Background/Purpose: OA affects the entire joint but most studies have focused on the disease process in a single tissue. In this study, we identified genes regulated during different stages of the development of surgically-induced OA by microarray using RNA isolated from the joint “organ” and analyzed the data using an unbiased computational modeling approach to discover the pathways active in the disease process.

Methods: 12 week-old male C57/BL6 mice underwent surgical destabilization of the medial meniscus (DMM) to induce OA or sham surgery as control. Joint tissues were collected for isolation of RNA (n=9 mice per group per time point) pre-surgery (time 0) and at 2, 4, 8, and 16 weeks after surgery and for histological analysis of OA severity (n=6 mice per group per time point). RNA was isolated from joint tissue collected from the medial half of the joint, including cartilage, meniscus, subchondral bone, and joint capsule with synovium. RNA was pooled from 3 mice for each Affymetrix microarray and 3 arrays were performed for each group at each time point. Signal log ratios (SLR) of DMM/sham were calculated using normalized array data. Genes passing detection, SLR (≥0.5 or ≤0.5 for at least one time point in all 3 pools), and consistency filters were used for computational modeling to identify patterns of gene expression by consensus clustering, network analysis using  jActiveModules (JAM) and functional classification using DAVID and KEGG.  

Results: Histological lesions of OA were present in the medial tibial plateaus (MTP) of the DMM knees beginning at the earliest (2 week) time point and became progressively more severe by 16 weeks. Osteophytes were cartilaginous at 2 weeks and became progressively ossified. A total of 427 genes passed the consistency and significance filters. There were more upregulated than downregulated genes at all time points except at 8 weeks (17 up­, 53 down) with the most upregulated at 4 weeks (336­ up,33 down) followed by 2 weeks (174 up­,12 down) and 16 weeks (84­ up,2 down). Clustering identified 27 clusters with 2 or more genes and DAVID analysis of clusters upregulated at 2 and 4 weeks included morphogenesis, differentiation, development, collagen, and ECM genes as well as transcription regulatory genes. Cell division and cytoskeleton genes were in a cluster highly down-regulated at 8 weeks while genes upregulated at 16 weeks included Prelp (involved in collagen binding), Col3a1 and fibromodulin. JAM analysis revealed 13 subnetworks with gene expression activity during the time course where the majority of up-regulated genes were up at 4 weeks. Prominent sub-networks included the TGF-β signaling pathway, ECM-receptor interactions (including thrombospondins, syndecans, and collagens) and Wnt and hedgehog signaling.         

Conclusion: The results support a phasic development of OA. Early matrix remodeling was associated with activation of TGFβ and Wnt/hedgehog signaling which may be drivers of cartilage degradation and osteophyte formation. The quiescent stage at 8 weeks suggests a temporary stabilization of the joint followed by activation of a more fibrotic process at 16 weeks. The findings suggest that specific therapies intended to slow disease progression may be most effective at specific stages of the disease.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/a-systems-biology-approach-to-elucidating-pathways-active-during-the-development-of-osteoarthritis/