- Email: [email protected]
E7 antisense RNA and induction of apoptosis in cervical cancer
Gynecologic Oncology 103 (2006) 820 – 830 www.elsevier.com/locate/ygyno
Adenovirus-mediated transfer of human papillomavirus 16 E6/E7 antisense RNA and induction of apoptosis in cervical cancer Katsuyuki Hamada a,⁎, Toshiro Shirakawa b , Akinobu Gotoh c , Jack A. Roth d , Michele Follen e a
c d
Department of Obstetrics and Gynecology, School of Medicine, Ehime University, Shitsukawa,Toon, Ehime 791-0295, Japan b Division of Infectious Disease Control, International Center for Medical Research and Treatment, Faculty of Medicine, Kobe University Graduate School of Medicine, Kusunoki, Chuo-ku, Kobe, Hyogo 650-0017, Japan Laboratory of Cell and Gene Therapy, Institute of Advanced Medical Sciences, Hyogo College of Medicine, Mukogawa, Nishinomiya, Hyogo 663-8501, Japan Department of Thoracic and Cardiovascular Surgery, The University of Texas, M.D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030, USA e Department of Gynecologic Oncology, The University of Texas, M.D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030, USA Received 22 March 2006 Available online 5 September 2006
Abstract Objective. In most cervical cancers, human papillomaviruses (HPVs) are identified. The E6 and E7 genes of HPVs encode proteins, that interfere with the function of the tumor suppressor proteins p53 and Rb. We are exploring the potential use of antisense HPV RNA transcripts for gene therapy for HPV-positive cervical cancers. Methods. Via a recombinant adenoviral vector, Ad5CMV-HPV 16 AS, we introduced the antisense RNA transcripts of the E6 and E7 genes of HPV type 16 into human cervical cancer SiHa cells harboring HPV 16. We then analyzed the effects of expression of these genes on cell and tumor growth. Results. HPV 16 E6/E7 antisense RNA was detected for 14 days in Ad5CMV-HPV 16 AS-infected cells. After infection, E6 and E7 protein expression was suppressed, and p53 and Rb protein expression increased. The Ad5CMV-HPV 16 AS-infected cells underwent apoptosis in vitro and in vivo. Cell growth and tumorigenicity were greatly suppressed. Ad5CMV-HPV 16 AS treatment significantly reduced the volumes of established subcutaneous tumors. Conclusion. Transfection of cervical cancer cells with HPV 16 E6/E7 antisense RNA in a form such as Ad5CMV-HPV 16 AS might be a potentially useful approach to the therapy of HPV 16-positive cervical cancer. Published by Elsevier Inc. Keywords: Recombinant adenovirus; HPV 16; Antisense; Cervical cancer; Apoptosis; E6; E7
Introduction Cervical cancer, the second most common malignancy in women worldwide, remains an important health problem for women [1]. Incidence rates are currently increasing in American women [2]. Human papillomavirus (HPV) infection has been identified as an etiologic agent in cervical cancer [1]. Only particular HPVs, such as HPV types 16 and 18, are associated with high risk of malignant conversion [3]. The viral DNA is randomly integrated into host cell genomic ⁎ Corresponding author. Fax: +81 899 60 5381. E-mail address: [email protected] (K. Hamada). 0090-8258/$ - see front matter. Published by Elsevier Inc. doi:10.1016/j.ygyno.2006.06.035
sequences, and HPV-specific RNA and proteins can be detected in the tumors and cell lines derived from these host cells [4]. The HPV genome contains seven early genes, two of which encode the oncoproteins E6 and E7. E6 can form complexes with the tumor suppressor p53 protein and promote p53 degradation through a ubiquitin-dependent protease system [5]. Inactivation of the p53 gene by allelic loss or by point mutation is infrequent in primary cervical cancer [6]. E7 binds to the tumor suppressor Rb protein, preventing Rb from binding to its normal substrate, the E2F transcription factor [7]. The expression of the viral E6 and E7 proteins in HPV-infected cervical cancer cells interferes with
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
p53 and Rb function despite expression of wild-type p53 and Rb alleles. E6 and E7 act independently or synergistically to potentiate the immortalization of their natural host cells, primary human squamous epithelial cells [8]. It has been demonstrated that the inactivation of the Rb and p53 proteins is an important step in cervical carcinogenesis [9]. Furthermore, the continued expression of the E6 and E7 regions of the viral genome appears to be necessary for the maintenance of the proliferative and malignant phenotypes of cervical cancer cells [10,11]. Gene expression can be selectively inhibited by introducing complementary antisense messages [12]. The efficacy of antisense inhibition has been demonstrated for a variety of genes, particularly oncogenes [12–14]. These intriguing findings lead to the possibility of using antisense messages to develop gene therapy techniques for human cervical cancers that contain HPV DNA. A plasmid expressing antisense HPV 18 RNA was shown to decrease the growth rate of the HPV 18-positive human cervical cancer cell line HeLa [15,16], but the low transduction efficiency of the plasmid would be a problem for human clinical therapeutic trials. Antisense oligonucleotides to E6 and E7 of HPV 16 and HPV 18 were shown to selectively inhibit the growth of HPV 16- and 18-positive cancer cells, respectively, but not HPV-negative cells [17,18]. However, when the oligonucleotides were withdrawn, the remaining cells recovered and grew as before; any therapy with oligonucleotides might therefore require continuous administration. Adenoviral vectors have many potential advantages over other viral vector systems and other techniques for introducing DNA into eukaryotic cells. Adenoviruses can be produced at very high titers, have high infectivity in eukaryotic cells, and have had a good clinical safety record in previous human and primate studies [19,20]. Several adenoviral vectors containing antisense expression units have shown growth inhibitory effects in vitro but not in vivo [21– 24]. There have been no reports of the efficacy of an adenoviral construct expressing antisense HPV RNA. In the study presented here, the E6/E7 region of HPV 16 in the antisense orientation was inserted into an adenoviral vector. We chose HPV 16, because it is identified in 60% of cervical cancers [3]. We evaluated the effects of the antisense HPV 16 E6/E7 construct (Ad5CMV-HPV 16 AS) on growth of human cervical cancer cells in vitro, ex vivo, and in vivo in a nude mouse model. We also evaluated whether the mechanism of the growth-suppressive effect that we observed is related to induction of apoptosis. Materials and methods Cell lines and culture conditions The human cervical cancer cell lines SiHa, CaSki, MS751, and C33A were obtained from the American Type Culture Collection. SiHa and CaSki are HPV 16-positive cell lines and have wild-type p53, MS751 is an HPV 18-positive cell line and has wild-type p53, and C33A is an HPV-negative cell line and has mutant p53 [25]. Cells were grown in RPMI medium supplemented with 10% heat-inactivated fetal bovine serum.
821
Generation of recombinant adenovirus The HPV 16 genome, cloned into pBR322, was provided by the Japanese Collection of Research Bioresources (JCRB, Tokyo). The entire E6/E7 region (nt 45 to 877) was amplified by PCR using an upstream primer with a NotI site and a downstream primer with a HindIII site. This fragment was cloned in antisense orientation into pAdE1CMV, a shuttle plasmid that contains the left end of Ad5 (0–16 map units [m.u.]) with the E1 region (1.25–9.2 m.u.) substituted by an expression cassette containing the cytomegalovirus (CMV) promoter, a multicloning site, and a simian virus (SV40) polyadenylation signal. The recombinant plasmid with the E6/E7 region in antisense orientation, pAdE1CMV-HPV 16 AS, was cotransfected with pJM17, a recombinant plasmid into 293 cells to generate the recombinant adenovirus, Ad5CMV-HPV 16 AS. An adenoviral vector, DL312, was used as the control.
Adenoviral infection Recombinant adenovirus was purified by double cesium-gradient ultracentrifugation as described elsewhere [26]. Viral stocks were propagated in 293 cells. Cells were harvested 36–40 h after infection, and cell debris was removed by subjecting the lysed cells to CsCl-gradient centrifugation. Concentrated virus was dialyzed and stored in aliquots at −80°C. Infection was carried out by adding the virus to high-glucose Dulbecco's modified Eagle's medium and to the cell monolayers. The viral titers were determined by plaque assays [26]. CAR expression was assessed by scoring X-gal-positive blue cells in each of three replicate dishes after the infection with Ad5CMV-LacZ. All cell lines inoculated with a single dose of Ad5CMV-LacZ at 10 MOI (multiplicity of infection) or greater exhibited 100% blue cells.
PCR analysis DNA was extracted using phenol–chloroform and precipitated with ethanol and sodium acetate. Total RNA was prepared by lysing cell monolayers in guanidinium isothiocyanate and centrifuging over a 5.7 M CsCl solution. RTPCR was performed from 6 μg of total RNA to generate cDNA. These DNA samples and cDNA products were then amplified by PCR, using primers for the antisense E6/E7 of HPV 16 and the polyadenylation signal of the adenovirus, yielding a 486-bp product. PCR products were analyzed on 2% agarose gels and stained with ethidium bromide.
Cell count assay Cells were plated at a density of 5 × 104/well in 12-well plates in triplicate. RPMI supplemented with 10% heat-inactivated fetal bovine serum was used as the growth medium. Cells were infected with either Ad5CMV-HPV 16 AS or DL312 control. Culture medium was used for mock infection. Cells were harvested using trypsin and counted. Cell viability was determined by trypan blue exclusion after harvest.
[3H]Thymidine incorporation assay Cells were cultured at 1 × 103 cells/well in 96-well flat-bottomed plates. Cells were infected with Ad5CMV-HPV 16 AS or DL312 and cultured in medium for 5 days. Each well was pulse-treated with 1 μCi [3H]thymidine (sp. act., 6.7 Ci/ mmol; Amersham Corp., Arlington Heights, IL) for an additional 24 h, after which the cells were harvested with a PhD cell harvester (Cambridge Technology Inc., Cambridge, MA). Individual filter discs were then processed for liquid scintillation counting. The data are presented as the means of quintuplicate samples.
Western blot analysis Total cell lysates were prepared by lysing cell monolayers in plates with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer after rinsing the cells with phosphate-buffered saline (PBS). Each lane was loaded with 10 μg of cell lysate protein as determined by BCA
822
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
protein assay (Pierce, Rockford, IL). After electrophoresis at 20 mA for 2 h, the proteins in the gels were transferred to Hybond-ECL membranes (Amersham). The membranes were blocked with 1% dry milk and 0.1% Tween 20 (Sigma Chemical Co., St. Louis, MO) in PBS and probed with a primary antibody, mouse anti-human p53 monoclonal antibody DO7 (DAKO Co., Carpinteria, CA) or rabbit anti-human Rb polyclonal antibody C-15 (Santa Cruz Biotechnology Inc., Santa Cruz, CA), and secondary antibodies, a biotinylated goat anti-mouse and anti-rabbit antibodies mix (DAKO). Then the membranes were incubated with horseradish peroxidase-conjugated streptavidin (DAKO). The membranes were developed according to the Amersham ECL protocol.
Immunoprecipitation analysis Cells were metabolically labeled for 12 h with 1 mCi of 35S Translabel (ICN Radiochemicals, Irvine, CA) per 10-cm dish in methionine and cysteine-free RPMI medium containing 10% dialyzed fetal calf serum. Cells were lysed at 4°C in 1 ml of radioimmunoprecipitation assay (RIPA) buffer [150 mM NaCl, 50 mM Tris–HCl (pH 8.0), 1% NP-40]. Insoluble debris was pelleted at 10,000×g for 10 min, and the supernatant was incubated with mouse monoclonal anti-HPV 18/16 E6 antibody (Oncogene Sciences, Uniondale, NY) or mouse monoclonal anti-HPV 16 E7 antibody (Triton Diagnostics, Alameda, CA), and protein G–Sepharose beads (Sigma). After they were washed with RIPA buffer, the beads were boiled in SDS sample buffer and loaded onto SDS-polyacrylamide gels.
Immunohistochemical analysis The Ad5CMV-HPV 16 AS-infected SiHa cells were analyzed to detect p53 protein expression. SiHa cells growing on coverslips were air dried and then fixed with 50% ethanol and 50% acetone at − 20°C for 20 min. Cells were treated with 3% H2O2 in methanol. Immunohistochemical staining was performed with the Vectastain Elite kit (Vector Laboratories Inc., Burlingame, CA). The primary antibody was a mouse anti-human p53 monoclonal antibody DO7 (DAKO), and the secondary antibody was a biotinylated goat anti-mouse IgG (Vector Labs). An avidin and biotinylated horseradish peroxidase macromolecular complex reagent (Vector Labs) was used to detect the antigen–antibody complex. The cells were stained with diaminobenzidine and then counterstained with Harris hematoxylin (Sigma).
Tumorigenicity assay SiHa, MS751, and C33A cells were infected with Ad5CMV-HPV 16 AS or DL312 at 10 MOI. An equal number of cells were treated with medium as a mock infection. Three hours after infection, the treated cells were harvested and rinsed twice with PBS. For each treatment, 5 × 106 cells in a volume of 100 μl were injected subcutaneously into a nude (nu/nu) female mouse (aged 4– 5 weeks; CLEA Japan). Experiments were reviewed and approved by institutional committees for animal care and use and for recombinant DNA research in School of Medicine, Ehime University. Seven mice were used for each treatment. The mice were examined every day and tumor formation and size were evaluated for 40 days. The tumors were measured every other day with calipers in two perpendicular diameters without knowledge of treatment groups. Tumor volume was calculated by assuming a spherical shape with the average tumor diameter calculated as the square root of the product of cross-sectional diameters.
Inhibition of tumor growth in vivo To determine inhibition of tumor growth in vivo, Ad5CMV-HPV 16 AS was injected into an established tumor in each female nude (nu/nu) mouse (CLEA). Experiments were reviewed and approved by institutional committees for animal care and use and for recombinant DNA research in School of Medicine, Ehime University. Five million SiHa, MS751, or C33A cells in 100 μl of PBS were injected into the right posterior flank of each mouse through an insulin syringe with a 28 1/2-gauge needle. Ten animals were used for each group. After 20– 25 days, tumors with a diameter of 5 to 6 mm had grown in all animals. Either 100 μl of Ad5CMV-HPV 16 AS or DL312 at 2 × 109 plaque-forming units (PFU) or PBS only was injected intratumorally on days 0, 1, 2, 3, 4, and 5. The tumors were measured every other day and tumor volume was calculated as described above.
Statistical analysis Values are expressed as the mean ± SD, and analyzed with the unpaired t-test. Statistical significance was set at P < 0.05.
Results
TUNEL assay The Ad5CMV-HPV 16 AS-infected cells and injected tissues were analyzed to detect apoptotic cells by the previously described TUNEL technique [27]. The slides were treated with 3% H2O2 in methanol. The nuclei of the cells were stripped from the proteins by incubation with 0.002% protease (Sigma). Next, the slides were covered with TdT (0.1 U/μl; USB, Cleveland, OH) and biotinylated dUTP (0.4 nM; Boehringer Mannheim Biochem., Indianapolis, IN) in TdT buffer (30 mM Tris, pH 7.2; 140 mM sodium cacodylate; 1 mM cobalt chloride) in a humid atmosphere of 37°C for 120 min. The reaction was terminated by transferring the slides to termination buffer (300 mM sodium chloride, 30 mM sodium citrate). Next, the slides were incubated with an avidin and biotinylated horseradish peroxidase macromolecular complex (Vector Laboratories Inc., Burlingame, CA) and stained with diaminobenzidine. Finally, the slides were counterstained with Harris hematoxylin (Sigma).
Flow cytometry analysis To measure numbers of apoptotic cells, ethanol-fixed cells on the third day after infection with DL312 or Ad5CMV-HPV 16 AS were incubated with propidium iodide (50 μg/ml) and ribonuclease (20 μg/ml) for 20 min at 37°C [28]. All measures were made with an EPICS profile (Coulter Corporation, Hialeah, FL) equipped with an air-cooled argon ion laser emitting 488 nm at 15 mW. A minimum of 10,000 events per sample were analyzed and FITC fluorescence was collected using a 525 BP filter. Coulter's cytologic program was used for data analysis. Mean peak fluorescence was determined for each histogram.
Generation of the recombinant adenovirus Ad5CMV-HPV 16 AS The entire E6/E7 region was amplified and cloned in antisense orientation into a shuttle plasmid vector to form the plasmid pAdE1CMV-HPV 16 AS. Ad5CMV-HPV 16 AS was generated after cotransfecting pAdE1CMV-HPV 16 AS and pJM17 into 293 cells. The expected genomic structure of Ad5CMV-HPV 16 AS is depicted in Fig. 1A. A 0.5-kb HPV 16 E6/E7 antisense-specific band amplified through the PCR confirmed the identity of Ad5CMV-HPV 16 AS (Figs. 1A and B). SiHa and CaSki cells (HPV 16-positive cervical cancer cell lines) were infected with either Ad5CMV-HPV 16 AS or a control adenoviral vector, DL312. Culture medium was used as the mock infection control. RNA was extracted 72 h after infection. In RT-PCR analysis, the transcript of antisense E6/E7 RNA from Ad5CMV-HPV 16 AS was detected in Ad5CMVHPV 16 AS-infected cells, while DL312-infected cells did not show any adenoviral antisense transcripts (Fig. 1C). Mockinfected cells also did not show any adenoviral antisense transcripts (data not shown). The time course of expression of antisense E6/E7 RNA from Ad5CMV-HPV 16 AS-infected CaSki cells is shown in Fig. 1D. Antisense E6/E7 RNA was
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
823
Fig. 1. (A) A map of Ad5CMV-HPV 16 AS genomic DNA, showing locations of the CMV promoter, HPV 16 E6/E7 antisense, the SV40 polyadenylation signal (pA), and the PCR primers. Primers define 486 bp of HPV 16 E6/E7 antisense. (B) The newly generated HPV 16 E6/E7 antisense recombinant adenovirus (Ad5CMV-HPV 16 AS) was identified by PCR analysis of the DNA samples. DNA templates used in each reaction were pAdE1CMV-HPV 16 AS, DL312, no DNA, a sample from the supernatant of Ad5CMV-HPV 16 AS-infected 293 cells after cytopathic effect, and the purified adenovirus (Ad5CMV-HPV 16 AS). (C) RT-PCR analysis for expression of the HPV 16 E6/E7 antisense RNA. DNA templates shown are from DL312-infected SiHa and CaSki cells with RT and without RT and from Ad5CMVHPV 16 AS-infected SiHa and CaSki cells with RT and without RT. RNA from each sample was extracted 72 h after incubation with adenoviruses at 10 MOI. (D) RTPCR analysis for expression of the HPV 16 E6/E7 antisense RNA. RNA was extracted from HPV 16 E6/E7 antisense-infected CaSki cells at 10 MOI before infection and on days 1, 3, 5, 7, 14, and 21.
expressed in the Ad5CMV-HPV 16 AS-infected CaSki cells for 14 days but not on day 21. Growth inhibition of cervical cancer cells by Ad5CMV-HPV 16 AS in vitro To examine whether the Ad5CMV-HPV 16 AS virus could inhibit growth of human cervical cancer cells, the cell lines SiHa, CaSki, MS751 (HPV 18-positive cervical cancer cell line), and C33A cells (HPV-negative cervical cancer cell line) were treated with Ad5CMV-HPV 16 AS, DL312, or mock
infection. The growth of the Ad5CMV-HPV 16 AS-infected SiHa and CaSki cells was greatly suppressed when estimated by [3H]thymidine incorporation and cell count assay 6 days after infection (Figs. 2A and B). In contrast, the growth of the Ad5CMV-HPV 16 AS-infected MS751 and C33A cells did not differ from that of mock- and DL312-infected cells (Figs. 2A and B). The time course of growth of SiHa, MS751, and C33A cells is shown in Fig. 2C. Growth of Ad5CMV-HPV 16 ASinfected MS751 and C33A cells was not significantly different from that of the mock- and DL312-treated cells. Growth of Ad5CMV-HPV 16 AS-infected SiHa cells started to decrease
824 K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830 Fig. 2. (A) Inhibition of the growth of SiHa, CaSki, MS751, and C33A cells as determined by [3H]thymidine incorporation assay. The cells were inoculated at densities of 1 × 103 cells/well in each 96-well plate 24 h before infection. At each indicated MOI, cells in five wells were infected for 5 days, pulse-treated with [3H]thymidine for an additional 24 h, trypsinized, harvested, and counted by liquid scintillation counter. The mean counts per minute for quintuplicate wells on day 6 after infection were plotted. Bars indicate standard deviation. (B) Inhibition of the growth of SiHa, CaSki, MS751, and C33A cells as determined by a cell count assay. The cells were inoculated at densities of 5 × 104 cells/well in each 12-well plate 24 h before infection. At each indicated MOI, cells in three wells on each well plate were infected. Six days after infection, cells were trypsinized and counted. The mean cell counts for triplicate wells were plotted. Bars indicate standard deviation. (C) Time course of growth of SiHa, MS751, and C33A cells as determined by a cell count assay. The cells were infected with Ad5CMV-HPV 16 AS or DL312 at 10 MOI in SiHa and MS751 cells and at 5 MOI in C33A cells. Culture medium alone was used as the mock infection control. Triplicate wells for each treatment were counted at each indicated time point after infection, and the mean cell counts were plotted. Bars indicate standard deviation. (D) Effects of treatment with Ad5CMV-HPV 16 AS on tumor growth of SiHa cells in nude mice. Mice were injected subcutaneously with 5 × 106 SiHa cells/mouse. After tumors with a diameter of 5–6 mm were obtained, 100 μl of Ad5CMV-HPV 16 AS or DL312 at 2 × 109 PFU or PBS alone was injected into tumors of 5 to 6 mm in diameter. Intratumoral injections were made on days 0, 1, 2, 3, 4, and 5. Ten mice made up each treatment group. The means of tumor volumes were plotted. Bars indicate standard deviation.
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
825
Fig. 2 (continued).
significantly compared with the mock- and DL312-infected cells 54 h after infection. Effect of Ad5CMV-HPV 16 AS on E6 and E7 expression To examine whether the Ad5CMV-HPV 16 AS virus could decrease the expression of E6 and E7 proteins, immunopreci-
pitation analysis was done using the mouse anti-human HPV 18/16 E6 antibody and mouse anti-human HPV 16 E7 antibody. The amounts of E6 and E7 proteins produced were compared following 72 h of infection with Ad5CMV-HPV 16 AS or DL312 at 10 MOI or mock infection with culture medium (Fig. 3A). E6 and E7 bands were observed in cellular extracts immunoprecipitated from mock- and DL312-infected SiHa
Fig. 3. (A) Immunoprecipitation analysis. SiHa cells were infected for 72 h with Ad5CMV-HPV 16 AS or DL312 at 10 MOI or mock infected. The 35S-labeled E6 and E7 proteins were immunoprecipitated with anti-E6 and anti-E7 antibodies from the cell lysates. Immunocomplexes were electrophoresed in SDS-PAGE and autoradiographed. (B) Western blot analysis. SiHa cells were infected as in (A). Cellular extracts were subjected to SDS-PAGE. Western blots were probed with antip53 and anti-Rb antibodies.
826
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
cells. Expression of E6 and E7 proteins was lower in the Ad5CMV-HPV 16 AS-infected SiHa cells than in mock- and DL312-infected cells. Effect of Ad5CMV-HPV 16 AS on p53 and Rb expression To determine the expression of the p53 and Rb proteins in the Ad5CMV-HPV 16 AS-infected SiHa cells, Western blot analysis was done using the mouse anti-human p53 antibody and the rabbit anti-human Rb antibodies. The amounts of p53 and Rb proteins produced were compared following 72 h of infection with Ad5CMV-HPV 16 AS or DL312 at 10 MOI or mock infection (Fig. 3B). The p53 and Rb proteins were highly expressed in Ad5CMV-HPV 16 AS-infected SiHa cells. Samples isolated from mock- and DL312-infected cells had very low levels of the p53 and Rb proteins. Immunohistochemical analysis was done using the mouse anti-human p53 antibody to determine the expression of the p53 protein in the Ad5CMV-HPV 16 AS-infected
SiHa cell. Immunohistochemical analysis of SiHa cells infected with Ad5CMV-HPV 16 AS at 10 MOI revealed staining of p53 protein in the nucleus 12 h after infection, whereas DL312-infected cells failed to show p53 staining (Fig. 4A). Induction of apoptosis by Ad5CMV-HPV 16 AS To characterize the effects of Ad5CMV-HPV 16 AS, the morphological changes in SiHa cells were examined after infection with Ad5CMV-HPV 16 AS or DL312 at 10 MOI. Within 36 to 72 h after infection with Ad5CMV-HPV 16 AS, an apparent morphological change occurred, with portions of the cell population rounding up and their outer membranes forming blebs. These changes are part of a series of histologically predictable events that suggest programmed cell death [27]. On the other hand, cells infected with the control adenovirus, DL312, grew normally with no histomorphological abnormalities.
Fig. 4. (A) Immunohistochemical staining of p53 in SiHa cells. Left, 12 h after DL312 infection at 10 MOI; right, 12 h after Ad5CMV-HPV 16 AS infection at 10 MOI. ×100. (B) TUNEL staining in SiHa cells. Left, 48 h after DL312 infection at 10 MOI; right, 48 h after Ad5CMV-HPV 16 AS infection at 10 MOI. ×100. (C) TUNEL staining in SiHa cell tumor. Left, 48 h after DL312 injection at 2 × 109 PFU; right, 48 h after Ad5CMV-HPV 16 AS injection at 2 × 109 PFU. ×100.
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
To determine whether the morphological changes seen in the Ad5CMV-HPV 16 AS-infected cells are related to the apoptotic process, SiHa cells were analyzed by the TUNEL method and flow cytometry. When the TUNEL method, which stains for broken DNA ends via a terminal deoxynucleotide transferase reaction, was performed, the DL312-infected SiHa cells and DL312-injected SiHa cell tumor did not stain, but nuclei in the Ad5CMV-HPV 16 AS-infected SiHa cells and the Ad5CMVHPV 16 AS-injected SiHa cell tumor stained 48 h after infection at 10 MOI (Fig. 4B) and 48 h after injection at 2 × 109 PFU (Fig. 4C). Flow cytometry analysis was performed to compare the amounts of apoptotic cells. The SiHa cells were infected with Ad5CMV-HPV 16 AS or DL312 at 10 MOI for 72 h and then fixed and stained with propidium iodide. The results of flow cytometry analysis are shown in Fig. 5. Apoptotic cells were evident in the Ad5CMV-HPV 16 AS-infected cells but not in the DL312-infected cells.
827
Table 1 Effect of pretreatment with Ad5CMV-HPV 16 AS on tumorigenicity of SiHa, MS751, and C33A cells in nude mice Cell line
Treatment
SiHa
Medium DL312 Ad5CMV- HPV 16 AS MS751 Medium DL312 Ad5CMV- HPV 16 AS C33A Medium DL312 Ad5CMV- HPV 16 AS
No. of tumors/ Mean tumor volume no. of mice (mm3 ± SD)
(%)
7/7 6/7 0/7
425 ± 130 294 ± 95 0±0
(100) (69) (0)
7/7 6/7 7/7
843 ± 276 726 ± 296 700 ± 197
(100) (86) (83)
7/7 5/7 6/7
1001 ± 417 635 ± 236 661 ± 169
(100) (63) (66)
Cells were treated with Ad5CMV-HPV 16 AS or DL312 at 10 MOI or mock infection for 3 h. Tumor sizes were determined 40 days after treated cells (5 × 106 cells/mouse) were injected.
Inhibition of tumorigenicity by Ad5CMV-HPV 16 AS To examine whether the Ad5CMV-HPV 16 AS virus could inhibit tumorigenicity of human cervical cancer cells, nude mice were injected subcutaneously with SiHa, MS751, and C33A cells to initiate tumor formation. Each mouse received one injection of 5 × 106 cells that had been infected with Ad5CMVHPV 16 AS or DL312 at 10 MOI or mock-infected with medium alone for 3 h. Tumors formed from the mock- and DL312infected SiHa cells; mice that received Ad5CMV-HPV 16 AStreated SiHa cells did not develop tumors (Table 1). Tumors formed from the mock-, DL312-, and Ad5CMV-HPV 16 ASinfected MS751 and C33A cells.
days 0, 1, 2, 3, 4, and 5. In the mice treated with PBS only or DL312, SiHa cell tumors continued to grow rapidly throughout the treatment, whereas in the mice treated with Ad5CMV-HPV 16 AS, tumor growth was greatly reduced (Fig. 2D). Six injections of Ad5CMV-HPV 16 AS reduced the size of SiHa cell tumors by 84% compared with six injections of PBS only, whereas six injections of DL312 did not reduce significantly tumor size. Six injections of Ad5CMV-HPV 16 AS did not suppress the growth of MS751 or C33A cell tumors. There was no significant difference in the body weights of any of the mice treated in this experiment. No animal died during this experiment.
Inhibition of tumor growth by Ad5CMV-HPV 16 AS in vivo
Discussion
To address the feasibility of HPV 16 E6/E7 antisense gene therapy for established tumors, the efficacy of Ad5CMV-HPV 16 AS in inhibiting tumor growth was evaluated in a tumorbearing nude mouse model using SiHa, MS751, and C33A cells. Tumors were allowed to grow for 25 days to a diameter of 5 to 6 mm. Mice (10 per group) received six intratumoral injections of PBS only, DL312, or Ad5CMV-HPV 16 AS on
In this study, we generated a recombinant adenovirus that expresses the antisense HPV 16 E6 and E7 genes. The virus, Ad5CMV-HPV 16 AS, contains the CMV promoter and the complete sequence of these two adjacent genes in the antisense orientation. RT-PCR analysis showed that Ad5CMV-HPV 16 AS expressed high levels of antisense HPV 16 E6/E7 RNA in the infected SiHa and CaSki cells. The time course of
Fig. 5. Flow cytometry analysis. Shown are the studies of the cell cycle of SiHa infected with DL312 and with Ad5CMV-HPV 16 AS at 10 MOI for 72 h. The data are presented as histograms in which number of cells was plotted against DNA content. Ap, apoptotic cells.
828
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
expression of HPV 16 E6/E7 antisense RNA showed that this expression continued in the Ad5CMV-HPV 16 AS-infected SiHa cells for 14 days. This result was similar to those of previous studies of p53 protein expression after transfection of Ad5CMV-p53 into non-small cell lung cancer cell lines [29] and cervical cancer cell lines [27]. The in vitro growth of SiHa and CaSki cells (HPV 16positive cervical cancer cell lines) that we transduced with the HPV 16 E6/E7 antisense RNA was significantly less than that of mock-infected and DL312-infected cells, suggesting that these results were not mediated by the infection process or virus itself. The effects of Ad5CMV-HPV 16 AS were specific to cells harboring HPV 16; there was no effect on the cervical cancer cell lines C33A, which has no HPV DNA, and MS751, which contains HPV 18. Overexpression of HPV 16 E6/E7 antisense RNA in cervical cancer cells completely suppressed the tumorigenicity of the HPV 16-positive SiHa cell line but not the HPV 16-negative MS751 or C33A cell lines in nude mice. This is the first report of the effect of exogenous antisense RNA on the tumorigenicity of cervical cancer cell lines. We also demonstrated that overexpression of antisense RNA was enough to overcome E6 and E7 protein expression in the HPV 16-positive cervical cancer SiHa cell line. Moreover, overexpression of HPV 16 E6/E7 antisense RNA efficiently reduced growth of established SiHa cell tumors in nude mouse models. However, overexpression of HPV 16 E6/E7 antisense RNA did not eradicated any tumors, because it did not reverse the expression of p53 and Rb enough to eradicate tumors. Therefore, additional treatment of adenovirus-Rb or adenovirusp53 in combination with Ad5CMV-HPV 16 AS might be required to eradicate tumors. Our data show that E6 and E7 protein expression in SiHa cells was strongly inhibited 3 days after the infection of Ad5CMV-HPV 16 AS. Similarly, the introduction of a plasmid expressing HPV 18 E6 and E7 antisense RNA has reduced E6 and E7 mRNA levels in HPV 18-positive HeLa cells [30]. Previously, E7 protein levels in HPV 16-positive SiHa and CaSki cells had been decreased using a plasmidmediated antisense RNA approach [11], antisense oligonucleotide approach [31], and siRNA approach [32]. E6 protein expression in a rabbit reticulocyte translation system had been inhibited by antisense oligonucleotides [31]. However, there had been no reports of significant inhibition of E6 protein expression in HPV-positive cell lines using an antisense RNA or oligonucleotide approach. We have clearly demonstrated the decrease of E6 protein expression by Ad5CMV-HPV 16 AS in the present study. In the previous report, antisense oligonucleotide levels decreased in cells after 4 h of incubation [17]. Furthermore, the suppression of E7 protein expression decreased after 24 h of incubation [18]. In this study, Ad5CMV-HPV 16 AS strongly suppressed E6 and E7 proteins for at least 3 days, and antisense E6/E7 RNA expression continued for 14 days. Thus, adenovirus-mediated transfer of antisense RNA transcripts into the cervical cancer cells showed longer expression of antisense RNA transcripts and stronger suppression of E6 and E7 proteins expression than did an antisense oligonucleotide approach.
E7 translation may be inhibited by antisense E6 RNA in addition to antisense E7 RNA. Antisense E6 oligonucleotides have been shown to decrease E7 protein expression [18]. In cervical cancers and cervical cell lines containing HPV, mRNAs encoding E6 and E7 proteins are transcribed from the same promoter in the form of a bicistronic transcript [33]. The bicistronic transcript is spliced to produce two shorter transcripts [34]. It has been proposed that it is a function of these spliced transcripts to stimulate E7 translation [33]. Because all three transcripts have the same 5′ oligonucleotide sequence, one would assume that antisense E6 oligonucleotides could affect all three. Our adenoviral vector, Ad5CMV-HPV 16 AS, includes both antisense E6 RNA and antisense E7 RNA, and our results show that the decrease of E7 protein expression in the Ad5CMV-HPV 16 AS-infected cells was more distinct than the decrease of E6 protein expression. Another antisense approach is use of the hammerhead ribozyme, which involves both catalytic and antisense effects. The hammerhead ribozyme affects the phenotype of HeLa cells and increases by a small amount the intracellular concentration of p53 protein [35]. In our study, Western blot and immunohistochemical analyses showed a marked increase of p53 protein in Ad5CMVHPV 16 AS-infected SiHa cells compared with control vector-infected cells. This marked increase of p53 protein may be due to a strong suppression of E6 protein by overexpression of antisense E6 RNA. The mechanism by which p53 protein inhibits growth may be related to the arrest of the cell cycle at G1 [36], apoptosis [27], or induction of another tumor suppressor gene, such as WAF1/ CIP1 [37]. The induction of apoptosis is one of several documented functions of p53 protein. Apoptosis is a selective process of physiological cell deletion. It seems that chromatin cleavage is the most characteristic biochemical feature of the process. To determine the role of apoptosis in the growth inhibition we observed, we used the TUNEL method and flow cytometry analysis. Our results showed apoptotic nuclear DNA fragmentation in the Ad5CMV-HPV 16 AS-infected SiHa cells in vitro. The TUNEL staining in the Ad5CMV-HPV 16 AS-injected SiHa cell tumors also showed apoptotic nuclear DNA fragmentation. These results strongly suggest that one of the mechanisms of growth suppression and cell death induced by Ad5CMV-HPV 16 AS in cervical cancer SiHa cells is apoptosis mediated by the p53 protein. In this study, the protein expression of Rb significantly increased after the infection with Ad5CMV-HPV 16 AS. This increase of Rb was accompanied by the decrease of E7. It is reported that antisense E7 oligonucleotides increase Rb protein expression as well as decrease E7 protein expression [18]. E7 protein is known to bind the product of the retinoblastoma susceptibility gene, Rb, thereby preventing Rb from regulating cell growth [38]. Inhibition of E7 protein expression allows Rb to perform its function of regulating cell growth [31]. E7 can directly interfere with cell cycle control by inducing S-phase entry by sequential activation of cyclin A and cyclin E gene
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
expression [39]. Both E6 and E7 play important roles in affecting specific though distinct checkpoint controls in G1 [40,41]. It has been suggested that the transforming function of E6 is independent of a direct association with p53, because mutants of E6 that fail to bind to p53 or promote p53 degradation are still able to cooperate with ras in an immortalization assay [40]. Therefore, in addition to restoration of p53 and Rb function, a p53- and Rb-independent pathway might contribute to the reduction of proliferation of cervical cancer cells by Ad5CMV-HPV 16 AS. The results of this study are noteworthy in several ways. First, a microscopic tumor model, in which antisense HPV 16 E6 and E7 oligonucleotides are injected into the tumor 10 days after inoculation with cervical cancer cells, had previously been developed for HPV 16-positive cervical cancer [18]. However, that tumor model is too small for evaluation and prediction of the effectiveness of antisense E6 and E7 in humans. In this study, a larger established tumor model involving tumors of 5 to 6 mm in diameter has been developed. Second, there are no previous reports of antisense plasmid vectors for in vivo experiments, because transduction efficiency of plasmid vectors is too low to allow assessment of tumor growth inhibition. Our use of an adenoviral vector resulted in efficient transduction and tumor reduction. Third, there are no previous reports of induction of apoptosis for antisense E6 treatment. This may be due to the distinct suppression of expression of E6 protein by the overexpression of antisense E6 RNA after the adenovirus-mediated transfer. Furthermore, overexpression of p53 by adenovirus-p53 in combination with Ad5CMV-HPV 16 AS might induce the potent induction of apoptosis and tumor reduction compared with Ad5CMV-HPV 16 AS only. Collectively, our new model of gene therapy for cervical cancer on the basis of Ad5CMV-HPV 16 AS is promising for future clinical trials. Acknowledgments We would like to sincerely thank Dr. M. Higuchi and Dr. T. Kamitani for helpful discussion. We would like to sincerely thank Mr. Keizo Oka for their help in preparing medium. This study was supported by a grant from the Ministry of Education, Science, Sports and Culture, Japan, and a grant from the Ministry of Health and Welfare for a Comprehensive Strategy for Controlling Intractable Cancer, Japan. This study was also supported by the Integrated Center for Science, Ehime University. References [1] Brinton LA. The epidemiology of human papillomavirus and cervical cancer. In: Munoz N, Bosch F, Shah KY, Meheus A, editors. IARC Sci Publ: Lyon, vol. 119; 1992. p. 3–23. [2] Mitchell MF, Hittelman WK, Lotan R, Nishioka K, Tortolero-Luna G, Richards-Kortum R, et al. Chemoprevention trials and surrogate end point biomarkers in the cervix. Cancer 1995;76:1956–77. [3] Durst M, Gissmann L, Ikenberg H, zur Hausen H. A papillomavirus DNA from a cervical carcinoma and its prevalence in cancer biopsy samples from different geographic regions. Proc Natl Acad Sci U S A 1983;80:3812–5.
829
[4] Steele C, Shillitoe EJ. Viruses and oral cancer. Crit Rev Oral Biol Med 1991;2:153–75. [5] Scheffner M, Huibregtse MM, Vierstra RD, Howley PM. The HPV-16 E6 and E6-AP complex functions as a ubiquitin–protein ligase in the ubiquitination of p53. Cell 1993;75:495–505. [6] Fujita M, Inoue M, Tanizawa O, Iwamoto S, Enomoto T. Alterations of the p53 gene in human primary cervical carcinoma with and without human papilloma infection. Cancer Res 1992;52:5323–8. [7] Schwarz E, Freese UK, Gissmann L, Mayer W, Roggenbuck B, Stremlau A, et al. Structure and transcription of human papillomavirus sequences in cervical carcinoma cells. Nature 1985;314:111–4. [8] Kaur P, McDougall JK, Cone R. Immortalization of primary human epithelial cells by cloned cervical carcinoma DNA containing human papillomavirus type 16 E6/E7 open reading frames. J Gen Virol 1989;70: 1261–6. [9] Howley PM, Scheffner M, Münger K. Oncoproteins encoded by the cancer-associated human papillomavirus target the products of the retinoblastoma and p53 tumor suppressor genes. Cold Spring Harbor Symp Quant Biol 1991;56:149–55. [10] Crook T, Morgenstern JP, Crawford L, Banks L. Continued expression of HPV 16 E7 protein is required for maintenance of the transformed phenotype of cells co-transformed by HPV-16 plus EJ-ras. EMBO J 1989;8:513–9. [11] Watanabe S, Kanda T, Yoshiike K. Growth dependence of human papillomavirus 16 DNA-positive cervical cancer cell lines and human papillomavirus 16-transformed human and rat cells on the viral oncoproteins. Jpn J Cancer Res 1993;84:1043–9. [12] Inouye M. Antisense RNA: its functions and applications in gene regulation—A review. Gene 1988;72:25–34. [13] Crooke ST. Therapeutic applications of oligonucleotides. Biotechnology 1992;10:882–6. [14] Stein CA, Cheng YC. Antisense oligonucleotides as therapeutic agents— Is the bullet really magical? Science 1993;261:1004–12. [15] von Knebel Doeberitz M, Oltersford T, Schwarz E, Gissmann L. Correlation of modified human papillomavirus early gene expression with altered growth properties in C4-I cervical carcinoma cells. Cancer Res 1988;48:3780–6. [16] Steele C, Sacks PG, Adler-Storthz K, Shillitoe EJ. Effect on cancer cells of plasmids that express antisense RNA of human papillomavirus type 18. Cancer Res 1992;52:4706–11. [17] Steele C, Cowsert LM, Shillitoe EJ. Effects of human papillomavirus type 18-specific antisense oligonucleotides on the transformed phenotype of human carcinoma cell lines. Cancer Res 1993;53:2330–7. [18] Tan TMC, Ting RCY. In vitro and in vivo inhibition of human papillomavirus type 16 E6 and E7 genes. Cancer Res 1995;55:4599–605. [19] Kucharczuk JC, Raper S, Elshami AA, Amin KM, Sterman DH, Wheeldon EB, et al. Safety of intrapleurally administered recombinant adenovirus carrying herpes simplex thymidine kinase DNA followed by ganciclovir therapy in nonhuman primates. Hum Gene Ther 1996;7: 2225–33. [20] Mitchell MF, Hamada K, Sastry J, Sarkar A, Tortolero-Luna G, Wharton JT, et al. Transgene expression in the rhesus cervix mediated by an adenovirus expressing beta-galactosidase. Am J Obstet Gynecol 1996;174: 1094–101. [21] Lu CY, Giordano FJ, Rogers KC, Rothman A. Adenovirus-mediated increase of exogenous and inhibition of endogenous fosB gene expression in cultured pulmonary arterial smooth muscle cells. J Mol Cell Cardiol 1996;28:1703–13. [22] Yamanishi Y, Maeda H, Hiyama K, Ishioka S, Yamakido M. Specific growth inhibition of small-cell lung cancer cells by adenovirus vector expressing antisense c-kit transcripts. Jpn J Cancer Res 1996;87:534–42. [23] Alemany R, Ruan SB, Kataoka M, Koch PE, Mukhopadhyay T, Cristiano RJ, et al. Growth inhibitory effect of anti-K-ras adenovirus on lung cancer cells. Cancer Gene Ther 1996;3:296–301. [24] Lee CT, Wu S, Gabrilovich D, Chen H, Nadaf-Rahrov S, Ciernik IF, et al. Antitumor effects of an adenovirus expressing antisense insulin-like growth factor I receptor on human lung cancer cell lines. Cancer Res 1996;56:3038–41.
830
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
[25] Yee C, Krishnan-Hewlett I, Baker CC, Schlegel R, Howley PM. Presence and expression of human papillomavirus sequences in human cervical carcinoma cell lines. Am J Pathol 1985;119:361–6. [26] Hamada K, Zhang WW, Alemany R, Roth JA, Wolf J, Mitchell MF. Adenovirus-mediated transfer of HPV 16 E6/E7 antisense RNA to human cervical cancer cells. Gynecol Oncol 1996;60:373–9. [27] Hamada K, Alemany R, Zhang WW, Hittelman WN, Lotan R, Roth JA, et al. Adenovirus-mediated transfer of a wild-type p53 gene and induction of apoptosis in cervical cancer. Cancer Res 1996;56: 3047–54. [28] Andreeff M, Darzynkiewicz Z, Sharpless TK, Clarkson BD, Melamed MR. Discrimination of human leukemia subtypes by flow cytometric analysis of cellular DNA and RNA. Blood 1980;55:282–93. [29] Zhang WW, Fang X, Mazur W, French BA, Georges RN, Roth JA. Highefficiency gene transfer and high-level expression of wild-type p53 in human lung cancer cells mediated by recombinant adenovirus. Cancer Gene Ther 1994;1:4–15. [30] Hu G, Liu W, Hanania EG, Fu S, Wang T, Deisseroth AB. Suppression of tumorigenesis by transcription units expressing the antisense E6 and E7 messenger RNA (mRNA) for the transforming proteins of the human papilloma virus and the sense mRNA for the retinoblastoma gene in cervical carcinoma cells. Cancer Gene Ther 1995;2:19–32. [31] Tan TMC, Gloss B, Bernard H-U, Ting RCY. Mechanism of translation of the bicistronic mRNA encoding human papillomavirus type 16 E6-E7 genes. J Gen Virol 1994;7:2663–70. [32] Niu X-Y, Peng Z-L, Duan WQ, Wang H, Wang P. Inhibition of HPV16 E6 oncogene expression by RNA interference in vitro and in vivo. Int J Gynecol Cancer 2006;16:743–51.
[33] Schneider-Gädicke A, Schwarz E. Different human cervical carcinoma cell lines show similar transcription patterns of human papillomavirus type 18 early genes. EMBO J 1986;5:2285–92. [34] Shirasawa H, Tanzawa H, Matsunaga T, Shimizu B. Quantitative detection of spliced E6-E7 transcripts of human papillomavirus type 16 in cervical premalignant lesions. Virology 1991;184:795–8. [35] Chen Z, Kamath P, Zhang S, Weil MM, Shillitoe EJ. Effectiveness of three ribozymes for cleavage of an RNA transcript from human papillomavirus type 18. Cancer Gene Ther 1995;2:263–71. [36] Diller L, Kassel J, Nelson CE, Gryka MA, Litwak G, Gebhardt M, et al. p53 functions as a cell cycle control protein in osteosarcomas. Mol Cell Biol 1990;10:5772–81. [37] El-Deiry WS, Harper JW, O'Connor PM, Velculescu VE, Canman CE, Jackman J, et al. WAF1/CIP1 is induced in p53-mediated G1 arrest and apoptosis. Cancer Res 1994;54:1169–74. [38] Dyson N, Howley P, Münger K, Harlow E. The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science 1989;243:934–7. [39] Zerfass K, Schulze A, Spitkovsky D, Friedman V, Henglein B, JansenDürr P. Sequential activation of cyclin E and cyclin A gene expression by human papillomavirus type 16 E7 through sequences necessary for transformation. J Virol 1995;69:6389–99. [40] Storey A, Massimi P, Dawson K, Banks L. Conditional Immortalization of primary cells by human papillomavirus type 18 E6 and EJ-ras defines an E6 activity in G0/G1 phase which can be substituted for mutations in p53. Oncogene 1995;11:653–61. [41] Chen JJ, Reid CE, Band V, Androphy EJ. Interaction of papillomavirus E6 oncoproteins with a putative calcium-binding protein. Science 1995; 269:529–31.
Adenovirus-mediated transfer of human papillomavirus 16 E6/E7 antisense RNA and induction of apoptosis in cervical cancer Katsuyuki Hamada a,⁎, Toshiro Shirakawa b , Akinobu Gotoh c , Jack A. Roth d , Michele Follen e a
c d
Department of Obstetrics and Gynecology, School of Medicine, Ehime University, Shitsukawa,Toon, Ehime 791-0295, Japan b Division of Infectious Disease Control, International Center for Medical Research and Treatment, Faculty of Medicine, Kobe University Graduate School of Medicine, Kusunoki, Chuo-ku, Kobe, Hyogo 650-0017, Japan Laboratory of Cell and Gene Therapy, Institute of Advanced Medical Sciences, Hyogo College of Medicine, Mukogawa, Nishinomiya, Hyogo 663-8501, Japan Department of Thoracic and Cardiovascular Surgery, The University of Texas, M.D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030, USA e Department of Gynecologic Oncology, The University of Texas, M.D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030, USA Received 22 March 2006 Available online 5 September 2006
Abstract Objective. In most cervical cancers, human papillomaviruses (HPVs) are identified. The E6 and E7 genes of HPVs encode proteins, that interfere with the function of the tumor suppressor proteins p53 and Rb. We are exploring the potential use of antisense HPV RNA transcripts for gene therapy for HPV-positive cervical cancers. Methods. Via a recombinant adenoviral vector, Ad5CMV-HPV 16 AS, we introduced the antisense RNA transcripts of the E6 and E7 genes of HPV type 16 into human cervical cancer SiHa cells harboring HPV 16. We then analyzed the effects of expression of these genes on cell and tumor growth. Results. HPV 16 E6/E7 antisense RNA was detected for 14 days in Ad5CMV-HPV 16 AS-infected cells. After infection, E6 and E7 protein expression was suppressed, and p53 and Rb protein expression increased. The Ad5CMV-HPV 16 AS-infected cells underwent apoptosis in vitro and in vivo. Cell growth and tumorigenicity were greatly suppressed. Ad5CMV-HPV 16 AS treatment significantly reduced the volumes of established subcutaneous tumors. Conclusion. Transfection of cervical cancer cells with HPV 16 E6/E7 antisense RNA in a form such as Ad5CMV-HPV 16 AS might be a potentially useful approach to the therapy of HPV 16-positive cervical cancer. Published by Elsevier Inc. Keywords: Recombinant adenovirus; HPV 16; Antisense; Cervical cancer; Apoptosis; E6; E7
Introduction Cervical cancer, the second most common malignancy in women worldwide, remains an important health problem for women [1]. Incidence rates are currently increasing in American women [2]. Human papillomavirus (HPV) infection has been identified as an etiologic agent in cervical cancer [1]. Only particular HPVs, such as HPV types 16 and 18, are associated with high risk of malignant conversion [3]. The viral DNA is randomly integrated into host cell genomic ⁎ Corresponding author. Fax: +81 899 60 5381. E-mail address: [email protected] (K. Hamada). 0090-8258/$ - see front matter. Published by Elsevier Inc. doi:10.1016/j.ygyno.2006.06.035
sequences, and HPV-specific RNA and proteins can be detected in the tumors and cell lines derived from these host cells [4]. The HPV genome contains seven early genes, two of which encode the oncoproteins E6 and E7. E6 can form complexes with the tumor suppressor p53 protein and promote p53 degradation through a ubiquitin-dependent protease system [5]. Inactivation of the p53 gene by allelic loss or by point mutation is infrequent in primary cervical cancer [6]. E7 binds to the tumor suppressor Rb protein, preventing Rb from binding to its normal substrate, the E2F transcription factor [7]. The expression of the viral E6 and E7 proteins in HPV-infected cervical cancer cells interferes with
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
p53 and Rb function despite expression of wild-type p53 and Rb alleles. E6 and E7 act independently or synergistically to potentiate the immortalization of their natural host cells, primary human squamous epithelial cells [8]. It has been demonstrated that the inactivation of the Rb and p53 proteins is an important step in cervical carcinogenesis [9]. Furthermore, the continued expression of the E6 and E7 regions of the viral genome appears to be necessary for the maintenance of the proliferative and malignant phenotypes of cervical cancer cells [10,11]. Gene expression can be selectively inhibited by introducing complementary antisense messages [12]. The efficacy of antisense inhibition has been demonstrated for a variety of genes, particularly oncogenes [12–14]. These intriguing findings lead to the possibility of using antisense messages to develop gene therapy techniques for human cervical cancers that contain HPV DNA. A plasmid expressing antisense HPV 18 RNA was shown to decrease the growth rate of the HPV 18-positive human cervical cancer cell line HeLa [15,16], but the low transduction efficiency of the plasmid would be a problem for human clinical therapeutic trials. Antisense oligonucleotides to E6 and E7 of HPV 16 and HPV 18 were shown to selectively inhibit the growth of HPV 16- and 18-positive cancer cells, respectively, but not HPV-negative cells [17,18]. However, when the oligonucleotides were withdrawn, the remaining cells recovered and grew as before; any therapy with oligonucleotides might therefore require continuous administration. Adenoviral vectors have many potential advantages over other viral vector systems and other techniques for introducing DNA into eukaryotic cells. Adenoviruses can be produced at very high titers, have high infectivity in eukaryotic cells, and have had a good clinical safety record in previous human and primate studies [19,20]. Several adenoviral vectors containing antisense expression units have shown growth inhibitory effects in vitro but not in vivo [21– 24]. There have been no reports of the efficacy of an adenoviral construct expressing antisense HPV RNA. In the study presented here, the E6/E7 region of HPV 16 in the antisense orientation was inserted into an adenoviral vector. We chose HPV 16, because it is identified in 60% of cervical cancers [3]. We evaluated the effects of the antisense HPV 16 E6/E7 construct (Ad5CMV-HPV 16 AS) on growth of human cervical cancer cells in vitro, ex vivo, and in vivo in a nude mouse model. We also evaluated whether the mechanism of the growth-suppressive effect that we observed is related to induction of apoptosis. Materials and methods Cell lines and culture conditions The human cervical cancer cell lines SiHa, CaSki, MS751, and C33A were obtained from the American Type Culture Collection. SiHa and CaSki are HPV 16-positive cell lines and have wild-type p53, MS751 is an HPV 18-positive cell line and has wild-type p53, and C33A is an HPV-negative cell line and has mutant p53 [25]. Cells were grown in RPMI medium supplemented with 10% heat-inactivated fetal bovine serum.
821
Generation of recombinant adenovirus The HPV 16 genome, cloned into pBR322, was provided by the Japanese Collection of Research Bioresources (JCRB, Tokyo). The entire E6/E7 region (nt 45 to 877) was amplified by PCR using an upstream primer with a NotI site and a downstream primer with a HindIII site. This fragment was cloned in antisense orientation into pAdE1CMV, a shuttle plasmid that contains the left end of Ad5 (0–16 map units [m.u.]) with the E1 region (1.25–9.2 m.u.) substituted by an expression cassette containing the cytomegalovirus (CMV) promoter, a multicloning site, and a simian virus (SV40) polyadenylation signal. The recombinant plasmid with the E6/E7 region in antisense orientation, pAdE1CMV-HPV 16 AS, was cotransfected with pJM17, a recombinant plasmid into 293 cells to generate the recombinant adenovirus, Ad5CMV-HPV 16 AS. An adenoviral vector, DL312, was used as the control.
Adenoviral infection Recombinant adenovirus was purified by double cesium-gradient ultracentrifugation as described elsewhere [26]. Viral stocks were propagated in 293 cells. Cells were harvested 36–40 h after infection, and cell debris was removed by subjecting the lysed cells to CsCl-gradient centrifugation. Concentrated virus was dialyzed and stored in aliquots at −80°C. Infection was carried out by adding the virus to high-glucose Dulbecco's modified Eagle's medium and to the cell monolayers. The viral titers were determined by plaque assays [26]. CAR expression was assessed by scoring X-gal-positive blue cells in each of three replicate dishes after the infection with Ad5CMV-LacZ. All cell lines inoculated with a single dose of Ad5CMV-LacZ at 10 MOI (multiplicity of infection) or greater exhibited 100% blue cells.
PCR analysis DNA was extracted using phenol–chloroform and precipitated with ethanol and sodium acetate. Total RNA was prepared by lysing cell monolayers in guanidinium isothiocyanate and centrifuging over a 5.7 M CsCl solution. RTPCR was performed from 6 μg of total RNA to generate cDNA. These DNA samples and cDNA products were then amplified by PCR, using primers for the antisense E6/E7 of HPV 16 and the polyadenylation signal of the adenovirus, yielding a 486-bp product. PCR products were analyzed on 2% agarose gels and stained with ethidium bromide.
Cell count assay Cells were plated at a density of 5 × 104/well in 12-well plates in triplicate. RPMI supplemented with 10% heat-inactivated fetal bovine serum was used as the growth medium. Cells were infected with either Ad5CMV-HPV 16 AS or DL312 control. Culture medium was used for mock infection. Cells were harvested using trypsin and counted. Cell viability was determined by trypan blue exclusion after harvest.
[3H]Thymidine incorporation assay Cells were cultured at 1 × 103 cells/well in 96-well flat-bottomed plates. Cells were infected with Ad5CMV-HPV 16 AS or DL312 and cultured in medium for 5 days. Each well was pulse-treated with 1 μCi [3H]thymidine (sp. act., 6.7 Ci/ mmol; Amersham Corp., Arlington Heights, IL) for an additional 24 h, after which the cells were harvested with a PhD cell harvester (Cambridge Technology Inc., Cambridge, MA). Individual filter discs were then processed for liquid scintillation counting. The data are presented as the means of quintuplicate samples.
Western blot analysis Total cell lysates were prepared by lysing cell monolayers in plates with sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer after rinsing the cells with phosphate-buffered saline (PBS). Each lane was loaded with 10 μg of cell lysate protein as determined by BCA
822
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
protein assay (Pierce, Rockford, IL). After electrophoresis at 20 mA for 2 h, the proteins in the gels were transferred to Hybond-ECL membranes (Amersham). The membranes were blocked with 1% dry milk and 0.1% Tween 20 (Sigma Chemical Co., St. Louis, MO) in PBS and probed with a primary antibody, mouse anti-human p53 monoclonal antibody DO7 (DAKO Co., Carpinteria, CA) or rabbit anti-human Rb polyclonal antibody C-15 (Santa Cruz Biotechnology Inc., Santa Cruz, CA), and secondary antibodies, a biotinylated goat anti-mouse and anti-rabbit antibodies mix (DAKO). Then the membranes were incubated with horseradish peroxidase-conjugated streptavidin (DAKO). The membranes were developed according to the Amersham ECL protocol.
Immunoprecipitation analysis Cells were metabolically labeled for 12 h with 1 mCi of 35S Translabel (ICN Radiochemicals, Irvine, CA) per 10-cm dish in methionine and cysteine-free RPMI medium containing 10% dialyzed fetal calf serum. Cells were lysed at 4°C in 1 ml of radioimmunoprecipitation assay (RIPA) buffer [150 mM NaCl, 50 mM Tris–HCl (pH 8.0), 1% NP-40]. Insoluble debris was pelleted at 10,000×g for 10 min, and the supernatant was incubated with mouse monoclonal anti-HPV 18/16 E6 antibody (Oncogene Sciences, Uniondale, NY) or mouse monoclonal anti-HPV 16 E7 antibody (Triton Diagnostics, Alameda, CA), and protein G–Sepharose beads (Sigma). After they were washed with RIPA buffer, the beads were boiled in SDS sample buffer and loaded onto SDS-polyacrylamide gels.
Immunohistochemical analysis The Ad5CMV-HPV 16 AS-infected SiHa cells were analyzed to detect p53 protein expression. SiHa cells growing on coverslips were air dried and then fixed with 50% ethanol and 50% acetone at − 20°C for 20 min. Cells were treated with 3% H2O2 in methanol. Immunohistochemical staining was performed with the Vectastain Elite kit (Vector Laboratories Inc., Burlingame, CA). The primary antibody was a mouse anti-human p53 monoclonal antibody DO7 (DAKO), and the secondary antibody was a biotinylated goat anti-mouse IgG (Vector Labs). An avidin and biotinylated horseradish peroxidase macromolecular complex reagent (Vector Labs) was used to detect the antigen–antibody complex. The cells were stained with diaminobenzidine and then counterstained with Harris hematoxylin (Sigma).
Tumorigenicity assay SiHa, MS751, and C33A cells were infected with Ad5CMV-HPV 16 AS or DL312 at 10 MOI. An equal number of cells were treated with medium as a mock infection. Three hours after infection, the treated cells were harvested and rinsed twice with PBS. For each treatment, 5 × 106 cells in a volume of 100 μl were injected subcutaneously into a nude (nu/nu) female mouse (aged 4– 5 weeks; CLEA Japan). Experiments were reviewed and approved by institutional committees for animal care and use and for recombinant DNA research in School of Medicine, Ehime University. Seven mice were used for each treatment. The mice were examined every day and tumor formation and size were evaluated for 40 days. The tumors were measured every other day with calipers in two perpendicular diameters without knowledge of treatment groups. Tumor volume was calculated by assuming a spherical shape with the average tumor diameter calculated as the square root of the product of cross-sectional diameters.
Inhibition of tumor growth in vivo To determine inhibition of tumor growth in vivo, Ad5CMV-HPV 16 AS was injected into an established tumor in each female nude (nu/nu) mouse (CLEA). Experiments were reviewed and approved by institutional committees for animal care and use and for recombinant DNA research in School of Medicine, Ehime University. Five million SiHa, MS751, or C33A cells in 100 μl of PBS were injected into the right posterior flank of each mouse through an insulin syringe with a 28 1/2-gauge needle. Ten animals were used for each group. After 20– 25 days, tumors with a diameter of 5 to 6 mm had grown in all animals. Either 100 μl of Ad5CMV-HPV 16 AS or DL312 at 2 × 109 plaque-forming units (PFU) or PBS only was injected intratumorally on days 0, 1, 2, 3, 4, and 5. The tumors were measured every other day and tumor volume was calculated as described above.
Statistical analysis Values are expressed as the mean ± SD, and analyzed with the unpaired t-test. Statistical significance was set at P < 0.05.
Results
TUNEL assay The Ad5CMV-HPV 16 AS-infected cells and injected tissues were analyzed to detect apoptotic cells by the previously described TUNEL technique [27]. The slides were treated with 3% H2O2 in methanol. The nuclei of the cells were stripped from the proteins by incubation with 0.002% protease (Sigma). Next, the slides were covered with TdT (0.1 U/μl; USB, Cleveland, OH) and biotinylated dUTP (0.4 nM; Boehringer Mannheim Biochem., Indianapolis, IN) in TdT buffer (30 mM Tris, pH 7.2; 140 mM sodium cacodylate; 1 mM cobalt chloride) in a humid atmosphere of 37°C for 120 min. The reaction was terminated by transferring the slides to termination buffer (300 mM sodium chloride, 30 mM sodium citrate). Next, the slides were incubated with an avidin and biotinylated horseradish peroxidase macromolecular complex (Vector Laboratories Inc., Burlingame, CA) and stained with diaminobenzidine. Finally, the slides were counterstained with Harris hematoxylin (Sigma).
Flow cytometry analysis To measure numbers of apoptotic cells, ethanol-fixed cells on the third day after infection with DL312 or Ad5CMV-HPV 16 AS were incubated with propidium iodide (50 μg/ml) and ribonuclease (20 μg/ml) for 20 min at 37°C [28]. All measures were made with an EPICS profile (Coulter Corporation, Hialeah, FL) equipped with an air-cooled argon ion laser emitting 488 nm at 15 mW. A minimum of 10,000 events per sample were analyzed and FITC fluorescence was collected using a 525 BP filter. Coulter's cytologic program was used for data analysis. Mean peak fluorescence was determined for each histogram.
Generation of the recombinant adenovirus Ad5CMV-HPV 16 AS The entire E6/E7 region was amplified and cloned in antisense orientation into a shuttle plasmid vector to form the plasmid pAdE1CMV-HPV 16 AS. Ad5CMV-HPV 16 AS was generated after cotransfecting pAdE1CMV-HPV 16 AS and pJM17 into 293 cells. The expected genomic structure of Ad5CMV-HPV 16 AS is depicted in Fig. 1A. A 0.5-kb HPV 16 E6/E7 antisense-specific band amplified through the PCR confirmed the identity of Ad5CMV-HPV 16 AS (Figs. 1A and B). SiHa and CaSki cells (HPV 16-positive cervical cancer cell lines) were infected with either Ad5CMV-HPV 16 AS or a control adenoviral vector, DL312. Culture medium was used as the mock infection control. RNA was extracted 72 h after infection. In RT-PCR analysis, the transcript of antisense E6/E7 RNA from Ad5CMV-HPV 16 AS was detected in Ad5CMVHPV 16 AS-infected cells, while DL312-infected cells did not show any adenoviral antisense transcripts (Fig. 1C). Mockinfected cells also did not show any adenoviral antisense transcripts (data not shown). The time course of expression of antisense E6/E7 RNA from Ad5CMV-HPV 16 AS-infected CaSki cells is shown in Fig. 1D. Antisense E6/E7 RNA was
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
823
Fig. 1. (A) A map of Ad5CMV-HPV 16 AS genomic DNA, showing locations of the CMV promoter, HPV 16 E6/E7 antisense, the SV40 polyadenylation signal (pA), and the PCR primers. Primers define 486 bp of HPV 16 E6/E7 antisense. (B) The newly generated HPV 16 E6/E7 antisense recombinant adenovirus (Ad5CMV-HPV 16 AS) was identified by PCR analysis of the DNA samples. DNA templates used in each reaction were pAdE1CMV-HPV 16 AS, DL312, no DNA, a sample from the supernatant of Ad5CMV-HPV 16 AS-infected 293 cells after cytopathic effect, and the purified adenovirus (Ad5CMV-HPV 16 AS). (C) RT-PCR analysis for expression of the HPV 16 E6/E7 antisense RNA. DNA templates shown are from DL312-infected SiHa and CaSki cells with RT and without RT and from Ad5CMVHPV 16 AS-infected SiHa and CaSki cells with RT and without RT. RNA from each sample was extracted 72 h after incubation with adenoviruses at 10 MOI. (D) RTPCR analysis for expression of the HPV 16 E6/E7 antisense RNA. RNA was extracted from HPV 16 E6/E7 antisense-infected CaSki cells at 10 MOI before infection and on days 1, 3, 5, 7, 14, and 21.
expressed in the Ad5CMV-HPV 16 AS-infected CaSki cells for 14 days but not on day 21. Growth inhibition of cervical cancer cells by Ad5CMV-HPV 16 AS in vitro To examine whether the Ad5CMV-HPV 16 AS virus could inhibit growth of human cervical cancer cells, the cell lines SiHa, CaSki, MS751 (HPV 18-positive cervical cancer cell line), and C33A cells (HPV-negative cervical cancer cell line) were treated with Ad5CMV-HPV 16 AS, DL312, or mock
infection. The growth of the Ad5CMV-HPV 16 AS-infected SiHa and CaSki cells was greatly suppressed when estimated by [3H]thymidine incorporation and cell count assay 6 days after infection (Figs. 2A and B). In contrast, the growth of the Ad5CMV-HPV 16 AS-infected MS751 and C33A cells did not differ from that of mock- and DL312-infected cells (Figs. 2A and B). The time course of growth of SiHa, MS751, and C33A cells is shown in Fig. 2C. Growth of Ad5CMV-HPV 16 ASinfected MS751 and C33A cells was not significantly different from that of the mock- and DL312-treated cells. Growth of Ad5CMV-HPV 16 AS-infected SiHa cells started to decrease
824 K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830 Fig. 2. (A) Inhibition of the growth of SiHa, CaSki, MS751, and C33A cells as determined by [3H]thymidine incorporation assay. The cells were inoculated at densities of 1 × 103 cells/well in each 96-well plate 24 h before infection. At each indicated MOI, cells in five wells were infected for 5 days, pulse-treated with [3H]thymidine for an additional 24 h, trypsinized, harvested, and counted by liquid scintillation counter. The mean counts per minute for quintuplicate wells on day 6 after infection were plotted. Bars indicate standard deviation. (B) Inhibition of the growth of SiHa, CaSki, MS751, and C33A cells as determined by a cell count assay. The cells were inoculated at densities of 5 × 104 cells/well in each 12-well plate 24 h before infection. At each indicated MOI, cells in three wells on each well plate were infected. Six days after infection, cells were trypsinized and counted. The mean cell counts for triplicate wells were plotted. Bars indicate standard deviation. (C) Time course of growth of SiHa, MS751, and C33A cells as determined by a cell count assay. The cells were infected with Ad5CMV-HPV 16 AS or DL312 at 10 MOI in SiHa and MS751 cells and at 5 MOI in C33A cells. Culture medium alone was used as the mock infection control. Triplicate wells for each treatment were counted at each indicated time point after infection, and the mean cell counts were plotted. Bars indicate standard deviation. (D) Effects of treatment with Ad5CMV-HPV 16 AS on tumor growth of SiHa cells in nude mice. Mice were injected subcutaneously with 5 × 106 SiHa cells/mouse. After tumors with a diameter of 5–6 mm were obtained, 100 μl of Ad5CMV-HPV 16 AS or DL312 at 2 × 109 PFU or PBS alone was injected into tumors of 5 to 6 mm in diameter. Intratumoral injections were made on days 0, 1, 2, 3, 4, and 5. Ten mice made up each treatment group. The means of tumor volumes were plotted. Bars indicate standard deviation.
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
825
Fig. 2 (continued).
significantly compared with the mock- and DL312-infected cells 54 h after infection. Effect of Ad5CMV-HPV 16 AS on E6 and E7 expression To examine whether the Ad5CMV-HPV 16 AS virus could decrease the expression of E6 and E7 proteins, immunopreci-
pitation analysis was done using the mouse anti-human HPV 18/16 E6 antibody and mouse anti-human HPV 16 E7 antibody. The amounts of E6 and E7 proteins produced were compared following 72 h of infection with Ad5CMV-HPV 16 AS or DL312 at 10 MOI or mock infection with culture medium (Fig. 3A). E6 and E7 bands were observed in cellular extracts immunoprecipitated from mock- and DL312-infected SiHa
Fig. 3. (A) Immunoprecipitation analysis. SiHa cells were infected for 72 h with Ad5CMV-HPV 16 AS or DL312 at 10 MOI or mock infected. The 35S-labeled E6 and E7 proteins were immunoprecipitated with anti-E6 and anti-E7 antibodies from the cell lysates. Immunocomplexes were electrophoresed in SDS-PAGE and autoradiographed. (B) Western blot analysis. SiHa cells were infected as in (A). Cellular extracts were subjected to SDS-PAGE. Western blots were probed with antip53 and anti-Rb antibodies.
826
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
cells. Expression of E6 and E7 proteins was lower in the Ad5CMV-HPV 16 AS-infected SiHa cells than in mock- and DL312-infected cells. Effect of Ad5CMV-HPV 16 AS on p53 and Rb expression To determine the expression of the p53 and Rb proteins in the Ad5CMV-HPV 16 AS-infected SiHa cells, Western blot analysis was done using the mouse anti-human p53 antibody and the rabbit anti-human Rb antibodies. The amounts of p53 and Rb proteins produced were compared following 72 h of infection with Ad5CMV-HPV 16 AS or DL312 at 10 MOI or mock infection (Fig. 3B). The p53 and Rb proteins were highly expressed in Ad5CMV-HPV 16 AS-infected SiHa cells. Samples isolated from mock- and DL312-infected cells had very low levels of the p53 and Rb proteins. Immunohistochemical analysis was done using the mouse anti-human p53 antibody to determine the expression of the p53 protein in the Ad5CMV-HPV 16 AS-infected
SiHa cell. Immunohistochemical analysis of SiHa cells infected with Ad5CMV-HPV 16 AS at 10 MOI revealed staining of p53 protein in the nucleus 12 h after infection, whereas DL312-infected cells failed to show p53 staining (Fig. 4A). Induction of apoptosis by Ad5CMV-HPV 16 AS To characterize the effects of Ad5CMV-HPV 16 AS, the morphological changes in SiHa cells were examined after infection with Ad5CMV-HPV 16 AS or DL312 at 10 MOI. Within 36 to 72 h after infection with Ad5CMV-HPV 16 AS, an apparent morphological change occurred, with portions of the cell population rounding up and their outer membranes forming blebs. These changes are part of a series of histologically predictable events that suggest programmed cell death [27]. On the other hand, cells infected with the control adenovirus, DL312, grew normally with no histomorphological abnormalities.
Fig. 4. (A) Immunohistochemical staining of p53 in SiHa cells. Left, 12 h after DL312 infection at 10 MOI; right, 12 h after Ad5CMV-HPV 16 AS infection at 10 MOI. ×100. (B) TUNEL staining in SiHa cells. Left, 48 h after DL312 infection at 10 MOI; right, 48 h after Ad5CMV-HPV 16 AS infection at 10 MOI. ×100. (C) TUNEL staining in SiHa cell tumor. Left, 48 h after DL312 injection at 2 × 109 PFU; right, 48 h after Ad5CMV-HPV 16 AS injection at 2 × 109 PFU. ×100.
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
To determine whether the morphological changes seen in the Ad5CMV-HPV 16 AS-infected cells are related to the apoptotic process, SiHa cells were analyzed by the TUNEL method and flow cytometry. When the TUNEL method, which stains for broken DNA ends via a terminal deoxynucleotide transferase reaction, was performed, the DL312-infected SiHa cells and DL312-injected SiHa cell tumor did not stain, but nuclei in the Ad5CMV-HPV 16 AS-infected SiHa cells and the Ad5CMVHPV 16 AS-injected SiHa cell tumor stained 48 h after infection at 10 MOI (Fig. 4B) and 48 h after injection at 2 × 109 PFU (Fig. 4C). Flow cytometry analysis was performed to compare the amounts of apoptotic cells. The SiHa cells were infected with Ad5CMV-HPV 16 AS or DL312 at 10 MOI for 72 h and then fixed and stained with propidium iodide. The results of flow cytometry analysis are shown in Fig. 5. Apoptotic cells were evident in the Ad5CMV-HPV 16 AS-infected cells but not in the DL312-infected cells.
827
Table 1 Effect of pretreatment with Ad5CMV-HPV 16 AS on tumorigenicity of SiHa, MS751, and C33A cells in nude mice Cell line
Treatment
SiHa
Medium DL312 Ad5CMV- HPV 16 AS MS751 Medium DL312 Ad5CMV- HPV 16 AS C33A Medium DL312 Ad5CMV- HPV 16 AS
No. of tumors/ Mean tumor volume no. of mice (mm3 ± SD)
(%)
7/7 6/7 0/7
425 ± 130 294 ± 95 0±0
(100) (69) (0)
7/7 6/7 7/7
843 ± 276 726 ± 296 700 ± 197
(100) (86) (83)
7/7 5/7 6/7
1001 ± 417 635 ± 236 661 ± 169
(100) (63) (66)
Cells were treated with Ad5CMV-HPV 16 AS or DL312 at 10 MOI or mock infection for 3 h. Tumor sizes were determined 40 days after treated cells (5 × 106 cells/mouse) were injected.
Inhibition of tumorigenicity by Ad5CMV-HPV 16 AS To examine whether the Ad5CMV-HPV 16 AS virus could inhibit tumorigenicity of human cervical cancer cells, nude mice were injected subcutaneously with SiHa, MS751, and C33A cells to initiate tumor formation. Each mouse received one injection of 5 × 106 cells that had been infected with Ad5CMVHPV 16 AS or DL312 at 10 MOI or mock-infected with medium alone for 3 h. Tumors formed from the mock- and DL312infected SiHa cells; mice that received Ad5CMV-HPV 16 AStreated SiHa cells did not develop tumors (Table 1). Tumors formed from the mock-, DL312-, and Ad5CMV-HPV 16 ASinfected MS751 and C33A cells.
days 0, 1, 2, 3, 4, and 5. In the mice treated with PBS only or DL312, SiHa cell tumors continued to grow rapidly throughout the treatment, whereas in the mice treated with Ad5CMV-HPV 16 AS, tumor growth was greatly reduced (Fig. 2D). Six injections of Ad5CMV-HPV 16 AS reduced the size of SiHa cell tumors by 84% compared with six injections of PBS only, whereas six injections of DL312 did not reduce significantly tumor size. Six injections of Ad5CMV-HPV 16 AS did not suppress the growth of MS751 or C33A cell tumors. There was no significant difference in the body weights of any of the mice treated in this experiment. No animal died during this experiment.
Inhibition of tumor growth by Ad5CMV-HPV 16 AS in vivo
Discussion
To address the feasibility of HPV 16 E6/E7 antisense gene therapy for established tumors, the efficacy of Ad5CMV-HPV 16 AS in inhibiting tumor growth was evaluated in a tumorbearing nude mouse model using SiHa, MS751, and C33A cells. Tumors were allowed to grow for 25 days to a diameter of 5 to 6 mm. Mice (10 per group) received six intratumoral injections of PBS only, DL312, or Ad5CMV-HPV 16 AS on
In this study, we generated a recombinant adenovirus that expresses the antisense HPV 16 E6 and E7 genes. The virus, Ad5CMV-HPV 16 AS, contains the CMV promoter and the complete sequence of these two adjacent genes in the antisense orientation. RT-PCR analysis showed that Ad5CMV-HPV 16 AS expressed high levels of antisense HPV 16 E6/E7 RNA in the infected SiHa and CaSki cells. The time course of
Fig. 5. Flow cytometry analysis. Shown are the studies of the cell cycle of SiHa infected with DL312 and with Ad5CMV-HPV 16 AS at 10 MOI for 72 h. The data are presented as histograms in which number of cells was plotted against DNA content. Ap, apoptotic cells.
828
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
expression of HPV 16 E6/E7 antisense RNA showed that this expression continued in the Ad5CMV-HPV 16 AS-infected SiHa cells for 14 days. This result was similar to those of previous studies of p53 protein expression after transfection of Ad5CMV-p53 into non-small cell lung cancer cell lines [29] and cervical cancer cell lines [27]. The in vitro growth of SiHa and CaSki cells (HPV 16positive cervical cancer cell lines) that we transduced with the HPV 16 E6/E7 antisense RNA was significantly less than that of mock-infected and DL312-infected cells, suggesting that these results were not mediated by the infection process or virus itself. The effects of Ad5CMV-HPV 16 AS were specific to cells harboring HPV 16; there was no effect on the cervical cancer cell lines C33A, which has no HPV DNA, and MS751, which contains HPV 18. Overexpression of HPV 16 E6/E7 antisense RNA in cervical cancer cells completely suppressed the tumorigenicity of the HPV 16-positive SiHa cell line but not the HPV 16-negative MS751 or C33A cell lines in nude mice. This is the first report of the effect of exogenous antisense RNA on the tumorigenicity of cervical cancer cell lines. We also demonstrated that overexpression of antisense RNA was enough to overcome E6 and E7 protein expression in the HPV 16-positive cervical cancer SiHa cell line. Moreover, overexpression of HPV 16 E6/E7 antisense RNA efficiently reduced growth of established SiHa cell tumors in nude mouse models. However, overexpression of HPV 16 E6/E7 antisense RNA did not eradicated any tumors, because it did not reverse the expression of p53 and Rb enough to eradicate tumors. Therefore, additional treatment of adenovirus-Rb or adenovirusp53 in combination with Ad5CMV-HPV 16 AS might be required to eradicate tumors. Our data show that E6 and E7 protein expression in SiHa cells was strongly inhibited 3 days after the infection of Ad5CMV-HPV 16 AS. Similarly, the introduction of a plasmid expressing HPV 18 E6 and E7 antisense RNA has reduced E6 and E7 mRNA levels in HPV 18-positive HeLa cells [30]. Previously, E7 protein levels in HPV 16-positive SiHa and CaSki cells had been decreased using a plasmidmediated antisense RNA approach [11], antisense oligonucleotide approach [31], and siRNA approach [32]. E6 protein expression in a rabbit reticulocyte translation system had been inhibited by antisense oligonucleotides [31]. However, there had been no reports of significant inhibition of E6 protein expression in HPV-positive cell lines using an antisense RNA or oligonucleotide approach. We have clearly demonstrated the decrease of E6 protein expression by Ad5CMV-HPV 16 AS in the present study. In the previous report, antisense oligonucleotide levels decreased in cells after 4 h of incubation [17]. Furthermore, the suppression of E7 protein expression decreased after 24 h of incubation [18]. In this study, Ad5CMV-HPV 16 AS strongly suppressed E6 and E7 proteins for at least 3 days, and antisense E6/E7 RNA expression continued for 14 days. Thus, adenovirus-mediated transfer of antisense RNA transcripts into the cervical cancer cells showed longer expression of antisense RNA transcripts and stronger suppression of E6 and E7 proteins expression than did an antisense oligonucleotide approach.
E7 translation may be inhibited by antisense E6 RNA in addition to antisense E7 RNA. Antisense E6 oligonucleotides have been shown to decrease E7 protein expression [18]. In cervical cancers and cervical cell lines containing HPV, mRNAs encoding E6 and E7 proteins are transcribed from the same promoter in the form of a bicistronic transcript [33]. The bicistronic transcript is spliced to produce two shorter transcripts [34]. It has been proposed that it is a function of these spliced transcripts to stimulate E7 translation [33]. Because all three transcripts have the same 5′ oligonucleotide sequence, one would assume that antisense E6 oligonucleotides could affect all three. Our adenoviral vector, Ad5CMV-HPV 16 AS, includes both antisense E6 RNA and antisense E7 RNA, and our results show that the decrease of E7 protein expression in the Ad5CMV-HPV 16 AS-infected cells was more distinct than the decrease of E6 protein expression. Another antisense approach is use of the hammerhead ribozyme, which involves both catalytic and antisense effects. The hammerhead ribozyme affects the phenotype of HeLa cells and increases by a small amount the intracellular concentration of p53 protein [35]. In our study, Western blot and immunohistochemical analyses showed a marked increase of p53 protein in Ad5CMVHPV 16 AS-infected SiHa cells compared with control vector-infected cells. This marked increase of p53 protein may be due to a strong suppression of E6 protein by overexpression of antisense E6 RNA. The mechanism by which p53 protein inhibits growth may be related to the arrest of the cell cycle at G1 [36], apoptosis [27], or induction of another tumor suppressor gene, such as WAF1/ CIP1 [37]. The induction of apoptosis is one of several documented functions of p53 protein. Apoptosis is a selective process of physiological cell deletion. It seems that chromatin cleavage is the most characteristic biochemical feature of the process. To determine the role of apoptosis in the growth inhibition we observed, we used the TUNEL method and flow cytometry analysis. Our results showed apoptotic nuclear DNA fragmentation in the Ad5CMV-HPV 16 AS-infected SiHa cells in vitro. The TUNEL staining in the Ad5CMV-HPV 16 AS-injected SiHa cell tumors also showed apoptotic nuclear DNA fragmentation. These results strongly suggest that one of the mechanisms of growth suppression and cell death induced by Ad5CMV-HPV 16 AS in cervical cancer SiHa cells is apoptosis mediated by the p53 protein. In this study, the protein expression of Rb significantly increased after the infection with Ad5CMV-HPV 16 AS. This increase of Rb was accompanied by the decrease of E7. It is reported that antisense E7 oligonucleotides increase Rb protein expression as well as decrease E7 protein expression [18]. E7 protein is known to bind the product of the retinoblastoma susceptibility gene, Rb, thereby preventing Rb from regulating cell growth [38]. Inhibition of E7 protein expression allows Rb to perform its function of regulating cell growth [31]. E7 can directly interfere with cell cycle control by inducing S-phase entry by sequential activation of cyclin A and cyclin E gene
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
expression [39]. Both E6 and E7 play important roles in affecting specific though distinct checkpoint controls in G1 [40,41]. It has been suggested that the transforming function of E6 is independent of a direct association with p53, because mutants of E6 that fail to bind to p53 or promote p53 degradation are still able to cooperate with ras in an immortalization assay [40]. Therefore, in addition to restoration of p53 and Rb function, a p53- and Rb-independent pathway might contribute to the reduction of proliferation of cervical cancer cells by Ad5CMV-HPV 16 AS. The results of this study are noteworthy in several ways. First, a microscopic tumor model, in which antisense HPV 16 E6 and E7 oligonucleotides are injected into the tumor 10 days after inoculation with cervical cancer cells, had previously been developed for HPV 16-positive cervical cancer [18]. However, that tumor model is too small for evaluation and prediction of the effectiveness of antisense E6 and E7 in humans. In this study, a larger established tumor model involving tumors of 5 to 6 mm in diameter has been developed. Second, there are no previous reports of antisense plasmid vectors for in vivo experiments, because transduction efficiency of plasmid vectors is too low to allow assessment of tumor growth inhibition. Our use of an adenoviral vector resulted in efficient transduction and tumor reduction. Third, there are no previous reports of induction of apoptosis for antisense E6 treatment. This may be due to the distinct suppression of expression of E6 protein by the overexpression of antisense E6 RNA after the adenovirus-mediated transfer. Furthermore, overexpression of p53 by adenovirus-p53 in combination with Ad5CMV-HPV 16 AS might induce the potent induction of apoptosis and tumor reduction compared with Ad5CMV-HPV 16 AS only. Collectively, our new model of gene therapy for cervical cancer on the basis of Ad5CMV-HPV 16 AS is promising for future clinical trials. Acknowledgments We would like to sincerely thank Dr. M. Higuchi and Dr. T. Kamitani for helpful discussion. We would like to sincerely thank Mr. Keizo Oka for their help in preparing medium. This study was supported by a grant from the Ministry of Education, Science, Sports and Culture, Japan, and a grant from the Ministry of Health and Welfare for a Comprehensive Strategy for Controlling Intractable Cancer, Japan. This study was also supported by the Integrated Center for Science, Ehime University. References [1] Brinton LA. The epidemiology of human papillomavirus and cervical cancer. In: Munoz N, Bosch F, Shah KY, Meheus A, editors. IARC Sci Publ: Lyon, vol. 119; 1992. p. 3–23. [2] Mitchell MF, Hittelman WK, Lotan R, Nishioka K, Tortolero-Luna G, Richards-Kortum R, et al. Chemoprevention trials and surrogate end point biomarkers in the cervix. Cancer 1995;76:1956–77. [3] Durst M, Gissmann L, Ikenberg H, zur Hausen H. A papillomavirus DNA from a cervical carcinoma and its prevalence in cancer biopsy samples from different geographic regions. Proc Natl Acad Sci U S A 1983;80:3812–5.
829
[4] Steele C, Shillitoe EJ. Viruses and oral cancer. Crit Rev Oral Biol Med 1991;2:153–75. [5] Scheffner M, Huibregtse MM, Vierstra RD, Howley PM. The HPV-16 E6 and E6-AP complex functions as a ubiquitin–protein ligase in the ubiquitination of p53. Cell 1993;75:495–505. [6] Fujita M, Inoue M, Tanizawa O, Iwamoto S, Enomoto T. Alterations of the p53 gene in human primary cervical carcinoma with and without human papilloma infection. Cancer Res 1992;52:5323–8. [7] Schwarz E, Freese UK, Gissmann L, Mayer W, Roggenbuck B, Stremlau A, et al. Structure and transcription of human papillomavirus sequences in cervical carcinoma cells. Nature 1985;314:111–4. [8] Kaur P, McDougall JK, Cone R. Immortalization of primary human epithelial cells by cloned cervical carcinoma DNA containing human papillomavirus type 16 E6/E7 open reading frames. J Gen Virol 1989;70: 1261–6. [9] Howley PM, Scheffner M, Münger K. Oncoproteins encoded by the cancer-associated human papillomavirus target the products of the retinoblastoma and p53 tumor suppressor genes. Cold Spring Harbor Symp Quant Biol 1991;56:149–55. [10] Crook T, Morgenstern JP, Crawford L, Banks L. Continued expression of HPV 16 E7 protein is required for maintenance of the transformed phenotype of cells co-transformed by HPV-16 plus EJ-ras. EMBO J 1989;8:513–9. [11] Watanabe S, Kanda T, Yoshiike K. Growth dependence of human papillomavirus 16 DNA-positive cervical cancer cell lines and human papillomavirus 16-transformed human and rat cells on the viral oncoproteins. Jpn J Cancer Res 1993;84:1043–9. [12] Inouye M. Antisense RNA: its functions and applications in gene regulation—A review. Gene 1988;72:25–34. [13] Crooke ST. Therapeutic applications of oligonucleotides. Biotechnology 1992;10:882–6. [14] Stein CA, Cheng YC. Antisense oligonucleotides as therapeutic agents— Is the bullet really magical? Science 1993;261:1004–12. [15] von Knebel Doeberitz M, Oltersford T, Schwarz E, Gissmann L. Correlation of modified human papillomavirus early gene expression with altered growth properties in C4-I cervical carcinoma cells. Cancer Res 1988;48:3780–6. [16] Steele C, Sacks PG, Adler-Storthz K, Shillitoe EJ. Effect on cancer cells of plasmids that express antisense RNA of human papillomavirus type 18. Cancer Res 1992;52:4706–11. [17] Steele C, Cowsert LM, Shillitoe EJ. Effects of human papillomavirus type 18-specific antisense oligonucleotides on the transformed phenotype of human carcinoma cell lines. Cancer Res 1993;53:2330–7. [18] Tan TMC, Ting RCY. In vitro and in vivo inhibition of human papillomavirus type 16 E6 and E7 genes. Cancer Res 1995;55:4599–605. [19] Kucharczuk JC, Raper S, Elshami AA, Amin KM, Sterman DH, Wheeldon EB, et al. Safety of intrapleurally administered recombinant adenovirus carrying herpes simplex thymidine kinase DNA followed by ganciclovir therapy in nonhuman primates. Hum Gene Ther 1996;7: 2225–33. [20] Mitchell MF, Hamada K, Sastry J, Sarkar A, Tortolero-Luna G, Wharton JT, et al. Transgene expression in the rhesus cervix mediated by an adenovirus expressing beta-galactosidase. Am J Obstet Gynecol 1996;174: 1094–101. [21] Lu CY, Giordano FJ, Rogers KC, Rothman A. Adenovirus-mediated increase of exogenous and inhibition of endogenous fosB gene expression in cultured pulmonary arterial smooth muscle cells. J Mol Cell Cardiol 1996;28:1703–13. [22] Yamanishi Y, Maeda H, Hiyama K, Ishioka S, Yamakido M. Specific growth inhibition of small-cell lung cancer cells by adenovirus vector expressing antisense c-kit transcripts. Jpn J Cancer Res 1996;87:534–42. [23] Alemany R, Ruan SB, Kataoka M, Koch PE, Mukhopadhyay T, Cristiano RJ, et al. Growth inhibitory effect of anti-K-ras adenovirus on lung cancer cells. Cancer Gene Ther 1996;3:296–301. [24] Lee CT, Wu S, Gabrilovich D, Chen H, Nadaf-Rahrov S, Ciernik IF, et al. Antitumor effects of an adenovirus expressing antisense insulin-like growth factor I receptor on human lung cancer cell lines. Cancer Res 1996;56:3038–41.
830
K. Hamada et al. / Gynecologic Oncology 103 (2006) 820–830
[25] Yee C, Krishnan-Hewlett I, Baker CC, Schlegel R, Howley PM. Presence and expression of human papillomavirus sequences in human cervical carcinoma cell lines. Am J Pathol 1985;119:361–6. [26] Hamada K, Zhang WW, Alemany R, Roth JA, Wolf J, Mitchell MF. Adenovirus-mediated transfer of HPV 16 E6/E7 antisense RNA to human cervical cancer cells. Gynecol Oncol 1996;60:373–9. [27] Hamada K, Alemany R, Zhang WW, Hittelman WN, Lotan R, Roth JA, et al. Adenovirus-mediated transfer of a wild-type p53 gene and induction of apoptosis in cervical cancer. Cancer Res 1996;56: 3047–54. [28] Andreeff M, Darzynkiewicz Z, Sharpless TK, Clarkson BD, Melamed MR. Discrimination of human leukemia subtypes by flow cytometric analysis of cellular DNA and RNA. Blood 1980;55:282–93. [29] Zhang WW, Fang X, Mazur W, French BA, Georges RN, Roth JA. Highefficiency gene transfer and high-level expression of wild-type p53 in human lung cancer cells mediated by recombinant adenovirus. Cancer Gene Ther 1994;1:4–15. [30] Hu G, Liu W, Hanania EG, Fu S, Wang T, Deisseroth AB. Suppression of tumorigenesis by transcription units expressing the antisense E6 and E7 messenger RNA (mRNA) for the transforming proteins of the human papilloma virus and the sense mRNA for the retinoblastoma gene in cervical carcinoma cells. Cancer Gene Ther 1995;2:19–32. [31] Tan TMC, Gloss B, Bernard H-U, Ting RCY. Mechanism of translation of the bicistronic mRNA encoding human papillomavirus type 16 E6-E7 genes. J Gen Virol 1994;7:2663–70. [32] Niu X-Y, Peng Z-L, Duan WQ, Wang H, Wang P. Inhibition of HPV16 E6 oncogene expression by RNA interference in vitro and in vivo. Int J Gynecol Cancer 2006;16:743–51.
[33] Schneider-Gädicke A, Schwarz E. Different human cervical carcinoma cell lines show similar transcription patterns of human papillomavirus type 18 early genes. EMBO J 1986;5:2285–92. [34] Shirasawa H, Tanzawa H, Matsunaga T, Shimizu B. Quantitative detection of spliced E6-E7 transcripts of human papillomavirus type 16 in cervical premalignant lesions. Virology 1991;184:795–8. [35] Chen Z, Kamath P, Zhang S, Weil MM, Shillitoe EJ. Effectiveness of three ribozymes for cleavage of an RNA transcript from human papillomavirus type 18. Cancer Gene Ther 1995;2:263–71. [36] Diller L, Kassel J, Nelson CE, Gryka MA, Litwak G, Gebhardt M, et al. p53 functions as a cell cycle control protein in osteosarcomas. Mol Cell Biol 1990;10:5772–81. [37] El-Deiry WS, Harper JW, O'Connor PM, Velculescu VE, Canman CE, Jackman J, et al. WAF1/CIP1 is induced in p53-mediated G1 arrest and apoptosis. Cancer Res 1994;54:1169–74. [38] Dyson N, Howley P, Münger K, Harlow E. The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science 1989;243:934–7. [39] Zerfass K, Schulze A, Spitkovsky D, Friedman V, Henglein B, JansenDürr P. Sequential activation of cyclin E and cyclin A gene expression by human papillomavirus type 16 E7 through sequences necessary for transformation. J Virol 1995;69:6389–99. [40] Storey A, Massimi P, Dawson K, Banks L. Conditional Immortalization of primary cells by human papillomavirus type 18 E6 and EJ-ras defines an E6 activity in G0/G1 phase which can be substituted for mutations in p53. Oncogene 1995;11:653–61. [41] Chen JJ, Reid CE, Band V, Androphy EJ. Interaction of papillomavirus E6 oncoproteins with a putative calcium-binding protein. Science 1995; 269:529–31.