Tissue Inhibitors of Metalloproteinases

By Daren B. Filsinger

The role of matrix metalloproteinases (mmps) and tissue inhibitors of metalloproteinases (TIMPs) is extremely complex and poorly understood. TIMPs, cannot simply be viewed as inhibitors of the extracellular matrix (ECM) degrading enzymes, MMPs, but as regulators of growth factors, tumor invasion, metastasis, angiogenesis, and apoptosis.1-3 The function of TIMPs in cancer is therefore aggressively researched and debated.

Matrix Metalloproteinases

MMPs belong to a family of zinc endopeptidases capable of degrading numerous components of the ECM. Most MMPs in cancer are probably not produced by the cancer cells themselves, but by local stromal and inflammatory cells. The cancer cells produce a stimulator for MMP release, known as tumor cell-derived collagenase stimulating factor (TCSP) or extracellular matrix metalloproteinase inducer (EMMPRIN).4 MMPs are then secreted as proenzymes and activated by cleavage and subsequent exposure of the substrate binding pocket which contains Zn++.

Tissue Inhibitors of Metalloproteinases

Currently four endogenous TIMPs are known (TIMP-1,-2,-3,-4). They are each approximately 25 kDa in size and have a highly homologous N-terminal domain. This domain is responsible for the irreversible inhibition of the active site of metalloproteinases. The C-terminal domains have less homology and confer different specificities and affinities for each MMP and other molecules. Blavier and associates have summarized the diverse functions of TIMPs as 1) either direct or indirect inhibition of vascular endothelial cell proliferation; 2) degradation of insulin-like growth factor-binding protein-3, releasing active IGF; 3) cell cycle and migration inhibition by intracellular signaling of an intact ECM, which is maintained by TIMPs; 4) indirect activation of MMP-2 by TIMP-2 via binding the metalloproteinase in close proximity to its membrane-bound activator enzyme; and 5) suppression of apoptosis by TIMP-1 in malignant Burkitt cells, potentially by a TIMP receptor.3 A recent study demonstrates TIMP-2 and synthetic matrix metalloproteinase inhibitors (MMPIs), but not TIMP-1, induce apoptosis in human T lymphocytes.5 Depending on the cell line and specific inhibitor, TIMPs and MMPIs can have either inhibitory or stimulatory effects on cell cycle. This summary demonstrates the complexity of endogenous TIMPs and the controversy of their therapeutic usefulness.

Because multiple MMPs are notably higher in cancerous tissues, therapeutically exerting an imbalance in favor of TIMPs is, at first glance, of potential benefit. The stage and type of cancer may dictate how effective TIMPs may be. Aggressive cancers and extensive metastasis may overwhelm the effects of medicinal TIMP. Furthermore, little is known about the clearance of TIMPs but a short half-life in serum prevents systemic use and has lead to an interest in TIMP modification to increase half-life and gene therapy to induce overexpression of TIMPs in cancer cells.3

Synthetic Matrix Metalloproteinase Inhibitors

In order to overcome the obstacles of endogenous TIMPs, research has turned to synthetic inhibitors, the MMPIs. The goals in mind are to create the ideal inhibitor with oral bioavailabilty, an increased half-life, and specific activity only blocking the matrix degrading activity of MMPs while eliminating potential adverse effects like apoptotic inhibition and growth factor activation. Synthetic inhibitors work by entering the substrate binding pocket of MMPs and act as a zinc chelator. Early MMPIs, like batimastat, were broad-spectrum and effective against all matrix metalloproteinases. The use of batimastat in blocking tumor invasion in pre-malignant intestinal tumors resulted in a 48% decrease in tumorigenesis in mice.6 The same study by Goss and colleagues demonstrated TIMP-1 as having far different effects, some of which were deleterious in mice. This again proves the complex nature of endogenous inhibitors. Newer synthetic inhibitors maintain a zinc chelating ability but are selective for specific MMPs. Whether this is of benefit has yet to be determined, because in most cancerous states mutiple MMPs are elevated and no specific matrix metalloproteinase has been linked to cancer. Groups like the gelatinases (MMP-2 and MMP-9), however, have been implicated.7

Recent Studies

Although most MMPIs have not reached human clinical trials, drugs like marimastat are leading the way in human trials determining efficacy, side effects, and dose, while other synthetic inhibitors are being modified and animal tested. Being noncytotoxic drugs, MMPIs have the potential to be used in combination therapy and elicit few side effects.

Some MMP inhibitors appear to eliminate one known side effect that others may elicit while maintaining anticancer effects. Preliminary results of a small study by Drummond and associates compared the effects of two different unnamed broad-spectrum inhibitors and two unnamed selective inhibitors on tumor growth (melanoma) and ability to cause tendinitis, a common side effect of some MMPIs.8 Only one of the four tested maintained anticancer efficacy and did not cause tendinitis. The inhibitor with these results was one of the broad-spectrum inhibitors with activity unlike most broad-spectrum inhibitors. It has the ability to block the release of TNF-alpha. The other broad-spectrum inhibitor, like most, lacks the ability to bind to tumor necrosis factor-alpha convertace (TACE), which itself is a metalloproteinase. Inhibiting the release of the cytokine, TNF-alpha, decreases the possibilty of developing inflammatory side effects like tendinitis.9 The results of this small study indicate that selective inhibitors may not be efficacious against tumor growth and anti-TACE activity is required to eliminate inflammatory side effects.

In a more extensive study, a potent, selective MMPI, AG3340, demonstrated that selective inhibitors are effective against tumor growth.7 AG3340 was tested against human prostate, colon, lung, breast, and glial tumors that were grafted to mice. AG3340 is a synthetic inhibitor which is specific for the gelatinases and MMP-13 and -14 and has anti-TACE activity. Dose-dependent inhibitions of growth were observed in colon and lung models, while in breast cancer models the minimal dose demonstrated maximal inhibition. Tumor growth was also inhibited in glioma and prostate tumor models. The most significant results appeared in colon, lung, and glioma models with 65-79% growth inhibition compared to controls (P < 0.05). With early treatment, AG3340 also reduced angiogenesis in colon tumors. Potentiated efficacy was observed with combination chemotherapy using AG3340 and either Taxol or carboplatin.


The role of matrix metalloproteinases in cancer has proven to be significant, taking part in angiogenesis, growth, metastasis, and apoptosis. The complexity of the MMPs and their endogenous and exogenous inhibitors leaves a great deal to be discovered. Improved understanding of the sources, target, and mechanisms of action of both MMPs and their inhibitors is essential for effective and safe treatment. Clinical studies, however, are promising and may lead to a novel, noncytotoxic treatment to be used in combination therapy. (Mr. Filsinger is a medical student at the State University of New York at Buffalo.)


1. Fowlkes JL, Enghild JJ, Suzuki K, et al. J Biol Chem 1994;269:25742-25746.

2. Valente P, Fassina G, Melciori A, et al. Int J Cancer 1998;75:246-253.

3. Blavier L, Henriet P, Imren S, et al. Ann NY Acad Sci 1999;878:108-119.

4. Kataoka H, DeCastro R, Zucker S, et al. Cancer Res 1993;53:3155-3158.

5. Lim MS, Guedez L, Stetler-Stevenson WG, et al. Ann NY Acad Sci 1999;878:522-523.

6. Hepner Goss KJ, Brown PD, Matrisian LM. Int J Cancer 1998;78:629-635.

7. Shalinski DR, Brekken J, Zou H, et al. Ann NY Acad Sci 1999;878:236-270.

8. Drummond AH, Beckett P, Brown P, et al. Ann NY Acad Sci 1999;878:228-235.

9. Gearing AJH, Beckett P, Christodoulou M, et al. J Leukoc Biol 1995;57:774-777.