Histomorphometric and Biochemical Correlates of Arterial Procollagen Gene Expression During Vascular Repair After Experimental Angioplasty
Background To determine the transcriptional, biochemical, and histomorphometric correlates of neointimal procollagen accumulation during arterial repair after balloon angioplasty of atherogenic vessels, rabbit iliac artery collagen content and the induction of α1(I) and α1(III) procollagen mRNA were assessed in normal vessels and at 2, 7, and 30 days after angioplasty.
Methods and Results Quantitative iliac artery histomorphometric neointimal collagen analysis was performed using a specific picrosirius red stain under polarized light. Arterial cross-sectional area reduction, total cellularity, and vascular smooth muscle cell density (per 104 μ2 of neointima) were quantified in routine and immunohistochemically stained sections (α-actin and RAM-11), from which biochemical concentrations of tissue protein, RNA, and DNA were also measured. Collagen comprised 0.23±0.1 mg/mg of total protein in the normal vessel wall and did not increase in vessels studied 2 and 7 days after angioplasty (0.26±0.06, 0.28±0.05 mg/mg of protein, P=NS). By 30 days after angioplasty, >50% of the protein concentration was collagen (0.55±0.11 mg/mg of protein, P=.02). Collagen-positive histological staining also increased significantly from 17±2% of the neointima at day 2 to 32±5% by day 30 (P=.01). The transcript regulatory signal for α1(I) procollagen mRNA was induced 2 days after angioplasty, peaking at 7 days for both α1(I) and α1(III), and returning to control levels 30 days after angioplasty. A significant luminal cross-sectional area reduction of the arterial wall was confirmed both by angiography and histomorphometry (P=.01). This was not associated with a significant change in α-actin (+) vascular smooth muscle cell density (38±7 nuclei per 104μ2 at day 2 and at day 30) or tissue DNA concentration (P=NS).
Conclusions We conclude that procollagen genes are transcriptionally activated early (2 to 7 days) after angioplasty vessel injury and that collagen subsequently constitutes a major biochemical and histological component of the proliferative neointima by 30 days after angioplasty. Alterations in pathways regulating procollagen metabolism may also contribute to the accumulation of extracellular matrix and growth of the neointima in the late repair phase after vessel wall injury.
Collagen is an important vascular extracellular matrix protein, providing architectural stability and tensile strength, while acting as a powerful promoter of platelet aggregation after endothelial disruption.1 After angioplasty, the disproportionate restenotic response to vascular injury is characterized by vascular smooth muscle cell (VSMC) migration and proliferation followed by a cellular shift from a contractile to synthetic phenotype, with abundant extracellular matrix accumulation contributing to luminal reduction and limitation of blood flow.2 Although the biological characteristics of neointimal vascular smooth muscle cells, endothelial cells and blood-borne elements (including platelets) have been intensively studied after angioplasty, considerably less is known about the temporal evolution of and molecular signaling for extracellular matrix elaboration during vessel repair after barotrauma injury.3
The predominant collagen subtypes found in the normal human arterial wall are type I (70% to 75%) and type III (20% to 25%), with type V comprising 1% to 2%.1 Minor alterations occur in the relative proportions of these proteins in vascular disease states such as atherosclerosis.4 The interaction between the collagenous extracellular matrix and VSMCs is important in the biology of the developing atheroma.5 Phenotypic modulation of VSMCs from a quiescent/contractile to a migratory/synthetic state may initially be associated with the detachment of cells from the surrounding matrix, with increased mitosis and migration of cells from the media toward the growing neointima.6 7 Synthesis of matrix macromolecules and fibrous plaque constituents occurs in concert with lipid deposition.2 8
Accelerated rates of collagen biosynthesis have been reported in atherosclerotic tissues in vivo.9 10 11 Recent reports have suggested that accumulation of extracellular matrix tissue after angioplasty is delayed until phase 3 of arterial repair (day 14 to 3 months), potentially mediated by growth factors such as transforming growth factor (TGF-β).12
The objective of the current study was to relate the degree and time-course of induction of collagen mRNA to the biochemical alterations of collagen concentration occurring during vascular repair after angioplasty. The detailed quantitative histomorphometric characteristics of the neointima including VSMC density and collagen-specific picrosirius red staining, were also derived at identical time points after angioplasty. It was hypothesized that augmented transcriptional activation of α1(I) and α1(III) procollagen would precede the accumulation of fibrillar collagens in the late stages of arterial repair. Posttranslational alterations in the metabolic characteristics of the diseased vessel (for example, changes in the degradation of collagen), which may also be responsible for matrix accumulation, were not evaluated.
Materials and Chemical Reagents
High cholesterol rabbit chow (2% cholesterol with 6% peanut oil) was supplied by ICN Biomedical. Vascular injury was created using a 3F Fogarty arterial embolectomy catheter (Baxter). Balloon angioplasty catheters (2.5 mm) were provided by Bard Co USCI. Angioplasty determination of vessel diameter was performed using digital calipers (Sandhill Scientific Inc). Sectioned arterial segments were pressure-fixed (at 80 mm Hg) in 10% buffered formalin, processed through graded alcohols and xylene, and embedded in paraffin. Each specimen was sectioned 5 μm thick, and contiguous sections were stained with hematoxylin and eosin, rabbit antimacrophage (RAM-11), picrosirius red, and smooth muscle–specific anti–α-actin antibodies (Shandon, Inc; prediluted: catalog No 460105). The latter staining method used the labeled streptavidin-biotin (LSAB) method using 3-amino-9-ethylcarbazole (AEC) as the chromogen. Positive and negative controls were stained appropriately. Tissue histomorphometric analysis and planimetry were done using a computerized Bioquant analysis program (R & M Biometrics, Inc). All chemical reagents used were of the highest grade commercially available and were obtained from Sigma Chemical Co, Pierce Chemical Co, and Baxter Scientific Products. L-[2,3-3H] hydroxyproline and [α-32P] dCTP were obtained from Amersham. All cDNA probes were obtained from the American Tissue Type Collection, Bethesda, Md.
Experimental Design of the Study
Male New Zealand White rabbits (Doe Valley Industries) weighing 3000 to 3200 g were acclimated to the St Louis University School of Medicine’s Comparative Medicine animal care facility for at least 1 to 2 days before initiating feeding with high cholesterol chow for 1 week before the creation of denuding vascular injury. After vascular endothelial stripping, high cholesterol chow was continued for another 2 weeks, at which point balloon angioplasty of both iliac arteries was performed and serum cholesterol was determined.13 At this stage, animals were randomized into three groups, according to the time they were killed: days 2, 7, and 30 (3 animals in each group). Another groups of 3 animals was killed without any interventions to serve as normal controls. Weights of all the animals were recorded at the time of arrival, before modeling, at angioplasty, and finally before they were killed.
All surgical procedures were performed between 8:00 am and noon (to control for any diurnal variations) and under strict aseptic conditions. Bilateral iliac artery stripping was performed by initially anesthetizing rabbits with an intramuscular injection of ketamine (35 mg/kg), xylazine (4 mg/kg), and ace-promace (0.1 mg/kg). Groin areas were shaved and prepared for cutdown incisions. A 1.5-cm segment of the distal femoral artery was dissected free of surrounding fascia and neurovascular elements. The femoral artery was secured by string sutures and an arteriotomy performed using a microdissecting scissor. A 3F balloon artery fogarty catheter was introduced into the vessel and passed up to the 20 cm mark on the catheter. Two retracting passes were made with the balloon inflated to denude the vessel.13 At the completion of the procedure, sutures proximal and distal to the arteriotomy were tied off. The skin incision was closed with continuous subcutaneous sutures.
On the day of angioplasty, the rabbits were initially anesthetized lightly with 3 mL of intravenous ketamine (20 mg/kg), followed by endothracheal intubation (3.5F endotracheal tube) and mechanical ventilation during the remainder of the procedure. Further anesthesia was maintained by delivery of gas anesthetic (halothane/oxygen) through the endotracheal tube. The proximal femoral arteriotomy was performed in similar fashion to the denuding injury. A 22-gauge catheter was initially introduced in the vessel, and a preangioplasty angiogram was obtained. The site of focal iliac artery narrowing was identified and subsequently angioplastied using a 2.5-mm balloon catheter. Each iliac artery was dilated twice at 6.0 atm of pressure for a minimum of 60 seconds each. Angiograms were obtained to verify both the position of the balloon catheter before dilatation as well as at the completion of the dilatation to confirm luminal enlargement of the vessel. Before arterial cannulation, 500 U of intravenous heparin and 150 mL of normal saline were also administered to the animal. Nitroglycerin (250 μg IA) was administered before postangioplasty angiography to reverse spasm.
On the day of the euthanization, after light anesthesia, intubation, and ventilation, both femoral arteries were exposed in a fashion similar to the angioplasty. A midline neck dissection was then performed to expose the right carotid artery, and a 4.0F Swan-Ganz catheter was introduced via this vessel and positioned in the abdominal aorta, just distal to the renal arteries. A preeuthanization angiogram of both iliac arteries was obtained by injection contrast dye (5 to 10 mL) through the distal port of this catheter. After the angiogram was obtained, further dissection to expose the distal abdominal aorta and its bifurcation of the iliacs was accomplished. The right iliac artery was ligated, rapidly removed, snap-frozen in liquid nitrogen, and stored at −80°C until further analysis. The animals were euthanized with an overdose of sodium pentothal. The left iliac artery was pressure-fixed (80 mm Hg) with formalin for 24 to 36 hours for further histomorphometric analysis.
Determination of Biochemical Composition of Normal and Diseased Arterial Segments
Precisely weighed (25 mg wet wt) tissue specimens from the iliac artery of the control and the angioplasty-injured vessels were homogenized with 475 μL of milliQ-H2O, using a Polytron PCU1 homogenizer fitted with a microgenerator, to make a 5% tissue homogenate.
Hydroxyproline Determination Two hundred microliters of the homogenate was removed to analyze the tissue hydroxyproline (HYP) concentration as an indirect estimate of the total tissue collagen concentration (HYP comprises 13.4% of collagen).14 Precipitation of tissue proteins was achieved by adding 400 μL of cold 100% ethanol (final concentration, 67% vol/vol) to the homogenate, which was kept on ice for 1 hour. Ethanol- precipitated tissue was then centrifuged for 20 minutes at 12 000g using a microcentrifuge. The supernatant was evaporated and reconstituted in 100 μL of milliQ-H2O. Both the supernatant and pellet were then reconstituted in 2 mL of 6N HCL and hydrolyzed (100°C, 24 hours). Before the initiation of hydrolysis, each sample was also spiked with 178 000 dpm of [3H]-hydroxyproline to serve as an internal standard for the quantitative recovery of HYP. After hydrolysis, hydrolysates were transferred into 12×75-mm glass tubes, and 300 μL of charcoal slurry (240 mg/mL) was added; this was vortexed thoroughly and centrifuged at 2000g for 10 minutes. Clear supernatant was then transferred into a evapomix tube and later dried in the evapotech drier and reconstituted in 1.0 mL of milliQ-H2O. A 100-μL aliquot from this resuspended sample was taken for [3H]-hydroxyproline recovery to gauge for count losses and correct for HYP recovery. Two 50-μL aliquots were taken to perform Woessner’s colorometric assay to determine tissue HYP concentrations.15 Data were expressed as mg of HYP/g of arterial wet weight.
Determination of Total Tissue Protein, DNA, and RNA Concentrations
Another 200 μL of the homogenate was precipitated with 2 mL of 0.2N PCA (percloric acid) and kept on ice for 1 hour. After centrifugation at 12 000g for 10 minutes, the supernatant was discarded and the pellet washed once with a further 2 mL of 0.2N PCA. The pellet was subsequently dissolved in 1 mL of 0.3N KOH and heated at 60°C for 1 hour. Fifty microliters of the KOH-dissolved sample was added to 950 μL of 0.15N NaCl (in a 1/20 dilution), and quantitative measurements of total tissue protein content were done by the method of Lowry et al.16 Further, two aliquots of 100 μL were taken, and total tissue DNA concentration was assessed using a minor modification of the microfluorometric technique DNA analysis described by Cesarone et al.17 The remaining sample was reprecipitated with 500 μL of 1.2N PCA, and total tissue RNA concentration was determined by the method of Munro and Fleck.18 Results were expressed as mg/g of arterial wet weight.
Total RNA was isolated from frozen arterial segments according to the method of Chomczynski and Sacci.19 RNA was quantified by absorbance at 260 nm, and the integrity was determined by examining the 28S and 18S rRNA bands in ethidium bromide–stained agarose gels. Total RNA was size-fractionated by electrophoresis under denaturing conditions, transferred to nylon membranes, and immobilized by UV irradiation. Equal amounts of RNA (5 μg) were loaded in each lane, quantified by their absorbance at 260 nm. The 260/280 OD ratio was also assessed to confirm the purity of the loaded RNA before running the agarose gel electrophoresis. Prehybridization and hybridization with an oligonucleotide probe were performed at 51°C in a solution containing 6×SSC, 1×Denhard’s solution, 0.1% SDS, 0.005% Na pyrophosphate, 20 μg/mL tRNA, and 100 μg/mL denatured salmon sperm DNA. Blots were washed in 6×SSC–0.1% SDS (65°C, 30 minutes). All blots were exposed to Kodak XAR-5 film with intensifier screens at −80°C for varying time periods. Band intensity was quantified by laser densitometry (LKB Ultrascan XL Enhanced Laser Densitometer interfaced to a personal computer running Gelscan XL version 1.2). The following cDNA clones were used: human collagen type I, α-1 (α1[I]), clone Hf67720 ; human collagen type III, α-3 (α1[III]) clone Hf93421 ; and human glyceraldehyde 3-phosphate dehydrogenase (GAPDH), clone pHcGAP.22 Hybridization conditions and washing steps were similar for all the probes used. The inserts were isolated and labeled by random primer extension with (α-32P)dCTP. Data were expressed as the densitometric intensity of the hybridization signals for α1(I) and α1(III) mRNAs relative to the constitutively expressed GAPDH mRNA.
Tissue Histomorphometric Analysis
Formalin-treated, pressure-fixed sections of the iliac artery were obtained, mounted on paraffin blocks, and sectioned 5 μm thick for histology. Sections were stained with routine hematoxylin and eosin and picrosirius red stains. The vector immunoperoxidase system with mouse anti–α-actin (SMC), and rabbit antimacrophage (RAM-11) monoclonal antibodies were used to identify the VSMCs and foam cell populations in these sections. Results were expressed as % cross-sectional area reduction and as total cell, RAM-11 (+), and α-actin (+) cell density per 104μ2 of neointimal area.
A computer software analysis program with digital microscopic planimetry was used to serially measure areas in the vessel wall including the vessel lumen, area of neointima, total α-actin (+), and RAM-11 (+) nuclei. A total of three to four iliac artery cross sections were mounted on each slide with each stain (picrosirius red, RAM-11, α-actin). Two slides were evaluated for each stain (a total of six to eight vessel cross sections). Percent neointimal collagen area and cellularity per 104μ2 were averaged from the two most severely narrowed vessel cross sections. Variation of these parameters within a given animal was <10%. Tissue collagen was identified with picrosirius red staining under polarized light (Fig 1⇓). The threshold level in picrosirius red polarization studies was standardized by evaluating the video pixel intensity within an area of maximum collagen birefringence using the adventitia of an unmodeled vascular control specimen stained with picrosirius red. The threshold pixel intensity of the control specimen was recorded, and all subsequent vascular specimens were peak to the same intensity level by the computer. Before this analysis, the counterpolarizing filters were adjusted to assure maximum birefringence of collagen in the neointima. Percent collagen staining was derived by dividing the threshold number by the total area of the neointima.
Immunostaining With Anti–Type I and Type III Collagen Antibodies
Type I and type III collagen were histologically localized in the neointima using vector ABC immunoperoxidase staining with diethyl immunoperoxidase (DAB) as the substrate. Type-specific antibodies used included a monoclonal antibody against human type III collagen IF8 (Ig-1) (provided by Dr Maurice Godfrey, Connective Tissue Disease Research Laboratory, University of Nebraska Medical Center, Omaha, Neb) and a polyclonal anti-human type I collagen antibody (Biodesign International, Kennebunk, Me). Both antibodies were initially tested by Western blot analysis against rabbit cardiac fibrillar collagen obtained by cyanogen bromide cleavage and SDS/PAGE separation of type I and III collagen fragments.
After incubation with goat serum for 20 minutes to block nonspecific binding with antigens, slides were incubated as in the antibody group with the primary specific antibody of interest, and the controls were incubated with rabbit serum for a period of 60 minutes. All subsequent steps including addition of the biotinylated secondary antibody, the vectastain ABC reagent, and substrate were similar in both groups. Slides were cross-stained with hematoxylin and mounted with crystal mount.
Iliac Artery Angiography and Angiographic Determination of Vessel Diameter
Iliac artery diameters were measured using digital calipers by a blinded expert observer before and after angioplasty and before the animals were killed. Arterial segments measured (in duplicate) included diameters proximal and distal to the angioplasty site and two sites within the angioplasty zone per se. An angiography background grid was used to correct for magnification artifacts. Angiographic results were expressed as preangioplasty to postangioplasty percentage (%) diameter stenosis of the angioplasty zone of the involved vessel between the postangioplasty study and time the animals were killed.
Measurement of Serum Cholesterol Levels
Approximately 2 weeks after being on the high cholesterol chow, rabbit ear veins were bled, and 5 mL of blood was collected to measure serum cholesterol levels using a Ciba Corning Express 550 Clinical Chemistry Analysis Machine. Since cholesterol levels were usually very high (>1000 mg/d) and beyond the range of standard testing, serial dilutions of the serum were performed before analysis. Cholesterol levels were expressed as mg/dL of serum.
Unless otherwise stated, all results were expressed as mean±1 SD. Normality was assessed using the Wilk-Shapiro test, and homogeneity of variance was established with Levine’s test. One-way ANOVA was used to compare the time-dependent variables. Further statistical testing was performed by unpaired t test. Data were analyzed using the prophet computer system (Division of Research Resources, NIH). A probability value of <.05 was required to reject the null hypothesis.
Effects of Diet on Body Weight and Serum Cholesterol Levels
All animals used in these experiments were adult rabbits of similar weight at the time of inclusion in the study (3000 to 3200 g). No significant gain in body weight of the rabbits was noted with aging or feeding high fat/cholesterol diet. At the time the animals were killed, some showed evidence of fatty liver infiltration on gross examination. One animal lost approximately 200 g of weight after the vascular injury. No surgical mortality was noted in any group.
Cholesterol levels ranged between 1200 and 2000 mg/dL. Cholesterol levels measured in animals after resumption of normal diet continued to be two- to threefold elevated compared with controls.
Effect of Balloon Angioplasty on Biochemical Composition of Arterial Tissue
Tissue RNA and DNA Concentration
Total tissue DNA and RNA concentrations were expressed as corrected mg/g weight of tissue in both normal and angioplastied vessels at the specified time points. The ratio of arterial tissue total RNA to protein concentrations (Fig 2⇓) in mg/g of wet weight tissue was significantly increased at day 2 (P=.03), day 7 (P=.002), and at day 30 (P=.01) after intervention compared with control levels, indicating enhanced tissue capacity for protein synthesis. The total tissue DNA concentration (Fig 3⇓) was also increased significantly (P<.05) compared with control animals but remained constant between days 2 and 30 (P=NS).
Tissue Collagen Concentration
Total tissue collagen concentration (Fig 4A⇓), as derived from the tissue HYP concentration, was essentially unchanged from normal control levels and at 2 and 7 days after angioplasty. However, 30 days after angioplasty, the arterial tissue collagen concentration had approximately doubled, indicating disproportionate accumulation of collagen at this stage of postinjury repair (P=.02). Arterial collagen became the most abundant protein of the vessel wall, comprising greater than 50% (Fig 4B⇓) of the vascular protein content by 30 days after angioplasty.
Effect of Balloon Angioplasty on Fibrillar Procollagen Gene Expression
Messenger RNA levels for α1(I) and α1(III) procollagen polypeptides were quantified in arterial segments at 2, 7, and 30 days after angioplasty and compared with the level of mRNA expression in control arterial segments. As is evident in Figs 5⇓ and 6⇓, mRNA levels for both type I and III procollagens (relative to the constitutively expressed mRNA encoding GAPDH) were substantially increased within 7 days after balloon angioplasty. These transcript levels returned to control levels 4 weeks after balloon injury.
Immunohistological Staining With Collagen-Specific Antibodies
Qualitative assessment of the immunostained vascular sections revealed anticollagen types I and III antibody staining within the neointima, which was predominantly localized in a location deep to the subendothelial vascular smooth muscle cell layer. Intense type I collagen staining was present adjacent to neointomal macrophages and the internal elastic lamina. Less intense immunostaining with both anticollagen antibodies was apparent in the loose matrix tissue interspersed between VSMCs. An example of intense immunostaining using an antibody to type I collagen is given in Fig 8⇓. Sporadic staining was observed adjacent to VSMCs with type III collagen antibody at 30 days after angioplasty.
Effect of Balloon Angioplasty on Vascular Histomorphometry
Growth of the neointima with associated luminal encroachment is described as percent cross-sectional area reduction, measured by dividing the area of intimal proliferation by area of intimal proliferation plus the area of the vessel lumen (Table 1⇓). This reduction increased from 0.19±0.07 at day 2 to 0.60±0.13 at 30 days after angioplasty (P=.01). Growth of the neointima was not associated with a similar increase in VSMCs or total cell density between days 2 and 30 (P=NS). Total number of cells did not differ significantly (data not shown) from the combined total of neointimal VSMCs and macrophage foam cells, indicating that very few other cell types (for example, fibroblasts) populated the neointima. The neointimal picrosirius red staining collagen content confirmed a significant increase in the collagen matrix content from 17±2% at day 2 to 32±5% at 30 days after angioplasty (P=.01).
Correlation Between Biochemical and Histological Collagen Concentration
To assess the relative correlation between the postangioplasty vascular neointimal collagen concentration derived by biochemical assays and that derived histologically (computerized digital thresholding), the values obtained from the two techniques in the same animals were correlated by linear regression. As shown in Fig 7⇓, a fair correlation (r2=.65) was seen over the 0- to 4-week period after angioplasty. As such, histological quantitation of collagen in the vascular segment with the most severely reduced cross-sectional area within the angioplasty zone is indicative of biochemical vascular collagen content, although these disparate techniques are not directly comparable.
Effect of Balloon Angioplasty on Iliac Artery Luminal Diameter
Angiography of iliac arteries in animals killed at earlier time points (days 2 an 7) did not reveal acute thrombotic closure or significant focal stenosis (Table 2⇓). The percent luminal diameter narrowing of vessels at later time points was diffuse and was most evident in the distal segments of the angioplasty zone. By day 30, there was a 28% to 35% mean luminal diameter reduction in angioplasty segments as compared with noninjured segments (P<.001).
Proliferative neointimal repair of injured vascular segments leads to postangioplasty restenosis, a major limitation of interventional procedures.25 26 The pathogenesis of this exuberant vascular healing process remains to be elucidated.25 Although many previous postmortem and experimental studies have shown that such hyperproliferative lesions have a high content of extracellular matrix, the quantitative and temporal relations between neointimal cellular proliferation and matrix biosynthesis have not been elucidated. In this report, a validated experimental model of accelerated atherosclerosis was used to assess the effects of a controlled vascular injury on the induction of procollagen gene expression at serial time points during postangioplasty arterial repair. Simultaneous biochemical, histomorphometric, and angiographic correlates of procollagen gene induction were quantitated in order to characterize postangioplasty neointimal matrix elaboration. An apparent alteration in the metabolic control of vascular interstitial fibrillar procollagen turnover led to the accumulation of collagen.
Our results indicated that after balloon angioplasty, mRNA levels encoding fibrillar procollagen polypeptides are increased as early as 2 to 7 days after vascular injury. These alterations in mRNA levels may be due to transcriptional activation of procollagen gene expression by smooth muscle cells and/or fibroblasts within the vessel wall. It is likely that this increase in fibrillar procollagen α1(I) and α1(III) mRNA is responsible for the increased collagen content of balloon-injured vessels observed 30 days after angioplasty. However, it is conceivable that other factors, including translational and posttranslational processing of newly synthesized procollagens, may have contributed to the accumulation of extracellular matrix proteins within the vascular segment. Although this investigation has defined the principal transcriptional events preceding increased collagen types I and III biosynthesis after vascular barotrauma injury, the posttranslational control of these matrix proteins has yet to be confirmed.
Liau and Chan28 have shown that VSMC cultures in serum-poor media demonstrate positive induction of collagen type III mRNA genes in a manner that is tightly coupled with cellular quiescence and a contractile phenotype. In our in vivo study, no correlation was observed between the degree of luminal patency and induction of type III mRNA genes. The degree of expression of these genes may not, therefore, reliably predict an attenuated restenotic biological response in vivo after angioplasty. Majesky and colleagues29 have shown in a rat carotid injury model that induction of extracellular matrix genes occurs in concert with the formation of neointima. This process requires the synthesis and assembly of a new extracellular matrix, in particular, types I and III collagen, fibronectin, and elastin. It has been postulated that progressive intimal thickening may be stimulated in an autocrine fashion by growth regulatory factors such as TGF-β and PDGF.
In a postmortem clinical histopathological study by Nobuyoshi et al,3 restenotic lesions of human coronary arteries had an abundance of extracellular matrix coupled with synthetic-appearing VSMCs. Using routine histology, in the late stages (beyond 6 months) the extracellular matrix within the neointima was much less abundant. By 2 years after the procedure, the histological findings were indistinguishable from that of a conventional atherosclerotic plaque. However, descriptions of vascular matrix constituents using routine staining and histological analysis from a single tissue section, without quantitation, special staining, or correlative biochemical data, may be limited.
Previous investigators have postulated that an intricate metabolic balance exists between intracellular procollagen synthesis, degradation (both intracellular and extracellular), and accumulation.30 Enhanced synthesis of this protein in cardiac tissue (as in thyrotoxicosis) is accompanied by a severalfold increase in both rapid intracellular and slower extracellular degradation of procollagens, without disproportionate accumulation of total collagen.14 In the present study, 30-day postangioplasty collagen mRNA signals were similar to basal levels, while corresponding biochemical and histological tissue collagen content was disproportionately higher, suggesting an imbalance in procollagen metabolism in vivo. The concordant early rise in the total tissue concentration of both RNA and DNA does reflect enhanced capacity for protein synthesis and cellular multiplication. Significant tissue polyploidy exists in the vessel wall, with extensive hyperplasia occurring in both the VSMCs and the nonmuscle cell population very early after angioplasty. While the phenomenon may clearly be multifactorial, including VSMC hyperplasia and hypertrophy and macrophage infiltration with release of mitogens and autocrine growth factors, a quantitative change in the density of VSMCs or foam cells was not observed (Table 1⇑).
Our data are similar in many respects to the Clowes (rat carotid denudation)31 and Hanke (rabbit carotid electrical injury)32 studies. Clowes et al demonstrated that smooth muscle cell proliferation peaked at 48 hours in the media and at 96 hours in the intima and that significant smooth muscle cell proliferation persisted at the intimal surface for 12 weeks. Tritiated thymidine incorporation declined to baseline at 4 weeks in the media and 8 weeks in the intima. We have demonstrated a significant early increase in vascular tissue DNA concentration early after angioplasty, which remained constant between days 2 and 30. Clowes et al also demonstrated no change in total arterial smooth muscle cell number between 2 and 12 weeks after denudation.
Hanke et al demonstrated a relatively linear increase in the number of intimal smooth muscle cells as intimal thickening (measured as cell layers) occurred but no significant change in VSMC density over time. Bromodeoxyuridine labeling of intimal cells demonstrated a peak in DNA synthesis at 3 and 7 days after electrical injury, with a return to baseline levels by 28 days. Our measurement of DNA concentration also increased significantly at days 2 to 7 after angioplasty and did not decrease thereafter. Our measurement of DNA concentration is not identical to the measurement of DNA synthesis by Hanke et al, possibly contributing to the differences in our day-30 observations. Also, vascular injury (angioplasty versus electrical stimulation) and repair may not have been identical in these models.
Despite the recognized limitations associated with the use of small animal models for the study of clinical postangioplasty-accelerated atherosclerosis, extracellular matrix deposition is a constant histological feature of both human and animal neointimal lesions. It is conceivable that the transcriptional signals controlling collagen biosynthesis are largely preserved between lower and higher species. The ease and rapidity in which the atheroproliferative process is reliably reproduced makes rabbit iliac artery angioplasty an extremely useful model to study vascular repair and restenosis.
Liptay et al33 have reported that neointimal macrophages colocalize with type I collagen. Figs 1⇑ and 8⇓ demonstrate a similar phenomenon early (day 7) after vascular injury. As the neointima matured (day 30), collagen was qualitatively distributed with subendothelial α-actin–positive VSMCs. Reckhter et al34 have observed no spatial correlation between macrophages and type I procollagen in mature human atheromata. This issue remains controversial, and our observations are not definitive.
This study has also not examined the relative biochemical alterations in the tissue composition of types I and type III collagens after angioplasty.27 Rather, the gross total tissue collagen concentration was derived from vascular hydroxyproline concentrations, which may have failed to detect minor differences in the amounts of types I and III collagen proteins after vascular injury. Atherosclerosis, a related disease process involving VSMC hyperplasia and enhanced collagen biosynthesis, occurs without any gross alteration in the relative proportions of the individual collagen subtypes observed in vivo.4 The relatively small increase in type I collagen in atherosclerosis is probably not the result of VSMC “transformation” to a fibroblastic cell type, nor is it associated with a predominance of type III collagen.
Histomorphometric measures of VSMCs, foam cells, and collagen density from the single worst histological vascular sections may not accurately reflect changes occurring in the immediately adjacent tissue sections. However, this approach has defined the major cellular and acellular characteristics of the angioplasty lesion at the site of the vessel wall of maximal neointimal reaction. The suboptimal (r2=.65) correlation between two indices of tissue collagen content (that is, picrosirius red polarization and biochemical hydroxyproline concentration) may reflect the methodological diversity of these analysis in an in vivo experimental setting.
In this validated atherogenic model of angioplasty vascular injury, serial induction of transcriptional signals for α1(I) and α1(III) procollagen and mRNA occurred and peaked at 7 days after angioplasty, preceding significant biochemically and histologically derived increases in neotinimal collagen matrix content. The simultaneous progression of angiographic luminal narrowing (“restenosis”) was not accompanied by an increase in vascular neointimal cell density in this model.
Future studies should be directed toward analyzing the turnover of extracellular matrix proteins (for example, collagen) in arterial tissue and identifying the effect on procollagen gene induction of selected pharmacological interventions. Little information is presently available concerning the cell types, proteolytic enzymes, and regulatory elements involved in the in vivo processing of collagen vascular tissue. The interaction between synthetic and degradative pathways may have relevance in the pathogenesis of accelerated atherosclerosis and vascular restenosis.
This study was supported by a grant provided by the American Heart Association, Missouri Chapter (grant G019), and NHLBI grant HL-43582. The authors extend special thanks to Dr Michael J. Davies, Professor of Cardiovascular Pathology, St George’s Hospital Medical School, London, for his assistance in providing the picrosirius red stain. The authors also thank Lori Gallini for providing excellent secretarial assistance.
Presented in part at the 42nd Annual Scientific Session of the American College of Cardiology (Young Investigator’s Award), March 1993, Anaheim, Calif.
- Received August 8, 1994.
- Revision received November 10, 1994.
- Accepted November 26, 1994.
- Copyright © 1995 by American Heart Association
Davies MJ, Woolf N, Katz DR. The role of endothelial denudation injury, plaque fissuring and thrombosis in the progression of human atherosclerosis. Atherosclerosis Rev.. 1991;23:105-113.
Nobuyoshi N, Kimura T, et al. Restenosis after PTCA: pathologic observation in 20 patients. J Am Coll Cardiol. 1991;17(2):433-439.
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.
Campbell RG, Campbell JH. The phenotypes of smooth muscle expressed in human atheroma. Ann N Y Acad Sci. 1990;6:143-158.
Raines E, Ross R. Mechanisms of plaque formation: cellular changes and the possible role of growth regulatory molecules. Atherosclerosis Rev. 1991;23:143-152.
Davies MJ, Woolf N. Atheroma: Atherosclerosis in Ischemic Heart Disease. Vol 1: The Mechanisms. London Science Press; 1990.
Opsahl WP, Deluca DJ, Ehrhart LA. Accelerated rates of collagen synthesis in atherosclerotic arteries quantitated in vivo. Arteriosclerosis. 1987;7:470-476.
Karim MA, Ferguson A, Wakim B, Samarel AM. In vivo collagen turnover during the development of thyroxine-induced left ventricular hypertrophy. Am J Physiol. 1991;260:C316-326.
Woessner JF. The determination of hydroxyproline in tissues and protein samples containing small proportions of this amino acid. Arch Biochem Biol. 1961;93:440-447.
Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein estimation with the folin phenol reagent. J Biol Chem. 1951;193:265-275.
Munro HN, Fleck A. The determination of nucleic acids. In: Glick, ed. Methods of Biochemical Analysis. New York, NY: Interscience; 1966:113-176.
Chu ML, Myers JC, Bernard MP, Ding JF, Ramirez F. Cloning and characterization of five overlapping cDNAs specific for the human proα1(I) collagen chain. Nucleic Acids Res. 1982;10:5925-5934.
Chu ML, Weil D, deWet D, Bernard M, Sippola M, Ramirez F. Isolation of cDNA and genomic clones encoding human pro-α1(III) collagen. J Biol Chem. 1985;260:4357-4363.
Tso J, Sun XH, Kao T, Reece KS, Wu R. Isolation and characterization of rat and human glyceraldehyde-3-phosphate dehydrogenase cDNAs: genomic complexity and molecular evolution of the gene. Nucleic Acids Res. 1985;13:2485-2502.
Junqueira L, Bignollas G, Brentani RR. Picrosirius staining plus polarization microscopy, a specific method for collagen detection in tissue sections. Histochem J. 1979;42:41-51.
Dressler FA, Karim MA, Miller DD. Immunohistochemistry of postangioplasty neointima: cellularity in the wake of the adluminal proliferative smooth muscle cell ‘wavefront.’ Circulation. 1992;86(suppl I):I-799. Abstract.
Casscells W. Migration of smooth muscle and endothelial cells: critical events in restenosis. Circulation. 1992;86:723-729.
Holmes DR, Vliestra RE, Smith HC, Vetrovec GW, Kent KM, Cowley MJ, Faxon DP, Gruentzig AR, Kelsey SF, Detre KM, Van Raden MJ, Mock MB. Restenosis after PTCA: a report from the PTCA registry of NHLBI. Am J Cardiol. 1984:53:77C-81C.
McCullagh KG, Duance VC, Bishop KA. The distribution of collagen types I, III and V in normal and atherosclerotic human aorta. Am J Pathol. 1980;130:45-55.
Liau G, Chan LM. Regulation of extracellular matrix RNA levels in cultured smooth muscle cells. J Biol Chem. 1989;264:10315-10320.
Majesky MW, Lindner V, Twardzik DR, Schwartz SM, Reidy MA. Production of transforming growth factor β1 during repair of arterial injury. J Clin Invest. 1991;88:904-910.
Laurent G. Dynamic state of collagen: pathways of collagen degradation in vivo and their possible role in regulation of collagen mass. Am J Physiol. 1987;252:C1-C9.
Hanke H, Strohschneider T, Oberhoff M, Betze E, Karsch VR. Time course of smooth muscle cell proliferation in the intima and media of arteries following experimental angioplasty. Circ Res. 1990;67:651-659.
Liptay MJ, Parks WC, Mecham RP, Roby J, Kaiser LR, Cooper JD, Botney MD. Neointimal macrophages co-localize with extraacellular matrix gene expression in human atherosclerotic pulmonary arteries. J Clin Invest. 1993;91:588-594.