β3-Integrins Rather Than β1-Integrins Dominate Integrin-Matrix Interactions Involved in Postinjury Smooth Muscle Cell Migration
Background—Smooth muscle cell (SMC) migration is a vital component in the response of the arterial wall to revascularization injury. Cell surface integrin–extracellular matrix interactions are essential for cell migration. SMCs express both β1- and β3-integrins. In this study, we examined the relative functional roles of β1- and β3-integrin–matrix interactions in postinjury SMC migration.
Methods and Results—Flow cytometry and fluorescence microscopy of migrating SMCs immunostained with anti-β1 and anti-αvβ3/5 antibodies (Abs) revealed expression of both β1- and β3-integrins, with β1 observed as linear streaks and β3 found in focal contacts. In a scrape-wound migration assay, anti-β1 Abs (92.0±10.7% of control, P=.1) and 0.5 mmol/L linear RGD (105±5% of control, P=.2) did not alter SMC migration at 48 hours after injury. β3-Blockade, however, via Abs (anti-β3/5 35.7±4.5% of control, anti-β3 61±12% of control, both P<.001) and cyclic RGD (0.5 mmol/L) (12±10% of control, P<.001) decreased migration. Neither β1- nor β3-inhibition altered postinjury [3H]thymidine incorporation. In the rat carotid injury model, local adventitial polymer-based delivery of radiolabeled linear or cyclic RGD led to uptake and retention of label, for both peptides, over a 72-hour period after injury. Local arterial wall β1-blockade via polymer-based delivery of linear RGD had no effect on SMC migration at 4.5 days (11.5±3.2 versus 12.8 SMCs per ×600 field [control], P=.6) or on neointimal thickening at 14 days (I/M area ratio, 0.664±0.328 versus 1.179±0.324 [control], P=.6) after injury. In contrast, local β3-blockade via cRGD limited migration (0.8±0.8 versus 12.8±4.4 SMCs per ×600 field [control], P<.01) and thickening (I/M area ratio, 0.004±0.008 versus 1.179±0.324 [control], P<.01).
Conclusions—In postinjury migrating SMCs, β3- rather than β1-integrin–matrix interactions are of greater functional significance in adhesive processes essential for SMC migration in vitro and in vivo. Blockade of dominant SMC integrin (β3)–matrix interactions may be a valuable approach for limiting injury-induced SMC migration and late arterial renarrowing.
The long-term success of percutaneous arterial revascularization remains limited by the development of restenosis.1 SMC migration and progressive arterial wall remodeling are vital components of the restenosis process.2 3 As such, an understanding of mechanisms operative in the regulation of postinjury SMC migration has potential for the development of new therapeutic approaches to limit restenosis.
A vital mechanism involved in cell migration is the interaction of the cell with its surrounding extracellular matrix microenvironment.4 5 Several receptor-ligand systems are involved in cell-matrix interactions. One system that has been shown to regulate cell phenotype and function is that of interactions between cell surface integrin and extracellular matrix. Integrins are a superfamily of transmembrane glycoprotein adhesion receptors consisting of two noncovalently linked subunits, α and β.6 7 To date, 14 α-subunits and 8 β-subunits have been described, with a resulting combination of more than 21 integrins with varying ligand specificity. Integrin-matrix interactions are functionally essential in many cell migration–dependent processes, including wound healing, metastasis, embryonic development, and immune defense.7 8 β1-Integrins traditionally have been demonstrated in many systems to be involved in cellular adhesion events essential for cell migration.9 10 Likewise, β3-integrins have been demonstrated to be involved in adhesive events essential for cell migration, although typically they are more important in adhesion to provisional matrixes associated with wound healing.11 12 Quiescent, uninjured SMCs have been demonstrated to express several integrins on their surfaces, including both β1-(α1β1, α2β1) and β3-(αvβ3) integrins.13 14 15
Several recent studies have suggested that β3-integrins are essential for SMC migration. In preliminary studies, we demonstrated that a cRGD peptide limited SMC migration in vitro.16 Others have shown that continuous local or systemic delivery of cRGD limited neointimal thickening after balloon arterial injury.17 18 Recently, the EPIC trial demonstrated that acute systemic infusion of abciximab (c7E3 Fab) directed against platelet glycoprotein IIb/IIIa, with cross-reactivity to β3-integrins (αvβ3), limited postangioplasty restenosis.19 Despite the recognition of the importance of both β1- and β3-integrins in cell migration, the relative functional roles of β1- versus β3-integrins in postinjury SMC migration remain unknown.
In this study, we hypothesized that after arterial wall balloon stretch injury, β3-integrin–matrix interactions are functionally of greater importance than β1-integrin–matrix interactions in adhesive processes essential for SMC migration. To test this hypothesis in the rat, we first characterized the expression of β1- and β3-integrins on migrating SMCs. We then examined the effect of preferential blockade of β1-integrin–matrix interactions versus β3-integrin–matrix interactions on SMC migration–dependent processes both in vitro and in vivo. We found that postinjury migrating SMCs express both β1- and β3-integrins on their surfaces. However, blockade of only β3-integrins is of functional significance in limiting SMC migration in the models studied both in vitro and in vivo. These observations suggest that SMC β3-integrins exert a functional predominance over β1-integrins in modulating integrin-matrix interactions essential for postinjury migration–dependent processes.
All animal experiments were performed in accordance with the “Position of the American Heart Association on Research and Animal Use” and the guidelines of the Animal Care Committee of the University of Arizona. The University of Arizona is accredited by the American Association of Laboratory Animal Care.
Overall Experimental Plan
To examine the relative roles of β1- and β3-integrins in SMC migration, the following in vitro and in vivo protocols were undertaken: in vitro, (1) demonstration of β1- and β3-integrin expression on migrating SMCs, (2) the effect of inhibition of β1- and β3-integrin–matrix interactions on SMC migration, and (3) the effect of inhibition of β1- and β3-integrin–matrix interactions on SMC proliferation; and in vivo, (1) demonstration of local arterial wall uptake of radiolabeled RGD peptides, (2) the effect of local arterial wall inhibition of β1- and β3-integrin–matrix interactions on SMC migration, and (3) the effect of inhibition of β3-integrin–matrix interactions on neointimal thickening at 14 days after injury.
Antibodies, Peptides, and Reagents
For β1-antibodies, hamster anti-rat β1-integrin subunit Ab (clone Ha 2/5) was obtained from Pharmingen, and rabbit anti-human α5β1-antiserum (anti-FNR) was from GIBCO/BRL. For β3-antibodies, mouse anti-rat β3-integrin subunit Ab (clone F11) was from Pharmingen and rabbit anti-human αvβ3/5 (anti-VNR) from GIBCO/BRL. RGD peptides used included cyclic Gly-Pen-Gly-Arg-Gly-Asp-Ser-Pro-Cys-Ala (GPenGRGDSPCA), linear Gly-Arg-Gly-Asp-Ser-Pro (GRGDSP), and linear Gly-Arg-Ala-Asp-Ser-Pro (GRADSP) inactive control. Peptides, fibronectin, and vitronectin were obtained from GIBCO/BRL.
Media SMCs were isolated from adult rat aorta and cultured in DMEM plus 10% FCS according to the explant method of Ross and Kariya.20 Cells were identified as vascular SMCs through their characteristic hill-and-valley growth pattern as well as via indirect immunofluorescence with monoclonal anti–SMC α-actin (Sigma Chemical Co).21 By this method, cultures of >95% SMC purity were routinely obtained. Studies were performed with cells at passage 3 to 4.
In Vitro Studies
Protocol 1: Expression of β1- and β3-Integrins by Migrating SMCs
Confirmation of Ab Cross-reactivity With Rat Integrins via Western Blotting
Rat SMC lysate was prepared in TNES buffer (50 mmol/L Tris-HCl, pH 7.4/1% Nonidet P 40/2 mmol/L EDTA/100 mmol/L NaCl/1 mmol/L PMSF/20 μg/mL leupeptin/20 μg/μL aprotinin). Postnuclear supernatants were fractionated by 7.5% SDS-PAGE under nonreducing conditions. Proteins were transferred to nitrocellulose electrophoretically. After blocking in 3% nonfat milk, membranes were probed with anti-integrin antibodies followed by species-appropriate peroxidase-conjugated secondary antibodies (1:20 000 Kierkegard and Perry Laboratories) and chemiluminescent substrate (Pierce).
Demonstration of Cell Surface Integrin Expression by Flow Cytometry
Subconfluent SMCs were harvested with trypsin-EDTA and resuspended in cold PBA (1.0% BSA/0.1% sodium azide in PBS), followed by incubation with primary antibodies at 4°C for 60 minutes. After a washing in PBA, cells were resuspended in appropriate phycoerythrin-conjugated goat anti-rodent immunoglobulin (Jackson Research Laboratories; 1:50 in PBA). After an additional 30 minutes at 4°C, cells were washed and analyzed by flow cytometry with a Becton-Dickinson FACScan. Mean fluorescence intensity was determined with LYSIS II software.
Integrin Localization via Immunofluorescence Microscopy
Dense monolayers of rat aortic SMCs were established on glass coverslips and scrape-wounded with a sterile wooden applicator as described,22 washed with sterile DMEM, and incubated for 24 hours. Coverslips were fixed, permeabilized, and blocked with 1 mg/mL BSA in PBS (wash buffer). Samples were incubated with anti–β1-subunit Ab or anti-VNR Ab (1:100 in wash buffer) for 1 hour at 25°C. Appropriate secondary antibodies were applied as rhodamine conjugates (1:100 in wash buffer) for 1 hour at 25°C. Incubation with preimmune mouse or rabbit serum followed by fluorescent secondary antibodies served as nonspecific background controls. SMCs migrating from the leading edge of the scrape wound were viewed with a Leitz Diaplan immunofluorescence microscope at ×1000 magnification under UV epi-illumination with a rhodamine filter cube.
Protocol 2: Effect of Inhibition of β1- and β3-Integrin–Matrix Interactions on SMC Migration In Vitro
Function-blocking Activity of Anti-integrin Reagents
Fibronectin- and vitronectin-coated substrates were prepared by addition of solutions of fibronectin or vitronectin (10 μg/mL in sterile PBS) to 24-well dishes. Rat aortic SMCs were harvested nonenzymatically with 0.54 mmol/L EGTA, washed in DMEM containing 2 mg/mL BSA, and resuspended in DMEM/Ham’s F12 1:1 with 4 mg/mL BSA at a final concentration of 1×105 cells/mL. Cells were preincubated with peptides (1 mmol/L) or antibodies (1:50) for 15 minutes at 37°C and then seeded at 2×104 cells/cm2 in 0.2 mL/well on fibronectin- and vitronectin-coated dishes and incubated for 1 hour at 37°C. Wells were gently rinsed with serum-free medium (37°C), fixed, stained with 0.1% toluidine blue (10 minutes), and examined microscopically. Adherent cells were counted in ten 100× fields to determine the number of adherent cells per square millimeter. The extent of cell adhesion for each sample group was expressed as the number of cells per square millimeter.
Quantification of Scrape Wound–Induced SMC Migration In Vitro
Dense monolayers of SMCs on tissue culture plastic maintained in DMEM+10% FCS or SMCs on fibronectin-coated dishes maintained in DMEM+0.2% BSA were wounded as described.22 Monolayers were washed with DMEM and incubated in either DMEM+10% FCS or DMEM+0.2% BSA (serum-free conditions) plus integrin inhibitory agents, ie, antibodies at 1:50 and peptides at 0.2–1.0 mmol/L, for 48 hours. Cells were then fixed, stained with 0.1% toluidine blue (10 minutes), and examined by bright-field microscopy (Zeiss, Axiovert, ×100) with an attached video capture and analysis system. A closed perimeter of the zone of migrating cells in the wound site, defined by the original wound edge, the microscopic field boundaries, and the leading edge of cells migrating into the wound zone, was traced with a digitizing tablet (Kurta) for each field, and the area encompassed by migrating cells was quantified. A mean area of coverage by migrating cells was determined from three random fields in each of the wells per sample group.
Protocol 3: Effect of Inhibition of β1- and β3-Integrin–Matrix Interactions on SMC Proliferation In Vitro
Quantification of SMC Proliferation
Rat aortic SMC monolayers were established in 24-well culture dishes, scrape-wounded, and treated with integrin inhibitors as described above. After injury, the medium in each well was removed and replaced with medium containing [3H]thymidine (1 μCi/well). After 18 hours of incubation at 37°C, samples were rinsed with PBS and fixed with cold 10% aqueous TCA for 30 minutes. TCA-precipitated material was solubilized in 1% SDS/0.3N NaOH, and radioactivity incorporated into TCA-precipitated material was quantified by liquid scintillation counting (Beckman model LS 6000LL). Protein content was determined with the Bio-Rad DC protein assay (Bio-Rad). Thymidine incorporation for each sample was expressed as cpm/μg total protein.
In Vivo Studies
Protocol 1: Local Arterial Wall Uptake of Radiolabeled RGD Peptides Delivered via Adventitial Polymer In Vivo
Radiolabeling of Linear and Cyclic RGD Peptides
The free primary amines of RGD peptides were acylated by use of N-succinimidyl 2,3-[3H]propionate ([3H]NSP) (Amersham, 99.0 Ci/mmol) according to the method of Bolton and Hunter.23 GRGDNP (linear) and GPenGRGDSPCA (cyclic) peptides (BRL, Life Technologies) dissolved in borate buffer (pH 8.5) were reacted with 300 μCi (3 nmol) of [3H]NSP in a final volume of 24 μL (60 minutes, 24°C). The reaction mixture was fractionated by silica gel TLC in an aqueous buffer system (15 mmol/L H2PO4/250 mmol/L NaCl). Tritium label was detected by autoradiography of TLC plates, and the silica gel corresponding to labeled peptide of the expected retardation factor (linear=0.36, cyclic=0.19) was scraped into Eppendorf tubes. Peptides were eluted into 15 mmol/L H3PO4, neutralized, and lyophilized. Samples were reconstituted in water, and aliquots were analyzed for purity by repeat TLC (>95%) and for radioactivity by liquid scintillation counting. Aliquots of TLC-purified, radiolabeled peptide were added to 1 mmol/L solutions of the corresponding nonradioactive peptide in 25% Pluronic F127 gel to achieve a final peptide specific activity of 5.4 Ci/mol and radiotracer concentration of 5.4 μCi/mL.
Rat Carotid Injury and Local Uptake of Radiolabeled RGD Peptides In Vivo
The rat carotid artery balloon stretch injury model of Clowes et al24 was used, with modification. Male Sprague-Dawley rats (350 to 400 g, Harlan, Indianapolis, Ind) were anesthetized (ketamine 80 mg/kg and xylazine 10 mg/kg IM), and the left common and external carotid arteries were exposed. A 2F Fogarty balloon catheter (Baxter) was inserted via the external carotid into the common carotid, advanced proximally, and pulled retrograde three times, with the balloon inflated sufficiently to encounter resistance. After injury, the balloon catheter was removed and the external carotid ligated. A 10-mm length of PE-350 tubing was placed around the common carotid, and radiolabeled peptide solutions in liquid Pluronic (25% in PBS) were instilled and allowed to gel within the PE-350 segment in contact with the adventitia. At 1, 24, and 72 hours after application, rats were killed and carotids were harvested and rinsed in iced saline to remove residual gel. In addition, samples of contralateral (untreated) carotid artery, heart, lung, liver, spleen, kidney, and skeletal muscle were also excised. Samples were weighed, solubilized (Soluene 350, Packard Instruments), and analyzed by liquid scintillation counting (Beckman model LS 6000LL). Retention of radiolabel in the arterial wall was expressed as dpm/mg tissue.
Protocol 2: Effect of Local Inhibition of β1- and β3-Integrin–Matrix Interactions on SMC Migration In Vivo
Quantification of Balloon-Injured SMC Migration In Vivo
Sprague-Dawley rats were anesthetized, the left carotid artery was balloon-injured, and cRGD in Pluronic, linear RGD in Pluronic, and Pluronic alone (sham treatment control) were applied as described above (n=5 rats per group). SMC migration in vivo was quantified according to a modification of the method of Jackson and Reidy.25 Four and one half days after injury, rats were heparinized and killed. Injured carotid arteries were washed with PBS, pressure-fixed with 2.5% glutaraldehyde in PBS at 100 mm Hg, excised, opened longitudinally while continuously submerged in 25% ethanol in water, and pinned to paraffin supports with the endoluminal surface exposed en face. Vessels were dehydrated through graded ethanol, followed by critical-point drying (Fissons Instruments) and sputter-coating with gold. Coated specimens were examined in a Hitachi 2300N scanning electron microscope at 20 kV at ×600. SMCs migrating onto the luminal surface were readily identifiable via protruding nuclei. SMC migration was quantified as mean number of cells (nuclei) emerging on the intimal surface/×600 field.
Protocol 3: Effect of Local Inhibition of β1- and β3-Integrin–Matrix Interactions on Neointimal Thickening After Injury In Vivo
Quantification of Postinjury Neointimal Thickening
Sprague-Dawley rats were anesthetized, the left carotid artery was balloon-injured, and Pluronic gels were applied as described. At 14 days after injury, rats were killed and the carotid arteries were pressure-fixed as above, sectioned, and stained with 0.1% toluidine blue. Samples were examined at ×63 magnification with the video microscopy system described above. Arterial cross sections were traced with a digitizing pad, and media and intima cross-sectional areas were quantified. The ratio of intima to media cross-sectional area (I/Marea), a morphometric index for neointimal thickening, was calculated for each sample, and mean I/Marea±SD were determined for each experimental group (n=5 per group).
Data are presented as mean±SD. Statistical comparisons were made with Student’s t test. Significance was defined at the P<.05 level. Retention of radiolabel in the arterial wall at different time points was analyzed by ANOVA with SPSS 6.1 software. Unless otherwise stated, all experimental determinations were performed in triplicate.
In Vitro Studies
Integrin Expression on Migrating Rat Aortic SMCs
As a first step, immunoblot analysis of whole-cell lysates was used to verify that the anti–human integrin Abs used in this study recognize specific rat integrin subtypes. Blotting with monoclonal β1- and β3-subunit–specific antibodies revealed signals of apparent molecular sizes of 115 and 90 kD, consistent with the expected sizes of these subunits under nonreducing conditions (Fig 1A⇓). Blotting with rabbit polyclonal antibodies to the human fibronectin and vitronectin receptors revealed signals for both α- and β-subunits, as expected (Fig 1B⇓). The apparent molecular sizes are consistent with previously reported data.26 The identity of the signal indicated by an asterisk is unclear at this time. It may represent nonspecific cross-reactivity, proteolytic degradation of the subunit during analysis, or detection of a nascent subunit with immature posttranslational modification giving it a lower apparent molecular size. Flow cytometric analysis using anti-rat β1- and β3-subunit–specific antibodies revealed readily detectable levels of surface expression for both β1- and β3-integrins with mean fluorescence intensities of 403.2 and 281.3, respectively, relative to immunoglobulin controls (2.5) (Fig 1C⇓). Immunofluorescence microscopy confirmed cell surface expression of both β1- and β3-integrin subunits on migrating SMCs (Fig 2⇓) by use of anti–β1-integrin subunit Ab and anti-VNR Ab. Differences in the patterns of integrin expression were noted for β1- and β3-containing integrins. β1-Integrins were typically seen in a fine linear, fibrillar, streaky pattern, widely distributed over the surface of migrating cells (Fig 2A⇓). β1 was not seen in aggregates, patches, or other organized adhesion plaque structures. β3 was seen prominently organized at the leading edge of migrating cells as brightly fluorescent patches consistent with focal contact morphology (Fig 2B⇓). β3-Subunit was also seen distributed over the entire cellular cytoplasmic membrane in a coarse punctate pattern (Fig 2B⇓).
Effect of Inhibition of β1- and β3-Integrin–Matrix Interactions on SMC Migration In Vitro
The selectivity of individual reagents as inhibitors of rat integrin function was examined before studies examining the effect of specific blockade on SMC migration. Preincubation of SMCs with anti–integrin β1-subunit Ab did not alter cell adhesion on vitronectin substrates (89.3±15.5% of control, P=.143) (Fig 3A⇓); however, adhesion on fibronectin was strongly inhibited (32.6±10.6% of control, P<.001). In contrast, preincubation of SMCs with anti-VNR Ab or mouse anti–rat β3 Ab resulted in near complete inhibition of cell adhesion on vitronectin-coated substrates (anti-VNR 7.2±8.2% of control adhesion, anti-β3 14±1% of control, P<.001) (Fig 3A⇓), with no decrease in adhesion on fibronectin (anti-VNR 108.1±33.2% of control, P=.4; anti-β3 96±1% of control, P=.8).
Specific RGD peptides had distinct effects on SMC adhesion to defined substrates as well. Pretreatment of cells with 1.0 mmol/L linear GRGDSP did not alter cell adhesion to vitronectin (86.0±20.1% of control, P=.3); however, it strongly inhibited cell adhesion to fibronectin-coated substrates (53.1±19.6% of control, P<.001) (Fig 4A⇓). In contrast, a significant decrease in cell adhesion to vitronectin-coated substrates was observed when SMCs were preincubated with 1.0 mmol/L cyclic GPenGRGDSPCA (cRGD) (1.8±2.4% of control adhesion, P<.001) (Fig 4A⇓). Pretreatment with cRGD, however, did not alter adhesion to fibronectin-coated substrates (102±19.4% of control adhesion, P=.1).
The effect of these inhibitory reagents on SMC migration after scrape-wound injury was then examined in vitro. Treatment of scrape-wounded SMC monolayers with anti–β1-integrin Ab did not alter 48-hours-postinjury SMC migration (92.0±10.7% of control, P=.1) (Fig 3B⇑). Treatment with anti-VNR Ab or mouse anti–rat β3 Ab after injury resulted in a significant decrease in the extent of SMC migration at 48 hours after injury (anti-VNR 35.7±4.5% of control, anti-β3 61±12%, P<.001) (Fig 3B⇑).
Postinjury treatment of scrape-wounded SMC monolayers with the linear RGD peptide GRGDSP as well as the inactive control peptide GRADSP had no effect on postinjury SMC migration over the 0.2 to 1.0 mmol/L concentration range demonstrated to be effective for the cyclic RGD peptide (Fig 4B⇑). The cyclic RGD peptide GPenGRGDSPCA (0.2 to 1.0 mmol/L), however, led to a significant reduction of migration at 48 hours after injury, with maximal inhibition at 1 mmol/L cyclic RGD to 1.0±1.0% of control migration (Fig 4B⇑).
To rule out the possibility that the observed predominant effect of β3 in postinjury SMC migration was due to preferential interaction of β3-integrins with serum-derived vitronectin, migration studies were also performed under serum-free conditions. Under serum-free conditions, 1 mmol/L linear GRGDSP led to a minor reduction in SMC migration, whereas 1 mmol/L cyclic RGD led to a dramatic reduction in SMC migration (linear RGD 81.5±7.5% of control, cRGD 0.3±35% of control, P<.01).
Effect of Inhibition of β1- and β3-Integrin–Matrix Interactions on SMC Proliferation In Vitro
Postinjury treatment of scrape-wounded SMCs with either anti-β1 or anti-VNR Ab did not alter proliferation, as measured by [3H]thymidine incorporation. Similarly, postinjury treatment of scrape-wounded SMCs with 1 mmol/L of either linear or cyclic RGD peptide did not alter [3H]thymidine incorporation. As a positive control, postinjury treatment of cells with the SMC mitogen platelet-derived growth factor resulted in increased proliferation, as indicated by a significant increase (140.1±11.7% of control, P<.001) in [3H]thymidine incorporation (Fig 5⇓).
In Vivo Studies
Local Arterial Wall Retention of Radiolabeled RGD Peptides In Vivo
Local delivery of radiolabeled RGD peptides to the adventitial surface of balloon-injured rat carotid arteries via hydrogel led to rapid uptake and retention in the arterial wall. At 1 hour, 1754.6±156 dpm/mg tissue and 1368.3±381.4 dpm/mg tissue were present for linear and cyclic RGD, respectively. Linear RGD remained detectable in the arterial wall at 24 and 72 hours after delivery at 333.5±261 and 358.5±266 dpm/mg (Fig 6⇓). Similarly, cyclic RGD remained detectable in the arterial wall at 24 and 72 hours after delivery at 192.5±37.2 and 169±114 dpm/mg. Comparison of treatment groups by ANOVA revealed somewhat greater label retention after application of the linear material (P=.05). At all time points, minimal label was detected in distant organ sites, including contralateral untreated carotid, skeletal muscle, heart, lung, liver, kidney, and spleen. At all time points, the amount detected in the treated artery for both groups was a minimum of 20 times the level detected in these organ sites.
Effect of Local Arterial Wall Inhibition of β1- and β3-Integrin–Matrix Interactions on SMC Migration In Vivo
Local delivery of linear RGD to post–balloon injury carotid arteries did not alter SMC migration in vivo (11.5±3.2 versus 12.8±4.4 SMCs per ×600 field for control untreated vessels, P=.6). In contrast, local delivery of cRGD led to a significant reduction in SMC migration in vivo (0.8±0.8 versus 12.8±4.4 SMCs per ×600 field for control untreated vessels, P<.01) (Fig 7⇓).
Effect of Inhibition of β3-Integrin–Matrix Interactions on Neointimal Thickening at 14 Days After Injury In Vivo
Local delivery of cyclic RGD via a periadventitial hydrogel led to a significant reduction in neointimal thickening at 14 days after balloon injury. In control rats treated with Pluronic gel alone, mean intima/media area ratios were 1.179±0.324. Pretreatment with gel containing linear RGD led to a mean intima/media area of 0.664±0.328, which was not significantly different from control rats treated with gel alone (P=.6) (Fig 8⇓). In contrast, for cRGD-treated rats, mean intima/media area ratios were 0.004±0.008, P<.01. Furthermore, there was no difference in neointimal thickening for rats with carotid injury without gel treatment, with mean intima/media area ratios of 1.185±0.334, versus rats treated with Pluronic gel alone.
The principal finding of this study is that blockade of SMC β3- rather than β1-integrin–matrix interactions limits postinjury SMC migration in vitro and in vivo. In this study, we demonstrated that migrating postinjury SMCs express both β1- and β3-integrins on their surface. Despite the presence of both β-integrin subtypes, β3-integrins rather than β1-integrins appear to exert a functional predominance in both in vitro and in vivo migratory behavior.
β-Integrin Expression on Postinjury Migrating SMCs
Previous studies examining SMC β1- or β3-integrin expression have detected expression of one or both of these β-subunits on rabbit, lamb, and human vascular SMCs.13 14 27 28 29 We observed that postinjury migrating rat SMCs also express both β1- and β3-integrins on their surface. Interestingly, a clear difference in the pattern of distribution of β1- compared with β3-integrins was observed (Fig 2⇑). β1-Integrins were found to be diffusely distributed over the surface of migrating SMCs, organized as fine linear streaks rather than highly organized in focal adhesions at the cell edge. β3-Integrins were observed both in a diffuse distribution over the cell surface and in highly organized distinct focal adhesions predominantly at the leading edge of migrating cells. In previous studies with lamb SMCs on the defined substrates of collagen, fibronectin, and laminin, it was observed that β1-integrins were highly organized into focal contacts distributed around the cell periphery.13 14 In contrast, β3 staining was less intense and less organized, with fewer focal contacts than β1-integrin. The differences in the pattern of distribution of β1- and β3-subunits that we observed can be explained by the differing conditions of the two studies. The finding of β1 in focal contacts was observed for SMCs, largely adherent, with various degrees of migration on defined substrates, ie, fibronectin.14 In the present study, SMCs were migrating on a plastic substrate covered with post–scrape injury cell membrane residua30 and with endogenously secreted matrix.31 This secreted matrix is composed of a composite of extracellular matrix proteins, including collagens, fibronectin, laminin, elastin, osteopontin, and proteoglycans, which are continuously remodeled over time. Our finding of β3 rather than β1 in adhesion plaques is consistent with what might be expected for migrating cells surrounded by such a composite transitional matrix typical of a wound. Furthermore, matrix components such as osteopontin, denatured collagen, and fibronectin, which become integrated in the postinjury transitional matrix, have also been demonstrated to facilitate organization of β3-integrins into focal contacts on the cell periphery.14 32
β1-Integrin–Matrix Interactions in SMC Migration
β1-Integrins have been demonstrated in many biological systems to be essential for cell migration–dependent processes.9 10 As such, it seemed feasible to postulate at the outset of this study that β1-integrins may play a vital role in SMC migration after mechanical injury. Because cell migration involves many adhesion-dependent events, including direct substrate adhesion, spreading, and locomotion, we first demonstrated that the specific anti-integrin antibodies and peptides used in this study were function-blocking. Adhesion assays performed on fibronectin and vitronectin substrates showed that anti–β1-integrin Ab and linear RGD peptide inhibited SMC adhesion only on fibronectin substrates (Figs 3A⇑ and 4A⇑). These observations underscored the supposition that fibronectin adhesion is predominantly a β1-integrin–dependent process and that anti–β1-integrin Ab and linear RGD peptide selectively inhibited β1-integrin function in rat aortic SMCs. We then used a scrape-wound injury model to examine the effects of β1-integrin blockade on SMC migration in vitro.22 This model differs from simple chemotactic assays in that migration requires adhesion to a solid substrate and traction-based crawling of cells. In this model, a complex matrix exists on the substratum, composed of residual secreted matrix, cellular debris, and underlying basement membrane remnants.30 This assay was chosen because it more closely parallels the complex matrix and integrin-matrix interactions that occur in vivo in the arterial wall after injury. When the outlined function-blocking reagents were used in the scrape-wound model, we observed that blockade of SMC β1-matrix interactions did not limit SMC migration (Figs 3B⇑ and 4B⇑).
Previous investigators have suggested that β1-integrins are essential specifically for SMC migration.14 28 In migration studies examining non–injury-associated SMC migration on single, specific, matrix substrates, blockade of β1-integrins resulted in some reduction in migration.14 In transwell migration assays examining human SMC migration toward type I collagen, blockade of β1-integrins limited SMC migration.28 Our observations differ from these studies. In addition to potential species-specific differences, our observations may be explained by differences in the migration assays used. We specifically examined migration on a solid substrate. In the scrape-wound assay, migration is stimulated by multiple factors, including locally released chemotactic factors, damaged cell and matrix components, and serum factors. Both of the previous studies examined migration either on or toward collagen type I, a more specific β1-ligand,33 under conditions in which no other matrix components were synthesized. Our studies specifically examined migration over a complex postinjury matrix composed of numerous matrix components as outlined above.
For examination of the effect of β1-blockade on SMC migration in vivo, we used a modification of the en face migration assay of Jackson and Reidy.25 We administered linear RGD peptide, which is described to have greater specificity for β1-integrins,34 locally to the arterial wall via an adventitial polymer vehicle. Despite delivery to the arterial wall of peptide, no significant limitation of SMC migration was observed (Fig 7⇑). Similarly, local linear RGD delivery did not limit neointimal thickening at 14 days after injury (Fig 8⇑). These in vitro and in vivo data suggest that β1-integrins, despite having a predominant role in cell migration in many systems, do not have a functionally predominant role in arterial SMC migration after mechanical injury.
β3-Integrin–Matrix Interactions in SMC Migration
β3-Integrins have also been described as being essential in cell migration processes. The β3-subunit is widely expressed on mesenchymal cells, including SMCs, endothelial cells, and osteoclasts, as well as on platelets.7 27 35 The expression of β3 on tumor cells has been shown to frequently correlate with increased invasive and metastatic potential.36 αvβ3 is a highly promiscuous receptor able to bind to a wide variety of extracellular matrix components, including vitronectin, von Willebrand factor, thrombospondin, osteopontin, tenascin, and denatured collagen.7 35 These matrix proteins are typically seen in a postinjury provisional matrix.35 37 38
In the present study, selective inhibition of rat SMC β3-integrin function by anti-VNR Ab, anti-rat β3-subunit Ab, and cyclic RGD peptide was demonstrated. Adhesion assays performed on fibronectin and vitronectin substrates revealed that β3-integrin blockade selectively inhibited SMC adhesion to vitronectin versus fibronectin (Figs 3A⇑ and 4A⇑). When these reagents were used in the scrape-wound model, we observed that blockade of SMC β3-integrin–matrix interactions limited postinjury migration in vitro. Consistent with our observation of the importance of β3-integrins in SMC migration was the observation of Clyman et al.14 Using a noninjury migration model, they observed that β3-integrin inhibition limited SMC migration on defined substrates. Similarly, the observations of Yue et al39 are consistent with our findings. They observed that treatment of SMCs with F11, a rat-specific anti-αvβ3 Ab, limited osteopontin-stimulated SMC migration in a concentration-dependent fashion.
Further support for a functional predominance for β3 in SMC migration is our finding of β3-subunit staining in focal contacts at the leading edge of the cell (Fig 2B⇑). The ability of a cell to form focal contacts correlates with its ability to assemble an organized cytoskeleton, a requirement for cell spreading and migration.40 In cell locomotion, an asymmetry of adhesion receptors is described, with vital receptors necessary for force generation seen at the leading edge.41 Our finding of β3 organized into focal contacts, asymmetrically distributed on the leading migratory edge, is consistent with a vital role for this integrin in substrate adhesion and force generation needed for forward locomotion.
Local Delivery of cRGD Peptides Limits In Vivo Migration and Progressive Intimal Thickening
Two previous studies have examined the effect of in vivo β3-integrin blockade, via cRGD administration, on postinjury neointimal thickening in the rabbit and hamster.17 18 Our findings in the rat model agree with these studies in that cRGD-mediated β3-blockade limited neointimal thickening at 14 days after injury (Fig 8⇑). Our study has gone beyond these other reports in several ways. To date, studies of integrin blockade have not examined the effect on SMC migration in vivo. The present study demonstrates for the first time that local arterial wall β3-blockade limits in vivo SMC migration. Furthermore, in the present study we have examined the effects of both of β1- and β3-blockade in the same experimental model, demonstrating that β1-blockade is of little in vivo functional significance, with β3-blockade limiting postinjury SMC migration (Fig 7⇑) and late neointimal thickening (Fig 8⇑).
The mode of drug administration used to achieve integrin blockade also differs in the present study. Previous studies have examined either prolonged intravenous systemic delivery or prolonged regional delivery of cRGD via an implanted osmotic pump. The present study specifically used polymer-based adventitial delivery, with a surrounding nonporous polyethylene capsule, to provide focused circumferential local arterial wall delivery of RGD peptides.
The present study also provides some insight into the treatment period that is sufficient for in vivo integrin blockade to achieve altered SMC migration and reduced neointimal thickening. In the present study, radiolabel from cRGD delivery was detectable in the arterial wall for 72 hours after injury. If this detectable label is due to persistence of intact functional peptide, then this study suggests that sustained local arterial wall treatment within the 72-hour period after injury is adequate to reduce SMC migration and neointimal thickening at 14 days in our model. This observation is consistent with studies that have demonstrated that the SMC migration in vivo occurs primarily within the first few days after injury.42
Our data demonstrating the significance of blocking SMC β3-integrin–matrix interactions as a means of limiting postinjury arterial renarrowing are consistent with the finding of recent clinical studies examining agents with SMC β3-inhibitory activity. In the EPIC trial, abciximab, with bispecific αIIbβ3- and αvβ3-inhibitory activity,14 43 was effective in limiting the need for late coronary revascularization at 6 months after PTCA, with durability of that result over 3 years.44 In contrast, in the IMPACT II trial, in which the anti-integrin KGD peptide integrilin, an agent without specific inhibitory activity against SMC αvβ3, was used, no reduction in the need for late coronary revascularization was observed in a 6-month angiographic substudy.43
The recent EPILOG study differed from EPIC in the durability of the need for repeat coronary revascularization.45 The reason for the disparity in late revascularization rates between the two trials remains unknown, although it may be related to differences in patient population, heparin dosage, stent usage, or other clinical factors. Our data, however, lend biological mechanistic support to the EPIC outcome.
Limiting SMC integrin-matrix interactions may also provide secondary benefits, such as limitation of deposition of lumen-occupying extracellular matrix components, through effects on protease activity.46 Blocking integrin-matrix interactions may also alter late constrictive remodeling events, although this may require sustained delivery of integrin inhibitors for periods longer than in the present study, such as that achievable via other polymer-based sustained-release drug delivery systems.47 48 Local therapy with αvβ3-inhibitory antibodies, peptides, and peptidomimetics should also reduce the risks of systemic administration of these agents, which frequently possess anti–platelet GP IIb/IIIa (αIIbβ3-integrin) activity and are associated with an increase in bleeding events.49 50
The role of a cRGD-mediated antiplatelet effect contributing to the observed in vivo limitation of SMC migration and late thickening was not directly excluded in this study. It is unlikely that local antiplatelet effects are predominantly responsible for our observations. Our in vitro findings of reduced SMC migration as a consequence of β3-integrin blockade, under conditions that are platelet independent, are consistent with and support our in vivo observations of limited SMC migration and late thickening. On the clinical level, specific inhibition of platelet GP IIb/IIIa with agents that do not cross-react with αvβ3, ie, the KGD peptide integrelin, have not limited postangioplasty restenosis,43 whereas clinical restenosis was reduced by abciximab, which has dual αIIbβ3- and αvβ3-inhibitory activity.14 43 Future studies will be necessary to address the effects of a possible antiplatelet component contributing to reduced in vivo SMC migration and neointimal thickening in this model.
From the present study, we are unable to determine the mechanism by which β3-integrins exert their functional predominance. Changes in the levels of surface expression, activation state, or clustering of β3 relative to β1 may be involved in β3 functional predominance. Alternatively, local growth factors, cytokines, or other nonintegrin adhesion receptors may be responsible for the observed predominance of β3-integrins. The possibility that mechanisms other than blockade of β3-mediated SMC migration could account for our observed in vivo decrease in en face–detectable SMCs after injury was not excluded by this study. SMC apoptosis may also be induced via αvβ3-blockade.51 The contribution of cell death, versus altered migration alone, to our in vivo observations remains unknown.
The present study does not prove whether radiolabel in the arterial wall 72 hours after treatment is due to retained intact peptide or to persistence of a metabolite or free label. Nevertheless, local arterial wall exposure to peptide for ≤72 hours in this model was sufficient to limit in vivo SMC migration and late intimal thickening.
In this study, we have demonstrated in the rat model that despite expression of both β1- and β3-integrins on postinjury migrating SMCs, only β3-integrin expression is of functional consequence in integrin-matrix interactions involved in SMC migration. As such, pharmacological manipulation of cell integrin-matrix interactions after interventional arterial wall injury may be an additional viable target for limiting injury- induced SMC migration and late arterial thickening and remodeling. In particular, targeting of dominant cell integrin-matrix interactions, ie, β3-integrins, involved in postinjury arterial wall healing may be a valuable approach for limiting injury-induced restenosis.
Selected Abbreviations and Acronyms
|anti-FNR||=||rabbit anti–human α5β1|
|anti-VNR||=||rabbit anti–human αvβ3/5|
|SMC||=||smooth muscle cell|
This work was supported by Arizona Disease Control Research Commission contract No. 9623, a grant from the Simpson Atherectomy Research Foundation, an educational grant from Focal, Inc of Lexington, Mass., and a Dean’s Research Council Award, University of Arizona.
- Received May 27, 1997.
- Revision received October 30, 1997.
- Accepted November 24, 1997.
- Copyright © 1998 by American Heart Association
Califf RM, Fortin DF, Frid DJ, Harlan WR III, Ohman EM, Bengtson JR, Nelson CL, Tcheng JE, Mark DB, Stack RS. Restenosis after coronary angioplasty: an overview. J Am Coll Cardiol. 1991;17:2B–13B.
Casscells W. Migration of smooth muscle and endothelial cells: critical events in restenosis. Circulation. 1993;86:723–729.
Isner JM. Vascular remodeling: honey, I think I shrunk the artery. Circulation. 1994;89:2937–2941.
Ettinger L, Doljanski F. On the generation of form by the continuous interactions between cells and their extracellular matrix. Biol Rev (Lond). 1992;67:459–489.
Ruoslahti E. Integrins. J Clin Invest. 1991;87:1–5.
Beauvais A, Erickson CA, Groins T, Craig SE, Humphries MJ, Theiry JP, Dufour S. Changes in the fibronectin-specific integrin expression pattern modify the migratory behavior of sarcoma S180 cells in vitro and in the embryonic environment. J Cell Biol. 1995;128:699–713.
Kirchberg K, Lange TS, Klein EC, Jungtaubl H, Heinen G, Meyer-Ingold W, Scharffetter-Kochanek K. Induction of β1 integrin synthesis by recombinant platelet-derived growth factor (PDGF-AB) correlates with an enhanced migratory response of human dermal fibroblasts to various extracellular matrix proteins. Exp Cell Res. 1995;220:29–35.
Leavesley DI, Ferguson GD, Wayner EA, Cheresh DA. Requirement of the integrin β3 subunit for carcinoma cell spreading or migration on vitronectin and fibrinogen. J Cell Biol. 1992;117:1101–1107.
Filardo EJ, Brooks PC, Demming SL, Damsky C, Cheresh DA. Requirement of the NPX motif in the integrin β3 subunit cytoplasmic tail for melanoma cell migration in vitro and in vivo. J Cell Biol. 1995;130:441–450.
Clyman RI, McDonald KA, Kramer RH. Integrin receptors on aortic smooth muscle cells mediate adhesion to fibronectin, laminin and collagen. Circ Res. 1990;67:175–186.
Slepian MJ, Massia SP. Local delivery of a cyclic RGD peptide inhibits neointimal hyperplasia following balloon injury. Circulation. 1993;88(suppl I):I-372. Abstract.
Matsuno H, Stassen JM, Vermylen J, Deckmyn H. Inhibition of integrin function by a cyclic RGD-containing peptide prevents neointima formation. Circulation. 1994;90:2203–2206.
Topol EJ, Califf RM, Weisman HF, Ellis SG, Tcheng JE, Worley S, Ivanhoe R, George BS, Fintel D, Weston M, Sigmon K, Anderson KM, Lee KL, Willerson JT, on behalf of the EPIC Investigators. Randomised trial of coronary intervention with antibody against platelet IIb/IIIa integrin for reduction of clinical restenosis: results at six months. Lancet. 1994;343:881–886.
Ross R, Kariya B. Morphogenesis of vascular smooth muscle in atherosclerosis and cell structure. In: Bohr DF, Somiyo AP, Sparks HY, eds. Handbook of Physiology: The Cardiovascular System, Section 2. 1988. Bethesda, Md: American Physiological Society; 1988:69–91.
Dubey RK, Roy A, Overbeck HW. Culture of renal arteriolar smooth muscle cells: mitogenic responses to angiotensin II. Circ Res. 1992;71:1143–1152.
Bolton AE, Hunter WM. The labeling of proteins to high specific radioactivities by conjugation to a 125I-containing acylating agent. Biochem J. 1973;133:529–539.
Pytela R, Pierschbacher MD, Ginsberg MH, Plow EF, Ruoslahti E. Platelet membrane glycoprotein IIb/IIIa: member of a family of Arg-Gly-Asp-specific adhesion receptors. Science. 1986;231:1559–1562.
Albelda SM, Buck CA. Integrins and other cell adhesion molecules. FASEB J. 1990;4:2868–2880.
Hoshiga M, Alpers CE, Smith LL, Giachelli CM, Schwartz SM. αvβ3 integrin expression in normal and atherosclerotic artery. Circ Res. 1995;77:1129–1135.
Liaw L, Skinner MP, Raines EW, Ross R, Cheresh DA, Schwartz SM, Giachelli CM. The adhesive and migratory effects of osteopontin are mediated via distinct cell surface integrins: role of αvβ3 in smooth muscle cell migration to osteopontin in vitro. J Clin Invest. 1995;95:713–724.
Pierschbacher MD, Ruoslahti E. Influence of stereochemistry of the sequence arg-gly-asp-xaa on binding specificity in cell adhesion. J Biol Chem. 1987;262:17294–17298.
Felding-Habermann B, Cheresh DA. Vitronectin and its receptors. Curr Opin Cell Biol. 1993;5:864–868.
Dejena ES, Colella S, Conforti G, Abbadini M, Gaboli M, Marchisio PC. Fibronectin and vitronectin regulate the organization of their respective Arg-Gly-Asp adhesion receptors in cultured human endothelial cells. J Cell Biol. 1987;107:1215–1231.
Brown MJ, Loew LM. Graded fibronectin receptor aggregation in migrating cells. Cell Motil Cytoskeleton. 1966;34:185–193.
Clowes AW, Schwartz SM. Significance of quiescent smooth muscle cell migration in the injured rat carotid artery. Circ Res. 1985;56:139–145.
Topol EJ. Platelet glycoprotein IIb/IIIa receptor antagonists in coronary artery disease. Eur Heart J. 1996;17:9–18.
Topol EJ, Ferguson JJ, Weisman HF, Tcheng JE, Ellis SG, Kleiman NS, Ivanhoe RJ, Wang AL, Miller DP, Anderson KM, Califf RM. Long-term protection from myocardial ischemic events in a randomized trial of brief integrin β3 blockade with percutaneous coronary intervention. JAMA. 1997;278:479–484.
Nip J, Rabbani SA, Shibata HR, Brodt P. Coordinated expression of the vitronectin receptor and the urokinase-type plasminogen activator receptor in metastatic melanoma cells. J Clin Invest. 1995;95:2096–2103.
Edelman ER, Adams DA, Karnovsky MJ. Effect of controlled adventitial heparin delivery on smooth muscle cell proliferation following endothelial injury. Proc Natl Acad Sci U S A. 1990;87:3773–3777.
Slepian MJ, Hubbell JA. Polymeric endoluminal gel paving: hydrogel systems for local barrier creation and site-specific drug delivery. Adv Drug Deliv Rev. 1997;24:11–30.
Barker PL, Bullens S, Bunting S, Burdick DJ, Chan KS, Deisher T, Eigenbrot C, Gadek TR, Gantzos R, Lipari MT, Muir CD, Napier MA, Pitti RM, Padua A, Quan C, Stanley M, Struble M, Tom JYK, Burnier JP. Cyclic RGD peptide analogues as anti-platelet anti-thrombotics. J Med Chem. 1992;35:2040–2048.