RA-specific expression profiles and new candidate genes

Objective

To identify rheumatoid arthritis- (RA)-specific profiles of differentially expressed genes.

Methods

Synovial tissues from RA and osteoarthritis (OA) patients and from normal joints were selected according to their disease-characteristic histology. Gene expression was analyzed using DNA microarrays (GeneChip; Unigene-array) and representational difference analysis (RDA). Data were validated on larger cohorts of patients by RT-PCR.

Results

Nine hundred and eighty genes were significantly regulated in RA synovial tissue as compared with non-RA. Specialized cluster analysis identified a set of 312 genes as sufficient of unequivocally discriminating RA from non-RA patterns (class discovery). Genes of highest regulation were associated with leukocyte activation (chemokines, chemokine receptors, B- and T-cell genes), endothelial and angiogenic activation, tissue destruction and remodelling [MMP-3, BMP-4, TIMPs]. Interestingly, a large set of genes was down-regulated in RA (TGF-β superfamily, apoptosis-related genes, transcription factors). Osteopontin-like genes (n = 46) – up-regulated in RA – and glutathione peroxidase-3-like genes (n = 85) – down-regulated in RA – yielded the highest correlation coefficients (>0.94). Megakaryocyte stimulating factor (MSF), down-regulated in a subset of RA, may hold the key to subclassification: a loss-of-function mutation in the MSF-encoding gene leads to synovial hyperplasia in camptodactyly–arthropathy–coxa vara-pericarditis syndrome, and, as in RA, also to pericardial involvement. A further candidate, vitamin-D3-up-regulated protein-1 (VDUP-1), is regulated like MSF and predisposes to premature coronary artery disease when mutated, again a feature of a subset of RA.

Conclusion

RA specific gene profiles were identified and are useful to improve diagnostics of the disease. Novel gene candidates not yet in the focus of RA pathogenesis have been identified that are likely to further the understanding of RA.