Jump to
Abstract
Extracellular matrix formation is the major component of the restenosis lesion that develops after balloon angioplasty. Although ex vivo studies have shown that the synthesis of collagen is stimulated early after balloon angioplasty, there is a delay in accumulation in the vessel wall. The objectives of this study were to assess collagen turnover and its possible regulation by matrix metalloproteinases (MMPs) in a double-injury iliac artery rabbit model of restenosis. Rabbits were killed at four time points (immediately and at 1, 4, and 12 weeks) after balloon angioplasty. In vivo collagen synthesis and collagen degradation were measured after a 24-hour incubation with [14C]proline. Arterial extracts were also run on gelatin zymograms to determine MMP (gelatinase) activity. Collagen turnover studies were repeated in a group of 1-week postangioplasty rabbits that were treated with daily subcutaneous injections of either a nonspecific MMP inhibitor, GM6001 (100 mg/kg per day), or placebo. Collagen synthesis and degradation showed similar temporal profiles, with significant increases in the balloon-injured iliac arteries compared with control nondilated contralateral iliac arteries immediately after angioplasty and at 1 and 4 weeks. Peak collagen synthesis and degradation occurred at 1 week and were increased (approximately four and three times control values, respectively). Gelatin zymography was consistent with the biochemical data by showing an increase of a 72-kD gelatinase (MMP-2) in the balloon-injured side immediately after the second injury, peaking at 1 week, and still detectable at 4 and 12 weeks (although at lower levels). In balloon-injured arteries, the MMP inhibitor reduced both collagen synthesis and degradation. Overall, at 1 week after balloon angioplasty, GM6001 resulted in a 33% reduction in collagen content in balloon-injured arteries compared with placebo (750±143 to 500±78 μg hydroxyproline per segment, P<.004), which was associated with a nonsignificant 25% reduction in intimal area. Our data suggest that degradation of newly synthesized collagen is an important mechanism regulating collagen accumulation and that MMPs have an integral role in collagen turnover after balloon angioplasty.
Extracellular matrix formation is a key aspect of the intimal lesion that forms after balloon angioplasty, ultimately contributing to ≈90% of the volume of the lesion.1 Collagen is the principal extracellular matrix component, but mechanisms that regulate its accumulation in the vascular lesion that forms after balloon angioplasty remain poorly understood. Previous work in rabbit arterial injury models by our group2 and others3 have shown that although collagen synthesis is stimulated early after balloon angioplasty, its accumulation in the extracellular space is temporally delayed relative to its increased synthesis. In ex vivo studies, we found that collagen synthesis was low and comparable to the control undamaged contralateral vessel immediately after the angioplasty injury but was stimulated at 1, 2, and 4 weeks and then returned to baseline levels by 12 weeks.2 Despite marked increases in synthesis rates (approximately fourfold) at weeks 1 and 2 after balloon angioplasty, collagen content was essentially unchanged up to 4 weeks but increased by ≈33% between 4 and 12 weeks. Although the overall effect of balloon angioplasty was net accumulation of collagen, these temporal discrepancies between increases in synthesis and content of collagen were unexpected, and we hypothesized that these differences could be attributed to alterations in degradation rates of newly synthesized proteins.
Collagens are major structural components of the extracellular matrix of the arterial wall, composing 20% to 50% of the dry weight,4 5 with predominance of types I and III (and minor amounts of IV, V, and VI) in the fibrous stroma of atherosclerotic plaques.6 7 There is now ample evidence in the literature, in a variety of tissues, that measured collagen synthesis levels may not be directly related to procollagen mRNA levels8 or to changes in collagen content.2 8 9 Procollagen production appears to be regulated at both transcriptional and posttranscriptional levels, and there is extensive posttranslational processing that occurs at intracellular and extracellular sites.10 11 Although there are reports of in vitro and in vivo studies of collagen turnover in several nonvascular tissues8 9 10 and in pulmonary hypertension,12 no such studies have been undertaken in the arterial wall after balloon injury. Degradation, in particular, deserves further study, since degradation of newly synthesized procollagens occurs significantly (10% to 90% in various cell culture systems and in vivo) before or soon after procollagen secretion.10 13 14
MMPs are a group of zinc enzymes that are responsible for degradation of extracellular matrix components, including fibronectin, collagen, elastin, proteoglycans, and laminin in normal embryogenesis, inflammation, wound healing, and tumor invasion.15 16
Emerging data on the role of MMPs in vascular disease such as smooth muscle cell migration17 18 and plaque rupture19 20 have suggested that these enzymes could also be important in regulating collagen turnover and net collagen accumulation. The extracellular pathways of interstitial collagen degradation are mediated initially by either interstitial collagenase (MMP-1)21 or MMP-222 (a gelatinase), with cleavage at a specific site in the collagen molecule rendering it susceptible to other neutral proteases (eg, gelatinases) in the extracellular space. There is little information on extracellular degradation and MMP activation in vivo.
The objectives of the present study were to assess collagen turnover and its possible regulation by MMPs in a rabbit model of restenosis. At specific times after iliac artery balloon angioplasty, we measured newly synthesized collagen and its degradation, using an adaptation of the proline-flooding assay.23 The presence and/or activity of collagenase and gelatinases was also assessed at these time periods. Finally, the effect of inhibition of MMP activity on collagen turnover was assessed by repeating the collagen turnover studies 1 week after balloon angioplasty in a group of animals treated with a nonspecific MMP inhibitor, GM6001. Evidence is presented in this report indicating that alterations of collagen degradation occur to a significant extent after balloon injury in a temporal pattern of activation similar to that of collagen synthesis. MMP activity, specifically MMP-2, is also stimulated after balloon angioplasty, and inhibition of MMP activity has profound effects on both collagen synthesis and degradation, resulting in diminished net collagen accumulation. These data suggest that collagen turnover, in part mediated by MMPs, is an important mechanism regulating vascular remodeling and collagen accumulation after balloon injury.
Materials and Methods
The Model
The animal experiments were performed in accordance with guidelines set out by the University of Toronto and approved by the St. Michael's Hospital animal care committee. We used a double-balloon normolipemic injury model using male New Zealand White rabbits weighing 3.6 to 3.8 kg as previously described.2 In this model, a fibrocellular intimal thickening is formed in response to the first balloon injury and serves as the “pathological substrate” for the second balloon angioplasty injury. Since this model has many similarities to fibrous (but not lipid-rich) coronary lesions, we feel it is a more appropriate model for the study of vessel wall extracellular matrix changes in response to balloon injury than are normal arteries that have not been previously damaged. Xylazine and ketamine were administered as an intramuscular injection at a dose of 0.3 mL/kg with additional injections (0.2 mL) given as required every 20 to 30 minutes. The animals underwent surgical cutdown, after which a 4F sheath (Cordis Corp) was placed in the right carotid artery. Heparin sulfate (1000 IU) and acetylsalicylic acid (50 mg) were injected intra-arterially to prevent early vessel occlusion. A 3F angioplasty balloon catheter with a balloon length of 40 mm was then passed under fluoroscopic guidance using an over-the-wire system into the right iliac artery. The initial 4 cm of the right iliac vessel immediately distal to the aortic bifurcation (and therefore a reliable and reproducible landmark for removal of the vessels at a later date) underwent dilation, whereas the left iliac artery served as an internal control. The injury consisted of four 1-minute inflations of the 3.0-mm-diameter balloon catheter serially inflated to 6, 8, 4, and 10 atm with 45 seconds between inflations. Gentle traction of the balloon catheter at 4 atm was performed to ensure complete endothelial denudation. At the end of the procedure, angiographic patency was confirmed, the balloon catheter was withdrawn, and the arteriotomy was repaired. After recovery, the rabbits were fed regular rabbit chow and water ad libitum. Angioplasty of the right iliac artery was repeated 3 weeks later through a left carotid arteriotomy using a similar inflation protocol (but without balloon traction). Animals were killed at the following time periods after the second injury: immediately, after 1 week, after 4 weeks, and after 12 weeks. At the time of death, the iliac arteries were isolated and removed under general anesthetic followed by a fatal intracardiac injection of thiopental.
Collagen Turnover Experiments
A modification of the proline flooding infusion technique23 was used to measure the synthesis and degradation of new collagen. Our measurements of intact collagen synthesis and collagen degradation were based on the following principles: Since the molecular weights of intact collagen and collagen degradation products are ≈100 and <30 kD, respectively, we used a 30-kD millipore filter to separate newly synthesized intact collagen from its degradation products. On the basis of these differences in molecular mass, the intact newly synthesized collagen would be present in the >30-kD fraction, and degraded collagen fragments would be found in the <30-kD fraction. The adequacy of separation of the proteins by molecular mass is shown in Fig 1⇓. The presence of collagen types I and III was confirmed by Western blot of the >30-kD fraction using unlabeled affinity-purified goat antibodies directed against the tertiary structure of types I and III collagen (Southern Biotechnology Assoc) (Fig 2⇓).
A, Filtrate (<30 kD) from rabbit arterial extract that was electrophoresed on 15% SDS–polyacrylamide gel and then stained with Coomassie brilliant blue. No proteins >30 kD were detected on this gel. Lanes are as follows: 1, molecular mass markers; 2, control (undilated) iliac artery; and 3, balloon-injured iliac artery. B, Rabbit arterial extract that was retained on 30-kD filter and subsequently separated on 10% SDS–polyacrylamide gel. No proteins <30 kD were detected on staining with Coomassie brilliant blue. Lanes are as follows: 1, molecular mass markers; 2, control (undilated) iliac artery; and 3, balloon-injured iliac artery.
Western blot of arterial extract retained on 30-kD filter. The presence of intact collagen type I (A) and collagen type III (B) was indicated by the presence of bands above 112 kD. This arterial extract was obtained from an animal immediately after angioplasty. Lanes are as follows: far right, molecular mass marker; middle, control (undilated) artery; and far left, balloon-injured artery.
Hydroxylation of proline residues to hydroxyproline in collagen is a posttranslational event. Because elastin, the only other hydroxyproline-containing protein in the vessel wall, accounts for only a small proportion of this hydroxyproline, the appearance of [14C]hydroxyproline in collagen-containing extracts has been used as a specific measure of the synthesis of this protein. Moreover, because conversion of [14C]proline to [14C]hydroxyproline takes place posttranslationally, fragments of collagen containing [14C]hydroxyproline must represent collagen that has been synthesized during the labeling period and then undergone a degradation process.
Collagen also contains a much higher proportion of proline than noncollagen proteins. The ratio of proline residues in collagen relative to noncollagen protein is 5.4 to 1.24 This, together with the fact that collagen is the major extracellular protein in the vessel wall, means that determinations of total [14C]proline incorporation ([14C]proline plus [14C]hydroxyproline) into extracts containing collagen are also a useful method for estimating collagen synthesis and degradation. When total activity (ie, [14C]proline plus [14C]hydroxyproline) was compared with [14C]hydroxyproline activity in the <30-kD fraction, [14C]hydroxyproline accounted for 50% to 60% of counts, even though the oxidation step used in the hydroxyproline assay effectively removed one carbon from the structure (and thus reduced the counts by 1/6). Thus total [14C]proline incorporation, while less specific than assays of [14C]hydroxyproline alone, had the advantages of increased sensitivity, simplicity, and convenience. Because of lower counts in the degradation (<30-kD) fraction in the temporal collagen turnover studies (see “Results”), we measured total [14C]proline incorporation in the <30-kD fraction to enhance our sensitivity to detect collagen degradation. The validity of these assumptions in determining collagen degradation was subsequently confirmed using the more specific [14C]hydroxyproline assay for newly synthesized collagen degradation products in the <30-kD fraction in the placebo- and GM6001-treated rabbits killed at 1 week after balloon angioplasty (see below).
In these collagen turnover studies, known amounts of [14C]proline (10 μCi/100 g body wt, Amersham) were infused for 10 minutes via an ear vein. Excess unlabeled proline (800 mg/kg) was infused at the same time as the tracer to ensure that saturated pools of proline were available for collagen synthesis. The animal was killed 24 hours after proline infusion with a fatal intracardiac injection of thiopental. The left and right iliac arteries were isolated and flushed with 5 mL of saline injected in the aorta just proximal to the bifurcation. The vessels were stripped of the surrounding tissue, and a 3.5-cm length was marked. After removal, the arteries were weighed, cut into 3- to 5-mm pieces, and extracted with 1 mL of 0.5 mol/L acetic acid for 24 hours at room temperature. The extract was centrifuged for 5 minutes at 4000g, and the supernatant was filtered through a 30-kD filter (Millipore) by centrifugation at 4000g. The residual tissue was homogenized and reextracted with 2 mL of 0.5 mol/L acetic acid at room temperature for 24 hours, and again the supernatant was filtered through a 30-kD filter. To determine intact collagen synthesis, a [14C]hydroxyproline assay was performed in the >30-kD fraction. The residue in this fraction was extracted with 1 mL of 0.5 mol/L acetic acid for 24 hours at room temperature. The extract was dried and hydrolyzed with 6N HCl at 110°C for 24 hours. The hydrolysate was dried and assayed for [14C]hydroxyproline using a modification of the method of Blumenkrantz and Asboe-Hansen,25 as a measure of collagen synthesis and expressed as disintegrations per minute per arterial segment.
To determine degradation of newly synthesized collagen, the first and second <30-kD filtrates were combined and dried, and total [14C]proline incorporation ([14C]proline plus [14C]hydroxyproline) was assessed by scintillation counting (see above) and expressed as disintegrations per minute per arterial segment. The number of animals killed at each time point were as follows: immediately, n=6; after 1 week, n=7; after 4 weeks, n=7; and after 12 weeks, n=6.
MMP Studies
Western Blot Analysis
The frozen tissues were pulverized under liquid nitrogen with a mortar and pestle. The fine powder was suspended in a cell lysis buffer containing 1% SDS, 1 mmol/L phenylmethylsulfonyl fluoride, and 10 mg/mL leupeptin in 50 mmol/L Tris-HCl (pH 7.6). The homogenate was centrifuged at 13 500g. An aliquot of the clear supernatant was estimated for protein content. Aliquots of the supernatant were diluted with electrophoresis sample buffer (0.5 mol/L Tris-HCl [pH 6.6], 3% SDS, 10% glycerol, 5% β-mercaptoethanol, and 0.2% bromophenol blue), boiled for 5 minutes, and electrophoresed on a 7.5% SDS-polyacrylamide gel. The resolved proteins were transferred electrophoretically (35 V overnight, 4°C) onto polyvinylidene difluoride sequencing membrane (0.2 μm, Bio-Rad). After transfer, the membrane was washed briefly with TBS (20 mmol/L Tris-HCl and 500 mmol/L NaCl, pH 7.5) and then blocked in TBS buffer containing 5% skim milk powder. The blot was then briefly rinsed with TTBS buffer (20 mmol/L Tris-HCl [pH 7.5], 500 mmol/L NaCl, and 0.05% Tween-20) and incubated in a solution of primary antibody diluted (1:1000 MMP-1, 1:400 MMP-2) in TTBS buffer for 18 hours at room temperature. In the MMP-1 studies, the blot was briefly rinsed in TTBS and incubated at room temperature for 6 hours in a solution of secondary antibody (alkaline phosphatase–conjugated donkey anti-sheep) diluted (1:2500) in TTBS buffer. The blot was washed twice in TTBS buffer, followed by a brief wash in TBS. The antigen-antibody complex was visualized using color development method (Bio-Rad immunoblot assay kit). In the MMP-2 study, after washing with TTBS buffer, the membrane was incubated with a 1:3000 dilution of alkaline phosphatase–conjugated goat anti-rabbit antiserum (Bio-Rad) in TTBS buffer containing 1% skim milk powder at room temperature for 2 to 4 hours. The blot was developed using a substrate solution for alkaline phosphatase, which consisted of nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate, according to the manufacturer's protocol (Bio-Rad).
The primary antisera used for immunoblotting of MMP-1 was a polyclonal antibody raised in sheep and directed against rabbit procollagenase26 27 (a generous gift from Dr C.E. Brinckerhoff, Dartmouth Medical School, Hanover, NH). The MMP-2 antiserum was a polyclonal antibody raised in rabbits and directed against human 72-kD type IV collagenase (Ab 45, a generous gift from Dr W.G. Stetler-Stevenson, National Institutes of Health, National Cancer Institute, Bethesda, Md). The MMP-1 Western blot analysis was carried out for all time periods in duplicate (total of 8 rabbits). The MMP-2 analysis was done in duplicate from 1-week postangioplasty specimens.
Gelatinase Zymography
MMP activity was measured by zymography according to the method of Cruz et al.28 After harvesting the iliac artery, the tissue was homogenized in 5 vol PBS, centrifuged, and frozen at −80°C. Aliquots of homogenized tissue were subsequently diluted in Laemmli buffer,29 and equal amounts of total protein from each arterial extract were loaded onto a 10% SDS–polyacrylamide gel and electrophoresed at a constant voltage of 200 V. Gelatinase activity (MMP-2 and 9) were determined using 0.1% gelatin substrates (Fisher Scientific), which were incorporated into the gels. After electrophoresis, the gels were washed in 2.5% Triton X-100, incubated overnight at 37°C in gel incubation buffer (50 mmol/L Tris and 15 mmol/L CaCl), and then stained with 0.5% Coomassie brilliant blue R-250 (Bio-Rad). After destaining with water for 4 hours, areas of clearing in the gels indicated the presence of enzymatic activity. Gels were calibrated with high and low molecular weight standards (Bio-Rad). Medium from chondrocyte cultures served as positive controls. To confirm MMP activity, positive samples were run on zymograms also incubated in the presence of 20 mmol/L EDTA (BDH) and 20 mmol/L 1,10-phenanthroline (Fluka Chemika), which are inhibitors of metalloproteinase activity. These MMP studies were carried out in 24 rabbits (6 at each time period).
MMP Inhibitor Studies
A nonspecific inhibitor, GM6001 (Galardin), was generously supplied by Glycomed Inc. The hydroxamic group of GM6001 binds to the critical active-site zinc atom present in the MMPs. The isobutyl group and tryptophan side chain of GM6001 may also bind to subsites on MMPs, which normally bind side chains of extracellular matrix proteins.30 To test the efficiency of GM6001, we incubated zymogram gels in GM6001, ranging from 0.5 mmol/L to 50 nmol/L, and this completely abolished gelatinolytic activity. We administered the same dose of GM6001 as previously used by Bendeck et al17 (100 mg/kg), which was prepared daily in 2 mL PBS. Control rabbits received daily subcutaneous injections of PBS. The suspension was injected daily for 9 days by a subcutaneous route beginning 2 days before the second injury until the animals were killed (1 week after the second injury). This particular time period was studied since it corresponded to the time of maximal collagen synthesis and degradation in our model (see “Results”).
Collagen Turnover Studies
In these studies, [14C]hydroxyproline was determined separately in the collagen synthesis (>30 kD) and the collagen degradation (<30 kD) fractions (see above). Residues in both fractions were hydrolyzed at 110°C for 24 hours in 6N HCl. After evaporation of HCl, the residue was assayed for [14C]hydroxyproline and was expressed as disintegrations per minute per arterial segment. In addition, total hydroxyproline content, a measure of total collagen, was determined in the >30-kD fraction using a colorimetric assay.31 These studies were performed in 14 rabbits (8 treated with placebo and 6 treated with GM6001).
Intimal Areas, Cell Density, and Proliferation Studies
In a separate group of 13 (7 placebo-treated and 6 GM6001-treated) rabbits, morphometry studies were carried in the injured iliac arteries using a GM6001 study protocol identical to that described above. Because of technical problems in processing, one placebo artery could not be analyzed. To assess for cell proliferation, BrdU (Sigma Chemical Co), a thymidine analogue, was injected subcutaneously (330 mg) at 48, 24, and 12 hours before the animals were killed. At the time of death, representative sections of iliac arteries over a 1.5-cm length were fixed in 4% paraformaldehyde for 2 hours and embedded in paraffin. Sections, 4 μm in thickness, were stained with hematoxylin/eosin and Movat-pentachrome and examined under light microscopy for evidence of injury. Cross-sectional area measurements of the intima and media were performed using an image analysis system linking a light microscope (Nikon Labophot) to a computer (Compix, Inc; C-imaging 1280 hardware with Simple 4.0 software). The images of the Movat-pentachrome–stained sections were viewed under low power (×4) light microscopy, and calibration was carried out using a reticle. The boundaries of the intima and media were identified and traced to allow for computer quantification of intimal and medial areas (mm2).
Intimal cell counting was performed using a previously validated method.2 32 Analyses were done on cross sections stained with hematoxylin under ×40 microscopic magnification. Random areas (encompassing 20% to 40% of the total intimal cross-sectional area) within the intima were selected, and cell nuclei were enhanced and counted after dynamic color thresholding. The average cell nuclear count within these known areas was used to calculate the cell density (cells/mm2), which, when multiplied by the previously measured total intimal area (from Movat-pentachrome sections), was used to calculate the total intimal cell count. Immunostaining against BrdU was determined in tissue sections fixed in paraformaldehyde for 2 hours and stained with a monoclonal antibody (Dakopatts). Endogenous peroxide was blocked with 1.5% hydrogen peroxide, and nonspecific binding was inhibited using 10% goat serum. BrdU antibody (1:50 dilution) was added for 45 minutes at 37°C, followed by a secondary antibody for 45 minutes. Sections were counterstained with aqueous hematoxylin. Small intestine was used as a positive control; liver, as a negative control.
Statistics
Data are expressed as mean±SD. To determine overall differences between groups subjected to balloon angioplasty and control groups in the collagen turnover studies, a two-way ANOVA with repeated measures was performed for each study. To determine specific differences between the curves representing the balloon-injured and control groups, paired t tests with a Bonferroni correction were performed at each time point. To find any temporal differences within either the balloon-injured or control group, a one-way ANOVA was performed. If significant differences were found within a group, a Tukey's studentized range test was performed to determine which time points were significantly different. In the GM6001 studies, paired t tests were performed to compare balloon-injured arteries with the contralateral control nondilated arteries. Unpaired two-tailed t tests were used to compare the placebo and GM6001 groups. All statistics were computed using SAS PC V6.04 (SAS Institute). Statistical significance was defined as P<.05.
Results
Collagen Turnover Studies
Significant increases in collagen synthesis were present in the balloon-injured iliac arteries compared with the control nondilated contralateral iliac artery immediately after angioplasty (second injury) and at 1 week and at 4 weeks, but no significant differences were evident at 12 weeks (Fig 3⇓, top). Peak collagen synthesis occurred at 1 week and was approximately four times the control values. Within the balloon-injured arteries, the 1-week levels of collagen synthesis were significantly increased compared with the other time periods.
Top, In vivo intact collagen synthesis per arterial segment based on [14C]hydroxyproline activity in the >30-kD fraction. Significant increases in synthesis in the balloon-injured arteries (BAs) compared with nondilated control arteries occurred immediately and at 1 and 4 weeks after the second balloon injury. Peak synthesis was present at 1 week. Bottom, In vivo collagen degradation of newly synthesized collagen per arterial segment based on total [14C]proline incorporation ([14C]proline plus [14C]hydroxyproline) in the <30-kD fraction. Significant increases in degradation in BAs compared with nondilated control arteries occurred immediately and at 1 and 4 weeks after the second balloon injury. Peak degradation was present at 1 week.
The degradation profile of collagen was similar to the synthesis (Fig 3⇑, bottom). Significant increases in collagen degradation in the balloon-injured artery, two to three times higher than that in the control side, were present immediately after balloon angioplasty and at 1 and 4 weeks. Peak increase in collagen degradation was measured at 1 week and was significantly increased compared with balloon-injured arteries at 4 and 12 weeks.
MMP Activity
Gelatin zymography supported the biochemical data by showing an increase of a 72-kD gelatinase (MMP-2) in the balloon-injured side compared with the control side (Fig 4⇓, left). Lytic bands were present at 72 and 62 kD, reflecting both the proenzyme and activated forms of MMP-2. Lytic activity on the gel was evident in the balloon-injured side immediately after the second injury, was maximal at 1 week, and remained detectable at 4 and 12 weeks (although at lower levels). The lytic activity present immediately after the second injury appeared to be due to the effects of the first injury, since similar bands were present in arterial extracts removed 3 weeks after the first injury (data not shown). There was no evidence of 92-kD gelatinase (MMP-9) activity in either balloon-injured or control nondilated arteries at any time points during the study. In the MMP-2 Western blot, a prominent band was present at 72 kD, confirming that the gelatinase activity evident on gelatin zymography was due to MMP-2 (Fig 4⇓, right). An additional band at 53 kD was also evident but appeared to be due to cross-reactivity between IgG in the vessel wall and the primary polyclonal antibody that was raised in rabbits. This was confirmed by a similar 53-kD band that appeared after electrophoresing an aliquot of the primary antibody and incubating with the secondary antibody (data not shown).
Left, Gelatin zymogram. In the balloon-injured arteries, two bands were evident at ≈72 and 62 kD, corresponding to the proenzyme and activated enzyme forms of MMP-2, respectively. The MMP-2 activity was induced early after balloon injury, was most intense at 1 week, and gradually decreased at 4 and 12 weeks. Minimal activity was present on the control side. The MMP-2 activity was abolished with pretreatment of the gel with EDTA and phenanthroline. MW indicates molecular mass markers (kD); T, balloon-injured arteries; C, control (nondilated) arteries; and IMM, immediately after angioplasty. Right, Western blot showing the presence of 72-kD gelatinase (type IV collagenase) in iliac arterial extracts. Lanes are as follows: 1, molecular mass markers; 2, control artery, immediately after angioplasty; 3, balloon-injured artery, immediately after angioplasty; 4, balloon-injured artery, 1 week after angioplasty; and 5, control artery, 1 week after angioplasty. In lanes 3 and 4, a band is present at 72 kD. An additional band is also evident at 53 kD, which represents IgG heavy chain (see text for details).
The MMP-1 Western blot analyses in all arterial extracts showed the presence of a band at ≈60 kD, confirming the presence of the procollagenase (Fig 5⇓). There was no apparent relationship between the intensity of this 60-kD band and specific time periods after injury or between balloon-treated and uninjured arteries.
Western blot showing the presence of procollagenase in iliac arterial extracts. A single band at ≈60 kD was evident at the four time points in all arteries, with no differences noted according to time period or balloon injury. Imm indicates immediately after the second balloon injury; C, control (contralateral noninjured artery); and T, treated (balloon-injured artery).
MMP Inhibitor Studies
Collagen Degradation
Significantly increased collagen degradation was evident in the placebo group in angioplastied arterial segments compared with the contralateral control artery (254±146 dpm [14C]hydroxyproline per segment versus 106±28, P<.02) (Fig 6⇓, top). However, in GM6001-treated arteries, there were no significant differences in collagen degradation between angioplastied and control arteries (133±51 and 82±15 dpm [14C]hydroxyproline per segment, respectively), suggesting inhibition of degradation by GM6001. However, a direct comparison of angioplastied treated arteries only indicated a trend toward reduction of degradation in GM6001-treated animals, despite an almost 50% reduction (P=.10).
Top, In vivo collagen degradation of newly synthesized collagen per arterial segment based on [14C]hydroxyproline activity in the <30-kD fraction. In the placebo-treated animals, significantly increased collagen degradation was evident in balloon-injured arteries (BAs) compared with contralateral control arteries. However, in GM6001-treated animals, there were no significant differences in BAs compared with control nondilated arteries, suggesting inhibition of degradation by GM6001. It should be noted that the [14C]hydroxyproline assay used in these GM6001 experiments is a more specific but less sensitive assay for collagen degradation than the total [14C]proline incorporation (Fig 3⇑, bottom). The collagen degradation results of the placebo animals in the [14C]hydroxyproline assay were entirely consistent with the 1-week values in the total [14C]proline incorporation assay shown in Fig 3⇑, bottom. Middle, In vivo intact collagen synthesis per arterial segment based on [14C]hydroxyproline activity in the >30-kD fraction in 1-week animals treated with GM6001 or placebo. Collagen synthesis was significantly increased in BAs compared with contralateral control arteries in both placebo- and GM6001-treated animals. GM6001-treated animals showed a significant reduction (≈45%) in collagen synthesis compared with placebo-treated animals in BAs. Collagen synthesis in the control (nondilated) arteries in both groups was not significantly different. Bottom, Total collagen per arterial segment based on arterial wall hydroxyproline content. Total collagen was significantly increased in BAs compared with the contralateral control arteries in both placebo- and GM6001-treated animals. However, GM6001 significantly reduced total collagen in BAs by 33%. Total collagen in the control arteries in both groups was not significantly different.
Collagen Synthesis
Collagen synthesis was significantly increased in balloon-injured iliac arteries compared with the contralateral control arteries in both placebo- and GM6001-treated animals (Fig 6⇑, middle). Collagen synthesis in the control (nondilated) arteries in both groups was not significantly different. However, the GM6001 significantly reduced collagen synthesis in balloon-injured arteries by ≈45% from 1199±380 to 683±278 dpm [14C]hydroxyproline per segment.
Collagen Content
Total collagen was significantly increased in balloon-injured iliac arteries compared with the contralateral control arteries in both placebo- and GM6001-treated animals (Fig 6⇑, bottom). However, GM6001 significantly reduced total collagen in balloon-injured arteries by 33%, from 750±143 to 500±78 μg hydroxyproline per segment (P<.004). Total collagen in the control arteries in both groups was not significantly different (placebo, 354 μg per segment; GM6001, 351 μg per segment).
Intimal Areas and Proliferation Rates
There was an ≈25% reduction in intimal cross-sectional area in GM6001-treated animals (0.35±0.15 versus 0.47±0.32 mm2) that was not statistically significant. Medial areas were very similar between the two groups (GM6001, 0.57±0.13 mm2; placebo, 0.54±0.10 mm2). There were no differences in intimal cell densities (GM6001, 5582 cells/mm2 intima; placebo, 5475 cells/mm2 intima). There were also no significant differences in BrdU-labeled intimal cells between GM6001-treated animals (195±89 cells per cross section, 7.5% labeling index) and placebo-treated animals (251±99 cells per cross section, 9.8% labeling index).
Discussion
In the present in vivo study, we have extended and confirmed our previous ex vivo observations of marked increase in collagen synthesis in the vessel wall in the first 4 weeks following balloon angioplasty.2 Furthermore, we have shown that collagen degradation is also stimulated early after balloon angioplasty and could account for the slow accumulation of collagen in the vessel wall, particularly in the initial weeks after balloon injury. In fact, the temporal profile of collagen degradation was remarkably similar to the collagen synthesis, suggesting the occurrence of reciprocal events between collagen synthesis and its degradation. Moreover, a specific MMP, MMP-2, is expressed early after balloon angioplasty, with a time profile consistent with the collagen degradation studies. This particular gelatinase can degrade intact type IV basement membrane collagen as well as denatured collagens type I and III (gelatin), which have been fragmented by interstitial collagenase (MMP-1).16 Recent work has also indicated that similar to interstitial collagenase, both avian and human MMP-2 can cleave soluble triple helical–type interstitial collagen with formation of 3/4- and 1/4-length collagen fragments.22 Thus, MMP-2 appears to possess both potent collagenase and gelatinase activity.
Collagen Turnover
In the collagen turnover studies, we used an in vivo [14C]proline incubation period of 24 hours rather than the 3-hour period previously used by Laurent and colleagues.12 23 Although synthesis of collagen was evident after a 3-hour incubation period, no detectable degradation of this newly synthesized collagen was apparent at this time. However, by 24 hours of incubation, both synthesis and degradation were measurable. Previous studies had indicated that plasma-specific activity of proline remains constant over the first 3 hours of infusion and then declines.23 Therefore, absolute rates of collagen synthesis and degradation over the 24-hour period used in our study could not be measured, but comparisons between groups of animals were permitted.
Both collagen synthesis and its degradation in normal uninjured rabbit iliac arteries were low. In the iliac arteries of the immediate group that underwent balloon angioplasty, collagen synthesis and degradation were increased, likely reflecting the effects of the first injury performed 3 weeks earlier. Similar to our ex vivo studies,2 we found a marked stimulation of collagen synthesis at 1 and 4 weeks after the second injury, ≈4.5 times and 2 times above that in the control noninjured contralateral iliac artery, respectively. This stimulation of collagen synthesis is time dependent, since control and balloon-injured arteries are not different at 12 weeks. Collagen degradation followed the same temporal pattern as collagen synthesis. This enhanced degradation, which was evident particularly at 1 week after balloon angioplasty, likely accounts for the delay in collagen accumulation in the vessel wall. This is supported by evidence of increased proteolytic activity based on our MMP studies.
MMPs
Several in vitro studies have shown that arterial smooth muscle cells from several species can produce various MMPs.33 34 35 Endothelial cells have been shown to produce gelatinases and TIMP.36 37 Human vascular smooth muscle cells in culture constitutively secrete a 72-kD gelatinase and TIMP-1 and TIMP-2 and, after cytokine stimulation, can produce a 92-kD gelatinase, interstitial collagenase, and stromelysin.38 Macrophage foam cells isolated from atherosclerotic rabbit aortas can synthesize stromelysin and interstitial as well as 92-kD gelatinase.18 However, our experiments were performed in a non–cholesterol-fed rabbit model, and few macrophages can be identified in these vascular lesions.2
MMPs are secreted from cells as latent enzymes, which are subsequently converted to active enzyme by proteases, including plasmin and stromelysin (MMP-3). The mechanisms controlling this activation are complex and only partially understood. Both gelatin zymography and Western immunoblotting indicated an upregulation of MMP-2 in injured arteries in our model. Procollagenase (Pro MMP-1) was present in the vessel wall of both injured and uninjured arteries with no obvious increase in injured vessels. The absence of marked detectable changes in procollagenase protein levels with injury or at specific time periods after the injury suggests that the procollagenase levels are not important sites of regulation. However, this does not preclude a significant role for MMP-1 in collagen turnover, since reliable measurements of the activated form of the enzyme are not available. There is evidence in the literature that potential activation of procollagenase by plasmin could be upregulated after balloon injury. It has previously been shown that tissue plasminogen activator and urokinase mRNA and activity are increased dramatically in the vessel wall early after balloon injury, with peak activity in the first week.39 It is possible that the increased generation of plasmin from these plasminogen activators could result in enhanced metalloproteinase activation and collagenolytic activity, but this remains speculative. The marked increase in injured arteries of MMP-2 lytic activity on zymography paralleled the temporal pattern of collagen degradation. These findings suggest a significant role of MMPs in mediating the process of collagen degradation, which was further supported by the results of the MMP inhibitor studies.
In vivo studies of MMPs in vascular disease are quite limited. Our results differed in several respects from results obtained using the rat carotid injury model.17 In normal uninjured rabbit iliac arteries, gelatinolytic activity was barely visible, whereas uninjured rat carotid arteries demonstrate activity at 70 and 62 kD. We did not note a prominent gelatinolytic band at 88 kD (MMP-9), although in the rat model this was present transiently from 1 to 4 days after injury (a time period not studied in our model). We found gelatinase lytic activity at both the 72- and 62-kD bands at all time periods studied, including the 12-week animals. In the rat carotid injury model, there was no increase in 72-kD gelatinase activity after carotid injury compared with the control condition, and the 62-kD band was only increased up to 8 days after injury. TIMPs were not assessed in that study and could be important regulators of matrix degradation. Previous work by Forough et al40 in the rat carotid artery injury model has shown an upregulation of TIMP-2 mRNA 24 hours after injury, with peak levels around day 5.
The involvement of MMPs in the pathogenesis of vascular lesions has only recently been studied. Bendeck et al17 have shown that suppression of MMP activity by GM6001 significantly inhibited smooth muscle cell migration into the intima after balloon injury of rat carotid arteries, with no effect on medial smooth muscle cell replication rates. Although there was a significant effect of GM6001 on neointimal formation at 1 week, a delayed compensatory increase in smooth muscle cell proliferation negated this initial beneficial response at 2 weeks.41 Galis et al20 have identified gelatinolytic and caseinolytic activity within the shoulders, core, and microvasculature of human atherosclerotic plaques, suggesting an important role for these enzymes in promoting plaque rupture and destabilization of atherosclerotic plaques.
The results of the present study further emphasize a new role for MMPs in vascular disease as an important mediator of the collagen turnover that occurs after balloon injury. We found important effects of MMP inhibition on both collagen degradation and collagen synthesis. Although it may be expected that MMP inhibition should directly inhibit collagen degradation and promote collagen accumulation, we found an even more profound effect of GM6001 on the inhibition of collagen synthesis. This resulted in an overall significant reduction of collagen accumulation in the vessel wall. It is also possible that this marked effect on collagen synthesis also accounted for some of the apparent inhibition of collagen degradation, since there was less newly synthesized collagen available for degradation in the GM6001-treated animals. Although the effects of the MMP inhibitor on net collagen accumulation may seem paradoxical, it should be noted that elastase inhibitors have also been shown to decrease migration of smooth muscle cells into the subendothelial space and inhibit collagen accumulation in pulmonary arteries of rats with pulmonary hypertension induced by monocrotaline.42
There are several possible explanations for the effects of the MMP inhibitor in the present study. First, since GM6001 was started 2 days before the second balloon injury, it could have significant effects on early and critical steps involved in the activation of smooth muscle cells, which could prevent transition to the synthetic phenotype and interrupt migration as well as matrix synthesis. Therefore, it is necessary to investigate further whether the effects of MMP inhibition initiated several days after angioplasty, beyond the stage of early smooth muscle cell activation and migration, would have a different effect on collagen accumulation. Although we observed an ≈25% reduction in intimal cross-sectional area at 1 week in the GM6001-treated animals, the effect of the MMP inhibitor may be underestimated after lesion formation due to the presence of the intimal lesion that forms in response to the first injury (0.27 mm2 in previous work from our lab2 ) before GM6001 administration. Since matrix accumulation occurs several weeks after the angioplasty, the full effect of GM6001 is likely not appreciated by studying the neointima at 1 week. It remains to be determined whether early administration of MMP inhibitors will have a sustained and perhaps even more significant effect on collagen accumulation and neointimal formation several weeks after angioplasty, considering the observations of Bendeck et al41 in the rat carotid model of a “catch-up phenomenon” at 2 weeks. Second, there may be additional effects of MMPs aside from matrix degradation that are active in vascular repair following balloon injury. Several cytokines and mediators that can potentiate the smooth muscle cell response to balloon injury can be inhibited by MMP inhibitors. A related hydroxamic acid MMP inhibitor has been shown to block the bioconversion of big endothelin to endothelin in anesthetized rats,43 presumably by inhibiting the activity of endothelin-converting enzyme. The conversion of big endothelin to endothelin is dependent on a neutral metallopeptidase present in vascular endothelial cells and smooth muscle cells, which is also sensitive to phosphoramidon, another metalloproteinase inhibitor that shares some structural similarity to the hydroxamic acid MMP inhibitor, GM6001.43 44 45 TNFα is a potent cytokine released by activated macrophages, smooth muscle cells, and T cells.46 TNFα, which has been immunolocalized in balloon-injured vascular lesions47 48 and in experimental atherosclerotic lesions of rabbits,49 mediates many of the “acute-phase” responses that promote smooth muscle cell activation.50 51 Recently, it has been shown that the processing of pro-TNFα to TNFα is mediated by a unique Zn2+ endopeptidase, which can be inhibited by an MMP inhibitor, including synthetic hydroxamic acid–based metalloproteinase inhibitors.52 53 54 Finally, it has also been shown that degradation products of collagen and gelatin are chemotactic for a variety of inflammatory cells that secrete cytokines, such as interleukin-1 and TNFα.55 56 57 By preventing the degradation of collagen and the formation of collagen degradation products, MMP inhibitors such as GM6001 could therefore limit the inflammatory response to injury that contributes to the development of the neointimal lesion.
In summary, collagen turnover appears to be an important mechanism that regulates accumulation of collagen in restenosis lesions after balloon angioplasty. Collagen synthesis and degradation of newly synthesized collagen are stimulated after balloon angioplasty and show a similar temporal profile. The significant reductions in collagen synthesis and collagen content and the inhibition of collagen breakdown in rabbits treated with an MMP inhibitor suggest an integral role of MMPs in collagen turnover after balloon angioplasty. This may have potential therapeutic implications in preventing restenosis in patients after balloon angioplasty.
Selected Abbreviations and Acronyms
| BrdU | = | bromodeoxyuridine |
| MMP | = | matrix metalloproteinase |
| TBS | = | Tris-buffered saline |
| TIMP | = | endogenous tissue inhibitor of matrix metalloproteinases |
| TNFα | = | tumor necrosis factor-α |
Acknowledgments
This study was funded by the Medical Research Council (Ottawa, Canada), the Heart and Stroke Foundation of Ontario, and the Connaught Fund, University of Toronto. Dr Strauss is a Research Scholar of the Heart and Stroke Foundation of Canada. Drs Batchelor, Ravi, and Natarajan are recipients of Research Fellowships from the Pettit Fund, University of Toronto. We would like to acknowledge the technical assistance of Enza Mancuso and the statistical help of Lois Adams. We are grateful to Dr Tony Cruz for technical advice and for providing the chondrocyte culture medium. We also would like to thank Dr Gregory Mishkel, Springfield, Ill, and ACS Canada for providing angioplasty catheters.
Footnotes
-
Reprint requests to Bradley H. Strauss, MD, PhD, Division of Cardiology, St. Michael's Hospital, 30 Bond St, Toronto, Ontario, Canada M5B 1W8. E-mail straussb@smh.toronto.on.ca.
- Received February 1, 1996.
- Accepted June 3, 1996.
References
This Issue
Jump to
Article Tools
Share this Article
- In Vivo Collagen Turnover Following Experimental Balloon Angioplasty Injury and the Role of Matrix Metalloproteinases
