Interferon regulatory factor 5 (IRF5) gene variants are associated with multiple sclerosis in three distinct populations

Abstract

Background: IRF5 is a transcription factor involved both in the type I interferon and the toll-like receptor signalling pathways. Previously, IRF5 has been found to be associated with systemic lupus erythematosus, rheumatoid arthritis and inflammatory bowel diseases. Here we investigated whether polymorphisms in the IRF5 gene would be associated with yet another disease with features of autoimmunity, multiple sclerosis (MS).

Methods: We genotyped nine single nucleotide polymorphisms and one insertion-deletion polymorphism in the IRF5 gene in a collection of 2337 patients with MS and 2813 controls from three populations: two case-control cohorts from Spain and Sweden, and a set of MS trio families from Finland.

Results: Two single nucleotide polymorphism (SNPs) (rs4728142, rs3807306), and a 5 bp insertion-deletion polymorphism located in the promoter and first intron of the IRF5 gene, showed association signals with values of p<0.001 when the data from all cohorts were combined. The predisposing alleles were present on the same common haplotype in all populations. Using electrophoretic mobility shift assays we observed allele specific differences in protein binding for the SNP rs4728142 and the 5 bp indel, and by a proximity ligation assay we demonstrated increased binding of the transcription factor SP1 to the risk allele of the 5 bp indel.

Conclusion: These findings add IRF5 to the short list of genes shown to be associated with MS in more than one population. Our study adds to the evidence that there might be genes or pathways that are common in multiple autoimmune diseases, and that the type I interferon system is likely to be involved in the development of these diseases.

Figure 1. Schematic illustration of the IRF5 gene with the positions of the analysed polymorphisms. The exons and the 3′-UTR are shown as boxes with the exons labelled 1-9 and the untranslated alternative exons in the 5′-end of IRF5 labelled as 1a, 1b and 1c. The translation initiation site is indicated by an arrow above the gene. The single nucleotide polymorphism (SNP) rs2004640 is located at the splice junction of alternative exon 1b, where it alters a consensus splice donor site that allows expression of IRF5 mRNA containing exon 1b. The SNP rs10954213 located in the 3′untranslated region (UTR) of the gene correlates with an altered length of the IRF5 3′UTR and thereby affects the stability of the IRF5 transcript. The SNPs rs2280714, located ∼6 kb downstream of the IRF5 gene, has been reported to be correlated with variation in IRF5 mRNA expression levels. A 5 bp insertion-deletion polymorphism (CGGGG indel) in the promoter region of IRF5, identified by sequencing the IRF5 gene in Swedish individuals, was included in our study because it is predicted to alter a binding site for the transcription factor SP1, which could affect the expression of IRF5. The SNP rs 12539741 was included because this SNP, and several other SNPs located 3′ of IRF5 that are in almost full linkage disequilibrium (LD) with it, show very strong association signals with systemic lupus erythematosus.

Figure 2. Linkage disequilibrium (LD) structure of the IRF5 gene. Pairwise r² values are shown for controls or founders in the Spanish, Swedish and Finnish cohorts, respectively. The three most strongly associated markers rs4728142, the CGGGG indel and rs3807306 are in relatively high LD with each other with pair-wise r² values >0.6.

Figure 3. Electrophoretic mobility shift assay images for the three most strongly associated variants of IRF5. Reactions loaded in lanes 1–4 (from left to right) for each allele contain: (1) labelled probe only; (2) labelled probe and nuclear extract; (3) labelled probe, nuclear extract and the unlabeled probe, which is added as a competitor in 100-fold excess; (4) labelled probe, nuclear extract, and a 100-fold excess of unlabelled probe for the other allele of the polymorphism added as a cross-competitor.

Figure 4. Result from analysis of the SP1 protein–promoter interaction using the proximity ligation assay. The probes included in the assay shown from left to right are: SP1 mut—negative control probe where four nucleotides within the SP1 consensus binding site have been altered; CGGGG indel 3× and 4×—probes with three or four copies of CGGGG; SP1—positive control probe with SP1 consensus binding site; No lysate—control reaction where no nuclear lysate was added to the reaction. The samples were analysed in duplicate and the result presented are mean values of signal/noise, where the signal from the proximity probe pairs in the sample is divided by the signal in the control reaction without cell lysate. The data shown are from a representative experiment repeated three times with similar results. In each individual experiment the signal obtained by the 4× CGGGG indel probe was 2.6±0.3 fold higher than the signal from the 3× CGGGG indel probe.