INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ABNORMAL PROLIFERATION of vascular smooth muscle cells (VSMC) in the arterial intima has been implicated in a variety of pathological processes including atherosclerosis, hypertension, and restenosis after injury (21). It is believed that VSMC proliferation is suppressed under physiological conditions, but can be rapidly induced by growth factors, such as platelet-derived growth factor and fibroblast growth factor, in some pathological and adaptive states, resulting in vascular hyperplasia (28). Vasoconstrictive hormones such as angiotensin II (ANG II) and arginine vasopressin (AVP) also play important roles in these processes. Our previous study (9) demonstrated that ANG II and AVP could promote G₁/S transition and DNA replication in VSMC, but failed to induce G₂/M transition and mitosis primarily because of the failure of cdc2 mRNA induction. The failure of cdc2 promoter activation is at least in part attributable to defective binding of free E2F-1 to the promoter. In contrast, fetal bovine serum (FBS), which contains several growth factors, can induce cdc2 mRNA expression by recruiting E2F-1 to the cdc2 promoter, which results in cell progression into M phase (9). These data suggest that E2F is one of critical determinants of cellular responses to various stimulants in VSMC.

E2F is a family of transcriptional factors that control G₁/S transition of eukaryotic cells by regulating transcription of various growth-related genes such as thymidine kinase, cdk2, cdc2, cdc6, cyclin E, and E2F-1 (2). The E2F family is composed of at least six structurally related proteins that bind to DNA as heterodimers with one of two differentiation-regulated transcription factor-1-polypeptide-1 subunits (27). E2F subunits are aligned into three distinct groups according to their structural and functional similarities. E2F-1, E2F-2, and E2F-3 are related by sequence and can functionally induce S phase progression when overexpressed in quiescent cells (24). In contrast, E2F-4 and E2F-5 cannot induce cell proliferation and are now known to act mainly as transcriptional repressors in a complex with RB family proteins (pRB, p130, and p107), to maintain cells in G₀/G₁ phase by suppressing genes with growth-promoting properties (32). Moreover, a new member of the E2F family, E2F-6, was recently identified by virtue of its affinity to E2F consensus sequences. Although recent studies suggest that E2F-6 is a dominant negative repressor of E2F-mediated transcription, its biological significance is yet to be determined (3).

Most of our knowledge on E2F functions is primarily obtained from experiments using fibroblasts and continuous cell lines such as Saos-2 (24, 32). It is therefore possible that each E2F subunit has distinct functions in other cell types, as revealed by studies using knockout mice (15, 23, 30, 35). Furthermore, recent investigations indicated that some E2F components govern functions other than cell cycle regulation, such as induction of apoptosis (18) and the maintenance of terminally differentiated state (16). These findings suggest that individual E2F proteins can exert different functions in cell type- and stimulator-dependent manners. This may greatly affect cellular responses to various stimulants and is therefore important for understanding the biological behavior of cells under both physiological and pathological conditions.

The present study was therefore undertaken to determine how E2F family proteins participate in biological responses of VSMC to vasoconstrictive hormones. We first examined the effects of ANG II and AVP on the expression patterns of E2F subunits in cultured rat VSMC. We then investigated the involvement of E2F proteins in transcriptional regulation of cell cycle control genes in VSMC.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Preparation of the cells and cell culture. Rat VSMC were isolated from male Sprague-Dawley rats, as described in previous studies (9, 29). The dispersed cells were resuspended in Eagle's minimal essential medium (MEM), pH 7.4, containing 1 µmol/l L-glutamine, 100 U/ml penicillin, and 10% FBS. The cells were kept in a humidified incubator under 95% air-5% CO₂ at 37°C. Phenotype of isolated VSMC was determined by morphological examination and immunostaining for -SM actin (using antibody 1A4, Sigma) and calponin (using antibody hCP, Sigma). The cells were subcultured with trypsin-EGTA treatment. After a day of subculture, the cells were synchronized for 48 h in serum-free Eagle's MEM to obtain quiescent cells (control group). The medium was then replaced with Eagle's MEM containing effectors, and the culture was continued for given culture periods. The "FBS-free" shown in the figure legends means an additional serum-deprived incubation after 48-h starvation.

Western blotting. Cells were washed with Tris-buffered saline (TBS) (25 mmol/l Tris · HCl, pH 8.0, and 150 mmol/l NaCl) and lysed for 30 min at 4°C with lysis buffer (50 mmol/l Tris · HCl, pH 8.0, 120 mmol/l NaCl, 0.5% Nonidet P-40, 100 mmol/l sodium fluoride, and 200 µmol/l sodium orthovanadate), containing 10 µg each of aprotinin, phenylmethylsulfonyl fluoride, and leupeptin (Sigma). Cell lysates were centrifuged, and the supernatants (40 µg/sample) were applied onto sodium dodecyl sulfate (SDS)-polyacrylamide gels. After electrophoresis, proteins were transferred onto nitrocellulose membranes (Amersham; Buckinghamshire, UK) in transfer buffer containing 25 mmol/l Tris · HCl, 192 mmol/l glycine, and 20% methanol. The blots were incubated with 1 µg/ml primary antibodies, washed with TBS containing 0.05% Tween 20, and probed with a 1:1,000 dilution of anti-mouse or anti-rabbit horseradish peroxidase-conjugated antibodies. The blots were incubated with an enhanced chemiluminescence substrate and exposed to photographic film to visualize immunoreactive bands.

Immunoprecipitation. Whole cell lysates (300 µg protein) were incubated with 1 µg each of rabbit polyclonal antibodies against E2F-4 (C-108), E2F-5 (E-19), pRB (C-15, Santa Cruz Biotechnology), and p130 (C-20) in 300 µl of lysis buffer at 4°C for 1 h. Immune complexes were collected on protein A Sepharose beads, washed three times in 0.5× lysis buffer, and resolved on 8% SDS-polyacrylamide gel electrophoresis, followed by immunoblotting with mouse monoclonal antibodies against E2F-4 (GG22-2A6; Upstate Biotechnology; Lake Placid, NY), E2F-5 (MH-5, Santa Cruz Biotechnology), pRB (XZ55; Pharmingen; San Diego, CA), and p130 (clone10, BD Transduction Laboratories).

Northern blotting. Total cellular RNA was isolated by cesium chloride ultracentrifugation. RNA samples (10 µg each) were electrophoresed in a 1% agarose gel containing 6% formaldehyde, 20 mmol/l 3-(N-morpholino)propanesulfonic acid, 5 mmol/l sodium acetate, and 1 mmol/l EDTA, and blotted onto nylon membranes. The membranes were hybridized with optimal cDNA probes, which were labeled with [³²P]dCTP (NEN). Partial cDNA fragments of rat E2F-1 (14), mouse E2F-2 (GenBank accession no. W53973), mouse E2F-3 (5), mouse E2F-4 (GenBank accession no. AA050824), rat E2F-5 (19), rat thymidine kinase (TK) (GenBank accession no. Y17296), rat p130 (21), and mouse cdc6 (13) were generated by reverse transcription-polymerase chain reaction using specific primer pairs (Table 1) and used as probes. Full-length cDNA of mouse dihydrofolate reductase (DHFR) was obtained from Japanese Cancer Research Resources Bank (7).

Electrophoretic mobility shift assay. Nuclear extract was prepared according to the method of Dignam et al. (6). Samples (5 µg protein each) were incubated with ³²P-labeled oligonucleotide probe in the presence of sonicated salmon sperm DNA. Incubations were carried out at room temperature for 30 min in a solution composed of (in mmol/l) 20 HEPES (pH 7.9) 0.1% Nonidet P-40, 40 KCl, 1 MgCl₂, 0.1 EDTA, 0.5 dithiothreitol, and 10% glycerol. The DNA protein complexes were resolved on a 4% polyacrylamide gel (acrylamide-to-bisacrylamide ratio, 86:1 wt/wt).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Expression patterns of E2F family proteins and mRNAs in rat VSMC. First, we screened for the expression of E2F family proteins in rat VSMC cultured with FBS and vasoconstrictive hormones (ANG II and AVP) using immunoblotting. After serum deprivation for 48 h to obtain quiescent cells (control), VSMC were restimulated by the addition of effectors at optimal concentrations. Optimal concentrations of FBS, ANG II, and AVP were determined as 10%, and 0.1 and 1 µmol/l, respectively, in our previous study (9). E2F-1 and E2F-5 were downregulated in quiescent cells, whereas the expression of E2F-2, E2F-3, and E2F-4 was stable (Fig. 1A, lane 3). In FBS-treated cells, E2F-1 was readily induced with a peak 24 h after restimulation (Fig. 1A, lane 4). Similarly, E2F-5 significantly increased in a time-dependent manner (Fig. 1A, lane 5). The signal intensity of E2F-1 was much stronger than that of E2F-5 at their respective peaks of expression. In contrast, vasoconstrictive hormones did not alter E2F-1 levels but stimulated the expression of E2F-3, which was slightly reduced by FBS (Fig. 1A, lanes 6-9). The abundance of E2F-2 and E2F-4 was not affected by these stimulants in our experimental conditions.

View larger version (50K): [in this window] [in a new window]   Fig. 1.   Effects of fetal bovine serum (FBS), angiotensin II (ANG II), and arginine vasopressin (AVP) on E2F family protein and mRNA expression in vascular smooth muscle cells (VSMC). Serum-starved VSMC (control) were cultured with each effector at its optimal concentration for the indicated periods and subjected to Western and Northern blot analysis. A: Western blotting revealed that E2F-1 migrated as a doublet of ~60 kDa. B: Northern blotting was carried out with cDNA fragments, which were amplified by reverse transcriptase-polymerase chain reaction (RT-PCR) with specific primer pairs (see Table 1) as probes. Ethidium bromide staining of 28S rRNA is shown as a loading control. Data shown are representative of three independent experiments.

Effects of FBS, ANG II, and AVP on binding patterns of E2F family proteins on E2F-1 promoter. We then investigated how these changes in expression patterns of E2F subunits affected transcription of E2F-dependent genes in VSMC. To this end, we isolated nuclear extracts from VSMC before and after 24 h of culture with FBS and ANG II and performed electrophoretic mobility shift assays using an oligonucleotide containing the putative E2F-binding site of the E2F-1 promoter as a probe. As shown in Fig. 2A, E2F complexes on the E2F-1 promoter were discerned as six bands (labeled bands A-F) and divided into two different patterns: an FBS-stimulated pattern (lane 3) and a nonstimulated pattern (lanes 1, 2, and 4). The specificity of these complexes was verified by cold competition assays (Fig. 2B). The identification of each band of the FBS-stimulated pattern was based on the results of antibody perturbation experiments using specific antibodies against E2F subunits (E2F-1 to 5) and RB family proteins (pRB, p107, and p130) (Fig. 3A). The anti-E2F-1 antibody decreased the intensity of band A, along with induction of a new band with retarded mobility (lane 2). Similarly, band E was supershifted by both anti-E2F-5 (lane 6) and anti-pRB antibodies (lane 7). The other bands were not affected by any of the antibodies used in this study. These results indicate that band A corresponded to free (uncomplexed) E2F-1 and band E represented the E2F-5/pRB complex. The nonstimulated pattern was similarly analyzed using nuclear extracts from ANG II-treated VSMC (Fig. 3B). The anti-E2F-3 antibody abolished band F and induced a new band with retarded mobility (lane 4). The anti-p130 antibody almost completely supershifted band B (lane 9). The other bands were not affected by any of the antibodies used in this study. These findings suggest that band F is free E2F-3 and that band B represents an E2F complex containing p130. The binding partner of p130 in ANG II-treated VSMC was identified as E2F-4 by coimmunoprecipitation experiments (see below). The same pattern was obtained in experiments using nuclear extracts from serum-starved VSMC (data not shown). Overall, the E2F-1 promoter is activated by E2F-1 itself in an autoregulatory manner in FBS-treated VSMC, which in turn facilitates cell proliferation. Upregulation of E2F-5 by FBS may be involved in the negative regulation of some E2F-dependent genes, including E2F-1, in concert with pRB in later phases of cellular responses to FBS. In addition, FBS greatly reduced the amounts of transcriptional repressor complex containing p130 (band B), which contribute to derepression of E2F-1 promoter. Vasoconstrictive hormones cannot transactivate the E2F-1 gene, despite the increase in E2F-3 expression and the decrease in E2F/p130 complex.

View larger version (52K): [in this window] [in a new window]   Fig. 2.   Effects of FBS and ANG II on E2F complexes on the E2F-1 promoter. A: nuclear extracts were isolated from rat VSMC cultured with each effector for 24 h and incubated with 32P-labeled oligonucleotides containing the E2F-binding site of the E2F-1 promoter. Positions of specific E2F complexes are indicated on the left of each panel (bands A-F). B: cold competition assays were carried out with samples from FBS-stimulated cells. C: 32P-labeled probe alone was electrophoresed. Data shown are representative of three independent experiments.

View larger version (58K): [in this window] [in a new window]   Fig. 3.   Identification of E2F complexes on the E2F-1 promoter. Nuclear extracts were isolated from VSMC cultured with FBS (A) and ANG II (B) for 24 h. Antibody perturbation experiments were carried out with specific antibodies against E2F and RB family proteins. Positions of specific E2F complexes are indicated on the left side of each panel (bands A-F). Arrows indicate supershifted bands. Coimmunoprecipitation studies (C): whole cell lysates from VSMC cultured with FBS and ANG II for 24 h were subjected to immunoprecipitation with polyclonal antibodies against the indicated proteins (IP), followed by immunoblotting with monoclonal antibodies against the indicated proteins (Blotting). Data shown are representative of three independent experiments.

View larger version (66K): [in this window] [in a new window]   Fig. 4.   Effects of FBS, ANG II, and AVP on TK and dihydrofolate reductase (DHFR) expression in cultured rat VSMC. Serum-starved VSMC (control) were cultured with each effector at its optimal concentration for the indicated periods and subjected to Northern blot analysis. Northern blotting was carried out with PCR product of TK cDNA and full-length DHFR cDNA as probes. Ethidium bromide staining of 28S rRNA is shown as a loading control.

Effect of FBS, ANG II, and AVP on binding patterns of E2F family proteins on cdc6 promoter. Specific induction of E2F-3 by ANG II and AVP strongly suggests the involvement of E2F-3 in certain cellular responses of VSMC to vasoconstrictive hormones involved in G₁/S transition. However, little is known about the actual role of E2F-3 in transcriptional regulation of the genes responsible for G₁/S transition in VSMC. To address this point, we repeated electrophoretic mobility shift assays to examine E2F binding to the promoters of several candidate genes, including TK, cdk2, cyclin E1, and cdc6. E2F-3 did not bind to the putative E2F binding sites of the promoter regions of TK, cyclin E1, and cdk2 genes (data not shown). On the cdc6 promoter, however, E2F complexes (marked A-E) were detectable in two different patterns: an FBS-stimulated pattern (lane 3) and a nonstimulated pattern (lanes 1, 2, and 4) (Fig. 5A). The specificity of these complexes was verified by cold competition assays (Fig. 5B). The identification of each band was done by antibody perturbation, as in the E2F-1 promoter. As shown in Fig. 6, anti-E2F-3 and p130 antibodies supershifted bands D and A, respectively, in ANG II-treated VSMC, suggesting that band D corresponds to free E2F-3 and band A represents a complex containing p130 and an E2F protein of unknown identity (probably E2F-4). The same pattern was obtained from experiments with serum-starved cells (data not shown). In contrast, no apparent supershift was observed in FBS-treated VSMC (data not shown). These results indicate that vasoconstrictive hormones can activate the cdc6 gene by recruiting free E2F-3 onto its promoter and downregulating a transcriptional repressor complex containing p130 (band A). In addition, FBS almost completely eradicated the E2F/p130 complex that may be responsible for activation of the cdc6 gene, if it occurs, without recruitment of transcriptionally active free E2Fs by FBS.

View larger version (55K): [in this window] [in a new window]   Fig. 5.   Effects of FBS and ANG II on E2F complexes on the cdc6 promoter. A: nuclear extracts were isolated from rat VSMC cultured with each effector for 24 h and incubated with 32P-labeled oligonucleotides containing the E2F-binding site of the cdc6 promoter. Positions of specific E2F complexes are indicated on the left (bands A-E). B: cold competition assays were carried out with samples from FBS-stimulated cells. C: 32P-labeled probe alone was electrophoresed. Data shown are representative of three independent experiments.

View larger version (90K): [in this window] [in a new window]   Fig. 6.   Identification of each band of ANG II-induced E2F complexes on the cdc6 promoter. Nuclear extracts were isolated from rat VSMC cultured with ANG II for 24 h. Antibody perturbation experiments were carried out with specific antibodies against E2F and RB family proteins. Positions of specific E2F complexes are indicated on the left (bands A-E). Arrows indicate supershifted bands. Data shown are representative of three independent experiments.

Effects of FBS, ANG II, and AVP on cdc6 and p130 expression in VSMC. As described above, the analysis of the cdc6 promoter suggested that the cdc6 gene is transactivated by both FBS and vasoconstrictive hormones, although the underlying mechanisms are somewhat different. To confirm this, we examined the expression of cdc6 in VSMC cultured with FBS, ANG II, and AVP. Immunoblots showed that the cdc6 protein was downregulated during serum starvation of VSMC (Fig. 6A, lanes 2 and 3). Both FBS and vasoconstrictive hormones increased the amounts of cdc6 protein in a time-dependent manner (Fig. 7A, lanes 4-9). Northern blotting revealed that cdc6 mRNA was under the detection limits in serum-starved cells but induced by FBS, ANG II, and AVP with a peak 12 h after restimulation (Fig. 7B). These results are compatible with the hypothesis drawn from the analysis of the cdc6 promoter.

View larger version (38K): [in this window] [in a new window]   Fig. 7.   Effects of FBS, ANG II, and AVP on p130 and cdc6 expression in cultured rat VSMC. Serum-starved VSMC (control) were cultured with each effector at its optimal concentration for the indicated periods and subjected to Western and Northern blot analyses. A: Western blotting revealed that p130 and cdc6 were detected as single bands of 130 and 62 kDa, respectively. B: Northern blotting was carried out with cDNA fragments, which were generated by RT-PCR with the indicated primers (Table 1) as probes. Ethidium bromide staining of 28S rRNA is shown as a loading control. Data shown are representative of three independent experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Elucidation of the molecular mechanisms of cell proliferation of VSMC is necessary to understand the pathological processes in atherosclerosis, hypertension, and restenosis after injury. Several investigators recently studied cell cycle regulation of VSMC under pathological conditions. Wei et al. (34) reported that VSMC proliferation after angioplasty in the carotid artery is associated with a temporally and spatially coordinated expression of cdk2, cyclins E and A, and proliferating cell nuclear antigen. Guo et al. (12) reported that the antiproliferative effect of nitric oxide on VSMC results, at least in part, from the repression of cyclin A gene transcription. Ihling et al. (17) demonstrated that transforming growth factor-₁ present in human atherosclerotic tissue exerts its growth suppressive activity on VSMC through blocking the activity of cdk2-cyclin E complex by upregulation of p27. There are some studies regarding the involvement of cyclin-dependent kinase inhibitors in cell cycle regulation in VSMC. Aoki et al. (1) reported that p53 and pRB negatively regulated the cell cycle of VSMC and that the former also plays a pivotal role in apoptosis associated with cell growth. Tanner et al. (33) demonstrated that p27^kip1 and p21^cip1 (but not p16^ink4) are potent inhibitors of VSMC growth. Although these findings provide important information on the molecular basis of VSMC proliferation in pathological states, the universal mechanisms regulating these cell cycle control genes remain unclear.

Vasoconstrictive hormones such as ANG II and AVP are known as enhancers of atherosclerosis under certain conditions in vivo. It is therefore clinically important to examine their effects on cell cycle regulation of VSMC. We (9) reported that vasoconstrictive hormones could promote G₁/S transition and DNA replication in VSMC but fail to induce mitosis. This may be implicated in the development of vascular hypertrophy leading to atherosclerosis. In contrast, FBS can induce mitosis in VSMC by upregulation of cdc2 mRNA. This difference stems from the fact that only FBS can recruit free E2F-1 to the cdc2 promoter, suggesting that E2F is one of the critical determinants of cellular responses to various stimulants in VSMC (9). However, little is known about the precise roles of E2F family proteins in transcriptional regulation of E2F-target genes in VSMC.

We investigated how E2F family proteins participate in biological responses of VSMC to vasoconstrictive hormones compared with FBS. We first examined the expression patterns of E2F family proteins and mRNAs in VSMC cultured with FBS, ANG II, and AVP. FBS induced upregulation of E2F-1 and E2F-5 at both mRNA and protein levels and slightly reduced the abundance of E2F-3 protein. In contrast, vasoconstrictive hormones increased the amounts of E2F-3 protein but not E2F-1 and E2F-5 without upregulation of E2F-3 transcripts. This can be explained by the increment in protein stability as described previously (8). Among E2F subunits, E2F-1, E2F-2, and E2F-3 are suppressed in quiescent cells and induced at late G₁ phase of the cell cycle by mitogenic stimuli (2, 24, 27). Our present data are in line with this notion but are also suggestive of stimulant-specific induction of distinct E2F subunits in the same cell type: E2F-1 and E2F-5 by FBS, and E2F-3 by vasoconstrictive hormones. The biological significance of this phenomenon will be discussed later. E2F-5 is usually expressed at nearly equivalent levels in both quiescent and proliferating cells (11, 19). Recent investigations (23) have indicated that E2F-5 plays some functional roles in terminally differentiated cells, best characterized by the dysfunction of the choroid plexus in E2F-5 nullizygous mouse. In our analysis, E2F-5 was induced by FBS in VSMC, and made a complex with pRB. The inducibility of E2F-5 is unique to VSMC and may be implicated in feedback regulation of certain cellular functions during cell proliferation, such as the inhibition of overexpression of target genes like E2F-1 in our study.

We then investigated the consequences of the stimulant-specific induction of distinct E2F subunits and its biological significance. E2F-1 gene was transactivated by FBS through the recruitment of free E2F-1 onto the E2F-1 promoter. In contrast, ANG II-induced binding of free E2F-3 did not result in activation of the E2F-1 promoter. Although it is currently unknown why and how E2F-1 and E2F-3 affect the same target differentially, such differences may be related to the divergence of cellular responses to stimulants. A similar conclusion was drawn from recent study (15) using E2F-1 and E2F-3 knockout mice, wherein the lack of E2F-1 and E2F-3 affects different subsets of E2F-dependent genes. Consistent with our data, E2F-1 is normally induced in E2F-3-deficient fibroblasts. Previous studies (15, 22) showed that E2F-3 is critical for transcriptional activation of genes that control the rate of cell proliferation such as TK, cyclin A, cyclin E, and cdc6. However, we could not detect E2F-3 binding to the promoters of TK, cyclin E, or cdk2 genes in this study. In contrast, free E2F-3 was recruited to cdc6 promoter by ANG II, which is at least in part implicated in transcriptional activation of the cdc6 gene by vasoconstrictive hormones. Because cdc6 plays a pivotal role in the timely initiation of DNA replication, the inducibility of cdc6 is closely related to the ability of vasoconstrictive hormones to allow the G₁/S transition in VSMC. Both FBS and ANG II decreased the amounts of a transcriptional repressor complex containing p130, a member of the RB family (4). FBS-induced activation of the cdc6 promoter was considered as primarily mediated through this mechanism because FBS could not upregulate E2F-3. We found that the decrease in the amounts of p130 underlays the downregulation of the E2F/p130 complex by FBS and vasoconstrictive hormones. This suggests that p130 has some roles in the regulation of G₁/S transition of VSMC, which occurs in both FBS- and hormone-treated cells, through suppressing E2F activity. Modulation of E2F, as well as its major regulator p130, may have therapeutic value against atherosclerosis, hypertension, bypass graft failure (25), and restenosis after vascular injury (26).