Background/Purpose

Systemic sclerosis (SSc) is a complex autoimmune disease and several genetic loci increasing SSc susceptibility have been identified with small to modest effect sizes. These loci even when taken together are not able to fully explain the heritability of SSc. We believe that genetic epistasis along with rare variants and gene-environment interaction may explain the missing heritability. In this study, we test two seemingly distinct loci, rs2004640 on IRF5 and rs2736340 on BLK both of which have previously been associated with SSc, for genetic epistasis.

Methods

In this study, we combined genetic data from rs2004640 variant on IRF5 and rs2736340 variant on BLK from 1024 patients with SSc and 694 unrelated healthy controls from University of Texas and a Spanish case–control series of 395 SSc patients and 443 healthy controls. Odds ratios (OR) and biologic interactions as departures from additivity or multiplicity were analyzed by logistic regression. To quantify the amount of interaction in terms of departure from additivity of effects, the relative excess risk due to interaction (RERI), proportion attributable to interaction (AP), and the synergy index (S) were calculated. Interaction between the two SNPs was evaluated using the cross-product of the risk factors in a logistic-regression model as interaction criteria (multiplicative interaction). We adjusted for gender and cohort in the analyses. Recoding of IRF5, BLK genotypes from protective effects to risk effects were done to produce the meaningful measures for departures of additivity. Statistical analyses were performed using SAS 9.3. Gene expression array of PBMCs from PAXgene tubes were analyzed with BRB-ArrayTools.

Results

Both IRF5:rs2004640 and BLK:rs2736340 variants show independent association with SSc. The OR of single effects was 1.40 for IRF5:rs2004640 and 1.44 for BLK:rs2736340 (Table 1).

We observed a significant multiplicative interaction between IRF5 GT/TT & BLK CT/TT genotypes as compared to IRF5 GG & BLK CC genotypes (p=0.02). The OR of joint effects for IRF5 GT/TT & BLK CT/TT genotypes was higher than the wildtype genotype (P=0.0003, OR=2.26; 95%CI 1.7-3.1). RERI<0, AP<0, and S<1 mean negative interaction and less than additivity. P value of S is significant, but 95% CI includes 1.

PBMC RNA gene expression arrays predicted more T & B cell pathways in SSc as compared to controls for IRF5 GT/TT & BLK CT/TT genotypes and PPAR-γ and WNT signaling pathways in SSc as compared to controls for IRF5 GG & BLK CC genotypes.

Conclusion

In an effort to explore genetic epistasis we demonstrate genetic interaction using multiple methodologies between two well-replicated and distinct loci on IRF5 and BLK. We also observed differences in the gene expression pathways based on the IRF5 and BLK genotypes. Additional studies are needed to test them in other autoimmune diseases and discern their role at a molecular level.