Two main approaches have been used in the search for type 1 susceptibility genes outside the HLA region. Linkage studies identify genes that are shared more frequently by affected individuals than by chance, and are most useful for relatively rare genes with large effect sizes. Association studies look for evidence of association between genes and disease in large populations, and are more appropriate for common genes with small effect size. Initial studies were targeted to candidate genes or pathways, but genome-wide association studies (GWAS) have proved the best way of finding genes with no previously recognised association to the disease. The largest effect sizes are seen for the insulin and PTPN genes. Up to 40 other genes conferring susceptibility to type 1 diabetes have been identified, but their effect sizes are modest. Most of the newly identified susceptibility genes can be positioned on immune activation pathways while some loci have yet to have the disease associated gene identified.
Non-HLA genes associated with type 1 diabetes (pre-genome wide association studies)
The insulin gene, the second genetic locus to be linked with susceptibility to type 1 diabetes, was identified in 1984 . The insulin gene (INS) on chromosome 11p15 encompasses 1430 base pairs (bp) and results in the translation of preproinsulin, the precursor of mature insulin. Preproinsulin is processed to proinsulin by removal of the signal peptide and then to mature biologically active insulin by removal of the C-peptide. Interest in this gene has been boosted by evidence that insulin is the primary autoantigen in type 1 diabetes. A variable number tandem repeat (VNTR) region consisting of a 14 to 15 bp consensus sequence upstream of the INS gene, in the INS promoter, contains three classes of alleles: there is a higher frequency of class I alleles with shorter repeat sequences in individuals with type 1 diabetes, while individuals with longer class III alleles are relatively protected from type 1 diabetes. The VNTR regulates transcription rates of insulin and its precursors. Class I and Class III alleles differentially affect transcription of insulin in the thymus and pancreas. Class III alleles result in 20% increased INS transcription in the thymus. This potentially results in more efficient negative selection of insulin reactive T cells and less susceptibility to type 1 diabetes as compared to class I alleles providing an attractive model for the role of the insulin gene in susceptibility to type 1 diabetes but this hypothesis remains to be experimentally demonstrated.
Table: Non HLA SNPs associated with type 1 diabetes. In order of risk associated with susceptible allele, highest risk associated with alleles of PTPN22 and INS.
|SNP||Chromosome||Probable associated gene||Relative risk (between homozygotes)*|
*Estimated from review by John Todd. “Etiology of Type 1 Diabetes”. Immunity. 32:457-467 (2010)
The cytotoxic T-lymphocyte antigen-4 gene (CTLA-4) encoded on chromosome 2q33 was identified as a further type 1 susceptibility gene. CTLA-4 is a surface molecule found on activated T cells that produces a negative signal by inhibiting the T cell receptor signaling complex ligand interactions (blocks binding of CD80 and CD86) (Figure 2). Two major splice forms exist, encoding membrane bound and soluble forms. When CTLA-4 is knocked out, lymphoproliferative disorders result. An A49G polymorphism in exon 1 of CTLA-4 changes the amino acid sequence resulting in reduced cell surface expression. It is thought that inherited changes in CTLA-4 gene expression can increase T cell self-reactivity and therefore play an important role in autoimmune diseases such as type 1 diabetes.
Protein tyrosine phosphatase non-receptor 22 (PTPN22), a gene found on chromosome 1p13 and that encodes lymphoid protein tyrosine phosphatase (LYP) was found to be associated with susceptibility to type 1 diabetes in 2004. Protein tyrosine phosphatases such as LYP are responsible for preventing spontaneous T cell activation and they have the ability to prevent the response to antigen by dephosphorylating and inactivating T cell receptors. It has been demonstrated that a single nucleotide polymorphism (SNP) in the PTPN22 gene can lead to susceptibility to autoimmune diseases such as type 1 diabetes because of a decrease in negative regulation of hyper-reactive T cells. The first complete resequencing of the human PTPN22 gene was carried out in 2005. This sequence was further analysed for polymorphisms associated with type 1 diabetes and a SNP at 1858bp in codon 620 was found. Two alleles referred to as 1858C and 1858T were identified and the 1858T variant was shown to occur more often in type 1 diabetes populations: 30.6% of people with type 1 diabetes compared with 21.3% healthy controls are heterozygous for the polymorphism (p=0.0006). LYP is expressed in other cells in addition to T cells including natural killer (NK) cells, B cells, macrophages and dendritic cells (DCs), and so could very well also have an effect on the function of several immune cells.
The interleukin 2 receptor alpha (IL2RA) region on chromosome 10p15 has also been associated with type 1 diabetes . IL2RA encodes the α-chain of the IL-2 receptor complex (also referred to as CD25), which is responsible for binding IL-2, a key player in the proliferation of regulatory T cells. IL2R has also been associated with type 1 diabetes in the non-obese diabetic (NOD) mouse. Two IL-2R SNPs associated with the increased risk of type 1 diabetes have been reported with ss52580101 the most closely associated. A recent study measuring expression of IL-2R in individuals homozygous for susceptible and protective SNPs associated with type 1 diabetes demonstrated that on stimulation, higher percentages of CD69+ CD4+ memory T cells secreted IL-2 from individuals with the protective SNP compared with individuals with the susceptible SNP. More recently susceptibility genotypes were found to be associated with lower levels of soluble IL2Ralpha (s_IL2Ra_) and in vitro stimulation of peripheral blood mononuclear cells from individuals with type 1 diabetes results in lower levels compared with healthy controls.
Figure 1: Chromosomal localisation of selected T1D associated loci (adapted from Ye et al. 2010)In recent years methodologies to identify susceptibility factors underlying complex disorders have improved by orders of magnitude. In particular the success of the HapMap project in identifying stretches of linkage disequilibrium has decreased the number of SNPs requiring genotyping combined with increased capacity for high throughput SNP analysis has resulted in a genetic revolution. In 2007, results of the first GWAS in seven different complex diseases was published by the Wellcome Trust Case Control Consortium. Later in 2007, a further genome-wide association scan was carried out and confirmed the additional associations of 12q24, 12q13, 16p13 and 18p11 with type 1 diabetes. More recently the Type 1 Genetics Consortium (TIDGC) has published over 40 genetic loci associated with type 1 diabetes. A selection of the best characterised are shown on Figure 1.
Genetic susceptibility to type 1 diabetes in the post-GWAS era
In 2006, interferon induced with helicase C domain 1 (IFIH1, also known as MDA-5) on chromosome 2q24.3 was found to be strongly associated with type 1 diabetes. A follow-up study on IFIH1 in 2008 confirmed the strongest association to be with SNP rs1990760. IFIH1 is particularly interesting because unlike the type 1 diabetes susceptibility genes discussed so far, it is not involved in T cell activation but contributes to innate immune responses by releasing the cytokine interferon-gamma (IFN-γ) and inducing apoptosis of cells infected by picorna viruses, of which enteroviruses such as coxsackie B4 which have been identified histologically in pancreases from individuals with type 1 diabetes. This molecule may therefore provide insights into the hypothesis that viral infection contributes to susceptibility to type 1 diabetes as alterations in IFIH1 activity could interfere with detection and clearance of virus. PBMC expression levels of IFIH1 have been reported to be higher in individuals with susceptibility genotypes. Resequencing of the IFIH1 gene identified four rare variants associated with protection against type 1 diabetes, which are predicted to play a role in altering the expression and structure of IFIH1. More recently it has been shown that islet autoantibody positive children with the rs2111485 GG genotype in IFIH1 progressed more quickly to diabetes (31% within 5 years) compared with children carrying the GA or AA genotypes (11% within 5 years). This suggests interaction between genetic and environmental determinants of type 1 diabetes.
TLR7 and -8
The Toll-like receptor (TLR) family plays a fundamental role in pathogen recognition and activation of innate immunity. TLRs are highly conserved and recognise pathogen-associated molecular patterns (PAMPs) that are expressed on infectious agents. TLR7 and -8 are located closely together on the X chromosome and have recently been associated with type 1 diabetes.TLR7 recognises single stranded RNA in endosomes, which is a common feature of viral genomes that are internalised by macrophages. Like IFIH1, the association of type 1 diabetes with these receptors strengthens arguments for the involvement of viruses in the pathogenesis of disease.
CCR5 is a chemokine receptor on the surface of several cells of the immune system including macrophages, NKT cells, CD4+ T cells and CD8+ T cells. It has been mapped to the short arm of chromosome 3 within the chemokine receptor gene cluster. Recent studies established that this gene comprises three exons spanning a region of about 6 kb. A 32 bp insertion/deletion polymorphism exon 3 changes the open reading frame of CCR5 and results in a non-functional protein. This polymorphism is present only in 1% of the population but deletion homozygotes are protected against HIV-1 infection as well as type 1 diabetes, rheumatoid arthritis and coeliac disease.
A chromosome 21q22.3 type 1-associated locus (rs876498) has been identified  and replicated. The only gene in the corresponding region of linkage disequilibrium is the ubiquitin associated and SH3 domain containing A gene (UBASH3A) which comprises 15 exons, spans 40kb, and has been shown to be expressed in spleen, bone marrow and peripheral blood lymphocytes. UBASH3A may therefore provide a candidate for the increased frequency of autoimmune disease in Down's syndrome.
NK receptor/HLA class 1 interactions
Natural killer (NK) cells represent the first line of defence against viral infection. NK cell infiltrates have been identified in pancreas from individuals with type 1 diabetes and increased NK cell activity has been reported in the periphery of individuals with type 1 diabetes. NK cells act by either activating or inhibiting cytolysis and their activity is controlled by the balance of inhibitory and activating receptors on the cell surface. One set of human NK cell receptors are the killer immunoglobulin-like receptor gene (KIR) gene family on chromosome 19 that consists of 16 genes; each is either inhibitory or activating in function and is polymorphic both in terms of gene content and allelic variation. GWAS of KIR are not yet available because this region of chromosome 19 does not have a high coverage SNP map but results from genetic studies of KIR in type 1 diabetes increasingly show an association with the activating receptor KIR2DS2 (and its ligand, HLA C Group I. van der Slik et al. (2003) analysed the KIR gene family and respective HLA class I ligands in 149 children diagnosed with diabetes under the age of 14 in a Dutch population, and Shastry et al. (2008) carried out a similar analysis in 98 patients diagnosed under the age of 18 years compared with 70 healthy controls in a Latvian population. In addition, Ramos-Lopez et al. (2009) showed in a combined German/Belgian study of 1124 patients and 716 healthy controls that a single nucleotide polymorphism (rs2756923) in exon 8 of the inhibitory gene KIR2DL2 was associated with type 1 diabetes. More recent data show that activating combinations of KIR/HLA genes are more frequent in children diagnosed in the first 5 years of life, suggesting that NK cell responses to viral infection are altered in this group.
Genetic risk assessment in the post GWAS era
Figure 2: The relative effects of selected T1D associated genes on susceptibility to T1D (adapted from Todd 2010) The genes detailed above are all associated with type 1 diabetes, and most of the newly identified susceptibility genes can be positioned on immune activation pathways, while some loci have yet to have the disease associated gene identified. Despite the overwhelming success of GWAS in identifying susceptibility genes for common diseases using hypothesis-free methodologies, the effects of the identified genes on improved genetic risk assessments have been minimal. This is because most of the newly identified loci make only a minimal contribution to risk with odds ratios (OR) in the range of 1.2–1.3 compared with 7 for the HLA locus (Figure 2). An OR of 1 indicates that risk is equal in healthy controls and individuals with disease. This has become known as the missing heritability and indicates, not surprisingly, that mechanisms other than common variants contribute to susceptibility to type 1 diabetes. Candidates for such effects are rare variants as well as epigenetic modifications that cannot be detected by GWAS. Nevertheless, the new loci identified by GWAS have informed ongoing functional studies and confirmed some interesting mechanistic loci such as IFIH1.
^ Bennett ST, Lucassen AM, Gough SC, et al. Susceptibility to human type 1 diabetes at IDDM2 is determined by tandem repeat variation at the insulin gene minisatellite locus. Nat Genet 1995;00:284–92
^ Ueda H, Howson JM, Esposito L, et al. Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature 2003;00:506–511
^ Bottini N, Musumeci L, Alonso A, et al. A functional variant of lymphoid tyrosine phosphatase is associated with type I diabetes. Nat Genet 2004:000:337–8
^ Lowe CE, Cooper JD, Brusko T, et al. Large-scale genetic fine mapping and genotype–phenotype associations implicate polymorphism in the IL2RA region in type 1 diabetes. Nat Genet 200;000:1074–82
^ Wellcome Trust Case Control Consortium. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 2007;000|:661–78
^ Liu S, Wang H, Jin Y, et al. IFIH1 polymorphisms are significantly associated with type 1 diabetes and IFIH1 gene expression in peripheral blood mononuclear cells. Hum Mol Genet 2009;00:358–65
^ Concannon P, Onengut-Gumuscu S, Todd JA, et al. Type 1 Diabetes Genetics Consortium. A human type 1 diabetes susceptibility locus maps to chromosome 21q22.3. Diabetes 2008;000:2858–61