Other HLA-encoded susceptibility determinants
The DR and DQ genes cannot completely explain the association between T1DM and the HLA gene region. Teasing out the contribution of other HLA loci, however, is complicated by the high level of linkage disequilibrium within the region, combined with the strong effects of DR and DQ. A number of different approaches have been employed to address this issue, including the analysis of case/control data matched for specific DR-DQ combinations, investigation of the transmission to affected and unaffected offspring of heterozygous markers from parents homozygous for DR-DQ alleles, conditional logistic regression and conditional haplotype analysis.These, and other methods, have identified additional disease risk markers at the HLA-A, -B, -C, and -DPB1 loci, in the “class III” gene region and in loci located telomeric and centromeric of the classical HLA gene region (see Figure 30.1).
A number of association studies have implicated variants of the DPB1 gene in T1DM risk, but results are conflicting between datasets. A recent analysis of high-resolution genotyping data in a large family-based cohort fromthe Type 1 Diabetes Genetics Consortium (T1DGC) showed that disease risk was increased by DPB1*0202 and DPB1*0301, while DPB1*0402 conferred protection from diabetes . These associations remained significant after relative predispositional analysis, adjusting for the effects of LD with the DR-DQ loci. The DPB1*0301 allele was found to increase disease risk both in the presence and absence of high-risk susceptibility DR-DQ alleles. In contrast the predisposing effect of DPB1*0202 was seen only in individuals carrying the high-risk and protective DR-DQ haplotypes, while DPB1*0402 conferred protection only in the absence of the high-risk and protective DR-DQ haplotypes.
Similar conditional analyses in the T1DGC dataset have identified HLA class I alleles that influence the risk of diabetes independently of DR and DQ [35,36]. The most significant disease associations reported by Noble and coworkers  were with HLA-B*5701 (protective) and B*3906 (predisposing). Increased disease risk was also conferred by A*0201, A*2402, B*1801, and C*0501, while A*1101, A*3201, A*6601, B*0702, B*3502, B*4403, C*0401, and C*1601 were all associated with protection. The predisposing effects of HLA-A*24, B*39, and B*18 and reduced disease risk conferred by A*11 were confirmed by Howson et al. . In addition, HLA-B*13 and B*50 were found to increase T1DM risk independently of DR and DQ, while A*01 conferred protection. It has been postulated that HLA class II molecules play an important role in the initiation of the autoimmune response in T1DM, while class I molecules mediate the later stages of β-cell destruction.
A number of other loci in the MHC have been reported to influence the risk of T1DM independently of DR and DQ, including the TNFA, AIF1, PRSS16, and ITPR3 genes (Figure 30.1), although the association with the latter marker could not be confirmed in an independent study and may be secondary to the effect of DQB1 . Single nucleotide polymorphisms (SNPs) mapping close to LTA and CFB have also been associated with disease on DR3 and DR4 haplotypes. Overall HLA genes contribute to 40–50%of the genetic risk of T1DM, with the strongest effect seen in individuals developing the disease at an early age. As the incidence of T1DM has risen over past decades, however, a temporal change has been observed in the impact of specific genotypes on disease risk . The proportion of patients with high-risk HLA genotypes has decreased, suggesting that greater environmental pressure is boosting disease penetrance in subjects with moderate- and low-risk HLA genotypes.
Other candidate genes for T1DM
The candidate gene approach has been used to identify a number of T1DM risk loci outside the HLA gene region. This method seeks disease associations with sequence variants at loci thought to play a biologic role in disease etiology. The insulin gene and immune response genes are the most logical candidates for T1DM and a large number of polymorphisms have been reported to be associated with disease risk over the last 30 years. Many of these have proved to be false positive results, artefacts caused by limited sample sizes and population stratification. Some, however, have been replicated in independent datasets and have functional data to support their role in disease etiology. These validated loci are described in the following sections and the role of the immune response genes in T-cell activation is illustrated in Figure 30.3.
The insulin gene (INS)
Outside the HLA locus, the strongest signal for T1DM risk maps within and upstream of the INS gene on chromosome 11p15.5 (OR∼2.0). It is unclear whether the etiologic variant is a variable number of tandem repeats (VNTR), located approximately 0.5 kb upstream of INS, or two noncoding SNPs that are in complete LD with the VNTR . Homozygosity for class I VNTR alleles (26–63 repeats) is associated with an increased risk of T1DM, while class III alleles (140–210 repeats) are generally dominantly protective.These markers appear to exert their influence by modulating insulin gene transcription in the thymus. INS expression from class I alleles is more than twofold lower compared with that driven by class III alleles . The resulting low level of insulin in the developing thymus of class I homozygotes is thought to lead to autoimmunity by impairing the deletion of insulin-specific T cells. In contrast, the higher level of insulin present in class III allele carriers promotes T-cell tolerance to the autoantigen and protects against T1DM. In support of this hypothesis, rare class III alleles that completely repress thymic insulin expression confer an increased risk of diabetes .