Targeting Pericellular Proteolysis in Vascular Disease
Proteolytic enzymes released from smooth muscle cells (SMCs) degrade extracellular matrix proteins, and this is thought to facilitate cell migration and neointimal thickening in restenosis and vein graft disease. The plasminogen activator (PA) and the matrix metalloproteinase (MMP) systems play important roles mediating these processes. In the PA system, tissue-type plasminogen activator (t-PA) and urokinase-type plasminogen activator (u-PA) cleave plasminogen to release plasmin (Figure A). Many components of the PA system including t-PA, u-PA, and the endogenous plasminogen activator inhibitor (PAI-1) are upregulated in diseased blood vessels,1 and recent studies using mice with targeted gene deletion point to a role for u-PA in mediating neointimal hyperplasia.2 u-PA may have a particularly important role in facilitating SMC migration, because it is localized to the cell surface by binding to the u-PA receptor (u-PAR). This potentiates the activity of u-PA by bringing the enzyme into close proximity to its surface-bound plasminogen substrate, permitting plasmin activation within a spatially constrained pericellular environment (Figure A). Plasmin can directly degrade some components of the extracellular matrix and has the potential to activate several of the matrix metalloproteinases including MMP-3, 9, 12, and 13. In fact, there is good evidence that the effects of u-PA/plasmin in tissue remodeling are mediated indirectly via activation of the MMPs.1
The MMPs are a family of enzymes that degrade many molecules of the extracellular matrix. MMPs are secreted in a latent zymogen form, with a propeptide tightly coupled to the enzyme active site. MMP zymogens are activated extracellularly when the propeptide is cleaved by plasmin or by other MMPs (Figure A). MMP activity can be inhibited by direct binding of endogenous tissue inhibitors of metalloproteinases (TIMPs) to the active site.3 MMP-1, 2, 3, 9, 12, and 13 are upregulated in diseased vessels, and MMPs produced by SMCs are thought to clear a path for migration of these cells from media or adventitia to the intima.4 Taken together, this data suggests cooperative roles for the PA and the MMP systems in mediating matrix degradation and smooth muscle cell migration; however, very few studies have addressed this directly.
In an article published in this issue of Circulation Research, Lamfers et al5 describe a strategy to inhibit both protease families simultaneously. They generated cDNA constructs encoding a novel hybrid protein TIMP-1.ATF, consisting of the MMP inhibitor TIMP-1 linked to the amino terminal fragment (ATF) of u-PA. The ATF of u-PA lacks enzymatic activity, but competes with native u-PA for binding to u-PAR, thus reducing the generation of plasmin (Figure B).6 Furthermore, this strategy was used to anchor TIMP-1 at the cell surface, and the authors postulated that this would neutralize cell-surface MMP activity. Human saphenous vein SMCs were transfected in vitro with the recombinant TIMP-1.ATF construct under the control of an adenoviral promoter. Overexpression of TIMP-1.ATF inhibited both MMP activity and plasmin generation by SMCs. Importantly, TIMP-1.ATF expression inhibited SMC migration in vitro. The effects of the TIMP-1.ATF hybrid protein were greater than the effects of either TIMP-1 or ATF expressed individually. This was probably due to a combined effect whereby plasmin-mediated activation of MMP zymogens was reduced, while TIMP-1 anchored at the cell membrane directly inhibited local MMP activity. Next, they transfected human saphenous veins with TIMP-1.ATF and studied neointimal formation in organ culture. TIMP-1.ATF significantly attenuated intimal formation measured after 4 weeks in culture, leading to a 69% decrease in intimal size.
These studies outline a strategy that simultaneously hits two proteases involved in cell migration and targets inhibition to the cell surface. The molecular mechanisms of migration are complex and multifaceted involving the extrusion of lamellopodia and formation of new adhesive contacts at the leading edge of the cell, followed by detachment and release of the cell tail. Precise control of cell-cell and cell matrix adhesions, reorganization of the cytoskeleton, and degradation of the extracellular matrix are required to accomplish this.7 There is convincing evidence for specific mechanisms that confine and concentrate proteases at the cell surface. In fact, MMP-1 is found discretely localized to the leading edge and the tail of SMCs migrating on collagen.8 Large multimolecular complexes that form on the cell surface include matrix molecules, matrix receptors, proteases, protease receptors, protease inhibitors, and degraded matrix products. These complexes link to the cytoskeleton, and to intracellular signaling pathways to coordinate cell migration.3 In short, all the action is localized at the cell surface.
The present studies of Lamfers et al5 suggest a mechanism whereby TIMP-1 anchored to the cell surface acts locally to inhibit surface bound MMPs, preventing the degradation of cell-matrix contacts and matrix molecules, thus attenuating the process of cell migration. MMPs bind to cell surface molecules including integrins, CD44, EMMPRIN, and the membrane-type MMPs (MT-MMPs).9–11⇓⇓ u-PA and plasminogen are also localized at the cell surface. Furthermore, there is crosstalk between u-PAR and integrin receptors; u-PAR binds to several integrins and also binds directly to vitronectin in the matrix.12 It is reasonable to postulate therefore that TIMP-1.ATF interferes with these interactions, but this has not yet been investigated.
Another limitation of the present study is that TIMP-1.ATF was not tested in an in vivo model. Past studies using systemic administration of MMP or PA inhibitors, or gene transfection to overexpress TIMPs or PAI-1 in the vessel wall in vivo reduced SMC migration, but it is not clear that long-term neointimal hyperplasia was inhibited in all cases, due to a catch-up effect caused by continued SMC proliferation.4 The fact that TIMP-1.ATF failed to inhibit SMC proliferation is a significant drawback. Thus, it seems that any strategy to limit neointimal hyperplasia must also target cell proliferation.
At this time, it is not clear whether the MMPs are directly involved in the control of cell proliferation. However, in this context, it is important to note that the MMPs and TIMPs have functions beyond their roles in matrix degradation. MMPs can activate cytokines, release growth factors from matrix, and degrade integrins and cadherins. TIMPs can inhibit these MMP-dependent processes, or independently stimulate cell growth or apoptosis.3 There is compelling evidence that MMPs may initiate tumor cell growth by releasing growth factors from the extracellular matrix, or by cleaving cadherins, thereby disrupting cell-cell adhesions, leading to β-catenin signaling and increased cell proliferation.13 This diversity of actions must be taken into account when attempting to elucidate the role of the MMP-TIMP system in any pathological process.
In conclusion, the data presented in this study suggests a unique strategy to curtail cell-surface proteolytic activity in the arterial wall. The results shed light on our understanding of the mechanisms of SMC migration and intimal thickening, and provide support for the hypothesis that pericellular proteolysis is an important component of these responses.
The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.
- ↵Carmeliet P, Moons L, Herbert JM, Crawley J, Lupu F, Lijnen R, Collen D. Urokinase but not plasminogen activator mediates arterial neointima formation in mice. Cir Res. 1997; 96: 829–839.
- ↵Galis ZS, Khatri JJ. Matrix metalloproteinases in vascular remodeling and atherogenesis: the good, the bad, and the ugly. Circ Res. 2002; 90: 251–262.
- ↵Lamfers MLM, Grimbergen J, Aalders MC, Havenga MJ, de Vries MR, Huisman LGM, van Hinsbergh VWM, Quax PHA. Gene transfer of the urokinase-type plasminogen activator receptor-targeted matrix metalloproteinase inhibitor TIMP-1.ATF suppresses neointima formation more efficiently than tissue inhibitor of metalloproteinase-1. Circ Res. 2002; 91: 945–952.
- ↵Lamfers ML, Wijnberg MJ, Grimbergen JM, Huisman LG, Aalders MC, Cohen FN, Verheijen JH, van Hinsbergh VW, Quax PH. Adenoviral gene transfer of a u-PA receptor-binding plasmin inhibitor and green fluorescent protein: inhibition of migration and visualization of expression. Thromb Haemost. 2000; 84: 460–467.
- ↵Li S, Chow LH, Pickering JG. Cell surface-bound collagenase-1 and focal substrate degradation stimulate the rear release of motile vascular smooth muscle cells. J Biol Chem. 2000; 275: 35384–35392.
- ↵Dumin JA, Dickeson SK, Stricker TP, Bhattacharyya-Pakrasi M, Roby JD, Santoro SA, Parks WC. Pro-collagenase-1 (matrix metalloproteinase-1) binds the α2β1 integrin upon release from keratinocytes migrating on type I collagen. J Biol Chem. 2001; 276: 29368–29374.
- ↵Yu Q, Stamenkovic I. Localization of matrix metalloproteinase 9 to the cell surface provides a mechanism for CD44-mediated tumor invasion. Genes Dev. 1999; 13: 35–48.
- ↵Ho AT, Voura EB, Soloway PD, Watson KL, Khokha R. MMP inhibitors augment fibroblast adhesion through stabilization of focal adhesion contacts and up-regulation of cadherin function. J Biol Chem. 2001; 276: 40215–40224.