Analysis of polymorphisms in 16 genes in type I diabetes that have been associated with other immune-mediated diseases
Smyth DJ., Howson JMM., Payne F., Maier LM., Bailey R., Holland K., Lowe CE., Cooper JD., Hulme JS., Vella A., Dahlman I., Lam AC., Nutland S., Walker NM., Twells RCJ., Todd JA.
Background: The identification of the HLA class II, insulin (INS), CTLA-4 and PTPN22 genes as determinants of type 1 diabetes (T1D) susceptibility indicates that fine tuning of the immune system is centrally involved in disease development. Some genes have been shown to affect several immune-mediated diseases. Therefore, we tested the hypothesis that alleles of susceptibility genes previously associated with other immune-mediated diseases might perturb immune homeostasis, and hence also associate with predisposition to T1D. Methods: We resequenced and genotyped tag single nucleotide polymorphisms (SNPs) from two genes, CRP and FCER1B, and genotyped 27 disease-associated polymorphisms from thirteen gene regions, namely FCRL3, CFH, SLC9A3R1, PAD14, RUNX1, SPINK5, IL1RN, IL1RA, CARD15, IBD5-locus (including SLC22A4), LAG3, ADAM33 and NFKB1. These genes have been associated previously with susceptibility to a range of immune-mediated diseases including rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), Graves’ disease (GD), psoriasis, psoriatic arthritis (PA), atopy, asthma, Crohn disease and multiple sclerosis (MS). Our T1D collections are divided into three sample subsets, consisting of set 1 families (up to 754 families), set 2 families (up to 743 families), and a case-control collection (ranging from 1,500 to 4,400 cases and 1,500 to 4,600 controls). Each SNP was genotyped in one or more of these subsets. Our study typically had approximately 80% statistical power for a minor allele frequency (MAF) >5% and odds ratios (OR) of 1.5 with the type 1 error rate, α = 0.05. Results: We found no evidence of association with T1D at most of the loci studied 0.02 <P < 1.0. Only a SNP in ADAM33, rs2787094, was any evidence of association obtained, P = 0.0004 in set 1 families (relative risk (RR) = 0.78), but further support was not observed in the 4,326 cases and 4,610 controls, P = 0.57 (OR = 1.02). Conclusion: Polymorphisms in a variety of genes previously associated with immune-mediated disease susceptibility and/or having effects on gene function and the immune system, are unlikely to be affecting T1D susceptibility in a major way, even though some of the genes tested encode proteins of immune pathways that are believed to be central to the development of T1D. We cannot, however, rule out effect sizes smaller than OR 1.5. © 2006 Smyth et al; licensee BioMed Central Ltd.