Abstract. Introduction. Risk factors for osteoarthritis include raised BMI and female gender. Whether these two factors influenced synovial gene expression was investigated using a triangulation and modelling strategy which generated 12 datasets of gene expression in synovial tissue from three knee pathologies with matching BMI groups, obese and overweight, and gender distributions. Methodology. Intra-operative synovial biopsies were immersed in RNAlater at 4oC before storage at -80oC. Total RNA was extracted using RNAeasy with gDNA removal. Following RT- PCR and quality assessment, cDNA was applied to Affymetrix Clariom D microarray gene chips. Bioinformatics analyses were performed. Linear models were prepared in limma with gender and BMI factors incorporated sequentially for each pathology comparison, generating 12 models of probes differentially expressed at FDR p<0.05 and Bayes number, B>0. Data analysis of differently expressed genes utilized Ingenuity Pathway Analysis and Cytoscape with Cluego and Cytohubba plug-ins. Results. Expression of 453 synovial genes was influenced by BMI and gender, 360 encode proteins such as HIF-1a, HSF1, HSPA4, HSPA5. Top canonical pathways include Unfolded protein response, Protein Ubiquiitation and Clathrin mediated endocytosis signalling linked by modulation of heat shock proteins, comparable to pathology dependent regulation. In addition BMI and gender modulate gene expression in the NRF2-mediated oxidative stress response pathway with down regulation of Glutathione-S-transferases potentially down regulating antioxidant defences. Conclusion. The enhanced risk of osteoarthritis induced by an elevated BMI and female gender maybe include differential expression of heat shock proteins and genes in the NRF2 pathway

Abstract. Introduction. It is increasingly evident that synovium may play a larger role in the aetiology of osteoarthritis. We compared gene expression in whole tissue synovial biopsies from end-stage knee osteoarthritis and knee trauma patients with that of their paired explant cultures to determine how accurately cultured cells represent holistic synovial function. Methodology. Synovial tissue biopsies were taken from 16 arthroplasty patients and 8 tibial plateau fracture patients with no osteoarthritis. Pairs of whole tissue fragments were either immediately immersed in RNAlater Stabilisation Solution at 4o C before transfer to -80o C storage until RNA extraction; or weighed, minced and cultured at 500mg tissues/5ml media in a humidified incubator at 37oC, 5% CO2. After sub-culturing total RNA was extracted using RNAeasy Plus Mini Kit with gDNA removal. Following RT-PCR and quality assessment, cDNA was applied to Affymetrix Clariom D microarray gene chips. Bioinformatics analyses were performed. Results. PCA analysis illustrates the clear separation of expression array data from cultured cells compared with their parental whole tissues and no segregation between cells derived from osteoarthritic or trauma tissues. A differentially expressed gene heat map demonstrated the hierarchical independence of cultured cells from their paired sample parental tissues. The biological pathways enriched by these gene expression differences emphasise the activities of macrophages and lymphocytes lost from culture. Conclusion. Adherent synovial cells grown from different knee pathologies lose the expression patterns characteristic of their originating pathology. Interpretation of data needs caution as the cells are not representative of whole synovium

Abstract. Introduction. Synovitis impacts osteoarthritis symptomatology and progression. The transcription factors controlling synovial gene expression have not been described. This study analyses gene expression in synovium samples from 16 patients with osteoarthritis with 9 undergoing arthroscopic and 8 knee trauma surgery for non-arthritic pathologies. Methodology. Intra-operative synovial biopsies were immersed in RNAlater at 4oC before storage at -80oC. Total RNA was extracted using RNAeasy. After purification, RT-PCR and quality assessment, cDNA was applied to Affymetrix Clariom D microarray gene chips. Bioinformatics analyses were performed. Linear models were prepared in limma with gender and BMI factors incorporated sequentially for each pathology comparison, generating 12 models of probes differentially expressed at FDR p<0.05 and Bayes number, B>0. Data analysis of differently expressed genes utilized Ingenuity Pathway Analysis and Cytoscape with Cluego and Cytohubba plug-ins. Results. Amongst the 2084 genes with significantly differential expression (DEG), 135 had transcription regulator capabilities and 121 a nuclear location. IPA analysis of OATKR and arthroscopic tissue comparison DEG identified 12 nuclear transcription factors linked to 31 DEG whose encoded proteins located within cytoplasmic and cell membrane compartments. All 12 were significantly up-regulated and acting in pathways up-regulating transcription of DNA and RNA, cell survival and angiogenesis while down-regulating senescence and apoptosis. NFE2L2, integral to the TGF-beta signalling pathway, was identified as a bottleneck gene. Conclusion. This analysis indicates the complexity of synovial gene expression regulation and offers target genes and pathways for evaluation during osteoarthritis pathogenesis