Physiological and pathological vascular remodeling in

Transcription

1 Targeted Disruption of the Matrix Metalloproteinase-9 Gene Impairs Smooth Muscle Cell Migration and Geometrical Arterial Remodeling Zorina S. Galis, Chad Johnson, Denis Godin, Richard Magid, J. Michael Shipley, Robert M. Senior, Eugen Ivan Downloaded from by guest on May 22, 2018 Abstract Matrix remodeling plays an important role in the physiological and pathological remodeling of blood vessels. We specifically investigated the role of matrix metalloproteinase (MMP)-9, an MMP induced during arterial remodeling, by assessing the effects of genetic MMP-9 deficiency on major parameters of arterial remodeling using the mouse carotid artery flow cessation model. Compared with remodeling of matched wild-type (WT) arteries, MMP-9 deficiency decreased intimal hyperplasia, reduced the late lumen loss, eliminated the correlation between intimal hyperplasia and geometric remodeling, and led to significant accumulation of interstitial collagen. Biochemical analysis of MMP-9 knockout (KO) arterial tissue and isolated smooth muscle cells (SMCs) confirmed the lack of MMP-9 expression or compensation by other gelatinases. To investigate potential mechanisms for the in vivo observations, we analyzed in vitro effects of MMP-9 deficiency on the migration, proliferation, and collagen gel contracting capacity of aortic SMCs isolated from MMP-9 KO and WT mice. Although proliferation was comparable, we found that MMP-9 deficient cells had not only decreased migratory activity, but they also had decreased capacity to contract collagen compared with WT cells. Thus, MMP-9 appears to be involved not only in degradation, but also in reorganization of collagenous matrix, both facets being essential for the outcome of arterial remodeling. Our results also establish MMP-9 as an attractive therapeutic target for limiting the effects of pathological arterial remodeling in restenosis and atherosclerosis. (Circ Res. 2002;91: -.) Key Words: matrix degradation cell migration restenosis atherosclerosis Physiological and pathological vascular remodeling in response to a variety of stimuli, including hemodynamic changes, inflammation, and injury, leads to reshaping of the vessel wall. Inappropriate remodeling is currently thought to be the main cause of most prevalent vascular pathologies of arteries, including atherosclerosis and restenosis. 1 Degradation of the matrix scaffold enables cell movement and general tissue reorganization, making specialized enzymes called matrix metalloproteinases (MMPs) 2 prime candidates for agents of vascular remodeling. 3 Although many studies have addressed and endorsed a role for MMPs in pathological remodeling, suggesting these as potential therapeutic targets in restenosis and atherosclerosis, due to the current lack of specific MMP inhibitors, such studies could not resolve the identity of the specific relevant MMP(s). We thus decided to explore the effects of MMP-9 genetic deficiency on the remodeling of carotid arteries in the flow cessation murine model, 4 which allows investigation of both formation of intimal hyperplasia and arterial geometrical remodeling, 5 the two main processes implicated in the stenotic remodeling of human arteries. Materials and Methods Animal Model MMP-9 expression was disrupted by deletion of most of the exon 2 of the MMP-9 gene in the 129/SvEv genetic background (Jackson Labs, Bar Harbor, Maine), as previously described. 6 Mouse carotid artery remodeling was triggered through ligation of the left common carotid proximal to its bifurcation, 4 in male six weeks old MMP-9 knockout (KO) and background matched wild-type (WT) mice. Carotid arteries were harvested 1, 3, 7, 14, and 28 days after surgery, fresh for biochemical analysis (5 animals for each time point), or after fixation at physiological pressure of 100 mm Hg using 10% formalin in phosphate-buffered saline (PBS) for morphological studies (4 to 5 animals per time point). Carotid artery specimens further fixed for 16 hours were embedded in paraffin. The Emory University Committee on Institutional Animal Care and Use approved all protocols. Original received April 9, 2002; revision received September 26, 2002; accepted September 30, From the Division of Cardiology (Z.S.G., C.J., D.G., R.M., E.I.), Emory University School of Medicine, Atlanta, Ga; the Department of Biomedical Engineering (Z.S.G., C.J., R.M.), Emory University/Georgia Institute of Technology, Atlanta, Ga; and the Pulmonary and Critical Care Division, Department of Medicine and Department of Cell Biology and Physiology (J.M.S., R.M.S.), Washington University School of Medicine, St Louis, Mo. Presented in part at the North American Vascular Biology Meeting, 2000, and the Second Annual Conference on Arteriosclerosis, Thrombosis, and Vascular Biology, Arlington, Va, May 11 13, 2001, and published in abstract form (Arterioscl Thromb Vasc Biol. 2001;21:667). Correspondence to Zorina S. Galis, PhD, Depts of Medicine and Biomedical Engineering, Emory University School of Medicine, 1639 Pierce Dr, WMB 319, Atlanta GA zgalis@emory.edu 2002 American Heart Association, Inc. Circulation Research is available at DOI: /01.RES

2 2 Circulation Research November 1, 2002 An expanded Materials and Methods section can be found in the online data supplement available at Morphometric and Statistical Analysis To minimize animal to animal variation, all morphometric measurements were performed at the vascular lesion apex, identified in each individual carotid artery using a procedure previously described in detail. 5 For each time point, morphometric measurements obtained from 4 to 5 individual carotid arteries were used to generate a mean value SEM. For multiple comparisons, we tested differences between groups using ANOVA, followed by a t test to reveal significant differences, with a value of P 0.05 considered significant. Linear regression, using intimal thickness as the predictor and EEL as the dependent variable, was performed for the day 1 to day 7 specimens (Minitab 13). Collagen Analysis We used a modification of the Escolar et al 7 technique to quantify collagen after specific staining of paraffin sections with Picrosirius red (online data supplement). Collagen quantification was done at the lesion apex for 3 to 5 carotid arteries per time point. Immunohistochemistry Tissue sections from the apex were rehydrated, deparaffinized, and blocked for endogenous peroxidase activity and then for nonspecific staining. We used anti-smc actin antibody (Sigma; 1:200) followed by the MOM kit (Vector Labs) to minimize nonspecific staining, and chromogenic reaction with the DAB kit (Vector Labs). Anti-Mac3 (Rat anti-mouse, Pharmingen, 1:500), followed by Rhodamine Red X goat anti-rat (Jackson Immunoresearch; 1:50) was used for detection of macrophages. For fluorescence microscopy, cell nuclei were counterstained with Hoechst (Sigma). In Vitro Experiments With Mouse Aortic SMCs Arterial SMC were obtained from outgrowths of aortic explants harvested from MMP-9 KO and WT mice (online data supplement). The capacity of SMCs to contract a collagen gel was estimated using water exclusion as a measure of gel contraction 8 (online data supplement). Migratory activity of WT or MMP-9 KO SMC was compared determining the total number of cells and the average distance traveled from the wound edge 24 hours after wounding of confluent monolayers in serum-free conditions (details in online data supplement). SDS-PAGE Zymography Levels of MMP-2 and MMP-9 were detected in tissue lysates of individual carotid arteries normalized by protein (20 g), run in parallel with prestained molecular weight markers (BioRad) in 10% SDS-PAGE gels containing 1% gelatin, as described previously. 5 The optical volume-density product of individual lytic bands was quantified using the Molecular Analyst program (BioRad) and compared with a mouse MMP-9 standard obtained from cultured WT mouse bone marrow macrophages. Data Analysis Biochemical data obtained for each carotid artery by densitometric analysis of gels were averaged for 4 to 5 individual carotid arteries per each time point. Collagen gel contraction and SMC migration data were resultant from 3 independent experiments. For all data, comparisons were made by ANOVA followed by Student s t test to compare the means of multiple groups using Microsoft Excel and Minitab Release 13. Means were considered significantly different if P Results The profile and evolution of gelatinolytic activity, detected by SDS-PAGE zymography in the ligated arteries of the 129/ SvEv WT strain used in the present experiments (Figure 1) indicated the early induction of MMP-9, similar to our previous observations in the remodeling carotid artery of the C57 Black 6J (C57BL/6J) mice. 5 As expected, no MMP-9 activity was detected at any time point in the MMP-9 KO. Importantly, we did not detect any compensatory increase of gelatinolytic activity by MMP-2 or other gelatinases, neither in the MMP-9 KO carotid artery tissue extracts (Figure 1) nor in the culture media of cultured MMP-9 KO SMCs. In vitro experiments with cultured SMCs (online Figure, available in the data supplement found at demonstrated a similar amount of MMP-2 was secreted, ie, levels of total MMP-2: in WT; in MMP-9 KO (respectively, pro-mmp-2: versus ; active MMP-2: versus ; P NS for all pairs). Morphological examination of the ligated carotid artery in the 129/SvEv WT indicated that arterial remodeling included formation of a neointima and late lumen loss, as reported previously in other mouse strains (Figure 2). 4,5 However, morphological examination of the remodeling MMP-9 KO and WT carotid arteries (Figure 2) suggested differences further investigated by comparing parameters commonly used to characterize arterial remodeling (summarized in online Table, available in the data supplement found at MMP-9 Deficiency Decreases the Late Lumen Loss The lumen size is of major importance for the proper function of a remodeling artery, thus we have investigated the ultimate effects of MMP-9 deficiency on the lumen size. In the WT mouse, although the increased size of the EEL did not reach statistical significance, as found previously in this model in the C57 BL/6 background, 9 formation of the neointima initially did not result in a lumen loss, suggesting early compensation through geometrical remodeling. On the other hand, in the MMP-9 KO mouse, the lumen tended to be smaller in spite of significantly decreased intimal thickening of the remodeling carotid arteries up to 7 days (Figure 2). This trend was later reversed such as at 28 days we found that the residual lumen, expressed by normalizing the lumen size to the value at day 0 (Table) was only 14% of initial size in WT carotid arteries, whereas this was nearly 50% of the initial lumen in the MMP-9 KO at 28 days. As a result, the remodeling carotid arteries had significantly larger lumen in the MMP-9 KO compared with the WT ( m 2 versus m 2 at 28 days; P 0.05). To explain this significant effect on the late lumen size, we further investigated the effect of MMP-9 deficiency on the two major processes that determine lumen size, intimal thickening, and geometric arterial remodeling. MMP-9 Deficiency Decreases Intimal Thickening We found that the neointima thickness tended to be smaller in the MMP-9 KO compared with the WT remodeling arteries (Figure 2). The difference increased with time, becoming statistical significant at 28 days ( m 2 in KO versus m 2 in WT, n 5; P 0.05). Cell type specific immunostaining showed that neointima contained mainly SMC -actin positive cells in both the MMP-9 KO and WT, whereas inflammatory cells were not detected in

3 Galis et al MMP-9 Deficiency and Arterial Remodeling 3 Figure 1. Gelatinolytic activity present in tissue extract of carotid arteries harvested from the carotid arteries of MMP-9 KO and WT mice at several time points after ligation (SDS-PAGE gelatin zymography). Presence of MMP-9 and MMP-2 is indicated by their lytic activity (white bands) and apparent molecular weight. MMP-9 activity is induced in the remodeling WT carotid artery as early as 1 day after ligation. As expected, MMP-9 expression is undetectable in the KO arterial tissue (or by isolated SMC, see online data supplement). No compensatory regulation of other gelatinases was detectable. Lower graphs illustrate quantification of MMP-2/9 levels obtained from averaging densitometric data obtained from 5 to 7 independent carotid arteries for each time point by SDS-PAGE gelatin gels. these lesions (Figure 3). An analysis of cell number in the neointima, performed by image analysis of Hoechst nuclear staining, confirmed a decreased number of intimal cells in the MMP-9 KO versus WT mice at all time points (online Table). MMP-9 Deficiency Decreases Migration of SMC The arterial intima of many species used as experimental models for remodeling, including the mouse, does not initially contain SMCs. Thus, cells need to migrate from the medial layer across the internal elastic lamina (IEL) to give rise to the neointima (Figure 3), offering a convenient experimental system for investigating factors that modulate SMC migration in vivo. Although such models, especially the rat carotid artery injury model, have been used for many years, methods that precisely quantify smooth muscle migration in vivo are still unavailable. We estimated SMC migration into the neointima for the first 3 days of remodeling using a simple model, as previously described in detail. 5 The number of intimal cells at any given time point was considered to be the net result of adding cells, through earlier migration from the medial layer and proliferation within the neointima, and loss of cells, through death and potentially through emigration. However, during this time interval, death of intimal SMCs based the nuclear staining pattern with Hoechst 10 was found to be negligible (not shown). Using the in vitro thymidine incorporation assay, we found that proliferation rates of cultured MMP-9 KO and WT SMCs were not significantly different (not shown), and we estimated that the doubling time was 24 hours. Using the actual values obtained from image analysis of tissue specimens of WT or MMP-9 KO, we estimated that approximately 3 times more cells had migrated into the intima of WT compared with the MMP-9 KO carotid arteries (38 in WT versus 12 in KO) within the first 3 days of remodeling. Difference is maintained if we consider that in vivo cell doubling time was shorter (eg, for 16 hours: 33 cells in WT versus 8 cells in KO). For a more accurate assessment of the effects of MMP-9 deficiency on SMC migration, we performed in vitro experiments using SMCs isolated from the medial layer of MMP-9 KO or WT aortas. We found that MMP-9 deficiency was associated with significant impairment of SMC migration ability (Figure 4), expressed either as the total number of migrating cells (61 5 for KO versus for WT; P 0.05), or as the average distance traveled by each cell during the assay (50 12 m for KO versus m for WT; P 0.05) (Figure 4). Taken together, these results

4 4 Circulation Research November 1, 2002 Figure 2. Comparison of overall remodeling of WT vs MMP-9 KO carotid artery at 0, 7, 14, and 28 days after ligation. Top, Characteristic morphology of elastin stained cross sections (bar 50 m). Bottom, Morphometric and statistical analyses of average lumen area and intimal thickness detected in such histological specimens. MMP-9 deficiency decreases the late (28 days) lumen loss and reduces intimal thickness in the remodeling mouse carotid artery. For each time point, mean values were obtained by averaging data for the apex of the lesions obtained in 5 different animals (*P 0.05). suggest that the inhibitory effect of MMP-9 deficiency on SMC migration was a major reason for decreased neointima formation. MMP-9 Deficiency Blunts Geometrical Remodeling of Carotid Artery At day 7, in spite of greater neointimal thickening in the WT arteries ( m for WT versus m for KO), their average lumen was larger than that of the MMP-9 KO (Figure 2), suggesting potential additional effects of MMP-9 deficiency on the early compensatory geometric remodeling. A statistical analysis of morphometric data collected from 0, 1, 3, and 7 day time points plotting a linear regression of the external elastic lamina (EEL, dependent variable) against intimal thickness (independent variable) (Figure 5) indicated that the size of neointimal thickening was a strong predictor of the EEL perimeter in the WT carotid artery (r , P 0.01). In contrast, neointimal thickness did not correlate with the EEL perimeter in the MMP-9 KO carotid arteries during early remodeling (r , P NS), suggesting that MMP-9 deficiency had impaired the physiological tendency for remodeling that compensates for development of intimal lesions. Effect of MMP-9 Deficiency Upon the Late Lumen Loss Day 0 Day 7 Day 14 Day 28 Wild Type (0.02) 0.96 (0.11) 0.56 (0.31) 0.14 (0.04) MMP-9 KO (0.05) 0.85 (0.19) 0.95 (0.15) 0.49 (0.13)* Normalized lumen area (SEM). *P 0.05.

5 Galis et al MMP-9 Deficiency and Arterial Remodeling 5 Figure 3. MMP-9 deficiency reduces the extent of intimal hyperplasia (fluorescence microscopy). Top, Morphology of cross sections of WT and MMP-9 KO carotid arteries 7 days after ligation compared with day 0 (left insets): cell nuclei counterstained blue with Hoechst, the internal elastic lamina (IEL, thin white arrow), and external elastic lamina (EEL, block white arrow) identified due to their green autofluorescence. Note the large number of intimal cells detected after 7 days on top of IEL in the WT artery, indicating the formation of neointima (thickness indicated by double red arrow). Right inset shows that these cells are positive for SMC -actin. In contrast, the sparse cell nuclei above the IEL in the MMP-9 KO carotid artery at 7 days suggest delay in neointima formation. Bottom, Immunohistochemical staining for macrophages, which were not detectable in either WT or MMP-9 KO carotid arteries even in fully developed lesions, suggesting that circulating monocytes are not a significant contributor to neointima formation in these mice. In contrast, macrophages were readily detectable in the neointima triggered by the same procedure in the ApoE KO carotid arteries (Positive control, red signal of Texas Red labeled secondary antibody). Negative control illustrates the level of fluorescence in specimens processed by omitting the primary anti-macrophage antibodies. MMP-9 Participates in the Metabolism and Organization of Interstitial Collagen We hypothesized that the impaired arterial geometrical remodeling in MMP-9 deficient animals may be mediated through effects on collagen metabolism and/or organization, thought to play an important role in the geometrical remodeling of arteries. Although constrictive arterial remodeling was found to bear resemblance to wound contraction, 11 characterized by increased collagen synthesis and tissue contraction by cells, degradation of the existing collagen scaffold is most likely necessary to allow for the outward or expansive remodeling of the arterial wall. Accumulation of fibrillar collagen in the remodeling mouse carotid artery, as indicated by analysis of Picrosirius red staining, was found to be most striking in the adventitial layer (Figure 6). The amount of adventitial collagen normalized by EEL perimeter increased significantly with time in the MMP-9 KO arteries (from at day 1, to arbitrary units/ m EEL at day 28; P 0.05), although it did not change significantly during the 28 days in WT arteries ( at 1 day to arbitrary units/ m EEL at 28 days; P NS). This accumulation of collagen seemed to be responsible for a doubling of the adventitial area in the MMP-9 KO carotid artery compared with that of WT at 28 days after ligation ( m 2 in KO versus m 2 in WT; P 0.05), because the adventitial cell number was similar ( cells/adventitia in KO versus cells/adventitia in WT; P NS). Accumulation of interstitial collagen within the wall might oppose the expansive remodeling, as suggested by the morphometric data in the MMP-9 KO arteries, or even constrict the arterial wall. However, at 28 days, in spite of a significant accumulation of collagen in the MMP-9 KO remodeling artery, the overall size (EEL) was comparable in the KO and WT arteries. This lack of tissue constriction in spite of collagen accumulation in the MMP-9 KO remodeling artery suggested the presence of an opposing mechanism. We thus investigated in vitro the hypothesis that MMP-9 deficiency impairs the capacity of arterial SMC to contract a collagen gel (Figure 6). Indeed, we found that mouse MMP-9 KO arterial SMC have a significantly impaired capacity to contract collagen gels in vitro when compared with WT SMC ( % by MMP-9 KO cells versus % for WT; P 0.05). Discussion The action of MMPs has been linked to the pathological remodeling of blood vessels that constitutes the basis for

6 6 Circulation Research November 1, 2002 Figure 4. Arterial SMC migration in vitro (fluorescence microscopy, red phalloidin staining of actin, nuclei counterstained with blue Hoechst) in a wound assay (white dotted line indicates wound edge). Inset, Positive identification of cultured cells by staining with anti- SMC -actin antibody (green fluorescence). After 6 hours, many more WT SMCs had migrated further from the wound edge compared with MMP-9 KO SMCs. Bottom, Quantification of results from a typical 24-hour experiment, demonstrating that MMP-9 deficiency significantly impairs the migratory capacity of arterial SMCs, measured both by quantification of total migrated cell number and the average distance migrated by each cell (n 4, *P 0.05). cardiovascular conditions with significant mortality and morbidity. Specifically, MMP-9 has been associated with the development of lesions after vascular surgical interventions, 12 appears to have a role in vein graft occlusion after coronary bypass surgery, 13,14 and seems to be associated with acute vascular syndromes, 15,16 suggesting that MMP-9 may be an attractive target for therapeutic intervention. However, lack of specificity, poor oral bioavailability, and unrelated side effects of currently available synthetic inhibitors do not allow the proper demonstration of MMP-9 involvement in pathological arterial remodeling. We thus decided to investigate the role of MMP-9 in arterial remodeling by using targeted genetic disruption of MMP-9 gene, which results in complete loss of MMP-9 expression and activity, 6 and the mouse carotid artery flow cessation model. In spite of its inherent limitations, this animal model was found to allow the in vivo investigation of arterial geometric remodeling and intimal hyperplasia formation, 5 the two major processes that control pathological remodeling of human arteries. Our previous study in the C57BL/6J mouse strain also suggested a role for MMP-9 in both these processes. Although following the general pattern previously observed in the C57BL/6J background, the remodeling of mouse carotid arteries in the 129/SvEv background seemed attenuated, consistent with recent observations regarding strain-related differences in the magnitude of the carotid artery remodeling response. 17 Several previous studies have examined the possibility that MMP-2 and MMP-9 may facilitate SMC migration based on their basement membrane-degrading and elastolytic activity, 18 thought to contribute to growth of intimal lesions in atherosclerosis and restenosis; however, human arteries already have intimal SMCs. Increased expression and activity of MMP-2 and MMP-9 induction were detected in several experimental models of vascular lesion formation, including those triggered by balloon injury in the carotid artery of rats and rabbits. 22 Increased MMP-2 and -9 activity was associated also with SMC migration in vitro. 23 Nonspecific MMP inhibitors, as well as neutralizing anti-mmp antibodies were shown to limit SMC migration in vitro 24 proliferation in vitro and in situ 21,25 and to delay formation of neointima in the rat model. 19,20 Differences observed in our current in vivo and in vitro studies using the specific inhibition of MMP-9 due to genetic deficiency support a major enabling role of MMP-9 in SMC migration in vitro and in situ. We suggest that impairment of SMC migration is the major reason for decreased formation of intimal hyperplasia, supporting the utility of specific MMP-9 inhibitors to limit postintervention and in-stent restenosis. Use of nonspecific MMP inhibitors was also found to decrease SMC proliferation. 21 However, we do not expect this to have happened in our study where we found no differences in cell proliferation between WT and KO in situ as assessed in situ by BrdU incorporation (online Table) and in classical in vitro thymidine incorporation experiments with isolated SMCs (data not shown). Recent studies, including our own, have highlighted the contribution of recruitment of bone marrow derived cells to formation of intimal lesions associated with postangioplasty restenosis, graft vasculopathy, and hyperlipidemia-induced atherosclerosis in the ApoE KO mouse. 26,27 However, immunohistochemical analysis of the lesions in this study, as well as in our previous studies 9 did not reveal a significant number of macrophages, suggesting that circulating cells are not a major contributor to intimal lesions that do not involve endothelial denudation or an

7 Galis et al MMP-9 Deficiency and Arterial Remodeling 7 Figure 5. Linear regression analysis of external elastic lamina (EEL) dependence on the thickness of neointima confirms a positive correlation in the remodeling WT carotid arteries but not in the MMP-9 KO (r , P NS), suggesting that MMP-9 deficiency had impaired the physiological tendency of geometrical remodeling to compensate for growth of neointima. immunological challenge in mice that are not prone to atherogenesis. Using an integrated approach to take into account the relation between changes in the EEL, intimal thickening, and the ultimate lumen size, we also found that MMP-9 deficiency affects arterial geometrical remodeling. Several studies have suggested that the inappropriate geometrical remodeling, either through lack of compensatory enlargement or due to the active shrinking of the remodeling artery, is in fact a more important determinant of lumen loss than intimal thickening We propose that MMP-9 deficiency effects on geometrical remodeling were most likely exerted through modulation of collagen metabolism and organization. The action of MMPs, and particularly of MMP-9, has been recently associated with the outward arterial remodeling. 5,31,32 MMP inhibition using the wide spectrum inhibitor marimastat was found to decrease the constrictive arterial remodeling after balloon angioplasty in a pig model. 33 These findings are consistent with our observations of the effects of specific MMP-9 inhibition on arterial geometrical remodeling in deficient mice. Compared with the WT, the MMP-9 KO remodeling artery showed a significant in situ accumulation of interstitial collagen, as assessed using quantification of the characteristic reaction with Picrosirius red, a histochemical stain that specifically reveals interstitial collagen 7 previously used to investigate in situ collagen accumulation in experimental and human arterial lesions. 34 Besides the potential implication of the participation of MMP-9 to geometric arterial remodeling via enhancing accumulation of collagen, we believe that this observation may have relevance to the biology of MMPs in vivo. Our finding suggests that MMP-9, known to degrade short collagens and elastin, may also participate directly or indirectly to the metabolism of interstitial collagen in vivo. Preliminary experiments suggest that the MMP-9KO and WT SMCs have comparable capacity for collagen synthesis in vitro (data not shown). If confirmed, this finding may indicate that MMP-9 has a larger in vivo array of substrates than previously inferred from in vitro experiments and, together with other previous observations reporting the collagen degrading capacity of MMP-2, 35 challenges the notion that interstitial collagenases are the only MMPs controlling the turnover of interstitial collagen in vivo. Limiting the knowledge regarding MMP in vivo substrates in general is the fact that our information is based largely on results from in vitro assays, testing purified enzymes against purified substrates, 2 which further emphasizes the importance of in vivo models. Another related issue was suggested by the lack of compensation by MMP-2 or other gelatinases in the MMP-9 KO, as demonstrated by gelatin SDS-PAGE zymography of tissue extracts and cell culture medium, suggesting that although MMP-2 and MMP-9 enzymatic activities may seem similar in vitro, these MMPs may have distinct activities in vivo. Interestingly, the accumulation of collagen in the carotid arteries of MMP-9 KO did not enhance the constrictive remodeling of arteries, as expected from other studies, 11 and the late lumen loss, an undesirable effect in human arteries. We propose that this lack of constriction effect is due to the

8 8 Circulation Research November 1, 2002 Figure 6. MMP-9 deficiency increases collagen accumulation in the remodeling arteries and impairs the SMC capacity to contract collagen gels. A, Collagen distribution 14 days after ligation in the remodeling arteries of WT and MMP-9 KO arteries, revealed by elastin staining combined with Picrosirius red. Two different magnifications of the same section are illustrated under direct and polarized light (all bars 50 m). Note the dense red staining indicating accumulation of interstitial collagen. Arrows indicate the levels of internal elastic lamina (IEL) and external elastic lamina (EEL). B, Left, Quantification of collagen content normalized by EEL perimeter confirms significant accumulation of collagen in the remodeling MMP-9 KO, compared with the WT carotid artery, supporting an in vivo role for MMP-9. B, Right, Quantification of in vitro collagen gel contraction by SMCs isolated from WT or MMP-9 KO aortas. MMP-9 deficient SMCs have a significantly impaired capacity to contact collagen gels (n 3, *P 0.05). fact that MMP-9 deficiency seems to impair the capacity of cells to compact a collagen matrix, as indicated by our in vitro studies with isolated arterial SMCs. Although this appears to be a novel function for MMP-9, deserving further detailed study, similar effects have been reported for MMP Mechanisms that may be relevant for the organization of a matrix scaffold and could be controlled by MMPs include the modification of matrix components so as to increase cell adhesion or to provide additional binding sites for other matrix components and the release of biological active factors from the matrix. Taken together, these experimental results suggest a key role for MMP-9 in vascular remodeling, through mechanisms that rely on degradation and reorganization of extracellular matrix. In addition, our results suggest that specific inhibition of MMP-9 may increase the collagen content of arteries, thus potentially their mechanical stability, while at the same time decreasing intimal hyperplasia and the late lumen loss. Because interstitial collagen is the major contributor to the mechanical strength of arteries in general and of atherosclerotic plaques, our observations may have significant implications regarding the role of MMP-9 in atherosclerotic plaque stability. They could provide mechanistic support for previous observations that correlated the weakening of the fibrous cap 37 with the increased MMP-9 activity in the vulnerable shoulders of atherosclerotic plaques, 38 supporting the contribution of MMP-9 to the destabilization of atherosclerotic plaques. These observations support the utility of specific MMP-9 inhibition as a therapeutic strategy in cardiovascular disease. Acknowledgments Funding for these studies was provided through the NIH RO1HL64689 and the American Heart Association Established Investigator Award No N to Dr Zorina Galis, and the HL29594, HL47328, and Alan A. and Edith L. Wolff Charitable Trust to Dr Robert M. Senior. Dr Eugen Ivan was supported through a NRSA Award Training grant No. T32HL and Chad Johnson through NSF Award EEC References 1. Pasterkamp G, de Kleijn DP, Borst C. Arterial remodeling in atherosclerosis, restenosis and after alteration of blood flow: potential mechanisms and clinical implications. Cardiovasc Res. 2000;45:

9 Galis et al MMP-9 Deficiency and Arterial Remodeling 9 2. Woessner JF. The matrix metalloproteinase family. In: Parks WC, Mecham RP, ed. Matrix Metalloproteinases. 1st ed. San Diego, Calif: Academic Press; 1998: Galis ZS, Khatri JJ. Matrix metalloproteinases in vascular remodeling and atherogenesis: the good, the bad, and the ugly. Circ Res. 2002;90: Kumar A, Lindner V. Remodeling with neointima formation in the mouse carotid artery after cessation of blood flow. Arterioscl Thromb Vasc Biol. 1997;17: Godin D, Ivan E, Johnson C, Magid R, Galis ZS. Remodeling of carotid artery is associated with increased expression of matrix metalloproteinases in mouse blood flow cessation model. Circulation. 2000;102: Vu TH, Shipley JM, Bergers G, Berger JE, Helms JA, Hanahan D, Shapiro SD, Senior RM, Werb Z. MMP-9/gelatinase B is a key regulator of growth plate angiogenesis and apoptosis of hypertrophic chondrocytes. Cell. 1998;93: Escolar JD, Tejero C, Escolar MA, Montalvo F, R G. Methodological contributions for the morphometric study of the lung: approximation to the ideal sample size and quantification of collagen fiber. Anat Rec. 1997;247: Nakagawa S, Pawelek P, Grinnell F. Extracellular matrix organization modulates fibroblast growth and growth factor responsiveness. Exp Cell Res. 1989;182: Ivan E, Khatri JJ, Johnson C, Magid R, Godin D, Nandi S, Lessner S, Galis ZS. Expansive arterial remodeling is associated with increased neointimal macrophage foam cell content: the murine model of macrophage-rich carotid artery lesions. Circulation. 2002;105: Pollman MJ, Hall JL, Gibbons GH. Determinants of vascular smooth muscle cell apoptosis after balloon angioplasty injury: Influence of redox state and cell phenotype. Circ Res. 1999;84: Geary RL, Nikkari ST, Wagner WD, Williams JK, Adams MR, Dean RH. Wound healing: a paradigm for lumen narrowing after arterial reconstruction. J Vasc Surg. 1998;27: Galis ZS, Sukhova GK, Kranzhöfer R, Clark S, Libby P. Macrophage foam cells from experimental atheroma constitutively produce matrixdegrading proteinases. Proc Natl Acad Sci U S A. 1995;92: Southgate KM, Mehta D, Izzat MB, Newby AC, Angelini GD. Increased secretion of basement membrane-degrading metalloproteinases in pig saphenous vein into carotid artery interposition grafts. Arterioscl Thromb Vasc Biol. 1999;19: Mavromatis K, Fukai T, Tate M, Chesler N, Ku DN, Galis ZS. Early effects of arterial hemodynamic conditions on human saphenous veins perfused ex vivo. Arterioscl Thromb Vasc Biol. 2000;20: Brown DL, Hibbs MS, Kearney M, Loushin C, Isner JM. Identification of 92-kD gelatinase in human coronary atherosclerotic lesions. Association of active enzyme synthesis with unstable angina. Circulation. 1995;91: Inokubo Y, Hanada H, Ishizaka H, Fukushi T, Kamada T, Okumura K. Plasma levels of matrix metalloproteinase-9 and tissue inhibitor of metalloproteinase-1 are increased in the coronary circulation in patients with acute coronary syndrome. Am Heart J. 2001;141: Kuhel DG, Zhu B, Witte DP, Hui DY. Distinction in genetic determinants for injury-induced neointimal hyperplasia, and diet-induced atherosclerosis in inbred mice. 2002;22: Senior RM, Griffin GL, Fliszar CJ, Shapiro SD, Goldberg GI, Welgus HG. Human 92- and 72-kilodalton type IV collagenases are elastases. J Biol Chem. 1991;266: Bendeck MP, Zempo N, Clowes AW, Galardy RE, Reidy MA. Smooth muscle cell migration and matrix metalloproteinase expression after arterial injury in the rat. Circ Res. 1994;75: Zempo N, Kenagy RD, Au YPT, Bendeck M, Clowes MM, Reidy MA, Clowes AW. Matrix metalloproteinases of vascular wall cells are increased in balloon-injured rat carotid artery. J Vasc Surg. 1994;20: Bendeck MP, Conte M, Zhang M, Nili N, Strauss BH, Farwell SM. Doxycycline modulates smooth muscle cell growth, migration, and matrix remodeling after arterial injury. Am J Path. 2002;160: Wang H, Keiser JA. Expression of membrane-type matrix metalloproteinase in rabbit neointimal tissue and its correlation with matrixmetalloproteinase-2 activation. J Vasc Res. 1998;35: Pauly RR, Passaniti A, Bilato C, Monticone R, Cheng L, Papadopoulos N, Gluzband YA, Smith L, Weinstein C, Lakatta EG, Crow MT. Migration of cultured vascular smooth muscle cells through a basement membrane barrier requires type IV collagenase activity and is inhibited by cellular differentiation. Circ Res. 1994;75: Kenagy RD, Hart CE, Stetler-Stevenson WG, Clowes AW. Primate smooth muscle cell migration from aortic explants is mediated by endogenous platelet-derived growth factor and basic fibroblast growth factor acting through matrix metalloproteinases 2 and 9. Circulation. 1997;96: Southgate KM, Davies M, Booth RF, Newby AC. Involvement of extracellular-matrix-degrading metalloproteinases in rabbit aortic smooth-muscle cell proliferation. Biochem J. 1992;288: Sata M, Saiura A, Kunisato A, Tojo A, Okada S, Tokuhisa T, Hirai H, Makuuchi M, Hirata Y, Nagai R. Hematopoietic stem cells differentiate into vascular cells that participate in the pathogenesis of atherosclerosis. Nature Medicine. 2002;8: Lessner SM, Prado HL, Waller EK, Galis ZS. Atherosclerotic lesions grow through recruitment and proliferation of circulating monocytes in a murine model. Am J Pathol. 2002;160: Clarkson TB, Prichard RW, Morgan TM, Petrick GS, Klein KP. Remodeling of coronary arteries in human and nonhuman primates. JAMA. 1994;271: Kakuta T, Currier JW, Haudenschild CC, Ryan TJ, Faxon DP. Differences in compensatory vessel enlargement, not intimal formation, account for restenosis after angioplasty in the hypercholesterolemic rabbit model. Circulation. 1994;89: Pasterkamp G, Wensing PJ, Post MJ, Hillen B, Mali WP, Borst C. Paradoxical arterial wall shrinkage may contribute to luminal narrowing of human atherosclerotic femoral arteries. Circulation. 1995;91: Mason DP, Kenagy RD, Hasenstab D, Bowen-Pope DF, Seifert RA, Coats S, Hawkins SM, Clowes AW. Matrix metalloproteinase-9 overexpression enhances vascular smooth muscle cell migration and alters remodeling in the injured rat carotid artery. Circ Res. 1999;85: de Smet BJGL, de Kleijn D, Hanemaaijer R, Verheijen JH, Robertus L, van der Helm YJM, Borst C, Post MJ. Metalloproteinase inhibition reduces constrictive arterial remodeling after balloon angioplasty: a study in the atherosclerotic Yucatan micropig. Circulation. 2000;101: Sierevogel MJ, Pasterkamp G, Velema E, de Jaegere PPT, de Smet BJGL, Verheijen JH, de Kleijn DPV, Borst C. Oral matrix metalloproteinase inhibition and arterial remodeling after balloon dilation: an intravascular ultrasound study in the pig. Circulation. 2001;103: Aikawa M, Rabkin E, Okada Y, Voglic SJ, Clinton SK, Brinckerhoff CE, Sukhova GK, Libby P. Lipid lowering by diet reduces matrix metalloproteinase activity and increases collagen content of rabbit atheroma: a potential mechanism of lesion stabilization. Circulation. 1998;97: Aimes RT, Quigley JP. Matrix metalloproteinase-2 is an interstitial collagenase. Inhibitor-free enzyme catalyzes the cleavage of collagen fibrils and soluble native type I collagen generating the specific 3/4- and 1/4- length fragments. J Biol Chem. 1995;270: Bullard KM, Mudgett J, Scheuenstuhl H, Hunt TK, Banda MJ. Stromelysin-1-deficient fibroblasts display impaired contraction in vitro. J Surg Res. 1999;84: Shah PK, Falk E, Badimon JJ, Fernandez-Ortiz A, Mailhac A, Villareal-Levy G, Fallon JT, Regnstrom J, Fuster V. Human monocytederived macrophages induce collagen breakdown in fibrous caps of atherosclerotic plaques: potential role of matrix-degrading metalloproteinases and implications for plaque rupture. Circulation. 1995;92: Galis ZS, Sukhova GK, Lark MW, Libby P. Increased expression of matrix metalloproteinases and matrix degrading activity in vulnerable regions of human atherosclerotic plaques. J Clin Invest. 1994;94:

10 Targeted Disruption of the Matrix Metalloproteinase-9 Gene Impairs Smooth Muscle Cell Migration and Geometrical Arterial Remodeling Zorina S. Galis, Chad Johnson, Denis Godin, Richard Magid, J. Michael Shipley, Robert M. Senior and Eugen Ivan Circ Res. published online October 10, 2002; Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX Copyright 2002 American Heart Association, Inc. All rights reserved. Print ISSN: Online ISSN: The online version of this article, along with updated information and services, is located on the World Wide Web at: Data Supplement (unedited) at: Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: Subscriptions: Information about subscribing to Circulation Research is online at:

11 Online supplemental data Materials and methods Morphometric and statistical analysis Apex of the lesion was identified by sectioning the entire length of each ligated carotid artery. Sections were examined with a Zeiss Axioskop microscope equipped with an image capture system. The group containing the 5 consecutive 5 µm sections displaying the smallest lumen was used for the morphometric analysis using Image Pro 3.0 software (Media Cybernetics). The perimeter of the lumen, internal elastic lamina (IEL), and external elastic lamina (EEL) were obtained by measuring the tracing of their contours on digitized images. We calculated the intimal, medial and adventitial thickness as average distances between the lumen and IEL, the IEL and EEL, and the EEL and the border between densely packed and surrounding loose tissue, respectively. Collagen staining and quantification. Five micron paraffin sections were deparrafinized and rehydrated in xylene/ethanol. Following rehydration the sections were stained for collagen with direct red (0.1% in saturated picric acid for 90 minutes; Sigma). The slides were washed twice in 0.01N HCl and then rinsed in dh20. Following the rinse, the slides were stained for elastin using an elastin staining kit (Sigma). The protocol was modified by omitting the Van Gieson's staining steps and proceeding directly to dehydration and coverslipping. The collected white- and polarized-light images of each section stained with Picrosirius red were processed using Microsoft PhotoDraw (version 2.0, Redmond, WA) in order to mask all non-carotid artery tissue. The polarizing filters were adjusted for each section to maximize the birefringence intensity. Using ImagePro 3.0, the red-channel from the polarized-light image was isolated and a threshold analysis was applied. The amount of collagen in a section was calculated as the area that appeared stained under white light and exceeded the threshold under polarized light. Mouse arterial SMC isolation and characterization. Arterial SMC were obtained from outgrowths of aortic explants harvested from MMP-9 KO and WT mice grown in DMEM

12 Ms. 4140/R1 supplemented with 10% fetal bovine serum (FBS, Sigma), 1% L-glutamine (Gibco) and penicillin-streptomycin (Gibco), and used within passages 3 to 8. Cell identity was confirmed by immunocytochemical detection of α-smooth-muscle-cell actin. Purity was >95% for all cultures. Collagen gel contraction assays. For each assay, six 0.5 ml aliquots of a mixture containing 1 mg collagen (rat-tail acid-soluble, Becton Dickinson) per ml DMEM and 250,000 MMP-9 KO or WT SMC/ml, or collagen alone, were allowed to gel at 37 0 C for 24 hours, after which 0.5 ml of DMEM with 10% FBS containing 0.5 µci tritiated water was added to each gel. After an additional 24 hours of contraction, collagen gels were solubilized in 1N NaOH. The capacity of each type of SMC to contract the collagen gel, and expressed as an average percentage decrease in the volume of six identical gels containing SMC compared to collagen gels with no cells added, normalized by the volume of cell-free collagen gels. Three independent experiments were performed. Cell migration assays. SMC were plated onto glass coverslips and allowed to grow to confluence in DMEM supplemented with 10% FBS. The culture medium was replaced with serum free DMEM 24 hours prior to wounding. Half of the confluent SMC monolayer was scraped off using a sterile cell scraper and cells were returned to the incubator for 24 hours, then washed with DMEM, fixed in 10% buffered formalin, and stained rhodamine-conjugated phalloidin (Molecular Probes). Nuclei were counterstained with Hoechst (Sigma Corp.). We calculated the total number of cells and the average distance traveled by cells from the wound edge, determined as the line of cell nuclei debris, clearly visible in the blue channel, in three independent 24 hour migration experiments using a Zeiss Axioscope fluorescent microscope equipped with an image capture system and ImagePro

13 Ms. 4140/R1 Supplemental Figure 7 Figure 7. Gelatinolytic activity secreted by aortic MMP-9 KO or WT mouse SMC (SDS-PAGE zymography) basally or after stimulation in vitro with phorpbol myristate acetate (PMA). Pro- MMP-2 is secreted basally by both WT and MMP-9 KO SMC with no significant differences in MMP-2 levels. No other gelatinases were detected. Stimulation with PMA induces MMP-9 secretion by WT, but not by MMP-9 KO. 3

14 Ms. 4140/R1 Table 2. Summary of morphological measurements WT and MMP-9 KO carotid arteries Measurement Group 0 Days 1 Day 3 Days 7 Days 14 Days 28 Days Intimal Thickness Wild Type 0 (0) 0 (0) 5.33 (0.74) (2.76) 78.2 (34.6) (15.2) MMP-9 KO 0 (0) 0 (0) 3.02 (0.38) 6.13 (1.82) 30.0 (11.7) 52.6 (16.2)* Medial Thickness Wild Type 30 (2) 33 (2) 33 (2) 42 (2) 47 (6) 37 (4) MMP-9 KO 22 (2) 30 (2) 34 (2) 49 (5) 37 (3) 43 (1) Adventitial Thickness Wild Type 0 (0) 0 (0) 10.8 (4.2) 34.1 (10.7) 39.9 (2.7) 20.9 (4.8) MMP-9 KO 0 (0) 0 (0) 7.9 (2.2) 28.1 (16.0) 34.0 (8.6) 35.5 (2.5)* EEL Perimeter Wild Type 1376 (42) 1165 (79) 1311 (72) 1459 (106) 1512 (145) 1385 (58) MMP-9 KO 1234 (60) 1262 (66) 1346 (69) 1212 (102) 1461 (102) 1366 (70) Total Cell # Wild Type n/a MMP-9 KO n/a Total Proliferation % Wild Type n/a 3.36% 7.97% 1.97% 1.85% 1.96% MMP-9 KO n/a 1.33% 4.08% 4.29% 3.43% 1.56% * P < 0.05 between MMP-9 KO and WT average value for the same time point 4