Background/Purpose: Genome-wide association studies (GWAS) have to date identified over 100 genomic loci at which single nucleotide polymorphisms (SNPs) confer an increased risk of developing rheumatoid arthritis (RA). These loci are enriched for lymphocyte-specific enhancer elements, consistent with a regulatory function of many causal variant(s). Epigenetic modifications have also been strongly implicated in RA pathogenesis, potentially impacting cell phenotype through altered gene expression. Here, we investigate the role of DNA methylation as a mediator of RA genetic risk.

Methods: CD4+ T- and B-lymphocytes were isolated by positive selection from fresh peripheral blood of drug-naïve patients attending an early arthritis clinic. Paired CD4+ T-lymphocyte-specific DNA and RNA were extracted from 43 RA and 60 disease control patients, respectively, and equivalent material from B-lymphocytes of 46 RA and 73 controls. Comparator groups were matched for age, sex, and acute phase response. Genotyping was performed using the Illumina Infinium Human CoreExome-24 array, and DNA methylation at >850,000 CpG sites quantified with the Illumina MethylationEPIC array. Gene expression profiling was measured using the Illumina Human HT-12 v4 BeadChip. Having first mapped genome-wide methylation quantitative trait loci (mQTLs) in cis (<1Mb), we focussed our analysis on known RA risk loci, and integrated paired normalised gene expression measurements for transcripts within 500kb of index CpGs. Finally, we performed an analysis of mQTLs acting in trans, seeking interactions between genotype and disease diagnosis to highlight RA-specific effects.

Results: CD4+ T lymphocyte cis-mQTLs colocalised with 30 independent (r2 < 0.8) RA-associated SNPs, whilst in B lymphocytes such mQTL effects were present at 31 RA SNPs. A high proportion of these variants (>80%) appeared to function as cis-mQTLs in both cell types. CpG sites subject to cis effects at risk loci were depleted in regions associated with cell type-specific repressed chromatin marks, with enrichment at enhancer regions and those flanking transcription start sites suggesting active roles in transcriptional regulation. Linear regression identified putative regulatory effects of these CpG sites on gene expression, and causal inference testing highlighted genes for which risk SNPs most likely modulate gene expression via CpG methylation. Such effects, robust to false discovery rate, were particularly prevalent in CD4+ T lymphocytes, implicating ANKRD55, ORMDL3, and FCRL3 amongst others as causal genes in this cell type. Similar effects were less robust in B-lymphocytes, albeit potentially implicating genes including CCR6. Our analysis of mQTLs acting in trans identified inter-chromosomal SNP-CpG associations, also revealing instances of differential effect sizes in RA patients and controls.

Conclusion: Here we demonstrate the utility of DNA methylation profiling as a tool for the prioritization of candidate genes following GWAS studies in RA, and highlight an important mechanism through which genetic variants may contribute to altered lymphocyte phenotype. The functional roles of highlighted genes in CD4+ T cells during RA pathogenesis await clarification.

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/lymphocyte-dna-methylation-as-a-mediator-of-genetic-risk-in-rheumatoid-arthritis/