Overexpression of Human Papillomavirus Type 16 Oncoproteins Enhances Hypoxia-Inducible Factor 1α Protein Accumulation and Vascular Endothelial Growth Factor Expression in Human Cervical Carcinoma Cells

Human papillomavirus (HPV), a group of small nonenveloped DNA viruses, are clinically classified as high and low risk. Epidemiologic evidences have convincingly showed a strong correlation between persistent infection with high-risk human papillomavirus types and human cervical cancers (1). The two principal viral oncogene products, namely, E6 and E7, have been reported to be persistently expressed during all stages of HPV-associated cervical cancers and are essential for HPV-induced malignant transformation and perseverance of cellular oncogenic phenotype in cervical carcinomas (1, 2).

The oncogenic properties of E6 and E7 are associated with their close interactions with a variety of host cell proteins including the apoptotic regulation (1–3). Expression of E6 oncoprotein leads to the rapid ubiquitination and degradation of p53, thereby abolishing p53-mediated apoptosis and other cellular functions such as cell cycle regulation (4). On the other hand, E7 binds to members of the pocket protein family such as pRB and causes the release of E2F-like transcription factors, thus leading to G₁-S transition of the cell cycle (5, 6). In addition, the presence of E6 and E7 oncoproteins has been shown to induce tumor angiogenesis by promoting the expression of a variety of proangiogenic factors, including fibroblast growth factor binding protein, an angiogenic switch molecule (7); basic fibroblast growth factor; transforming growth factor-β; tumor necrosis factor-α; interleukin-8 (8, 9); and vascular endothelial growth factor (VEGF; refs. 8–11). However, the underlying mechanisms by which high-risk HPV type 16 E6 and E7 oncoproteins induced up-regulation of VEGF expression and enhancement of tumor angiogenic activity remain unclear.

Hypoxia-inducible factor-1 (HIF-1), the master transcription factor in response to hypoxia, is composed of two subunits, HIF-1α and HIF-1β, wherein only HIF-1α is tightly regulated by low oxygen tension. Under normoxic conditions, HIF-1α is hydroxylated at two prolyl residues (Pro⁴⁰² and Pro⁵⁶⁴) within an oxygen-dependent degradation domain by Fe²⁺ and O₂-dependent HIF-1 prolyl hydroxylases, resulting in ubiquitination and subsequent degradation via the 26S proteasome (12, 13). However, under hypoxic conditions, prolyl hydroxylation is inhibited, leading to rapid accumulation of HIF-1α. Other factors independent of hypoxia can also promote HIF-1α protein accumulation via translational or post-translational mechanisms (14, 15), including the inactivation of pVHL (16) and phosphatase and tensin homologue (PTEN, deleted on chromosome 10; ref. 17), the activation of Ha-ras, v-Src, and HER2 (18–20), and the stimulation by several inflammatory cytokines, hormones, and growth factors.

Several studies have shown that many human cancers and their metastases express an elevated level of HIF-1α protein, which is closely associated with a more aggressive tumor phenotype (21–23). The elevated expression of HIF-1α not only serves a useful intrinsic biological marker of hypoxia (24–27), but also an independent indicator of tumor aggressiveness, malignant progression, and prognosis in cervical cancer patients (26–31). Built upon these findings and epidemiologic evidences, we hypothesize that the expression of high-risk HPV-16 E6 and E7 viral oncoproteins may contribute, at least in part, to an elevated expression of HIF-1α in cervical cancers.

Resveratrol (3,4′,5-tri-hydroxystilbene), a natural polyphenolic phytoalexin found in various plants, including grapes, berries, and peanuts, has recently been considered a potential cancer chemopreventive agent. Recent studies have shown that resveratrol has potent inhibitory effects on tumor angiogenesis (32), and the underlying mechanisms seem to be associated with the suppression of HIF-1α and VEGF expression in different cancer cell lines (33, 34). In the present study, we showed for the first time that overexpression of HPV-16 E6 and E7 oncoproteins significantly promoted HIF-1α protein accumulation and VEGF expression in human cervical cancer cells. The viral oncoproteins induced up-regulation of HIF-1α activity, and its proangiogenic activity could be reversed by resveratrol. These findings suggest that high-risk HPV-16 oncoproteins can promote tumor angiogenesis possibly via up-regulating HIF-1α–mediated VEGF expression in human cervical cancers. Inhibition of HIF-1α and VEGF expression by resveratrol could serve a new therapeutic approach in chemoprevention and treatment of human cervical cancers.

Reagents.Trans-3,4,5′-trihydroxystilbene (resveratrol; Sigma), LY294002, and PD98059 (Calbiochem) were dissolved in 100% DMSO. Antibodies include HIF-1α monoclonal antibody (BD Transduction Laboratories), total and phosphorylated p42/p44 mitogen-activated protein (MAP) kinases (MAPK; Thr²⁰²/Tyr²⁰⁴) or Akt (Ser⁴⁷³) antibodies (New England Biolabs), β-actin monoclonal antibody (Sigma), and horseradish peroxidase–conjugated secondary antibodies (Pierce). Transfection reagents (LipofectAMINE 2000 and OligofectAMINE) were from Invitrogen Corporation.

Cell culture. Human cervical carcinoma cell lines (C-33A and HeLa; American Type Culture Collection) were cultured in RPMI 1640 supplemented with 10% fetal bovine serum, penicillin (100 units/mL), and streptomycin (100 μg/mL; Invitrogen). All cultures were maintained at 37°C in a humidified atmosphere with 5% CO₂.

Transient transfection. Hemagglutinin-tagged pSG5 plasmid vectors harboring HPV-16 E6, 16 E7, and their mutants were generated as described previously (3). The expression of HPV-16 E6 and E7 oncoproteins in transfected cells was confirmed using specific monoclonal antibody against hemagglutinin (3). Briefly, C-33A and HeLa cells at 70% to 80% confluency were transiently transfected for 4 h with different pSG5-HPV-oncogene constructs using LipofectAMINE 2000. Twenty-four hours post-transfection, the transfected cells and the conditioned media were harvested for further analysis. Transfection with empty pSG5 vector or pSG5 HPV-16 E6 and 16 E7 mutants served as controls. Cells exposed to LipofectAMINE 2000 or OligofectAMINE alone served as mock transfection controls. To observe the effect of resveratrol on HPV oncoprotein–induced HIF-1α and VEGF expression, transfected cervical cancer cells were exposed to different concentrations of resveratrol for 16 h. HIF-1α and VEGF protein levels were determined by Western blot and ELISA, respectively.

siRNA (or RNA interference). HIF-1α small interfering RNA (siRNA) duplexes (5′-AGAGGUGGAUAUGUG UGGGdTdT-3′ and 5′-CCCACACAUAUCCACCUCUdTdT-3′) were synthesized and annealed (Pharmacon Research, Inc.) as described previously (35). Cells were transfected for 4 h with HIF-1α siRNA duplexes (200 nmol/L) using OligofectAMINE. siRNA targeted to an irrelevant mRNA served as a nonspecific control.

Western blot analysis. Transfected and nontransfected cells were lysed with buffer containing 50 mmol/L Tris-HCl (pH, 7.5), 5 mmol/L EDTA, 150 mmol/L NaCl, 0.5% Triton X-100, 10 mmol/L sodium fluoride, 20 mmol/L β-mercaptoethanol, 250 μmol/L sodium orthovanadate, 1 mmol/L phenylmethylsulfonyl fluoride, and complete protease inhibitor cocktail (Sigma) and incubated at 4°C for 1 h. The lysates were ultrasonicated and centrifuged at 12,000 × g for 10 min. Protein concentrations were determined by the bicinchoninic acid method. Approximately 100 μg of protein was separated on 8% to 10% SDS-polyacrylamide gel and electroblotted onto nitrocellulose membranes (Hybond ECL, Amersham Pharmacia). After blocking with TBS/5% skim milk, the membrane was incubated overnight at 4°C with antibody against special primary antibodies, followed by incubation with horseradish peroxidase–conjugated secondary antibodies (1:1,000; Pierce). As a loading control, the blots were stripped and reprobed with an anti–β-actin antibody (1:4,000).

Immunofluorescence studies. Transfected or nontransfected (mock transfection) C-33A and HeLa cells in eight-well Lab-Tek II Chamber Slide System (Nalge Nunc Int.; 70% confluence) were fixed with 4% paraformaldehyde in PBS followed by incubation in 0.5% Triton X-100 in PBS. After several washes, cells were blocked with 3% bovine serum albumin in PBS followed by incubation with a mouse monoclonal anti-human HIF-1α antibody (1:200) at 4°C overnight in a humidified chamber. The cells were subsequently exposed to Fluor 488–conjugated goat anti-mouse immunoglobulin G (1:2,000; 0.5 μg/mL; Molecular Probes) and treated with mounting solution (ImmunoMount, Shandon). The labeled HIF-1α–expressing cells were detected under a fluorescence microscope. Cells incubated with fluorescein-conjugated secondary antibodies in the absence of primary antibodies served as negative controls.

Real-time PCR analysis for HIF-1α and VEGF mRNA levels. Total RNA was isolated from cells using TRIzol reagent (Invitrogen). Real-time PCR analysis of HIF-1α and VEGF mRNA levels was done using iScript One-step RT-PCR Kit with SYBR Green (Bio-Rad) according to the manufacturer's instructions. The following primers were designed for real-time PCR: for HIF-1α, forward 5′-GTT TAC TAA AGG ACA AGT CAC C-3′ and reverse 5′-TTC TGT TTG TTG AAG GGA G-3′ (36); human VEGF: forward 5-TCT ACC TCC ACC ATG CCA AGT-3 and reverse 5-GAT GAT TCT GCC CTC CTC CTT-3 (37); and β-actin: forward 5′-TCA AGA TCA TTG CTC CTC CTG-3′ and reverse 5′-CTG CTT GCT GAT CCA CAT CTG-3′ (38). All the primers were synthesized by GenoMechanix, LLC. The thermocycling conditions were as follows: 50°C for 10 min, 95°C for 5 min, followed by 40 cycles at 95°C for 10 s, and 53°C for 30 s (HIF-1α) or 60°C for 30 s (VEGF). The relative HIF-1α and VEGF mRNA levels were normalized to β-actin. The experiment was repeated in triplicate.

ELISA. The concentration of VEGF protein in the conditioned media of untreated and treated cells was determined using human VEGF ELISA Development kit (Peprotech Inc.) according to the manufacturer's instructions. Results were normalized to cell number (2 × 10⁵). The experiment was repeated in triplicate.

In vitro angiogenesis assay. An in vitro angiogenesis assay kit was employed according to the manufacturer's instructions (Chemicon International). Briefly, 96-well cell culture plates were coated with ECMatrix followed by seeding of human umbilical vascular endothelial cells (HUVEC; 5 × 10³ cells per well) onto the surface of the polymerized ECMatrix. Conditioned media derived from C-33A cells transfected with pSG5, pSG5 HPV-16 E6 and E7 alone or cotransfected with HIF-1α siRNA were added to cells, and tubule formation was observed under a phase-contrast microscope.

Statistical analysis. Data are presented as the mean ± SD for three separate experiments. A paired Student's t test, one-way ANOVA, Bonferroni, and Dunnett T3 were employed for statistical analysis using SPSS 11.0 for Windows software. P < 0.05 was considered to be statistically significant.

Expression of HPV oncoproteins enhanced HIF-1α protein accumulation in cervical cancer cells. To explore whether expression of HPV oncoproteins could contribute to an elevated expression of HIF-1α in cervical cancer, we transiently transfected two human cervical carcinoma cell lines, C-33A and HeLa cells, with pSG5 plasmid constructs expressing HPV-16 E6 or E7 oncoproteins and examined HIF-1α expression. Transfection with empty pSG5 vector or 16 E6, 16 E7 mutants served as controls. Our results showed that expression of HPV-16 E6 and E7 significantly enhanced protein accumulation in both C-33A and HeLa cells (Fig. 1A, lanes 3 and 4), whereas cells transfected with pSG5 empty vector (Fig. 1A, lane 2), 16 E6 and 16 E7 mutants (Fig. 1A, lanes 5 and 6), or mock-transfected (Fig. 1A, lane 1) controls showed minimal HIF-1α expression. Similar results were further confirmed using immunofluorescence studies in C-33A (Fig. 1B) and in HeLa (Fig. 1C). Enhanced HIF-1α signals were localized primarily in the nucleus of HPV-16 E6- and E7-transfected cells. To determine whether the increase of HIF-1α was regulated at the transcriptional level, we measured mRNA level by real-time PCR. As shown in Fig. 1D, expression of HPV oncoproteins had no obvious effects on HIF-1α mRNA levels in both 16 E6- and E7-transfected cervical cancer cells as compared with mock transfection or pSG5 vector controls. We next questioned whether HPV oncoproteins could induce HIF-1α protein accumulation by increasing its stability using cycloheximide as an inhibitor of protein synthesis. Our studies estimated a similar half-life of ∼5 to 7 min for HIF-1α protein in cells transfected with either HPV-16 E6, E7, or mock-transfected controls (P > 0.05; Fig. 1E and F). These results suggest that the apparent oncoprotein-induced multifold up-regulation of HIF-1α protein is probably independent of protein stabilization, but may be attributed to an altered regulation at the translational level. Further studies are needed to clarify the underlying mechanism.

Up-regulation of VEGF expression by HPV oncoproteins in cervical cancer cells is HIF-1α dependent. VEGF is an immediate downstream target gene of HIF-1α and plays a critical role in tumor angiogenesis (39). Previous studies have shown that expression of HPV-16 E6 and E7 oncoproteins promote tumor angiogenesis possibly by up-regulating VEGF expression (10, 40–42). Our results also showed that expression E6 and E7 led to a significant increase in VEGF secretion in the conditioned media of both C-33A and HeLa cells (P < 0.01), compared with mock-transfected or pSG5 vector-transfected control cells (Fig. 2A and B). VEGF secretion did not increase in cells expressing either 16 E6 or E7 mutant constructs and was similar to the mock-transfected controls (Fig. 2A and B). The results were further supported by real-time PCR analysis, which indicated that expression of HPV oncoproteins significantly increased VEGF mRNA levels as compared with transfection controls or 16 E6 and E7 mutant constructs (Fig. 2C and D).

To investigate whether the increase of VEGF expression induced by HPV oncoproteins is mediated by HIF-1α, C-33A cells were transfected with HIF-1α siRNA or a nonspecific siRNA, followed by cotransfection with pSG5-HPV-16 E6 or E7 constructs. Our results showed that HPV-16 E6- and E7-induced accumulation of HIF-1α protein was partially suppressed by cotransfection with HIF-1α siRNA but not by nonspecific siRNA (Fig. 3A). As expected, the increased VEGF mRNA expression and protein secretion induced by HPV-16 E6 or E7 were also significantly inhibited by cotransfection with HIF-1α siRNA (P < 0.01, Fig. 3B and C). These results indicated that HPV oncoprotein-induced up-regulation of VEGF expression is HIF-1α dependent.

HPV-16 E6 and E7 induced HIF-1α protein accumulation and VEGF expression via ERK1/2 and PI3K/Akt signaling pathways. Recent studies have shown that the HPV-16 oncoproteins can activate phosphoinositide-3-kinase (PI3K)/Akt and MAP/extracellular signal-regulated kinase (ERK) kinase (MEK)/ERK1/2 signaling pathways in different types of cells (11, 43, 44). Consistent with these findings, overexpression of HPV-16 E6 or E7 also led to an increased level of phosphorylated Akt and ERK1/2 in C33A cells (Fig. 4A). To study whether these signaling pathways are involved in HPV-16 E6- and E7-induced HIF-1α accumulation and VEGF expression, C-33A cells were transfected with pSG5-HPV-16 E6 or E7 for 4 h, followed by exposure to different concentrations of PD98059, a specific pharmacologic inhibitor of ERK1/2, or LY294002, a specific inhibitor of PI3K, for 16 h. Our results showed that treatment with both ERK1/2 and PI3K inhibitors led to a concentration-dependent decrease in HPV-16 E6- and E7-induced HIF-1α protein accumulation (Fig. 4B and C) and VEGF secretion (P < 0.01; Fig. 4D and E). In addition, our results indicated that pretreatment with PD98059 or LY294002 had no obvious effects on the basal levels of both HIF-1α protein and VEGF expression in C-33A cells (data not shown). These results suggest that HPV oncoproteins promoted HIF-1α protein accumulation and VEGF secretion possibly through the activation of ERK1/2 and PI3K/Akt signaling pathways.

Resveratrol inhibited HPV-16 oncoprotein-induced HIF-1α protein accumulation and VEGF expression in cervical cancer cells. Previous studies have shown that resveratrol can inhibit HIF-1α and VEGF expression induced by hypoxia or growth factors (33, 34). To explore whether resveratrol had similar inhibitory effects on HPV oncoprotein-induced HIF-1α protein accumulation and VEGF expression, C-33A cells transfected with pSG5-HPV oncoprotein constructs were exposed to different concentrations of resveratrol for 16 h. Our results showed that resveratrol significantly inhibited HPV-16 E6- and E7-induced HIF-1α protein accumulation in a concentration-dependent manner (Fig. 5A), but had no obvious effects on HIF-1α mRNA levels (Fig. 5C). Similar results were further confirmed in another cervical cancer cell line, HeLa cells (Fig. 5B and C). Consistent with the findings above, resveratrol also inhibited HPV oncoprotein-induced VEGF expression at both mRNA and protein levels (P < 0.01), but had no effect on the basal level of VEGF (P > 0.05; Fig. 5D and E). Mechanistically, we found that resveratrol inhibited HPV-16 E6- and E7-induced activation of both Akt and ERK1/2 (Fig. 4A). 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay showed minimal effects of resveratrol on cell viability (data not shown), thus ruling out the possibility that the inhibitory effects of resveratrol were due to its cellular toxicity.

In vitro capillary tube formation stimulated by expression of HPV-16 oncoproteins in cervical cancer cells was HIF-1α dependent. To further support the essential role of HPV-16 oncoproteins in angiogenesis in cervical cancers, an in vitro angiogenesis model was employed to evaluate the capillary tube formation of HUVECs stimulated by the conditioned media derived from C-33A cells transfected with HPV-16 E6 and E7 constructs. Our results showed that HPV-16 oncoprotein-transfected C-33A cells greatly stimulated HUVECs to form capillary tubelike structures on Matrigels as compared with mock-transfected cells (Fig. 6). The stimulation of capillary tube formation induced by HPV-16 E6- and E7-transfected cervical cancer cells was significantly abrogated by cotransfection with HIF-1α siRNA or by treatment with 100 μmol/L of resveratrol, but not by nonspecific siRNA (Fig. 6). Taken together, these results suggested that the enhanced in vitro capillary tube formation induced by expression of HPV oncoproteins in cervical cancer cells was HIF-1α dependent.

Increased angiogenesis has been correlated with tumor progression and metastasis, including cervical carcinomas (42, 45, 46). Previous studies have reported a possible correlation between VEGF expression and high-risk HPV in low-grade cervical intraepithelial neoplasia and early invasion (42, 45, 46). Consistent with the in vivo findings, HPV oncoproteins, 16 E5, E6, and E7, have been shown to induce VEGF expression in several established cervical cancer cells (8–11, 47). In the present study, we showed for the first time that expression of HPV-16 oncoproteins promoted HIF-1α protein accumulation (Fig. 1) in C-33A and HeLa cells and led to a significant increase in VEGF mRNA expression and protein secretion (Fig. 2). Moreover, inhibition of HIF-1α by cotransfection with HIF-1α siRNA significantly inhibited HPV oncoprotein–induced VEGF expression in these cells (Fig. 3). Functionally, expression of HPV-16 E6 and E7 in cervical cancer cells can stimulate HUVECs to form capillary tubule structures in vitro, and these angiogenic effects were suppressed by cotransfection with HIF-1α siRNA (Fig. 6). Collectively, these findings suggest that the up-regulation of VEGF expression induced by HPV-16 oncoproteins is HIF-1α dependent. Most recently, López-Ocejo et al. (10) have shown that E6 oncoprotein induces VEGF expression by direct promoter transcriptional activation in a p53-independent manner and identified an E6 response region between site −194 and −50 bp upstream to the transcription initiation site of VEGF promoter. Further studies are needed to explore the interplay between HIF-1α and HPV-16 oncoproteins in the context of VEGF promoter regulation.

It is known that hypoxia activates HIF-1 by stabilizing HIF-1α protein through inhibiting HIF prolyl-4-hydroxylase 2 (PHD2; ref. 48), whereas under normoxic conditions, growth factors, cytokines, and oncogenes activate HIF-1α through both translational (protein synthesis) and posttranslational (protein stabilization) mechanisms (15, 49). In the present study, our results showed that expression of HPV-16 E6 and E7 oncoproteins had no effects on HIF-1α mRNA levels (Fig. 1D) and HIF-1α protein stabilization (Fig. 1E and F), suggesting that the oncoprotein-induced HIF-1α protein may be attributed by an altered regulation at the translational level. Furthermore, the abundant localization of HIF-1α signals in the cell nucleus of HPV-transfected cells further support an increase in protein synthesis. In addition, several studies have shown that HPV oncoproteins can activate Raf/MEK/MAPKs, PI3K/Akt, and epithelial growth factor receptor kinase signaling pathways (11, 50, 51), which are involved in the up-regulation of VEGF expression (11). Similarly, we showed that expression of HPV-16 E6 and E7 led to the activation of both Akt and ERK1/2 (Fig. 4A), and blocking PI3K/Akt and ERK1/2 signaling pathways significantly inhibited HPV-16 E6- and E7-induced HIF-1α protein accumulation and VEGF protein secretion (Fig. 4B-E), suggesting that these two signaling pathways are involved in HPV-16 E6- and E7-induced HIF-1α protein accumulation and VEGF protein secretion. However, more studies are required to gain further insights into the detailed molecular mechanisms involved in HPV oncoprotein-induced HIF-1α protein accumulation.

To date, it has been well established that persistent infection of the cervix with high-risk HPVs is a necessary cause of cervical cancers (1); however, the prognosis for patients diagnosed with HPV-associated carcinomas remains poor (52). Due to the wide range of biological functions of HIF-1α pathway in cancer biology especially in tumor angiogenesis, HIF-1α has emerged as a potential novel molecular target for chemoprevention and therapy of cancers, including cervical cancers (53). An increasing body of evidence has indicated that inhibition of HIF-1α pathway-induced angiogenesis by synthetic small molecules or naturally occurring functional food components may provide an effective means of chemoprevention of a variety of cancers (53, 54). Resveratrol, a natural antioxidant and polyphenol, has been considered a potential chemopreventive agent (55, 56). Recent studies have shown that resveratrol can potently inhibit HIF-1α activity by suppressing its synthesis as well as by promoting degradation (33, 34). In this study, we showed that resveratrol remarkably inhibited HPV-16 oncoprotein-induced HIF-1α protein accumulation and VEGF expression (Fig. 5) and activation of both Akt and ERK1/2 (Fig. 4A) in cervical cancer cells possibly via the PI3K/Akt and ERK1/2 signaling pathways. In addition, treatment with resveratrol dramatically suppressed in vitro capillary and tubelike structure formation by HUVECs stimulated by cervical cancer cells transfected with pSG5-HPV-16 E6 and E7 constructs (Fig. 6). Collectively, these data suggest that resveratrol potently inhibits HPV-16 oncoprotein-induced angiogenesis by suppressing HIF-1α/VEGF expression in cervical cancer cells. Because the progression of cervical cancer relies on tumor angiogenic activity, the use of resveratrol may enhance chemotherapy by directly suppressing its vascular support.

In summary, we have shown, for the first time to our knowledge, that high-risk HPV type 16 oncoproteins, E6 and E7, can induce HIF-1α protein accumulation and HIF-1α–dependent VEGF expression via ERK1/2 and PI3K/Akt signaling pathways, and such effects can be potently inhibited by resveratrol and siRNA specific for HIF-1α. These findings provided evidence that HPV-16 oncoproteins promote tumor angiogenesis possibly via up-regulating HIF-1α–dependent VEGF expression and may contribute, at least in part, to an elevated expression of HIF-1α in human cervical cancer patients. Our study also suggests that resveratrol can suppress tumor angiogenic activity in vitro and may be a potential antiangiogenic agent in the treatment of high-risk HPV-associated human cervical cancers.