Background/Purpose

The molecular mechanisms underlying the topographic differences in the susceptibility of synovial joints to develop rheumatoid arthritis (RA) are unknown. Positional embryonic expression of Hox genes along proximal-distal and anterior-posterior body axes is critical for proper limb development. Adult skin fibroblasts retain the positional embryonic Hox code and exhibit major anatomic differences in their transcriptome, defining their unique positional identities. Synovial fibroblasts (SF) in the joints of RA patients drive joint destruction and inflammation locally. We hypothesized that SF from different joints show a joint specific, positional gene expression pattern, which can predispose joints to develop certain types of arthritides, like RA or osteoarthritis (OA).

Methods

SF were derived from knees, shoulders and metacarpophalangeal joints (MCPs) of RA and OA patients (n=9 each) undergoing joint replacement surgery. SF were obtained also from front paws, ankles and knees of wildtype (wt) and TNF transgenic (TNFtg) mice (n=7 each). Total RNA was extracted and RNA sequencing was performed with the Illumina HiSeq 2000 sequencing system followed by hierarchical clustering. Functional annotation clustering of mRNAs was done using Database for Annotation, Visualization, and Integrated Discovery (DAVID). Positionally expressed RNAs were validated by qPCR.

Results

Unsupervised hierarchical cluster analysis showed clustering of SF according to anatomic joint localization rather than disease. The positional embryonic HOX code was retained in SF, clearly differentiating between different joints. Among the Hox cluster residing long noncoding RNAs, HOTTIP was expressed in distal, MCP-derived SF and HOTAIR in posterior, knee-derived SF. Several positionally expressed mRNAs, e.g.HOXC8 and HOXD13, were differentially expressed in MCP-derived RA and OA SF. DAVID analysis showed positional enrichment of GOTERM limb development, anterior/posterior patterning, cartilage development, extracellular region part, cell adhesion, regulation of transcription. While some outliers where found when clustering was based on mRNA expression, clustering of SF into knee, shoulder and MCPs was perfect when based on miR expression. For example, miR-24 was positionally expressed in shoulder, miR-34c in MCP, and miR-137 in knee-derived SF, irrespective of disease. The positional expression of these miRs was confirmed in wt and TNFtg mice. Interestingly, miR-204 and miR-146a were positionally expressed in MCPs of OA but not of RA patients. These miRs were indeed positional also in wt mice but their MCP specific expression in humans correlated to ankle specific expression in wt mice. In addition, their expression was significantly changed in ankles of TNFtg compared to wt mice.

Conclusion

SF from joints of different anatomic sites exhibit particularly different mRNA and miR expression patterns suggesting that functionally unique subsets of SF populate different joints. The existence of positionally imprinted “risk” signatures of SF may account for the susceptibility of certain synovial joints to develop RA in humans and mice and may have major implications for synovial disease pathways operating early in RA.

---

« Back to 2014 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/joint-specific-positional-differences-in-coding-and-noncoding-transcriptome-of-synovial-fibroblasts-as-a-determinant-of-the-susceptibility-of-synovial-joints-to-rheumatoid-arthritis/