Background/Purpose: In many rheumatologic and autoimmune diseases African Americans (AA) have a higher prevalence and greater disease severity than other ethnicities. We posit that population-specific natural selection have influenced allele frequencies at some loci and thereby contribute to the higher prevalence and severity of systemic lupus erythematosus (SLE) in AA. The Gullah population of coastal South Carolina and Georgia has lower European admixture and higher ancestral homogeneity from the Sierra Leone (SL) area in Far-West Africa than other AA populations. The shorter genetic distance between the Gullah and SL suggests that population genetic signals, such as regions under recent selection, may be more easily detected in the Gullah than in other AA populations. The goal of this study was to leverage the evidence for recent natural selection to increase the statistical power to detect novel SLE-predisposing loci in the Gullah.

Methods: We computed a genome-wide association study (GWAS) on 151 Gullah SLE cases and 119 Gullah controls adjusting for admixture estimates. We also computed the likelihood ratio test for allelic differences between 277 Gullah population controls and 400 SL population controls adjusting for admixture. The intersection of SNPs from the case-control and test of selection that met standard quality control were 193,008 SNPs. We combined the evidence of association with SLE in the Gullah from the case-control analysis with evidence of selection by summing the chi-square statistic from a SNPs allele frequency test with its case-control association chi-square to create an overall chi-square statistic with 2 df (Ayodo et al, 2007). We used HaploReg for functional annotation of variation.

Results: Several genomic regions with multiple SNPs with suggestive evidence for association with SLE (P<5.0E-04) were found, including an intergenic region at 1p31.1, TLE4, and PCNXL4. When combined with the allele frequency differences, XCR1-CCR1 showed multiple SNPs with P<0.0005. However, several other regions showed multiple SNPs with increased significance (P<0.005) when the association and selection evidence was combined, including the HLA, PKHD1, GAS2L3 and FHOD3. Interestingly, the strength of significance increased for several SNPs in the PCNXL4 region. This was also the region with the highest regulatory potential, with multiple marks of active promoters and transcription start sites in different immune-related cell lines.

Conclusion: We report novel loci with suggestive evidence for association with SLE in a genetically homogeneous AA population, by combining evidence of natural selection with association to detect SLE risk variants in Gullah AA. These loci merit focused association replication efforts in SLE and this approach is generalizable to several other rheumatic diseases with ethnic differences in prevalence.  Clearly the link with evidence of selection will only increase the power for a subset of SLE-predisposing to those that may have been historically adaptive.  Such distinctions help not only identify new risk loci but may help provide insight into the evolution of rheumatic and autoimmune diseases in humans.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/integrating-evidence-for-genetic-association-and-natural-selection-helps-detect-new-systemic-lupus-erythematosus-risk-loci-in-african-americans/