1-11 CARTILAGE BIOLOGY IN OSTEOARTHRITIS: ACTIVATION OF MATRIX DEGRADING ENZYMES

Abstract

While osteoarthritis (OA) is a disease involving all joint tissues, progressive cartilage erosion remains a pathognomonic feature and a major indicator of the need for replacement surgery. Through the use of genetically-modified mice, our understanding of the molecular mechanisms that drive cartilage erosion in OA has greatly improved. Factors regulating two particular pathways have emerged as being critical in modulating OA cartilage pathology: (1) chondrocyte differentiation/hypertrophy, and (2) proteolysis of key matrix proteins (aggrecan and collagen) by metalloproteinases. These two pathways may converge as chondrocyte hypertrophic differentiation is associated with upregulation of a number of significant matrix degrading enzymes (ADAMTS and MMPs). However, data from MMP-13 KO mice may suggest it is the MMP activity rather than chondrocyte hypertrophy per se, that is the critical step in OA-cartilage degradation. Despite localised pathology we found wide-spread upregulation of chondrocyte MMP1, 2, 3, 13 and 14 mRNA in OA, suggesting local proenzyme activation regulates focal cartilage breakdown. We identified the serine proteinase, activated protein C (APC) but not its zymogen (PC), as a potential regulator of MMP activation in cartilage. PC and the proteins necessary for its activation (EPCR, TM) are expressed by chondrocytes, and PC/APC is increased in areas of MMP activation in OA. In healthy cartilage, an inflammatory stimulus (IL-1, TNF) is required for APC effect. In contrast, APC alone induces aggrecan and collagen proteolysis by MMPs in OA human cartilage. The effect of APC was still augmented by IL-1 stimulation in OA cartilage, even where IL-1 itself had no significant effect on degradation. APC did not regulate expression of MMPs or TIMPs but increased activation of pro-MMP2, 9 and 13 present in OA human cartilage. We found that APC cannot directly activate recombinant pro-MMP2, 9 or 13, and therefore investigated whether it was acting through PARs. Chondrocyte PAR expression is increased in OA and by inflammatory cytokines, and PARs are implicated in regulating MMP expression and activation. We found no difference in vitro in basal or IL-1 stimulated aggrecan or collagen proteolysis, MMP activity or mRNA expression in PAR-1 or PAR-2 KO compared with wild type (WT) cartilage, suggesting neither PAR alone is critical for cartilage degradation in mice. We confirmed this in vivo, observing no change in cartilage erosion compared to WT in PAR-1 or -2 KO 8 weeks after DMM-induced OA, a time when MMP-13 activity is critical for cartilage breakdown. In WT cartilage in the presence of IL-1, APC induced aggrecan and collagen proteolysis, and pro-MMP2, 9 and 13 activation. There was no diminution in APC action in PAR-1 or -2 KO cartilage, indicating its effect was not dependent on either receptor. Surprisingly, APC administration in WT mice commencing 1 week after DMM surgery and continuing twice weekly until sacrifice (week 8) significantly reduced cartilage degradation. This chondroprotection was not associated with altered cartilage aggrecan loss nor the appearance of the MMP-cleaved aggrecan neoepiope DIPEN. Initiating APC treatment 4 weeks after DMM surgery failed to alter cartilage degradation. These in vivo results implicate the effects of APC on other tissues in regulating OA cartilage pathology, and suggest that early intervention, perhaps to modify acute joint inflammation, may be required prevent long-term cartilage degradation after joint injury. Local activation of MMPs is a critical step in cartilage degradation in OA, providing a potential therapeutic target. However, understanding the interaction between all joint tissues, and the role of inflammation in the initiation and progression of cartilage degradation in post-traumatic OA requires further investigation.