461 Characterization of Mutations Causing ATP8B1 Deficiency and Pharmacological Restoration of ATP8B1 Abundance and Cell Surface Expression

Abstract

attributable risk of PBC, and thus performed a multiplicative interaction study of the two variants. Methods: The rs231725 SNP of CTLA4 and the rs1800629 SNP of TNF were previously genotyped in 416 PBC patients and 404 controls by commercially available TaqMan assays. Interaction between the 2 SNPs was assessed by logistic regression under 2 models, considering either a (1) dominant (i.e. AA or AG genotype) or (2) recessive (i.e. AA genotype) affect of the CTLA4 allele with a dominant affect of the TNF allele (i.e. AA or AG genotype). Results: The TNF-CTLA4 interaction was significant under both models; CTLA4 dominant effect model (overall fit p=0.02), interaction odds ratio 2.33 (95% CI 1.28-4.26), p=0.006; CTLA4 recessive effect model (overall fit p=0.002), interaction odds ratio 5.00 (95% CI 1.75-14.25), p=0.002. Conclusions: There is an interaction between these two alleles of the TNF and CTLA4 genes that results in an increased risk of PBC. This affect is significant when just one of each of the offending alleles is carried (i.e. dominant CTLA4 effect model), a finding that is of wider scope than the previously described association of the individual CTLA4 SNP, which required the presence of two risk alleles for association. Moreover, the interaction is even stronger under the recessive CTLA4 effect model, further demonstrating that carriage of the TNF variant allele amplifies the risk attributed to CTLA4.