Diabetologia (1996) 39:474-480 Diabetologia �9 Springer-Verlag 1996

ICA69 is expressed equally in the human endocrine and exocrine pancreas M. I. Mally 1, V. Cirulli 1, A. Hayek 1, T. Otonkoski 1,2

1 The Whittier Institute, Department of Pediatrics, University of California San Diego, California, USA 2 Transplantation Laboratory and Children's Hospital, University of Helsinki, Helsinki, Finland

Summary Islet cell autoantigen 69 kDa (ICA69) has been reported as a polypeptide antigen expressed in pancreatic beta cells, and autoimmunity against this antigen has been associated with insulin-dependent diabetes mellitus. We have studied the cell type spec- ificity and ontogeny of ICA69 gene expression in man. The ICA69 gene was expressed in all adult hu- man tissues. The level of expression was three-to five-times higher in the pancreas than in the brain, li- ver, intestine, kidney, spleen, lung or adrenal glands. Pancreatic ICA69 expression increased with age, adult levels being five times higher than the levels present at 13 weeks of gestation. Total RNA from four separate preparations of isolated human islets revealed levels of ICA69 mRNA similar to those found in the pancreas as a whole, although another is- let antigen, glutamic acid decarboxylase 65, was highly

enriched in the islets. In situ hybridization and im- munohistochemical staining of sections of the fetal and adult pancreas revealed expression of the ICA69 gene and protein throughout the acinar, ductal, and islet tissue, but not in the mesenchyme. Analysis of ICA69 mRNA levels in human cell lines indicated expression in neural, endothelial and epithelial cells, but not in fibroblasts. In conclusion, ICA69, although highest in the pancreas, is widely distributed in other human tissues, excluding connective tissue. Within the human pancreas, ICA69 is not enriched in the is- lets or in the beta cells. [Diabetologia (1996) 39: 474-480]

Keywords ICA69, pancreatic islets, antigen, insulin- dependent diabetes mellitus, pathogenesis, fetal de- velopment.

Insulin-dependent diabetes mellitus (IDDM) is char- acterized by specific autoimmune destruction of pan- creatic islet beta cells eventually leading to insulin de- ficiency [1]. Immune responses against several beta- cell antigens have been reported, such as insulin [2], glutamic acid decarboxylase (GAD) [3], carboxypep- tidase H [4], glucose transporter 2 (GLUT2) [5] and

Received: 22 August 1995 and in revised form: 23 October 1995

Corresponding author: Dr. T. Otonkoski, Transplantation Lab- oratory, University of Helsinki, R O. Box 21 (Haartmaninkatu 3), FIN-00 014 Helsinki, Finland Abbreviations: GAD, Glutamic acid decarboxylase; BSA, bo- vine serum albumin; GLUT2, glucose transporter 2; bp, base pair; PBS, phosphate buffered saline; IDDM, insulin-depen- dent diabetes mellitus; HAEC, human aortic endothelial cells; ICA, islet cell autoantibodies.

islet gangliosides [6]. It is likely that all of these im- mune responses do not imply a direct pathogenetic role for these antigens, but rather reflect the epiphe- nomenal development of antibodies against beta-cell antigens exposed during the autoimmune process. One of the recently found IDDM-associated antigens is a 69 kDa protein which has been hypothesized to associate to bovine serum albumin (BSA)-induced beta-cell immunity [7-9]. The structural gene (named ICA1) encoding this protein (named ICA69) was cloned and found to encode a 483-amino acid protein not related to any known sequence except for two short regions homologous with BSA [10]. The ICA69 nucleotide sequence is identical to p69, an an- tigen identified by screening a rat islet cDNA expres- sion library with anti-BSA antibodies [11]. Although the gene was found to be expressed in several tissues, in the rat pancreas immunoreactive ICA69 was only

M. I, Mally et al.: ICA69 expression

de tec t ed in the be t a cells, suggest ing its role as a be ta - cell au toan t i gen of po ten t i a l pa thogene t i c impor - tance [10].

The studies r e p o r t e d he re were u n d e r t a k e n to charac te r i ze the tissue dis t r ibut ion and o n t o g e n y of I C A 6 9 in h u m a n tissues. Unexpec ted ly , our f indings indica te tha t in the h u m a n panc reas the gene and its p ro t e in p roduc t are not p re fe ren t i a l ly expressed in the islet cells.

Materials and methods

Human tissues and cell lines. Human fetal tissues at 13- 24 weeks' gestational age were obtained through non-profit or- gan procurement centres (Advanced Bioscience Resources, Oakland, Calif., USA; Anatomic Gift Foundation, Laurel, Md., USA). Informed consent for tissue donation was ob- tained by the procurement centres. Tissues were shipped on ice in RPMI 1640 medium (Irvine Scientific, Irvine, Calif., USA) containing 10 % pooled normal human serum. Upon ar- rival, the organs were carefully dissected free of extraneous tis- sue, immediately snap-frozen in liquid nitrogen, and stored a t - 70~ Adult pancreatic tissue was kindly provided by Dr. J. Chang (Immune Response Corp., Carlsbad, Calif., USA). To- tal RNA isolated from adult tissues was obtained from Clon- tech Laboratories (Palo Alto, Calif., USA). Pancreatic tissues were also obtained at autopsy (4-8 h post-mortem) of a 36- week preterm infant and an ll-month-old child. Isolated adult pancreatic islets were kindly provided by Dr. C. Ricordi (Uni- versity of Miami, Fla., USA) and Dr. D. Scharp (Washington University, St. Louis, Mo., USA).

The following human cell lines were harvested after culture in RPMI 1640/10 % fetal calf serum in 5 % COJ95 % air: pri- mary foreskin fibroblasts (Bassenger cells), tracheal epithelial cells (9HTEO), embryonic kidney epithelial cells (293ATCC), glioblastoma cells (HS683), cervix carcinoma cells (HeLa), osteosarcoma cells (SaOS2) and fibrosarcoma cells (HT-1080). In addition, RNA was prepared from the fol- lowing primary cultured human cells: NEU, fetal brain cell cul- tures consisting of approximately 15 % neurons and 85 % glial cells (kindly provided by Dr. E Gage, University of California San Diego, Calif., USA), HAEC, primary aortic endothelial cells (Clonetics, San Diego, Calif., USA) and HAF, fibroblasts grown from the adult pancreas (passage number 3).

RNA isolation and analysis. Total cellular RNA was isolated by the acid guanidinium thiocyanate method [12] and quantitated spectrophotometrically. A multiprobe RNase protection assay using the cyclophilin gene as an internal control was used to quantitate ICA69 mRNA as described previously in detail [13]. Yeast tRNA (10 ~Lg) was included as a negative control. The ICA69 probe was a 319 base pair (bp) fragment encom- passing amino acids 378-483 in the full length cDNA sequence [10I (provided by Dr. H.-M. Dosch, The Hospital for Sick Chil- dren, Toronto, Ontario, Canada). The GAD65 probe consisted of 5 ' untranslated sequences and the sequence encoding amino acids 1-80 in the full-length cDNA [14] (provided by Dr. A.J. Tobin, University of California, Los Angeles, Calif., USA). The cyclophilin probe (provided by Dr. D. Bergsma, Smith Kline Beecham Pharmaceuticals, King of Prussia, Pa., USA) corresponded to 5 ' untranslated sequences and amino acids 1-28 in the full-length cDNA sequence [15]. Probes were sub- cloned into either the pBluescript (Stratagene, La Jolla, Calif., USA) or pGEM (Promega, Madison, Wis., USA) riboprobe

475

vectors by standard procedures [16]. Target RNAs were quanti- tated in autoradiographs by scanning densitometry (LKB Ult- roScan XL Laser) and integrated using GelScan XL software (Pharmacia LKB, Piscataway, N.J., USA). Different exposure times were used to ensure that all signals fell within the linear range of the densitometer. The probe-specific mRNA signals were normalized to the cyclophilin signal in each sample to ac- count for differences in sample loading between lanes.

In situ hybridization of ICA69 and insulin. Sense and antisense probes for both insulin and ICA69 were transcribed from cDNAs in the presence of [35S]-uridine triphosphate using the appropriate RNA polymerase. The ICA69 antisense probe was the same as that used for the RNase protection assays. The ICA69 control (sense) probe was of equal size (300 bp) and corresponded to amino acids 1-100 in the cDNA sequence [10]. The insulin antisense and sense probes corresponded to amino acids 1-87 in preproinsulin mRNA [13]. In situ hybrid- ization was performed according to the method of Simmons et al. [17]. Consecutive 5-~tm paraffin sections of human fetal and 10-~tm frozen sections of adult pancreas were hybridized with the ICA69 and insulin probes. Slides were coated with Ko- dak NTB-2 autoradiograph emulsion (Eastmon Kodak, New Haven, CT, USA) and exposed in sealed boxes at 4 ~ The ex- posure time was 3-7 days for insulin and 4 weeks for ICA69.

lmmunofluorescence. Snap-frozen human fetal (20 weeks' ges- rational age) and adult pancreata were used to prepare 8%tin- thick cryostat sections which were mounted on glass slides, dried at room temperature for 1 h, fixed in freshly made 4 % formaldehyde (from paraformaldehyde) for 20 rain at 4~ and permeabilized in 0.1% saponin for 5 min. Reactive groups generated by formaldehyde were saturated by 50 retool/1 gly- cine. Non-specific binding was blocked by incubation in phos- phate buffered saline (PBS) containing 2 % normal donkey se- rum (Jackson ImmunoResearch Labs, West Grove, Pa., USA) and 1% BSA (fraction V, Sigma, St. Louis, Mo., USA) for 1 h at room temperature. Following extensive washes the sections were incubated for 1 h at room temperature with a mixture of primary antibodies: IgG fraction (5 ~g/ml) of a sheep anti- human insulin polyclonal antiserum (The Binding Site, Bir- mingham, UK) and IgGs (5 gg/ml) from a rabbit anti-ICA69 polyclonal antiserum (provided by Dr. M. Pietropaolo, Univer- sity of Colorado, Denver, Col., USA). In separate sections, IgGs from normal sheep serum and from pre-immune rabbit serum were used as controls for the specificity of the primary antibodies. After several washes the sections were incubated with a cocktail of secondary antibodies (5 ~tg/ml; 1 h at room temperature): LissamineRhodamine (LRSC)-conjugated af- finity-purified donkey anti-sheep IgG(H+L) (preadsorbed on chicken, guinea pig, hamster, horse, human, mouse, rabbit and rat serum proteins) and fluorescein isothiocyanate (FITC)-conjugated affinity-purified donkey anti-rabbit IgG(H + L) (preadsorbed on bovine, chicken, goat, guinea pig, hamster, horse, human, mouse, rat and sheep serum pro- teins) (both from Jackson ImmunoResearch). Following six washes of 5 min each with PBS (0.2 % donkey serum, 0.1% BSA), sections were mounted in slow fade medium (Molecular Probes, Eugene, Ore., USA), and viewed on a Zeiss Axiovert 35 M microscope (Carl Zeiss, Thornwood, NY, USA), using a 40X 1.3 NA objective lens, equipped with a laser scanning con- focal attachment (MRC-1000, Bio-Rad Laboratories, Cam- bridge, Mass., USA). Fluorescent images relative to each mar- ker were collected using an argon/krypton mixed gas laser. Colour composite images were generated using Adobe Photo- shop 3.0 (Adobe Systems, Inc., Mountain View, CA, USA) running on a Quadra 950 computer (Apple Inc., Cupertino,

476 M.I. Mally et al.: ICA69 expression

B

6

5: 5 r-" ~. 4 O "6 3 O 2

�9 16-week fetal

[ ] 22-week fetal

[ ] Adult

I n H I I I I I i I i

> .- r

Fig. 1A, B. ICA69 expression in various human fetal and adult tissues. A A representative RNase protection assay for the si- multaneous detection of ICA69 and cyclophilin. The two probes were hybridized with 10 ,ug of total RNA (65 h expo- sure). Identification of lanes: M, molecular weight markers (nucleotides shown on left); lanes 1-3, brain (16-week, 21- week, adult); lanes 4-6, liver ; lanes 7-9, kidney; lanes 10-11, adrenal (16-week, adult); lanes 12-13, lung (16-week, 21- week); lanes 14-15, pancreas (16-week, adult); lane 16, adult islets; lane 17, fetal spleen; lane 18, fetal ventricle; lane 19 fetal jejunum; lane 20, yeast tRNA (negative control); R undigested probes. Sizes of protected fragments are indicated with the ar- rows (ICA69:319 nucleotides, cyclophilin: 135 nucleotides). B Densitometrically determined ICA69 mRNA levels derived from the assay shown in A

CA, USA), and printed on a Tektronix Phaser II-SDX co/our printer (Beaverton, OR, USA).

Statistical analysis

The correlation between age and ICA69 expression level was studied with simple regression analysis using Statview 4.1 soft- ware for the Macintosh (Abacus Concepts, Berkeley, Calif., USA).

Results

Tissue distribution, cell type specificity and ontogeny of ICA69 gene expression. Using the RNase protect ion assay, ICA69 gene expression was detected in all tis- sues studied (brain, liver, kidney, adrenal, lung, pan- creas, spleen, ventricle and small intestine). The high-

8 -

6- E

4- ;-~ 57

~-- 2- M9 <

~ < < eq m <

Fig.2. ICA69 expression in cultured cells. Densitometrically determined band intensities are presented using 10-~g total RNA derived from the various cell lines, and from adult pan- creas and brain tissue as a control. HFR human fetal pancreas; ICCs, islet-like cell clusters; ML, monolayer. For explanation of the cell lines, see Methods. n. d., not detectable

est signal was detected in the adult pancreas, in which ICA69 expression was approximately five times high- er than in the other adult tissues (Fig. 1). In the pan- creas, liver, kidney and adrenals the m R N A levels were higher in the adult than in the fetal tissue, whereas fetal and adult levels were equal in the brain.

Of the cultured cells tested, ICA69 m R N A was highest in the isolated adult pancreatic islets. This le- vel, however, was not higher than that found in the pancreas as a whole (Fig. 2). No expression was de- tected in fibroblasts derived from the adult pancreas (HAF cells), suggesting that I C A 6 9 m R N A was mainly localized in the epithelial cells. Fetal pancre- atic epithelial cells, cultured either as islet-like cell clusters or monolayers, contained less than 50 % of the ICA69 m R N A found in adult islets. An even low- er level of expression was detected in primary fetal brain cells (NEU cells, consisting of approximately 15 % neurons) and aortic endothelial cells ( H A E C cells). No ICA69 expression was detected in the pri- mary fibroblasts (HAF or Bassenger cells), glioblas- toma cells (HS683) or HeLa cells, and the expression was barely detectable in the other epithelial cell lines (293ATCC and 9HTEO, Fig. 2).

Transcriptional expression of ICA69 in the pan- creas was positively correlated with age (r = 0.69, p < 0.01). A two- to three-fold increase was evident between the 13th and 24th gestational weeks and the adult pancreas contained up to 7 times more ICA69 m R N A than the early fetal pancreas (Fig. 3).

M. I. Mally et al.: ICA69 expression

7

r -

E o 6~

C

r "

. m r " r ~

o o > . ,

o

r < o

6

5

4

3

2

1

0 10

�9 �9

8 0 I,

o l o

/g i I i i I

months Gestational age (weeks)

Fig.& Ontogeny of pancreatic ICA69 expression. Each point represents one pancreas. Data represent densitometrically de- termined ratios of ICA69 and cyclophilin band intensities de- rived from a single RNase protection assay using 5 gg of total RNA

477

thelial localization of ICA69, not limited to the islet cells (data not shown).

ICA69 antigen was detected simultaneously with insulin by double immunofluorescence in the fetal human pancreas. A representative field is shown in Figure 6, where insulin-positive cells are detected in small clusters in panel A. ICA69-specific immunore- activity, shown in panel B, is clearly detectable in all epithelial cells of the pancreas, with strong staining of ductal cells, whereas the surrounding mesenchyme is negative. The same immunofluorescence protocol was used to stain sections of adult human pancreas. An islet is identified by the anti-insulin antibody in Figure 7A. Strong ICA69-specific immunoreactivity is recorded in both beta cells and the surrounding aci- nar cells (Fig. 7B). In both the fetal and the adult pan- creas, incubation of a separate section with anti-insu- lin antibody and pre-immune rabbit serum (obtained from the same animal used to generate the anti- ICA69 antibody) did not show any significant immu- noreactivity for ICA69 (Figs. 6E, 7E).

Fig.4. Comparison of ICA69 and GAD6s expression in the same adult pancreas and islet RNA samples. There is no differ- ence between the levels of ICA69 m R N A in the islets and total pancreas (/eft), whereas the levels of GAD6s mRNA are 5.0- and 6.3-times higher in islet samples 1 and 2 than in the total pancreas (right)

Lack of islet specificity in pancreatic ICA69 expres- sion. Quantitation of ICA69 transcriptional levels us- ing RNase protection assays revealed similar expres- sion levels in total pancreas and isolated islets (Figs. 2, 4). The ratio of islet to pancreas expression ranged from 0.6 to 1.1 in four separate preparations of iso- lated islet RNA analysed simultaneously with total pancreas RNA. In comparison, GAD65 expression was highly enriched in the same samples (islet/pan- creas ratio 4.9 to 9.1). Representative experiments are shown in Figure 4.

Localization of ICA69 gene expression in the var- ious pancreatic cell types was accomplished by in situ hybridization. In the fetal pancreas the gene was expressed widely throughout the epithelial com- ponents of the pancreas (Fig. 5C), contrasting with the scattered clusters of insulin expressing cells (Fig.5D). Although islet cells did express ICA69, there was no evidence for clustering of ICA69 mR- NA in insulin-expressing cells (Fig.5E, F). Adult pancreas essentially showed a similar pattern of epi-

Discussion

Our results indicate that the gene encoding ICA69, initially reported as an islet cell autoantigen with a beta-cell specific expression pattern in the rat pan- creas [10], is widely transcribed and translated in hu- man pancreatic epithelial cells, including both exo- and endocrine tissue. Transcripts of the gene were de- tected in all human tissues studied but the level of ex- pression was clearly highest in the pancreas. Thus, ICA69 may be involved in some aspects of pancreatic physiology. Based on our findings, vascular endothe- lium and peripheral nerve endings might represent the cellular compartments expressing ICA69 and ac- count for the low ICA69 transcription levels detected in organs other than the pancreas. Fibroblasts and other stromal cells do not express ICA69. The sub- cellular localization and function of the ICA69 pro- tein remain unknown.

The RNase protection assay used in this study is more sensitive than the Northern blot method used previously [10]. Consequently, we have detected low level gene expression (e.g. in the spleen), which might not have been detectable with a less sensitive method. Our results on the expression of the ICA69 gene are in agreement with previously published data [10], demonstrating high expression in the pan- creas although present in a variety of other tissues. In previous studies, the expression levels in total pan- creas and isolated islets have not been compared. Thus, the equally high expression in pancreatic exo- crine and endocrine cells could have remained un- noticed.

Our results concerning the localization of ICA69 protein in the human pancreas are in apparent dis-

478 M.I. Mally et al.: ICA69 expression

M. I.Mally et al.: ICA69 expression 479

Fig.6A-E Double immunofluorescence for insulin and ICA69 antigen in the human fetal pancreas. Panel A shows insulin- positive cells in islet-like clusters scattered throughout the tis- sue. ICA69-specific immunoreactivity is shown in panel B, where all epithelial cells appear highlighted, including the duc- tal epithelium. Merging of red and green fluorescence is shown in panel C, indicating co-localization of ICA69 and insulin. A representative field from a separate section incubated with anti-insulin antibody and IgGs from preimmune rabbit serum is shown in panels D, E and E Calibration bar = 50 btm

Fig.7A-F. Double immunofluorescence for insulin and ICA69 antigen in the human adult pancreas. Panel A shows an islet as identified by an anti-insulin antibody. ICA69-specific immu- noreactivity of the same field is displayed in B, where the fluo- rescence is distributed in both insulin-producing cells and sur- rounding acinar tissue. Merging of the two fluorescent signals is shown in C. Incubation of a sequential section with anti-insu- lin antibody and IgGs from preimmune rabbit serum identified the same islet stained in panel D. Note the lack of reactivity for the preimmune IgGs (E, F). Calibration bar = 60 btm

Fig.5A-H. In situ hybridization with anti-sense (A,C,E) and sense (G) probes for ICA69 on the left and insulin on the right (B,D,E anti-sense probe; H, sense probe) in consecutive sec- tions of a 20-week human fetal pancreas. C and D (dark-field of same areas as in A and B) illustrate the contrasting pattern of ICA69 and insulin gene expression: there is no evidence for clustering of ICA69 mRNA in the islets, although the mRNA is concentrated in epithelial cells. E and F show a higher mag- nification of the same areas, outlined in A and B. Widespread pattern of specific ICA69 gene expression is evident, in con- trast to the highly localized pattern of insulin transcription. Some insulin-expressing cells may also express ICA69. Expo- sure time was 7 days for insulin and 28 days for ICA69. Origi- nal magnification 363 x in A, B, C, D, G and H; 1451 x in E and F

a g r e e m e n t with p rev ious findings, which indica ted a beta-cel l specific express ion pa t t e rn of I C A 6 9 p ro te in in the ra t panc reas [10]. In this regard , the possibi l i ty of a species d i f fe rence b e t w e e n ra t and m a n in the cel l - type t a rge t ed express ion of this p ro t e in mus t be t aken into account . Indeed , a s imilar s i tuat ion has b e e n descr ibed for a n o t h e r islet ant igen, G A D , which appea r s to be beta-cel l specific in the ra t bu t not in the h u m a n panc reas [18,19]. A l t h o u g h the 67 k D a G A D i so fo rm is highly expressed in ra t islets, it can- not be found in h u m a n islets [18]. O u r ev idence for equa l express ion of I C A 6 9 in the endocr ine and exo- crine panc rea s is based on consis tent resul ts ob t a ined with th ree s epa ra t e me thods : 1) quan t i t a t ion of I C A 6 9 t ranscr ip t ional levels using to ta l panc reas vs i so la ted islets; 2) in situ hybr id iza t ion and 3) specific

480

immunof luorescence . Thus, we conclude that in the huma n pancreas ICA69 is un i formly expressed in all epithelial cells, including the be ta cells.

It is obvious that I D D M pat ients have ant ibodies against ICA69 when studied on Wes te rn blots [10]. Howeve r , the re levance of ICA69 as an au toan t igen in I D D M has been quest ioned, since the ant igen could not be immunoprec ip i t a t ed by I D D M sera and the occur rence of ICA69 autoant ibodies was not dif- f e ren t be tween I D D M pat ients and contro l subjects in a subsequent s tudy [20]. Ant ibodies to ICA69 also occur f requent ly in pat ients with rheumato id arthrit is [21]. Thus, ICA69 may be a general ly immunogen ic pept ide when re leased f rom several types of cells. The results of the present study, indicating that ICA69 is un i formly expressed in pancrea t ic endo- crine and exocr ine cells, m a k e it even more difficult to envision a pa thogene t i c role for ICA69 in I D D M .

In conclusion, the data p r e sen t ed in this s tudy indi- cate that a l though ICA69 is dis t r ibuted widely in hu- man tissues, it is most highly expressed in the pan- creas. Most importantly, we show that in the pancreas ICA69 gene and pro te in are equal ly found in the exo- crine and endocr ine compar tmen t . This unusual ex- pression pa t t e rn suggests a h i ther to unknown role for ICA69 in pancrea t ic physiology. The possible role of ICA69 as an au toan t igen in I D D M remains to be elucidated.

Acknowledgements. We thank Ms. A. D. Lopez for skilled tech- nical assistance, Dr. M. Pietropaolo for providing the ICA69 antisera, and Drs. H.-M. Dosch, A.J. Tobin, and D. Bergsma for providing the cDNAs encoding ICA69, GAD65, and cyc- lophilin, respectively. We also thank Dr. O. Vaarala for helpful discussions during the preparation of the manuscript, and Dr. M.H. Ellisman for advice and allowing the use of his confocal microscope at the Microscopy and Imaging Resource of the University of California, San Diego. This study was supported by NIDDK Grant RO1-DK 39087, The Herbert O. Perry Fund, The Deutz Family, The Sigrid Jusdlius Foundation, The Paulo Foundation, The Foundation for Diabetes Research and The Foundation for Pediatric Research (Finland). V. Cirulli was a recipient of a fellowship from the Swiss National Research Fund. T. Otonkoski is the recipient of a Juvenile Dia- betes Foundation International Career Development Award.

References

1. Eisenbarth GS (1986) Type I diabetes mellitus. A chronic autoimmune disease. N Engl J Med 314:1360-1368

2. Palmer JR Asplin CM, Clemons Pet al. (1983) Insulin anti- bodies in insulin-dependent diabetics before insulin treat- ment. Science 222:1337-1339

3. Baekkeskov S, Aanstoot H-J, Christgau S e t al. (1990) Identification of the 64 K autoantigen in insulin-dependent diabetes as the GABA-synthesizing enzyme glutamic acid decarboxylase. Nature 347:151-156

M. I. Mally et al.: ICA69 expression

4. Castano L, Russo E, Zhou L, Lipes MA, Eisenbarth GS (1991) Identification and cloning of a granule autoantigen (carboxypeptidase H) associated with type I diabetes. J Clin Endocrinol Metab 73:1197-1201

5. Inman LR, McAllister CT, Chen L et al. (1993) Autoanti- bodies to the GLUT2 glucose transporter of B cells in insu- lin-dependent diabetes mellitus of recent onset. Proc Natl Acad Sci USA 90:1281-1284

6. Dotta F, Colman PG, Lombardi D, et al. (1989) Ganglio- side expression in human pancreatic islets. Diabetes 38: 1478-1483 Martin JM, Trink B, Daneman D, Dosch H-M, Robinson B (1991) Milk proteins in the etiology of insulin-dependent diabetes mellitus (IDDM). Ann Med 23:447-452 Karjalainen J, Martin JM, Knip Met al. (1992) A bovine al- bumin peptide as a possible trigger of insulin-dependent di- abetes mellitus. N Engl J Med 327:302-307 Cheung R, Karjalainen J, Vandermeulen J, Singal DR Dosch HM (1994) T cells from children with IDDM are sensitized to bovine serum albumin. Scand J Immunol 40: 623-628 Pietropaolo M, Castafio L, Babu S et al. (1993) Islet cell au- toantigen 69 kDa (ICA69). Molecular cloning and charac- terization of a novel diabetes-associated autoantigen. J Clin Invest 92:359-371 Miyazaki I, Gaedigk R, Hui MF et al. (1994) Cloning of hu- man and rat p69 cDNA, a candidate autoimmune target in type 1 diabetes. Biochem Biophys Acta 1227:101-104 Chomczynski R Sacchi N (1987) Single-step method of RNA isolation by guanidium thiocyanate-phenol-chloro- form extraction. Anal Biochem 162:156-159 Mally MI, Otonkoski T, Lopez AD, Hayek A (1994) Devel- opmental gene expression in the human fetal pancreas. Ped Res 36:537-544 Karlsen AE, Hagopian WA, Grubin CE et al. (1991) Clon- ing and primary structure of a human islet isoform of gluta- mic acid decarboxylase from chromosome 10. Proc Natl Acad Sci USA 88:8337-8341 Haendler B, Hofer-Warbinek R, Hofer E (1987) Comple- mentary DNA for human T-cell cyclophilin. Embo J 6: 947-950 Sambrook J, Fritsch EF, Maniatis T (1989) In: Molecular Cloning: a laboratory manual. Cold Spring Harbor Labora- tory Press, New York, pp 1.53-1.86 Simmons DM, Arriza JL, Swanson LW (1989) A complete protocol for in situ hybridization of messenger RNAs in brain and other tissues with radiolabeled single-stranded RNA probes. J Histotechnol 12:169-181 Petersen JS, Russel S, Marshall MO et al. (1993) Differen- tial expression of glutamic acid decarboxylase in rat and human islets. Diabetes 42:484-495 Vives-Pi M, Somoza N, Vargas F et al. (1993) Expression of glutamic acid decarboxylase (GAD) in the a, [3 and 6 cells of normal and diabetic pancreas: implications for the pathogenesis of type I diabetes. Clin Exp Immunol 92: 391-396 Lampasona V, Ferrari M, Bosi E, Pastore MR, Bingley PJ, Bonifacio E (1994) Sera from patients with IDDM and healthy individuals have antibodies to ICA69 on Western blots but do not immunoprecipitate liquid phase antigen. J Autoimmun 7:665-674 Martin S, Kardorf J, Schulte B et al. (1995) Autoantibodies to the islet antigen ICA 69 occur in IDDM and in rheuma- toid arthritis. Diabetologia 38:351-355

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.