1
Wei et al. (R96-832/R1)
Temporally and Spatially Coordinated Expression of
Cell Cycle Regulatory Factors After Angioplasty
Geeyeon Laura Wei, BA; Kevin Krasinski, BA; Marianne Kearney, BS; Jeffrey M. Isner, MD; Kenneth Walsh, PhD; Vicente Andrés, PhD
From the Departments of Medicine (Cardiology) and Biomedical Research, St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston, MA 02135
Running title: Cell cycle control after angioplasty.
Corresponding author:
Vicente Andrés, PhD
St. Elizabeth’s Medical Center
736 Cambridge Street
Boston, MA 02135
Tel: (617) 562-7509
Fax: (617) 562-7506
E-mail:
Abstract
Intimal hyperplasia following angioplasty results in part from the migration and proliferation of vascular smooth muscle cells (VSMCs). However, the cell cycle regulatory networks underlying injury-induced VSMC proliferation are largely unknown. In the present study, we examined the kinetics of expression and activity of cell cycle regulatory factors after angioplasty in rat and human arteries. Cell lysates were prepared from uninjured rat carotid arteries and at different time points after balloon denudation. Marked induction of the proliferating cell nuclear antigen (PCNA), the G1/S cyclin-dependent kinase cdk2, and its regulatory subunits, cyclin E and cyclin A, occurred between 1-2 d following angioplasty, was sustained up to 10 d post-injury and then declined. Induction of these factors correlated with increased cdk2-, cyclin E- and cyclin A-dependent kinase activity, indicating the assembly of functional cdk2/cyclin E and cdk2/cyclin A holoenzymes in the injured arterial wall. Immunohistochemical analysis revealed early expression of cdk2, cyclin E and PCNA within the media of injured carotid arteries. At later time points, expression of these markers declined to basal levels in the media, but was detected within the intimal lesion. Thus, VSMC proliferation after angioplasty in the rat carotid artery is associated with a temporally and spatially coordinated expression of cdk2, cyclins E and A, and PCNA. Analysis of human arteries also revealed expression of these factors in VSMCs within restenotic lesions. Thus, cdk2 and its regulatory cyclins may be suitable targets to limit human restenosis.
KEY WORDS: angioplasty, restenosis, cell cycle control.
At homeostasis, vascular smooth muscle cells (VSMCs) are postmitotic and express markers of the differentiated phenotype. However, mature VSMCs can undergo phenotypic modulation in response to several environmental stimuli. Pathological VSMC proliferation is thought to play a central role during atherosclerosis and restenosis Owens, 1995 #735; Fuster, 1992 #709; Ross, 1993 #177. Therefore, understanding the fundamental basis of VSMC proliferation is of obvious interest. Antisense oligonucleotides against several protooncogenes Shi, 1994 #382; Simons, 1992 #192; Bennett, 1994 #376 and positive cell cycle control genes Abe, 1994 #852; Morishita, 1993 #763; Morishita, 1994 #762; Morishita, 1994 #847, or arterial gene transfer of negative cell cycle control genes Chang, 1995 #413; Chang, 1995 #718 and growth stimulatory genes Nabel, 1993 #764; Nabel, 1993 #765 affected injury-induced VSMC proliferation in different animal models of vascular remodeling. However, the kinetics of expression of the endogenous cell cycle regulatory factors in response to vascular injury remain largely unknown.
Progression through the mammalian mitotic cycle is controlled by multiple holoenzymes comprised of a catalytic cyclin-dependent protein kinase (cdk) and a cyclin regulatory subunit Heichman, 1994 #713; King, 1994 #714; Morgan, 1995 #453; Motokura, 1993 #730; Nurse, 1994 #399. Functional cdk/cyclin holoenzymes are presumed to phosphorylate target protein substrates that facilitate cell cycle progression Graña, 1995 #772; Peeper, 1994 #712; Weinberg, 1995 #446. Different cdk/cyclin complexes are orderly activated at specific phases of the cell cycle. Progression through the first gap-phase (G1) requires both cyclin D-dependentcdk4 and cdk6, and cdk2/cyclin E holoenzymes. Functional cdk2/cyclin A complexes are required for DNA synthesis (S-phase) and, subsequently, cdc2/cyclin A and cdc2/cyclin B pairs are assembled and activated during the second gap-phase (G2) and mitosis (M-phase), respectively. Recent evidence has been provided suggesting the requirement of cdk2 for entry into mitosis as a positive regulator of cdc2/cyclin B kinase activity Guadagno, 1996 #776.
In this study, we examined the regulation of cell cycle control proteins following vascular injury. Our findings demonstrate the temporally and spatially coordinated induction of PCNA, cdk2 and its regulatory subunits, cyclin E and cyclin A, in balloon-injured rat carotid arteries. Upregulation of these proteins was associated with the formation of functional cdk2/cyclin E and cdk2/cyclin A complexes. Immunohistochemical staining of human restenotic lesions also revealed the expression of cdk2, cyclin E and PCNA in SM -actin immunoreactive cells. Thus, induction of cdk2 and its regulatory cyclin subunits in the vessel wall may contribute to intimal thickening during restenosis.
Materials and Methods
Antibodies. The following antibodies were used in this study: (a) Rabbit polyclonal antibodies sc-397 (anti-p21), sc-163 (anti-cdk2), sc-481 (anti-cyclin E), sc-198 (anti-human cyclin E), and sc-751 (anti-cyclin A) (Santa Cruz Biotechnology); (b) Mouse mAb anti-smooth muscle -actin (SM -actin) (Enzo Diagnostics) and anti-SM -actin conjugated to alkaline phosphatase (Sigma Chemicals); (c) Mouse mAb anti-proliferating cell nuclear antigen (PCNA) (Signet Laboratories).
Human vascular tissue. Coronary and peripheral atherosclerotic lesions were retrieved percutaneously by therapeutic directional atherectomy as previously described Pickering, 1993 #263. Six coronary lesions were obtained from patients undergoing percutaneous revascularization for the first time (primary lesions). Another five coronary and four peripheral lesions were identified at the site of a previous angioplasty and were therefore designated as restenotic lesions. Tissue specimens were fixed by immersion in 100% methanol overnight and then embedded in paraffin and cut in 5-m sections. Specimens were estained with hematoxylin/eosin and elastic tissue trichrome for histopathological examination.
Rat model of balloon injury.Acute endothelial denudation of the left common carotid artery was performed essentially as described by Clowes et al. Clowes, 1983 #312. Male Sprague-Dawley rats weighing 400-500 g were anesthetized with sodium pentobarbital (intraperitoneal injection, 45 mg/kg body wt; Abbot Laboratories). The bifurcation of the left common carotid artery was exposed through a midline incision and the left common, internal, and external carotid arteries were temporarily ligated. A 2F embolectomy catheter (Baxter Edwards Healthcare Corp.) was introduced into the external carotid and advanced to the distal ligation of the common carotid. The balloon was inflated with saline and drawn towards the arteriotomy site three times to produce a distending, deendothelializing injury. After withdrawing the catheter, the proximal external carotid artery was ligated and blood flow was restored to the common carotid artery by release of the ligatures. The right uninjured carotid artery was used as control tissue. At the indicated times post-injury, rats were euthanized with sodium pentobarbital (intraperitoneal injection, 100 mg/kg body wt), and both injured and uninjured common carotid arteries were perfused with saline and dissected free of the surrounding tissue. For immunohistochemistry, tissue specimens were fixed by immersion in 100% methanol overnight and then embedded in paraffin. Arteries for the preparation of cell extracts were quickly frozen in liquid nitrogen and stored at -80 oC until further manipulation.
Preparation and analysis of arterial extracts. To minimize animal-to-animal and procedure variability, whole cell extracts were prepared from the pooled tissue from 8-11 rat carotid arteries at each time point. Arteries were homogenized in 1ml of ice-cold lysis buffer [50 mmol/L Tris-Cl (pH 7.6), 100 mmol/L NaCl, 5% NP-40, 30 mmol/L NaF, 1 mmol/L PMSF, 10% glycerol, 2 g/ml leupeptin] using a Tissumizer Mark II homogenizer (Tekmar). Tissue lysates were centrifuged at 4 oC for 10 minutes in a microfuge set at maximum speed, and the supernatant was stored at -80 oC in small aliquots.
Western immunoblot analysis (50 g protein) was performed as previously described Andrés, 1996 #527. The following dilutions of primary antibodies were used: 1:5 (anti SM a-actin), 1:100 (anti-cyclin A, anti-p21), 1:200 (anti-PCNA), 1:250 (anti-cyclin E), and 1:500 (anti-cdk2). For histone H1 kinase assays, arterial lysates (75 g protein) were incubated at 4 oC for 1.5-2 h under constant rotation in 0.5 ml of lysis buffer containing 0.15 g of the indicated antibodies and 25 l of Protein A/G PLUS-Agarose beads (Santa Cruz). Immunocomplexes were washed three times with lysis buffer and twice with kinase buffer [40 mmol/L Tris-Cl (pH 7.6), 20 mmol/L MgCl2, 2 mmol/L DTT]. Subsequently, the beads were resuspended in 30 l of kinase buffer containing 2 g of histone H1 (Boehringer Mannheim), 7 mol/L ATP, and 5 Ci of [-32P]ATP. The reaction mixtures were incubated at 30 oC for 30 minutes and then separated on 12% SDS/polyacrylamide gels Laemmli, 1970 #620.Gels were stained with Coomassie Blue (Sigma Chemicals), dried, and autoradiographed. Quantification of the signal was performed by counting individual histone H1 bands in a scintillation counter. Background cpm determined from regions of the dried gel that did not contain protein were substracted.
Immuhistochemistry.Methanol-fixed, paraffin-embedded rat and human specimens were cut in 5-µm sections. Sections were deparaffinized and rehydrated with PBS. Immunohistochemical staining was performed using a HistoMark universal strepatavidin/biotin kit according to the recommendations of the manufacturer (Kirkegaard & Perry Laboratories, Inc.). The following dilutions of primary antibodies (in PBS containing 2% normal goat serum, 0.02% sodium azide) were used: 1:500 (cdk 2), 1:150 (cyclin E sc-481), 1:100 (cyclin E sc-198), 1:50 (PCNA), and 1:30 (SM -actin). Rat carotid and human coronary arteries were immunostained with cyclin E antibodies sc-481 and sc-198, respectively. Incubation with primary antibodies was performed either at 37 oC for 1 h, or overnight at 4 oC. For peptide competition, a 10-fold mass excess of either cdk2 or cyclin E immunogenic peptide was added to the corresponding primary antibody 1-2 h prior to incubation of sections. SM -actin-alkaline phosphatase immunocomplexes were directly stained with Fast Red substrate (Bio Genex). Other primary immunocomplexes were detected with species-appropriate biotinylated secondary antibodies and strepatavidine-peroxidase. For double immunohistochemistry of human specimens, sections were first stained for PCNA, cyclin E and cdk2 as described above and peroxidase activity was detected with 0.05% (wt/vol) 3, 3'-diaminobenzidine tetrahydrochloride dihydrate substrate. Specimens were then washed with PBS, incubated with anti SM -actin antibody conjugated to alkaline phosphatase and stained with Fast Red substrate. The specimens were mounted with glycerol gelatin aqueous mounting media (Sigma Chemicals) and examined on an Olympus Vanox-T microsocope (Olympus America, Inc.). Pictures were recorded on Kodak Gold Plus film (Eastman Kodak Co.).
Results
Kinetics of expression and activity of cell cycle regulators after angioplasty in the rat carotid artery.To begin to elucidate the molecular mechanisms underlying the proliferative response of VSMCs to arterial injury, we analyzed the kinetics of expression and activity of cell cycle regulatory factors in control and balloon-injured rat carotid arteries. Intimal hyperplasia in this model of vascular remodeling results primarily from an excesive proliferative response of VSMCs, which also undergo dedifferentiation and migration Clowes, 1986 #779; Clowes, 1983 #312; Clowes, 1985 #777. Emergence of VSMCs from quiescence involves the transition from G0 to G1 and S-phase of the cell cycle. Since the cdk inhibitory protein p21Cip1 has been involved in withdrawal from the cell cycle in a variety of differentiated cell types Andrés, 1996 #527; Guo, 1995 #433; Halevy, 1995 #422; Parker, 1995 #423; Jiang, 1994 #393; Guo, 1995 #433; Halevy, 1995 #422; Parker, 1995 #423; Zhang, 1995 #788; Andrés, 1996 #527; Liu, 1996 #789; Macleod, 1995 #631; Poluha, 1996 #783, we hypothesized that downregulation of p21Cip1 might be involved in VSMC proliferation following angioplasty. However, p21Cip1 was undetectable in control rat carotid arteries and up to 60 h post-angioplasty under conditions where its expression was detected in cultures of quiescent skeletal myocytes (Fig. 1A). Thus, these findings provided no indication that p21Cip1 is involved in the maintenance of the postmitotic state of VSMCs in vivo.
Positive regulators of cellular proliferation that were induced following angioplasty included cdk2 and its cognate cyclins E and A (Fig. 1B). Following the accumulation of cyclin E within 1 d after injury, cdk2 and cyclin A expression was upregulated at 2 d post-injury. Western blot analysis at earlier time points revealed induction of cyclin E within 8h, which also preceded the upregulation of cdk2 and cyclin A (Fig. 2). Expression of these proteins was sustained up to 10 d and then declined at 18 d following angioplasty (Fig. 1B). Of note, the temporal pattern of expression of cdk2 and its cyclin subunits correlated with the expression of PCNA (Fig. 1B), a marker of cell growth Hall, 1990 #826; Bravo, 1980 #821; Bravo, 1986 #820; Bravo, 1987 #814; Almendral, 1987 #815; Jaskulski, 1988 #816.
We next sought to test whether cdk2 and cyclins E and A assembled into functional holoenzymes in the injured arterial wall. To this end, cdk2-, cyclin E- and cyclin A-containing protein complexes were harvested from control and injured rat carotid arteries using specific antibodies, and histone H1 kinase activity was assayed in the immunoprecipitates. A marked induction of kinase activity at 36 h post-injury was observed in anti-cdk2 (Fig. 2A), anti-cyclin E (Fig. 2B) and anti-cyclin A (Fig. 2C) immunoprecipitates, which correlated with upregulation of cdk2, cyclin E and cyclin A protein levels. Analysis at 60 h post-angioplasty disclosed diminished cdk2-, cyclin E- and cyclin A-dependent kinase activity. Of note, cdk2 and cyclin A protein levels were similar at 36 and 60 h post-injury. Thus, downregulation of cdk2- and cyclin A-dependent kinase activity between 36 and 60 h post-injury might result from post-translational inactivation of cdk2/cyclin A holoenzymes.
Spatially coordinated expression of cdk2, cyclin E and PCNA following vascular injury in the rat carotid artery. To further characterize the kinetics of expression of cell cycle regulatory proteins following angioplasty, we performed immunohistochemical analysis in control and injured rat carotid arteries. Cdk2 expression was upregulated in medial VSMCs at 36 h post-injury (Fig. 3A, 3B). At later time points when intimal hyperplasia was first noted, cdk2 was detected in the intimal lesion, but was expressed at low or undetectable levels in the media (Fig. 3C). By 2 wk, when a prominent intimal lesion was formed, cdk2 expression was largely confined to the luminal surface (Fig. 3D, 3E). As shown in Fig. 4, cyclin E upregulation was also noted in the media at 36 h after angioplasty and declined thereafter. Expression of cyclin E was detected in the emerging intimal lesion (Fig. 4E), and then was predominantly seen in the luminal surface of the intima at 2 wk (Fig. 4G). The specificity of cdk2 and cyclin E signal in these immunohistochemical analyses was demonstrated in control experiments, in which preincubation of the primary antibodies with an excess of the corresponding immunogenic peptide abroggated the signal (Fig. 3F, 4B, 4D, 4F, 4H, and data not shown).
To investigate whether expression of cdk2 and cyclin E was spatially correlated with VSMC proliferation following balloon angioplasty, we examined the expression of PCNA by immunohistochemistry. PCNA expression was undetectable in uninjured vessels (Fig. 5A) and its expression was upregulated at 36 h post-injury (Fig. 5B). At later time points, expression of PCNA was detected throughout the emerging neointima (Fig. 5C), and then became limited to the luminal surface (Fig. 5D). Adjacent sections incubated with control non-immune mouse IgG disclosed lack of staining (Fig. 5E-H). Taken together, these results indicate that VSMC proliferation is correlated with the spatially coordinated expression of cdk2 and cyclin E following angioplasty in the rat carotid artery.
Expression of cell cycle regulators in human atherectomy specimens.The data thus far suggest that induction and activation of cdk2-containing holoenzymes may be involved in VSMC accumulation following injury in the rat carotid artery. Since the biology of injury-induced VSMC proliferation may differ in animals and human arteries, we next sought to examine the expression of cell-cycle regulatory proteins in human atherosclerotic lesions obtained by directional atherectomy (Table 1). Each of the primary lesions analyzed were hypocellular. In marked contrast, extensive foci of hypercellularity were observed in all restenotic lesions. Other histopathological findings are summarized in Table 1.
Sections were stained with anti SM -actin antibody to identify VSMCs. All restenotic specimens and 4/6 (67%) primary lesions analyzed disclosed SM -actin immunoreactive cells (Table 1). Of these, eight restenotic (89%) and one primary (25%) specimen revealed expression of cdk2 and cyclin Ein regions of the plaque that stained positive for SM -actin and PCNA (Table 1, and Fig. 6A-D). Preincubation of the cdk2 and cyclin E antibodies with an excess of the corresponding immunogenic peptide abroggated the signal (Fig. 6E, F), demonstrating the specificity of cdk2 and cyclin E immunostaining. Lack of PCNA staining in human VSMCs correlated with undetectable levels of cdk2 and cyclin E in 3/4 (75%) primary lesions and 1/9 (11%) restenotic specimen (Table 1, and Fig. 7). Thus, in agreement with our findings in balloon-injured rat carotid arteries, abundant expression of cdk2 and cyclin E was detected in regions of human restenotic lesions that contained SM -actin and PCNA immunoreactive cells. Further evidence for the expression of proliferation markers within human restenotic VSMCs was provided using a sequential double immunostaining approach, which disclosed expression of PCNA, cdk2 and cyclin E in SM -actin immunoreactive cells (Fig. 8).
Discussion
This study demonstrates the temporally and spatially coordinated induction of positive regulators of cell growth following balloon angioplasty in the rat carotid artery. Induction of cdk2 and its regulatory subunits, cyclin E and cyclin A, was associated with the formation of functional cdk2/cyclin E and cdk2/cyclin A holoenzymes in injured arteries. Both the time course and spatial distribution of these proteins correlated with the expression of the proliferation marker PCNA. Cdk2 and cyclin E were also detected in regions of human restenotic lesions that contained SM -actin and PCNA immunoreactive cells. Thus, cdk2 and its regulatory cyclins may be suitable targets for restenosis therapy.
Temporally and spatially coordinated induction of positive regulators of cell cycle progression after angioplasty in the rat carotid artery.The main goal of the present study was to elucidate regulatory mechanisms underlying the proliferative response of VSMCs to arterial injury. Endothelial denudation in the rat carotid artery induces a well-characterized proliferative and migratory response of VSMCs Clowes, 1983 #312; Clowes, 1985 #777; Clowes, 1986 #779. The first response to injury (0-3 d) consists of VSMCs proliferation in the media. Subsequently, migration of VSMCs from the media into the intima is observed (3-14 d), and later cell proliferation appears to account for intimal lesion growth between 1-3 wk after injury. Several lines of evidence indicate that cdk2 and its regulatory subunits, cyclins E and A, may contribute to intimal lesion growth in this animal model of arterial injury. First, Western blot analysis and kinase assays revealed a rapid induction and activation of cdk2 and cyclins E and A between 1 and 2 d post-injury. Expression of these proteins, which correlated with PCNA expression, was sustained up to 10 d following angioplasty and then declined by 18 d. Secondly, immunohistochemical analyses at 36 h post-injury revealed cdk2 and cyclin E expression in medial VSMCs. Subsequently, cdk2 and cyclin E expression was detected in VSMCs within the emerging intimal lesion, and became low or undetectable within the media. At later time points when a prominent hyperplastic response was noted, expression of cdk2 and cyclin E was largely confined to the luminal surface of the intima. That expression of cdk2 and cyclin E was spatially correlated with VSMC proliferation following angioplasty was suggested by PCNA immunohistochemistry, which displayed essentially the same spatial pattern of expression. These findings are consistent with previous [3H]thymidine and 5-bromodeoxyuridine pulse-labeling experiments in rats Belknap, 1996 #782; Clowes, 1983 #312; Majesky, 1987 #865 demonstrating a peak of VSMC proliferation in the media of balloon-injured carotid arteries between 33-48 h, which then declined rapidly thereafter to return to baseline levels. Proliferative activity was noted throughout the newly formed intimal lesion, and subsequently became limited to the luminal surface. Taken together, these studies demonstrate a striking temporal and spatial correlation between the kinetics of VSMC proliferation and the kinetics of expression and activity of cdk2, cyclin E and cyclin A during injury-induced vascular remodeling in the rat carotid artery. In this regard, antisense cdk2 oligonucleotides have been shown to inhibit intimal hyperplasia in the rat carotid model of balloon angioplasty Morishita, 1994 #762; Abe, 1994 #852.