Abstract
AbstractBackgroundType 1 diabetes is an autoimmune disease, which leads to insulin dependence. We investigated genetic evidence to support the repurposing of seven drugs, already licensed or in clinical phases of development, for prevention of type 1 diabetes.MethodsWe obtained genome-wide association study (GWAS) summary statistics for the risk of type 1 diabetes, whole-blood gene expression, and serum protein levels, and investigated genetic polymorphisms near seven potential drug target genes. We used colocalization to examine whether the same genetic variants that are associated with type 1 diabetes risk were also associated with the relevant drug target genetic proxies, and Mendelian randomization to evaluate the direction and magnitude of the associations. Furthermore, we performed Mendelian randomization analysis restricted to functional variants within the drug target genes.FindingsColocalization revealed that the blood interleukin (IL)-2 receptor subunit alpha (IL2RA) and IL-6 receptor (IL6R) gene expression levels within the corresponding genes shared the same causal variant with type 1 diabetes liability (posterior probabilities 100% and 96.3%, respectively). Odds ratios (OR) of type 1 diabetes per 1-SD increase in the genetically proxied gene expression ofIL2RAandIL6Rwere 0.22 (95% confidence interval [CI] 0.17-0.27) and 1.98 (95% CI 1.48-2.65), respectively. Using missense variants, genetically proxied tyrosine kinase 2 (TYK2) expression levels were associated with type 1 diabetes risk (OR 0.61, 95% CI [0.54-0.70]).InterpretationOur findings support the targeting of IL-2, IL-6R and TYK2 signaling in prevention of type 1 diabetes. Further studiesshouldassess the optimal window to intervene to prevent type 1 diabetes.FundingKyllikki ja Uolevi Lehikoisen säätiö, Foundation for Pediatric Research, Finnish Cultural Foundation, JDRF International; The University of Oulu & The Research Council of Finland Profi 326291; European Union’s Horizon 2020 research and innovation program under grant agreement no. 848158 (EarlyCause).Research in contextEvidence before this studyWe searched PubMed for GWAS, Mendelian randomization, colocalization, and other studies for genetic evidence of efficacy of targeting IL2RA, IL2RB, IL6R, IL6ST, IL23A, TYK2, JAK2, and JAK3 signaling in prevention of type 1 diabetes. We searched studies with combinations of “IL2RA”, “IL2RB”, “IL6R”, “IL6ST”, “IL23A”, “JAK2”, “JAK3”, “TYK2”, “colocalization”, "mendelian randomization", or “GWAS” and "Diabetes Mellitus, Type 1"[Mesh].A recent large GWAS by Robertson et al. observed five conditionally independent genome-wide significant single-nucleotide polymorphisms (SNPs) associated with the risk of type 1 diabetes in intergenic and intronic areas nearIL2RA. Another large recent GWAS by Chiou et al. found six independent SNPs associated with the risk of type 1 diabetes nearIL2RA. A few smaller candidate gene studies also reported associations between the risk of type 1 diabetes and eight other loci near IL2RA. Epigenetic studies reported that a few type 1 diabetes risk variants nearIL2RAare associated with changes in methylation of DNA near promoters ofIL2RAin white blood cells and B-cells.Both above mentioned GWAS found associations between two missense mutations inTYK2gene (rs12720356(A>C), rs34536443(G>C)) and the risk of type 1 diabetes. The same variants protected against type 1 diabetes and some other autoimmune diseases in a Finnish FINNGEN biobank study. Smaller candidate gene studies found that the rs2304256 A/A genotype was protective against type 1 diabetes in a Southern Brazilian and in a European population. In a study of Japanese population, similar association was not seen, possibly due to differences in allele frequencies and incidence of type 1 diabetes between populations. In the same study, TYK2 promoter haplotype with genetic polymorphisms in promoter region and exon 1 decreased the promoter activity and increased the risk for type 1 diabetes. The minor 358Ala allele of the IL6R SNP rs2228145 (A>C) was found to be associated with a reduced risk of developing T1D.In addition to loci near or at IL2RA and TYK2 loci, the GWAS by Robertson et al. reported an association between rs2229238 nearIL6Rand the risk of type 1 diabetes. Furthermore, a candidate gene study found that the GG haplotypes of variants rs11171806 and rs2066808 (in strong LD) nearIL23Awas protective against T1D.We found no studies in which colocalization between the loci near the reviewed genes and type 1 diabetes was reported nor estimates from Mendelian randomization on the possible causal relationship. However, Robertson et al. studied the genetic evidence for therapeutic potential of their GWAS hits using a priority index algorithm, which combined information from GWAS of type 1 diabetes, expression quantitative loci in immune cells, evidence from chromatin conformation in immune cells, as well as protein-protein interactions and genomic annotations. Among the reported GWAS hits, priority index ranked IL2RA as the top target, followed by TYK2 (the 3rd target), JAK2 (11th), IL23 (25th), JAK3 (35th), and IL6R (50th).Added value of this studyWe found that the risk of type 1 diabetes and whole-blood gene expression ofIL2RAandIL6Rcolocalized to rs61839660 nearIL2RAand rs10908839 nearIL6R, respectively. Using these loci as instruments ofIL2RAandIL6Rexpression, Mendelian randomization indicated that whole bloodIL2RAexpression is associated with decreased risk of type 1 diabetes (OR 0.70 per 1 SD increase) whereas IL6R expression is associated with increased risk (OR 1.079 per 1 SD increase). Furthermore, when a missense variant rs2304256 inTYK2was used as an instrument forTYK2expression, Mendelian randomization indicated that an SD increase inTYK2expression was predicted to decrease the risk of type 1 diabetes (OR 0.61).Implications of all the available evidenceWe found genetic evidence supporting the efficacy of targeting IL-2, IL-6 and TYK2 signalling in prevention of type 1 diabetes. Further research is needed to address when would be the most optimal time to intervene in the pathogenesis of type 1 diabetes.
Publisher
Cold Spring Harbor Laboratory