Survivin (SVV) is a multifunctional protein that has been implicated in the development of neointimal hyperplasia. Nuclear SVV is essential for mitosis, whereas in mitochondria SVV has a cytoprotective function. Here, we investigated the effects of RNA interference (RNAi)-mediated SVV knockdown on cell cycle kinetics, apoptosis, migration, and gene expression in primary cultured vascular smooth muscle cells (VSMCs) from the human saphenous vein. Primary Human VSMCs were obtained from saphenous veins and cultured under standard conditions. SVV knockdown was achieved by either small interfering RNA or lentiviral transduction of short hairpin RNA, reducing SVV gene expression by quantitative PCR (>75%, P < 0.01) without a loss of cell viability. Subcellular fractionation revealed that RNAi treatment effectively targeted the nuclear SVV pool, whereas the larger mitochondrial pool was much less sensitive to transient knockdown. Both p53 and p27 protein levels were notably increased. SVV RNAi treatment significantly blocked VSMC proliferation in response to serum and PDGF-AB, arresting VSMC growth. Cell cycle analysis revealed an increased G2/M fraction consistent with a mitotic defect; 4′,6-diamidino-2-phenylindole staining confirmed an increased frequency of polyploid and abnormal nuclei. In a transwell assay, SVV knockdown reduced migration to PDGF-AB, and actin-phalloidin staining revealed disorganized actin filaments and polygonal cell shape. However, apoptosis (DNA content and annexin V flow cytometry) was not directly induced by SVV RNAi, and sensitivity to apoptotic agonists (e.g., staurosporine and cytokines) was unchanged. In conclusion, RNAi-mediated SVV knockdown in VSMCs leads to profound cell cycle arrest at G2/M and impaired chemotaxis without cytotoxicity. The regulation of mitosis and apoptosis in VSMC involves differentially regulated subcellular pools of SVV. Thus, treatment of VSMC with RNAi targeting SVV might limit the response to vascular injury without destabilizing the vessel wall.
- cell cycle
reminiscent of the phenotype of cancer cells and in contrast to quiescent vascular smooth muscle cells (VSMCs), neointimal VSMCs express a proliferative, migratory, and apoptosis-resistant profile. Understanding the molecular mechanisms underlying this phenotypic switch in VSMCs is central to controlling the healing response of blood vessels, a fundamental problem that limits the effectiveness of many cardiovascular interventions.
Survivin (SVV) is a unique member of the inhibitor of apoptosis protein family that functions at the nexus of cellular survival and growth (1). Absolutely required for the execution of mitosis, SVV gene expression in nontransformed cells is tightly regulated at the G2-M transition point (11). The functions of SVV as an inhibitor of apoptosis protein have been demonstrated in cancer models using dominant negative mutants and antisense or proteomic approaches (2, 15, 16).
A recent study (5) has demonstrated distinct subcellular pools of SVV that mediate its disparate activities and are differentially regulated. In cultured cancer cells, cellular SVV is predominantly localized to the mitochondrial compartment during oxidative stress (5). Interestingly, mitochondrial SVV is protected from degradation by the chaperone proteins heat shock protein (Hsp)60 and Hsp90 (6, 8). In contrast, nuclear SVV is rapidly ubiquitinated and degraded after mitosis.
Little is known about the functions of SVV in vascular cells. In previous studies (3, 17), we have demonstrated that SVV expression is upregulated in the vessel wall after diverse forms of mechanical or dietary injury. Local modulation of SVV expression by adenoviral delivery of wild-type or dominant negative SVV constructs resulted in correspondingly increased or decreased wall thickening in a rabbit vein graft model (20). These studies have suggested that SVV is a highly relevant molecular target for the control of the vascular injury response.
We hypothesized that targeted knockdown of SVV gene expression in VSMCs would reduce the proliferative response and induce apoptosis. RNA interference (RNAi) experiments demonstrated a profound mitotic defect in treated VSMCs without, however, an immediate loss of viability or a significant change in the apoptotic threshold. Furthermore, treated cells demonstrated markedly impaired chemotaxis to the prototypic agonist PDGF-AB. These findings are consistent with a model of differentially regulated subcellular pools of SVV that mediate distinct functions in VSMCs. RNAi-mediated SVV gene knockdown is cytostatic in VSMCs and may offer a therapeutic strategy to control neointimal growth without acutely destabilizing the vessel wall.
Cells and Cell Culture
Primary human VSMCs were harvested from discarded saphenous vein segments (from patients undergoing bypass surgery procedures under an Institutional Review Board-approved protocol at the University of Californa, San Francisco, CA) by explant methods, as previously described, and cultured in DMEM (Invitrogen) containing FBS (Invitrogen) in 5% CO2 at 37°C. All experiments were performed on cells that were passage 5 or less.
Table 1 shows the list of reagents used in the present study.
RNAi-Based SVV Gene Knockdown
Small interfering RNA.
The validated SVV and matching control small interfering (si)RNA oligonucleotides used were as follows: SVV siRNA ID:2734 and Silencer Negative Control no. 1 siRNA AM4635 (Applied Biosystems). For further validation and to exclude nonspecific cellular effects of the siRNA sequences, we also obtained additional SVV and control siRNA sequences from another vendor (Invitrogen) as follows: Alexa 647-labeled Stealth Select RNAi siRNA Oligo ID HSS179403 targeting SVV and Stealth RNAi siRNA Negative Control LO GC. Results obtained with the Invitrogen siRNA sequences were similar across all of the assays used.
Dharmafect no. 1 (Dharmacon) and Lipofectamine RNAi Max (Invitrogen), respectively, were used as transfection reagents. VSMCs were seeded 24 h before transfection, with 50–70% confluency on the day of transfection. A 50 nM siRNA solution was prepared in serum-free media (Opti-Mem, Invitrogen) and combined with serum-free media containing Dharmafect no. 1. After a brief incubation, the siRNA-Dharmafect no. 1 solution was added to the cells.
Lentiviral short hairpin RNA.
SVV-specific and control short hairpin (sh)RNA constructs (validated sequences) were obtained from the Broad Institute (Massachusetts Institute of Technology, Cambridge, MA) targeting the following nucleotide sequence: 5′-CCGCATCTCTACATTCAAGAA-3′ (SVV cDNA, NM_001168.2, 172–192 nt). All shRNA constructs contained a puromycin resistance gene cassette (purR).
Recombinant lentiviral preparations were assembled in the Lentiviral RNAi Core facility of the University of California (San Francisco, CA).
Lentiviral experiments were carried out at a calculated multiplicity of infection of 5. The virus particles from the stock solution were prepared in DMEM with 10% FBS in the presence of polybrene (8mg/ml, Millipore, Billerica, MA) and then added to the cultured cells. After infection (48 h), puromycin (2 μg/ml, Sigma) was added to the growth medium as a selection antibiotic for 48 h. After the puromycin selection, cells were grown for 24–48 h in regular growth media (DMEM with 10% FBS).
Analysis of Gene Expression
VSMCs were lysed with TRIzol (Invitrogen), and total RNA was extracted using a PureLink RNA Mini kit including DNAse treatment (Invitrogen). cDNA was synthesized with a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems).
Power SYBR green Mastermix (Applied Biosystems) using a MJ Research Opticon 2 engine Real-Time-PCR system (Bio-Rad Laboratories) was used for amplification. The respective primer sequences are shown in the Supplemental Material (Supplemental Table S1).1 In addition to survival in puromycin-containing selection medium, successful integration of shRNA in treated cells was demonstrated by real-time RT-PCR analysis of purR gene expression.
GAPDH served as a housekeeping gene, and the ΔΔCt method, where Ct is threshold cycle, was used to calculate relative mRNA expression levels.
Cell Viability and Proliferation Assays
Cell viability was quantified through the assessment of mitochondrial activity using Alamar blue assays (Invitrogen). In brief, 5,000 VSMCs were plated per well in a 96-well plate and grown under standard conditions for 24 h. To assess viability, Alamar blue [10% (wt/vol) in FBS-containing media, Trek Diagnostics] was added to the cells 16 h after the respective RNAi treatment, and, 3 h later, optical density was determined by the use of excitation/emission wavelengths of 570/600 nm in a cytofluorometer (Fluoroskan Asent FL, Thermo Fisher Scientific). The optical density values of the RNAi control and RNAi SVV groups were normalized to the no treatment group. A minimum of three independent experiments were conducted.
Proliferation was quantified by plating 25–30,000 VSMCs into 6 well plates (Nunc), and cells were cultured under standard conditions. After SVV RNAi treatment, cells were either cultured in regular media containing FBS (1% for SVV siRNA or 10% for SVV shRNA) or serum-free media containing 100 ng/ml PDGF-AB. VSMCs were trypsinized at the respective time points and counted using trypan blue and a hemocytometer. A minimum of three separate experiments was carried out for each condition, and cell counts were normalized to the cell count at the time of plating.
Total Cellular Protein Isolation
Cells were washed with PBS (Invitrogen) and trypsinized. The cell suspension was pelleted at 2,000 g for 5 min. The supernatant was aspirated, and lysates were prepared by adding ice-cold RIPA buffer containing a cocktail of proteinase inhibitors (BD Biosciences, Clontech). The lysate was pipetted up and down for 45 s. Lysates were pelleted at 2,000 g for 5 min at 4°C. Protein concentrations were determined using the BCA protein assay (Pierce).
VSMCs were transfected with siRNA (Ambion BIRC5 or nontargeting control, 50 nM) for 48 h in Opti-MEM I reduced serum media. Cells were harvested with trypsin, neutralized with FBS, and washed with ice-cold PBS. VSMCs were separated into cytosolic, membranous, and nuclear fractions using the Thermo Scientific Subcellular Protein Fractionation kit (catalog no. 78840) following the manufacturer's instructions. Protein lysates were quantified using the Pierce BCA protein assay kit (product no. 23225), and equal amounts of each fraction were loaded for Western blot analysis. Membranes were probed for SVV (Santa Cruz Biotechnology), Hsp90 (a cytosolic protein control, Cell Signaling), and histone H3 (a nuclear protein control, Cell Signaling).
For the isolation of the mitochondrial protein fraction, we used the Qproteome mitochondria isolation kit (Qiagen). In brief, cells were washed, pelleted, and then resuspended in ice-cold lysis buffer. The lysate was centrifuged at 1,000 g for 5 min at 4°C. Next, the supernatant, which contained cytosolic proteins, was removed, and the cell pellet was resuspended in 200 μl ice-cold disruption buffer. Complete cell disruption was achieved by sonication. The lysate was then centrifuged at 1,000 g for 10 min at 4°C, and the supernatant was transferred to a clean Eppendorf tube. The pellet was then resuspended in 100 μl mitochondrial storage buffer. The supernatant was spun at 6,000 g for 10 min at 4°C. The retrieved pellet contained the mitochondrial fraction. After purification, the pellet was resuspended in mitochondrial storage buffer and frozen at −80°C until further use.
Western Blot Analysis
Protein concentrations were measured using the standard BCA protein assay method (Pierce). Equal amounts of protein were loaded in each lane. Proteins were separated by gel electrophoresis using the XCell SureLock Mini-Cell System and precast NuPAGE Novex 4–12% bis-Tris gels (both Invitrogen). MES buffer served as the running buffer. Proteins (20–40 μg) were loaded per lane, and the gel was run at 130 V for 55 min. Proteins were transfered onto polyvinylidene difluoride (PVDF) membranes at 25 V for 50 min using Tris-glycine buffer (containing 20% methanol) in a Bio-Rad Trans-Blot Semi-Dry System.
PVDF membranes were then blocked in 5% milk in Tris-buffered saline (TBS)-Tween (TBS-T) for 45 min at room temperature. The primary antibody was prepared in 5% milk in TBS-T. PVDF membranes were incubated in the primary antibody solution at 4°C overnight. PVDF membranes were washed three times with TBS-T and then incubated in the secondary antibody solution (also in 5% milk in TBS-T) for 3 h at 4°C followed by a wash. Amersham ECL Plus Western Blotting Detection reagents (GE Healthcare) served as the chemiluminescence kit.
The following primary antibodies were used: goat anti-SVV (sc-8807, Santa Cruz Biotechnology, 1:200), mouse anti-GAPDH (sc-47724, Santa Cruz Biotechnology, 1:10,000), mouse anti-cytochrome c (sc-13560, Santa Cruz Biotechnology, 1:200), rabbit anti-p27 (1:200, Santa Cruz Biotechnology), mouse anti-p53 (1:200, Santa Cruz Biotechnology), rabbit anti-Hsp90 (a cytosolic protein control, Cell Signaling), mouse anti-histone H3 (a nuclear protein control, Cell Signaling), and mouse anti-β-actin (1:5,000, Sigma).
Detection of Apoptosis
For the detection of apoptosis, annexin V and propidium iodide staining was performed using the Annexin V-FITC/Propidium Iodide Apoptosis kit (Biovision). Apoptosis was induced in VSMCs by stimulation with a cytokine cocktail consisting of 400 U/ml human IFN-γ (R&D Systems, Minneapolis, MN), 400 U/ml human TNF-α (Pierce Endogen), and 100 U/ml human IL-1β (Pierce Endogen) for 24 h. Staurosporine (20 nM) in DMEM (Invitrogen) served as a positive control. VSMCs were washed, trypsinized, and pelleted. The supernatant was removed, and cells were resuspended in the provided staining buffer containing the annexin V-FITC antibody and propidium iodide. Analysis was carried out using flow cytometry (FACScan, Becton Dickinson). Data analysis was performed with FloJo software (Tree Star).
Cell Cycle Analysis
VSMCs were grown in standard growth medium and then stimulated with either different concentrations of FBS or PDGF-AB (100 ng/ml, Sigma). Cells were fixed with 70% ethanol, stained with propidium iodide (0.5 mg/ml, Roche Diagnostics), and then analyzed for DNA content by flow cytometry (FACScan, Becton Dickinson). Data analysis was performed with FloJo software (Tree Star).
Nuclear Morphology, Actin-Phalloidin Staining, and Imaging
Cells on microscope slides (Nunc) were stained with Alexa fluor 488 phalloidin (Invitrogen) following the manufacturer's protocol. In brief, cells were washed with PBS and then fixed with 4% formaldehyde for 15 min at room temperature. Cells were then washed three times with PBS for 5 min each. 4′,6-Diamidino-2-phenylindole (DAPI) staining was performed using Vectashield mounting medium (Vector Laboratories). Cell images were captured with a Zeiss Axiovert 200M digital fluorescence microscope (Carl Zeiss).
Transwell Migration Assay
One day before the migration assay, RNAi-treated cells were cultured in DMEM containing 0.1% serum. On the day of the experiment, cells were trypsinized and seeded onto the membranes (30,000 cells/insert, 6.5-mm diameter) in serum-free media (Transwell, Costar, Corning). Serum-free media (600 μl) containing PDGF-AB (50 ng/ml) as a chemoattractant was added to the lower chamber of the transwell. Assays were stopped after 16 h. Membranes were washed three times in PBS, and cells in the upper chamber were harvested using cotton swabs and thorough brushing. Membranes were then fixed in methanol (Sigma) for 15 min and subsequently mounted onto a microscope slide using DAPI-containing Vectashield mounting medium. Each experimental condition was carried out in duplicate. An average of five to eight random high-power (×10) pictures were taken per membrane, and the nuclei were counted manually.
In each individual experiment, counts were normalized to the untreated control group. Differences in migration were subsequently expressed as relative migration comparing treatment groups (control siRNA or shRNA and SVV siRNA and shRNA) with noninfected cells.
Quantitative RT-PCR data were compared using one-way ANOVA with a Bonferroni post hoc correction or unpaired t-tests as appropriate. For densitometry analysis, unpaired and paired t-tests were used as appropriate. For cell cycle analysis and proliferation and migration assays, ANOVA with a Bonferroni post hoc correction was used. Statistical significance was considered at P < 0.05. Experiments were repeated a minimum of three times.
Effects of RNAi on SVV Gene Expression and Protein Levels
RNAi treatment significantly reduced SVV gene expression in treated VSMCs by quantitative real-time PCR (Fig. 1A). Similar knockdown was achieved with lentiviral shRNA treatment (data not shown). After puromycin selection, SVV gene knockdown in lentiviral shRNA-treated cells increased to 90% (Fig. 1B).
Interestingly, total cellular SVV protein levels did not change significantly at 72 h after SVV siRNA transfection despite the reduction in mRNA (Fig. 1C). Transduction of VSMCs with SVV shRNA vectors also did not change cellular SVV protein levels after 72 h (data not shown). However, after puromycin selection (48 h) and an additional 48 h of cultivation, shRNA-treated cells demonstrated a moderate reduction of total SVV protein levels compared with controls (Fig. 1C).
We next performed subcellular fractionation experiments to determine the effects of SVV RNAi on specific compartments. Treatment with siRNA markedly reduced nuclear SVV protein levels (by ∼80%) while exerting minimal effects on cytosolic levels (Fig. 1D, left). The cytosolic protein fraction was further divided into mitochondrial and nonmitochondrial fractions, demonstrating that the majority of cellular SVV in VSMCs resides in a mitochondrial pool, which was less sensitive to RNAi knockdown. Densitometry analysis revealed that, on average, SVV levels within the mitochondrial fraction were ∼30 times higher than in the nonmitochondrial fraction (Fig. 1D, right).
Notably, SVV siRNA treatment lead to significant reciprocal increases in levels of the cell cycle inhibitory proteins p27kip1 and p53 (Fig. 1E).
RNAi-Mediated SVV Knockdown Does Not Induce Apoptosis in VSMCs
Cell viability was not altered after either mode of RNAi treatment (Fig. 2, A and B). Neither siRNA nor shRNA treatments led directly to VSMC apoptosis. Furthermore, the susceptibility of VSMCs to proapoptotic stimuli (e.g., inflammatory cytokines and staurosporine) was unchanged by RNAi treatment, as measured by annexin V and DNA content flow cytometry (Fig. 2C and Supplemental Fig. S1).
SVV Knockdown Is Cytostatic in VSMCs
SVV siRNA and shRNA treatment dramatically reduced VSMC proliferation in response to both serum (P < 0.05) and PDGF-AB (P < 0.05) stimulation (Fig. 3, A and B). Compared with both nontransfected and control-transfected conditions, SVV knockdown resulted in a lack of any discernible proliferative response 72–168 h after serum or growth factor exposure. The growth curves in these conditions were flat, consistent with a profound cytostatic, but not cytotoxic, effect.
Effect of SVV Gene Knockdown on Cell Cycle Kinetics in VSMCs
Cell cycle analysis revealed an increased G2/M fraction associated with SVV gene knockdown. The effects on the G2/M fraction were clearly demonstrated at 48 h after exposure to PDGF-AB (27% SVV siRNA vs. 17% control siRNA, P < 0.01) or 0.1% serum (25% SVV siRNA vs. 16% control siRNA, P < 0.05; Fig. 3, C and D). Consistent with a profound mitotic defect, SVV RNAi consistently induced an increase of polyploidy and an overall increase of abnormal nuclear morphology in treated VSMCs irrespective of the SVV siRNA or shRNA sequence used (Fig. 4).
SVV Gene Knockdown Reduces the VSMC Migration Response to PDGF-AB
SVV siRNA treatment significantly reduced VSMC chemotaxis toward PDGF-AB (74% reduction vs. controls, P < 0.01) in a transwell migration assay (Fig. 5). Similar results were observed after SVV shRNA treatment (Supplemental Fig. S2). Actin-phalloidin staining revealed a strong association between abnormal nuclear configuration, a disorganized actin filament system, and polygonal cell shape (Fig. 5A). These findings suggest that the reduced migration of treated cells may be at least partly related to the induction of polyploidy.
The results of this study are consistent with a model of differentially regulated subcellular pools of SVV in VSMCs, paralleling observations that have been made in cancer cells (4, 7). Acute gene knockdown using RNAi selectively targeted cell cycle-dependent SVV expression, which is critical for mitosis. The larger mitochondrial pool of SVV is less sensitive to short-term RNAi treatment, suggesting different kinetic regulation of SVV protein levels in this compartment. These findings have important implications for therapeutic targeting of SVV in vascular injury. RNAi-mediated SVV gene knockdown produced a dramatic cytostatic effect in VSMCs without a loss of viability, an attractive profile for treatment of a bypass graft or stented artery.
During mitosis, nuclear SVV functions along with other chromosomal passenger proteins to insure faithful chromatid separation by the spindle apparatus (18). Forming a complex along with AuroraB and the inner centromere protein, SVV is required for the activation of AuroraB, which phosphorylates other targets, including histone H3, vimentin, and centromere protein-A, during chromatid separation and cytokinesis (9). Depletion of nuclear SVV by RNAi resulted in evidence of mitotic failure in VSMCs, with increased DNA content and polyploidy. A recent study (12) has demonstrated that polyploidy of VSMCs is commonly seen in the aging vasculature and in hypertension, where it may be associated with cellular hypertrophy.
Interestingly, RNAi-mediated SVV silencing also significantly reduced VSMC migration. This observation could be related to the observed polyploidy and hypertrophy of SVV RNAi-treated cells. Of note, recent studies have described how SVV promotes cancer cell motility independent of cell cycle and apoptosis regulation. Akt has been implicated in SVV-mediated cell migration (13, 14). In addition, a prior study (19) has shown that p27kip1 inhibits migration in VSMCs and endothelial cells, and, therefore, the increased levels of p27kip1 observed after SVV knockdown may provide an alternative mechanism for these observations.
Mitochondrial SVV associates in a complex with Hsp60 and Hsp90 and inhibits apoptosis by stabilizing the mitochondrial pore complex (10). Increased levels of SVV are associated with apoptosis resistance, a characteristic not only of cancer cells but also of neointimal VSMCs. A prior study (10) in cancer cells demonstrated that targeting of this SVV-Hsp complex within the mitochondria results in an acute loss of mitochondrial membrane function and apoptotic death. These experiments suggest that this mitochondrial protein complex is critical for the regulation of the apoptotic threshold in cells. Our data suggest that mitochondrial SVV levels in VSMCs are relatively insensitive to short-term transcriptional repression by SVV siRNA treatment, similar to findings in tumor cell lines (2). Furthermore, our data illustrate that most SVV resides within the mitochondria of VSMC and that SVV siRNA treatment primarily reduces nuclear SVV levels. Future experiments will seek to examine the phenotype of VSMCs in which long-term stable SVV knockdown has been achieved. Alternative proteomic approaches may be more relevant for targeting apoptosis resistance via the mitochondrial SVV-Hsp pathway (16).
Our previous studies demonstrated a dramatic upregulation of SVV expression in the vessel wall postinjury, primarily within VSMCs. Importantly, SVV expression in these vascular injury models appeared to have both cell cycle-associated and cell cycle-independent aspects, as judged by only partial localization to proliferating cells (i.e., Ki-67 expression) (3, 17). Our present model of SVV trafficking in VSMCs suggests differentially regulated pools with distinct functions relevant to neointimal disease. Mitogenic stimuli (e.g., growth factors or serum) result in cell cycle-dependent SVV expression at G2/M, which is primarily localized to the nucleus to support mitosis. In addition, VSMCs stimulated by mechanical injury, growth factors, or cytokines also demonstrate a marked upregulation of cell cycle-independent SVV transcription, which contributes to a mitochondrial pool that is stabilized by Hsp90 and has a cytoprotective, antiapoptotic function (17, 20).
In summary, short-term knockdown of SVV gene expression in VSMCs produces a cytostatic phenotype without an acute loss of viability. These findings are most likely explained by differentially regulated subcellular pools of SVV that have distinct critical roles relevant to cellular growth, survival, and function. We believe that these findings have direct relevance to the manipulation of the vascular injury response and represent distinct therapeutic targets for reducing neointima formation.
This work was supported by National Heart, Lung, and Blood Institute Grant HL-085157 (to M. S. Conte) and funding from the Foundation for Accelerated Vascular Research and a Lifeline Student Fellowship Award from the American Vascular Association.
No conflicts of interest, financial or otherwise, are declared by the author(s).
The authors thank Dr. Dario C. Altieri (University of Massachusetts) for critical review of the manuscript.
↵1 Supplemental Material for this article is available at the American Journal of Physiology-Heart and Circulatory Physiology website.
- Copyright © 2011 the American Physiological Society