CD1d-expressing cells present lipid Ag to CD1d-restricted NKT cells, which play an important role in immune regulation and tumor rejection. Lymphoid enhancer-binding factor-1 (LEF-1) is one of the regulators of the Wnt signaling pathway, which is a powerful regulator in cellular growth, differentiation, and transformation. There is little evidence connecting Wnt signaling to CD1d expression. In this study, we have identified LEF-1 as a regulator of the expression of the gene encoding the human CD1d molecule (CD1D). We found that LEF-1 binds specifically to the CD1D promoter. Overexpression of LEF-1 in K562 or Jurkat cells suppresses CD1D promoter activity and downregulates endogenous CD1D transcripts, whereas knockdown of LEF-1 using LEF-1–specific small interfering RNA increases CD1D transcripts in K562 and Jurkat cells but there are different levels of surface CD1d on these two cell types. Chromatin immunoprecipitation showed that the endogenous LEF-1 is situated at the CD1D promoter and interacts with histone deacetylase-1 to facilitate the transcriptional repressor activity. Knockdown of LEF-1 using small interfering RNA potentiates an acetylation state of histone H3/H4, supporting the notion that LEF-1 acts as a transcriptional repressor for the CD1D gene. Our finding links LEF-1 to CD1D and suggests a role of Wnt signaling in the regulation of the human CD1D gene.

Abstract CD1d expressing cells present lipid antigen to CD1d-restricted natural killer T (NKT) cells, which play an important role in immune regulation and tumor rejection. Lymphoid enhancer-binding factor-1 (LEF-1) is one of the regulators of the Wnt signaling pathway, which is a powerful regulator in cellular growth, differentiation, and transformation. There is little evidence connecting Wnt signaling to CD1d expression. In this study, we have identified LEF-1 as a regulator of the expression of the gene encoding the human CD1d molecule (CD1D). We found that LEF-1 binds specifically to the CD1D promoter. Over-expression of LEF-1 in K562 or Jurkat cells suppresses CD1D promoter activity and down regulates endogenous CD1D transcripts, whereas knockdown of LEF-1 using LEF-1-specific siRNA increases CD1D transcripts in K562 and Jurkat cells but different levels of surface CD1d on these two cell types. Chromatin immunoprecipitation showed that the endogenous LEF-1 factor is situated at the CD1D promoter and interacts with histone deacetylase-1 to facilitate the transcriptional repressor activity. Knockdown of LEF-1 using siRNA potentiates an acetylation state of histone H3/H4, supporting that LEF-1 acts as a transcriptional repressor for CD1D gene. Our finding links LEF-1 to CD1D and suggests a role of Wnt signaling in the regulation of human CD1D gene.

CD1d presents lipid Ags to a specific population of NK T cells, which are involved in the host immune defense, suppression of autoimmunity, and the rejection of tumor cells. The transcriptional control mechanism that determines the regulation and the tissue distribution of CD1d remains largely unknown. After investigating 3.7 kb 5' upstream of the coding region, we found that human gene encoding CD1d molecule (CD1D) has TATA boxless dual promoters with multiple transcription initiation sites. The proximal promoter is located within the region of -106 to +24, and the distal promoter is located within the region of -665 to -202 with the A of the translational start codon defined as +1. The longest 5'-untranslated region derived from 5'-RACE and apparently generated by the distal promoter has 272 bp in length covering the genomic sequence of the proximal promoter. The region covering the proximal promoter gave a much higher luciferase activity in Jurkat cells than in K562 cells, whereas it was in reverse for the region covering the distal promoter, indicating a cell type sp. act. of the two promoters. Transcription factor SP1 plays a crucial role in the function of the proximal promoter. The analysis of the CD1D promoter region indicates that IFN-gamma, NF-IL-6, and T cell factor 1/lymphoid enhancer-binding factor 1 are most likely involved in the regulation of CD1d expression. The illustration of the dual CD1D gene promoters will help to reveal the regulatory factors that control CD1d expression and its tissue distribution for a better understanding of the cross-regulation between CD1d and NK T cells.

CD1d presents lipid antigen to a conserved population of natural killer (NK) T cells, which participate in host immune defense, tumor cell rejection and suppression of autoimmunity. The levels of human CD1d expression vary significantly between individuals. To understand such variation, we sequenced the region up to 1.7 kb 5' upstream of the translation start site and partially through exon 2 in 44 white Americans. We also studied two tagged single nucleotide polymorphisms ( SNP) in 112 white Americans, 60 African-Americans, 88 Europeans, and 84 Chinese people from the region. Six SNP present in the region (-836C --> T, - 773C --> T, - 764C --> G, - 713A --> T, - 365A --> G and +363A --> G) were found to be in a complete linkage disequilibrium and comprised three haplotypes. Haplotype 1 had - 836C, - 773C, - 764C, - 713A, - 365A and +363A. Haplotype 2 had -836C, - 773T, - 764C, - 713A, - 365A and +363A. Haplotype 3 had - 836T, - 773C, - 764G, - 713T, - 365G and +363G. - 773C --> T and - 764C --> G can serve as the tagged SNP to differentiate the three haplotypes. The frequency of haplotype 1 was significantly higher in African Americans than in the other three ethnic groups, whereas the frequency of haplotype 3 was significantly higher in the Chinese people than those in the other three groups. The finding of the three haplotypes provides a genetic marker for CD1d and facilitates the study of the functional role of the genetic variations in human CD1d expression and regulation.

The most compelling case for autoimmune mediated hypogonadism occurs when ovarian failure is part of an autoimmune polyglandular syndrome (APS). In patients with the rare, recessively inherited type 1 APS (APS-1), characterized by the triad of chronic mucocutaneous moniliasis, hypoparathyroidism, and Addison’s disease, primary amenorrhea (elevated pituitary gonadotropins) or oligomenorrhea and infertility are constant features. Ovarian failure is associated with autoantibodies to steroid hormone secreting cells in the adrenal cortex, Leydig cells of the testes, granulosa /thecal cells of the Graffian follicles, corpus luteum, and the syncytiotrophoblast of the placenta. These autoantibodies react with 3 P450 enzymes involved with steroidogenesis, namely, 21-hydroxylase (adrenal specific), 17α-hydroxylase, and the side chain cleavage enzyme. Recently the 14 exon, APS-1 (autoimmune regulator or AIRE) gene has been cloned (chr. 21p22.3), and multiple mutants discovered. Parents who are obligatory heterozygotes for a single mutant gene lack clinical features of APS-1. They also do not develop APS-1 autoantibodies. Thus, hypogonadal patients without features of APS-1 are unlikely to have AIRE gene mutations. In the more common APS-2/3, characterized by combinations of autoimmune thyroid disease, immune mediated type 1 diabetes, vitiligo, pernicious anemia, and Addison’s disease (type 2, not type 3), ovarian disease may be seen. In primary hypogonadism outside of the context of an APS, these autoantibodies are rare.

Information on genetic susceptibility to Graves' disease in African Americans is limited. We studied DRB1, DQB1, DRB3 subtypes, DQA1*0501, DQA1*0201, and CTLA-4 polymorphisms in 49 African American patients with adult onset Graves' disease and 47 racially-matched controls using PCR-based sequence-specific priming methods. There were no significant differences in DRB1 or DQB1 allelic frequencies or CTLA-4 polymorphisms between patients and controls. However, we found that the frequency of DRB3 was significantly increased in the patients (75.5% vs. 57.4%, P = 0.006, X2 = 3.52), especially for the DRB3*0202 subtype (53.1% vs. 23.4, P = 0.003, X2 = 8.91). In this one respect, the finding was in concordance with our previous observations in Caucasian patients with adult-onset Graves' disease. In addition, whereas the frequency of DQA1*0501 was increased (P = 0.018, X2 = 5.63) in our patients, the haplotype of DRB3/DQA1*0501, or DRB3*0202/DQA1*0501 was found to be more strongly associated (P = 0.008, X2 = 7.0; P = 0.0008, X2 = 11.34, respectively). These data suggest that DRB3*0202, particularly when found with DQA1*0501 in a haplotype is a susceptible gene(s) for Graves' disease in adult African Americans. Considering these data with those in Caucasian patients, our results would suggest that the primary Graves susceptible locus is likely DRB3 and not DRB1.

Anti–idiotypic antibodies may regulate the immune system and influence pathogenic autoimmunity. We investigated idiotype-anti–idiotype interactions in sera of patients with primary biliary cirrhosis (PBC), normal subjects and animals immunized with pyruvate dehydrogenase complex (PDC) or its derivatives. IgG autoantibody to the E2 subunit of PDC (PDC–E2) was derived by affinity–purification from sera of 12 patients with PBC, and F(ab)2 was prepared (anti-PDC–E2–F[ab]2). This was used as a reactant by enzyme–linked immunosorbent assay (ELISA) with sera from patients with PBC, normal subjects, or immunized animals. Results were that IgG antibody to anti-PDC–E2–F(ab)2 was detectable at low concentration in 12 PBC sera (mean optical density [OD] ± SD: 1.02 ± 0.26), and also in 19 normal sera (mean OD ± SD: 0.97 ± 0.35) using a serum dilution of 1:20; background OD was 0.09 to 0.10, whereas antisera from animals immunized with PDC or PDC–E2 were nonreactive. There was a significant inverse correlation (r = -.59, P = .04) between the levels of anti-PDC–E2 in PBC sera (but not normal sera), and anti–idiotypic antibody reactive with anti-PDC–E2–F(ab)2. Anti–idiotypic antibody existed as a complex with anti-PDC–E2, because the removal of anti-PDC–E2 from serum resulted in decreased reactivity to anti-PDC–E2–F(ab)2. Reactivity between PDC–E2 and anti-PDC–E2 from PBC serum was not inhibited by normal sera, indicating that anti–idiotypic antibody from normal sera with anti-PDC–E2 reacts with the framework of F(ab) rather than the paratope. The conclusions are that PBC and normal sera contain IgG class anti–idiotypic antibodies to anti-PDC–E2, the characteristic autoantibody in PBC. Anti-PDC–E2 in immunized animals does not contain an idiotype cross–reactive with human anti-PDC–E2. Anti–idiotypic antibody in PBC is complexed with anti-PDC–E2 and in part accounts for immune complexes demonstrable in PBC. Anti–idiotypic antibody in PBC may regulate levels of anti-PDC–E2.

Graves' disease is known to be HLA-D associated; however, the primary loci involved remain unclear. We examined HLA genotypes of DRB1 and DQB1 plus DRB3 subtypes using PCR-based sequence-specific priming in two groups of North American (Gainesville, FL; and Toronto, Canada) Caucasian patients with Graves' disease. We stratified patients into those with either early age at onset (<20 yr; 13.1 +/- 4.8 yr; n = 30) and later age at onset of disease (38.8 +/- 9.7 yr; n = 62) and compared the results to 192 normal controls. As expected, we found that DRB1*03 was associated with Graves' disease, but at a higher odds ratios for early-onset than later-onset patients (3.7 vs. 2.2). The frequency of DRB1*08 was also increased in both groups of patients, but significantly so only in patients with early-onset Graves' (P = 0.001; chi2 = 10.8). DRB3 was highly associated with Graves' in both groups of patients (P = 0.009; chi2 = 6.83 and P = 0.0015; chi2 = 10.1, respectively); however, the subtypes of DRB3 revealed differential susceptibilities. Whereas the frequencies of both DRB3*0101 and DRB3*0202 were increased over the entire cohort, that of DRB3*0301 was not. Significant P values were found for DRB3*0101 in patients with early-onset and for DRB3*0202 in patients with later onset of Graves' disease. When the haplotypes of DRB1*03-DRB3 of all subtypes were removed for analysis (all DRB1*03 positive also had DRB3*0101), the frequency of DRB3*0202 remained significantly higher in the patients with later onset of Graves' disease than in controls (P = 0.0043; chi2 = 8.13), but DRB3 was no longer positively associated with the early-onset group. In addition, we found that DRB1*07 was negatively associated with both groups of patients (P = 0.024; chi2 = 5.10 and P = 0.0085; chi2 = 6.93). These data suggest that DRB3*0202 is more likely to be the primary susceptible locus than DRB1*03 for patients with later onset of Graves' disease.

Autoimmune polyendocrinopathy syndrome type 1 (APS-1; MIM# 240300) is a rare autosomal recessively inherited disease characterised by destructive autoimmune diseases of endocrine glands. The gene responsible for APS-1, known as AIRE (for autoimmune regulator), was recently identified and contains motifs suggestive of a transcription regulator, To date, nine APS-1-associated mutations have been identified in the AIRE gene, including two common mutations R257X and 1094-1106del. In addition to these two mutations, we report seven novel mutations in 16 APS-1 patients from North America, We found that 1094-1106del and R257X were the most common mutations in this population of mixed geoethnic origin, accounting for 17/32 and 4/32 alleles, respectively, Haplotype analyses suggest that both are recurrent mutations, occurring on several different haplotypes with closely linked markers. All the novel mutations appear to be rare, occur ring in only single APS 1 families. After examining all coding sequences and exon/intron boundaries of the AIRE gene, the other APS 1 allele remained unidentified in three patients, Genotype phenotype correlations for APS-1 remain difficult, suggesting that other genetic or environmental factors, or both, influence the clinical presentation and disease progression in individual APS 1 patients. Hum Mutat 13:69-74, 1999. (C) 1999 Wiley-Liss, Inc.

Anti-idiotypic antibodies may regulate the immune system and influence pathogenic autoimmunity. We investigated idiotype-anti-idiotype interactions in sera of patients with primary biliary cirrhosis (PBC), normal subjects and animals immunized with pyruvate dehydrogenase complex (PDC) or its derivatives. IgG autoantibody to the E2 subunit of PDC (PDC-E2) was derived by affinity-purification from sera of 12 patients with PBC, and F(ab)(2) was prepared (anti-PDC-E2-F[ab](2)). This was used as a reactant by enzyme-linked immunosorbent assay (ELISA) with sera from patients with PBC, normal subjects, or immunized animals. Results were that IgG antibody to anti-PDC-E2-F(ab)(2) was detectable at low concentration in 12 PBC sera (mean optical density [OD] +/- SD: 1.02 +/- 0.26), and also in 19 normal sera (mean OD +/- SD: 0.97 +/- 0.35) using a serum dilution of 1:20; background OD was 0.09 to 0.10, whereas antisera from animals immunized with PDC or PDC-E2 were nonreactive. There was a significant inverse correlation (r = -.59, P = .04) between the levels of anti-PDC-E2 in PBC sera (but not normal sera), and anti-idiotypic antibody reactive with anti-PDC-E2-F(ab)(2). Anti-idiotypic antibody existed as a complex with anti-PDC-E2, because the removal of anti-PDC-E2 from serum resulted in decreased reactivity to anti-PDC-E2-F(ab)(2). Reactivity between PDC-E2 and anti-PDC-E2 from PBC serum was not inhibited by normal sera, indicating that anti-idiotypic antibody from normal sera with anti-PDC-E2 reacts with the framework of F(ab) rather than the paratope. The conclusions are that PBC and normal sera contain IgG class anti-idiotypic antibodies to anti-PDC-E2, the characteristic autoantibody in PBC. Anti-PDC-E2 in immunized animals does not contain an idiotype cross-reactive with human anti-PDC-E2. Anti-idiotypic antibody in PBC is complexed with anti-PDC-E2 and in part accounts for immune complexes demonstrable in PBC. Anti-idiotypic antibody in PBC may regulate levels of anti-PDC-E2.

Naturally occurring autoantibodies are ubiquitous and may serve physiological functions. We examined the relationship of natural and disease-associated autoantibodies in the context of autoantibodies to dihydrolipoamide acetyltransferase, the 74 kDa E2 sub-unit of the mitochondrial pyruvate dehydrogenase complex (PDC-E2), characteristic of primary biliary cirrhosis (PBC). We tested for natural autoantibodies to PDC-E2 in normal sera, and compared epitopes recognised by natural and disease-associated autoantibodies. Methods included affinity purification of anti-PDC-E2 from normal and PBC sera, ELISA and immunoblotting, capacity of antibodies to inhibit the enzyme function of the pyruvate dehydrogenase complex (PDC), use of F(ab)2fragments of anti-PDC-E2 in inhibition assays, and testing affinity purified anti-PDC-E2 on peptide fragments of PDC-E2. We found that natural auto-antibodies to PDC-E2 of IgG class were demonstrable in all healthy human sera (10/10). However, their reactivity differed from that of disease-associated autoantibodies, in that anti-PDC-E2 from normal serum failed to inhibit the catalytic activity of PDC; and F(ab)2fragments from PBC sera potently blocked the binding of anti-PDC-E2 from PBC sera to PDC-E2, but not the binding of natural anti-PDC-E2 to PDC-E2. Immunoblotting on fragments of PDC-E2 using affinity-purified preparations from PBC sera and normal sera failed to provide evidence for gross differences in epitope reactivity. We conclude that normal human sera contain natural IgG autoantibodies to the immuno-dominant inner lipoyl domain of PDC-E2, as seen characteristically in PBC. However, there is evidence for differences in fine epitope recognition.

Immune-mediated (type 1) diabetes is a multigenetic, autoimmune disease that is environmentally influenced, with worldwide incidence rates that vary greatly (1Atkinson M Maclaren N N. Engl. J. Med. 1994; 331: 1428-1436Crossref PubMed Scopus (922) Google Scholar). Whereas the inductive events remain speculative, a viral etiology has been frequently implicated (7Maclaren N Atkinson M Mol. Med. Today. 1997; 3: 76-83Abstract Full Text PDF PubMed Scopus (44) Google Scholar). Conrad and colleagues have proposed that the pancreatic β cell damage is mediated by autoreactive T lymphocytes activated systemically by a retroviral superantigen (Sag). This hypothesis is based on three observations: the enrichment of Vβ7 in the receptor repertoires expressed by T lymphocytes in the pancreas of two recently diagnosed diabetic patients (3Conrad B Weidmann E Trucco G Rudert W.A Behboo R Ricordi C Rodriquez-Rilo H Finegold D Trucco M Nature. 1994; 371: 351-355Crossref PubMed Scopus (306) Google Scholar); the subsequent cloning of a superantigen encoded by the envelope (env) gene of a species of the endogenous retrovirus HERV-K10 (IDDMK1,222); and the demonstration that the Env protein of this virus mediates a Vβ7-biased Sag effect on lymphocytes irrespective of their HLA phenotype (4Conrad B Weissmahr R.N Böni J Arcari R Schüpbach J Mach B Cell. 1997; 90: 303-313Abstract Full Text Full Text PDF PubMed Scopus (355) Google Scholar). In that report, the onset of diabetes appeared to coincide with a demonstrable IDDMK1,222 viremia that was not detected in controls. We tried to confirm this latter finding in view of its potential importance, but could not. In our studies, plasma samples derived from 8 newly diagnosed diabetic children of 5–15 years of age (within one month of their clinical diagnoses) and 4 with established type 1 diabetes, as well as 12 age-matched nondiabetic controls, were processed using the protocol that Conrad et al. described in their epidemiological study (4Conrad B Weissmahr R.N Böni J Arcari R Schüpbach J Mach B Cell. 1997; 90: 303-313Abstract Full Text Full Text PDF PubMed Scopus (355) Google Scholar). The Southern blot of the reverse-transcribed (RT) polymerase chain reaction (PCR) products we obtained revealed that most plasma samples had evidence of LTR-flanked nucleotide sequences (Figure 1A) whether they were from patients or controls. This was not unexpected in that NCBI database searches suggest that the oligonucleotide primers used for this study could amplify a sizable proportion of the 10,000 HERV-K LTRs located within the human genome (6Löwer R Löwer J Kurth R Proc. Natl. Acad. Sci. USA. 1996; 93: 5177-5184Crossref PubMed Scopus (611) Google Scholar). We concluded that these RT-PCR data could be explained if most of the plasma samples contained either IDDMK1,222-like RNA, genomic DNA, or both. In order to minimize the impact of possible contamination with genomic DNA, all RNA samples were incubated with DNase prior to their amplification using separate oligonucleotide primers complementary to the IDDMK1,222 env gene, in order to reduce the likelihood that small amounts of genomic DNA could bias the results. As a further precaution, RNA samples were also processed without the RT step, to detect remnant genomic DNA. The Southern blot of these RT-PCR products revealed that 7 of 12 patients with type 1 diabetes and 6 of 12 control subjects had evidence of the IDDMK1,222-like env RNA sequences in the plasma (Figure 1B) without demonstrable contaminating genomic DNA (Figure 1C). This RT-PCR study suggests that HERV-K viremia appears to be a common human occurrence but not a finding specifically associated with the onset of type 1 diabetes. In further studies, we cloned the IDDMK1,222 env from peripheral blood genomic DNA of a patient with type 1 diabetes and sequenced 38 clones amplified using PCR primers complementary to the putative IDDMK1,222 env gene. A family of seven (RIC1 to -7) closely related HERV-K10-like sequences were found suggesting that several HERV-K10-like sequences similar to IDDMK1,222 are present in the human genome. All of the sequences show high degrees of homology with that of HERV-K10 with identical sequences between IDDMK1,222 and RIC1 (GenBank accession numbers: AF084864, AF084865, AF086866, AF086867, AF086868, AF086869, and AF086870). Our findings, taken together with those of Conrad et al., rather strongly suggest that IDDMK1,222 should be considered to be one of the multiple variants of HERV-K10 rather than a specific entity. On the nucleotide level, there is greater than 90% homology between HERV-K10 and IDDMK1,222 env gene. The amino acid variability of the env and pol sequences of HERV-K10, and the IDDMK1,222 is less than 5% (4Conrad B Weissmahr R.N Böni J Arcari R Schüpbach J Mach B Cell. 1997; 90: 303-313Abstract Full Text Full Text PDF PubMed Scopus (355) Google Scholar) This is similar to the variance observed between the other HERV-K10-like clones isolated from genomic DNA (5Löwer R Tönjes R.R Korbmacher C Kurth R Löwer J J. Virol. 1995; 69: 141-149Crossref PubMed Google Scholar), which is markedly less than that observed in env genes of other RNA viruses. Furthermore, it is unlikely that IDDMK1,222 is a functional virus as it has type 1 HERV-K genome with a 292-nucleotide deletion at the pol–env boundary (5Löwer R Tönjes R.R Korbmacher C Kurth R Löwer J J. Virol. 1995; 69: 141-149Crossref PubMed Google Scholar). Whereas expression of RNA transcripts of IDDMK1,222-like proviruses appears not to have specificity for diabetes, it is conceivable that the virus could provoke a pathogenic immune response in an individual genetically predisposed to type 1 diabetes, by a Sag-mediated expansion of an autoreactive T cell subset, or by other mechanisms. To evaluate whether patients with type 1 diabetes produced an immune response to IDDMK1,222, immunoblots were performed using recombinant IDDMK1,222 env (∼20 kDa of RIC clone 1) and HERV-K Gag proteins. Western blot studies were performed using the bacterial lysates with patient sera at a 1 in 250 dilution. Seroreactivity to the IDDMK1,222 Env protein was observed in 8 of 19 diabetic and 7 of 17 control patients (Figure 2C). Two of four of the authors (N. M. and M. L.) tested were found to be strongly positive for Env antibodies. Further dilutions of patient sera to detect the highest titers of Env antibodies did not bias the findings toward the diabetic patients nor alter the lack of specificity of the study. Since immunoreactivity to IDDMK1,222 proteins may have been more evident prior to diagnosis of diabetes at an earlier point in the natural histories of the patients, we also studied 19 type 1 diabetic relatives who were positive for islet cell autoantibodies (ICA) and 20 without ICA since the former group only are at increased risk for the disease. Ten of the ICA+ group were found positive for the Env antibodies compared to 11 of the ICA− relatives. For the HERV-K Gag immunoblot study, the “GH” human teratoma-derived virus HTDV/HERV-K-producing cell line (kindly provided by Dr. Roswitha Löwer, Paul-Ehrlich Institute, Langen, Germany) was used as the source of HERV-K Gag protein as this agent is a HERV-K10-related virus (5Löwer R Tönjes R.R Korbmacher C Kurth R Löwer J J. Virol. 1995; 69: 141-149Crossref PubMed Google Scholar). Reactivity to the 80 kDa HTDV Gag protein was used as evidence for reactivity to HERV-K10 Gag (2Boller K Janssen O Schuldes H Tönjes R.R Kurth R J. Virol. 1997; 71: 4581-4588PubMed Google Scholar). Seroreactivity to HTDV Gag was detected in 6 of 16 diabetic patients and 5 of 13 controls. Further, immunoreactivity to the 80 kDa HTDV Gag protein was observed in approximately a third of healthy laboratory personnel (data not shown). Our studies suggest that IDDMK1,222 and HERV-K10-like viremias are common in man and that many healthy subjects develop immune responses to them, as demonstrated by their seroreactivity to IDDMK1,222 Env and/or HTDV Gag proteins. However, we have not assessed T lymphocyte reactivity to IDDMK1,222 proteins nor assessed whether the IDDMK1,222 Env protein has Sag activity. Nevertheless, we were unable to confirm the absolute specificity reported by Conrad and colleagues of IDDMK1,222 viremia in patients with acute onset diabetes (4Conrad B Weissmahr R.N Böni J Arcari R Schüpbach J Mach B Cell. 1997; 90: 303-313Abstract Full Text Full Text PDF PubMed Scopus (355) Google Scholar).∣∣To whom correspondence should be addressed. Supported by RO1 HD 19469-14.

Our objective was to ascertain the frequency of antibodies to glutamic acid decarboxylase (GAD) in Europids and four Asian ethnic groups with insulin-dependent diabetes mellitus (IDDM) to gain insight into why the prevalence and incidence of IDDM varies so widely among ethnic and/or geographically diverse population groups. The subjects in this study were Europid (n = 49), Japanese (n = 16), Thai (n = 7), Korean (n = 21), and Chinese (n = 13) persons with IDDM with a duration ranging from 5 to 14 years. There were similar numbers of healthy controls matched for each ethnic group. A validated radioimmunoprecipitation assay used GAD from pig brain radiolabeled with 125I using chloramine T. Islet cell cytoplasmic antibodies measured by indirect immunofluorescence were expressed as Juvenile Diabetes Foundation units. The prevalence of antibodies to GAD, compared with Europids (63%), was much lower in all Asian populations with IDDM: Japanese (31%), Thai (29%), Korean (5%), and Chinese (27%). The mean level of antibodies to GAD, however, among diabetics from each population who gave a positive reaction, was similar. For all groups, the prevalence of antibodies to GAD was much higher than that of islet cell cytoplasmic antibodies. Almost all IDDM subjects positive for islet cell antibodies had antibodies to GAD, but the converse did not hold. A radioimmunoprecipitation assay for antibodies to GAD applied to serum from subjects with IDDM in various ethnic groups showed that Europids with IDDM had a much higher prevalence of such antibodies than did Asians. This held for all ethnic groups, and particularly Koreans. Thus, among different populations, there may be etiologic heterogeneity of IDDM.(ABSTRACT TRUNCATED AT 250 WORDS)