Effect of Maternal Blood Phenylalanine Level on Mouse MaternalPhenylketonuria Offspring

Sangbun Cho and J. David McDonald1

Molecular Genetics and Metabolism 74, 420–425 (2001)doi:10.1006/mgme.2001.3255, available online at http://www.idealibrary.com on

Department of Biological Sciences, Wichita State University, 1845 Fairmount, Wichita, Kansas 67260-0026

e Nove

Published onlin

The genetic mouse model BTBR-Pahenu2 was usedto more thoroughly investigate the pathogenesis ofmaternal phenylketonuria (MPKU). More specifi-cally, it was used to examine the effect of maternalblood phenylalanine (PHE) level on the pregnancyoutcome of MPKU offspring as determined by cer-tain key measures of development at birth (i.e.,head circumference, weight, and crown-rumplength of offspring). In this study, we clearly ob-served that elevated maternal blood PHE levels,whether they were caused by the maternal diet orthe maternal genotype, were responsible for fetalabnormalities. As in human MPKU, significant re-ductions (P < 0.0001) in birth weight, crown-rumplength, and head circumference were seen in off-spring gestated under the condition of high mater-nal blood PHE levels. These findings strongly sug-gest that there are sufficient similarities betweenhuman MPKU and MPKU in this mouse model toestablish it as a very promising model for futurestudies designed to characterize human MPKUmore thoroughly. © 2001 Elsevier Science

Key Words: disease models; animal; phenylketonu-rias/metabolism; phenylalanine/blood; maternal-fe-tal exchange; pregnancy complications.

Phenylketonuria (PKU, McKusick No. 261600) isa hereditary disease characterized by severe neona-tal brain damage due to a lack of the hepatic enzymephenylalanine hydroxylase (PAH, EC 1.14.16.1).Without this enzyme, conversion of phenylalanine(PHE) to tyrosine (TYR) is blocked and PHE accu-mulates in the tissues. This syndrome is inherited

1 To whom correspondence should be addressed. Fax: (316)978-3772. E-mail: david.mcdonald@wichita.edu.

4201096-7192/01 $35.00© 2001 Elsevier ScienceAll rights reserved.

mber 30, 2001

as an autosomal recessive trait. Unless treated,PKU leads to severe mental retardation, neuropsy-chiatric symptoms, and defects in pigmentation(1,2).

Thirty years after the discovery of classical PKU,a diet therapy was developed that could preventmental retardation in this disorder (3–5). Shortlythereafter, with the advent of routine newbornscreening for PKU (6), it seemed possible to preventmental retardation in all individuals born withPKU. Indeed, the combined success of these twoactions has resulted in a large number of PKU indi-viduals who have been spared the severe mentalretardation of PKU. Most of these individuals arementally normal and are now of childbearing age.Unfortunately, many of them have discontinued thePHE-restricted diet since childhood and currentlyhave high blood PHE levels (7,8). It is now clear thatwomen with PKU who were not treated prior toconception are at increased risk of having a preg-nancy that results in serious fetal damage. The se-vere metabolic imbalance created by PKU in themother produces a toxic environment in her uterus,a syndrome called maternal PKU (MPKU). MPKUhas become a significant public health problem thatmust be resolved to prevent a second-generationrebound in the incidence of pathology associatedwith PKU (9). Unless treated, MPKU results in ter-atogenic effects on the fetus; typically microcephaly,mental retardation, intrauterine growth retarda-tion, congenital heart defects, low birth weight,spontaneous abortion, and stillbirth (10).

MPKU is capable of causing embryopathy and

fetopathy even in non-PKU offspring. In contrast toneonatal PKU, the sole determining factors for ma-

g/L PH

ternal PKU seem to be maternal diet and maternalgenotype, regardless of fetal genotype (11). The pre-cise etiology of this syndrome is unknown; however,it has been widely acknowledged that the increasedconcentration of PHE and/or its metabolites are theprimary suspects for the teratogenic factor(s) inMPKU (12). This response may be related to directtoxicity in certain fetal organs by excess PHE de-rived from a positive transplacental gradient. Due tothis positive PHE concentration gradient across theplacental barrier into the fetal compartment, thefetus is exposed to 1.5–2 times the PHE concentra-tion in the maternal blood stream, yielding an evengreater risk for the developing fetus than antici-pated on the basis of the maternal value alone (13).

PKU has been recognized as a major challenge fora number of years and has therefore been highlystudied. Nevertheless, surprisingly little is knownabout the pathophysiologic mechanism(s) of PHEtoward the fetus. This is in large part due to thetechnical and ethical limitations associated withclinical research. To more thoroughly investigatePKU pathogenesis and to explore potential thera-peutic actions, mouse PKU models were generated(14,15). Among these, the genetic mouse modelBTBR-Pahenu2 displays genetic, biochemical, andphysiological characteristics that are strikingly sim-ilar to human PKU and therefore seems appropriateto function as a model of this disease. The overallgoals of the research described in this manuscriptwere to use the BTBR-Pahenu2 mouse to examine theeffect of maternal blood PHE level on the pregnancyoutcome of MPKU offspring by observing severalkey measures of development at birth (i.e., headcircumference, weight, and crown-rump length ofoffspring).

MATERIALS AND METHODS

Production of Fetuses for Examination

TAExperimental Design to Produce Variation of

CrossesTest cross

/ BTBR-Pahenu2/Pahenu2 3 B

Dietary treatment group T0, Normal maternal blooMaternal dietary condition Amino acid diet, 1 0.625Maternal blood PHE range (mM) 90–120, in all dams

A MOUSE MATERNAL P

All animal manipulations were approved by theWichita State University Animal Care and Use

Committee and all BTBR-Pahenu2 mutant and BTBRcontrol mice were obtained either from a colonymaintained by Dr. McDonald in the animal holdingfacilities of Wichita State University or from theJackson Laboratory (Bar Harbor, ME).

The mice were housed in polycarbonate microiso-lator cages (Lab Products, Inc, Maywood, NJ). Theirenvironment was maintained at a constant temper-ature of 22–24°C, 50% relative humidity, and a 12-hlight cycle (0600–1800 h). As depicted in Table 1,homozygous mutant (BTBR-Pahenu2/Pahenu2) fe-males were crossed to heterozygous (BTBR-Pahenu2/1) males to provide homozygous and het-erozygous mutant offspring gestated in the sameuterus. Control litters were produced by crossingheterozygous (BTBR-Pahenu2/1) females with ho-mozygous mutant (BTBR-Pahenu2/Pahenu2) males.Females in mating cages were checked every morn-ing for the presence of a copulatory plug. Noon onthe day of plug detection was designated 0.5 dayspostcoitum (dpc). Upon plug detection, females wereassigned to one of the two experimental dietary con-ditions (i.e., T0 vs T3) as described in Table 1. Fe-males were provided with drinking water ad libitumwith an appropriate PHE concentration (0.625g/L:T0 or 6 g/L:T3) to produce the desired maternalblood PHE level. A synthetic amino acid-defined dietdevoid of PHE was provided to supply other nutri-tional needs (Teklad Research Diet No. TD97152,Harlan-Teklad, Madison, WI) and contains all es-sential amino acids except PHE. Defined diet ingre-dients and amounts (g/kg) were L-alanine (3.5), L-arginine-HCl (12.1), L-asparagine (6.0), L-asparticacid (3.5), L-cystine (3.5), L-glutamic acid, F.B.(40.0), glycine (23.3), L-histidine HCL z H2O (4.5),L-isoleucine (8.2), L-leucine (11.1), L-lysine-HCl(18.0), L-methionine (8.2), L-phenylalanine (0.0), L-proline (3.5), L-serine (3.5), L-threonine (8.2), L-tryptophan (1.8), L-tyrosine (5.0), L-valine (8.2), su-crose (498.38), corn starch (150.0), corn oil (100.0),

1nal Genotype and Maternal Blood PHE Level

ahenu2/1Control cross

/ BTBR-Pahenu2/1 3 BTBR-Pahenu2/Pahenu2

level T3, High maternal blood PHE level in mutant damsE Amino acid diet, 16 g/L PHE

.1200, in mutant dams

421LKETONURIA MODEL

BLEMater

TBR-P

d PHE

HENY

cellulose (fiber) (30.0), mineral mix, AIN-76 (35.0),calcium phosphate dibasic (4.5), vitamin mix, Tek-

MCD

lad (10.0), ethoxyquin (antioxidant) (0.02). Giventheir level of consumption of the PHE-containingdrinking water, this supplies about 6.25 mg PHE/day (for T0) and about 60 mg PHE/day (for T3). Forcomparison purposes, a mouse kept on a normalmouse diet consumes about 40 mg PHE/day.

At 18.5 dpc, the pregnant female mice were eu-thanized by cervical dislocation and their abdominalcavity was opened to expose the internal organs. Thedescending aorta was severed to obtain a blood sam-ple to determine maternal blood PHE concentration.Beginning at the ovary end of the left uterine hornand proceeding toward the cervix, the fetuses wereindividually dissected from the uterine lining.Trunk blood from each fetus was collected with acapillary tube and was used for fetal amino acidanalysis. All blood PHE and TYR concentrationswere determined by the commercial testing service,Neogen, Inc. (Pittsburgh, PA), using isotope-dilutiontandem mass spectrometry (16). Fetal tissue sam-ples were flash-frozen in liquid nitrogen and storedat 220°C. DNA was extracted from fetal tissues asdescribed in reference 17 and was used for sexingand genotyping the fetuses.

Spontaneous Abortion Incidence, HeadCircumference, Crown-Rump Length,and Birth Weight

Females were checked daily for evidence of com-plete or partial litter loss and scored accordingly; apositive score resulted from either complete or par-tial litter loss before term. Crown-rump length wasdetermined using an Iwanson Spring Caliper as de-scribed in Ref. (18). Maximum head circumferencewas measured by wrapping a thread around thefetus’ head (18). Head circumference measurementswere done in triplicate for each fetus and the meanvalue of the three measurements was used.

PCR-Based Genotyping and Sexing of Fetuses

The identity of the induced mutation of Pahenu2

has been characterized by DNA sequence analysis(19). The mutation in the Pahenu2 line creates a newrecognition site for the restriction enzyme Alw26I(Promega Co., Madison, WI) and the isoschizomerBsmA1 (New England Biolabs Inc., Beverly, MA) byaltering the sequence from nucleotides 831 through835 from GTCTT to GTCTC. To detect this restric-tion fragment length polymorphism, approximately

422 CHO AND

100 ng of genomic DNA was used as template forPCR amplification with gene-specific oligonucleotide

primer pairs as in Ref. (19). Digestion with eitherAlw26I or BsmA1 yielded a distribution of frag-ments with a distinct banding pattern for each of thethree possible genotypes.

The genomic DNA isolated from tissues also wasused for the sexual identification of each fetus. Prim-ers were designed to amplify the SMC gene, whichgives a Y-specific band that is distinct in size fromthe X-specific band; females gave a single band andmales gave two bands because of an SMC introndifference between the X and Y genes (20). SMColigonucleotide primers were SMCX-1, 59-CCGCT-GCCAAATTCTTTGG-39; and SMC4-1, 59-TGAAG-CTTTTGGCTTTGAG-39 (Genosys Biotechnologies,Woodlands, TX).

Statistical Analysis

Fetal developmental parameters were analyzedby a two-way factorial analysis of variance(ANOVA). Three distinct hypotheses were tested:Ho

1, no difference in developmental trait due to ma-ternal genotype; Ho

2, no difference in developmentaltrait due to maternal diet; and Ho

3, no significantinteraction between maternal genotype and mater-nal diet.

x2 homogeneity tests were used to determine if therelative frequencies of discrete variables (e.g., fetalgenotype, spontaneous abortion incidence) weresimilar between control and test conditions. All sta-tistical analyses were conducted using the Statisti-cal Analysis System (SAS, 1992). A test was judgedsignificant if P , 0.05.

RESULTS

Head Circumference, Crown-Rump Length, andBirth Weight

Preliminary data analysis indicated there was nosignificant difference between female and male fe-tuses regarding developmental parameters. There-fore, all data for each variable from both females andmales were pooled (Table 2). Mean values of fetalbirth weight, crown-rump length, and head circum-ference differed significantly between diet treat-ments (Tables 2 and 3). In all cases, fetal traits weresignificantly smaller for mutant mothers on the T3

diet vs mutant mothers on T0 diet. Fetal crown-rump length was also significantly affected by ma-ternal genotype (Tables 2 and 3); fetuses of homozy-

ONALD

gous mothers exhibited significantly smaller traits.In no case was there a significant interactions affect

HENY

between maternal diet and genotype on fetal traits(e.g., addition of high PHE in the diet had a similareffect on fetuses of both maternal types). As shownin Table 2, fetuses exposed to high PHE in the dietand a maternal PKU uterine environment exhibitedthe lowest trait values whereas fetuses of normalmaternal genotypes that were not exposed to excessdietary PHE exhibited the largest fetal traits.

Fetal PHE and TYR Concentrations

Significant maternal diet and maternal genotypeeffects were observed on offspring blood PHE andTYR concentrations (Tables 4 and 5). As shown,mean values for fetal PHE were larger with increas-ing PHE background from either maternal genotypeor diet. The PHE level exhibited by fetuses of ho-mozygous mothers and high PHE diet correspondsto classical PKU blood PHE levels of greater than1200 mM. Maternal genotypes also affect fetal TYR;fetuses of homozygous mothers exhibited smallerTYR values than fetuses of heterozygous mothers.Reduced levels of TYR are typical of classical PKUconcomitant with elevated blood PHE levels. All T3

vs T0 comparisons for blood PHE and TYR weresignificant to the P # 0.004 level with the exceptionof the T3 vs T0 comparison of PHE among BTBR-Pahenu2/1 dams. Interestingly, a two-by-two factorialANOVA indicated a significant interaction of geno-type and diet for TYR and PHE level. The interac-tion is illustrated by the greater than additive effecton fetal concentrations observed among fetuses ofhomozygous mothers and high PHE diet. The aver-age ratio of fetal-to-maternal PHE and TYR was2.14 6 0.06 and 2.69 6 0.08, respectively.

TABLE 2Mean Values for Fetal Developmental Traits

Maternaldiet

Maternal genotype

BTBR-Pahenu2/Pahenu2 BTBR-Pahenu2/1

T3 (80) T0 (85) T3 (17) T0 (5)

BW (g) 0.70 6 0.02 0.88 6 0.02 0.86 6 0.03 0.91 6 0.04CR (cm) 1.77 6 0.03 1.95 6 0.02 2.07 6 0.03 2.18 6 0.05HC (cm) 1.75 6 0.02 1.94 6 0.02 1.86 6 0.03 1.89 6 0.02

Note. BW, birthweight; CR, crown-to-rump length; HC, headcircumference. Mean values for body size measurements arelisted with the standard error of the mean.

A MOUSE MATERNAL P

Initially, there was concern that PHE in thedrinking water might lead to avoidance (i.e., ho-

mozygous mutant females might drink less waterbecause of the PHE addition). However, there wereno differences in drinking water consumption be-tween homozygous and heterozygous females. Aver-age water consumption was 9–10 ml per day forboth T3 and T0 levels. Further, there was concernthat fetal genotype would contribute to fetopathy.This concern appears to have been substantiated.The genotype test of approximately half of the fe-tuses from the test T3 cross revealed that 38% offetuses were mutant homozygotes and 62% of fetalgenotype were heterozygotes. The x2 test revealedthat 38% was different from 62%, indicating theratio of fetal genotype was not 1:1.

Spontaneous Abortion Incidence and Fetal Loss

Spontaneous abortion was noted in a T3 mutantfemale who had already started to deliver threeaborted fetuses at 18 dpc. An emergency hysterec-tomy was performed and, when the abdominal cav-ity was opened, one fetus was left in the right uter-ine horn and was found to be dead. Anotherspontaneous abortion was noted in a control T0 mu-tant female. When the abdominal cavity was openedat 18.5 dpc, the right uterine horn was greatly di-lated, indicating possible previous fetal occupancy.One fetus remained in the right uterine horn and,although dead, was developed to late developmentalterm.

Other forms of fetal loss, indicated by uterine re-sorption sites and stillbirths, were also noted. Re-sorption sites indicate early fetal death, where theconceptus is in the process of being reabsorbed bythe uterine wall. On the other hand, stillbirth isindicative of late fetal death. In order to determinewhether any classes are significantly different from

TABLE 3Results of Factorial ANOVA to Determine Signif-

icance of Differences of Mean Fetal Traits betweenMaternal Genotypes and Maternal Diets (T0 and T3)

Trait Source DF F P

BW (g) Genotype 1181 3.79 0.0531Diet 1181 38.60 0.0001Genotype and diet 1181 1.96 0.1635

CR (cm) Genotype 1183 21.34 0.0001Diet 1183 30.53 0.0001Genotype and diet 1183 0.46 0.4998

HC (cm) Genotype 1183 0.17 0.6813

423LKETONURIA MODEL

Diet 1183 40.55 0.0001Genotype and diet 1183 2.39 0.1240

4 6 3

) are li

others in the frequency of fetal loss, the data forresorption sites, spontaneous abortion, and still-birth were grouped together and compared by the x2

test. Approximately 5 times greater fetal losses wereobserved in mutant females than in nonmutant fe-males and the x2 test revealed that this was a sig-nificant difference in fetal loss (P , 0.001).

DISCUSSION

From what is known so far about human maternalPKU, the syndrome results from a combination ofmaternal genotype and environmental factors. Withregard to the genotypic basis, a mutant maternalgenotype is responsible for fetal abnormalities. Alloffspring are equally affected, suggesting that thefetal genotype does not contribute to teratogenicity.With regard to the environmental basis, a treatmentthat reduces the maternal blood PHE levels duringpregnancy can ameliorate the teratogenic effects ofmaternal PKU. In this study, we clearly see thatelevated maternal blood PHE levels, produced by acombination of mutant maternal genotype and highdietary PHE concentration, are responsible for theteratogenic birth defects when examined by several

TABLE 5Results of Factorial ANOVA to Determine Signif-

icance of Differences of Mean Fetal Blood PHE andTYR Concentrations between Maternal Genotypesand Maternal Diets (T0 and T3)

Amino acid Source DF F P

PHE Genotype 1127 81.12 0.0001Diet 1127 375.95 0.0001Genotype and diet 1127 52.36 0.0001

TYR Genotype 1127 104.97 0.0001

TABlood PHE and TYR Concentr

Diet

BTBR-Pahenu2/Pahenu2

T3 (52) T

PHE (mM) 2525.14 6 107.73 230.1TYR (mM) 137.63 6 4.97 108.4

Note. Mean values for blood PHE and TYR concentration (mM

424 CHO AND

Diet 1127 29.53 0.0001Genotype and diet 1127 15.21 0.0002

developmental parameters of pregnancy outcome.First, significantly reduced birth weights were notedamong offspring born to mothers with high bloodPHE levels. Second, crown-rump length was signif-icantly lower among offspring born to mothers withhigh blood PHE levels compared those born to moth-ers with low blood PHE levels. Third, significantlyreduced head circumference was also indicatedamong offspring born to mothers with high bloodPHE levels. The pregnancy endpoints of birthweight, head circumference, and crown-rump lengthof offspring were inversely affected by increasingmaternal blood PHE level during gestation. Thesefindings strongly suggest that in mouse PKU, as inhuman PKU, mutant maternal genotype and di-etary PHE combine to produce a maternal PKUsituation.

Although human studies suggest that the soledetermining factors of MPKU seem to be the mater-nal diet and maternal genotype, these conclusionsare based on data that are, in some important re-spects, difficult to control. It is not possible, there-fore, to completely exclude the contribution of thefetal genotype. Therefore, we thought it prudent totest for the presence or lack of a fetal genotypecontribution to fetopathy by producing half homozy-gous and half heterozygous offspring within thesame MPKU uterus. Therefore, the genotypes of thecrosses were as follows: test cross, / BTBR-Pahenu2/BTBR-Pahenu2 3 ? BTBR-Pahenu2/1; control cross, /BTBR-Pahenu2/1 3 ? BTBR-Pahenu2/Pahenu2. Thesecrosses produce a genotypically mixed population offetuses within the same uterus, which allows us toinvestigate for possible fetal genotype effects. It isclear from our data that all offspring surviving toterm seemed to be equally affected regardless oftheir genotype. However, total fetal losses (including

4s in Test vs Control Offspring

aternal genotype

BTBR-Pahenu2/1

T3 (17) T0 (5)

.26 112.18 6 18.60 92.26 6 15.80

.88 247.07 6 17.99 130.84 6 24.75

sted with the standard error of the mean.

ONALD

BLEation

M

0 (57)

5 6 60

MCD

spontaneous abortions, resorption sites, and still-births) were significantly increased in mutant moth-

HENY

ers. Further, of those fetuses surviving to term,there was a significant deficit of homozygous mu-tants. Therefore, in this study it appears that mu-tant fetuses are at a selective disadvantage duringgestation. It is important to note that extensive ex-perience with the control cross (i.e., / BTBR-Pahenu2/1 3 ? BTBR-Pahenu2/Pahenu2) in colonymaintenance reveals no deficit of homozygous off-spring (McDonald, unpublished). This suggests thatmutant fetuses are not inherently less able to sur-vive to term but may be at a selective disadvantageunder the adverse conditions of gestation in theuterus of a mutant mother.

In summary, this study represents proof of prin-ciple that the BTBR-Pahenu2 mouse model will beuseful for maternal PKU modeling studies. Thestudy is preliminary in the sense that it representsan opening foray into the use of this model to studythe relationship between maternal blood PHE andbirth defects among the offspring. Future studieswill more thoroughly survey the offspring for otherpathologic endpoints seen in human maternal PKU,including heart defects, neuropathology, and post-natal growth deficits.

ACKNOWLEDGMENTS

This material is based upon work supported by the NationalScience Foundation under Grant EPS 9550487. Gabriele Whiteand Mark Haefele are thanked for technical assistance in geno-typing and sexing PCR determinations. Dr. Karen Brown isthanked for assistance with the statistical analysis. Dr. DonaldChace, of Neogen Inc., Pittsburgh, Pennsylvania, is thanked fortechnical assistance in amino acid determinations.

REFERENCES

1. Trefz FK, Schmidt H, Bartholome K, Mahl M, Matthis P,Pecht G. Differential diagnosis and significance of varioushyperphenylalaninemias. In Inherited Diseases of AminoAcid Metabolism (Bockel H, Wachtel U, Eds.). Stuttgart:Thieme, pp 86–100, 1985.

2. Scriver CR, Kaufman S, Eisensmith RC, Woo SLC. Thehyperphenylalaninemias. In The Metabolic and MolecularBases of Inherited Disease, 7th ed. (Scriver CR, Beaudet A,Sly W, Valle D, Eds.). New York: McGraw-Hill, pp 1015–1075, 1995.

3. Bickel H, Gerrard J, Hickmans EM. Influence of phenylala-nine intake on phenylketonuria. Lancet 2:812–813, 1953.

A MOUSE MATERNAL P

4. Woolf LI, Griffiths R, Moncrieff A. Treatment of phenylke-tonuria with a diet low in phenylalanine. Brit Med J 1:57–64, 1955.

5. Armstrong MD, Tyler FH. Studies on phenylketonuria. I.Restricted phenylalanine intake in phenylketonuria. J ClinInvest 34:565–580, 1955.

6. Guthrie R, Susi A. A simple phenylalanine method for de-tecting phenylketonuria in large populations of newborninfants. Pediatrics 32:338–343, 1963.

7. Koch R, Levy HL, Hanley W, Matalon R, Rouse B, Trefz F,de la Cruz F. Outcome implications of the internationalmaternal phenylketonuria collaborative study (MPKUCS).Eur J Pediatr 155 (Suppl 1):162–164, 1996.

8. Hanley WB, Koch R, Levy HL, Matalon R, Rouse B, Azen C,de la Cruz F. The North American maternal phenylketonu-ria collaborative study, developmental assessment of theoffspring: Preliminary report. Eur J Pediatr 155 (Suppl 1):169–172, 1996.

9. Kirkman H. Projections of a rebound in frequency of mentalretardation from phenylketonuria. Appl Res Ment Retard3:319–328, 1982.

10. Lenke RR, Levy HL. Maternal phenylketonuria and hyper-phenylalaninemia: An international survey of the outcomeof untreated and treated pregnancies. N Engl J Med 303:1202–1208, 1980.

11. Levy H, Lobbregt D, Sansariq C, Snyderman S. Comparisonof non-phenylketonuria sibs from untreated pregnancies ina mother with phenylketonuria. Am J Med Genet 44:439–442, 1992.

12. Levy H. Maternal phenylketonuria. Review with emphasison pathogenesis. Enzyme 38:312–320, 1987.

13. Schoonheyt WE, Clarke JTR, Hanley WB, Johnson JM, Le-hotay DC. Feto-maternal plasma phenylalanine concentra-tion gradient from 19 weeks gestation to term. Clin ChimActa 225:165–169, 1994.

14. McDonald J, Bode V, Dove W, Shedlovsky A. A mouse mu-tant deficient in phenylalanine hydroxylase. Proc Natl AcadSci USA 87:1965–1967, 1990.

15. Shedlovsky A, McDonald J, Symula D, Dove W. Mouse mod-els for human PKU. Genetics 134:1205–1210, 1993.

16. Chace DH, Millington DS, Terada N, Kahler SG, Roe CR,Hofman LF. Rapid diagnosis of phenylketonuria by quanti-tative analysis for phenylalanine and tyrosine in neonatalblood spots by tandem mass spectrometry. Clin Chem 39:66–71, 1993.

17. Moore D. Preparation and analysis of DNA. In CurrentProtocols in Molecular Biology(Ausubel FM, Brent R, King-ston RE, Moore D, Siedman JG, Smith JA, Struhl K, Eds.).New York: Wiley Interscience, pp 2.0.1–2.1.7, 1998.

18. Saad AY. Postnatal effects of nicotine on incisor develop-ment of albino mouse. J Oral Pathol Med 19:426–429, 1990.

19. McDonald JD, Charlton CK. Characterization of mutationsat the mouse phenylalanine hydroxylase locus. Genomics39:402–405, 1997.

425LKETONURIA MODEL

20. Agulnik AI, Longepied G, Ty MT, Bishop CE, Mitchell M.Mouse H-Y encoding Smcy gene and its X chromosomalhomolog Smcx. Mamm Genome 10:926–929, 1999.