Background/Purpose:

Genome wide association studies (GWAS) have identified numerous rheumatoid arthritis (RA) risk loci in patients of European and Asian ancestry but the causal variants have rarely been identified. These studies have also identified susceptibility loci that are common to various autoimmune diseases. The need to identify casual variants prompted the design of the Immunochip, a custom Illumina Infinium high density genotyping array, with ~196 000 single nucleotide polymorphisms (SNPs) from 186 loci previously associated with 12 autoimmune diseases. Little is known about the genetics of RA in black South Africans. We aimed to test associations with known genetic loci and to identify novel risk loci in black South Africans with RA.

Methods:

Consenting black RA patients fulfilling the 1987 ACR criteria for RA, >18 years at disease onset, and seen at a single centre in Johannesburg, South Africa were studied and compared to ethnically and geographically matched controls. Genotyping of 716 samples was performed using the Immunochip. Only samples with ≥94% call rate and individual SNPs with ≥95% call rate, were polymorphic, in Hardy-Weinberg equilibrium (p≥5×10^-7) and with a minor allele frequency of ≥0.02 were analyzed. The cohort was further pruned for relatedness and ancestral outliers. After quality control, 117 353 SNPs were tested for association in 263 cases and 365 controls. The statistical analysis was performed using PLINK vers1.07. A p value = <5x10^-8 was considered significant based on previous genome wide association studies.

Results:

The strongest associations were found in the MHC region with 64 SNPs reaching statistical significance. The most significant associations were in the intergenic region of the HLA DRB1- HLA DQA1 alleles (rs3104413, OR=3.88, p=5.49×10^-21; rs3129769, OR=3.91, p=4.60×10^-21; rs6931277, OR=3.97, p=1.03×10^-21). There were 2 non HLA SNPs that reached genome wide significance, rs9283487 (OR=3.25, p=7.96 x10^-18) in the PRKRA gene, and rs35198051 (OR=0.34, p=2.94×10^-8) in the intergenic region of C11orf76 and LOC100133306. In addition there were suggestive associations of 2 SNPs on chromosome 1, rs12739262 (OR=0.36, p=1.36×10^-7) in the PPP1R12B gene and rs12738883 (OR=0.36, p=1.69×10^-7) in the intergenic region of PLD5 and LOC400723, both SNPs reached statistical significance in the rheumatoid factor (RF) positive subgroup (n=240/254) (rs12739262, OR=0.33, p=9.10×10^-8 and rs12738883, OR= 0.33, p=9.10×10^-8).

Conclusion:

In keeping with previous studies the HLA class II region confers the strongest genetic risk for RA in black South Africans. We also found 2 novel SNPs in the overall cohort and a further 2 SNPs in the RF positive subgroup. The association of the PRKRA gene with RA is of interest because of its recent association with Sjogren’s Syndrome. Further studies in larger cohorts of African patients are required to validate these findings and to better understand the functional role of these novel loci in the pathogenesis of RA in black South Africans.