Genome-wide association study meta-analysis identifies seven new rheumatoid arthritis risk loci

Abstract

To identify new genetic risk factors for rheumatoid arthritis, we conducted a genome-wide association study meta-analysis of 5,539 autoantibody-positive individuals with rheumatoid arthritis (cases) and 20,169 controls of European descent, followed by replication in an independent set of 6,768 rheumatoid arthritis cases and 8,806 controls. Of 34 SNPs selected for replication, 7 new rheumatoid arthritis risk alleles were identified at genome-wide significance (P < 5 x 10(-8)) in an analysis of all 41,282 samples. The associated SNPs are near genes of known immune function, including IL6ST, SPRED2, RBPJ, CCR6, IRF5 and PXK. We also refined associations at two established rheumatoid arthritis risk loci (IL2RA and CCL21) and confirmed the association at AFF3. These new associations bring the total number of confirmed rheumatoid arthritis risk loci to 31 among individuals of European ancestry. An additional 11 SNPs replicated at P < 0.05, many of which are validated autoimmune risk alleles, suggesting that most represent genuine rheumatoid arthritis risk alleles.

Figure 1. Association with RA risk across 4 loci

Regional association plots show strength of association (-Log(P-value)) versus chromosomal position (kilobases, kb) for all SNPs across 1 Megabase (Mb) regions centered on the newly validated SNPs (labeled). P_GWAS values are plotted with red/white diamonds for all SNPs, shaded by the degree of LD (r², see legend) with the validated SNP (larger red diamond). P_overall in combined analysis of GWAS and replication collections is plotted with the blue diamond. Local recombination rates estimated from HapMap CEU (centi-Morgans per Mb, blue line) are plotted against the secondary y-axis, showing recombination hotspots across the region. Labeled green arrows below the plots indicate genes and their orientations. (a) 2p14, SPRED2 locus. (b) 5q11, IL6ST-ANKRD55 locus. (c) 5q21, C5orf13 locus. (d) 10p15, IL2RA locus.

Figure 1. Association with RA risk across 4 loci

Regional association plots show strength of association (-Log(P-value)) versus chromosomal position (kilobases, kb) for all SNPs across 1 Megabase (Mb) regions centered on the newly validated SNPs (labeled). P_GWAS values are plotted with red/white diamonds for all SNPs, shaded by the degree of LD (r², see legend) with the validated SNP (larger red diamond). P_overall in combined analysis of GWAS and replication collections is plotted with the blue diamond. Local recombination rates estimated from HapMap CEU (centi-Morgans per Mb, blue line) are plotted against the secondary y-axis, showing recombination hotspots across the region. Labeled green arrows below the plots indicate genes and their orientations. (a) 2p14, SPRED2 locus. (b) 5q11, IL6ST-ANKRD55 locus. (c) 5q21, C5orf13 locus. (d) 10p15, IL2RA locus.

Figure 1. Association with RA risk across 4 loci

Regional association plots show strength of association (-Log(P-value)) versus chromosomal position (kilobases, kb) for all SNPs across 1 Megabase (Mb) regions centered on the newly validated SNPs (labeled). P_GWAS values are plotted with red/white diamonds for all SNPs, shaded by the degree of LD (r², see legend) with the validated SNP (larger red diamond). P_overall in combined analysis of GWAS and replication collections is plotted with the blue diamond. Local recombination rates estimated from HapMap CEU (centi-Morgans per Mb, blue line) are plotted against the secondary y-axis, showing recombination hotspots across the region. Labeled green arrows below the plots indicate genes and their orientations. (a) 2p14, SPRED2 locus. (b) 5q11, IL6ST-ANKRD55 locus. (c) 5q21, C5orf13 locus. (d) 10p15, IL2RA locus.

Figure 1. Association with RA risk across 4 loci

Regional association plots show strength of association (-Log(P-value)) versus chromosomal position (kilobases, kb) for all SNPs across 1 Megabase (Mb) regions centered on the newly validated SNPs (labeled). P_GWAS values are plotted with red/white diamonds for all SNPs, shaded by the degree of LD (r², see legend) with the validated SNP (larger red diamond). P_overall in combined analysis of GWAS and replication collections is plotted with the blue diamond. Local recombination rates estimated from HapMap CEU (centi-Morgans per Mb, blue line) are plotted against the secondary y-axis, showing recombination hotspots across the region. Labeled green arrows below the plots indicate genes and their orientations. (a) 2p14, SPRED2 locus. (b) 5q11, IL6ST-ANKRD55 locus. (c) 5q21, C5orf13 locus. (d) 10p15, IL2RA locus.

Figure 2. Previously validated autoimmune SNPs tested in our replication study

Eighteen SNPs tested in our replication samples were in LD (defined as r² > 0.3) with a validated autoimmune risk allele. Of these, 5 were validated as RA risk alleles in our study (P_overall < 5×10⁻⁸, inner most circle), 6 were suggestive associations (P_replication <0.05 but P_overall > 5×10⁻⁸), and 6 demonstrated no evidence of association in our replication samples (P_replication ≥0.05). For the 12 SNPs with suggestive or no evidence of association, each SNP is plotted by the strength of association with RA risk in the replication samples; those closer to the inner circle have more significant P_replication. All of the RA risk alleles confer risk in the same direction as the validated autoimmune risk alleles (when the same allele or a near perfect proxy was tested). We include the following as “autoimmune” diseases in our study, listed on the outside of the circle, although these reflect diseases along the autoimmuneinflammatory spectrum: systemic lupus erythematosus (SLE), celiac disease, Crohn’s disease, multiple sclerosis (MS), psoriasis, and type 1 diabetes (T1D); other autoimmune diseases are not included (e.g., autoimmune thyroiditis). The haplotype tagged by the IL2RA SNP, rs706778, is associated with T1D and MS^,; the SH2B3 SNP is associated with both T1D and Celiac disease^,. The CCR6 SNP rs3093023 is in partial LD with a SNP associated with Crohn’s disease (rs2301436, r² = 0.48). The AFF3 SNP has an equivocal association with T1D (where the associated SNP is rs9653442). The IL2-IL21 SNP tested in our study, rs13119723, is in LD (r²=0.67) with a SNP previously implicated in both Celiac disease and RA (rs6822844)^,, but only in partial LD (r²=0.09) with a T1D SNP (rs4505848). The CD19-NFATC2IP SNP tested in our study, rs8045689, was selected because of GRAIL; it is in partial LD (r²=0.38) with a SNP associated with T1D (rs4788084). The TNIP1 SNP in our study, rs6889239, is in strong LD with an SLE SNP (rs7708392, r²=0.91), but not in LD with another TNIP1 SNP associated with psoriasis (rs17728338, r²<0.01). The ZEB1 SNP (rs2793108) was from the May 2009 release of T1D base, although this SNP did not appear in a subsequent publication. The PXK SLE SNP was tested in this study but not shown, as it is in weak LD with the RA risk SNP (r² = 0.15 between rs6445975 and rs13315591); the RA SNP was selected because of P_GWAS<10⁻⁶, not because of its association with another autoimmune disease. Note that there are SNPs associated with RA and other autoimmune diseases not shown; we only include those SNPs tested as part of the current study.