Background/Purpose: Substantial epidemiologic evidence demonstrates that SLE disproportionately affects minority patients in terms of incidence, prevalence, and disease severity. European ancestry has been associated with a lower risk of developing renal disease in SLE. We developed an approach to identify novel SLE risk variants that may influence disease severity by searching for single nucleotide polymorphisms (SNPs) with high ethnic variability worldwide and testing them in our multiethnic SLE cohort. 

Methods: The Human Genome Diversity Project (HGDP) characterized the allele frequency of 650,000 SNPs in 938 people from 53 populations worldwide. Our multiethnic cohort of SLE patients was genotyped on the Illumina Immunochip (I-chip), covering over 160,000 SNPs across 185 autoimmune genes plus select non-autoimmune genes and ancestry informative markers. There were 32,907 SNPs common to the HGDP and I-chip. We selected the top 2% of those SNPs that met both an absolute mean frequency difference and a t-test criteria between Europeans and Non-Europeans. We tested these SNPs in our multiethnic cohort of 1427 SLE patients for severe disease outcomes (renal disease by ACR renal criterion, severe renal disease on biopsy or end-stage renal disease, and production of dsDNA antibodies) stratified by ethnicity: Caucasian, African American, Hispanic, and Asian. Logistic regression was performed, adjusting for disease duration and gender. Ethnic strata were refined via STRUCTURE analysis of 878 I-chip SNPs, excluding subjects with substantial ancestry outside of their self-identified ethnicity.

Results: This approach identified 13 SNPs with ethnicity-disease outcome associations (unadjusted p value < 0.001). Two SNPs, rs2099365 and rs2163882, which are in high linkage disequilibrium (LD) (r2 = 0.98) in Hispanics, also had significant false discovery rate p values (0.014) for association of renal disease in Hispanics (odds ratio 2.90). Both SNPs are 20kb from the 3’ end of the endomucin (EMCN) gene, a glycoprotein involved in cell adhesion. These SNPs are predicted to disrupt 3-4 regulatory motifs. The EMCN gene is a non-autoimmune gene on the I-chip, not previously associated with SLE, but recently reported to be associated with susceptibility to RA. The association of the 2 SNPs with renal disease in Hispanics is consistent with an estimated 5.7 odds of renal disease per 100% Amerindian ancestry (p = 3.5e-06) compared to 0.23 odds per 100% European ancestry (p=1.1e-22) based on I-chip data. To validate the EMCN association results, we examined available EMCN gene SNPs in a large cohort of Hispanic patients, confirming an association with renal disease (5 SNPs with p values < 0.001, most significantly 4.5e-05).

Conclusion: Here we describe an approach to identify novel SLE severity risk variants. Testing these candidate alleles in our multiethnic SLE cohort identified 2 SNPs in high LD that were significantly associated with renal disease among Hispanic SLE patients. These SNPs are near the EMCN gene, a gene never before associated with SLE. Additional investigation is warranted to further validate these results and better characterize the association of this gene with renal disease in SLE.

---

« Back to 2014 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/identifying-novel-lupus-severity-risk-variants-through-identification-of-alleles-with-high-ethnic-variability-worldwide/