| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Diabetes Research Institute (A.P., M.Z., C.R.) and the Department of Neurology (C.T.M.), University of Miami School of Medicine, Miami, Florida 33136; the First Department of Internal Medicine, Nagasaki University School of Medicine (E.K.), Nagasaki, Japan; the Barbara Davis Center for Childhood Diabetes, University of Colorado Health Sciences Center (E.K., L.Y., S.B., G.S.E.), Denver, Colorado 80262; the Section of Endocrinology, Department of Internal Medicine, Yale School of Medicine (M.S.), New Haven, Connecticut 06510; the Division of Immunogenetics, Children’s Hospital of Pittsburgh, University of Pittsburgh (M.P., R.P.F., M.T.), Pittsburgh, Pennsylvania 15213; the Human Genetics Department, Roche Molecular Systems (M.A., J.A.N.), Alameda, California 94501; and the Children’s Hospital Oakland Research Institute (J.A.N., H.A.E.), Oakland, California 94609
Address all correspondence and requests for reprints to: Alberto Pugliese, M.D., Diabetes Research Institute, University of Miami School of Medicine, 1450 NW 10th Avenue, Miami, Florida 33136. E-mail: apuglies{at}mednet.med.miami.edu
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In contrast, HLA-DR2 haplotypes are rarely observed among patients with type 1 diabetes. Most of the HLA-DR2-associated type 1 diabetes in Caucasians is accounted for by three neutral haplotypes: DQA110102/DQB110502/DRB111601, DQA110103/DQB110601/DRB111501, and DQA110103/DQB110601/DRB111502 (6, 16, 17, 18, 19, 20). The DQA110102/DQB110602/DRB111501 haplotype, the most common DR2 haplotype among Caucasians, is the only DR2 haplotype conferring dominant and almost absolute protection from type 1 diabetes among Caucasian and other racial groups. Indeed, patients carrying this haplotype are extremely rare (4, 6, 16, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30). The characterization of both common and rare recombinant DR2 haplotypes observed in patients with type 1 diabetes also indicates that DQB110602 is the only class II allele exclusively found on diabetes-protective DR2 haplotypes. This suggests that protection is mostly, although not exclusively, conferred by DQB110602 (which together with DQA110102 codes for a protective DQ heterodimer) (18, 31, 32). Moreover, our previous studies indicate that DQB110602 confers strong diabetes protection even among islet cell antibodies (ICA)-positive first degree relatives of patients with type 1 diabetes (33). In fact, approximately 7% of ICA-positive relatives identified through autoantibody screening carry DQB110602, and in our family study, none of such relatives has developed diabetes on follow-up. Because of the apparent dominance of the protective effect (19, 20) and the extremely low risk ascertained in prospective studies (33, 34), ICA/DQB110602-positive first degree relatives are presently excluded from receiving treatment in the Diabetes Prevention Trial–Type 1, a major ongoing trial involving several centers in the United States (35).
Of interest, a few novel variants of the DQB110602 and DQB110603 alleles were recently reported by Hoover et al. (36) in four rare patients originally typed as DQB110602-positive with conventional sequence-specific oligonucleotide (SSO) typing techniques (20). Such variants, not distinguished by the initial panel of SSO probes, appear to be permissive for the development of diabetes. This observation suggests the hypothesis that DQB110602 may confer absolute protection and may never occur in patients with type 1 diabetes. If this hypothesis is, in fact, correct, all or most of those rare patients typed as DQB110602 with conventional techniques could carry one of the above diabetes-permissive variants instead of the protective DQB110602. It is also conceivable that such variants may occur in the previously described ICA/DQB110602-positive relatives (33); if so, this could have significant prognostic implications, as their risk of developing diabetes may be higher than previously estimated.
In addition, a mitochondrial DNA mutation (mtDNA A3243G) has been recently associated with the development of late-onset diabetes in ICA-positive individuals and was also reported in patients with various forms of diabetes (37, 38, 39, 40, 41). The presence of such a mutation could be associated with the development of diabetes in ICA/DQB110602-positive individuals.
We therefore investigated the occurrence of diabetes-permissive DQB110602 variants and the mtDNA A3243G mutation in the eight ICA-positive relatives previously identified (33) and in six rare patients with type 1 diabetes, all typed as DQB110602-positive with conventional techniques.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Unaffected ICA/DQB110602-positive first degree relatives (n
= 8) were identified through autoantibody screening of family members
of patients with type 1 diabetes at the Joslin Diabetes Center and the
Barbara Davis Center for Childhood Diabetes, as previously reported
(33). All relatives were Caucasians. Their clinical characteristics and
updated follow-up information are shown in Table 1
. ICA positivity was defined as
measurements of 20 Juvenile Diabetes Foundation units or more on at
least two occasions, and the ICA titer shown in Table 1 is the highest
level observed for each individual.
|
) according to the criteria of the
National Diabetes Data Group (42). The diagnosis was confirmed in four
patients by the presence of autoantibodies against islet cell antigens
(serum was unavailable for testing the other two patients). Table 2
shows the autoantibody profiles for the four patients tested (two
Caucasians and two African-Americans). One of the African-American
patients (no. 17865) also had stiff-man syndrome (SMS) and developed
diabetes at age 35 yr (43, 44). SMS is a rare neurological disorder
reportedly associated with type 1 diabetes and autoimmune responses to
the GAD65 autoantigen (glutamic acid decarboxylase, 65-kDa isoform), as
in this case (45, 46). This patient was particularly interesting and
unusual because he developed diabetes despite having DQB110602, an
allele that we reported as diabetes protective even among patients with
SMS (43, 44). The remaining two Caucasian patients with DQB110602 were
identified through the Human Biological Data Interchange
repository, a collection of families with type 1 diabetes (47); as
already mentioned, no serum samples from these patients were available
for autoantibody testing.
|
Genomic DNA samples were extracted from heparinized blood samples with phenol/chloroform. HLA typing for DQA1, DQB1, and DRB1 alleles was performed on all subjects by the PCR and hybridization with SSO probes as previously described (4, 33). For one patient, DRB1 alleles were determined by direct sequencing. The second exon of the DRB1 gene was amplified using forward primers for the DRB1115/16 or DRB1104 subtypes paired with a generic DRB1 reverse primer in separate PCR reactions. PCR products were cloned and sequenced with the same strategy described below for DQA110102 and DQB110602 alleles.
Sequencing exon 2 of the DQB110602 and DQA110102 alleles
The second exon of both the DQA1 and DQB1 genes was sequenced, because this exon encodes the extracytoplasmic polymorphic regions of the HLA-DQ molecule. The DQA1 gene was amplified from genomic DNA by PCR with primers 0102ex2-F 5'-CT GAC CAC GTT GCC TCT TGT-3' and 0102ex2-R 5'-ATT GGT AGC AGC GGT AGA GTT-3'. Primers were selected at the 5'- and 3'-termini of the second exon (intron sequences are not currently available) and amplify a 261-bp product (annealing, 58 C). The second exon of the DQB1 gene was amplified with primers 0602ex2-F 5'-TC CCC GCA GAG GAT TTC GTG T-3' and 0602ex2-R 5'-TCC TGC AGG GCG ACG ACG CTC ACC TCT CC-3' (annealing, 60 C). These primers amplify a 302-bp product including 8 and 18 bp of intronic sequences at the 5'- and 3'-ends, respectively, and allow distinction of the polymorphism reported at codon 9 by Hoover et al. and Williams et al. (36, 48). Primers MADQ4 5'-CCT GAC TGA CCG GCC GGT GAT-3' and UG71 5'-ACA TGT AAA ACG ACG GCCAGT TCT CCT CTG CAG GAT CCC GC-3' were also used and allowed discrimination of the codon 9 polymorphism as well (299-bp product; annealing, 55 C). The PCR products were cloned into the TA Cloning Vector (Invitrogen, San Diego, CA). DNA samples extracted from transformed colonies were screened by PCR using sequence-specific primers for the DQB110602 and DQA110102 alleles (PCR-SSP) as described by Olerup et al. (49). DNA samples from DQB110602- and DQA110102-positive colonies were then sequenced (both strands, using the SP6 and T7 primers). PCR products generated with the MADQ4/UG71 primer pair were sequenced directly using MADQ4 as primer and dye terminators FS (ABI). The other strand of each product PCR was also sequenced as a control using the dye primer FS kit (ABI, Foster City, CA) containing the -21M13 primer. Sequence analysis was performed on an ABI 373 automated sequencer.
Typing for the mtDNA A3243G gene mutation
Typing for the A3243G transfer ribonucleic acidLeu
(UUR) mutation was performed by PCR-restriction fragment length
polymorphism as previously described (50). In brief, approximately 0.5
µg genomic DNA was subjected to PCR amplification (94 C for 60
s, 55 C or 60 s, 72 C for 45 s; 25 cycles) using primers for
the mtDNA light strand positions 3116–3134 and heavy strand positions
3353–3333 (51). The DNA fragment produced by the PCR reaction was
digested with the restriction enzyme HaeIII for 2 h at
37 C. The A
G transition at position 3243 creates a new
HaeIII site that is diagnostic for the mutation (50). The
digestion products were electrophoresed on a 12% nondenaturing
polyacrylamide gel and stained with ethidium bromide. This technique
can detect mitochondrial mutations with low degree of heteroplasmy in
peripheral blood (as low as 5%, below which any significant influence
on the mitochondrial function in more relevant tissues is
unlikely).
Autoantibody testing
The eight unaffected ICA/DQB110602-positive relatives were tested for the presence of ICA with a standardized assay as previously described (33, 52). ICA positivity was defined as measurements of 20 Juvenile Diabetes Foundation units or more on at least two occasions. Serum samples from these relatives and the other subjects studied were also tested for the presence of autoantibodies against three major type 1 diabetes autoantigens, such as insulin, GAD65, and the tyrosine phosphatase-like ICA512 (or IA-2). Insulin autoantibodies (IAA) were determined with a fluid phase RIA using 600 µL serum, with duplicate determinations with and without unlabeled insulin for competition (53). The interassay coefficient of variation for the IAA assay is 10.3% at low positive values. The assay had a specificity of 91% and a sensitivity of 49% for new-onset patients less than age 30 yr in the 1995 Immunology of Diabetes Society (IDS) Combinatorial Workshop. Autoantibodies to GAD65 were measured in triplicate by RIA, using in vitro transcribed and translated GAD65 (clone provided by A. Lernmark) (54, 55). Autoantibody-bound GAD65 was precipitated with protein A-Sepharose. The interassay coefficient of variation of this assay is 6.5%. Assay specificity was 99%, and sensitivity was 83.7% for new-onset patients less than age 30 yr in the IDS Combinatorial Workshop. Autoantibodies to ICA512 were measured using the in vitro transcribed and translated labeled product of a clone termed ICA512bdc (amino acids 256–979 of ICA512/IA-2) (55, 56, 57, 58). Autoantibodies were measured in triplicate with protein A-Sepharose precipitation. This assay gave a specificity of 100% and a sensitivity of 74.4% for new-onset patients less than age 30 yr at the IDS Combinatorial Workshop. GAD65 and ICA512 autoantibody levels are expressed as an index calculated from the counts per min for the test sample and the positive and negative control samples. GAD65 and ICA512 autoantibodies were determined simultaneously with differential labeling (35S for ICA512 and 3H for GAD65) in an automated 96-well ß-counter.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We studied eight unaffected, ICA-positive first degree relatives
(Table 1
) and six rare patients with type 1 diabetes (Table 2), all
previously typed as DQB110602 positive with standard SSO typing
techniques. We performed direct sequence analysis of the second exon of
their DQB1 and DQA1 alleles to investigate whether any of the
diabetes-permissive DQB110602 variants recently described (36) occurred
among these subjects. All subjects carried conventional DQB110602 and
DQA110102 exon 2 sequences. Thus, all subjects carry DQA1 and DQB1
alleles coding for a diabetes-protective HLA-DQ heterodimer.
Typing for the mtDNA A3243G gene mutation
None of the subjects studied carried the mtDNA A3243G mutation reportedly associated with the development of late-onset diabetes in ICA-positive individuals.
Additional follow-up and autoantibody testing in ICA/DQB110602-positive first degree relatives
All of the previously identified ICA/DQB110602-positive relatives
were tested to determine whether they express other autoantibodies
besides ICA. The upper limits of normal are 42 nU/mL for IAA and
indexes of 0.032 and 0.071 for GAD65 and ICA512 autoantibodies,
respectively. Autoantibody levels shown in Tables 1 and 2 are the mean
levels for each individual. As illustrated in Table 1, seven of eight
relatives have GAD65 autoantibodies but only two of eight and one of
eight have autoantibodies against insulin and ICA512, respectively.
None of these relatives have developed type 1 diabetes to date with
further extended follow-up (Table 1; mean follow-up ±
SD, 8.55 ± 4.93 yr; a mean increase of 2.6 yr in
follow-up length since our first report) (33).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Among ICA/DQB110602-positive relatives, 7 of 8 have GAD65
autoantibodies but only 2 of 8 and 1 of 8 have autoantibodies against
insulin and ICA512, respectively (Table 1). Thus, ICA reactivity
appears to be mainly directed against GAD65 in most
ICA/DQB110602-positive relatives (59, 60). In contrast, GAD65
autoantibodies were detected only in 2 of 13 ICA-positive patients from
Japan with the mtDNA A3243G mutation, none of whom had DQB110602 (37).
As diabetes risk also correlates with the number of autoantigens
targeted by antiislet immune responses (61), it is apparent that
ICA/DQB110602-positive relatives express a low risk phenotype with only
limited loss of tolerance to islet cell autoantigens (60). Indeed,
despite ICA positivity, none of these relatives has developed type 1
diabetes to date during further extended follow-up. Moreover, most
relatives (5 of 8) are now over 40 yr old. Together, these observations
confirm that ICA/DQB110602-positive relatives are a distinct group of
individuals with extremely low risk of diabetes, even at an older age.
These findings validate our previous report that DQB110602
significantly protects from type 1 diabetes even in the presence of
ICA/GAD65 autoantibodies (33), an observation that has significant
implications for the design of prevention trials.
We also investigated the occurrence of diabetes-permissive DQB110602 variants in six rare patients with type 1 diabetes and DQB110602 (as defined with conventional typing techniques). Although a number of patients with DQB110602 may suffer from rare genetic forms of diabetes rather than type 1 diabetes [e.g. Wolfram’s syndrome, type 1 autoimmune polyendocrine syndrome, type 1B diabetes, and maturity onset diabetes of the young (MODY)], the clinical diagnosis of type 1 diabetes was confirmed in four of six patients by the presence of autoantibodies against ICA. The diagnosis was exclusively based on clinical criteria for the two patients identified through the Human Biological Data Interchange repository of families with type 1 diabetes (42). We did not observe the diabetes-permissive variants reported by Hoover et al. (36), and all six patients carry normal DQB110602 and DQA110102 sequences. Moreover, none of the patients has the A3243G mtDNA mutation. Thus, our findings do not support the hypothesis that the above variants may commonly occur and be specifically associated with diabetes development in conventionally typed DQB110602-positive subjects.
Our results also imply that the protective effect associated with DQB110602 is not absolute, as type 1 diabetes may develop in extremely rare cases in individuals carrying the DQB110602 allele. Perhaps other polymorphisms regulating the phenotypic expression of DQB110602, including polymorphisms in the DQA1/DQB1 promoter regions, or other genetic loci/environmental factors may modulate the protective effect associated with DQB110602. For instance, patient 17865, an African-American with diabetes and SMS, has the DRB111503 allele instead of the DRB111501 allele. The DRB111501 allele, usually found in linkage disequilibrium with DQB110602 on protective DR2 haplotypes, may theoretically contribute to the protective effect and, conversely, the DQB110602/DRB111503 combination may be less protective. However, the analysis of recombinant haplotypes suggests that most of the protection derives from the DQB110602 allele (18, 31, 32); moreover, one of the ICA-positive relatives (no. 1271) carries the unusual combination of DQB110602 in cis with the diabetes-predisposing DRB110301 allele (DR3) and has not developed diabetes despite having ICA/GAD65 autoantibodies for more than 12 yr. We also identified a patient with type 1 diabetes carrying DQB110603 in cis with DQA110102 and DRB111501, once again suggesting that DQB110602 is central to diabetes protection (data not shown).
In contrast to our findings, none of the five DQB110602-positive
patients with type 1 diabetes originally described by Baisch et
al. (20) were found to carry DQB110602 by Hoover et al.
(36); one patient was initially mistyped and had DQB110603, and four
patients carry alleles with a sequence closely related to that of
DQB110602 (n = 1) and DQB110603 (n = 3). Of note, DQB110603
is usually found on haplotypes bearing the DQA110103 and DRB111301
(DR6) alleles, and such haplotypes are significantly less protective
than the DQA110102/DQB110602/DRB111501 (DR2) haplotype. It remains
unclear whether the DQB110603 variants described by Hoover (termed
0603a, 0603b, and 0603c) are linked to DRB111501 (DR2) or DRB111301
(DR6) alleles; however, in all instances those variants were found in
cis with DQA110103, and based on linkage disequilibrium, one
would predict that none of those patients carries a DRB111501 (DR2)
haplotype. The allele variant directly related to DQB110602, also
reported by Williams et al. and now termed DQB110611 (48),
differs from DQB110602 at codon 9, where a TA substitution
determines a change in the amino acid sequence (PheTyr). Molecular
modelling studies suggest that such sequence variation may alter the
peptide-binding site of the HLA-DQ molecule (36), and it is also of
interest that the same codon 9 polymorphism is shared with the most
common susceptibility alleles (DQB110302 and DQB110201) and other
diabetes-permissive alleles (DQB110502, DQB110604, and DQB110603c) (36, 62). However, the same polymorphism is also found in alleles that are
not associated with increased susceptibility (i.e. DQB110603
and DQB110301), and susceptible alleles such as DQB110401 share codon 9
sequence with DQB110602 instead. It should also be noted that none of
the diabetes-permissive variants differs from DQB110602 at positions 57
and 70, the combined variation of which reportedly modulates peptide
binding and diabetes susceptibility (36, 62).
As regards the frequency of the diabetes-permissive variants, DQB110611
is extremely rare, as it was detected in only 1 of 30 African-Americans
and 0 of 62 non-Hispanic/Hispanic whites with DQB110602 (48). No
frequency data are available for the other variants (DQB110603a,
DQB110603b, and DQB110603c), but one would speculate that these alleles
are probably very rare, as they would have been reported much earlier
had they occurred at a significant frequency in the population. This is
consistent with our findings confirming the presence of DQB110602 in
all of the relatives and patients studied. Nonetheless, given the
significant prognostic implications associated with DQB110602,
the current typing protocols should be modified to allow the
unequivocal identification of DQB110602 when typing relatives of
patients with type 1 diabetes. This can be easily achieved
with sequence-specific primers that selectively amplify DQB110602
(PCR-SSP) with a strategy based on the protocols of Olerup et
al. (49) and Williams et al. (48), as illustrated in
Table 3.
|
In conclusion, we show that all the ICA/DQB110602-positive relatives previously described carry conventional DQB110602 and DQA110102 sequences coding for a protective heterodimer. Such relatives appear to have a low risk of type 1 diabetes, as also confirmed by their autoantibody profiles and extended follow-up data presented. Thus, our findings provide additional confirmation for the dramatic protective effect associated with DQB110602, even among relatives with autoantibody positivity. However, our finding that DQB110602 is present in rare patients with type 1 diabetes suggests that protection is not absolute, perhaps because of other unknown genetic or environmental factors that may modulate the protective mechanisms associated with DQB110602. Finally, although putative diabetes-permissive DQB110602 variants appear to be extremely uncommon, we suggest a simple typing strategy, based on the protocols of Olerup et al. (49) and Williams et al. (48), that can be easily applied to allow the unequivocal identification of the DQB110602 allele when typing results are to be used for predicting type 1 diabetes.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received August 7, 1998.
Revised February 3, 1999.
Accepted February 8, 1999.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
, and DQß gene polymorphism in 23
DR2-positive, insulin-dependent diabetes mellitus patients. Hum
Immunol. 33:140–147.[CrossRef][Medline]
24JQ T cells in type 1 diabetes. Nature. 391:177–180.[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  M. J. Redondo, S. Babu, A. Zeidler, T. Orban, L. Yu, C. Greenbaum, J. P. Palmer, D. Cuthbertson, G. S. Eisenbarth, J. P. Krischer, et al. Specific Human Leukocyte Antigen DQ Influence on Expression of Antiislet Autoantibodies and Progression to Type 1 Diabetes J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1705 - 1713. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. A. Al-Jenaidi, S. F. Wakim-Ghorayeb, A. Al-Abbasi, M. R. Arekat, N. Irani-Hakime, P. Najm, K. Al-Ola, A. A. Motala, and W. Y. Almawi Contribution of Selective HLA-DRB1/DQB1 Alleles and Haplotypes to the Genetic Susceptibility of Type 1 Diabetes among Lebanese and Bahraini Arabs J. Clin. Endocrinol. Metab., September 1, 2005; 90(9): 5104 - 5109. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. L. Edghill, A. L. Gloyn, K. M. Gillespie, A. P. Lambert, N. T. Raymond, P. G. Swift, S. Ellard, E. A.M. Gale, and A. T. Hattersley Activating Mutations in the KCNJ11 Gene Encoding the ATP-Sensitive K+ Channel Subunit Kir6.2 Are Rare in Clinically Defined Type 1 Diabetes Diagnosed Before 2 Years Diabetes, November 1, 2004; 53(11): 2998 - 3001. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. M. Al-Harbi, A.-J. Abbassi, H. Tamim, F. al-Jenaidi, M. Kooheji, M. Kamal, S. al-Mahroos, F. al-Nasir, A. A. Motala, and W. Y. Almawi Specific HLA-DRB and -DQB Alleles and Haplotypes Confer Disease Susceptibility or Resistance in Bahraini Type 1 Diabetes Patients Clin. Vaccine Immunol., March 1, 2004; 11(2): 292 - 296. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M A Kelly, M L Rayner, C H Mijovic, and A H Barnett Molecular aspects of type 1 diabetes Mol. Pathol., February 1, 2003; 56(1): 1 - 10. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. P. Krischer, D. D. Cuthbertson, L. Yu, T. Orban, N. Maclaren, R. Jackson, W. E. Winter, D. A. Schatz, J. P. Palmer, and G. S. Eisenbarth Screening Strategies for the Identification of Multiple Antibody-Positive Relatives of Individuals with Type 1 Diabetes J. Clin. Endocrinol. Metab., January 1, 2003; 88(1): 103 - 108. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Stenstrom, B. Berger, H. Borg, P. Fernlund, J. S. Dorman, and G. Sundkvist HLA-DQ Genotypes in Classic Type 1 Diabetes and in Latent Autoimmune Diabetes of the Adult Am. J. Epidemiol., November 1, 2002; 156(9): 787 - 796. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Yu, D. D. Cuthbertson, N. Maclaren, R. Jackson, J. P. Palmer, T. Orban, G. S. Eisenbarth, and J. P. Krischer Expression of GAD65 and Islet Cell Antibody (ICA512) Autoantibodies Among Cytoplasmic ICA+ Relatives Is Associated With Eligibility for the Diabetes Prevention Trial-Type 1 Diabetes, August 1, 2001; 50(8): 1735 - 1740. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Greenbaum, D. A. Schatz, D. Cuthbertson, A. Zeidler, G. S. Eisenbarth, and J. P. Krischer Islet Cell Antibody-Positive Relatives with Human Leukocyte Antigen DQA10102, DQB10602: Identification by the Diabetes Prevention Trial-Type 1 J. Clin. Endocrinol. Metab., March 1, 2000; 85(3): 1255 - 1260. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
