Matrix metalloproteinase-9-mediated Syndecan-4 shedding correlates with osteoarthritis severity

Abstract

Osteoarthritis (OA) is a degenerative joint disease, characterized by the extracellular matrix (ECM) breakdown of articular cartilage, as well as by changes in the synovial membrane, subchondral bone and synovial fluid composition. The heparan sulphate proteoglycan Syndecan- 4 (Sdc4) has been shown to play a crucial role in the progression of OA. In an animal model of OA, based on the destabilization of the medial meniscus (DMM), Sdc4 deficiency and antibody-mediated blocking of Sdc4 ameliorated clinical signs of OA. Interleukin-1 (IL-1) signaling is repeatedly correlated with OA pathology. It was shown that IL-1¯, binds directly to Sdc4 and that Sdc4 regulates the IL-1 receptor type I (IL-1R1) presentation at the membrane surface and thereby modulates IL-1 signaling. Moreover Sdc4 is a target for extracellular domain shedding. Several studies show that Sdc4 shedding is crucial in physiological, as well as pathophysiological, processes. Thus, I investigated the involvement of Sdc4 shedding in the development and progression of OA. Shed Sdc4 levels were analyzed in knee synovial fluid of OA patients using a specific ELISA. The severity of OA in the individual patients was assessed by the radiographic Kellgren-Lawrence score (KL 1-4). The levels of shed Sdc4 correlated with the KL score, and the receiver operating characteristic (ROC) curve presented shed Sdc4 as a suitable biomarker for discriminating early (KL 1-2) and late (KL 3-4) OA. Levels of the matrix metalloproteinases (MMP)-2 and -9 as potential sheddases, analyzed by specific ELISAs, showed an increase depending on the KL score forMMP-9, whereas levels ofMMP-2 did not change. However, assessment with the ROC curve showed thatMMP-9 was less suitable to discriminate between early and late OA. Further, levels ofMMP-9 significantly correlated with levels of shed Sdc4, suggestingMMP-9 to be a potential sheddase for Sdc4 in OA.MMP-2 levels did not correlate with shed Sdc4 levels. Expression levels of Sdc4 and MMP-2 mRNA in articular cartilage and synovial membrane did not change depending on the KL score, whereasMMP-9 mRNA levels were upregulated with increasing KL score in the synovial membrane. MMP-9 mRNA expression in cartilage, however, did not change in context of OA severity. To elucidate the role of Sdc4 shedding in OA, Sdc4 shedding was induced in chondrocytes by 4-Aminophenylmercuric acetate (APMA), an organomercurial activator of MMPs, and IL-1¯ stimulation. The addition of an MMP-9 inhibitor reduced basal, as well as APMA- and IL-1¯-induced, Sdc4 shedding. An inhibitor for MMP-2 did not influence Sdc4 shedding, whether basal or induced. In IL-1¯-induced Sdc4 shedding the inhibition ofMMP-9 had more prominent effects with higher concentrations of IL-1¯. The samewas apparent when MMP-9 was knocked down by siRNA. APMA-induced increase of Sdc4 shedding attenuated IL-1 signaling. The addition of the soluble ectodomain of Sdc4 had no effect on IL-1 signaling. Thereby, the possibility of shed Sdc4 being a decoy receptor for IL-1 was ruled out. In conclusion, this study presents shed Sdc4 as a possible biomarker for knee OA severity for the first time, whileMMP-9 was shown to be a sheddases in Sdc4 shedding in OA.Moreover, it is shown that IL-1 signaling can be modulated by Sdc4 shedding, displaying a possible role for the increased shedding in OA.