P105 : ALLELE-, HAPLOTYPE- AND GENOTYPE-LEVEL ASSOCIATIONS WITH CROHN’S DISEASE IN JEWISH AND NON-JEWISH EUROPEAN AMERICANS

Abstract

We studied the association between HLA-A,-B, -C, DRB1, DQA1, DQB1, DPA1 and DPB1 variation and Crohn’s Disease (CD), an inflammatory bowel disease, with a case-control study of Jewish and non-Jewish European American CD patients. 1352 Crohn’s Disease patients and healthy unrelated controls of European ancestry, determined with a panel of ∼300 AIMs, were HLA genotyped via reverse format SSO strip, Luminex rSSO and Roche 454 NGS genotyping. Patients and controls were matched for either self-reported Jewish or non-Jewish ethnicity. 474 Jewish CD patients and 89 Jewish controls, and 619 non-Jewish CD patients and 170 non-Jewish controls were genotyped. Typing ambiguity was reduced using the ImmPort Ambiguity Reduction Tool, and allele-names were analyzed at the two-field level. Adherence to expected Hardy–Weinberg equilibrium proportions (HWEP) at each locus was tested with PyPop, and the Bridging Immunogenomic Data-Analysis Workflow Gaps (BIGDAWG) R package was used to perform case-control analyses. Extended haplotypes were inferred from significantly associated two-locus haplotypes. HLA-B deviated from expected HWEP ( p = 0.0001) in non-Jewish CD patients, due to an excess of B ∗ 08:01 homozygotes (13 obs, 5.37 exp, p = 0.0010). Non-Jewish patients and controls differed significantly at the DRB1 locus ( p = 0.00013); DRB1 ∗ 01:03 is predisposing (OR 5.19 p = 0.0021), while ∗ 11:04 is protective (OR 0.3 p ∗ 07:01 ∼ DQA1 ∗ 02:01 ∼ DQB1 ∗ 02:02 is predisposing (OR 3.13 p = 0.0014) and A ∗ 11:01 ∼ DRB1 ∗ 11:04 is protective (OR 0.09 p = 0.0004). In the non-Jewish cohort, the inferred A ∗ 02:01 ∼ B ∗ 51:01 ∼ DRB1 ∗ 01:03 ∼ DQA1 ∗ 01: 01 ∼ DQB1 ∗ 05:01 ∼ DPA1 ∗ 01:03 ∼ DPB1 ∗ 04:01 and A ∗ 01:01 ∼ B ∗ 57:01 ∼ DRB1 ∗ 07:01 ∼ DPA1 ∗ 02:01 extended haplotypes predispose, and the C ∗ 07:01 ∼ DRB1 ∗ 11:04 ∼ DQA1 ∗ 05:01 ∼ DQB1 ∗ 03: 01 ∼ DPA1 ∗ 01:03 ∼ DPB1 ∗ 03:01 extended haplotype is protective. The associations with DRB1 ∗ 01:03 and DRB1 ∗ 07:01 are confirmatory, but the protective role of DRB1 ∗ 11:04 is novel; this allele differs from the neutral DRB1 ∗ 11:01 allele by a single amino-acid change (V86G). These associations extend to class I ∼ class II haplotypes. HWEP analysis suggests that the B ∗ 08:01+B ∗ 08:01 genotype is predisposing in non-Jewish patients, while there is no difference in B ∗ 08:01 frequency between patients and controls.