In this study, we evaluated the effects of a novel MMP-13 inhibitor on cartilage and joint destruction using different models of RA—the SCID mouse co-implantation model, the CIA model and the AIA model. We are aware that animal models cannot reflect the complexity of any human disease. Therefore, we used three different models of RA since each model is focused on a specific process within the disease and does not display the whole spectrum of pathological changes that occur in humans.11 12
We used the CIA and the AIA model, which are strongly driven by inflammation. Moreover, we analysed the effects of the MMP-13 inhibitor on human cells using the SCID mouse co-implantation model. This focuses specifically on the invasion of human RASF into human cartilage and is therefore relevant for the analysis of processes that interfere with destruction.13
The specific MMP-13 inhibitor decreased cartilage destruction in both, the SCID mouse and the CIA model. In the AIA model, the MMP-13 inhibitor did not show a significant effect on the destruction process. Both the CIA and the AIA are animal models in which proinflammatory processes have a major role, which are associated with excessive expressions of other matrix degrading enzymes. This might explain why the specific MMP-13 inhibitor was more effective in the non-inflammatory SCID mouse co-implantation model.
Collagenase-1 (MMP-1) also has a key role in the process of cartilage destruction in RA. It has been shown that inhibition of MMP-1 by ribozymes has a protective effect on cartilage erosion in the SCID mouse co-implantation model.14
Both MMP-1 and MMP-13 contribute to the process of cartilage destruction in RA and can therefore be targeted for treatment. However, MMP-13 is five times more effective at degrading collagen type II than MMP-1,15
suggesting that small changes in the expression level of MMP-13 may have severe consequences for the cartilage integrity. Cleaved fragments of collagen type II again can further induce MMP-13 gene expression and result in a cycle of collagen degradation.
The analysis of MMP-1 in mouse and rabbit models of arthritis is problematic because MMP-1 is present in rabbits and humans, but functionally absent in mice, whereas MMP-13 is present in all.16 17
Murine collagenase-like A (McolA) and McolB are counterparts of the human interstitial collagenase (MMP-1) with a high percentage of identities (58% in amino acids and 74% in nucleotides) but only present during mouse embryogenesis. Recombinant McolA displays proteolytic activity against type I and type II fibrillar collagens, although its specific activity versus fibrillar type I collagen is much lower than that described for human MMP-1. McolB is apparently devoid of collagenolytic activity.18
Thus, we cannot exclude the possibility that the missing effect of the MMP-13 inhibitor in the AIA model is due to the induction of MMP-1 in a highly inflammatory milieu. Since in the non-inflammatory SCID mouse model, MMP-1 is only expressed constitutively by the implanted human cells, this might account for the higher efficiency of the MMP-13 inhibitor in this model. Even though MMP-13 in rabbits is 90% and in mouse is 86% homologous to the human MMP-13 protein, it might be possible that the inhibitor designed against the human MMP-13 cannot inhibit the collagenases of other species to the same extent.17
The majority of previously developed MMP inhibitors are non-specific and inhibit a large number of MMPs. Broad-spectrum MMP inhibitors based on hydroxamate, a chemical chelator that does not show specificity for zinc, are known to have severe side effects due to inhibition of non-target metalloproteinases. The new mode of action of our MMP inhibitor has the advantage of targeting single MMPs in a highly specific manner. Therefore, it is likely that side effects like painful joint stiffening due to musculoskeletal syndrome do not occur.7
Recently, newly developed MMP inhibitors have comprised piperazine-based hydroxamic acids, and carboxylic and phosphinic acids, but the drawback of these inhibitors is that they are not selective for MMP-13 and that they have a short half-life in vivo.19–21
Only a few of the previously developed MMP inhibitors reached the clinical testing phases but none of these is specific for MMP-13. In RA, the MMP inhibitor Trocade (Hoffmann-La Roche, Basel, Switzerland) with selectivity for the collagenase MMP-1 and the gelatinases (MMP-2 and MMP-9) was not efficacious.3
Another MMP-inhibitor BB-2827 (British Biotech, Oxford, UK) a collagenase-targeting hydroxamate, was studied in clinical phase I.22
To date, only the non-chelating MMP-13 inhibitor from Alantos Pharmaceuticals (Heidelberg, Germany) has reached preclinical testing phase.11
Physiologically MMP-13 activity is controlled by naturally occurring inhibitors such as α-macroglobulins and tissue inhibitors of metalloproteinases.20 23
However, these natural inhibitors do not specifically inhibit MMP-13 to the extent that would be necessary for therapeutic intervention.
In normal human tissues, MMP-13 is scarcely detected, but it has a pivotal role in the pathogenesis of RA and osteoarthritis.24
MMP-13 is synthesised as a pro-enzyme and must be processed by proteolytic cleavage at the N-terminus to generate the active form. In vivo, MMP-14 (MT1-MMP) and MMP-2 (gelatinase A) are two of several enzymes responsible for the activation of proMMP-13 and both are overexpressed in RA. In transgenic mice, excessive MMP-13 activity can result in articular cartilage degradation.25
In RA synovial tissues, MMP-13 is detected in fibroblasts, chondrocytes, macrophages and vascular endothelial cells.26–28
Most importantly, MMP-13 is detected at sites of joint destruction.29 30
To analyse the anti-destructive effects of this novel MMP-13 inhibitor, we decided to use a prophylactic approach for all three animal models and therefore we applied the drug before the appearance of the disease, which provides more information about the pathogenesis of the disease.
In RA, joint damage is often considered to be a direct result of the inflammatory synovitis.3
Thereby, the synovial fibroblasts keep an activated status mediated by the inflammatory milieu and continuously express matrix degrading enzymes and/or activate other cells like chondrocytes, resulting in enhanced destruction of the cartilage.31
This implies that by actively treating the synovial inflammation, subsequent joint damage would be reduced and the long-term outcome of patients with RA would improve. We observed that the specific MMP-13 inhibitor was effective in destructive models but less effective in inflammatory models. An alternative is that the mechanisms causing inflammation and those leading to joint destruction are parallel processes, but only indirectly related. Clinical studies show that radiological progression of RA occurs in spite of reduction in serological markers of inflammation.32 33
Experimental studies also fail to show a close correlation between inflammation and cartilage damage.34 35
Attempts to improve the outcome of RA should not merely concentrate on controlling inflammation, but should also aim to reduce the associated connective tissue damage. Therefore, there is a true need for the development of additional anti-destructive drugs such as selective MMP-13 inhibitors—that may target specifically activated synovial fibroblasts.