USE OF A DIFFERENTIAL SUBTRACTION METHOD TO IDENTIFY GENES THAT CHARACTERIZE THE PHENOTYPE OF CULTURED RHEUMATOID-ARTHRITIS SYNOVIOCYTES

Objective. To identify the genes that characterize the distinctive phe notype of cultured rheumatoid arthritis (RB) fibroblastoid synoviocyte s. Methods. A representational difference method was used to subtract complementary DNA (cDNA) from cultured RA fibroblastoid synoviocytes w ith cDNA from noninflammatory osteoarthritis synoviocytes. The genes w ere identified by DNA sequencing, and their relative expression was de termined By Northern blot analysis.Results. Twenty-four genes were ide ntified, including novel genes such as a human homolog of mouse semaph orin E and one homologous to N-acetylglucosamine-6-sulfatase. Eleven o f these genes were constitutively overexpressed in the rheumatoid syno viocyte line, including a chemokine, stromal cell-derived factor 1, an d several genes capable of mediating synoviocyte-leukocyte interaction s, including vascular cell adhesion molecule 1 and Mac-2 binding prote in. Three genes (lumican, biglycan, and insulin-like growth factor bin ding protein 5) encoded extracellular matrix components, suggesting th at distinct stromal-synoviocyte interactions mag be mediated by this p henotype. Two interferon-inducible genes of unknown function were also found, emphasizing the presence of activation-like features in the ph enotype. Conclusion. A general method for the identification of differ ences in patterns of gene expression revealed that cultured RA fibrobl astoid synoviocytes overexpress certain proinflammatory genes that are potentially relevant to lymphocyte and monocyte entry and interaction s, The features of the genes identified in these mesenchymal cells sug gest that they facilitate localization of immune reactions to the join t through leukocyte chemokinesis, cell-cell adhesion, and matrix speci alization. The further characterization of these genes should help in resolving whether this phenotype is the consequence of modulation and imprinting by an inflammatory milieu or, more likely, whether it refle cts the intrinsic lineage characteristics of intimal lining synoviocyt es.