The described strategy for genetic prediction of Type 1 diabetes mellitus relies on the convenient genotyping procedure and could be applied in large scale screening projects such as DIPP.

Several studies provide evidence that in addition to the DQ-DR genes, HLA contains another uncharacterized gene or genes associated with type 1 diabetes. Our aim was to investigate the effect of this gene independently of the DQ-DR genes and to localize it with a matched case-control study. More than 1,400 patients and 30,000 control individuals from Finland were studied. They were first genotyped for the selected alleles of the HLA-DQB1, -DQA1, and -DRB1 genes. For the DR3/4(0404) genotype, 75 patients and 181 control subjects were stratified, and 241 patients and 354 controls were stratified for the DR3/4(0401) genotype. Ten microsatellite markers in the HLA class III and I regions (D6S273, TNFa, C12A, STR MICA, MIB, C125, C143, C245, C3211, and MOGc) and selected alleles of the HLA-A and HLA-B genes were studied. In the DR3/4(0404)-stratified group, we found that markers located between C12A and C143 near the HLA-B gene confer a strong additional diabetes association. This was confirmed by the population differentiation test in both DR3/4(0404)-and DR3/4(0401)-stratified groups. Our data indicate that an additional gene associated with type 1 diabetes is located in the 240-kb region near HLA-B. We excluded STR MICA polymorphism as a mutation responsible for diabetes association. Diabetes 49:2217-2221, 2000 H LA localization of the genes predisposing to type 1 diabetes is well established. DQB1, DQA1, and DRB1 genes are considered primary and major loci associated with type 1 diabetes. However, evidence is accumulating that the HLA class I-III region contains gene(s) associated with an increased risk for type 1 diabetes (1-4) or its earlier age at onset (5-8). We searched for an additional HLA gene associated with type 1 diabetes among Finnish patients and control subjects stratified for the DQ-DR alleles.Initially, samples from >1,400 type 1 diabetic patients and >30,000 control subjects were collected and genotyped for DQB1 (9). DQB1*02/0302 subjects were then studied for the DQA1*05 and *0201 alleles and DR4 subtypes (of which only DRB1*0401 and DRB1*0404 are frequent among Finns). Thus, a group stratified for the DQA1*05-DQB1*02/ DQB1*0302-DRB1*0404 [DR3/4(0404)] genotype (75 patients and 181 control subjects) and a group stratified for the DQA1*05-DQB1*02/DQB1*0302-DRB1*0401 [DR3/4(0401)] genotype (241 patients and 354 control subjects) were selected. These subjects are all DR3/DR4 haplotype heterozygotes: one group carries DR404 and the other DR401 haplotype. Thus, within each group we excluded the diabetogenic effect of the known DQ-DR markers.In these groups, selected alleles of the HLA-B and HLA-A genes and 10 microsatellites (D6S273, TNFa, C12A, STR MICA, MIB, C125, C143, C245, C3211, and MOGc) were studied to map a non-class II HLA gene associated with type 1 diabetes susceptibility. These cover distance of ~2,300 kb and include the whole HLA class I region and the class I-III boundary (Fig. 1).We found that in the DR3/404-and DR3/401-stratified groups there was only one allele at each microsatellit...

The high incidence of insulin-dependent diabetes mellitus (IDDM) in Finland contrasts strikingly with the low rates in the neighbouring populations of countries in the Eastern Baltic region: Estonia, Latvia and Russia. To evaluate the possible contribution of genetic factors to these differences, the frequencies of HLA-DQB1 alleles and relevant DQB1-DQA1 or DQB1-DRB1 haplotypes associated with IDDM risk or protection were analysed among IDDM patients and control subjects from these four populations. An increased frequency of HLA-DQB1*0302, DQB1*02-DQA1*05 and DQB1*0302-DRB1*0401 was observed in subjects with IDDM in all studied populations, whereas the prevalence of DQB1*0301 and DQB1*0602 and/or *0603 was decreased among patients. The degree of IDDM risk associated with HLA alleles analysed here did not differ significantly between the populations. Comparisons of the distribution of IDDM-related HLA alleles and haplotypes in the background populations revealed its consonance with IDDM incidence. The combined frequency of high risk genotypes was significantly higher among Finns than in other populations studied. Our data support the hypothesis that variance in the dispersion of HLA alleles is the genetic basis of variation of IDDM incidence observed in the Eastern Baltic region.

We determined the distribution of DR4 subtypes in 309 DQB1*0302-positive haplotypes found in insulin-dependent diabetes mellitus (IDDM) patients and 70 control haplotypes present only in healthy family members. An increased frequency of DRB1*0401 allele (74.4% vs. 55.7%, P = 0.003) and a decrease of DRB1*0404 allele (23.6% vs. 40.0%, P = 0.0064) was revealed. A further analysis of extended haplotypes demonstrated strong linkages between various B alleles and DRB1*04 subtypes. HLA-B39 was more frequent in DRB1*0404-DQB1*0302-positive IDDM haplotypes compared with control ones (37.0% vs. 14.3%, P = 0.049), suggesting an involvement of the region telomeric to HLA-DRB1 in the susceptibility to IDDM.

This is the largest and most comprehensive study of the VDR gene in GD to date and these data suggest that these polymorphisms of the VDR gene do not contribute to GD susceptibility in the UK.