Hg (32–48)

HCO3− = 14 mmol/L (22–28)

White cell count: 19,500/mm3

Sodium: 128 mmol/L (134–148)

Potassium: 5.0 mmol/L (3.5–5.0)

HCO3−: 15 mmol/L (22–28)

Chloride: 90 mmol/L (96–106)

Plasma glucose: 28.2 mmol/L (3.6–6.1)

BUN: 8.9 mmol/L (2.6–6.43)

Creatinine: 176.8 µmol/L (80–132)

Serum ketones: Strongly positive

HbA1C: 12.6% (4.4%–5.8%)

Urinanalysis: Glucose +5, ketones +3

Chest radiograph: Normal

*Numbers in parentheses indicate the normal range.

Table 32-2. Diagnostic Criteria for DiabetesMellitus1. Symptoms of diabetes plus a casual plasma glucose

concentration ≥11.1 mmol/L (200 mg/dL). Casual isdefined as any time of day without regard to the timesince the last meal.The classic symptoms of diabetesinclude polyuria, polydipsia, and unexplained weightloss. Or

2. Fasting plasma glucose ≥7.0 mmol/L (126 mg/dL).Fasting is defined as no caloric intake for at least 8 h.Or

3. Two-hour postload glucose ≥11.1 mmol/L(200 mg/dL) during a standardized oral glucosetolerance test (OGTT).The test should be performedusing a glucose load containing the equivalent of 75 gof anhydrous glucose dissolved in water.

In the absence of unequivocal hyperglycemia, thesecriteria should be confirmed by repeat testing on adifferent day. The third measure (OGTT) is notrecommended for routine clinical use.

Table 32-3. Guidelines for Interpreting the 75-gOral Glucose Tolerance Test (OGTT)• 2-h postload glucose < 7.8 mmol/L

(140 mg/dL) = normal glucose tolerance

• 2-h postload glucose 7.8–11.0 mmol/L(140–199 mg/dL) = impaired glucose tolerance

• 2-h postload glucose ≥11.1 mmol/L(200 mg/dL) = provisional diagnosis of diabetes

chronic exposure to hyperglycemia can resultin permanent chemical modifications to collagenand connective tissue.This process accounts formany of the end-stage complications of diabetes.

HbA1C makes up about 80% of the glycatedHbA1.The HbA1C fraction represents hemoglo-bin molecules that have undergone irreversibleglycation.The proportion of hemoglobin mole-cules that have undergone such a reaction isa direct reflection of the average prevailingblood glucose concentration during the life ofthe hemoglobin molecule (i.e., about 3months). Measuring HbA1C proves to be an ac-curate and reliable method for estimatingglycemic control over the preceding 2–3months. Current guidelines do not allow forHbA1C to be used as a diagnostic test for dia-betes,but it is an essential tool for assessing theeffectiveness of diabetes treatment. Currenttreatment guidelines recommend an HbA1C

level of 6. 5%–7.0% for patients with well-controlled diabetes (ADA, 2004b).

C-Peptide

The connecting peptide, a 31-amino acid chainbetween the α- and β-chain of the insulin mole-cule, is an excellent measure of endogenous in-sulin secretion in healthy individuals. However,C-peptide concentrations are difficult to inter-pret in the setting of new diabetes because ofconsiderable overlap between normal indivi-duals and those with type 1 or type 2 diabetes,depending on the duration and degree of meta-bolic control of the disease. The vast majorityof children with type 1 diabetes do not pro-duce any detectable C-peptide 2 years after di-agnosis. A diagram showing how C-peptide isderived from proinsulin during the productionof insulin is shown in Figure 32-2.

Antibodies

Antibodies to islet cells, insulin, and glutamicacid decarboxylase are present in many pa-tients with newly diagnosed type 1 diabetes. In

Type 1 Diabetes Mellitus 347

Figure 32-1. (A) Schematic representation of the nonenzymatic glycation of proteins,including hemoglobin, resulting in glycated HbA1C. (B) Chemical structure of advancedglycation end products in tissues exposed to chronic hyperglycemia.The R groups des-ignate tissue proteins that have become cross-linked and frequently become dysfunc-tional as a result of this process.Adapted from Brownlee (1992).

combination, they can be very sensitive andspecific in predicting the risk of disease, butthey are currently limited only to research ap-plications and are not routinely used in clinicalpractice.


Glucose Homeostasis

An understanding of the symptoms and bio-chemical results in the illustrative case requiresan overview of the homeostatic mechanismsregulating blood glucose concentration. Plasmaglucose concentrations are tightly controlled bya balance between the actions of hormones, en-zymes, and substrates that either raise or lowerblood glucose levels. Glucose homeostasis de-pends on a balance between glucose produc-

tion by the liver and glucose utilization by bothinsulin-dependent tissues (mainly fat and mus-cle) and insulin-independent tissues (such asbrain and kidney). In normal individuals, thisbalance helps to keep the blood glucose con-centration in a narrow range between 3.9 and6.1 mmol/L.

Glucose is the primary fuel for the centralnervous system (CNS).A constant supply of glu-cose at sufficient levels is critical for normalfunctioning of the CNS. On the other hand, theprevention of hyperglycemia is important toavoid loss of calories that would occur whenexcess glucose is spilled into the urine (the glu-cose concentration at which proximal tubularglucose transporters are saturated and excessglucose begins to be spilled into the urine isabout 10 mmol/L) and excess glycation of pro-teins.

Plasma glucose concentrations are deter-


Figure 32-2. A schematic representation of the production of insulin and C-peptidefrom proinsulin in the pancreatic β-cell. C-peptide acts as an indicator of endogenous in-sulin secretion, even in people who inject exogenous insulin.

mined by a net balance between glucose re-leased into the circulation and glucose taken upfrom plasma. Normal regulation of glucose lev-els depends largely on three factors: (1) the abil-ity of the pancreatic β-cell to secrete insulinboth acutely and in a sustained fashion; (2) theability of insulin to inhibit hepatic glucose pro-duction and promote glucose utilization in mus-cle and other tissues; and (3) the ability ofglucose to enter cells in the absence of insulin(glucose effectiveness).

In the fasting state, approximately 80% ofglucose uptake by tissues is non-insulin medi-ated, with the CNS the main consumer of glu-cose. In the setting of reduced insulin levels,pyruvate dehydrogenase activity in muscle isdecreased, thus limiting glucose oxidation. Atthe same time, lack of insulin promotes catabo-lism of muscle protein and the release of ala-nine and glutamine.The alanine is taken up bythe liver and provides a major substrate for glu-coneogenesis, thus ensuring a constant supply

of glucose to the brain during periods of caloricdeprivation (Figure 32-3).The glutamine is me-tabolized by the gut and kidney, and in the lat-ter organ, also serves to provide ammonium ionsto buffer the metabolic acids produced duringketogenesis.

The roles of hepatic glycogen metabolismand gluconeogenesis in the regulation of bloodglucose are best illustrated by a description ofnormal glucose homeostasis in fed and fastedstates. It can be divided into five phases, as de-picted in Figure 32-4. Glucose utilization is plot-ted against time in a person who ingests 100 gof glucose and then fasts for 40 days (Ruder-man, 1975).

• Phase 1, Absorptive Phase: For 3–4 h afterglucose ingestion, blood glucose is derivedprincipally from exogenous carbohydrate.Concentrations of insulin and glucose areincreased, and glucagon is depressed. Ex-cess glucose is stored in liver and muscle


Figure 32-3. Schematic representation of fuel mobilization during fasting. Catabolismof muscle proteins provides alanine for gluconeogenesis and glutamine for utilization bythe gut and kidney, while branched chain amino acids are primarily oxidized within themuscle. Breakdown of adipocyte triacylglycerols provides glycerol and free fatty acids(not shown); the free fatty acids provide fuel for liver, muscle and most other peripheraltissues. The liver utilizes both alanine and glycerol to synthesize glucose which is re-quired for the brain and for red blood cells (not shown). Adapted from Besser andThirner (2002).

as glycogen or is converted to lipid.The ab-sorptive phase is the only phase in whichthe liver is a net glucose sink.

• Phase 2, Postabsorptive Phase: By approxi-mately 4 h after feeding, insulin and glu-cose return to their basal states and theliver produces glucose from stored glyco-gen.The brain is the major user of glucoseduring the postabsorptive phase, whilemuscle and adipose tissue use glucose ata reduced rate.There is an increase in therelease of alanine which is used as sub-strates for gluconeogenesis.

• Phases 3 and 4, After 12–14 h of starvation,the ability of the liver to carry out gluco-neogenesis is enhanced secondary to a de-crease in insulin and an increase inglucagon. In addition, the release of aminoacids from muscle continues to be in-creased. At this point, liver glycogen is de-pleted,but the brain is not yet using ketonebodies, and the demand for gluconeogene-sis is at its peak. Thus, this is the time ofgreatest susceptibility for hypoglycemia be-cause of impaired gluconeogenesis.

• Phase 5, Prolonged Starvation: In the laterpart of phase 4 and in phase five, plasmalevels of ketone bodies increase and ke-tone bodies partially replace glucose as afuel supply for the brain.This, in turn, de-creases the demand for hepatic gluco-neogenesis and acts to conserve muscleprotein.

Fatty Acid and Ketone BodyMetabolism

Most peripheral tissues can use ketone bodies(acetoacetate and β-hydroxybutyrate) as wellas nonesterified fatty acids and glucose formetabolic fuel. Ketone bodies are produced bythe liver during gluconeogenesis. They are animportant fuel for the brain during starvation,and are thus crucial for protein conservationduring prolonged fasting. A near-total defi-ciency of insulin, such as occurs in untreatedtype 1 diabetes, can, however, yield danger-ously high concentrations of ketone bodiessecondary to an imbalance between their he-patic production and peripheral utilization.This pathologic ketosis causes severe meta-bolic acidosis due to the marked excess of theweak organic acids, acetoacetate, and β-hydroxybutyrate, which can be fatal if nottreated promptly.

A schematic depiction of the intracellularpathway for β-oxidation of long-chain fattyacids and, ultimately, ketogenesis is shown inFigure 32-5. Long-chain fatty acids are activatedprior to entering the mitochondria by convert-ing them to acyl-CoA derivatives. Because themitochondrial membrane is impermeable toCoA and its derivatives, a specific transportsystem is required.As shown in Figure 32-5, thissystem has three components: (1) the enzymecarnitine acyltransferase I (CAT I), which trans-fers the activated acyl unit from fatty acyl-CoA


Figure 32-4. Phases of glucose homeostasis in a normal human during a prolongedfast. Adapted from Ruderman et al. (1975).

Figure 32-5. β-oxidation and ketogenesis in the liver.The rate-limiting step in fatty acidoxidation and subsequent ketone body production is the activity of carnitine acyltrans-ferase I (CAT I).The activity of CAT I is inhibited by malonyl-CoA. Insulin deficiency re-sults in inhibition of acetyl-CoA carboxylase, decreased levels of maloyl-CoA, and thusincreased activity of CAT-I.Adapted from Foster and McGarry (1983).

351

to carnitine; (2) a translocase system for facilitat-ing the exchange diffusion of fatty acyl-carnitineinto the mitochondrial matrix in exchange forcarnitine; and (3) a second transferase called car-nitine acyltransferase II (CAT II), which catalyzesthe transfer of the fatty acyl unit from fatty acyl-carnitine back to CoA prior to β-oxidation.Thefatty acylation of carnitine by CPT 1 is the rate-limiting intracellular event of fatty acid oxidationand is inhibited by malonyl-CoA.

Most tissues oxidize the acetyl-CoA pro-duced during β-oxidation to CO2 and water viathe TCA cycle. During fasting, however, theliver utilizes the intermediates of the TCA cycleas gluconeogenic substrates. Under these con-ditions, the liver converts acetyl-CoA to ketonebodies (acetoacetate and β-hydroxybutyrate)(Figure 32-5). Most other peripheral tissues canoxidize ketone bodies by the pathway shownin the figure. After entering the mitochondria,acetoacetate reacts with succinyl-CoA to formacetoacetyl-CoA, a reaction that is catalyzedby 3-oxoacid-CoA transferase. Alternatively,acetoacetyl-CoA is formed by direct activationof acetoacetate by the enzyme acetoacetyl-CoA synthetase. Acetoacetyl-CoA is thencleaved to form two molecules of acetyl-CoAby acetoacetyl-CoA thiolase.As noted earlier in

the discussion of phases of glucose metabo-lism, ketone bodies are only utilized by thebrain during prolonged fasting and starvation.

Type 1 Diabetes Mellitus

Type 1 diabetes (formerly referred to as juvenile-onset diabetes mellitus and insulin-dependent di-abetes mellitus) results from destruction of betacells and a complete or near total absence of in-sulin synthesis. Insulin is the primary hormoneresponsible for regulating glucose metabolismand in signaling for the utilization and storage ofbasic nutrients. As shown in Table 32-4, insulinacts as a powerful anabolic hormone, and it isalso a potent inhibitor of the catabolic processesevoked by the counterregulatory hormones (i.e.,glucagon, epinephrine, cortisol, and growth hor-mone).Although the important target tissues forinsulin are liver, muscle, and fat, insulin haspleiotropic effects on cell growth and metabo-lism in many tissues (Kahn,2001).

The development of type 1 diabetes is theculmination of a chronic autoimmune destruc-tion of the pancreatic β-cells that occurs overmany years. This process results in severe, andultimately complete, insulin deficiency. In theabsence of insulin, fasting hyperglycemia is


Table 32-4. Physiological Consequences of the Action of InsulinVariable Action Tissue

Carbohydrate Metabolism

Glucose transport Increase Muscle, adipose tissue

Glycolysis Increase Muscle, adipose tissue

Glycogen synthesis Increase Liver, muscle, adipose tissue

Glycogen degradation Decrease Liver, muscle, adipose tissue

Gluconeogenesis Decrease Liver & kidney

Lipid Metabolism

Lipolysis Decrease Adipose tissue

Synthesis of fatty acids and triglycerides Increase Liver, adipose tissue

Synthesis of very low density lipoprotein Increase Liver

Lipoprotein lipase activity Increase Adipose tissue

Fatty acid oxidation Decrease Muscle, liver

Cholesterol formation Increase Liver

Protein Metabolism

Amino acid transport Increase Muscle, liver, adipose tissue

Protein synthesis Increase Muscle, liver, adipose tissue

Protein degradation Decrease Muscle

Urea synthesis Decrease Liver

primarily due to an unrestricted increase in he-patic glucose production. Gluconeogenic pre-cursors are elevated to supply the necessarysubstrate to support gluconeogenesis. In addi-tion, insulin deficiency results in the unre-strained lipolysis and increased ketogenesis thatleads to diabetic ketoacidosis.

Type 1 diabetes accounts for approximately10% of all patients diagnosed with diabetes mel-litus. It is a major chronic disease of childrenand is now being recognized with increasingfrequency in adults. In the absence of insulin,the resulting metabolic derangements in acutediabetic ketoacidosis eventually lead to comaand death.

Approximately 90% of diabetics have type 2 di-abetes mellitus rather than type 1. Type 2 dia-betes (previously called maturity-onset diabetes),is characterized by insulin resistance. These pa-tients initially exhibit impaired glucose uptakeinto tissues and a compensatory increase in in-sulin secretion.Although type 2 diabetes usuallyoccurs in people over 40 years of age, its inci-dence has been increasing markedly in youngerindividuals over the past decade.Type 2 diabetesis often accompanied by hypertension and dys-lipidemia (abnormalities in blood lipoproteins)and most of these patients are obese. Long termcompensatory increase in insulin secretion fre-quently leads to pancreatic failure and most pa-tients with type 2 diabetes eventually requireinsulin. Certain ethnic groups currently exhibithigher rates of type 2 diabetes than the generalpopulation. For example, half of the adult PimaNative Americans in the U.S. Southwest are nowdiabetic.Overall, the rate of diabetes for AmericanIndians and Alaska Natives is more than twice therate for the U.S.population as a whole.

The typical pancreatic lesion of type 1 dia-betes is the selective loss of almost all β-cells,whereas other islet cell types (α, δ, and pancre-atic polypeptide cells) remain intact. The mostcommon mechanism for β-cell destruction isthought to be autoimmune-mediated inflamma-tory damage. Prospective family studies stronglysupport a genetic basis for susceptibility to thisautoimmune process and suggest that the under-lying immune abnormalities precede clinical in-sulin deficiency by many years. However, not allspontaneous type 1 diabetes is the result of au-toimmune mechanisms.

It has been long-recognized that heredity is amajor factor in diabetes. Identical twins whoshare all the same genes have a much greaterrisk of diabetes than fraternal twins who may

share only 50% of the same genes: concordancein identical twins is about 25–50% for type 1 dia-betes. The lack of complete concordancestrongly suggests that environmental factorsalso play a role in the development of clinical di-abetes.Environmental factors that have been im-plicated include certain foods (including cow’smilk), common viruses, and vaccines, but thereis little evidence for any specific association.Theexception is exposure to wild-type rubella virusduring the first trimester of pregnancy.As manyas 20% of the children born after prenatal expo-sure to rubella later develop type 1 diabetes.

Islet Cell Antibodies and OtherImmunological Markers

Circulating IgG antibodies specific for islet cellantigens are found in 70%–80% of individualswith type 1 diabetes at the time of diagnosis.These antibodies do not appear to play a rolein β-cell destruction, but they serve as usefulmarkers for immunological autoreactivity.Some of the specific targets for these anti-bodies include insulin and glutamic acid decar-boxylase. Islet cell antibodies are detectablefrom infancy or early childhood in 3%–8% ofthe first-degree relatives of patients with type1 diabetes. About of half of these individualswill eventually develop clinical diabetes. Thelifetime risk of developing type 1 diabetes inthese individuals is related to the titer of isletcell antibodies, but antibody screening is cur-rently used only in the research settings to de-fine risk among the relatives of affectedindividuals.

Two rodent models of genetically deter-mined autoimmune β-cell destruction and dia-betes are available to provide some insight tothe human disease.The disease process has sev-eral discrete steps, each of which may be sub-ject to genetic or pharmacological control inthe future.The process is thought to be initiatedby release of antigens from the β-cells. Thiscould be the result of genetic defects, infectiousagents, or β-cell toxins.These antigens are thenpresented to the immune system, resulting inlymphocytic activation. Normal mice react tothe antigen exposure with a self-limited im-mune response, but, in diabetes-prone rodents,this exposure results in self-perpetuating au-toimmune destruction. Once initiated, proin-flammatory cytokines, highly reactive oxygenradicals and nitric oxide are released, causingdirect cellular damage.


Clinical Manifestations

The clinical manifestations of type 1 diabetesare the end result of persistent hyperglycemia.Sustained glucose levels above the renalthreshold (approximately 10 mmol/L) leads toan osmotic diuresis and polyuria, which inturn leads to dehydration and a hyperosmolarstate.Weight loss is universally present due toa combination of fluid loss and catabolism ofmuscle and fat stores. Excessive production ofketoacids consumes buffer (i.e., HCO3

−) andlowers the serum pH.The combination of en-ergy deprivation, dehydration, and sleep depri-vation due to nocturia leads to fatigue andmalaise. Vision is often affected, with patientscomplaining of blurry vision or inability to fo-cus.This sign is attributable to the deposition ofexcess glucose and sorbitol in the lens, whichresults in osmotic swelling and a distortion oflight refraction. Sorbitol arises from the actionof nearly ubiquitous enzymes, known as aldosereductases, that have a low affinity for glucosebut that convert glucose to sorbitol when glu-cose concentrations are elevated. These symp-toms eventually resolve with treatment.

At diagnosis, new patients with type 1 dia-betes may exhibit ketoacidosis with resultantacid-base and electrolyte disturbances. Themost common abnormality is hyponatremia,which is often due to the movement of waterto the extravascular space. Volume contrac-tion may lead to elevations in blood urea nitro-gen (BUN) and creatinine, as well as milderythrocytosis. Leukocytosis may also exist inthe absence of infection, and serum triglyc-erides and urine glucose are almost univer-sally elevated.

Newly diagnosed patients with type 1 dia-betes may present with acute metabolic decom-pensations, ketone production, and metabolicacidosis, a condition known as diabetic ketoaci-dosis (DKA).Although β-cell destruction occursgradually, acute physical or emotional stress canacutely create a demand for increased insulinproduction.

The pathophysiology of DKA is as follows. Se-vere insulin deficiency leads to hyperglycemiaand hyperosmolality. Unopposed glucagon ac-tion then leads to accelerated glycogenolysis,gluconeogenesis, and lipolysis (Figure 32-3).During gluconeogenesis, the acetyl-CoA pro-duced from during oxidation of fatty acids inthe liver is converted to acetoacetate and β-hydroxybutyrate. Ketones are weak organic

acids, and their accumulation in the serumleads to metabolic acidosis.The combination ofdehydration and acid-base abnormalities ulti-mately leads to severe electrolyte disturbances.

Most patients with DKA appear ill and weak.They are usually hypotensive and have poorskin turgor, indicating severe dehydration. If thepatient is able to give a history, symptoms ofpolyuria, polydipsia, and weight loss are invari-ably present. The breath may have a classic“fruity odor” due to the excretion of acetone inexpired breath.Acetone arises from the sponta-neous, nonenzymatic decarboxylation of ace-toacetate:

CH3-C-CH2-COOH → CH3-C-CH3 + CO2

Tachycardia is usually present, and breathingcan be deep and labored (Kussmaul breathing).Anorexia, nausea, and vomiting may also bepresent. Abdominal pain may be severe with-out underlying pathology. Patients are generallylethargic, but the level of consciousness canrange from alert to coma.

Diabetics are prone to many long-term com-plications. Chronic hyperglycemia (even if par-tially controlled) affects many different organsystems. Individuals with diabetes are at in-creased risk of heart attack, stroke and compli-cations related to poor circulation. Neuropathyis one of the most common complications of di-abetes. The combination of nerve damage andpoor blood flow can result in foot complicationsso severe as to lead to amputation. Diabetes canalso damage the kidneys (nephropathy), result-ing in kidney failure or a reduced ability tocarry out their normal filtration function. Thiscomplication develops in 20–30% of peoplewith type 2 diabetes, and in the U.S. accountsfor more than 50% of cases of end-stage renaldisease. Retinopathy (eye disease) is anothercomplication of diabetes. It is caused by nar-rowing, hardening, swelling, hemorrhaging orsevering of the capillaries of the retina and canlead to blindness. Gastroparesis (delayed gastricemptying) is also common in individuals withboth type 1 and type 2 diabetes.

THERAPY

While insulin replacement will be the primarilyaspect of long-term management of this pa-tient, the acute problem is the diabetic keto-


O O��

acidosis. Correction of hyperglycemia is not theprimary therapeutic approach to the medicalmanagement of DKA, although such treatmentwill ultimately result in a reduction in osmoticdiuresis and dehydration. Fluid replacement,restoration of tissue perfusion, and correctionof electrolyte imbalance are the primary goalsof DKA management. The critical componentsin the medical management of DKA are summa-rized in Figure 32-6.

Fluid Replacement

The patient discussed in the illustrative casepresented with orthostatic hypotension, poorskin turgor, dry mucous membrane, a ketoticodor to the breath, elevations in BUN andcreatinine, and ketoacidosis. She had severe ex-tracellular volume depletion, which can be esti-mated using the following clinical criteria:

1. Loss of greater than 10% of extracellularfluid (ECF) volume results in an ortho-static increase in pulse without a changein blood pressure.

2. A 15%–20% loss of ECF volume manifestsas an orthostatic drop in blood pressure ofmore than 15 mm Hg systolic and morethan 10 mm Hg diastolic.This amounts toapproximately 3–4 L of fluid loss.

3. A greater than 20% loss of ECF volumecauses supine hypotension, usually afterlosing more than 4 L of fluid.

The intravenous administration of isotonicsaline (0.9% NaCl) should be used to restoreECF and intracellular fluid volumes. Infusion ofisotonic saline restores the extracellular vol-ume deficit. Isotonic saline also restore intra-cellular volume deficits in patients with DKAand hypotonicity. Aggressive hydration itself


Figure 32-6. Summary of the critical components in the medical management of diabeticketoacidosis.

causes reduction in counterregulatory hor-mone concentrations and in blood glucose, andthis remains the cornerstone of DKA therapy.Once the patient has been stabilized, rehydra-tion should be gradual over a period of36–48 h order to prevent cerebral edema. Anaccurate record of fluid input and outputshould be maintained to help assess volumestatus. Oral rehydration should be avoided untilthe patient is hemodynamically stable andvomiting has stopped (Foster and McGarry,1983;Waldhausl et al., 1979).

Insulin Administration andCorrection of Ketoacidosis

DKA is a positive anion gap metabolic acidosisassociated with the accumulation of β-hydroxy-butyrate and acetoacetate. Lactic acidosis sec-ondary to cardiac or renal failure, hypoxia, poortissue perfusion, shock, or sepsis may also con-tribute to the anion gap in DKA.A normal aniongap (AG) is 12 ± 2 mEq/L.The anion gap (AG) iscalculated using the following formula: AG =([Na+ + K+] − [Cl− + HCO3

−]). In our illustrativecase, the anion gap was 28, indicating severemetabolic acidosis.

As ketoacidosis is corrected, the anion gapwill normalize. Although fluid replenishmentimproves insulin sensitivity and promotestissue perfusion, simultaneous administrationof a low-dose insulin infusion is shown tocause a faster decline in ketonemia. Regular in-sulin should be initially infused at a rate of0.1 U/kg/h intravenously. Intravenous insulinboluses are not recommended in an effort toavoid hypokalemia, hypoglycemia, and cerebraledema.With a low-dose insulin infusion, steady-state insulin concentrations are achievedwithin 20 to 30 min. Insulin infusion shouldnot be terminated prematurely because offalling blood glucose levels approaching hypo-glycemia. Since hyperglycemia is correctedmore rapidly than ketoacidosis, dextroseshould be added to fluids once the blood glu-cose concentration is less than 13.9 mmol/L,continued insulin administration will helpclear the ketonemia, and dextrose infusion willhelp maintain plasma glucose concentrationsin the ideal range of 6.1–13.9 mmol/L. Oncethe ketoacidosis has resolved and the patient iseating, intravenous insulin can be safelyswitched to subcutaneous insulin (Wagneret al., 1999).

Correction of Electrolyte Imbalances

Sodium: When patients present with hyper-natremia and elevated serum osmolality, theyare suffering from severe fluid deficits. Depend-ing on the patient’s hemodynamic stability, fluidtherapy should generally be instituted as amoderate-to-slow intravenous infusion of 0.9%normal saline over a period of 48–72 h to avoidcerebral edema.Patients with evidence of circu-latory compromise will require more aggressivefluid resuscitation. Estimated plasma osmolalityand corrected serum sodium concentrationsare calculated using the following formulas:

Estimated plasma osmolality= 2(Measured Na+ mEq/L)

+ (Glucose mg/dL/18)+ (BUN mg/dL/2.8)

Normal serum osmolality= 280–290 mOsm/kg H2O.

Corrected serum sodium= Measured Na+

+ [(Glucose in mg/dL− 100)/100] × 1.6

Normal serum sodium = 135–145 mEq/L

Using these formulas, our illustrative case hadan estimated serum osmolality of (256 + 28.4 +8.9 = 293.3 mOsm/kg H2O) and a correctedserum sodium of (128 + [4.12 × 1.6] = 135meq/L), suggesting that the losses of fluid wereisotonic.

Potassium: Hyperglycemia shifts water andpotassium from the intracellular to the extra-cellular compartment, and this shift is aug-mented by acidosis and protein breakdown.Loss of potassium occurs during osmotic diure-sis and is amplified by the secondary hyperal-dosteronism resulting from volume contraction.Aldosterone,a steroid hormone produced by theadrenal cortex, regulates fluid volume by caus-ing the renal retention of sodium (and the con-sequent excretion of potassium) as a result ofincreased angiotensin II availability. Most DKAsubjects actually have a total body potassiumdeficit of 500–700 meq/L. Once fluid replace-ment and insulin infusions have commenced,there is a shift of potassium back into the intra-cellular compartment, resulting in a rapid fall inserum potassium. Intravenous potassium ad-ministration should usually begin with insulin


infusions but should not exceed 40 mEq/L inthe first hour and 20–30 meq/L/h thereafter.Forrare patients who present with hypokalemia(<3.3 mEq/L), insulin administration should bewithheld until the potassium level is greaterthan 3.3 mEq/L (Kitabchi et al., 2001; White,2003).

Phosphate:An intracellular shift of phosphateoccurs along with potassium as fluid rehydra-tion commences.The phosphate deficit can alsobe worsened with correction of the metabolicacidosis. Controlled, randomized studies haveshown that routine phosphate repletion is notnecessary, but some practitioners think it pru-dent to provide supplemental phosphate ifserum phosphate levels are less than 1 mEq/L,potentially reducing the risk of seizure or tissueischemia. During intravenous phosphate admin-istration, serum calcium concentrations shouldbe monitored carefully to avoid hypocalcemiaand tetany (Fisher and Kitabchi, 1983).

Bicarbonate: Current guidelines state thatacidosis need not be corrected using exoge-nous bicarbonate therapy. Intravenous adminis-tration of fluids and insulin are sufficient tocorrect the metabolic acidosis and to regener-ate bicarbonate. Some practitioners believe,however, that bicarbonate repletion should beprovided in the setting of severe acidosis withpH less than 6.9.

Management of Type 1 DiabetesMellitus

Effective management of type 1 diabetes in-volves a team approach in which physicianscollaborate with certified diabetes educators,nutritionists, and other health care profession-als (such as psychologists and social workers)to achieve the therapeutic goals of acceptableglycemic control and prevention of complica-tions. The requirements for effective manage-ment of type 1 diabetes are shown in Table32-5.Apart from diet and exercise, instituting in-tensive diabetes therapy using either multipledaily insulin injections or continuous subcuta-neous insulin infusion (insulin pump) therapymay help in achieving these goals.

Human recombinant insulins and their ana-logues are classified as long- or short-actinginsulins based on their pharmocokinetic charac-teristics.The pharmocokinetic properties of thecommon insulin preparations are given in Table32-6. Possible approaches include (1) injection

of rapid-acting insulins at mealtime combinedwith intermediate-acting insulin at breakfast anddinner or at bedtime, (2) injection of rapid-acting insulin with meals combined with a long-acting insulin once a day, or (3) continuoussubcutaneous insulin infusion administration us-ing an external insulin pump.All of these insulinregimens depend on routine home blood glu-cose monitoring and the application of rationalinsulin algorithms for effectiveness.

The Diabetes Control and ComplicationsTrial (DCCT) has shown that intensified dia-betes therapy resulted in HbA1C reductions ofapproximately 2% compared to conventionaltherapy, and that patients practicing intensifieddiabetes therapy enjoyed significant reductionsin the feared microvascular complications of di-abetes (DCCT Research Group, 1993).As such,the current aim of diabetes therapy is toachieve glycemic control that is as close to nor-mal as possible while maintaining an accept-able quality of life for those patients for whomthe therapy is prescribed.The DCCT also demon-


Table 32-5. Requirements for Intensive InsulinTherapy

Patient motivation and interest

Diet compliance and carbohydrate counting

Frequent home blood glucose monitoring (generallyfour to six times per day)

Establishment of target blood glucose levels

Management of hypo- and hyperglycemia

Regular contact and support from diabetes team,including physician, diabetes educator, and dietitian

Family and social support

Table 32-6. Insulin Preparations and TheirPharmocokinetic PropertiesPreparation Onset Peak Duration

Rapid ActingRegular 0.5–1 h 2–4 h 6–8 hInsulin lispro 15 min 1 h 3–4 hInsulin aspart 15 min 1 h 3–4 h

Intermediate actingNPH 1–3 h 6–8 h 12–16 hLente 1–4 h 6–10 h 14–18 h

Long actingUltralente 1 h 8–10 h 16–24 hGlargine 1–2 h No peak 24 h

NPH, neutral protamine Hagedon.

strated, however, that patients using intensifieddiabetes therapy experienced hypoglycemia ata rate that was three times greater than patientsemploying conventional therapy. Thus, hypo-glycemia is the single most important limitingfactor in the glycemic management of type 1 di-abetes.

Prevention of Diabetic Ketoacidosis

Individuals with type 1 diabetes are also at riskof subsequent ketoacidotic crises. One com-mon cause of DKA is noncompliance with in-sulin therapy. Infections, (including subclinicalinfections) and physiological stressors (such asmyocardial infarction) can also precipitateDKA. Since the mortality rate associated withDKA can be as high as 10%, successfully educat-ing patients about the early warning signs ofthe condition and providing a plan for early in-tervention is crucial to its effective prevention.

Case Resolution

At presentation, the patient was hemodynami-cally unstable. She had a rapid heart rate and ex-hibited orthostatic hypotension (a fall in bloodpressure on assuming an upright posture) andslow mentation. Her fluid deficit was greaterthan 4 L. Her lab work revealed hyponatremia(low serum sodium), hyperkalemia (high serumpotassium), and a severe anion gap metabolicacidosis with dehydration.Her anion gap was 28,and her corrected serum Na+ was 135 mEq/L.

The patient was treated with 0.9% NaCl at arate of 1L/h. Orders were given to sample forserum electrolytes, glucose, BUN, Cr, and pHevery 2 h initially and then every 4 h thereafter.Intravenous insulin was started at the rate of0.1 µ/kg/h, and 20 mEq of KCl were added toeach liter of intravenous fluid.The rate was ad-justed so that plasma glucose declined at a rateof 2.8–3.9 mmol/L/h. Once the patient’s bloodglucose was less than 13.9 mmol/L, she wasswitched to intravenous fluids containing 5%dextrose with 0.45% NaCl to maintain herblood glucose concentration between 8.3 and11.1 mmol/L. Blood and urine cultures werealso obtained but failed to reveal an underlyinginfection.

Once the patient was clinically stable andthe acidosis had resolved, the insulin infusionwas changed to subcutaneous insulin, and shewas discharged home after 4 days in the hospi-tal and plenty of diabetes home care training.

She is currently doing well with a HbA1C of6.9% on a regimen of insulin glargine at bed-time and insulin lispro at meals, and she is be-ginning to develop an interest in obtaining aninsulin pump. She looks forward to a long andhealthy life with type 1 diabetes.

QUESTIONS

1. Why was the patient tachycardic?2. What is the main reason for administering

insulin in the treatment of diabetic keto-acidosis (DKA)?

3. List three serious adverse events that mayresult from the over zealous administra-tion of insulin during the treatment ofDKA?

4. Why is volume resuscitation the mainstayof DKA therapy?

5. What is the best way to avoid life-threatening hypokalemia during the treat-ment of DKA?

6. What is the single most important factorfor this patient’s health as she attempts toavoid diabetes-related organ damage overthe decades to come?

7. What would the blood levels of glucoseand ketones be after 48 h of fasting in aperson who was carnitine deficient? Ex-plain your reasoning.

8. Name another disease of sugar metabo-lism, aside from diabetes mellitus, that in-volves a glycation reaction as a componentof the pathophysiology.

BIBLIOGRAPHY

American Diabetes Association: Clinical practicerecommendations. Diagnosis and classificationof diabetes mellitus. Diabetes Care 27:S5–S10,2004a.

American Diabetes Association: Clinical practicerecommendations. Standards of medical care indiabetes. Diabetes Care 27:S15–S35, 2004b.

Besser GM, Thorner MO (eds): ComprehensiveClinical Endocrinology. 3rd ed.Mosby,St. Louis,MO, 2002.

Brownlee M: Glycation products and the patho-genesis of diabetes complications. DiabetesCare 15:1835–1843, 1992.

DCCT Research Group: The effect of intensivetreatment of diabetes on the development andprogression of long-term complications ininsulin-dependent diabetes mellitus. N Engl JMed 329:977–986, 1993.


Eisenbarth GS: Classification, diagnostic tests, andpathogenesis of Type I Diabetes Mellitus, inBecker KL (ed): Principles and Practice of En-docrinology and Metabolism. 3rd ed. Lippin-cott Williams and Wilkins, 2001, pp. 1307–1314.

Fein IA, Rackow EC, Sprung CL, Grodman R: Rela-tion of colloid osmotic pressure to arterial hy-poxemia and cerebral edema during crystalloidvolume loading of patients with diabetic ke-toacidosis. Ann Intern Med 96:570–575, 1982.

Fisher JN, Kitabchi AE: A randomized study ofphosphate therapy in the treatment of diabeticketoacidosis. J Clin Endocrinol Metab 57:177–180, 1983.

Foster DW, McGarry JD: The metabolic derange-ments and treatment of diabetic ketoacidosis.N Engl J Med 309:159–169, 1983.

Kahn CR: Glucose homeostasis and insulin action,in Becker KL (ed): Principles and Practice ofEndocrinology and Metabolism. 3rd ed. Lippin-cott Williams and Wilkins, 2001, pp. 1303–1306.

Kitabchi AE, Umpierrez GE, Murphy MB, et al.:Management of hyperglycemic crises in pa-

tients with diabetes (technical review). Dia-betes Care 24(suppl 1):131–153, 2004.

Morris LR, Murphy MB, Kitabchi AE: Bicarbonatetherapy in severe diabetic ketoacidosis. Ann In-tern Med 105:836–840, 1986.

Position Statement of the American Diabetes Asso-ciation. Hyperglycemic crises in diabetes. Dia-betes Care 27(suppl 1):S94–S102, 2004.

Ruderman NB,Aoki TT, Cahill GF: Gluconeogenesisand its disorders in man, in Hanson R, MehlmanM (eds): Gluconeogenesis. Wiley, New York,1975, p. 515.

Wagner A, Risse A, Brill HL, et al.: Therapy of se-vere diabetic ketoacidosis: zero-mortality undervery-low-dose insulin application. DiabetesCare 22:674–677, 1999.

Waldhausl W, Kleinberger G, Korn A, Dudcza R,Bratusch-Marrain P, Nowatny P: Severe hyper-glycemia: effects of rehydration on endocrinederangements and blood glucose concentra-tion. Diabetes 28:577–584, 1979.

White NH: Management of diabetic ketoacidosis.Endocrinol Metabol Dis 4:343–353, 2003.


CASE REPORTS

Patient 1

A 2-day-old neonate was referred to the pedi-atric endocrinology team with ambiguous ex-ternal genitalia, including clitoral hypertrophyand fusion of the labia.The baby had been deliv-ered spontaneously at 38 weeks of gestation af-ter an uneventful pregnancy and presentedwith no obvious clinical manifestations otherthan apparent masculinization of the externalgenitalia/genital ambiguity. A sample was ob-tained for genetic testing, which subsequentlyconfirmed a 46XX chromosomal makeup. Ex-amination of serum electrolytes indicated mildhyponatremia (decreased sodium concentra-tion) and hyperkalemia (elevated potassium).Ablood sample was also sent to a central labora-tory for hormone analyses.The results, obtained5 days after the patient was sent home, showedthat blood levels of cortisol (hydrocortisone)and aldosterone were well (<4 fold) below nor-mal, whereas levels of androgens, including de-hydroepiandrosterone (DHEA), DHEA-sulfate(DHEAS), androstenedione, and testosterone,were elevated more than 5-fold. Serum levels of17-hydroxyprogesterone (17-OHP) exceeded10,000 ng/mL (>100-fold above normal), whileestradiol and estrone were barely detectableand thus normal for age. Plasma levels of reninwere also elevated.

Four days later, the neonate was broughtby ambulance to the emergency room barelyconscious, severely dehydrated. and apparently

in hypovolemic (abnormally low plasma fluidvolume) shock, signs consistent with “adrenalcrisis.” Laboratory tests revealed severe hypo-natremia and hyperkalemia. Serum levels ofcortisol and aldosterone were well below nor-mal, while those of DHEAS, renin, and 17-hydroxyprogesterone were markedly elevated.The patient was diagnosed with classic (severe)21-hydroxylase (P450c21) deficiency congenitaladrenal hyperplasia (CAH). After restoration offluids and electrolytes, the patient was treatedwith cortisol (hydrocortisone) at doses that re-stored levels of DHEA, DHEAS, androstenedione,testosterone, and 17-OHP to within the normalrange.

Surgery was subsequently performed to cor-rect the labial fusion and clitoral hypertrophy.The child has developed normally and is beingclosely monitored for hormone replacementdose requirements to avoid either iatrogenicCushing’s syndrome from excess glucocorti-coid or premature epiphyseal closure and shortstature from unsuppressed adrenal androgenproduction.

Patient 2

The patient, a 7-year-old male, was brought tothe clinic by his parents, who were concernedabout apparent precocious puberty, includingonset of maturation of secondary sex character-istics (e.g., increased pubic hair, penile develop-ment). The boy’s height was greater thanexpected (51 inches, three standard deviations

33

Congenital Adrenal Hyperplasia:P450c21 Steroid Hydroxylase Deficiency

GERALD J. PEPE and MIRIAM D. ROSENTHAL

360

above mean for chronological age based onstandardized growth curves), and he appearedto be rather husky,consistent with increased so-matic maturation.

Blood pressure and other vital signs werenormal.Although serum electrolytes were alsonormal, blood levels of DHEA, DHEAS, an-drostenedione, and testosterone were elevatedthreefold, while cortisol levels were twofoldlower than normal. Serum levels of aldos-terone were normal and those of were reninonly slightly elevated.As in patient 1, however,the blood levels of 17-hydroxyprogesterone(>10,000 ng/mL) were extremely elevated com-pared to normal (<100 ng/mL). Blood levelsof leuteinizing hormone were low and wereonly slightly increased following injection ofgonadotropin-releasing hormone, thus indicatingthat the patient was truly prepubertal, and thatthe elevated androgens were not gonadal butmore likely adrenal in origin.The patient was di-agnosed with CAH due to classic P450c21 hy-droxylase deficiency and subsequently treatedwith exogenous glucocorticoid (hydrocorti-sone),which suppressed excess adrenal 17-OHPand androgen production.

DIAGNOSIS

Sexual differentiation is initiated very early infetal development (for review, George and Wil-son, 1994; Jost, 1972).Although genetic sex de-termines whether the fetal bipotential gonadwill develop into an ovary (46XX) or testes(46XY), sexual differentiation of internal ductsystems and external genitalia is markedly influ-enced by hormones (Fig. 33-1). In the male, pro-duction of testosterone by the fetal testes isrequired to induce development of the Wolffianduct into epididymis, ductus deferens, ejacula-tory ducts, and seminiferous tubules and differ-entiation of the primordia of the externalgenitalia as male. Differentiation of the primor-dia of the external genitalia as female and ofthe internal duct system (i.e., Müllerian ducts)into fallopian tubes, uterus, and cervix isthought to occur by a default mechanism andthus spontaneously without a requirement forsteroid hormones. In females, the Wolffian ductwill spontaneously regress in the absence ofhigh concentrations of testosterone, whereas inmales the production of anti-Müllerian hor-mone by the fetal testes induces regression ofthe Müllerian ducts. In genetic female fetuses,

therefore, exposure to elevated levels of andro-gens in utero can result in various degrees ofmasculinization of the primordia of the exter-nal genitalia as was noted in patient 1.

The most common cause of genital ambigu-ity is virilizing congenital adrenal hyperplasia(CAH),a group of autosomal recessive disorderscaused by deficiency of enzymes essential forsteroid hormone synthesis. It is now well estab-lished that 90%–95% of CAH cases are causedby a deficiency of the steroidogenic enzyme21-hydroxylase (P450c21, CYP21). First de-scribed in the mid-19th century by De Crec-chio (1865), our understanding of this diseasewas not forthcoming until the mid-20th cen-tury, when the recessive nature of the genetictrait and identification of the hormonal abnor-malities were recognized (Speiser and New,1987). The underlying defect among patientswith CAH due to 21-hydroxylase deficiency isfailure of the adrenal gland to synthesize corti-sol efficiently. Reduced blood levels of cortisollead to activation of the hypothalamic-pituitaryaxis and thus subsequent increased secretion ofcorticotropin-releasing hormone and corti-cotropin (ACTH) (Fig. 33-2).

In classic salt-wasting patients (such as patient1), deficiency of 21-hydroxylase also precludesadequate production of the physiologically im-portant mineralocorticoid aldosterone. Patientswith classic simple virilization (such as patient2) have a milder form of the enzyme deficiencyand are able to produce aldosterone in levels suf-ficient to maintain electrolyte balance. In bothforms of CAH, levels of 17-hydroxyprogesterone,the substrate for 21-hydroxylase, is markedly ele-vated. The adrenal responds to the buildup of17-OHP by increasing synthesis of steroid hor-mones that lack a 21-OH group, namely, theandrogens DHEA and DHEAS, which on secre-tion are further metabolized to the more activeandrogenic hormones testosterone and dihy-drotestosterone.The net effect of increased an-drogen production is thus prenatal virilizationof females and rapid somatic growth, with earlyclosure of epiphyseal plates in both sexes.

The incidence of classic CAH-P450c21 in theUnited States ranges from 1:10,000 to 1:18,000,with approximately 75% of patients with CAH-P450c21 exhibiting classic salt wasting (Whiteand Speiser, 2000). Aldosterone is essential fornormal sodium homeostasis and acts toenhance sodium absorption and potassium ex-cretion (Fig. 33-3). In the relative absence of al-dosterone, there is an increase in sodium loss

Congenital Adrenal Hyperplasia 361

via the kidney as well as other organ systems(colon, sweat glands), which can compromisemaintenance of adequate blood pressure(Bondy, 1985). Moreover, because salt-wastingpatients also have a more marked deficiency incortisol synthesis, they are more prone to crisesthat can be life-threatening. For example, in theabsence of glucocorticoids, cardiac output candecrease, leading to a decrease in glomerular fil-tration, inability to excrete water, and, conse-quently, hyponatremia. Salt-wasting CAH isoften more difficult to diagnose in males sincethe external genitalia are normal, and there is

usually no suspicion of adrenal insufficiencyprior to the initial presenting crisis.

In addition to the two classic forms ofP450c21 deficiency (salt wasting and simplevirilizing), there is another more commonform (occurrence 1:500 to 1:1000), which istermed nonclassic P450c21 CAH. This form ismuch less problematic physiologically, withpatients asymptomatic at birth but exhibitingvarious degrees of problems associated withandrogen excess later in life, including acne,hirsutism, oligomenorrhea, and perhaps infer-tility. In women, these symptoms and signs can


Figure 33-1. Summary of sexual differentiation: impact of genetic sex on differentiationof the bipotential gonad and role of the hormonal milieu on development of the primor-dia of the internal duct system and external genitalia in utero.

be confused with the more common polycysticovarian syndrome.

Other, more rare, forms of congenital adrenalhyperplasia are the result of deficiencies ofother steroidogenic enzymes, particularly17-hydroxylase (P450c17) and 11β-hydroxylase(Bondy, 1985;White and Speiser, 2000).As withP450c21 deficiency, impaired activity of eitherof these other enzymes also results in de-creased cortisol production,enhanced ACTH se-cretion, and thus increased tropic drive to theadrenal. Deficiency of 11β-hydroxylase is char-acterized by increased production of adrenalmineralocorticoids (e.g., corticosterone, aldos-terone), resulting in sodium retention, potas-sium excretion, and increased blood pressure.In addition, adrenal androgens are produced in

excess, and genital ambiguity is often seen in af-fected females.

By contrast, deficiency of 17-hydroxylase re-sults in impaired ability of the gonads (as wellas the adrenals) to synthesize androgens (males)or estrogen (females); patients subsequentlymanifest with problems typically associatedwith primary hypogonadism. Deficiency of 17-hydroxylase is similar to that of 11β-hydroxylasein that there is adrenal overproduction of min-eralocorticoids, resulting in sodium retentionand high blood pressure. Differential diagnosisis usually facilitated by examination of the com-plete profile of adrenocortical steroid hor-mones.


Steroid Biosynthesis

Steroid hormones are members of a group ofcompounds having in common the cyclopen-tanoperhydrophenanthrene nucleus, a hydratedfour-ring system in which the five-sided ring Dis attached to three six-sided rings of phenan-threne (Fig. 33-4).The carbon atoms of the ringstructure are numbered 1–17, while the addi-tional carbons atoms are 18–21. Compoundshaving 21 carbons are derivatives of pregnane,and their biological activity is progestational,glucocorticoid, or mineralocorticoid. Progestinssuch as progesterone are involved in the supportof pregnancy. Glucocorticoids provide the re-sponse to stress and are involved in carbohydrate


Figure 33-2. Hypothalamic-pituitary-adrenal axis.Corticotropin-releasing hormone (CRH) acts to in-crease pituitary ACTH secretion, which enhancessteroid hormone synthesis by the adrenal, includ-ing production of cortisol (hydrocortisone),which controls CRH/ACTH secretion via negativefeedback.

Figure 33-3. Summary of the renin-angiotensin-aldosterone system.Aldosterone secre-tion is controlled by several factors, including increased K+,ACTH, or angiotensin II.Al-dosterone acts to increase Na+ retention by both the kidney and colon.Aldosterone alsopromotes renal K+ excretion, which contributes to maintenance of Na+/K+ balance. Inthe absence of aldosterone, Na+ is lost, K+ is enhanced, the extracellular fluid volume isreduced, and mean arterial pressure and renal perfusion pressure are decreased.As a re-sult, renin secretion is increased, leading to increased formation of angiotensin II, whichpromotes vasoconstriction and aldosterone secretion.

metabolism by promoting gluconeogenesis.Mineralocorticoids are involved in regulation ofelectrolyte and water balance through their ef-fect on ion transport in the renal tubules. Com-pounds having 19 carbon atoms are derivativesof androstane and biological activity is andro-genic (producing masculine characteristics),whereas steroids with 18 carbons are deriva-tives of estrane and exhibit estrogenic (femalesex hormone) activity.

All steroid hormones are derived from cho-lesterol, a sterol molecule comprised of 27 car-bon atoms (Fig. 33-5). Cholesterol can besynthesized in the steroid-producing glands(adrenal, gonads) or brought to these glandsbound to low-density lipoprotein (LDL) follow-ing synthesis in the liver (see Chapter 14, thisvolume).To serve as a substrate, cholesterol inthe cytosol must be transported to the innermitochondrial membrane where the P450 cho-lesterol–side-chain cleavage enzyme complex(CYP11A, cholesterol desmolase, side-chaincleavage enzyme, P450scc) is located. Recentstudies indicated that steroid acute regulatoryprotein (StAR) transports cholesterol to the in-ner mitochondrial membrane (Lin et al., 1995;Tuckey et al., 2002). In the mitochondria, cho-lesterol is converted by P450scc to the 21-carbon molecule pregnenolone (Fig. 33-5),which is the common precursor for all othersteroid hormones.

Within the adrenal gland, steroids are synthe-sized by the outer layer of cortical cells, whilethe central core of medullary cells is innervatedby sympathetic preganglionic neurons, whichstimulate it to produce and release epineph-rine. The adrenal cortex is comprised of three

histologically and metabolically distinct zonesof cells, including an outer glomerulosa, anintermediate zone known as the fasciculata,and an internal layer termed the reticularis.In the adrenal, StAR is synthesized in responseto ACTH (zona fasciculata) or calcium (zonaglomerulosa).

Synthesis of cortisol (also known as 11,17,21-hydroxyprogesterone) occurs in the zona fasci-culata of the adrenal. Pregnenolone is firstconverted by the enzyme 17-hydroxylase to 17-hydroxypregnenolone; the latter is convertedby the enzyme 3β-hydroxysteroid dehydroge-nase (3βHSD) to 17-hydroxyprogesterone (17-OHP) which is then hydroxylated sequentiallyat carbon 21 (P450c21) and then carbon 11(P450c11) (Fig. 33-2). Since patients withP450c21 deficiency cannot utilize 17-OHP forcortisol synthesis, the levels of this metabolicintermediate increase dramatically.The inabilityto form cortisol results in increased release ofpituitary ACTH, increased synthesis/uptake ofcholesterol, and, consequently, continued pro-duction of 17-OHP (Figs. 33-2 and 33-5).

Aldosterone production is initiated by 21-hydroxylation of pregnenolone in the adrenalglomerulosa, where there is no 17-hydroxylaseactivity. The product of hydroxylation ofpregnenolone by P450c21 is 21-hydroxy-progesterone, also known as deoxycorticos-terone (DOC). The quantity of aldosteroneproduced by the adrenal is only 1% of that ofcortisol; for this reason, not all patients withclassic P450c21 deficiency manifest with saltwasting.

Androgens produced by the adrenal glandare synthesized in the inner zone of cortical


Figure 33-4. The four-ring cyclopentanoperhydrophenanthrene nucleus identifyingthe four rings and the numbering of the carbon atoms of a 21-carbon steroid such as cor-tisol.The ring carbons are numbered 1–17, while those of the side chain are 18–21.

Figure 33-5. Summary of the pathway and enzymes catalyzing steroid hormone synthesis from precursor choles-terol. Reproduced with permission from White and Speiser (2000). Copyright 2000,The Endocrine Society

cells, known as the reticularis, which lack3βHSD. In these cells, 17-hydroxypregnenoloneis converted to DHEA by P450c17, an enzymethat, in the reticularis, also exhibits a lyase func-tion which produces a C19 steroid by cleavageof two carbons. Thus, in patients with classicP450c21 deficiency and increased ACTH driveto the adrenal, levels of DHEA (and its sulfateDHEAS) are increased. Importantly, the C19 ste-roid DHEA can be hydroxylated by 3βHSD inother regions of the adrenal to more active an-drogens, including androstenedione. In the go-nads, androstenedione can be converted totestosterone by 17-hydroxysteroid dehydroge-nase and then to the very active androgendihydrotestosterone by the enzyme 5α-reduc-tase. The presence of such active androgens,normally not secreted in a female (genetic sex46XX) fetus, during the period when the exter-nal genitalia are developing is responsible forthe masculinization observed in patients withclassic P450c21 deficiency.

P450c21 Hydroxylase

The human P450c21 protein is composed of494 amino acid residues and exhibits a molecu-lar mass of 52 kd (Higashi et al., 1986). It is amember of the cytochrome P450 (CYP) super-family of proteins, all of which are monooxyge-nases in that they incorporate one of two atomsof molecular oxygen into the lipophilic sub-strate.The name P450 is derived from the ab-sorbance maximum at approximately 450 nmof the reduced form of the heme prostheticgroup. In addition to characterizing many ofthe reactions of steroidogenesis, mammaliancytochrome P450 enzymes are involved in catal-ysis of numerous other reactions, including oxi-dation of xenobiotics (exogenous substrates)such as drugs. Some cytochrome P450 en-zymes, including P450scc and 17-hydroxylase,are mitochondrial enzymes; others, includingP450c21 and the 17α-hydroxylase/17,20 lyase(P450c17), are localized on the endoplasmicreticulum.

The P450c21 gene (CYP21) is located in theHLA major histocompatibility complex on theshort arm of chromosome 6 (6p21.3) and iscomprised of 10 exons spaced over 3.1 kb (Car-roll et al., 1985). CAH due to P450c21 defi-ciency is inherited as a monogenic autosomalrecessive trait closely linked to the HLA com-plex. Thus, siblings with 21-hydroxylase defi-ciency are almost always HLA identical. Indeed,

before cloning of CYP21, HLA typing was a ma-jor means to perform prenatal diagnosis.

The 6p21.3 region of chromosome 6 alsocontains a pseudogene (CYP21P), a nonfunc-tional DNA sequence approximately 30 kbfrom CYP21P. The exon and intron nucleotidesequences of CYP21 and CYP21P are 98% and96% identical, respectively. Some 40–50 distinctmutations that cause 21-hydroxylase deficiencyhave been characterized (New and Wilson,1999;White and Speiser, 2000). More than 90%of them appear to be the result of two types ofmeiotic interactions between CYP21 andCYP21P. One type of interaction involves mis-alignment and unequal crossing over, resultingin large-scale DNA deletions.The other type ofinteraction, called gene conversion, results inthe transfer of deleterious mutations fromCYP21P to CYP21. CAH is unusual among gene-tic diseases in the high proportion of mutant al-leles generated by intergenic recombination(White and Speiser, 2000). Indeed, the high rateof de novo mutations in CYP21 may explain thehigh carrier frequency (1%–2%) of CAH despitethe fact that classic CAH is potentially lethal ifuntreated.

The various CAH phenotypes (salt wasting,simple virilizing, and nonclassic) represent aspectrum of severe to moderate to mild loss ofCYP21 enzymatic activity. Attempts have beenmade to correlate the genotype of the individ-ual (specific mutations) with this CAH pheno-type.As described by White and Speiser (2000),patients are usually compound heterozygotesfor two different mutations, and the phenotypeof each patient correlates quite well with thenature of the less severely impaired allele. Fac-tors other than 21-hydroxylase activity may alsoinfluence phenotype. Genetic variations in an-drogen biosynthesis or sensitivity may influ-ence clinical expression of androgen excess.Similarly, both genetic and nongenetic factorsmay contribute to the expression of salt wast-ing.

THERAPY

Glucocorticoid Replacement

Patients with CAH require lifelong gluco-corticoid replacement therapy to correct thedeficiency in cortisol secretion. This in turnsuppresses ACTH production, excess adrenal17-hydryoxyprogesterone, and excess androgenproduction. Hydrocortisone is the cortico-


steroid of choice for children with CAH be-cause of its short half-life and easy dose adjust-ment (Marshall and New, 2003). Olderindividuals may be treated with synthetic gluco-corticoids such as prednisone or dexametha-sone. Treatment efficacy is determined bymonitoring of 17-hydroxyprogesterone and an-drostenedione levels.

The therapeutic goal is to provide the mini-mum dosage that will effectively suppress ex-cess adrenal function. Excess glucocorticoidtreatment can be detrimental in children, result-ing in growth suppression. Excess glucocorti-coids can also result in Cushing’s syndrome,which is characterized by a cluster of symp-toms including so-called moon face, weightgain, central obesity, fatigue, muscle atrophy, os-teoporosis, increased thirst and urination,hyper-glycemia, and changes in mental status.

Stress Dosing and Medical Alerts

Patients with classic CAH do not adequately in-crease endogenous cortisol production in re-sponse to stress and thus require additionalexogenous glucocorticoid in a number of situa-tions, including febrile illnesses and surgery un-der general anesthesia. Increased requirementscan be 2–3 times normal dosage during physio-logical stress and as much as 5–10 times normaldosage for the first 24–48 h during and after asurgical procedure. Families should be given in-jection kits with age-appropriate dosages ofhydrocortisone for emergency use and trainedin their appropriate use. In addition, patientswith CAH, like others who are receiving cortico-steroid replacement therapy, should wear aMedic-Alert identifying bracelet to alert healthcare providers in an emergency situation.

Mineralocorticoid Replacement

Infants with salt-wasting CAH require miner-alocorticoid replacement therapy, usually withfludrocortisone (9α-fluorohydrocortisone).In addition, they require sodium chloride(1–2 g/day) since the sodium content of bothbreast milk and common infant formulas isonly sufficient to meet the requirements ofhealthy infants (White and Speiser, 2000).Older children often acquire a taste for saltyfoods and do not require daily sodium chloridetablets. Plasma renin activity may be used tomonitor mineralocorticoid and sodium re-placement.

Other Therapeutic Approaches

Several additional pharmacological approaches,currently still considered experimental, werereviewed by White and Speiser (2000). In addi-tion, these authors reviewed the correctivesurgical procedures that have been used fornewborns with ambiguous genitalia. They alsodiscussed the need for appropriate psychologi-cal counseling for families of genetic femalesborn with virilized external genitalia.

Prenatal Treatment and Diagnosis

In pregnancies at risk for a child affected withclassic P450c21 deficiency, prenatal hormonetherapy can be used to eliminate the virilizingeffects of excess androgen in utero (David andForest, 1984; Mercado et al., 1995).As reviewedby New and Wilson (1999), prenatal maternaltreatment with dexamethasone (a potent syn-thetic glucocorticoid) has been shown to pre-vent or significantly reduce virilization ofgenitalia in female neonates with CAH and pre-clude the need for genitoplasty.Dexamethasoneis used because it can cross the placenta and,unlike cortisol itself, is not metabolized to less-active forms by the enzyme 11β-hydroxysteroiddehydrogenase (Pepe and Albrecht, 1995).To beeffective, treatment must be initiated on or be-fore the ninth week of pregnancy.

Because of the need to initiate therapy earlyin fetal development, dexamethasone treatmentof at-risk pregnancies must be initiated prior toprenatal diagnosis. Amniocentesis cannot beperformed until after week 15; chorionic villussampling can be performed earlier (weeks10–12) but has a higher pregnancy loss rate(1%–2%). In either case, initial karyotyping isperformed; if the fetus is male, then the dexa-methasone treatment is discontinued (Mercadoet al., 1995). If the fetus is female, then DNAanalysis is required; hydroxyprogesterone lev-els will be uninformative since the dexametha-sone treatment suppresses amniotic fluidadrenocortical hormones. As reviewed byWhite and Speiser (2000), the molecular analy-sis is complicated by need to amplify PCR with-out amplifying the highly homologous CYP21Ppseudogene, which already carries most of themutations of interest.

Although studies to date have not identifieduntoward effects of the dexamethasone treat-ment on either the mother or the fetus, there is apotential for long-term/long-lasting effects of


glucocorticoid treatment (e.g.,programming fetalorgan system development/function) and thus in-creasing risk for development of diseases in adult-hood. For this reason, current protocols forprenatal therapy include prenatal diagnosis anddiscontinuation of treatment as soon as the fetuscan be identified as male or unaffected female.

QUESTIONS

1. Why do severe forms of CAH affect syn-thesis of both cortisol and aldosterone?

2. Why are males with classic P450c21 defi-ciency at more risk than females?

3. What is the mechanism by which impairedsynthesis of cortisol actually enhances syn-thesis of 17-hydroxyprogesterone (17-OHP)?

4. Why does a deficiency of 21-hydroxylaseresult in increased fetal synthesis of andro-gens?

5. Why is it important to initiate prenataltherapy for CAH early in pregnancy?

6. Why does excess hydrocortisone treat-ment increase blood glucose?

7. Why do patients with 21-hydroxylase defi-ciency exhibit low blood pressurewhereas those with 17α-hydroxylase defi-ciency exhibit high blood pressure?

8. Why are ACTH levels elevated in patientswith P450c21, P450c17, and P45011β ste-roid hydroxylase deficiency?

9. Why are deficiencies in adrenal steroidenzymes collectively termed congenitaladrenal hyperplasia?

BIBLIOGRAPHY

Bondy PK: Disorders of the adrenal cortex, in Wil-son JD,Foster DW (eds):Textbook of Endocrinol-ogy. Saunders,New York,1985,pp.816–890.

Carroll MC, Campbell RD, Porter RR: Mapping ofsteroid 21-hydroxylase genes adjacent to com-plement component C4 genes in HLA, the majorhistocompatibility complex in man. Proc NatlAcad Sci USA 82:521–525, 1985.

David M, Forest MG: Prenatal treatment of congen-ital adrenal hyperplasia resulting from 21-hydroxylase deficiency. J Pediatr 105:799–803,1984.

De Crecchio L: Sopra un caso di apparenz virile inuna donna Morgagni 7:154–188, 1865.

George FW,Wilson JD: Sex determination and dif-ferentiation, in Knobil E, Neill JD (eds): ThePhysiology of Reproduction. 2nd ed. RavenPress, New York, 1994, pp. 3–28.

Higashi Y,Yoshioka H,Yamane M, et al.: Completenucleotide sequence of two steroid 21-hydroxy-lase genes tandemly arranged in human chro-mosome: a pseudogene and a genuine gene.Proc Natl Acad Sci USA 83:2841–2845, 1986.

Jost A:A new look at the mechanisms controllingsex differentiation in mammals. Johns HopkinsMed J 130:38–53, 1972.

Lin D, Sugawara T, Strauss JF 3rd, et al.: Role of ste-roidogenic acute regulatory protein in adrenaland gonadal steroidogenesis. Science 267:1821–1831, 1995.

Marshall I, New MI: Adrenal disorders, in New M(ed.): Pediatric Endocrinology. Endotext.org,2003, Chapter 8.Available at: http://www.endo-text.com/pediatrics/pediatrics8/pediatrics-frame8.htm.

Mercado AB,Wilson RC, Cheng KC, et al.: Prenataltreatment and diagnosis of congenital adrenalhyperplasia owing to steroid 21-hydroxylase de-ficiency. J Clin Endocrinol Metab 80:2014–2020, 1995.

Migeon CJ: Editorial: comments about the need forprenatal treatment of congenital adrenal hyper-plasia due to 21 hydroxylase deficiency. J ClinEndocrinol Metab 70:836–837, 1990.

New MI, White PC: Genetic disorders of steroidhormone synthesis and metabolism. BaillieresClin Endocrinol Metab 9:525–554, 1995.

New MI,Wilson RC: Steroid disorders in children:congenital adrenal hyperplasia and apparentmineralocorticoid excess. Proc Natl Acad SciUSA 96:12790–12797, 1999.

Pepe GJ,Albrecht ED: Actions of placental and fe-tal adrenal steroid hormones in primate preg-nancy. Endocr Rev 16:608–648, 1995.

Speiser PW, New MI: Genotype and hormonal phe-notype in nonclassical 21-hydroxylase deficiency.J Clin Endocrinol Metab 64:86–91,1987.

Tuckey RC, Headlam MJ, Bose HS, Miller WL: Trans-fer of cholesterol between phospholipid vesi-cles mediated by the steroidogenic acuteregulatory protein (StAR). J Biol Chem 277:47123–47128, 2002.

White PC: Genetic diseases of steroid metabolism.Vitam Horm 49:131–195, 1994.

White PC, Speiser PW: Congenital adrenal hyper-plasia due to 21-hydroxylase deficiency. EndocrRev 21:245–291, 2000.


http://www.endotext.com/pediatrics/pediatrics8/pediatricsframe8.htm



Index

ABCA1, 162–165Abetalipoproteinemia, 298

biochemical perspective on, 293–298case report on, 290–291diagnosis of, 291–293, 291ttherapy for, 298

Acanthosis nigricans, 247Acid hydrolase, 173Acid phosphatase, 172Acute coronary pathophysiological process,

spectrum of, 62fAcute coronary syndrome, 55–63Adeno-associated viruses, 158Adrenal gland, 363f, 364Adrenal hyperplasia. See Congenital adrenal

hyperplasiaAdrenocorticotropic hormone (ACTH), 361, 363fAdrenoleukodystrophy. See X-linked adrenoleukody-

strophyAdrenomyeloneuropathy, 146–147ALD protein, 148–149Ammoniatelic organisms, 198Ankyrin, 70α1-Antitrypsin deficiency

biochemical perspectives on, 47–48, 49f, 50case report on, 42–44, 43fdiagnosis of, 44–47, 45ftherapy of, 50–52

Aplastic crises, 67–68Apolipoprotein B assembly, 292t, 293–296Arginase, 201Argininosuccinate lyase, 201Argininosuccinate synthetase, 200–201Aspartate-glutamate transporter, 201.

See also Citrin

Beta-carotene. See β-CaroteneBilirubin encephalopathy, 241Bilirubin metabolism, 236–238

genetic diseases of, 240hepatic, 238–240, 239fin intestine, 240

Bilirubin-UDP-glucuronyltransferase, 239, 240

Biliverdin reductase, 237–238Biopterin synthesis and recycling, 208–209, 210fBiotin cycle, 139, 140fBiotin holocarboxylase synthetase deficiency, 138Biotin supplementation, 142Biotin-dependent carboxylases, 136fBiotinidase deficiency

biochemical perspectives on, 139–141case report on, 134–135, 135fdiagnosis of, 135, 137–139therapy for, 142

Body mass index, 245, 246f, 247Bone marrow transplantation, 150Bone modeling and remodeling, 331

Calcium deficiency. See RicketsCalcium homeostasis, 329, 330fCalcium intake, 326Calcium receptor, 329Carbamoylphosphate synthetases, 200Carbon monoxide, 23

biology of, 236–237Cardiac biomarkers

clinical utilization of, 61–65immunoassay perspectives on monitoring, 57–61

Cardiac troponins, 57–64Carnitine acyltransferase I, 350, 351fCarnitine and fatty acid oxidation, 103–104, 103fCarnitine and HMG-CoA lyase deficiency, 223–224Carnitine deficiency, systemic

biochemical perspectives on, 103–105case report on, 101–102diagnosis of, 102–103therapy for, 105

Carnitine palmitoyl transferase reaction, 103, 104fβ-Carotene, 316fCeredase, 178Chitinase, 172Chitotriosidase, 172Cholestasis in infancy, 44Cholesterol transport, reverse, 163–164Cholesterol-lowering treatments, 157–158, 157f.

See also Lovastatin

369

Cirrhosis, 44Citrin, 195, 197. See also Aspartate-glutamate

transporter13C-mixed triglyceride breath test, 285–286, 286fCobalamin. See Vitamin B12

Coenzyme Q-10 (CoQ10), 98Collagen, 37, 37f, 331–332, 331fCollagen matrix synthesis pathway, 35–36, 36fCongenital adrenal hyperplasia

biochemical perspectives on, 363–364, 366case reports on, 360–361diagnosis of, 361–368therapy for, 366–368

Coronary artery disease, 165Corticotropin-releasing hormone, 363fCortisol, 259, 364C-peptide, 347, 348fCreatine kinase isoenzymes, 54–61, 64–65Crigler-Najjar syndrome types I and II, 240CYP21, 366Cystathionine-β-synthase, 227–232, 228fCysteine, 227, 228f, 262

D-biotin, 137fdfxr gene, 10Diabetes Control and Complications Trial, 358Diabetes mellitus, type 1, 345, 352–353

biochemical perspectives on, 348–354case report on, 345–346, 346tdiagnosis of, 346–348therapy for, 354–358

Diabetic mothers, effects on infants, 115, 116fDichloroacetate, 87, 98

E1α gene, mutations in, 82, 84–85Electron transport chain (Etc), 93, 94Emphysema, 44, 45

Familial hypercholesterolemiabiochemical perspectives on, 154–157case report on, 152–153diagnosis of, 153–154therapy for, 157–158, 157f

Fasting, metabolism of fuels during, 248, 250f,349, 349f

Fatty acid metabolism, 81, 113, 114f, 308–309, 311f,350–351

Fatty acid oxidation, 103–104, 103fFavism, 131Folate deficiency, 302f“Folate trap,”308, 310fFragile X mental retardation (FMR1) gene,

5, 6, 9–11, 9fFragile X syndrome, 7

case report on, 3–6, 4fdiagnosis of, 5–6, 12molecular perspectives on, 7–12therapy for, 14–15

Free radicals, 262initiation of, 340, 340f

Gallstones, 68Gaucher cells, 170Gaucher disease

biochemical perspectives on, 173–177case report on, 167–168, 169f, 170fdiagnosis of, 168, 170–173, 170ttherapy for, 177–179

Genetic anticipation, 12Genetic diseases associated with dynamic

mutations, 12, 13t, 14Globin gene clusters, 21–22, 22fGlucocerebrosidase, 173, 175–176, 178–179

N-linked oligosaccharides in, 173, 176, 177fGluconeogenesis, 221–222. See also HypoglycemiaGlucose homeostasis, 348–350, 350fGlucose 6-phosphate dehydrogenase deficiency

biochemical perspectives on, 126–129case reports on, 123–124diagnosis of, 124–126therapy for, 131–132

Glucose transporters, 109, 110Glucose-fatty acid cycle, 81Glutathione, 126–128Glycerol trierucate oil, 150Glycolipids, 184–185, 185fGrowth charts, CDC, 246f

HbA1C, glycated, 346–347, 347fHelicobacter pylori (H.pylori), 304–305Heme oxygenases, 236Hemochromatosis. See Hereditary

hemochromatosisHemoglobin genes, 21–22, 22fHemoglobin transition states, 23, 24fHemolytic anemia, 124

acute, 129–130congenital nonspherocytic, 130–131

Hemolytic crises, 67–68Hemolytic syndromes, 124Hemosiderosis, transfusion-induced, 28Hepatic stellate cells, 340, 341fHereditary hemochromatosis

biochemical perspectives on, 336–341case presentation of, 335diagnosis of, 335–336therapy of, 341–342

Hereditary spherocytosis, 66–67biochemical perspectives on, 68f, 69–73case report on, 66–68diagnosis of, 67–68, 67ftherapy for, 73–74

HFE, 337–340High-density lipoprotein (HDL), 162–166Hippuric acid, 202fHMG-CoA lyase deficiency, 217

biochemical perspectives on, 220–224case report on, 217–220diagnosis of, 220therapy for, 224–225

HMG-CoA reductase, 156–157

370 Index

Homocysteine. See HyperhomocysteinemiaHomocystinuria, 227Hydroxyurea, 27–28Hyperammonemia, 197, 222Hyperammonemic encephalopathy, 201Hyperbilirubinemia, neonatal, 130

biochemical perspectives on, 236–241case report on, 234diagnosis of, 234–236treatment of, 241–242

Hypercortisolemia, 259Hyperhomocysteinemia

biochemical perspective on, 227–232case report on, 226diagnosis of, 227therapy for, 232–233

Hyperphenylalanemia. See PhenylketonuriaHypoglycemia, neonatal

biochemical perspectives on, 109–117case report on, 107–108diagnosis of, 108–109therapy for, 117–118

Hypolactasia. See Lactose intoleranceHypophosphatemia, 331Hypothalamic-pituitary-adrenal axis, 363fHypothyroidism, 236

I-cell disease (mucolipidosis II)biochemical perspectives on, 186–191case history of, 181, 182fdiagnosis of, 181–186, 185ttherapy for, 191–192

Ileal brake, 281Immunodeficiency and malnutrition, 260Insulin. See Diabetes mellitus; HypoglycemiaInsulin growth factor II (IGF-II), 190–192Insulin preparations, pharmacokinetic

properties of, 357tIntrinsic factor, 303, 306Iron. See Hereditary hemochromatosisIslet cell antibodies, 353

Jaundice, neonatalconditions that may produce, 235trisk factors for, 241t

Kernicterus, 241Ketoacidosis. See Diabetes mellitusKetogenesis, 221–223, 351fKetone body metabolism, 350–351Kwashiorkor. See Protein-energy malnutrition

Lactase, 270Lactic acidosis. See Mitochondrial encephalomyopathy,

lactic acidosis, and strokelike episodesLactose intolerance

biochemical perspective on, 268–275case report on, 266diagnosis of, 266–269, 270ftherapy for, 275–276

Leucine, catabolism of, 220–224, 221f. See alsoHMG-CoA lyase deficiency

Lipoprotein classes, 154, 154tLipoprotein metabolism, 296, 297fLorenzo’s oil, 150Lovastatin, 150Low-density lipoprotein (LDL), 153–157, 155f, 165Low-density lipoprotein (LDL) receptor, 154–158Lundh test, 284Lyonization, 128

Malaria, P. falciparum, 21Malnutrition. See Protein-energy malnutritionMannose 6-phosphate receptors, 190–191Marasmus. See Protein-energy malnutritionMELAS. See Mitochondrial encephalomyopathy, lactic

acidosis, and strokelike episodesMembrane skeleton, 69Menadione, 99Methanogenesis, 272Methionine metabolism, 227, 228f. See also Vitamin B12

deficiencyMethylmalonic acid, 3034-Methylumbelliferyl-β-D-glucopyranoside (MUGIc), 172Microsomal triglyceride transfer protein, 293,

294f–295f, 296Mineralocorticoid replacement, 367Mitochondrial encephalomyopathy, lactic acidosis,

and strokelike episodes (MELAS)biochemical perspectives on, 93–98case report on, 89–90diagnosis of, 90–93therapy for, 98–99

Mitochondrial inheritance, 94–96mtDNA, 94, 95f, 97Mucolipidosis II. See I-cell diseaseMyelin, 309Myocardial infarction. See Cardiac biomarkers

N-acetylglutamate, 200N-acetylglutamate synthase, 197, 200N-carboxybiotin, 137fNitric oxide (NO), 23Nitrogen disposal, waste, 202, 202fNonalcoholic steatohepatitis (NASH), 248Nucleotide-binding folds, 164

Obesitybiochemical perspectives on, 248–252case report on, 245–247, 251, 252diagnosis of, 246, 248therapy for, 252–253

Okazaki fragment slippage, 14, 14fOral glucose tolerance test, 346Organic cation transporter 2, 104Ornithine transcarbamoylase, 197–198, 200Osmotic fragility test, 67, 68fOsteogenesis imperfecta

biochemical perspective on, 34–39case report on, 30–31

Index 371

Osteogenesis (continued )diagnosis of, 31–34therapy for, 39–40

Oxidative phosphorylation, 93, 94, 96Oxidative stress, 262. See also Free radicals

Pancreatic enzymes, 280–285Pancreatic exocrine insufficiency in chronic

pancreatitis, 282–286biochemical perspective on, 280–286case report on, 278–279diagnosis of, 279–280therapy for, 286–288

Pancreolauryl test, 285, 285fParathyroid hormone, 328–331P450c21 hydroxylase. See Congenital adrenal

hyperplasiaPentose phosphate pathway, 126, 128fPeriodic acid Schiff, 46Pernicious anemia, 303Peroxisomal function and peroxisomal diseases, 147Phenylacetylglutamine, 202fPhenylalanine. See PhenylketonuriaPhenylketonuria

biochemical perspective on, 205–213case report on, 204diagnosis of, 204–205therapy for, 213–215

Phosphoenolopyruvate carboxykinase,111–112, 113f

Phosphotransferase, 186–190Photobilirubin, 237, 238f, 240–241Plasmodium falciparum malaria, 21Procollagen, 37, 331Propionyl-CoA metabolism, 308–309, 311fProtease inhibitors. See α1-Antitrypsin deficiencyProtein-energy malnutrition

biochemical perspectives on, 258–262case reports on, 255–256diagnosis of, 257–258therapy for, 262–264

Pseudo-Hurler polydystrophy, 186, 188Pyruvate dehydrogenase complex deficiency

biochemical perspectives on, 78–85case report on, 77–78diagnosis of, 78therapy for, 84t, 85–87

Pyruvate dehydrogenase kinase, 80Pyruvate dehydrogenase phosphatase, 80–81

Reactive oxygen species, 340Red cell membrane, 68f, 69–70Renin-angiotensin-aldosterone system, 361, 363fRetinoic acid, gene transcription regulated by,

317, 320, 320fRetinoic acid receptors, 318, 320fRetinoid X receptors, 318, 320fRetinoids. See Vitamin ARetinol-binding protein, 315Riboflavin, 99Rickets, calcium-deficiency

biochemical perspectives on, 326–332case report on, 323–326, 324f, 324tdiagnosis of, 324–325therapy of, 332–333

S-adenosylhomocysteine, 227S-adenosylmethionine, 227Secretin-cerulein test, 284SERPINs (serine protease inhibitors), 48Sickle cell anemia

biochemical perspectives on, 19–26case history of, 17–18diagnosis of, 18–19, 20ftherapy for, 26–28

Sodium benzoate, 202Sodium phenylacetate, 202Spherocytosis. See Hereditary spherocytosisSphingolipid catabolism, pathway of, 173, 174fSphingolipidosis. See Gaucher diseaseSteatohepatitis, nonalcoholic, 248Steatorrhea, 282, 287–288Steroid hormones, 363–366Sterol regulatory element-binding proteins, 156

Tangier diseasebiochemical perspectives on, 162–165case reports on, 159–160diagnosis of, 161–162therapy for, 165–166

Tetrahydrobiopterin. See PhenylketonuriaThiamine, 86–87Thyrotropin (TSH), 113Transcobalamins, 302–303, 306–308, 307f, 308fTransferrin, 337Triglyceride breath test, 13C-mixed, 285–286, 286fTriglyceride transfer protein, microsomal, 293,

294f–295f, 296Troponins. See Cardiac troponins

Urea cycle disordersbiochemical perspectives on, 198–201case report on, 195diagnosis of, 195–198therapy of, 201–203

Ureotelic organisms, 198

Vascular cell adhesion molecule 1, 26Very long chain fatty acids. See X-linked

adrenoleukodystrophyVitamin A deficiency

abetalipoproteinemia and, 293, 297, 298biochemical perspective on, 315–321case reports on, 313–315diagnosis of, 313–315therapy for, 321

Vitamin B1. See ThiamineVitamin B2. See RiboflavinVitamin B12 (cobalamin), 305Vitamin B12 deficiency

biochemical perspectives on, 303–309case report on, 300–301

372 Index

diagnosis of, 301–303therapy for, 309–310

Vitamin D, 326–329Vitamin E, 296–298Vitamin K deficiency and abetalipoproteinemia,293,298Vitamin K3. See Menadione

Xerophthalmia, clinical classification of, 314tX-linked adrenoleukodystrophy

biochemical perspectives on, 147–149case report on, 144–145diagnosis of, 145–146therapy for, 149–150

Index 373

glew_clinical studies in medical biochemistry 3rd ed

Documents