In silico validation of a cartilage specific circadian clock: mutation of BMAL1 increased MMP expression

Purpose: Osteoarthritis (OA) is a complex, progressive degenerative cartilage disease involving remodeling of joint tissues. Several signaling pathways are involved in the pathogenesis of osteoarthritis. The intrinsic circadian clock is also known to be involved in OA pathophysiology. Circadian rhythm, which is the regulation of activity of man in rhythmic fashion by light and dark across a 24h period, is governed not only centrally but also peripherally by the same molecular events. Changes in physiology and gene expression are regulated in this fashion. In cartilage, a number of genes, including MMP13, have been shown to be regulated with circadian expression. We previously showed that a computational model built using the software tool ANIMO, allowed us to understand the complex network of major signaling pathways involved in pathogenesis of osteoarthritis. In this study, we built a cartilage specific computational model to understand the complex molecular network of the circadian clock in osteoarthritis.

Methods: Based on known molecular events of the circadian clock, a cartilage specific computational model was built in ANIMO (Figure 1). Using BMAL1 as output, the network was linked with MMP13 (catabolic) and COL2A (anabolic) gene expression, which are important indicators of cartilage homeostasis. This computational model was used as a tool to predict dynamics of the network. Our model predictions were validated by protein and gene expression studies by western-blot, immunofluorescence and qPCR in primary (healthy and osteoarthritic) chondrocytes by synchronizing them for circadian rhythm in culture via a 2h treatment with 50% FBS/ 100 nM dexamethasone. Cells were then cultured over 72h in culture medium containing 0.5 % FBS and samples were withdrawn at an interval of 4h.

Results: The model predicted that the system would exhibit an oscillating behavior over a period of 48h. The periodicity, phase and shape of the predicted oscillating system was quantified by mRNA expression data from literature. When compared with mRNA expression data over 24h at 4h interval for CRY, PER, COL2A, BMAL1 and MMP13, the dynamics of these proteins except COL2A matched to a great extent. Furthermore, the time dependent variability in protein level was confirmed experimentally in cultured chondrocytes by western blot and immunofluorescence. BMAL1 expression was found to be lower in osteoarthritic chondrocytes whereas pro-MMP13 expression was higher. Finally, the model predicted that the BMAL1 silencing leads to higher expression of MMP13 (Figure 2), which was tested by BMAL1 mutation in cultured chondrocytes.