advances in islet cell biology...original article advances in islet cell biology from stem cell...

ORIGINAL ARTICLE

Advances in Islet Cell BiologyFrom Stem Cell Differentiation to Clinical Transplantation: Conference Report

Fouad Kandeel, Craig V. Smith, Ivan Todorov, and Yoko Mullen

Summary: The 3rd Annual Rachmiel Levine Symposium entitled“Advances in Islet Cell Biology—From Stem Cell Differentiation toClinical Transplantation” was organized by the Department of Dia-betes, Endocrinology and Metabolism at the City of Hope NationalMedical Center, with the support of the Southern California Islet CellResources Center, American Diabetes Association–David ShapiroResearch Fund, Ross Foundation, the National Center for ResearchResources (NCRR), and the National Institute of Diabetes and Diges-tive and Kidney Diseases (NIDDK) of the National Institutes ofHealth. The symposium was held at the Hilton Anaheim Hotel in Ana-heim, CA, in October 2002, and was attended by nearly 400 partici-pants from 23 countries and 30 U.S. states. The symposium consistedof 11 sessions focusing on 3 areas: (1) pancreas and islet cell differ-entiation and islet generation, (2) � cell biology and insulin synthesisand/or secretion, and (3) pancreatic islet transplantation in patientswith type I diabetes. Thirty-nine world experts lectured on the mostcurrent information in each field. Fifty-three abstracts were selectedfor presentation and discussed at the poster session. The first author ofeach of the top 10 posters received a Young Investigator TravelAward provided by the National Center for Research Resources andthe Southern California Islet Cell Resources Center. The symposiumalso offered special Meet the Professor sessions, which gave the at-tendees an opportunity to closely interact with the participatingspeakers of the day.

(Pancreas 2003;27:e63–e78)

The Rachmiel Levine symposium is held annually in hismemory (1910–1998). Dr. Levine served as the Executive

Medical Director at the City of Hope National Medical Centerin Duarte, CA, from 1971 to 1991. Originally from EasternPoland, he immigrated to Canada where he attended McGillUniversity in Montreal and received his medical degree with

honors in 1936. Following medical school, Dr. RachmielLevine relocated to Michael Reese Hospital in Chicago, IL,and commenced his studies in diabetes research. In 1949, hegained recognition as the first scientist to discover the role ofinsulin in glucose metabolism. Known as the “Levine effect,”Dr. Levine’s transport theory was the first to describe insulinas the key regulatory factor for glucose transport into cells. Hisaccomplishments continued at the City of Hope where he de-veloped the Diabetes Program. Dr. Levine’s enthusiastic sup-port and encouragement also led to the synthetic production ofhuman insulin (Humulin®), the first genetically engineeredhealth care product from Escherichia coli bacteria, by Dr. Ar-thur D. Riggs and Dr. Keiichi Itakura at the City of Hope.1 Dr.Levine contributed more than 60 years of his life to diabetesresearch, and, as a result, he is often referred to as the “Fatherof Modern Diabetes Research.”

The new era of pancreatic islet transplantation wasopened by the successful restoration of euglycemia in patientswith type 1 diabetes by the Edmonton group in Alberta,Canada.2 Their successes have been reproduced by severalother transplant centers and have increased in every aspect ofislet transplantation. To reflect this, two-thirds of the 2002Levine symposium program provided updated information onpancreas development, differentiation of islet and/or � cellsand gene expression, islet cell signaling networks, and ex vivogeneration of � cells. The remaining one-third of the programcovered issues related to clinical islet transplantation, includ-ing recent advances in islet isolation and culture technologies,strategies in immune tolerance induction, update of the Im-mune Tolerance Network (ITN) Trial, and regulatory issues inislet transplantation. The 3-day program ended with a discus-sion of metabolic lessons learned from islet transplantation.

ADVANCES IN PANCREAS MORPHOGENESISAND ENDOCRINE STEM-CELL RESEARCH

The mammalian pancreas is composed of both endo-crine and exocrine tissues. Four endocrine cell types are con-tained in islets of Langerhans, which comprise 1%–2% of thecellular mass of the adult pancreas. The remaining tissue isorganized into acini, which synthesize digestive hydrolases,and ducts, which secrete a bicarbonate fluid to flush the acinarsecretions to the intestine. A comprehensive scheme of pan-

Received for publication June 16, 2003; revised manuscript accepted June 17,2003.

From the Leslie and Susan Gonda (Goldschmied) Diabetes and Genetic Re-search Center and Southern California Islet Cell Resources Center, City ofHope National Medical Center/Beckman Research Institute, Duarte, Cali-fornia

Reprints: Dr. Yoko Mullen, Department of Diabetes, Endocrinology and Me-tabolism, City of Hope National Medical Center, 1500 E. Duarte Rd., Du-arte, CA 91010, U.S.A. (e-mail: [email protected]).

Copyright © 2003 by Lippincott Williams & Wilkins

Pancreas • Volume 27, Number 3, October 2003 e63

creatic development must include both the resolution of theendocrine and exocrine tissues, as well as the multiple celltypes contained in each tissue. In the classic description of pan-creatic morphogenesis, the molecular mechanisms underlyingdorsal and ventral pancreatic development are similar, isletand acinar cells derive from ductal cells, and the hormone-expressing cells initially detected in the nascent pancreaticbuds are precursors for later endocrine cells. However, the re-sults of recent studies have revised this original frameworkconsiderably.

Pedro Luis Herrera Merino (Geneva, Switzerland)provided a clear picture of pancreas morphogenesis. Duringembryogenesis, the pancreatic endoderm is exposed sequen-tially to distinct mesodermal cell populations. The dorsal en-

doderm contacts the notochord, aorta, and pancreatic mesen-chyme. The ventral endoderm lies adjacent to the septumtransversum and cardiogenic mesoderm and then contacts ven-tral veins and mesenchyme. Pancreatic morphogenesis beginsin response to signals from these mesodermal tissues and isorchestrated by a cascade of transcription factors expressed byspecific cells in a specific sequence. Two essential homeodo-mains contain the transcription factors Hlxb9 and Pdx1/Ipf1,which are expressed before bud formation (∼E9.5). However,Hlxb9 expression becomes negative within 2 days of bud for-mation and then reappears in differentiating islet cells. Pdx1expression continues in duodenal and dorsal stomach epithe-lium (E11.5), and in the absence of Pdx1, growth and morpho-genesis of the bud epithelia is suspended. The inductive signals

FIGURE 1. Morphogenesis of the mouse pancreas and the expression of key genes during embryonic development. Table 1summarizes the role of each individual gene. FGF4, fibroblast growth factor 4; FGF2, fibroblast growth factor 2; Shh, Sonichedgehog. Graphics adapted with permission from Curr Opin Genet Dev. 2002;12:540–547.12–15

Kandeel et al Pancreas • Volume 27, Number 3, October 2003

e64 © 2003 Lippincott Williams & Wilkins

required for the dorsal and ventral foregut endoderm to de-velop into the pancreas are provided by adjacent mesodermalstructures, such as the notochord, dorsal aorta, cardiogenic me-soderm, and septum transversum. The expression of Sonichedgehog (SHH) in a region of the foregut is repressed by thenotochord signal, leading the dorsal foregut endoderm to be-come the pancreas, while the absence of the repressor signalleads to intestine formation.

The expression of key pancreatic transcription factorssuch as Pdx1, Pbx1, PTF1–p48, and Neurogenin 3 (Ngn3) isrequired throughout development of the pancreas. AlthoughPdx1 expression is not essential for the initiation of pancreaticbudding from the endoderm, the expression of Pdx1 and Ngn3is essential for regulating pancreatic morphogenesis and dif-ferentiation. Pbx1 is a member of the TALE class of homeodo-main transcription factors that regulate the activity of otherhomeobox factors and is essential for normal pancreatic devel-opment and function. PTF1–p48 is the pancreas-specific sub-unit of the heteromeric bHLH protein complex PTF1 thatbinds and activates the transcriptional enhancers of the acinarhydrolytic enzyme genes. The major role of p48 may be inacinar cell differentiation that begins at ∼E13.5. However, p48is expressed as early as E9.5 and has been shown to play a morefundamental role in endocrine and exocrine pancreatic devel-opment. In the nascent buds (E9.5–10), glucagon- and insulin-positive cells are detected by immunostaining, but whetherthese cells are islet precursors has yet to be determined. Pro-liferation of the pancreatic epithelium and mesenchymebranching, morphogenesis, and fusion of the dorsal and ventralbuds occur by E12.5 and produce an epithelial tubular complexcontaining the precursor cells for islets, acini, and ducts. Thefirst cells differentiating in the embryonic pancreas are endo-crine cells (glucagon-positive cells) followed by insulin-positive cells. These cells are found in duct epithelia, migrateinto the interstitium, and form endocrine cell clusters. Morpho-genesis of mouse pancreas is diagrammatically presented inFigure 1,12–15 and genes expressed during pancreas develop-ment and their possible roles are summarized in Table 1.

Using different cell tracing approaches, Dr. Herrera’sgroup has delineated pancreatic endocrine cell lineages. Usingthe Cre/Lox approach, which permits the detection of the de-scendants after they no longer express the reporter genes, thegroup has shown that (1) all pancreatic cells derive from cellsexpressing Pdx1, (2) islet cells derive from the cells expressingNgn3 and Pax6, while exocrine cells derive from cells express-ing p48, (3) glucagon-positive cells derive from a lineage dif-ferent from insulin-positive cells, and early glucagon-positivecells are not the precursors of � cells, and (4) early pancreaticpolypeptide-positive cells are necessary for � cell develop-ment. Interestingly, early glucagon-positive cells also expressPdx1, but adult glucagon-positive cells do not.

Michael S. German (San Francisco, CA) discussed �cell differentiation in transgenic mice, which has been used for

determination of the origin of pancreatic endocrine cells. Hehas shown that endocrine cells derive from cells positive forPdx1 and Ngn3, although Ngn3 expression occurs only duringembryogenesis. Ngn3 turns on several genes, including PAX4and NKX2.2.

The research goal of Christopher V. E. Wright and hiscolleagues (Nashville, TN) is to achieve a complete under-standing of the gene networks and cell interactions underlyingpancreas differentiation and maintenance as reflected by thetitle of his presentation “Global and Tissue-Specific Require-ments for Transcription Factors in Pancreas Development.”Using Cre/LoxP-mediated gene inactivation and lineage trac-ing, this group has shown that Pdx1 is required for outgrowthand differentiation of pancreatic buds. Pdx1 is required in adultinsulin-producing cells, and a low level of Pdx1 expression(relative to � cells) is essential for differentiation in exocrinecells. The PTF–p48b gene, which is required in adult acinarcells, is expressed in a subregion of the endodermal Pdx1-expression domain. Their recent studies using genetic recom-bination-based lineage tracing in normal tissue and after geneinactivation have shown that PTF–p48 is not exocrine specificbut expressed in progenitors of ducts, islets, and acinar cells,and he speculated that these cells may be multipotent stemcells. Inactivation of PTF1–p48 gene causes pancreatic pre-cursors to convert to duodenal epithelium. Thus, PTF1–p48must be an essential molecular trigger for acquisition of pan-creas fate from the foregut.

Islet morphogenesis and stem-cell markers were alsodiscussed by Luc Bouwens (Brussels, Belgium). He suggestedthat mammalian cells retain the capacity for dedifferentiation,cell cycle reentry, and transdifferentiation. His talk was basedon the observation of islet neogenesis in the rat pancreas at thelate gestation stage and the duct-ligated pancreas in adults.Distinct findings in the duct-ligated pancreas of rats treatedwith gastrin included the appearance of a high number ofsingle insulin-positive cells and the presence of cell clustersformed by a few insulin-positive cells, which may indicate is-lets budding from precursors or islet neogenesis. In the surviv-ing acinar cells, a metaplastic conversion took place after ductligation to become regenerative epithelial cells. In culture, aci-nar cells lose the ability to produce acinar-specific molecules,transdifferentiate into a duct-like phenotype, and display fetalpancreatic cell characteristics. Duct cells and dedifferentiatedexocrine cells have shown far greater proliferative potentialthan islet cells, and their plasticity, as Dr. Bouwens speculated,plays a significant role in the rapid increase in islet numbers inthe pancreas during late gestation and duct ligation–mediatedislet regeneration. He also speculated that enzymatic dissocia-tion of pancreatic tissue may have similar metaplasia-inducingeffects as seen in duct ligation. He showed data demonstratingthat nestin is expressed in the pancreatic mesenchymal “stel-late”-type cells and neovascular and/or angiogenic endothe-

Pancreas • Volume 27, Number 3, October 2003 Advances in Islet Cell Biology

© 2003 Lippincott Williams & Wilkins e65

TABLE 1. Transcription Factors Involved in Mouse Pancreas Morphogenesis: Family, Embryonic Day, and Expression Site.

TranscriptionFactors Family

EmbryonicDay Expression Site

HNF-3�/Foxa2 3a 5.5–6.5 Expressed in the anterior part of the early primitive streak and subsequently confinedin a definite area of the endoderm

Adult* All endocrine cellsHNF-3�/Foxa2 3 7.5 In the endoderm

Adult All pancreatic cellsHlxb9 1b 8.0 Expressed in the epithelium of the forgut-midgut junction

12.5 Expression decreases13.5 Expression increases again in the dorsal and ventral bud

by 17.5 Restricted to islet cellsAdult Restricted to � cells

Pdx1/Ipf1 1 8.5 Uniformly expressed in the pancreatic bud and marks the territory of the pancreas;expression is later restricted to � cells

Adult � cellsPbx1 Homeodomain FHNF6 1 9.5 Expressed in pancreatic epithelial cells and later the expression is restricted to the

exocrine and ductal cellsPax4 1 9.5 Expressed by pancreatic progenitors

Adult Low or no expression in pancreasPax6 1 9.0 Expressed in the pancreas

10.5 Pax6 and glucagon are coexpressed15.5 Pax6 is coexpressed with insulin or glucagon

Adult All mature endocrine cellsNkx2.2 1 9.5 Expressed by pancreatic progenitors and subsequently restricted to �, �, and PP cells;

� cells are Nkx2.2 negativeAdult �, �, and PP cells

Isl1 1 9.0 Expressed in the dorsal pancreatic epithelium and in the mesenchyme, that surroundsthe dorsal evagination of the gut endoderm

Adult All endocrine cellsNgn3 2 9.0–9.5 In the pancreas anlage

15.5 Reaches peak expression and then decreases18.5 Very little expression

Adult Negative in islets, but expressed by duct and periductal cells, which suggests cellprecursor population

NeuroD1/BETA2 2 9.5 Pancreas dorsal budAdult All endocrine cells

PTF1-p48 2 9.5 Pancreatic progenitorsAdult Exocrine cells

Nkx6.1 1 10.5 Broadly expressed in the pancreatic budsBy 15.5 Expression is restricted to insulin+ cells and some scattered ductal and periductal cellsAdult Only in � cells

*Adult = after birth.aThe winged helix family.bThe homeodomain family.cThe basic helix-loop-helix (bHLH) family.IAPP, islet amyloid peptide; PC1/3, prohormone convertase (PC); PP, pancreatic polypeptide.References 12, 13, 15.



TABLE 2. Transcription Factors Involved in Mouse Pancreas Morphogenesis: Function and Observation in TargetedDisruption Mutant (�/�) Mice.

TranscriptionFactors Function Observation in Targeted Disruption Mutant (−/−) Mice

HNF-3�/Foxa2 Lack of foregut morphogenesis; mice die about E11HNF-3�/Foxa2 May be involved in glucagon gene transcriptional activity Reduction of glucagon mRNA by 70%, without reduction

of � cellsHlxb9 Required for the formation of the dorsal pancreas and

terminal differentiation into � cells; however, earlydevelopment of ventral bud does not require Hlxb9

Agenesis of dorsal pancreatic lobe without Pdx1 or Isl1;the ventral lobe developed, expressed Pdx1, andsubsequently contained all 4 types of endocrine cells; inaddition, the ventral pancreatic lobe contained 65% fewercells (E18.5)

Pdx1/Ipf1 Key role in pancreas morphogenesis and in maintaining theinsulin production and glucose-sensing system in � cells

Pancreatic agenesis; � cell specific inactivation of Pdx1 ledto loss of � cell phenotype and diabetes

Pbx1 Interacts with Pdx1 and the Pdx1/Pbx1 complexes;necessary for the expansion of the pancreatic buds butnot for the specification of the different pancreatic cells

Impairment of both endocrine and exocrine portions of thepancreas

HNF6 Specifies pancreatic epithelial cells; is necessary for thenormal increase in endocrine cells; stimulates Ngn3 geneexpression

Exocrine portion of the pancreas was normal; on E9.5,glucagon expression was standard but on E12.5 insulin+cells were absent and glucagon+ cell number decreasedby 85%; after birth, �, �, �, and PP cells were markedlyreduced; at birth, Ngn3 expression lessened and thepancreas showed abnormal architecture; mice werehypoglycemic at birth and then became glucoseintolerant

Pax4 Differentiation of � and � cells. Not necessary for thedevelopment of early insulin+ cells, but required for �and � cell maturation and maintenance.

On E13.5, insulin+ cells were negative and markers for �cells were absentMice appeared normal at birth but died in 3–5 days; inthe pancreas, mature � and � cells were absent and �-cellnumbers increased

Pax6 Required for the regulation of islet cell count, isletmorphology, and islet hormone synthesis levels.

In the pancreas, � cells were absent; �, �, and PP cells weremarkedly reduced and islets were disorganizedThe pancreas of Pax4 and Pax6 double mutant micecontained no mature endocrine cells

Nkx2.2 Required for terminal differentiation of � cells andmaintenance of Nkx6.1

Arrest of � cell differentiation and no maturation of � cells;decreased � and PP cell count and incompletedifferentiation of � cells; some � cell markers such asIAPP and PC1/3 were also present; Nkx6.1 were apparentbut disappeared by E18.5

Isl1 Required for early formation of the pancreas anddifferentiation of islet cells but not for exocrine cells

No formation of the dorsal pancreatic mesoderm, failure ofexocrine cell differentiation in the dorsal pancreatic bud,and loss of differentiated islet cells

Ngn3 Involved in islet cell development of endocrine precursorsbut turns off before the final differentiation

At birth, the −/− mouse pancreas contained an exocrineportion but no endocrine cellsThe pancreas of mice overexpressing Ngn3 showed amarked increase of endocrine cell formation at very earlystage of development (E8.5)

NeuroD1/BETA2 Important in the expression of differentiated endocrine cellproduct and for insulin gene transcription; it is inducedby Ngn3

Reduction of � (40%), � (75%), and � (20%) cells at birthdue to increased apoptosis; the remaining � cellscontinued to produce insulin

PTF1-p48 Required for epithelial cells of the bud to commit tobecome the pancreas; in later stages of development, it isrequired for exocrine development

Lack of exocrine portion of the pancreas; endocrine cellsformed but reside in the spleen

Nkx6.1 Necessary for expansion and final differentiation of �-cellprogenitors

Insulin+ cells were present at the same level as wild typeuntil E12.5

Afterward, �-cell numbers severely decreased, but �, �,and PP cell count remained normal; Ngn3+ progenitor cellswere present, but insulin+ cells reduced by 95%

*Adult = after birth.aThe winged helix family.bThe homeodomain family.cThe basic helix-loop-helix (bHLH) family.IAPP, islet amyloid peptide; PC1/3, prohormone convertase (PC); PP, pancreatic polypeptide.References 12, 13, 15.

lium of fetal and adult regenerated rat pancreas, but not in pan-creatic epithelial cells, including both ducts and islets.

Cell-cell and cell-matrix interactions play a critical rolein the morphogenesis of islets. Vincenzo Cirulli (LaJolla, CA)reported that �v�3 and �v�5 integrins, which were known tocoordinate epithelial cell adhesion and motility, were ex-pressed in pancreatic duct cells and clusters of undifferentiatedepithelial cells emerging from the duct. His team detected, invitro, the presence of these molecules at the tip of actin fila-ments. Blocking these molecules interferes with cell spreadingand migration. In vivo, blocking the activities of �v�3 and�v�5 integrins inhibited islet formation, decreased the numberof � cells, and increased the number of � cells. These resultsindicate that integrin molecules may also signal cell differen-tiation. Their recent studies have focused on diffusible factors,such as cytokines, chemokines, and growth factors that couldbe stored in the extracellular matrices. Netrin-1 is a secretedprotein that exhibits chemoattractant properties in the devel-oping central nervous system, where it mediates axon guid-ance. Netrin-1 is found to be expressed by a discrete popula-tion of epithelial cells in the fetal pancreas, to bind efficientlyto extracellular matrix components such as collagen IV, and tomediate cell migration through �6�4 and �3�1 integrins,which act as novel receptors for Netrin-1 in epithelial cells.Based on these findings, he suggested the existence ofintegrin/Netrin-1 interactions as novel adhesive and/or migra-tory cues, possibly regulating developmental processes in themammalian pancreas.

Gladys Teitelman (Brooklyn, NY) has used an insulin-treated streptozotocin (STZ) diabetic mouse model to charac-terize islet-derived multipotential stem and/or progenitor cells.She discovered that the restoration of euglycemia by insulininjections immediately following � cell depletion by STZ dra-matically increases � cell neogenesis. In this model, new �cells are generated from 2 different sets of intraislet � cell pre-cursors expressing embryonal traits: 1 cell lineage expressedsomatostatin and Pdx1 and the other GLUT2. Dr. Teitelmanspeculated that after STZ exposure, both � and � cells revert toan immature phenotype, characteristic of these cells during de-velopment. Thus, � cells resume Pdx1 expression, initiate in-sulin expression, downregulate somatostatin synthesis, and be-come monospecific � cells, while � cells start to expressGLUT2. In addition, the GLUT2 expression is presumably ac-tivated in precursor cells residing in the islets as they differen-tiate into � cells. Discrepancy between her speculation and thefindings by other investigators as to whether glucagon-positivecells are precursors of � cells was not resolved.

Construction of a functional genomic library enriched inendocrine transcripts expressed during fetal development is anessential tool in these areas of research. Klaus H. Kaestner(Philadelphia, PA) represented the NIDDK-sponsored consor-tium entitled “Functional Genomics of the Developing Endo-crine Pancreas” and summarized its activities over the past 3

years. The 3 major objectives of the consortium are (1) to de-velop and sequence cDNA libraries of mouse and human en-docrine pancreas, (2) to construct cDNA microarrays enrichedfor clones expressed in the endocrine pancreas, and (3) to de-velop bioinformatics tools for the diabetes community (exem-plified in the Endocrine Pancreas Consortium Database, EP-ConDB (www.cbil.upenn.edu/EPConDB/). The EPConDBwebsite contains information on the libraries and sequencesgenerated by the consortium, as well as data from microarrayexperiments. The library construction makes use of the Ngn3knockout mice, which lack all endocrine cells, to construct asubtracted library enriched in endocrine transcripts expressedduring fetal development. Unique nonredundant mRNA as-semblies identified so far were 7,700 in mice and 8,000 in hu-mans. A pilot microarray of close to 4000 unique elements wasconstructed based on a bioinformatics approach, and a com-bined pancreas-enriched microarray with more than 11,000 el-ements is currently in the process of printing. The full-lengthsequencing of the 7700 nonredundant mouse pancreas se-quences and high-throughput expression profiling of pancreas-enriched transcripts are also in progress. Information on thelibraries, sequences, and data from microarray experiments areavailable on the EPConDB website.

RECENT DEVELOPMENT IN �-CELL INSULINSYNTHESIS AND SECRETION RESEARCHThe conversion of proinsulin to insulin occurs at the 2

processing sites marked by pairs of basic amino acids, primar-ily after packaging into insulin secretory granules.3 HowardW. Davidson (Denver, CO) presented new findings on en-zymes involved in the processing of proinsulin to insulin. Pro-insulin processing requires 3 enzymes: prohormone conver-tase 1 (PC1), prohormone convertase 2 (PC2), and carboxy-peptidase E (CPE). Processing is essentially restricted tonascent and maturing secretory granules. PC1-mediated cleav-age of the B-C junction (Lys-Thr-Arg↓Glu-Ile) likely occursfirst. The exposed basic residues are then removed by CPE andsubsequently cleaved by PC2 at the A-C junction (Leu-Gln-Lys-Arg ↓Gly-Ile) to release the mature hormone. The ioniccompositions of the various secretory compartments and thebiochemical properties of the enzymes regulate this order.Thus, PC1 is activated in the endoplasmic reticulum, but isessentially nonfunctional prior to delivery to the trans-Golginetwork/nascent granule, due to a higher pH and lower cal-cium ion concentration of the earlier compartments, and pos-sibly by binding of the propeptide or of pro SAAS, an endog-enous PC1 inhibitor expressed in neuroendocrine cells. Theactivation of PC2 is prevented by the requirement of an acidicpH for its autoactivation and by a noncompetitive inhibitor,pro7B2. Furthermore, PC2 binding to the A-C junction likelyrequires some structural alteration of the precursor.

Peter Arvan (Bronx, NY) provided a comprehensive re-view on secretory protein trafficking from post-Golgi secre-



tory protein packaging to exocytosis, including a modeltermed the “cisternal maturation model of the Golgi complex”and the “sorting-by-retention” hypothesis.4 GH4C1 cells lackPC1 and PC2. By inducing stable PC1 expression in theGH4C1 cells, his group has shown that PC1 is involved in stor-ing insulin in secretory granules. The yeast prohormone con-vertase Kex2p mediates intracellular retention of insulinthrough endoproteolysis that express recombinant insulin-containing fusion protein. Dr. Arvan’s team found a family ofgenes in yeast that affect the activity of Kex2p processing ofproproteins in the secretory pathway. They found evidencethat a number of soluble secretory proteins enter the stimulus-dependent insulin secretory pathway of � cells, indicating arelative lack of selectivity at this step. Syntaxin 6 (Syn6), aSNARE protein involved in vesicle fusion, is localized to im-mature secretory granules (and endosomes), but not mature �cell granules, and is known to be required for mature granulebiogenesis. To examine the role of Syn6 either full-lengthSyn6 or a dominant-negative membrane-anchorless construct,Syn6t, was used to express the gene in INS-1 cells using anecdysone-inducible gene expression method. The expressionof Syn6t inhibited the lysosomal delivery via the endocyticpathway. However, Syn6t expression impaired neither secre-tory granule formation, proinsulin processing, nor secreta-gogue-stimulated �-granule exocytosis. These results have in-dicated that the Syn6 SNARE protein fulfills a specific role inendosomal maturation but plays a less important and indirectrole in the regulated, constitutive, and lysosomal biogenesispathway.

Guy A. Rutter (Bristol, UK) presented new imaging ap-proaches that permit monitoring dynamic molecular eventstaking place in single living cells. Gene expression and insulinsecretion were monitored using bioluminescent probes, suchas firefly luciferase, which were inserted into cells by micro-injection or by adenoviral technology. Dr. Rutter showed thatthese methods had potential to turn a single cell into a living“test tube” by achieving specific “knockout or knockdown” ofkey signaling elements. He stated that quantitative measure-ments of mRNA levels at a single cell level would also becomefeasible before long. The key finding in this field over the past5 years is that added insulin can mimic the effects of glucose tostimulate expression of the preproinsulin (PPI) gene. Conse-quently, insulin secretion plays a central role in the activationof gene expression by glucose in the � cells. Hyperglycemiastimulates the transcription of the PPI gene and of at least 20other genes in � cell to ensure normal insulin production andrelease. In single � cells, PPI promoter induction is suppressedif insulin secretion is blocked, whereas the blockade of pro-moter induction is reversed if insulin is replaced in the cells.Inactivation of the insulin receptor leads to a significant sup-pression of glucose and/or insulin stimulation of the PPI pro-moter with no evident effect on the IGF-1 receptor expression.Inactivation of IGF-1 receptor has no impact on glucose and/or

insulin stimulated PPI gene transcription, demonstrating thespecificity for insulin. Inhibition of type I PI 3K activity byoverexpression of the regulatory �p85 subunit blocks glucose-induced expression of the PPI and liver type pyruvate kinasepromoters. Expression of constitutively active PI 3K catalyticsubunits mimics the effect of glucose on PPI promoter activ-ity. Their results suggest that the effects of glucose may bemediated in part through an autocrine action of released insulinon � cell insulin receptors.

Insulin receptor signal transduction plays a critical rolein regulating � cell function, notably the acute first-phase in-sulin release in response to glucose. Michel Bernier (Balti-more, MD) presented evidence that � cell insulin resistancemay be related to abnormal regulation of tyrosine phosphory-lation events associated with an alteration in protein tyrosinephosphatase (PTPase) expression and/or activity. A subset ofPTPases is implicated in the regulation of insulin signaling ma-chinery in � cells, including regulation of the insulin receptorand insulin receptor substrates (IRS)-1 and -2. Insulin bindingto the � cell insulin receptor triggers the receptor autophos-phorylation, leading to subsequent phosphorylation of manycellular substrates on tyrosine residues including IRS-1 andIRS-2. The formation of competent signaling complexes at theplasma membrane then initiates a sequence of phosphorylationevents critical for processes as diverse as metabolism, proteinsynthesis, cell growth and proliferation, and resistance to ap-optosis. In vivo, various types of tissue-specific knockout micehave shown the importance of insulin signaling at the level ofthe � cells for glucose-stimulated insulin secretion, indicatinginsulin resistance in � cells related to abnormal regulation oftyrosyl dephosphorylation. He suggested that the design of se-lective PTPase inhibitors might be of therapeutic value in themanagement of disorders associated with insulin resistance.PTPases present in the � cells, and their function and specific-ity were also discussed.

NEW FINDINGS IN ISLET-CELLSIGNALING NETWORK

Insulin secretion is mediated by complex signal trans-duction events and unraveling this signaling system is emi-nently important for understanding � cell biology. John D.Scott (Portland, OR) pointed out that a fundamental principleof signal transduction is that parallel signaling pathways areassembled from a common repertoire of enzymes, and are ableto propagate diverse physiological responses. The key featureof this mechanism is that separate signaling pathways are or-ganized into localized transduction units, each tailored to re-spond optimally to a particular signal. One of the key events isprotein-protein interaction maintained by anchoring adaptersand scaffolding molecules that provide molecular glue. Suchproteins hold signal transduction units together. Dr. Scott dis-cussed some new experimental tools used by his team. Theycombine both gene-array and molecular biologic technologies



in gene expression analysis for the dissection and manipulationof intracellular molecular events and the disruption or inhibi-tion of protein anchoring. These types of highly sophisticatedand broad ranging biochemical approaches were useful toolsfor their investigations in dissecting scaffolding molecules todevelop functional domains. He concluded that this type ofinvestigation should provide answers for biologic questions re-lated to signaling network and contribute to the understandingof dynamic biologic activities.

Ravi Iyengar (New York, NY) utilizes both computa-tional models and laboratory experiments and integrates theresults to unravel complex cellular events. He demonstratedthe value of such approaches in his presentation on “SignalingNetworks and Coordinated Regulation of Cellular Re-sponses.” In secretory cells, the activity of intracellular ma-chines, involved in a transcriptional or translational apparatus,must be controlled to optimally respond to extracellular sig-nals. This is achieved by a signaling network that can receive,process, and integrate these signals and also can control differ-ent cellular machines. Localization of signaling events to a de-fined intracellular space contributes to the selective regulationof cellular machines. Using computational models and bench-top experiments, his team has investigated the MAP-kinase1,2/protein kinase C network. They have shown that the sys-tem is flexibly designed to operate as either a monostable orbistable system, depending on the concentration of MAP-kinase phosphatase. Brief stimulus with low concentrations ofMAP-kinase phosphatase induces a bistable behavior and sus-tained MAP-kinase phosphatase activation, whereas higherlevels of stimulus induce a monostable state, which regulatesacute responses to stimulus without sustained activation. Theyalso found that selective forms of the cAMP-phosphodiester-ase type 4 (PDE4) can regulate the microdomains of cAMPconcentrations in response to extracellular stimulation, whichin turn regulate the activity of protein kinase A. It is the balanceof MAP-kinase and protein kinase A effects PDE4 activitiesthat define the boundaries of cAMP microdomains and the du-ration of local and distal protein kinase A action.

For a number of years, Melanie H. Cobb (Dallas, TX)and her team have investigated the regulation and function ofMAP kinases in � cells. She provided several lines of evidencesupporting her conclusion that the MAP kinases ERK1 andERK2 are important components of the signaling mechanismsthat control homeostasis in � cells. ERK1 and ERK2 contrib-ute to both basal and glucose-stimulated insulin gene transcrip-tion. Of several transcription factors that bind to the insulinpromoter, ERK2 phosphorylates the major insulin transcrip-tion factors Beta2/NeuroD1 and PDX1. Phosphorylation ofBeta2 at serine 274 by ERK1 and ERK2 is required for itsmaximal transactivation activity in response to glucose. Simi-larly, PDX1 transactivation activity is reduced by mutating the2 sites phosphorylated by ERK2 and by adding a MEK inhibi-tor. The mechanisms of ERK1 and ERK2 activation in � cells

appear to include a Ras family small G protein, B-Raf, andMEK1. ERKs are activated in a calcium-dependent mannereven at subthreshold glucose concentrations, and GLP-1 alsoactivates ERK2 through a PKA-dependent mechanism.

Morris F. White (Boston, MA) reported that insulin-receptor substrate proteins were involved in the developmentof � cell dysfunction. Irs2 −/− knockout mice show insulinresistance, progressive glucose intolerance, � cell loss and de-creased islet insulin contents, followed by the development ofovert diabetes. Overexpression of the Irs2 gene in the � cells ofnormal mice has no effect on glucose metabolism, but in-creased glucose-stimulated insulin release islet cell mass andinsulin content in vivo. Irs2 overexpression in the � cells ofIrs2 −/− knockout mice resulted in increased blood insulinlevels, increased food intake and weight gain (up to 30%), andnormal insulin tolerance. However, the mice remained insulinresistant. The Pdx1 expression in the pancreas decreased be-fore the onset of diabetes in Irs2 −/− knockout mice as com-pared with that in wild-type or Irs1 knockout mice. The ex-pression of Irs2 in the islets upregulated the Pdx1 level in bothwild-type and Irs2 knockout mice. These results suggest thatPdx1 is regulated by the Irs2 branch of the insulin/IGF signal-ing system. Examination of Pdx1 gene dosage effects in Irs2knockout mice have revealed that Pdx1 haploinsufficiency ac-celerates diabetes onset (newborn versus 8–10 weeks), whiletransgenic expression of Pdx1 restores the � cell mass andfunction and prevents the onset of diabetes. Thus, Dr. Whitesuggested that dysregulation of Pdx1 expression by functionalmechanisms might be one of the common links between ordi-nary type 2 and MODY diabetes.

� CELL BIOLOGY AND PROMOTION OF �CELL GROWTH

Daniel Pipeleers (Brussels, Belgium) overviewed sev-eral lines of ongoing research at the Diabetes Research Centerof the Free University of Brussels. Their studies, using single �cell preparations, showed a high heterogeneity in response toglucose challenge. Glucose responsiveness was also found tobe age-dependent and � cells from donors older than 50 yearsof age showed significantly diminished insulin responses.Such age-related changes may be associated with amyloid de-posit and lipid droplets in the islet cells. In the discussion oftheir islet isolation results, Dr. Pipeleers mentioned the pres-ence of a considerable number of insulin-positive cells scat-tered as small cell clusters, some of them in association withductal structures, which were usually lost after islet purifica-tion. He also explained the role of ductal cells in � cell neo-genesis based on their observation of insulin-positive cellsemerging from the duct epithelia when islets were transplantedinto athymic mice.

Several other speakers discussed the presence of � cellprecursors in the pancreatic duct epithelium and their role ingeneration of � cells. Riccardo Perfetti (Los Angeles, CA)



observed, both in vitro and in vivo, � cell neogenesis from theduct epithelium in islet culture and islet recipients treated withglucagon like peptide-1 (GLP-1). He illustrated the effect ofGLP-1 on islet-cell regeneration, as well as on the inhibition of� cell apoptosis. Based on these findings, he suggested that theincreased � cell mass mediated by GLP-1 is due to a combi-nation of its effect as a growth factor and apoptosis inhibitor.His studies have also shown that GLP-1 enhances the tran-scription of various genes, including insulin, glucose trans-porter, and hexokinase-1 and also increases insulin secretion ina glucose-dependent fashion. In vivo, GLP-1 has been shownto improve � cell function in aging and obese animals.

Aaron I. Vinik (Norfolk, VA) discussed a factor thatinduces islet stem cell differentiation. INGAP, a factor recog-nized in their earlier investigation on regenerating pancreatictissue in the hamster, is found to improve and/or reverse STZ-induced diabetes in rats. A 15-amino-acid sequence withinINGAP is identified as a biologically active core that stimu-lates ductal cell proliferation and differentiation into � cells.INGAP is also thought to prevent � cell apoptosis.

Alberto Hayek (LaJolla, CA) discussed his team’s workon islet proliferation from human fetal pancreata and adult is-lets. Endocrine cells in the human fetal pancreas have beenshown to develop from Ngn3-positive cells, and cells co-stained with glucagon and insulin are often detected in ductepithelia. During normal development, insulin-positive cellsmigrate from the duct, aggregate, and form islets. In Dr.Hayek’s laboratory, islet-like cell clusters (ICC) were con-structed by aggregating fetal pancreatic cells prepared by en-zymatic dissociation. Twenty percent to 25% of ICCs con-structed from a 16- to 20-week pancreas are endocrine cells,and the remaining 60% to 70% are undifferentiated cells. Dr.Hayek’s group found that, in vitro, fetal cells did not surviveeven with the help of various growth factors, but in vivo inathymic mice, ICCs continued to differentiate. After 3 monthsin athymic mice, the insulin content of the ICC graft becameequivalent to that of adult islets. However, despite their abilityto grow, mature, and perhaps acquire normal function, Dr.Hayek pointed out that fetal ICCs do not meet the current FDAregulation in 2 areas: (1) multiple donors are required to con-struct sufficient ICCs for glycemic control and (2) fetal ICCsare incapable of responding to high glucose with insulin re-lease at the time of transplantation.

To promote functional maturation of ICCs in vitro, histeam tested a variety of growth factors and culture methods,but without success. Nicotinamide clearly promotes ICC dif-ferentiation but the effect is insufficient; betacellulin increasesprecursor proliferation; activin enhances � cell differentiationbut without effect on functional maturation; exendin-4 in-creases PDX1 expression but has no effect on insulin produc-tion. However, when ICC recipients were treated for 10 days,exendin-4 significantly increased � cell mass and shortenedthe period required for functional maturation of fetal � cells.

Thus, ICCs in the exendin-4 treated mice released insulin inresponse to high glucose stimulation 8 weeks after transplan-tation, approximately 4 weeks earlier than the untreated con-trols. Unlike fetal ICCs, adult human islets cells grew well inmonolayers when hepatocyte growth factor (HGF) was addedto the culture medium, but the insulin expression was lost inthe replicated cells. Even when monolayer-grown cells wereaggregated, these cells failed to express insulin in vitro or invivo following transplantation.

In the search for � cell precursor marker(s), Dr. Hayek’steam has conducted a series of investigations using fetal hu-man pancreata to determine whether nestin is a potential can-didate marker. Their results provide several lines of undeni-able evidence demonstrating that nestin is not a marker for �cell precursors. In the human fetal pancreas, nestin does notcolocalize with insulin. Nestin is totally negative in pancreaticepithelial cells, and its expression is only found in the associ-ated mesenchyma. To further confirm this, the team isolated apure population of nestin-positive cells from a 24-week pan-creas, expanded them ex vivo, and then used the cells to con-struct ICCs. These ICCs were transplanted into athymic miceand removed after 12 weeks. Mice were bled at the time ofgraft removal for human C-peptide testing. Immunostainingrevealed the total absence of endocrine cells, and no C-peptidewas detected in the blood. These results demonstrate that nes-tin is not a marker for pancreatic endocrine cells.

RECENT PROGRESS IN � CELL GENERATIONFROM STEM CELLS AND BY BIOENGINEERING

One of the major problems that will prevent the applica-tion of islet transplantation in a wider range of type 1 diabeticpatients is clearly the shortage of donor islets. Generating re-newable sources of � cells is 1 solution for this problem. In thiscontext, the use of either embryonic or adult stem cells repre-sents an interesting alternative.

Bernat Soria (Alicante, Spain), one of the leading in-vestigators in this field, presented several approaches that havebeen used to differentiate stem cells into insulin secreting cells,with special focus on the “gating technology” used in theirlaboratory. He stated that the key is knowing how to reprogramthe information of the nuclei and how to select cells expressinga desired phenotype. Because of their plasticity, embryonicstem cells (ESCs) need to be coerced into becoming a certaincell type. The method used in Dr. Soria’s laboratory was to firstgenerate ESC lines and then transfect them with a constructthat expresses the neomycin selection system under the controlof the regulatory regions of human insulin gene. The constructalso contains a hygromycin resistance gene under the controlof the constitutive phosphoglycerate kinase promoter to allowthe selection of single clonal ESC during the expansion stage.The selection strategy was combined with a final maturationprocess in Petri dishes to form cell aggregates. Lowering glu-cose concentration and adding nicotinamide during this stage



were instrumental for phenotypic changes. Clones producingvariable amounts of insulin were obtained, but only 1% of ap-proximately 800 clones showed any insulin production andonly 2 clones produced as much as 0.5–0.8 ng insulin/µg ofprotein. When aggregates were implanted into the spleen ofSTZ-diabetic athymic mice, the best insulin-containing clonereversed hyperglycemia in approximately half of the recipientsand the transplanted cells were detected in the spleen and liver.Thus, the principle of their method was shown to work. Strat-egies similar to their method were used by other investigatorsfor stem cells from bone marrow, endoderm, and human intes-tinal epithelia with certain amounts of success. Dr. Soria con-cluded that increasing the insulin production in higher num-bers of clones was the next major problem to be solved.

Development of surrogate � cells using bioengineeringtechniques is another potential source for cell therapy for dia-betes. Christopher B. Newgard (Durham, NC) is a pioneer inthis field and has conducted a series of investigations for thepurpose of obtaining detailed information of � cell function asa prerequisite for the development of surrogate � cells. In thissymposium, he discussed the studies conducted by an interdis-ciplinary approach, focusing on the biochemical mechanismsthat increase insulin secretion in response to metabolic fuel andthat protect � cells from immune destruction. Using INS-1 celllines with different glucose responsiveness, these investigatorshave shown differences in the expression profiles of a series ofgenes, an important role for pyruvate carboxylase-mediatedpyruvate cycling pathways for glucose-stimulated insulin re-lease, and specific impairment of pyruvate cycling in islets ex-posed to elevated lipid concentrations. To identify genes thatprotect islets from the toxic effect of cytokines and/or reactiveoxygen species (ROS), Dr. Newgard and his team exposedINS-1 cells for an extended period to the cytokines IL-1� andIFN-� to select particular resistant cell populations. Then, spe-cific genes expressed by the resistance cell populations weredetermined. Bcl-2 was found to provide clear protectionagainst cytotoxicity mediated by ROS but relatively poor pro-tection against cytokine damage. Cells selected by the cyto-kine-selection procedure were protected strongly against cyto-kines and weakly against ROS. When bcl-2 expressing cellswere exposed to the cytokine selection procedure, the selectedcell lines gained protection against both ROS and cytokines.These results suggested that the selection process resulted incytokine resistance via a different pathway independent of thepresence or absence of background bcl-2 expression. The cy-tokine-selected cells without bcl-2 expression strongly sup-pressed the iNOS expression and NO production in response tocytokines, while this suppression did not occur in the selectedcells with bcl-2 overexpression.

Pamela R. Itkin-Ansari (La Jolla, CA) reported on thegenetically engineered pancreatic endocrine cell lines devel-oped in her laboratory. These cell lines were generated to ex-press growth stimulatory genes using cultured adult human �

cells. These cells proliferated indefinitely, but growth stimu-latory genes had an adverse effect on differentiated cell func-tion. Restoration of � cell differentiation was achieved in 1 cellline by expressing PDX1 and stimulating multiple signalingpathways through cell-cell contact and the GLP-1 receptor.This cell line was able to restore � cell function and reverseSTZ-induced diabetes in athymic mice.5 In a cell line devel-oped from fetal islets, the expression of PDX1 and cell aggre-gation led to somatostatin expression,6 but subsequent over-expression of NeuroD1 in these cells induced insulin gene ex-pression. These results indicated that PDX1 and NeuroD1 actin synergy on native chromatin to promote insulin gene expres-sion. In addition, the IAPP gene, which was not a knowndownstream target of NeuroD1, was also induced. NeuroD1suppressed somatostatin expression, reminiscent of the switchin hormone expression during islet regeneration in vivo, whereNeuroD1 is localized in � cells but not in � cells. Thus, resultshave suggested that NeuroD1 may be a component of a geneexpression and/or transcription factor cascade that is involvedin hormone switching and terminal differentiation of endo-crine cells. The team has used this model for studying mecha-nisms and manipulation of endocrine cell fate in the humanpancreas.

ADVANCES IN UNDERSTANDINGDIABETES AUTOIMMUNITY

Type 1 diabetes is an autoimmune disease tightly asso-ciated with the HLA class II genes and the presence of anti-bodies against self-antigens. Both CD4 and CD8 cells are in-volved in the pathogenesis of the disease, but the mechanismsof disease pathogenesis have remained obscure. However, it isknown that type 1 diabetes is regulated by the genetic back-ground of individuals and by the environment. Linda S.Wicker (Cambridge, UK) presented her studies on the searchfor potential genes involved in the development of diabetes inNOD mice, an extensively studied mouse model for thehuman disease. Many such susceptibility genes, termed Iddgenes, have been assigned to specific regions of the genome.Transgenic NOD mice expressing the Idd3 or Idd5 genes havereduced insulitis and are able of restoring self-tolerance. Idd3expression also reduces the development of autoantibodies di-rected against insulin and islets and increases transplantationtolerance. The candidate genes for Idd5.1 are the ctla4 or icosgenes, both encoding for molecules that are critical to the func-tion of T cells. Interestingly, it seems that sequence variationsof the ctla4 gene exist between susceptible and resistant mousestrains, which may alter the expression of ctla4 isoforms pro-duced during immune responses. Other candidate Idd genes,such as 4-1BB, FRAP, and Idd10, may also be involved in thenormal function of T cells. Idd gene identification may in-crease the chances of finding 1 or more targets for therapies. Itis also likely that combination therapy that involves targeting



more than 1 critical pathway leading to diabetes could hold themost promise in treating or preventing the disease.

ISSUES IN CLINICAL ISLET TRANSPLANTATIONSince 1993, pancreatic islet and other types of somatic

cell therapy have become the subject to regulation by the U.S.Food and Drug Administration (FDA) as biologic products.7

The new era of islet transplantation also has brought increasedgrant support by the National Institutes of Health (NIH) andJuvenile Diabetes Research Foundation (JDRF) for both isletisolation and clinical procedures. The NIH also supported thedevelopment of a Collaborative Islet Transplantation Registry(CITR) for collection of clinical transplantation data and hassupported the development of a national program for islet iso-lation and distribution for clinical transplantation trials termedIslet Cell Resources (ICR) Centers. Ten such centers werefunded in 2001. Program officers were invited to cover theseareas in clinical islet transplantation.

Regulatory Issues in ClinicalIslet Transplantation

The FDA has published its expectations in detail regard-ing investigational new drug (IND) application for allogeneicislets.8 Darin J. Weber from the Division of Cellular andGene Therapies and Division of Biologic Evaluation and Re-search (CBER) of the FDA (Rockville, MD) discussed theregulatory issues that the FDA had encountered in the reviewprocess of more than 30 IND applications submitted for islettransplantation. Many of these IND applications had beenplaced on “clinical hold.” The hold for a phase 1 clinical studymeans that a given islet IND submission was considered defi-cient in safety information in 1 or more areas, such as manu-facturing, preclinical data, and/or clinical protocol design.Common reasons included (1) failure to submit data on thepreparation of high quality islets, (2) lack of data supportingpreclinical data for novel combinations of immunosuppressivedrugs or sites of implantation, and (3) deficiency in followinggood clinical practices (GCP). The FDA has been taking aflexible “stepwise” approach for the application of the regula-tory requirements. Since safety is of vital importance in aphase 1 study, the assurance of aseptic manufacturing and ap-propriate microbiological safety testing is imperative. Assum-ing safety considerations are appropriately addressed, it is notobligatory to prepare the product in a facility that is fully com-pliant with the current good manufacturing practices (cGMP),nor is it expected that the product will be fully characterized orthat the manufacturing process will be optimized prior to thephase 1 study. Since the products are still under developmentin the IND process, it is expected that the manufacturing pro-cess will continue to be optimized, methods to fully character-ize the product will be undertaken, and product specificationswill be refined based on data collected. For early-stage clinicalstudies, changes to the manufacturing process are acceptable

when they are minor and are likely to improve the safety with-out changing the inherent characteristics of the final product.For more significant changes, such as moving from a freshproduct to one that has been cultured, the FDA may ask that anew IND application be submitted. When changes in productmanufacturing are proposed, data should be available toclearly demonstrate how the changes affect the final productprepared under the different processes, in regulatory parlancereferred to as “product comparability.” The FDA also recog-nizes that some flexibility in islet preparation protocols may benecessary due to the inherent variability of donor pancreata.However, it is imperative that islet IND sponsors collect suf-ficient manufacturing data during ongoing clinical studies todemonstrate product comparability. As product developmentand clinical trials advance, increased compliance with lot re-lease testing, product characterization and cGMP require-ments must be implemented. Issues associated with isletpreparation must be resolved prior to the submission of a bio-logics license application (BLA). It is expected that only asingle, well-defined manufacturing process supported by datawill be chosen that meets established lot release specificationsby the time a BLA is submitted.

Thomas Eggerman (Bethesda, MD) from the NationalInstitute of Diabetes and Digestive and Kidney Diseases ex-plained the mission of the NIH-supported national Collabora-tive Islet Transplantation Registry. CITR was established inSeptember 2001, to expedite and promote the progress andsafety in islet and/or � cell transplantation through the collec-tion, analysis, and communication of comprehensive and cur-rent data on all islet and/or � cell transplantations performed inthe United States and Canada. The coordinating center is lo-cated at the EMMES Corporation, which provides support forlogistics, data capture, quality control monitoring, statisticaldesign and analysis, and other registry activities. Inclusion ofthe word “collaborative” in the name of the registry empha-sizes the importance of collaboration among all areas that havea connection to islet transplantation, including collaborations(1) among transplant centers; (2) with other initiatives, pro-grams, and networks of the National Institutes of Health, e.g.,the Human Pancreatic Islet Cell Resource, the Immune Toler-ance Network, U.S. Pancreas Transplant Registry, and the Ju-venile Diabetes Research Foundation; (3) with the diabetescare community, the health insurance industry, the Centers forMedicare and Medicaid Services, Health Resources and Ser-vices Administration, and the FDA; (4) with United Networkfor Organ Sharing (UNOS) and the Canadian Organ Replace-ment Register (CORR); and (5) with Islet Transplant Registry(ITR) in Giessen. Dr. Eggerman also talked about membershipcriteria, CITR organization, and the progress made in the de-velopment of web-based forms for data collection, includingthe CITR Internet Data Entry System for quality controlchecks. CITR took steps to avoid duplication of efforts at theTransplant Centers, such as data on U.S. cadaver donors that



could be directly imported from UNOS, and developed a shortRetrospective Summary Form to capture islet transplant dataperformed between January 1996 and December 1998 to fillthe gap between data collected by the ITR and CITR. CITRalso closely works with the Coordinating Center for the 10 ICRCenters sponsored by the National Center for Research Re-sources, where detailed data on islet isolation and processingtechniques are expected to be collected. CITR also assists theU.S.-based sites in making use of the resources provided by theGeneral Clinical Research Center (GCRC) program of the Na-tional Center for Research Resources (NCRR) to develop in-house islet transplant databases and to facilitate data transfersbetween the center and CITR. CITR invited 17 transplant cen-ters to register that are either currently performing or plan toperform islet transplantation.

Richard Knazek from the National Center for ResearchResources (Bethesda, MD) explained the role of NCRR-supported Islet Resources Centers in the advancement of islettransplantation. The Islet Cell Resources Centers are fundedby the NIH and JDRF and comprise an interactive group ofacademic laboratories charged with the pursuit of 2 majorgoals: (1) to provide pancreatic islets to investigators for use inFDA- and IRB-approved clinical transplantation and (2) to op-timize the harvest, purification, function, storage, and ship-ment of islets and to develop tests that characterize islet qualityand predict the effectiveness of islets after transplantation intodiabetic patients. The ICR Steering Committee reviews andapproves applications for the use of islets in clinical transplan-tation trials and allocates the supply of islets to investigatorsbased on the scientific merit and feasibility of proposed stud-ies. Investigator eligibility for receipt of ICR islets and contactinformation are detailed in the ICR Policy and Proceduresdocument, which is available online (www.ncrr.nih.gov/clinical/cr_icr.asp).

Larry A. Couture (Duarte, CA) discussed issues in-volved in designing and operating a multibiotherapeutic prod-uct facility in an academic environment through his extensiveexperience with the Center for Biomedicine and Genetics(CBG) facility at the City of Hope National Medical Center.The CBG is a 22,000-square foot manufacturing facility thatcomprises 3 production zones of 12 production rooms andsuites and a variety of support areas. This facility currentlyserves as the islet production site of the Southern CaliforniaIslet Cell Resources Center and also contains sites for otherclinically ready activities including the production of plasmidDNA, adenoviral vectors, monoclonal antibodies or other re-combinant proteins, and for ex vivo cell manipulations. Dr.Couture emphasized that academic institutions should be ableto establish the infrastructure to allow translation of virtuallyany biotherapeutic technology from the bench-side into theclinical setting without primary regard for size of patient popu-lation, market viability, or other factors critical to the commer-cial enterprise operation.

Update in Islet Isolation andCulture Technologies

Rodolfo Alejandro (Miami, FL) reviewed the islet iso-lation process step by step and pointed out new technologiesthat have recently been introduced. For pancreas preservation,the 2-layer method, developed for cold storage of the pancreasby Yoshikazu Kuroda and his colleagues at the Kobe Univer-sity9 has been shown not only to preserve better islets but alsoassist in the recovery of islets from stress.10 This method pre-serves harvested pancreata for longer periods, permits ship-ment of a donor organ from a distant site, and also providessome flexibility in the isolation schedule. In Miami, pancreatashipped in University of Wisconsin solution was often trans-ferred and maintained by the 2-layer method for an extendedperiod before initiating islet isolation. Their collaboration withthe Islet Transplant team at the Baylor College of Medicine inHouston, TX, is a good example of the advantages provided bythis method. Another major reason for the success is the as-signment of a single surgeon to the organ procurement and thusemphasis on the importance of organ harvesting and handlingprocedures.

Dr. Alejandro also described Miami’s experience of sig-nificant lot-to-lot variability of Liberase, the enzyme solutionused for pancreas digestion, and said that it is difficult to iden-tify good lots. The Miami group uses 70,000 U/vial of neutralprotease and 2000 Wunsch U/vial as a guideline for the selec-tion. Although the concept of the islet isolation process hasremained the same, there are different preferences in almostevery step of islet processing from 1 center to another and 1investigator to another. For example, there are various opin-ions on the addition of DNase and/or Pefabloc to the Liberasesolution, using a syringe versus the perfusion method for pan-creas distention, processing the whole pancreas versus sec-tioned tissue for digestion, using manual versus machine shak-ing in the digestion process, using discontinuous versus con-tinuous gradient centrifugation for islet purification, and usingOptiPrep (Accurate Chemical & Scientific Co., Westbury,NY), a new reagent for islet purification in the United States. InMiami, most of isolations require “islet rescue,” involving asecond discontinuous density gradient, following the firstCOBE purification.11 Individual preference also extends totransplantation procedures, such as transplantation of freshversus cultured islets and the use of a bag versus syringe forislet infusion into the portal vein.

The Memphis group originally developed a serum-freeculture medium for human islets. A. Osama Gaber (Mem-phis, TN) explained that the purpose of culturing islets is todevelop technology that permits the collection of islets from allavailable pancreata, regardless of their quality and islet yields.Since pancreata rejected for whole organ or islet transplanta-tion (discarded pancreata) are expected to be of poor qualityand provide low islet yields, islets harvested for these organs



must be accumulated by culturing or cryopreservation. In hiscenter, the islet recovery after culture (>50% after 2–3 months)is better than those after cryopreservation. Islets cultured forextended periods stained faintly or negative for insulin andother hormones. However, when these islets were transplantedinto SCID/NOD mice, 39 of the 58 grafts were viable withdetectable amounts of C-peptides in recipient blood. Isletsfrom discarded pancreata function better when transplanted af-ter having been maintained in culture than if freshly trans-planted. The tradeoff of culturing islets is a slow recovery offunction in vivo, after transplantation. Dr. Gaber reported thereversal of diabetes was achieved in all cases with islets cul-tured for 1 month and over 70% with those cultured for longerperiods.

Pre- and Post-transplantation Islet QualityIslet transplantation success depends greatly on the

health of the � cells. Gordon C. Weir (Boston, MA) reportedon the potential damage to islets and/or � cells throughout theislet transplant process. Like many other investigators, Dr.Weir noticed that some assays used for the assessment of isletquality were inadequate, and standardized assessment meth-ods were necessary for prediction of posttransplantation isletperformance. His isolation center routinely performs morpho-logic studies of 500–800 islets using light microscopy or elec-tron microscopy. These studies provide valuable and interest-ing information, such as distorted mitochondria, presence oflipid droplets and amyloid deposit, and differences in �- andnon–� cell ratio. Measurements of oxygen consumption(O2/DNA) by islet preparations have also shown some prom-ise. He suggests that proteonomics, gene arrays (especially thedetection of the key stress gene expression), nuclear magneticresonance, calcium imaging, and mitochondrial membrane po-tential measurements are also potentially useful tools for isletevaluation. Islet damage during transplantation is primarilydue to hypoxia, glucose toxicity, and the generation of inflam-matory products. Hyperglycemia increases oxygen consump-tion, and a loss of small vessels and delayed angiogenesis ac-celerates hypoxia. Chronic hypoxia may also reduce � cellnumbers. In islet transplants, the normal intraislet cell-cell re-lationships (such as � cells residing downstream of the bloodsupply from � cells) may be destroyed, causing altered insulinsecretory responses to glucose. Transplanted � cells understress would also be subject to glucose toxicity, since the geneexpression profiles of � cells are known to change under stress.Such changes would likely contribute to a reduction in glucoseresponsiveness and an increased susceptibility to apoptosis.Dr. Weir suggested the use of arginine stimulation tests forassessing post-transplantation � cell functional assessments.

Arne Andersson (Uppsala, Sweden) also pointed outthe importance of developing strategies to reduce harmfulevents on transplanted islets. He suggested that treatment withgrowth factors would be an effective method for increasing �

cell mass in the graft, and in vitro studies provide usefulscreening methods for selecting growth factors for in vivostudies. He produced data on graft site–dependent factors,which control posttransplantation graft growth, and donorage–dependent � cell replication. Following transplantation,endothelial cells in the islets dedifferentiate. The nerve cells,which initially survive, eventually disappear and are replacedmainly by sympathetic nerves located along the newly formedblood vessels. Previous studies have shown that grafted isletsrevascularize and that the density of newly formed blood ves-sels is similar to that in the endogenous islets. However, Dr.Andersson’s team found that the newly formed vessels weremostly located around the islets in the connective tissue sepa-rating islets from each other, with very few vessels found in-side of the islets. In the pancreas, the oxygen pressure in theislets is double that of the exocrine parenchyma. Immediatelyafter transplantation, under the kidney capsule, the oxygenpressure in mouse (and human) islets is about 25% of that mea-sured in native islets and remains at this level for over 6months. Under hyperglycemia, the oxygen pressure is shownto further decrease. Although hyperglycemia and hypoxia perse can increase insulin mRNA levels in cultured islets, the graftvascular dysfunction and hypoxia may likely lead to somemetabolic dysfunction or functional alteration of transplantedislets. Using a microdialysis technique, the grafted islets dis-play an increased degree of anaerobic metabolism, 3 timeshigher than that of cultured islets, as detected in his studies.

Recent Progress and Future Issues in ClinicalIslet Transplantation

Bernhard J. Hering (Minneapolis, MN) updated the re-sults of the NIH Immune Tolerance Network trials and dis-cussed new developments in immunosuppressive regimensthat may have potential in islet transplantation. Since 1999, 29islet allografts were performed by 9 institutions of the ITN. Allbut 2 grafts have lowered insulin requirements, and 25 patientsare insulin free for more than 1 year. The total islet numbersused for transplantation ranged from 8000 to 15,000 IEQ/kg ofrecipient body weight. Among the 25 patients, 5 received isletsfrom a single donor and the remaining from 2 to 3 donors. Theuse of the 2-layer method for pancreas preservation and cul-tured islets for transplantation has become a common proce-dure. Since islet autotransplants with >300,000 IEQ haveachieved over 15 years of insulin independence in depancre-atectomized patients, islet allografts should work if optimalimmunosuppressive regimens that have minimal adverse ef-fect on islet function are developed to control allo- and auto-immune responses. Results first reported by the Edmontongroup2 and subsequently confirmed by other institutions indi-cate that sequential islet allotransplantation from 2 to 3 donorsconsistently reversed insulin dependency in type 1 diabetic re-cipients treated with a steroid-free immunosuppressive regi-men that included interleukin-2 receptor antibody (dacli-



zumab), sirolimus, and a reduced dose of tacrolimus. This regi-men has been tested in all 9 ITN centers. Dr. Hering discusseda long list of newer agents as substitutions for diabetogeniccalcineurin inhibitors, e.g., cyclosporin A and tacrolimus, thatmay be worth testing toward the goal of achieving restorationof insulin independence with single donor islet transplantation.Such nondiabetogenic agents should preferably work syner-gistically with sirolimus or everolimus. Candidate agents in-clude mycophenolate mofetil, FTY720, CTLA4-Ig, LEA29Y,and anti-CD154 monoclonal antibody. Strategies that focus onperitransplant deletion or inhibition of autoreactive and allore-active T cells as well as the inhibition of macrophage activa-tion are necessary for enhancement of islet engraftment. Can-didate agents for these purposes include the FcR nonbindinganti-CD3 antibody, hOKT3�1 (AlaAla), polyclonal T-cell an-tibody and anti-CD52 monoclonal antibodies. A third strategyhe suggested is peritransplant administration of inhibitors ofinflammatory cytokines. Protocols to facilitate peripheral tol-erance induction by temporary immunosuppression have beeninvestigated to instigate activation-induced cell death of anti-donor and anti-self T cells to induce regulatory T cells thatcontrol memory T cells and to monitor these critical events.Strategies for induction of central tolerance may become morefeasible with costimulatory blockade and sirolimus. Althoughacute complications have been minor, the experience withnewer protocols is limited and monitoring needs to be ex-tended to assess long-term risks.

R. Paul Robertson (Seattle, WA) discussed metabolicissues in islet transplantation based primarily on clinical dataobtained from islet autotransplantation. Since 1980, islet auto-transplantations were successfully performed in patients withchronic, painful pancreatitis, following total pancreatectomy.In these patients, stable euglycemia was achieved with300,000–500,000 IEQ showing that a half million islets aresufficient to maintain intact acute insulin responses to intrave-nous glucose challenge and normal levels of HbA1c for up to16 years. Even 125,000 IEQ were able to induce partial butgood glucose control, a number well below the estimated 8000IEQ/kg currently required for allotransplantation in type 1 dia-betes. Thus, the challenge in islet allotransplantation is to re-duce the islet number required to achieve euglycemia to500,000 IEQ. Possible reasons for such differences includepreexisting autoimmunity, the use of cadaveric pancreata (in-volving long ischemic times), alloimmunity, and/or the use ofimmunosuppressive agents. He stated that the 2-layer pancreaspreservation method could preserve and/or improve islet func-tion as indicated by post-transplantation C-peptide levels ob-tained with a smaller islet mass. Dr. Robertson and his teammeasured �-cell function during hypoglycemia by step-clampfunction tests in islet-transplant recipients. In both auto-andalloislet recipients, there was no increase in glucagon duringhypoglycemia, despite the presence of glucagon in the hepaticvenous blood (indicating the presence of � cells in the grafts).

Similarly, negative glucagon responses were obtained in dogstransplanted with islets into the liver. In contrast, glucagon lev-els increased in dogs transplanted with islets into the peritonealcavity. Epinephrine response did not improve in islet recipi-ents with type 1 diabetes even after years of good glucose con-trol. In contrast, type 1 diabetic patients transplanted with awhole pancreas showed good glucagon responses and attenu-ated epinephrine responses. This raises a question about unde-tected hypoglycemic unawareness in islet transplant recipi-ents. In addition, he pointed out that immunosuppressiveagents are present in the liver at higher concentrations. Someof the existing results are inexplicable at present. Dr. Robert-son also raised the question of whether the liver is an optimalsite for islet grafts.

Induction of Specific Immune Tolerance:Potential Approaches Developed inAnimal Studies

Ali Naji (Philadelphia, PA) emphasized the importanceof maintaining active, open-minded basic research for fore-casting and solving emerging problems in clinical islet trans-plantation by referring to earlier animal studies that revealed anumber of key issues in clinical islet transplantation. Dr. Najipointed out that autoimmunity presents unique immunologicproblems including (1) need for immunosuppressive regimensdirected for both auto- and alloimmune responses, (2) majorhistocompatibility complex (MHC) compatibility between do-nor and recipient is beneficial for immune reaction directedagainst alloantigens but does not improve graft susceptibilityto autoimmunity, and (3) the presence of autoantibodies influ-ences graft survival. Culturing islets prior to transplantation isknown to reduce islet immunogenicity and consequently thealloimmune reaction, but it would have little effect on autoim-mune islet destruction. The thymus presents a microenviron-ment for islet survival and protection from both allo- and au-toimmunity. The Philadelphia team is currently investigatingintrathymic islet allotransplantation in a nonhuman primatemodel. Because autoimmune diabetes is T cell–mediated, theparticipation of B lymphocytes in the development of autoim-mune diabetes has long been neglected. Dr. Naji suggested apotential role of B cells in antigen presentation, since diabetesdevelops in only a very low percentage (1 out of 400) of NODB cell knockout mice.

The goal of transplant immunology is to induce specificimmune tolerance for permanent survival of allografts. Soji F.Oluwole (New York, NY) discussed experimental strategiesfor tolerance induction that may have potential for clinical ap-plication. Mechanisms involving T-cell tolerance are dividedinto 2 categories: (1) central deletional tolerance and (2) pe-ripheral tolerance, which includes strategies to mediate activa-tion-induced cell death, anergy, immune deviation, and/or theinduction of regulatory T cells. These mechanisms are not mu-tually exclusive. The introduction of donor antigens into the



thymus of mice induces specific T-cell tolerance, but whetherthe adult human thymus can shape the T-cell repertoire in thesame manner as the rodent thymus remains unclear. Mixedbone marrow chimerism allows the coexistence of both donorand recipient hematopoietic stem cells with minimal graft ver-sus host disease incidence. The bone marrow stem cells mi-grate to the thymus where they mediate the deletion of bothrecipient- and donor-reactive T cells. In nonhuman primates,anti–CD-154 monoclonal antibody (MAB) combined withsublethal total body irradiation was shown to be a promisingapproach and may be used clinically. Although a non-myeloablative protocol induces tolerance, it is also associatedwith the development of donor-specific alloantibodies thatmay lead to chronic graft rejection. The B7 molecules on an-tigen-presenting cells (APCs) and their ligands on T cells pro-vide critical costimulation signals required for clonal expan-sion of specific T cells. Blockage of the CD28/B7 pathwayusing CTLA4-Ig has been known to prevent graft rejection, butno permanent graft acceptance has been achieved. TheCD40/CD154 blockade induced islet allograft acceptancewhen combined with donor antigen pretreatment. However,this method was ineffective in NOD mice, and sublethal totalbody irradiation and bone marrow transplantation were alsoneeded. Other mechanisms for costimulatory pathway block-age, i.e., CD28 or B7 (ICOS and B7RP-1) or tumor necrosisfactor family member LIGHT, are also known to induce allo-specific T-cell anergy. In nonhuman primates anti–CD-154MAB was shown to be more effective than CTLA4-Ig. Twomajor problems associated with the use of anti–CD154 MABare the generation of donor-specific alloantibodies and the de-velopment of thromboembolic complications. Moreover, ad-ditional donor cell treatment appears to be needed to provideadequate direct signaling via MHC molecules to achieve graftacceptance.

Dr. Oluwole also discussed critical roles played by den-dritic cells (DCs) in central and peripheral tolerance. Since inthe thymus T cell receptor (TCR)-MHC/self-peptide presentedby self-DCs regulates T-cell development, a presentation ofpeptides by self-DCs to developing T cells or colonization withdonor DCs in the thymus induces acquired thymic tolerance.DCs are also key modulators of the immune responses thatinfluence T cell differentiation. Immature myeloid DCs caninduce allogeneic T cell hyporesponsiveness, while maturemyeloid DCs induce robust allostimulatory responses. Adop-tive transfer of donor-type immature myeloid DCs induces do-nor-specific regulatory T cells and prolongs allograft survival.Immature DCs may also stimulate the generation of effector Tcells in vivo that involve the induction and maintenance of pe-ripheral transplant tolerance via activation-induced cell death.Treatment with immature DCs was also shown to result in thegeneration of IL10– and transforming growth factor-�−producing regulatory T cells. At present, no reports are

available that have evaluated immature DCs therapy in a non-human primate model.

To control self-reacting T cells in the periphery, clonaldeletion and anergy in the thymus is not sufficient. Severallines of new evidence suggest that the presence of a distinctimmunoregulatory T cell is required to downregulate the acti-vation and/or proliferation of self-reactive or allospecific Tcells. These T cells express CD4, CD25, and CTLA-4 and arisefrom the thymus. CD4- and CD25-positive T cells, whenstimulated via TCRs in vitro, suppress antigen-specific andpolyclonal activation of other T cells. At least 2 mechanismsare involved in this suppressor function: (1) the regulatory Tcell directly inhibits the functional state of APCs resulting infailure to engage other T cells and (2) the regulatory T cellinhibits IL2 production by blocking the costimulatory signalthrough direct cell-to-cell contact via TCRs. Dr. Oluwole con-cluded that clinical transplant tolerance is only possible with adeliberate introduction of donor antigens into the recipientcombined with a short course of immunosuppression.

David Sharp (Irvine, CA) gave an overview of methodsfor immunoisolation of islets. He presented data demonstrat-ing the advantages of polyethylene glycol (PEG) coating.PEG-coated allogeneic islets transplanted into pancreatecto-mized, nonhuman primates functioned over 55 days withoutadditional immunosuppressants. Transplanted animals are offinsulin, and the recovered capsules contained islets that stainedfor both insulin and glucagon. Work is also in progress usingPEG-coated islets transplanted into STZ-induced diabetic non-human primates.

CONCLUSIONThe 3rd Annual Levine Symposium, held in October

2002, presented an excellent comprehensive review of up-to-date and newly generated data in the quickly developing fieldof islet biology and transplantation. The 3-day conference of-fered physicians and scientists with varying interests a thor-ough review of all aspects of the field. Topics covered at theconference included the latest information on pancreas and is-let cell morphogenesis, insulin synthesis and secretion, the is-let cell signaling network, stem cell differentiation, diabeticautoimmunity, islet isolation and culturing technologies, clini-cal islet transplantation, the efficacy of islet transplantation,and characterization of remaining metabolic abnormalitiespost–islet transplantation. Presentations and discussionshelped highlight current advances in the above topics and de-fined future challenges facing scientists and clinicians who arededicated to advancing the understanding of cellular develop-ment of the endocrine pancreas, pathophysiology of immune-mediated � cell destruction in type 1 diabetes, and the role ofislet transplantation in the treatment of patients with this dev-astating disease. The recruitment of young scientists to attendthis intensive review of available data and future challengeswas made possible by the generous travel awards provided by



the NCRR of the NIH and will undoubtedly help in the devel-opment of their careers. Generous support from the David Sha-piro (a victim of type 1 diabetes) Research Fund and theAmerican Diabetes Association helped personify the objec-tives of the 2002 Levine Symposium. The symposium pro-vided a forum for leading scientists and clinicians to shareideas and form collaborations that will help accelerate the de-velopment of a long-awaited cure for type 1 diabetes.

The 4th Annual Levine Symposium will be held on No-vember 4–8, 2003, at the Sheraton Universal Hotel in Univer-sal City, CA. The 4-day meeting will offer updates in type 1and type 2 diabetes research and focus on cellular and physi-ological aspects of diabetes and its complications and the latestprogress in metabolism and energy balance, obesity and mac-rovascular complications, oxidative stress and other mecha-nisms of diabetic complications, signal transduction and ad-vanced glycation end products, diabetic nephropathy, � cellbiology, islet cell generation, islet immunology, and islettransplantation. The 2003 Levine Symposium will provide anopportunity for scientists interested in both type 1 and type 2diabetes to interact and share ideas for the development andadvancement of future therapeutic strategies to combat theselife-threatening and epidemic diseases. Further informationon the 2003 Levine Symposium can be found online(levinesymposium.coh.org).

ACKNOWLEDGMENTThe organizers gratefully acknowledge the support of

the 3rd Annual Levine Symposium by the Southern CaliforniaIslet Cell Resources Center (Duarte, CA), National Center forResearch Resources (Bethesda, MD), National Institute ofDiabetes and Digestive and Kidney Diseases (Bethesda, MD),American Diabetes Association–David Shapiro ResearchFund (Alexandria, VA), Ross Foundation (Ketchun, ID),Roche Laboratories, Inc. (Nutley, NJ), Synthecon, Inc. (Hous-ton, TX), KRONUS, Inc. (Boise, ID), Fujisawa Healthcare,Inc. (Deerfield, IL) Medtronic MiniMed (Northridge, CA),GlaxoSmithKline (Research Triangle Park, NC), Takeda Phar-maceuticals North America (Lincolnshire, IL), Metrika, Inc.(Sunnyvale, CA), Bio—Rad Laboratories, Inc. (Hercules,

CA), WB Saunders/Mosby/Churchill (Philadelphia, PA), Dis-etronic (St. Paul, MN), and Islet Technology, Inc. (Minneapo-lis, MN). The authors acknowledge valuable assistance byKevin Ferreri, Ph.D., and Chih-Pin Liu, Ph.D., for the prepa-ration of this report and the editorial services provided byKaren J. Ramos and Jeannette Hacker.

REFERENCES1. Goeddel DV, Kleid DG, Bolivar F, et al. Expression in Escherichia coli of

chemically synthesized genes for human insulin. Proc Natl Acad SciU S A. 1979;76:106–110.

2. Shapiro AM, Lakey JR, Ryan EA, et al. Islet transplantation in sevenpatients with type 1 diabetes mellitus using a glucocorticoid-free immu-nosuppressive regimen. N Engl J Med. 2000;343:230–238.

3. Steiner DF, Cunningham D, Spigelman L, et al. Insulin biosynthesis: evi-dence for a precursor. Science. 1967;157:697–700.

4. Arvan P, Castle D. Sorting and storage during secretory granule biogen-esis: looking backward and looking forward. Biochem J. 1998;332:593–610.

5. Dufayet de la Tour D, Halvorsen T, Demeterco C, et al. � cell differen-tiation from a human pancreatic cell line in vitro and in vivo. Mol Endo-crinol. 2001;15:476–483.

6. Itkin-Ansari P, Demeterco C, Bossie S, et al. PDX-1 and cell-cell contactact in synergy to promote � cell development in a human pancreatic en-docrine precursor cell line. Mol Endocrinol. 2000;14:814–822.

7. U.S. Food and Drug Administration. Federal Register: Application ofCurrent Statutory Authorities to Human Somatic Cell Therapy Productsand Gene Therapy Products; Notice. 1993: 53248.

8. Weber D, McFarland RD, Irony I. Selected Food & Drug Administrationreview issues for regulation of allogeneic islets of Langerhans as somatictherapy. Transplantation. 2002;74:1816–1820.

9. Kuroda Y, Kawamura T, Suzuki Y, et al. A new, simple method for coldstorage of the pancreas using perfluorochemical. Transplantation. 1990;49:1027–1028.

10. Matsumoto S, Qualley SA, Goel S, et al. Effect of the two-layer (Univer-sity of Wisconsin solution-perfluorochemical plus O2) method of pan-creas preservation on human islet isolation, as assessed by the EdmontonIsolation Protocol. Transplantation. 2002;74:1414–1419.

11. Kirlew TJ, Ricordi C, Khan AA, et al. Successful human islet transplan-tation following discontinuous gradient “Rescue” purification. Trans-plantation. 2002;74:abst 414.

12. Schwitzgebel VM. Programming of the pancreas. Mol Cell Endocrinol.2001;185:99–108.

13. Kim SK, MacDonald RJ. Signaling and transcriptional control of pancre-atic organogenesis. Curr Opin Genet Dev. 2002;12:540–547.

14. Edlund H. Pancreatic organogenesis—developmental mechanisms andimplications for therapy. Nat Rev Genet. 2002;3:524–532.

15. Herrera PL. Adult insulin- and glucagon-producing cell differentiate fromtwo independent cell lineages. Development. 2000;127:2317–2322.



advances in islet cell biology...original article advances in islet cell biology from stem cell...

Documents