Article Text

OP0071 Positional Coding and Noncoding Transcriptomes of Synovial Fibroblasts in Joint Specific Patterns of Arthritis
  1. M. Frank Bertoncelj1,
  2. M. Trenkmann2,
  3. M. Armaka3,
  4. G. Russo4,
  5. A. Bratus4,
  6. C. Kolling5,
  7. B.A. Michel6,
  8. R.E. Gay6,
  9. C.D. Buckley7,
  10. G. Kollias3,
  11. S. Gay1,
  12. C. Ospelt1
  1. 1Center of Experimental Rheumatology, University Hospital Zurich, Schlieren, Switzerland
  2. 2Education & Research Centre, St. Vincent's University Hospital, Dublin, Ireland
  3. 3Institute of Immunology, Biomedical Sciences Research Center Alexander Fleming, Vari, Greece
  4. 4Functional Genomics Center Zurich, ETH Zurich and University of Zurich
  5. 5Schulthess Clinic Zurich
  6. 6Center of Experimental Rheumatology, University Hospital Zurich, Zurich, Switzerland
  7. 7Center for Translational Inflammation Research, University of Birmingham, Birmingham, United Kingdom


Background Molecular mechanisms underlying differential topographic susceptibility of joints to rheumatoid arthritis (RA) are unknown. Positional expression of Hox genes along the body axis regulates limb development. Adult skin fibroblasts retain their positional embryonic Hox code and exhibit major anatomic differences in transcriptome profiles which define their unique positional identities.

Objectives We hypothesized that synovial fibroblasts (SF), which locally drive joint destruction in RA, show positional gene expression patterns predisposing certain joints to develop site specific pathologies; e.g. symmetrical affection of small joints of the hands and feet by RA compared to asymmetrical large joints in SpA.

Methods SF were isolated from metacarpophalangeal (MCP), shoulder and knee joints of RA and osteoarthritis patients and from healthy human knees. Murine SF were obtained from front paws, ankles and knees of wild type (wt) and TNFα transgenic arthritic C57BL/6 mice. RNA sequencing (Illumina HiSeq 2000) of human MCP, shoulder and knee SF (n=21) was followed by hierarchical clustering, pathway analysis (Metacore) and qPCR. Positional epigenetic regulation and function of HOTAIR were studied in SF treated with TNFα or silenced for HOTAIR by chromatin immunoprecipitation, qPCR and ELISA.

Results Unsupervised hierarchical cluster analysis of RNASeq data showed clustering of human SF according to anatomic joint localization rather than primary disease. Small noncoding RNA conferred the strongest positional identity to SF – 21 out of 21 SF accurately clustered by the joint of origin when based on small RNA expression. Meanwhile, 2/6 arthritic, but none of healthy knee SF clustered with shoulder SF when based on long noncoding RNA (lncRNA) expression, and 2 knee and 1 shoulder arthritic SF clustered with MCP SF when based on mRNA expression, suggesting that lncRNA and mRNA may reflect weakening of positional identity of SF during the course of arthritis. Several arthritis-relevant pathways, such as chemotaxis, cell adhesion and the Jak-Stat inflammatory pathway were enriched in a joint-specific manner. In particular, matrix-degrading pathways were enriched in MCP compared to knee SF. Hox-residing RNA exhibited strict positional pattern in human and mouse. E.g. miR-196a and HOTAIR were specifically expressed in posterior knee SF and HOTTIP in anterior distal MCP/front paw SF. TNFα significantly repressed HOTAIR levels in SF. Repression of HOTAIR in anterior or TNFα-treated SF coincided with significant enrichment of H3K27me3 repressive histone mark at the HOTAIR promoter. Basal and TNFα-induced production of matrix metalloproteinases 1, 3, 13 and 14 was significantly enhanced in HOTAIR-silenced SF.

Conclusions Large positional differences in the transcriptome of SF suggest that functionally unique subsets of SF populate different joints. The existence of epigenetically imprinted positional “risk” signatures, like positional HOTAIR expression, may confer greater susceptibility to certain joints of developing destructive arthritis, such as RA.

Disclosure of Interest M. Frank Bertoncelj Grant/research support from: euroTEAM, BTCure, CABMM, IAR, M. Trenkmann: None declared, M. Armaka: None declared, G. Russo: None declared, A. Bratus: None declared, C. Kolling: None declared, B. Michel: None declared, R. Gay: None declared, C. Buckley: None declared, G. Kollias: None declared, S. Gay: None declared, C. Ospelt Grant/research support from: Grant/research support from: euroTEAM, BTCure, CABMM, IAR

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.