The matrix metalloproteinases (MMPs) are a family of Zn2+-dependent endopeptidases that regulate the degradation of ECM and play a pivotal role in many physiological and pathological processes of different tissues [2–4]. Indeed, the timely breakdown of ECM is essential for embryonic development, morphogenesis, reproduction, and tissue resorption and remodeling [5].

MMPs are categorized into the following groups: collagenases (MMP-1, MMP-8, and MMP-13), gelatinases (MMP-2 and MMP-9), stromelysins (MMP-3, MMP-10, and MMP-11), matrilysin (MMP-7), metalloelastase (MMP-12), and membrane-type matrix metalloproteinases (MT-MMP 1, 2, 3, and 4) [6].

Several studies dealing with the pathomechanisms of OA refer to MMP-1, -3, -7, -8, and -13 [7–12] whereas a smaller number of investigations have pointed out the pathogenic role of gelatinases in OA. Indeed, the ECM of articular cartilage is primarily composed of type II collagen and aggrecan that are not the primary substrates of gelatinases. These gelatinases are best known for their involvement in pulmonary [13], myocardial [14], and neoplastic disease [15] but they are emerging as important proteases implicated in the OA progression.

MMP-2 participates in ECM degradation with a wide range of substrates. Indeed, it is able to degrade type I, IV, V, VII, and X collagens, laminin, elastin, fibronectin, and proteoglycans [26–28]. Normal adult articular chondrocytes express significant amount of MMP-2 both in vivo and in vitro suggesting this metalloproteinase is involved in physiological collagen turnover of human adult articular cartilage [29].

MMP-9 (gelatinase B, 92kDa type IV collagenase) was first purified from human macrophages [20]. Its expression is limited to osteoclasts, macrophages, trophoblasts, hippocampal neurocytes, and migrating keratinocytes [22]. In particular MMP-9 and cathepsin K are considered the most abundant proteases in osteoclasts [30]. MMP-9 is controlled by growth factors, chemokines, and other stimulatory signals [31]. It is secreted as an inactive precursor form named pro-MMP-9 that forms a tight complex with TIMP-1 and TIMP-3. The complex of pro-MMP-9 and TIMP-1 [24, 32], the plasmin, and the complex plasminogen/MMP-3 are activators of pro-MMP-9 [26, 33].

Gelatinase B has been shown to dissolve extracellular matrix, initiating and promoting new vessel formation [34]. Furthermore, this enzyme is known to cleave native type IV, V, VII, and X collagens and elastin, as well as the products of collagens types I, II, and III after proteolysis by collagenases (Table 1) [3].

Current data suggest that during OA, the activity of gelatinases is higher on the subchondral bone rather than on cartilage ECM [39]. Indeed, MMP-2 is capable of cleaving type I and other fibrillar collagens [40] that are uncommon in the ECM of articular cartilage but are present in the ECM of subchondral bone. Using a specific gelatinase inhibitor, Hill et al.showed that both the gelatinases participate in the degradation of the organic matrix of bone [41]. Mansell and Bailey investigating the cancellous bone metabolism during OA, reported an increased potential for collagen degradation in presence of increased levels of both pro- and active MMP-2 [42]. What is evident from this study is that OA cancellous bone is metabolically active compared with normal tissue. Such differences in turnover might result in altered joint morphology, which in turn might exacerbate the osteoarthritic process (Figure 1). Osteoclasts constitutively express MMP-2, and synthesize MMP-9, MMP-3, and TIMP-1 in response to IL-1α stimulation, and during OA the increased levels of osteoclast-derived MMPs might contribute to osteoclast lacunar resorption [43]. This hypothesis concurs with the demonstration of higher plasma levels of MMP-9 in patients with rapidly destructive hip OA in comparison patients with OA or normal controls [44]. The higher detected amount of MMP-9 could be explained by the wide number of osteoclasts, which are one source of MMP-9, observed around necrotic bone in subchondral areas in rapidly destructive hip OA [45, 46]. Nor should be excluded the enhanced production of MMP-9 by synovial cells of patients with this kind of OA [47]. A direct route into the bloodstream via the subchondral microcirculatory system and an indirect route from synovial fluid into circulation could explain the higher plasma levels of MMP-9 in OA [48].

The work of Bellido et al.strengthens the role of the subchondral bone as a key player in the puzzle of OA development [49]. They found raised subchondral MMP-9 levels in patient suffering from OA demonstrating a clear increase in local bone resorption with a decreased thickness of the subchondral plate. The subchondral plate that separates articular noncalcified cartilage from the bone marrow cavity consists of calcified cartilage and subchondral lamellar bone layers [50]. Any impairment in subchondral bone quality makes this organ not able to receive and properly distribute loads from and/or to the articular cartilage. Thus, changes at subchondral bone may aggravate cartilage damage. Indeed, the authors observed a direct correlation between subchondral structural parameters and cartilage damage evaluated with the Mankin's scale. The presence of subchondral bone resorption pits composed by MMP-producing cells derived from bone marrow has been previously evidenced together with their contribution to cartilage degradation through the invasion of this tissue [51].

High levels of VEGF and gelatinases were confirmed in osteoarthritic fluid but MMP-2 and -9 levels were not significantly associated with VEGF expression [59]. Differently, it was previously demonstrated that the MMP-9 levels correlated with synovial fluid VEGF levels and with the pattern of vascularity found in the synovial membrane tissue of patients with arthritis. Moreover, synovial membrane explants stimulated with VEGF increased supernatant MMP-9 levels by 2-fold from baseline [60]. Recently, Li et al. [61] demonstrated that higher MMP-9 expression was found in case of severe OA in comparison with early OA and this expression correlated in a direct manner with vascular invasion. These findings together suggest a possible relationship among gelatinases and the angiogenesis noted with OA development, and it is tempting to speculate that MMP-9 may be therapeutic target for angiogenesis inhibition.

New regulatory mechanisms of gelatinase expression have been proposed. The increase of gelatinases in OA might be induced by abnormal mechanical pressure applied to the articulation. Indeed, cyclic compression on osteoblasts from osteoarthritic subchondral bone increases the expression of genes coding for MMP-9. Conversely, MMP-2 was not modulated by compression, suggesting that this is not a mechanosensitive gene [62]. An interesting article demonstrated that TGF-β1 protects articular cartilage by downregulating the expression of MMP-9 of chondrocytes and synoviocytes in OA, which may delay the biological behavior of this disease. The authors found a negative correlation between the expression of MMP-9 protein and TGF-β1 protein, and between the expression of MMP-9 mRNA and TGF-β1 mRNA in the specimens of osteoarthritic cartilage [63].