Overexpression of human papillomavirus (HPV E6 and HPV E7) oncogenes in human cervical cells results in the development of cancer, and E6 and E7 proteins are therefore targets for preventing cervical cancer progression. Here, we describe the silencing of E6 and E7 expression in cervical carcinoma cells by RNA interference. In order to increase the efficacy of the RNA interference, HPV pseudovirions coding for a short hairpin RNA (shRNA) sequence were produced. The results indicated the degradation of E6 and E7 mRNAs when shRNA against E6 or E7 were delivered by pseudovirions in HPV-positive cells (CaSki and TC1 cells). E6 silencing resulted in the accumulation of cellular p53 and reduced cell viability. More significant cell death was observed when E7 expression was suppressed. Silencing E6 and E7 and the consequences for cancer cell growth were also investigated in vivo in mice using the capacity of murine TC1 cells expressing HPV-16 E6 and E7 oncogenes to induce fast-growing tumors. Treatment with lentiviruses and HPV virus-like particle vectors coding for an E7 shRNA sequence both resulted in dramatic inhibition of tumor growth. These results show the ability of pseudovirion-delivered shRNA to produce specific gene suppression and provide an effective means of reducing HPV-positive tumor growth. [Mol Cancer Ther 2009;8(2):357–65]
The role of human papillomaviruses (HPV) in human cancer has been shown in the last 15 years, including their significance as targets for therapy. HPV infection is responsible for approximately half a million cases of cervical cancer each year (1). HPV are small DNA viruses with a genome of ∼8 kb that naturally infect skin and mucosal epithelia. A group of HPVs including HPV-16, HPV-18, HPV-31, HPV-33, HPV-45, and others has been designated as “high-risk,” and persistent infection of the cervix with these high-risk HPVs is considered to be the necessary cause of cervical cancer (2, 3). High-risk HPVs encode E5, E6, and E7 proteins, the constant expression of which is needed for a cell to remain malignant (4). Cervical cancer is attributed to persistent infection with high-risk HPVs, and the E6 and E7 proteins of these types play a major role in HPV-induced carcinogenesis by interfering with host cell regulatory proteins (5).
The transforming properties of the E6 and E7 proteins are due to their capacity to bind to and inactivate the p53 tumor suppressor protein and members of the retinoblastoma tumor suppressor protein family, respectively (6–8). The E6 protein binds to E6AP, a ubiquitin ligase, and uses E6AP to target p53 for ubiquitination and thus activates proteasome-mediated degradation. In parallel, E7 contributes to cell transformation by binding to the human pRb protein and E2F transcription factors, resulting in the dissociation of pRb from E2F and premature cell progression into the S phase of the cell cycle (9). These viral oncogenes are therefore considered to be the ideal targets for the development of novel therapeutic strategies for cervical cancer.
There are currently no specific antiviral compounds that are active against HPV. The therapeutic advantage of silencing the expression of E6 and E7 genes by RNA interference has been shown to induce senescence in HPV-positive cell lines associated with a significant increase in p53 levels (10–12). The development of novel strategies termed small interfering RNAs (siRNA) has been described for endonucleasic cleavage and degradation of homologous mRNA (13). One important development was the finding that siRNA could be genetically encoded in an organism by expressing a short hairpin RNA (shRNA), consisting of a double-stranded sequence of 21 to 29 nucleotides, and a short loop region. When transcribed by the cell, this shRNA, which is converted by endogenous nucleases into a short RNA, is recognized by the RNA-induced silencing complex and used to target mRNA for degradation. Various DNA vector–based systems expressing siRNA in mammalian cells have been described. However, the in vivo use of shRNA therapy in cancer has been limited by obstacles related to effective delivery into the nuclei of target cancer cells.
It has been shown that heterologous expression of the major capsid protein of papillomaviruses (L1), as for the major structural proteins of other viruses, leads to the formation of virus-like particles (VLP; refs. 14–17). Papillomavirus infection in animal models and in humans can be prevented by vaccination using L1 VLPs as antigen (18, 19). The recent development of vaccines for papillomaviruses has represented a major advance in preventive methods for human cancer. These vaccines prevent infection by HPVs but they seem to have little or no therapeutic effect (20) because the L1 gene is not expressed in transformed cells (21, 22). Other methods of managing HPV-associated cancers are therefore needed to treat already infected women and men. Whereas most investigations concerning VLPs deal with vaccine development, it has also been shown that HPV VLPs can be used to package unrelated plasmids to form HPV pseudovirions, and they thus represent a valuable gene therapy delivery system (23–27). Because these artificial viruses (or pseudovirions) are synthetic, nonreplicating viruses, they are safe systems for gene delivery.
The main aims of this study were to construct an HPV-based shRNA expression system and to explore HPV-mediated RNA interference for effective E6 and E7 gene silencing. We show here that HPV pseudovirions expressing E6 and E7 shRNA in cervical cancer cells resulted in the depletion of E6 and E7 expression and suppression of cancer cell growth, suggesting that HPV VLP could be valuable to deliver plasmids encoding shRNA. This novel approach to achieving effective reduction in cervical tumor growth may prove to be valuable for the treatment of cervical cancer.
Materials and Methods
Cell Lines, Cell Culture, and Cell Transfection
The human cervical carcinoma cell lines CaSki (ATCC CRL-150) and C33-A (ATCC HTB-31; American Type Culture Collection) were grown at 37°C in a humidified atmosphere with 5% CO2 in DMEM supplemented with 10% FCS (Invitrogen), 1% penicillin/streptomycin, and 1% sodium pyruvate. CaSki cells were infected by HPV-16 and express E6 and E7, whereas C33-A cells were negative for HPV.
Murine cell line TC1 (ATCC CRL-2785) was cotransformed by HPV-16 E6/E7 oncoproteins and c-Has-Ras. These cells were grown in RPMI 1640 with 10 mmol/L of HEPES, 1 mmol/L of sodium pyruvate supplemented with 2 mmol/L of nonessential amino acids and 10% FCS. Murine 293FT cells (Invitrogen) were derived from the 293T cell line (ATCC CRL 11268). They express the SV40 large T antigen and were cultured in the presence of 500 μg/mL of geneticin (Invitrogen).
One day before transfection, cells were trypsinized and seeded into six-well plates. Cells were then transfected with plasmids encoding shRNA using Lipofectamine 2000 (Invitrogen), HPV-31 pseudovirions, or lentivirus coding for shRNA. Cells were harvested for analysis at various times as indicated in the results.
Design and Production of Plasmids Expressing E6- and E7-Specific shRNA
In order to produce the pENTR/U6 entry clone, the BLOCK-iT U6 entry vector was used. We first designed and synthesized complementary DNA oligos (Invitrogen), each containing four nucleotide overhangs necessary for directional cloning. Complementary sequences of shRNA corresponding to E6 and E7 siRNA are described in Fig. 1B. E6-2 and E7-2 sequences were selected using the “shRNA designer” software from Invitrogen. E6-1 and E7-1 sequences were those described by Jiang and Milner (12). We used LacZ and the control shRNA sequence described by Jiang and Milner (12) as controls. All sequences were BLAST-confirmed for specificity.
Equal amounts of the forward and reverse strand oligos were annealed to generate the double-stranded oligos by incubation in annealing buffer at 95°C for 4 min. After generating double-stranded oligos, they were ligated into the pENTER/U6 vector (Invitrogen). The plasmids were sequenced, propagated, and purified using the Qiagen plasmid midi kit (Qiagen, France). Entry clones were used for transient RNA interference analysis.
Production of E6 and E7 shRNA-expressing HPV Pseudovirions
HPV-31 VLPs were expressed in Sf21 cells infected with a recombinant baculovirus encoding codon-optimized HPV-31 L1 and L2 genes (28). Cells were incubated at 27°C for 72 h (23, 26). Cells were harvested by centrifugation, resuspended in PBS containing 0.5% NP40, and allowed to stand at room temperature for 30 min. Cell lysates were then centrifuged at 14,000 × g for 15 min at 4°C. The nuclear fraction was further resuspended in PBS and sonicated. The fraction was then loaded on top of a preformed cesium chloride gradient and centrifuged at equilibrium in a Beckman SW28 rotor (24 h, 27,000 rpm, 4°C). L1-positive fractions were pooled in PBS and centrifuged (SW28 rotor, 3 h, 28,000 rpm, 4°C). VLPs were resuspended in 0.15 mol/L of NaCl.
Pseudovirions were generated as previously described with some modifications (23). Briefly, 1 μg of HPV-31 VLPs were incubated in 50 mmol/L of Tris-HCl buffer (pH 7.5) containing 20 mmol/L of DTT and 1 mmol/L of EGTA for 30 min at room temperature. At this stage, expression plasmids encoding shRNA, luciferase, or green fluorescent protein (100 ng) were added to the disrupted VLPs. The preparation was then diluted with increasing concentrations of CaCl2 up to a final concentration of 5 mmol/L, with or without ZnCl2 (10 nmol/L). ZnCl2 was used because it has been reported that ZnCl2 enhances the assembly of HPV capsomers into VLPs (29). Pseudovirions were then dialyzed against PBS 1× overnight and stored at 4°C before use.
The presence of capsomers, VLPs, and pseudovirions was analyzed by electron microscopy. For this purpose, samples were applied to carbon-coated grids, negatively stained with 1.5% uranyl acetate and observed at ×50,000 nominal magnification using a JEOL 1010 electron microscope.
Production of E6 and E7 shRNA Expressing Lentivirus
Lentivirus production was done using the BLOCK-iT lentiviral RNA interference expression system (Invitrogen). Briefly, an LR recombination reaction between the pENTR/U6 plasmid encoding E7 shRNA and plenti6/BLOCK-iT-DEST was done to generate the plenti6/BLOCK-iT-DEST expression construct. Lentivirus was produced by transfecting 293FT cells with 9 μg of the ViraPower Packaging Mix and 3 μg of plenti6/BLOCK-iT-DEST expression plasmid DNA using LipofectAMINE 2000 (Invitrogen).
Cell culture supernatants were collected at 48 h posttransfection. Titers of E7 shRNA expressing lentivirus were determined by infecting TC1 cells with serial dilutions of lentivirus. The lentiviral stock titer was 1 × 10 6 TU/mL. Stably transduced TC1 cells were selected by placing cells under blasticidin selection (10 μg/mL).
Analysis of E6 and E7 mRNA Levels by Reverse Transcription PCR
CaSki cells (106) transfected with shRNA pseudovirions were washed with PBS 1× and then mRNA was isolated using the Dynabeads mRNA direct kit (Dynal France SA) according to the instructions of the manufacturer. Single-stranded cDNAs were synthesized from mRNAs by reverse transcription for 1 h at 42°C in 1× incubation buffer containing 250 μmol/L of each deoxynucleotide triphosphate, 5 μmol/L oligo(dT) 20, 25 units of RNase inhibitor, and 20 units of avian myeloblastosis virus reverse transcriptase (Roche Diagnostics). cDNA samples were subjected to PCR amplification with forward and reverse primers specific to HPV-16 E6, E7, or glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The primers used were E6 forward, 5′ CACCAAAAGAGAACTGCAATGT 3′; and reverse, TTGCTGTTCTAATGTT GTTCCA; E7 forward, 5′GGAGATACACCTACATTGCATGA3′; and reverse, GGGGCACACAATTCCTAGTG; glyceraldehyde-3-phosphate dehydrogenase forward, 5′ACAGTCCATGCCATCACTG CC3′; and reverse, 5′GCCTGCTTCACCACCTTCTTG3′. PCR was set up with 200 μmol/L of deoxynucleotide triphosphate, 2 μmol/L of each specific primer, and 1 unit of Taq polymerase (Invitrogen) in a GeneAmp 9700 thermocycler (Pekin Elmer Applied Biosystems, France) programmed for 25 cycles at 50°C, 55°C, or 62°C for E7, E6, and GAPDH, respectively. PCR products were visualized on 2% agarose gels and analyzed with GelDoc system (Bio-Rad).
Detection of p53
Cells (2 × 105) were washed with PBS 1×, dissolved directly in 50 μL of SDS gel loading buffer, and incubated for 10 min at 95°C. Fifteen microliters of each sample were separated on 12% SDS-PAGE gels. Separated proteins were electroblotted onto nitrocellulose membrane for antibody detection. Human p53 protein was detected using monoclonal antibody DO-1 (Santa Cruz Biotechnologies) and endogenous β-actin was detected using a polyclonal antibody (Sigma). Bound antibodies were visualized using an alkaline phosphatase–conjugated anti-mouse IgG antibody (Sigma) with nitroblue tetrazolium and bromochloroindolylphosphate (Sigma) as substrates.
Detection of β-Galactosidase and Luciferase Activity
Detection of β-galactosidase in CaSki, C33-A, and TC1 cells transfected with β-galactosidase plasmid and LacZ shRNA was undertaken in cells washed with PBS 1× and fixed with PBS containing 2% formaldehyde and 0.2% glutaraldehyde. After 10 min at room temperature, cells were washed and incubated with β-galactosidase revelation solution (2 mmol/L MgCl2, 4 mmol/L potassium ferrocyanide, 4 mmol/L potassium ferricyanide, and 1 mg/mL X-gal). Blue cells showing β-galactosidase activity were counted.
The detection of luciferase gene expression was measured by luminescence assay (Firefly luciferase assay kit; Interchim). The luminescence was integrated over 10 s (Victor2, Wallac; Perkin-Elmer) and the results were expressed as counts per second (cps) per well.
Cell Viability Assays
Transfected and untransfected cells were trypsinated and then seeded into 96-well plates. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (0.5 mg/mL) was added to cells seeded in 100 μL of DMEM without FCS and incubated at 37°C. After 2 h, 100 μL of isopropanol and 0.4 mmol/L of HCl were added and absorbance was measured at 540 nm. Cell viability was determined as the ratio between the absorbance obtained in test wells and the absorbance obtained in untreated cells.
Apoptosis was detected using the anti-ssDNA/APOSTAIN (Abcys). Briefly, TC1 cells (2 × 105) grown in six-well plates were transduced with 10 μg of pseudovirions containing 1 μg of E6-1, E6-2, E7-1, E7-2 shRNAs, and control shRNA. After 2 days of infection, supernatants were harvested and the cells were fixed with ice-cold ethanol. After centrifugation, 5 × 105 fixed cells were resuspended in 250 μL of formamide for 5 min at room temperature, then incubated at 75°C for 10 min. After this step, 2 mL of PBS containing 1% of nonfat dried milk were added. After 15 min, cells were centrifuged and the pellet was resuspended in 100 μL of PBS containing anti-ssDNA monoclonal antibody F7-26. After 15 min of incubation and centrifugation, the cell pellet was resuspended in 100 μL of PBS containing fluorescence-conjugated anti-mouse IgM (20 μg/mL in PBS and 1% nonfat dry milk). After 15 min of incubation, cells were rinsed, centrifuged, and resuspended in 500 μL of PBS containing 1 μg/mL of propidium iodide. Negative controls were treated with mouse IgM instead of the specific primary antibody. Cells were then analyzed using a Coulter XL flow cytometer and with the Expo32 software (Beckman Coulter, France).
Apoptosis was also investigated using the Caspase-Glo 3/7 assay kit (Promega). Cells were transduced as previously with different shRNA. Two days after transduction, 2 × 105 cells were transferred to each well of a white-walled 96-well plate (Perkin-Elmer) and 100 μL of a caspase luminogenic substrate was added. After 1 h incubation at room temperature, the plates were read on a Luminoscan Ascent luminometer (Thermo Electron) for luminescence.
Assessment of Antitumor Effects in a Mouse Model
To investigate the effects of E6 and E7 shRNA pseudovirions on the tumorigenicity of TC1 cells, six groups (five mice/group) of 6-week-old female C57BL6 mice (CERJ, Le Genest St Isle, France) were subcutaneously inoculated with TC1 cells (2 × 105). Before administration to mice, TC1 cells were transfected with pseudovirions (10 μg VLP/1 μg E6-1 or E7-1 shRNA), with a lentivirus encoding E7-1 shRNA (multiplicity of infection, 50), or with 1 μg of E7-1 shRNA with Lipofectamine. A control group received TC1 cells without treatment (mock), and one control group received TC1 cells treated with HPV VLPs. After 3 weeks, mice were sacrificed and tumors were excised and weighed.
The antitumor efficacy of the E7 shRNA-encoding pseudovirions was also investigated in mice with TC1 cell tumors. Two groups (10 mice/group) of 6-week-old female C57BL6 mice were s.c. inoculated with TC1 cells (2 × 105). Seven days after inoculation, palpable tumors had formed in all mice, and pseudovirions containing shRNA were directly injected into each tumor at a dose of 20 μg VLP/2 μg E7-1 or 20 μg VLP/2 μg control shRNA every 2 days for 2 weeks. After 3 weeks, mice were sacrificed and tumors were extracted and weighed. All animal studies were approved by the regional animal ethics committee (CREEA).
Data were expressed as mean ± SE. Statistical analysis was done using Student's t test, and P < 0.05 were considered significant.
Design of shRNAs Directed at the E6 and E7 Proteins of HPV-16
Six sequences were designed to promote specific silencing. Two of these sequences, i.e., siE6-1 (138–159) and siE7-1 (101–119), have already been described by Jiang and Milner (12) as siRNA, two other sequences, siE6-2 (421–441) and siE7-2 (148–167), corresponding to E6 and E7, respectively, were selected using the Invitrogen shRNA designer software, and two sequences (LacZ shRNA and Jiang's control shRNA) were used as controls (Fig. 1A and B). These sequences were annealed and inserted into pENTR/U6.
To investigate the effectiveness of the pseudovirion system for delivery of shRNA, TC1, CaSki, and C33-A cells were transfected with a plasmid coding for β-galactosidase, with or without the pENTR/U6 LacZ plasmid coding for LacZ shRNA. In the presence of LacZ shRNA, a decrease in β-galactosidase expression was observed in all cell lines investigated, with 80%, 83%, and 81% inhibition in TC1, CaSki, and C33-A cells, respectively.
Production of Pseudovirions Encoding shRNA by Assembly of L1 Capsomers into VLPs in the Presence of ZnCl2
HPV VLPs were used to encapsidate plasmids encoding for E6 and E7 shRNAs. To generate these pseudovirions, we used the disassembly-reassembly method as previously described, with the modification of adding ZnCl2 during the reassembly process. Briefly, purified VLPs were incubated in a buffer containing EGTA and DTT, and in these conditions, VLPs were completely disaggregated into structures resembling capsomers (Fig. 2A). E7-1 shRNA plasmid was then added and the preparation was diluted in a buffer containing DMSO and CaCl2 with or without ZnCl2 in order to refold the VLPs. The presence of ZnCl2 clearly increased the reassembly of capsomers into structures resembling pseudovirions (Fig. 2A). In the presence of ZnCl2, the capsomers also assembled into tubular structures of 24 nm in diameter with lengths varying from 120 to 280 nm (Fig. 2A). The role of ZnCl2 in the production of pseudovirions containing a plasmid coding for luciferase was evaluated by comparing the ability of the pseudovirion preparations obtained with or without ZnCl2 to transduce 293FT cells. Luciferase activity of 9,981 cps was obtained with pseudovirions generated in the presence of ZnCl2, whereas it was only 845 cps with pseudovirions obtained without ZnCl2. Thus, a 12-fold increase in luciferase activity was observed when the L1 capsomers were reassembled into VLPs in the presence of ZnCl2 (Fig. 2B). In addition, the capacity of such HPV pseudovirions to transduce CaSki, C33-A, and TC1 cells was investigated. Such pseudovirions transduced all cell lines investigated (Fig. 2B). However, a higher level of luciferase expression was observed in TC1 cells (7,990 cps) than in CaSki cells (5,232 cps) or C33 cells (4,016 cps). In the absence of pseudovirions, luciferase activity of 57, 84, 51, and 41 cps was observed in 293FT, TC1, CaSki, and C33 cells, respectively.
Transduction of CaSki Cells by E6 and E7 shRNA Pseudovirions Induced Increased Gene Silencing and Inhibition of Cell Growth
The efficacy of gene transfer of pseudovirions in cells was investigated using pseudovirions containing a plasmid encoding green fluorescent protein. Eighty percent of TC1 cells expressed green fluorescent protein as detected by flow cytometry (data not shown). We evaluated the efficacy of E6 and E7 shRNAs by assaying their ability to interfere with protein expression. For this purpose, we first obtained the full-length cDNAs for E6 and E7 (and GAPDH as control) by reverse transcription PCR. The results showed that E7 mRNA decreased in the presence of shRNA directed against E7, a high level of interference being obtained with the Jiang sequence, whereas a lower decrease was observed with the E7-2 sequence and no decrease with control shRNA (Fig. 3A). Similar results were obtained for the detection of E6 mRNA in cells treated with E6 shRNA. A slight decrease in E6 mRNA was observed when cells were treated with E7 shRNA.
Because E6 is expressed in CaSki cells at levels too low to be detected by Western blot analysis using the antibodies available, as reported by Gu et al. (30) and Yamato et al. (31), E6 shRNA activity was screened on the basis of its ability to restore p53 expression. Our results showed that the p53 level detected by Western blotting increased after expression of E6-1 shRNA over 24 hours (Fig. 3B), and decreased with time. In the presence of E7-1, E7-2, and E6-2 shRNA, the increase in p53 expression was lower than that observed with E6-1 shRNA. p53 expression was not detected when cells were treated with LacZ shRNA. These results suggest that the level of E6 protein was reduced by shRNA treatment, and that p53 not only accumulated but was also functionally active.
To verify whether inhibition of E6 and E7 expression could induce a reduction in cell viability, a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide test was used in shRNA transfected and nontransfected CaSki and TC1 cells. Our findings showed that E6 and E7 silencing induced a decrease in cell viability (Fig. 4). This reduction was greater with E7 shRNA (70% cell growth inhibition). No inhibition of cell growth was observed when HPV-negative cervical cells such as C33-A were transfected with the same shRNA, indicating that cell growth inhibition by E6 and E7 shRNAs was specific. The reduction in cell growth was similar for E6-1 and E6-2 and for E7-1 and E7-2 shRNAs.
Transduction of TC1 Cells by E6 and E7 shRNA Pseudovirions did not Induce Apoptosis
An apoptosis assay based on the increased sensitivity of DNA in apoptotic cells to thermal denaturation was carried out to investigate whether cell death was due to apoptosis. In this assay, DNA is denatured by heating in the presence of formamide and stained with monoclonal antibody F7-26 specific to ssDNA. Flow cytometry did not reveal a significantly increased number of apoptotic cells with E6-1, E6-2, E7-1, and E7-2 shRNA-transfected TC1 cells compared with control shRNA-transfected cells. To further assess the existence or absence of apoptosis, a caspase assay which measures the enzymatic activity of caspases 3 and 7 was also done on TC1 cells transduced by E6 and E7 shRNA pseudovirions. No increase in caspase activity was observed with both E6 and E7 shRNAs. The results suggested that E6 and E7 shRNAs induce reduction in TC1 cell growth but not through the induction of apoptotic cell death.
Transfection of TC1 Cells In vitro by E6 and E7 shRNA Pseudovirions Induced Reduction in Tumor Growth in Mice
The efficacy of the E6 and E7 shRNA pseudovirions was also investigated in vivo using the TC1/mice model. HPV pseudovirions coding for E6-1 and E7-1 shRNA, HPV VLP alone and a lentivirus encoding for E7-1 shRNA were used to transduce TC1 cells in vitro. After 24 hours, nontransfected and transfected TC1 cells were s.c. injected into C57BL6 mice.
Mice were sacrificed 21 days later and tumors were excised and weighed. In control mice (mock and VLP alone), tumor weight ranged from 1.09 to 2.12 g (mean, 1.64 g).
No reduction in tumor weight was observed with TC1 cells transfected with E7-1 shRNA and Lipofectamine (mean, 1.61 g). A 42% reduction in mean weight of tumors was observed with E6-1 HPV pseudovirions (P < 0.10) and a 70% to 87% reduction in mean size with the E7-1 lentivirus and E7-1 HPV pseudovirions, respectively (P < 0.05 and P < 0.001; Fig. 5).
Intratumoral Transduction of TC1 Cell Tumors by E7 shRNA Pseudovirions Induced a Reduction in Tumor Growth in Mice
The in vivo efficacy of the E7 shRNA-encoding pseudovirions was also investigated for their capacity to reduce tumor growth in mice with TC1 cell tumors. E7-1 shRNA was selected for these studies based on its in vitro efficacy. E7-1 shRNA pseudovirions were directly injected in tumors every 2 days over 2 weeks. After 1 week, no tumor growth reduction was observed in E7-1 shRNA-treated mice compared with the control mice (Fig. 6A). A decrease in tumor growth was then clearly observed during the second week of treatment; with a mean tumor weight of 1.05 g in mice treated with control shRNA and 0.49 g in mice treated with the E7-1 shRNA pseudovirions at the end of the second week (P = 0.045).
Overexpression of HPV E6 and E7 oncogenes in cervical cells results in the development of cancer. Tumor cells are typically resistant to growth-suppressive signals and apoptosis, and the viral oncogenes E6 and E7 provide such suppression in HPV-induced cancers. HPV-induced cervical cancer therefore seems to be an ideal candidate for therapy using RNA interference, a process which is several orders of magnitude more effective than antisense and ribozyme strategies (13). In addition, E6 and E7 genes are not present in normal cells, and thus, RNA interference–based therapy would not affect them. A number of groups have recently shown that suppression of E6 or E7 results in the activation of p53 and retinoblastoma tumor suppressor protein pathways in cervical cancer cells (12, 31–36). The major obstacle to the clinical development of siRNA is poor cellular uptake and bioavailability. In this study, we used a shRNA approach to suppress E6 and E7 gene expression, generating HPV pseudovirions to deliver shRNA. We investigated the efficacy of two shRNA for HPV-16 E6 (E6-1 and E6-2) and two for E7 (E7-1 and E7-2) in silencing E6 and E7, respectively. These shRNA plasmids were encapsidated into HPV-31 VLP to form pseudovirions and then used to transfect cervical cancer cells or cells expressing HPV-16 E6 and E7 proteins. It should be noted that the addition of ZnCl2 increased the production of pseudovirions, in agreement with the increase in aggregation of capsomers into VLPs in the presence of ZnCl2 as shown by Hanslip et al. (29). E6 mRNA levels decreased slowly when E6-2 was used, whereas a dramatic decrease was observed with E6-1. Inhibition of the E6 mRNA was more effective using E6-1 than E6-2. We also showed that a greater decrease in E7 mRNA levels was obtained with E7-1 than E7-2, in agreement with the fact that different siRNA within the same gene have different silencing abilities (37, 38).
Transfection of cervical cancer cells with E6 shRNA pseudovirions induced the accumulation of p53 in cells, as reported with other RNA interference systems (10, 11). A slight decrease in E6 mRNA was observed when cells were treated with E7 shRNA, in agreement with the results of Jiang and Milner (12) and Yamato et al. (33), and explaining the lower increase in p53. This result is in agreement with the fact that HPV-16–positive tumors expresses E6 and E7 protein from a single polycistronic mRNA, suggesting that targeting the E7 messenger sequence may result in the slight abolition of E6 mRNA. We thus showed in this study that shRNA delivered by pseudovirions could effectively and specifically suppress targeted oncogene E6 and E7 expression and lead to CaSki and TC1 cell death. As shown by Jiang and Milner (12) and by Yamato et al. (33), targeting E6 with shRNA resulted in weak growth inhibition, whereas E7 silencing induced more significant inhibition. This E7 inhibition could be caused by a weak increase in p53 and by this E7-1 sequence, leading to the dephosphorylation of pRb, as described by Jiang and Milner (12), and inducing the modification of cell growth. To show that cells transduced by E6 and E7 shRNA HPV pseudovirions had become apoptotic, the cells were investigated for DNA fragmentation and for the activation of caspases 3 and 7. Apoptosis was not observed in TC1 cells, in contradiction with results obtained by Jiang and Milner (12) and Sima et al. (32), in CaSki cells. One of the specific features of TC1 cells, a mouse hepatocyte cell line, is that they are transformed by E6 and E7, and also by H-Ras. In a majority of studies, p21Ras protein seems to deliver an antiapoptotic signal (39, 40), leading to speculation that in TC1 cells, the p21Ras protein may act by blocking the apoptosis induced by E7 down-regulation.
Efficacy was also investigated in vivo using the TC1/mouse model (41). Pseudovirions coding for LacZ, E6-1, and E7-1 shRNAs were used to transduce TC1 cells. The results indicated that TC1 cells treated with E7-1 shRNA pseudovirions showed retarded tumor formation, similar to tumors treated with E7-1 shRNA lentivirus, and at much higher levels than those observed in mice treated with E7-1 shRNA Lipofectamine. However, it is not clear if the in vivo effect of E7 shRNA was due to the decrease of E6 or E7 mRNA. The in vivo experiment showed that the effect of E6 shRNA on tumor growth was more limited than the effect of E7 shRNA (Fig. 5). We and others (12, 35) have shown that E6 shRNA only decreases the E6 mRNA, and that E7 shRNA decreases both E6 and E7 mRNA. Thus, it could be hypothesized that the reduction in size observed with E7 shRNA was due in part to the effect of E7 mRNA silencing.
In addition, pseudovirions coding for E7-1 shRNA were investigated for their capacity to reduce tumor growth when injected intratumorally in mice with TC1 tumors. The results confirmed that E7-1 shRNA pseudovirions are able to reduce tumor development, and thus suggest that HPV pseudovirion delivery of shRNA is an effective way to achieve suppression of E6/E7 oncogene expression and induce inhibition of tumor growth in vivo. Our results suggest that in TC1 cells, the reduction in tumor size observed in mice by treatment with E7 shRNA pseudovirions must be mainly due to growth arrest. To increase efficacy, coadministration of E6 and E7 shRNA/pseudovirions should be investigated.
In conclusion, HPV pseudovirions encoding shRNA may provide another means of cervical cancer therapy, and using shRNA/HPV pseudovirions seems to be a more promising strategy than using siRNA alone. The presence of L1 capsids would also induce local immunization against L1 neutralizing epitopes, thus conferring protection in cases of HPV reinfection after the death of E7-expressing cells. These pseudovirions thus represent a new step in designing rational molecular cancer therapy using RNA interference, at the same time, offering protection against reinfection by the causative agent of cervical cancer.
Disclosure of Potential Conflicts of Interest
No potential conflict of interest relevant to this article was reported.
Grant support: “Ligue contre le Cancer” (Comités départementaux d'Indre et Loire et de l'Indre) and by “Vaincre la Mucoviscidose” (P. Coursaget), a doctoral grant from INSERM and the Région Centre (E. Alvarez), and by a postdoctoral grant from the Ligue contre le Cancer and from the Zonta Club (L. Bousarghin).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.