Differential Expression of Matrix Metalloproteinases After Stent Implantation and Balloon Angioplasty in the Hypercholesterolemic Rabbit
Background—Intimal hyperplasia is the principal mechanism of in-stent restenosis. Matrix metalloproteinases (MMPs) play a key role in intimal growth after balloon angioplasty (BA). Little is known, however, about MMP expression after stent implantation (ST). We investigated whether MMP9 and MMP2 are differentially expressed after ST and BA.
Methods and Results—Hypercholesterolemic rabbits underwent ST and BA in the right and left iliac arteries, respectively. The expression of MMPs and their inhibitors (TIMPs) was studied at various time points in the injured arteries by use of zymography, reverse transcription–polymerase chain reaction, and immunohistochemistry. MMP2, but not MMP9, was constitutively expressed in uninjured arteries. MMP9 expression was rapidly induced after injury, whereas the increase in MMP2 expression was delayed. At all time points, pro-MMP9 activity and MMP9 mRNA levels were ≥2-fold (ANOVA, P=0.002) and ≥3-fold (P<0.0001) higher after ST than after BA, respectively. Active MMP9 was detected only after ST. Although the increases in MMP2 mRNA levels were of similar magnitudes after ST and BA, pro-MMP2 activity was slightly higher 7 and 30 days after ST, and MMP2 activity was ≥2-fold higher 7 to 60 days after ST (P=0.002). No difference in TIMP expression was observed between stented and balloon-injured arteries. Cellular distributions of MMPs and TIMP1 were similar after ST and BA. Early inflammatory cell recruitment and 30-day intimal growth were more severe after ST.
Conclusions—Stent implantation results in more intense and sustained expression of MMP9 and activation of MMP2 than balloon angioplasty.
Stent implantation has become the principal revascularization technique used by interventional cardiologists worldwide.1 Although stents reduce restenosis rates in highly selected coronary artery lesions, essentially via an improved immediate luminal gain, in-stent restenosis still remains a frequent and unsolved clinical issue in the “real world.”2 Anatomically, in-stent restenosis results almost exclusively from intimal hyperplasia, whereas constrictive remodeling of the arterial wall is abrogated by the radial force of the metallic stent.3 Experimental4 and clinical5 studies have emphasized that neointimal area is ≈2-fold greater after stent implantation than after balloon angioplasty. The biological basis of in-stent intimal hyperplasia and the extent to which it compares to common—ie, post–balloon angioplasty—intimal hyperplasia, however, are unclear.
Platelet deposition, leukocyte recruitment, and intimal smooth muscle cell (SMC) proliferation are landmarks in the early response of the arterial wall to stent implantation.4 5 6 Intimal cells secrete abundant quantities of extracellular matrix, which composes the bulk of the intimal lesion at later time points.5 Early inhibition of leukocyte recruitment with anti–Mac-1 antibodies6 or interleukin-104 and acceleration of endothelium regrowth7 reverse in-stent intimal hyperplasia in rabbits.
Matrix metalloproteinases (MMPs), and among them MMP9 (gelatinase B) and MMP2 (gelatinase A), play a key role in extracellular matrix degradation, which is required for cell migration into the intima after arterial injury.8 MMPs are therefore considered important contributors to intimal growth and restenosis.9 We hypothesized that MMP expression is greater after stent implantation than after balloon angioplasty and may participate in the more robust intimal hyperplasia observed after stenting.
Models of Intimal Hyperplasia
New Zealand White rabbits weighing 3.6 to 4.0 kg were fed a 0.3% cholesterol diet starting 14 days before angioplasty. Bilateral iliac artery balloon angioplasty was performed with a 3-mm-diameter balloon (3X1-minute inflation, 10 atm); then a 15-mm-long Crown stent (Cordis) mounted over the balloon was implanted in the right iliac artery only (30-second inflation, 10 atm), as described.4 The animal protocol was approved by the Bichat University Institutional Animal Care and Use Committee. Animals were euthanized by pentobarbital overdose at 1 (n=4), 3 (n=4), 7 (n=13), 30 (n=4), and 60 (n=3) days after injury. At each time point, right (stent) and left (balloon angioplasty) iliac arteries were harvested, flushed with ice-cold saline, cleaned of any adipose tissue, and divided into 2 or 3 segments. The proximal and distal segments were used for reverse transcription–polymerase chain reaction (RT-PCR) and gelatin zymography, respectively. In animals killed 1, 7, and 30 days after injury, an additional segment was taken from the center of the injury site and used for morphometric analyses and immunohistochemistry. Unless indicated, stent struts were not removed before analysis so as to prevent the loss of MMP originating from adhering cells. The iliac arteries from 3 sham-operated (no arterial injury), hypercholesterolemic rabbits were used as negative controls. No significant variation in cholesterol levels was observed between baseline (620±150 mg/dL) and death (610±210, 595±150, 605±200, 625±115, and 618±175 mg/dL at 1, 3, 7, 30, and 60 days, respectively).
Organ Culture and SDS-PAGE Zymography
Arterial segments were cut into rings 1 mm long and incubated for 24 hours in 1 mL serum-free DMEM (BioMedia) at 37°C in humidified 5% CO2/95% air. Preliminary experiments demonstrated that organ culture did not induce microscopically detectable tissue injury (data not shown). The conditioned medium was collected and stored at −20°C. Gelatinolytic activities of 20-μL samples of conditioned medium were measured as described.10 Results are expressed in densitometric units/mg wet wt. To verify the metalloproteinase nature of the detected enzymes, identical gels were incubated in the presence of 30 mmol/L EDTA, an inhibitor of MMPs, or 1 mmol/L Pefabloc, a serine protease inhibitor.
Because binding of MMPs to their inhibitors (TIMPs) may have reduced MMP activity in conditioned media, samples were subjected to reduction in 2.5 mmol/L dithiothreitol and alkylation in 2.5 mmol/L iodoacetamide before electrophoresis.11 In addition, reverse zymography was performed to measure TIMP1 and TIMP2 inhibitory activities in conditioned media, as described.12
Densitometric analyses of scanned gelatinolytic bands was performed with NIH Image 1.55 software.
Reverse Transcription–Polymerase Chain Reaction
Arterial segments were immediately frozen in liquid nitrogen and stored at −80°C. Procedures used for total RNA extraction and RT-PCR have been described.12 All RNA samples were run in parallel during the same PCR reaction, allowing for head-to-head comparisons between stented and balloon-injured arteries. The primers used to measure rabbit MMP9, MMP2, TIMP1, TIMP2, and GAPDH mRNA levels are listed in the Table⇓. Results are expressed in arbitrary units and adjusted for GAPDH mRNA levels.
Morphometry and Immunohistochemistry
Stented and nonstented arterial segments were fixed in 4% paraformaldehyde and processed as described.4 Morphometric analyses were performed on 3 hematoxylin-phloxin-safran–stained cross sections for each artery.4
Arterial specimens retrieved 7 days after injury were used for immunohistochemistry. Stent struts were gently removed with microforceps. Then arterial segments were embedded in OCT compound, frozen in liquid nitrogen–chilled isopentane, and stored at −80°C. Four-micrometer cross sections were obtained from each block and immunostained with mouse monoclonal antibodies (1:50 dilution, all from Oncogene unless indicated) directed against (1) MMP9 (Ab-1); (2) MMP2 (Ab-3); (3) TIMP1 (Ab-1); (4) TIMP2 (Ab-2); (5) RAM-11, a marker of rabbit macrophage cytoplasm (Dako); and (6) smooth muscle α-actin (HHF-35, Enzo Diagnostics), as described.4 For negative control experiments, primary antibodies were omitted.
Data are expressed as mean±SD. MMP/TIMP activities and mRNA levels in stented versus balloon-injured arteries were compared by 2-way ANOVA (Statview 5.0, SAS Institute Inc), which tested the effect of injury type, time, and the interaction of the two. When ANOVA indicated an overall difference between stented and balloon-injured arteries, specific differences at each time point (intergroup differences) and temporal differences within each injury group were tested by 1-way ANOVA and post hoc Fisher’s protected least significant difference test. Morphometric data in stented versus balloon-injured arteries were compared by a nonparametric Mann-Whitney U test. A value of P<0.05 was considered significant.
Intense pro-MMP9 activity (98 kDa) was detected as soon as 1 day after arterial injury. At all time points, pro-MMP9 activity was ≥2-fold higher in stented arteries than in balloon-injured arteries (stent versus balloon, P=0.002; time effect, P<0.0001; interaction, P=0.17). Sixty days after stent implantation, pro-MMP9 activity was still ≈20-fold greater than control levels, whereas it was not different from control levels in balloon-injured arteries.
In 4 of 13 stented arteries studied 7 days after injury (but in none at other time points), a faint gelatinolytic band was observed at 80 kDa, which presumably represents the active form of MMP9.13 In no case was the active form of MMP9 detected in balloon-injured arteries.
There was no significant difference in pro-MMP2 activity between balloon-injured, stented, and uninjured arteries until day 7. At this time, pro-MMP2 activity increased in injured arteries, with slightly higher activities measured in stented arteries 7 and 30 days after injury (stent versus balloon, P=0.006; time effect, P<0.0001; interaction, P=0.23).
An additional gelatinolytic band at 60 kDa, representing the active form of MMP2,13 was observed 3 days after arterial injury. MMP2 activity peaked at day 30, then rapidly decreased. Seven to 60 days after arterial injury, MMP2 activity was ≥2-fold higher in stented than in balloon-injured arteries (stent versus balloon, P=0.002; time effect, P<0.0001; interaction, P=0.02). Sixty days after injury, MMP2 activity was still detectable in stented arteries but not in balloon-injured arteries.
All gelatinolytic activities were inhibited by EDTA but not by Pefabloc, suggesting that MMPs accounted for these activities (data not shown). Reduction-alkylation before gelatin zymography resulted in an ≈20% increase in pro-MMP9 activity, suggesting that part of pro-MMP9 was bound to its specific TIMP. This increase in pro-MMP9 activity, however, was similar in stented and balloon-injured arteries (data not shown). Reduction-alkylation had no effect on pro-MMP2 and MMP2 activities.
Background MMP9 mRNA levels were measured in uninjured arteries. One day after injury, MMP9 mRNA levels increased dramatically, remained stable until day 30, then decreased (Figure 2⇓). At each time point, MMP9 mRNA levels were ≥3-fold higher in stented arteries than in balloon-injured arteries (stent versus balloon, P<0.0001; time effect, P=0.0003; interaction, P=0.2). Sixty days after injury, high MMP9 mRNA levels were still present in stented arteries (≈5-fold greater than baseline) but not in balloon-injured arteries.
Constitutive MMP2 expression was found in uninjured arteries. MMP2 mRNA levels decreased slightly 1 day after injury, then increased significantly over time. Although MMP2 mRNA levels were slightly higher in stented arteries than in balloon-injured arteries between day 3 and day 30, no significant difference was observed (stent versus balloon, P=0.3; time effect, P<0.0001; interaction, P=0.9).
TIMP Inhibitory Activity and Expression
Both reverse zymography and RT-PCR demonstrated constitutive expressions of TIMP1 and TIMP2 (Figure 3⇓). There was a nonsignificant increase of TIMP1 inhibitory activity after injury, with no difference between stented and balloon-injured arteries (stent versus balloon, P=0.3; time effect, P=0.06; interaction, P=0.2). TIMP2 activity decreased significantly 1 day after injury and returned to baseline levels 30 days after injury. No difference in TIMP2 activity was observed between stented and balloon-injured arteries (stent versus balloon, P=0.3; time effect, P=0.002; interaction, P=0.8).
After arterial injury, both TIMP1 and TIMP2 mRNA levels increased transiently, with no difference between stented and balloon-injured arteries (TIMP1: stent versus balloon, P=0.8; time effect, P=0.004; interaction, P=0.9; TIMP2: stent versus balloon, P=0.8; time effect, P<0.0001; interaction, P=0.7).
Morphometry and Immunohistochemistry
The anatomy of the arterial wall appeared normal at baseline, ie, after 14 days on the hyperlipidemic diet (data not shown). An intense recruitment of inflammatory cells with morphological features of polymorphonuclear leukocytes (PMNs) or monocytes was observed 1 day after injury in both stented and balloon-injured arteries (Figure 4⇓). The number of adherent PMNs and monocytes was significantly higher after stenting (74.5±21.1 versus 21±8.2 per arterial section, P<0.0001). Thirty days after injury, a circumferential neointima was present in both stented and balloon-injured arteries, with a larger intimal area (2.4±0.5 versus 1.5±0.1 mm2, P=0.001) and intima/media ratio (4.2±0.8 versus 1.8±0.4, P<0.0001) in stented arteries. Typically, the intima of stented arteries was composed of a thin, fibromuscular layer covering a large, foam cell–rich outermost layer. No such compartmentalization of the intima was observed in balloon-injured arteries.
Immunostainings of adjacent cryosections revealed the constitutive expression of MMP2 and TIMP1 in medial SMCs of uninjured arteries (data not shown). MMP9 was found only in injured arteries and colocalized with intimal macrophages, intimal/medial SMCs, and adventitial fibroblasts (Figure 5⇓). MMP2 was detected in intimal/medial SMCs and in adventitial fibroblasts. TIMP1 immunostaining was present in intimal/medial SMCs. TIMP2 was not detectable by immunohistochemistry. The cellular distribution of MMP9, MMP2, and TIMP1 was similar in stented and balloon-injured arteries.
In the present study, we compared the impact of stent implantation and balloon angioplasty on the expression and activity of MMP9 and MMP2 in the iliac arteries of hypercholesterolemic rabbits. MMP9 expression was rapidly induced after arterial injury but more intensely and more durably after stent implantation. Stented arteries secreted higher levels of pro-MMP9 for ≥60 days after injury. In addition, active MMP9 was detected only in stented arteries. In contrast, the increase in constitutive MMP2 expression was delayed and of similar magnitude in stented and balloon-injured arteries. The activation and, to a lesser extent, the production of pro-MMP2, however, were more pronounced in stented arteries. The expression of TIMPs did not differ between stented and balloon-injured arteries. The net increased proteolytic activity observed in stented arteries was associated with a more severe recruitment of inflammatory cells at day 1 and an enhanced neointimal growth at day 30 after arterial injury.
Previous studies performed in mice,14 rats,15 16 17 18 19 and rabbits20 21 22 have demonstrated that arterial injury is a potent activator of the MMP system. In the normocholesterolemic rabbit, balloon angioplasty of the iliac artery upregulates the transcription of the MMP2 gene and stimulates the transformation of pro-MMP2 into active MMP2. MMP9 activity, however, remains undetectable.21 23 In fact, in this model, only the combination of balloon injury and low flow results in detectable pro-MMP9 and MMP9 activities.21 In contrast, balloon injury combined with sustained atherogenic diet results in substantial pro-MMP9 activity, in addition to high MMP2 activity in both its pro-MMP2 and activated forms.22
Our study brings new insight to the field of MMP response to arterial injury, in that we have included metallic stents, the principal technique used for percutaneous coronary interventions,1 in the design of our model. The main result of the present study is that overall, stents are more potent stimuli of MMP expression than balloon angioplasty. Given the key role of MMPs in cell migration and intimal growth15 and the recent findings that in-stent restenosis results almost exclusively from intimal hyperplasia,3 it is likely that MMPs are important contributors to in-stent restenosis. Indeed, we found that intimal area was ≈1.6-fold larger in stented arteries than in balloon-injured arteries.
Whether the more severe increase in MMP expression observed after stenting results from the recruitment of a larger population of MMP-expressing inflammatory cells or from a more pronounced upregulation of MMP expression in SMCs is difficult to determine. We were not able to fully investigate the cellular distribution of MMPs in stented arteries because, on the one hand, methyl methacrylate–embedded tissues do not lend themselves to standard immunohistochemistry protocols,24 and on the other hand, immunostaining of arterial cryosections requires that stent struts be retrieved before sectioning. This resulted in severe tissue damage (data not shown), precluding quantitative analysis of immunostained areas. Hence, no attempt was made to measure the specific contribution of each cell type to MMP9 and MMP2 activities. It is likely, however, that the more massive recruitment of PMNs observed 1 day after stent implantation participates in the early increase in MMP9 expression in stented arteries.25 Alternatively, MMP9-immunopositive SMCs were present in the media 7 days after arterial injury, whereas no MMP9 staining was detectable in uninjured arteries. Hence, MMP9 gene expression is induced in SMCs after arterial injury and may contribute to the higher MMP9 expression found in stented arteries.
We also provide evidence that MMP activation occurs after stenting. Activation of latent MMP9 was found only in stented arteries. However, it was inconstant (4 of 13 arteries), transient (≤7 days), and hardly quantifiable by zymography. In contrast, active MMP2 was detectable in both stented and balloon-injured arteries, but with much higher levels in stented arteries 7 to 60 days after injury.
Several mechanisms may be involved in the differential expression/activation patterns of MMPs in stented versus balloon-injured arteries. MMPs are regulated at 3 different levels: gene transcription, activation of latent pro-MMPs, and inactivation of MMPs by TIMPs.26 Proinflammatory cytokines—eg, interleukin-1—are key contributors to in-stent intimal hyperplasia4 and may be responsible, at least in part, for the induction of MMP expression in stented arteries.26 Recent studies suggest that reactive oxygen species27 or cell-cell28 or cell-matrix29 interactions may play a role as well.
Using both reduction-alkylation studies and direct measurements of TIMP expression, we provide evidence that the higher levels of MMP activities in stented arteries do not result from lower levels of TIMPs. Rather, the role of soluble activators of MMP9, such as plasmin,14 and the expression of MT1-MMP, a potent activator of pro-MMP2,30 deserve to be investigated.
Our study may have important implications for the understanding of the pathophysiology of in-stent restenosis and hence, the prevention of this phenomenon. Several authors have investigated the efficacy of MMP inhibitors on intimal growth after balloon injury. In rodents, synthetic MMP inhibitors only partially inhibited neointima formation,9 17 whereas overexpression of TIMPs, either via gene transfer31 or in genetically engineered mice,32 was more protective. N-Acetylcysteine has potent inhibitory effects on MMP9 in foam cells27 and deserves to be tested in vivo. On the basis of our findings that upregulation of MMP expression and MMP activation occur at much higher levels after stent implantation than after balloon angioplasty, the impact of MMP inhibitors on restenosis should be considered in stent models.
Dr Feldman was supported by grants from the Fondation de l’Avenir (ET6-167) and the Fondation de France (97003880 and 98004139). Dr Scheuble was supported by a grant from the Fédération Française de Cardiologie. Dr De Benedetti was supported by a grant from the Fonds National Suisse de la Recherche Scientifique. We are grateful to Cordis France for providing the Crown stents. We wish to thank Isabelle Prevost, Michelle Saadoun, Elisabeth Soustre, Laurence Le Men, and Graziella Ferrec for expert technical assistance.
- Received September 20, 2000.
- Revision received February 7, 2001.
- Accepted February 21, 2001.
- Copyright © 2001 by American Heart Association
Topol EJ, Serruys PW. Frontiers in interventional cardiology. Circulation. 1998;98:1802–1820.
Mehran R, Dangas G, Abizaid AS, et al. Angiographic patterns of in-stent restenosis: classification and implications for long-term outcome. Circulation. 1999;100:1872–1878.
Hoffmann R, Mintz GS, Dussaillant GR, et al. Patterns and mechanisms of in-stent restenosis: a serial intravascular ultrasound study. Circulation. 1996;94:1247–1254.
Feldman LJ, Aguirre L, Ziol M, et al. Interleukin-10 inhibits intimal hyperplasia after angioplasty or stent implantation in hypercholesterolemic rabbits. Circulation. 2000;101:908–916.
Farb A, Sangiorgi G, Carter AJ, et al. Pathology of acute and chronic coronary stenting in humans. Circulation. 1999;99:44–52.
Rogers C, Edelman ER, Simon DI. A mAb to the beta2-leukocyte integrin Mac-1 (CD11b/CD18) reduces intimal thickening after angioplasty or stent implantation in rabbits. Proc Natl Acad Sci U S A. 1998;95:10134–10139.
Dollery CM, McEwan JR, Henney AM. Matrix metalloproteinases and cardiovascular disease. Circ Res. 1995;77:863–868.
Bendeck MP, Irvin C, Reidy MA. Inhibition of matrix metalloproteinase activity inhibits smooth muscle migration but not neointimal thickening after arterial injury. Circ Res. 1996;78:38–43.
Gonzalez W, Fontaine V, Pueyo ME, et al. Molecular plasticity of vascular wall during Nω-nitro-l-arginine methyl ester–induced hypertension: modulation of proinflammatory signals. Hypertension. 2000;36:103–109.
Southgate KM, Davies M, Booth RF, et al. Involvement of extracellular-matrix-degrading metalloproteinases in rabbit aortic smooth-muscle cell proliferation. Biochem J. 1992;288:93–99.
Lijnen HR, Van Hoef B, Lupu F, et al. Function of the plasminogen/plasmin and matrix metalloproteinase systems after vascular injury in mice with targeted inactivation of fibrinolytic system genes. Arterioscler Thromb Vasc Biol. 1998;18:1035–1045.
Bendeck MP, Zempo N, Clowes AW, et al. Smooth muscle cell migration and matrix metalloproteinase expression after arterial injury in the rat. Circ Res. 1994;75:539–545.
Zempo N, Koyama N, Kenagy RD, et al. Regulation of vascular smooth muscle cell migration and proliferation in vitro and in injured rat arteries by a synthetic matrix metalloproteinase inhibitor. Arterioscler Thromb Vasc Biol. 1996;16:28–33.
Webb KE, Henney AM, Anglin S, et al. Expression of matrix metalloproteinases and their inhibitor TIMP-1 in the rat carotid artery after balloon injury. Arterioscler Thromb Vasc Biol. 1997;17:1837–1844.
Jenkins GM, Crow MT, Bilato C, et al. Increased expression of membrane-type matrix metalloproteinase and preferential localization of matrix metalloproteinase-2 to the neointima of balloon-injured rat carotid arteries. Circulation. 1998;97:82–90.
Strauss BH, Robinson R, Batchelor WB, et al. In vivo collagen turnover following experimental balloon angioplasty injury and the role of matrix metalloproteinases. Circ Res. 1996;79:541–550.
Bassiouny HS, Song RH, Hong XF, et al. Flow regulation of 72-kD collagenase IV (MMP-2) after experimental arterial injury. Circulation. 1998;98:157–163.
Aikawa M, Rabkin E, Okada Y, et al. Lipid lowering by diet reduces matrix metalloproteinase activity and increases collagen content of rabbit atheroma: a potential mechanism of lesion stabilization. Circulation. 1998;97:2433–2444.
Strauss BH, Chisholm RJ, Keely FW, et al. Extracellular matrix remodeling after balloon angioplasty injury in a rabbit model of restenosis. Circ Res. 1994;75:650–658.
Nagase H, Woessner JF Jr. Matrix metalloproteinases. J Biol Chem. 1999;274:21491–21494.
Galis ZS, Asanuma K, Godin D, et al. N-Acetyl-cysteine decreases the matrix-degrading capacity of macrophage-derived foam cells: new target for antioxidant therapy? Circulation. 1998;97:2445–2453.
Xie B, Laouar A, Huberman E. Fibronectin-mediated cell adhesion is required for induction of 92-kDa type IV collagenase/gelatinase (MMP-9) gene expression during macrophage differentiation: the signaling role of protein kinase C-beta. J Biol Chem. 1998;273:11576–11582.
Dollery CM, Humphries SE, McClelland A, et al. Expression of tissue inhibitor of matrix metalloproteinases 1 by use of an adenoviral vector inhibits smooth muscle cell migration and reduces neointimal hyperplasia in the rat model of vascular balloon injury. Circulation. 1999;99:3199–3205.
Lijnen HR, Soloway P, Collen D. Tissue inhibitor of matrix metalloproteinases-1 impairs arterial neointima formation after vascular injury in mice. Circ Res. 1999;85:1186–1191.