Background/Purpose: Epigenetics participates in the pathogenesis of rheumatoid arthritis (RA). Epigenetic marks, gene expression and DNA polymorphisms have been investigated but the analyses are limited to few marks and simple combination methods. In the present study, we used a novel algorithm (EpiSeq) to integrate epigenomes for RA FLS in an unbiased fashion combining whole genome histone modifications, open chromatin, RNA expression and DNA methylation. By focusing on the chromatin states of regulatory elements and using a new computational platform, we identified unexpected pathways that contribute to the pathogenesis of RA.

Methods: We applied multiple omics technologies on 11 RA and 11 osteoarthritis [OA] FLS: ChIPseq for histone modifications, ATACseq for open chromatin, RNAseq for transcriptomes and whole genome bisulfite sequencing (WGBS) for DNA methylation. The complex multidimensional relationships were addressed with our novel unbiased method, EpiSig, which is a flexible framework for integrative analysis of any type of sequencing data and identifies epigenomically co-modified regions that share similar patterns. Common epigenetic modification patterns were identified on a global genome scale and the genome was segmented into regulatory/functional elements. The epigenetic state of each element was defined and differentially modified epigenetic regions (DMER) between RA and OA were identified. Pathway evaluation used Ingenuity Pathway Analysis. Bostatistical analyses used Benjamini-Hochberg corrections.

Results: 218 genome-wide datasets were generated across FLS samples, including 152 histone modification datasets (H3K27ac, H3K4me1, H3K4me3, H3K36me3, H3K27me3, H3K9me3), 22 DNA methylation datasets, 22 open chromatin datasets and 22 transcriptome datasets. Eighteen epigenomic states were defined with distinct FLS chromatin signatures, including 4 promoter states, 6 enhancer states, 2 transcribed states, 2 states with zinc finger protein genes and 4 repressed states. The genome was segmented into 5 kb regions, and 125 epigenetic clusters sith similar epigenetic patterns were determined across FLS. We identified regions that were differentially marked when comparing OA and RA. 13 clusters with significant DMER enrichment in RA were identified, which were grouped into biological pathways based on genes associated with these promoters and enhancers. Among the pathways that were significantly different in RA FLS, Phospholipase C Signaling, p53 Signaling, Integrin Signaling, and Protein Kinase A signaling were particularly notable. Other pathways were unexpected, such as Huntington’s Disease Signaling (HDS). To biologically validate the HDS pathway, we showed that one key member of HDS, namely HIP1, is expressed in FLS. HIP1 deficiency induced by siRNA knockdown decreased cultured RA FLS invasion into an artificial matrix by 56% (p<0.001).

Conclusion: We developed the first high-resolution global epigenomic landscape for RA and using a novel method to prioritize RA-specific biological pathways. Biologic validation of one unanticipated target in HDS pathway confirms that this unbiased method can identify novel therapeutic targets.

« Back to 2017 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/complete-epigenetic-landscape-of-rheumatoid-arthritis-fibroblast-like-synoviocytes-reveals-unanticipated-critical-pathogenic-pathways/