1650 Human Tumor Growth Is Inhibited by a Vaccinia Virus Carrying the E2 Gene of Bovine Papillomavirus Viviana Valadez Graham, Biol. 1 Gerd Sutter, D.V.M., Ph.D.2 Marco V. José, Ph.D.3 Alejandro Garcı́a-Carranca, Ph.D.1 Volker Erfle, D.V.M., Ph.D.2 Norma Moreno Mendoza, Ph.D.4 Horacio Merchant, Ph.D.4 Ricardo Rosales, Ph.D.1 1 Department of Molecular Biology, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, México City, México 2 GSF-Center for Enviromental and Health Research, Institute of Molecular Virology, Neuherberg-Muenchen, Germany. 3 Department of Biophysics and Biomathematics, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, México City, México. 4 Department of Cell Biology, Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, México City, México. BACKGROUND. Papillomavirus is the etiologic agent associated with cervical carcinoma. The papilloma E2 protein is able to regulate negatively the expression of E6 and E7 papilloma oncoproteins. Therefore, a new, highly attenuated vaccinia virus known as modified vaccinia virus Ankara (MVA), which carries the papillomavirus E2 gene, was used for the treatment of tumors associated with human papillomavirus. METHODS. Analysis of expression of the E2 gene from the recombinant vaccinia virus was performed by reverse transcription–polymerase chain reaction of RNA isolated from infected cells. Detection of the E2 protein was done by immunoprecipitation from proteins labeled with [35S]-methionine, isolated from infected cells. The therapeutic effect of the MVA E2 recombinant virus over human tumors was tested in nude mice bearing tumors generated by inoculation of HeLa cells. Series of 10 nude mice with tumors of different sizes were injected with MVA, MVA E2, or phosphate-buffered saline. Tumor size was monitored every week to assess growth. RESULTS. The MVA E2 recombinant virus efficiently expressed the E2 protein in BS-C-1 cells. This protein was able to repress, in vivo, the papillomavirus P105 promoter, which controls the expression of the E6 and E7 oncoproteins. In nude mice the MVA E2 virus reduced tumor growth very efficiently. In contrast, tumors continued to grow in mice treated with MVA or PBS. The life expectancy of MVA E2-treated mice was also increased three- to fourfold compared with that of animals that received MVA or PBS. CONCLUSIONS. The growth of human tumors was efficiently inhibited by the MVA E2 recombinant vaccinia virus. The absence of side effects in treated animals suggested that the MVA E2 virus is a safe biologic agent that could in the future be used in humans for the treatment of cervical carcinoma. Cancer 2000;88:1650 – 62. © 2000 American Cancer Society. KEYWORDS: papillomavirus, vaccinia virus, MVA strain, cervical carcinoma. Supported in part by grant 3080N (from CONACyT, Mexico) and by grant IN211394 (from DGAPA– UNAM, México City, México). The authors thank Dr. Peter Howley for providing the pC59 plasmid, Dr. Jaime Berumen for providing the BPV-1 E2 antiserum, and Dr. Carlos Rosales for reading and commenting on the manuscript. They also thank Miriam Guido and Rosa Maria Dominguez for technical assistance. Address for reprints: Ricardo Rosales, Ph.D., Department of Molecular Biology, Instituto de Investigaciones Biomédicas–UNAM, Apto. Postal 70228, Cd. Universitaria, México, D.F.– 04510, Mexico. Received June 30, 1999; revision received December 1, 1999; accepted December 1, 1999. © 2000 American Cancer Society I n humans, neoplasic transformation has been linked to the presence of human papillomaviruses (HPVs). These viruses can induce diseases, from warts up to condylomas, and lesions that can progress to malignant neoplasia. Approximately 1 million people are infected with HPV every year. It is well known that cervical carcinoma correlates with the presence of HPVs, particularly types 16, 18, 31, 33, and 35.1–11 In contrast, HPV types 6, 11, 42, and 43 are found only in the anogenital tract and are not associated with cervical carcinoma.1,3,12,13 Cervical carcinoma is a serious health problem. In developing countries, 50,000 women die annually of this malignancy. Normally, HPVs infect and replicate themselves in the form of an unintegrated circular episome in keratinocytes of genital mucosa and perigenital skin. The papillomavirus E1 and E2 gene products regulate viral DNA replication. The E2 gene product can also activate or Papilloma E2 Gene in Tumor Therapy/Valadez et al. repress transcription of different HPV promoters.14,15 In particular, the papillomavirus protein E2 is known to down-regulate the P105 promoter from HPV type 18, which controls transcription of the E6 and E7 oncogenes. These oncoproteins are expressed in a variety of cervical human tumors.16 The best-studied HPVs, types 16 and 18, persist extrachromosomally in precancerous lesions, but they are often integrated into the cellular genome, causing transformation of the cell.17 Integration of HPV into the cell genome leads to a disruption and inactivation of the papillomavirus E2 gene. This event then results in derepression of the E6 and E7 oncogenes. Expression of these oncogenes appears to be a critical step in the maintenance of the transformed stage and progression to invasive carcinoma. This mechanism, however, is not necessarily the only one operating to induce the malignant stage, as suggested by recent studies in which HPV integration was not detected in at least 30% of the analyzed tumors.13,18 Due to the strong dependence of cervical carcinoma on infection by HPVs, it is thought that the induction of a protective stage against these viruses could help prevent the appearance of cervical tumors. Based on this idea, different strategies to develop a safe vaccine or immunotherapeutic agent against cervical carcinoma have been assessed. Other methods, such as radiotherapy and chemotherapy, have, of course, been used to reduce papillomas and cancer tumors. However, these methods work efficiently only during the first stages of tumor development. Later, it becomes very difficult to treat cervical tumors due to large tumor sizes and the negative side effects that anticancer drugs may have. Being aware of these difficulties in treating cervical carcinoma, we decided to use a viral vector based on an attenuated vaccinia virus known as modified vaccinia virus Ankara (MVA). This virus was developed and tested as a safe smallpox vaccine.19 It was also found to be avirulent for normal or immunosuppressed individuals and not to have negative side effects in 120,000 humans inoculated for priming vaccination.20 –22 One reason for the safety of MVA is that this virus has its viral expression and recombination mechanisms impaired. Moreover, MVA is capable of infecting most, if not all, the human cell lines tested up to now. It has also been demonstrated that MVA is an excellent vector for expressing foreign genes, such as the Escherichia coli Lac Z or the phage T7 polymerase,23,24 under nonpermissive conditions in infected cells. Because of these characteristics, the most successful current strategy for vaccine development involves the use of vaccinia virus vectors. The approach of expressing a foreign protein via 1651 vaccinia virus vectors has already been used to protect animals from other virus infections.25,26 Treatment of rats with live recombinant vaccinia viruses expressing tumor specific antigens from polyoma virus could prevent cognate tumor development and in some cases could also induce regression of preexisting tumors.25 In addition, an MVA recombinant virus expressing the hemagglutinin and the nucleoprotein genes of influenza virus was found to protect mice fully against a lethal influenza virus challenge.26 Based on this strategy, we decided to investigate the potential use of a recombinant MVA virus for the treatment of cervical carcinoma. In order to do this, we constructed an MVA recombinant virus by inserting the E2 gene from bovine papillomavirus (BPV-1) into the MVA genome. The E2 gene was chosen because the E2 protein represses expression of E6 and E7 proteins and can also induce p53-mediated apoptosis in HeLa cells. This protein is also able to arrest cell growth independently of the transcriptional repression of the endogenous viral E6 and E7 genes.15,27,28 The new recombinant virus was named MVA E2 and its effects over human tumor growth in immunosuppressed animals were investigated. We found that treatment with this recombinant MVA E2 virus resulted in human tumor growth arrest in nude mice. In contrast, only a slight inhibition of tumor growth was observed in animals inoculated with the parental MVA virus. In control animals just inoculated with phosphate-buffered saline (PBS), tumors continued to grow indefinitely. These data indicated that the recombinant MVA E2 virus has the capacity of inducing elimination of papilloma tumors from animals, and therefore it becomes a potential new therapeutic agent for cervical carcinoma. METHODS Mice Nude male mice (Mus musculus) age 8 weeks were purchased from Taconic Laboratory (New York, NY) and kept under “sterile” conditions in isolated cages. Cells and Viruses Monkey kidney (BS-C-1), human carcinoma (HeLa), and chicken embryo fibroblast (CEF) cells were grown in Dulbecco modified Eagle medium (DMEM) (Gibco BRL, Gaithersburg, MD) supplemented with 10% fetal calf serum (FCS) (Gibco BRL, Gaithersburg, MD). Cells were maintained in a humidified air-5% CO2 atmosphere at 37 °C. 3T3 F4 cells, which contain integrated copies of the ␤-galactosidase gene under the control of the P105 promoter of HPV type 18, were also grown in DMEM with 10% FCS. Vaccinia virus strain MVA and MVA E2 recombi- 1652 CANCER April 1, 2000 / Volume 88 / Number 7 FIGURE 1. Schematic map of the MVA genome and the plasmid pIIIgpt dsP E2 used for insertion of the bovine papilomavirus E2 gene. Hind III restriction endonuclease sites within the genome of MVA are indicated at the top (kb: kilobase pairs). Letters at the top represent the MVA fragments generated by Hind III digestion. DNA sequences (flank 1 and flank 2) adjacent to deletion III within the A fragment of the vaccinia genome were cloned into the plasmid PIIIgptdsP to allow recombination into the MVA genome. A cassette with back-to-back copies of a strong synthetic vaccinia virus early/late promoter (dsP) is shown. P 7.5 refers to well-characterized vaccinia virus early/late promoter. A cassette with the E2 gene was inserted in front of one of the vaccinia virus early/late promoters. The E. coli guanine phosphoribosyl transferase (gpt) gene under control of the P 7.5 promoter is also shown. nant viruses were routinely propagated and titrated by endpoint dilution in CEF cells to obtain the 50% tissue culture infectious dose (TCID50). Plasmid Construction A cDNA copy of the bovine papillomavirus E2 gene flanked by Bam HI and AscI sites was amplified by the polymerase chain reaction (PCR) using DNA from plasmid pC59 (donated by Dr. Peter Howley from the National Institutes of Health, Bethesda, MD)29 as a template. This plasmid contains an open reading frame that covers part of the E1 and E2 papilloma genes. A set of primers, GS78-CAGCAGGGATCCAGGATGGAGACAGCATGCGAACGT and GS79-CAGCAGGCGCGCCCATCATTGGTGGTGCGCCTT, were used to amplify the E2 gene from the pC59 plasmid. The PCR product was cloned into the unique BamHI and AscI sites of the pIIIgpt dsP plasmid, which carries two copies of the strong synthetic vaccinia virus early/late promoter23 (Fig. 1). The resulting plasmid (pIII gpt dsp E2) was purified by cesium chloride gradients and used to construct the MVA E2 recombinant virus. Construction of Recombinant Vaccinia Virus To generate a recombinant MVA virus, monolayers of nearly confluent CEF cells in 6-well plates were infected with MVA at a multiplicity of 0.05 tissue culture infectious dose (TCID50) of MVA per cell. Ninety minutes after infection, cells were transfected with 10 g of plasmid DNA (pIIIgpt dsP E2) using Lipofectin re- agent (Gibco BRL, Gaithersburg, MD) as recommended by the manufacturer. At 48 hours after infection, cells were harvested and processed as previously described.23 Recombinant MVA virus expressing the E2 protein was selected by 6 consecutive rounds of plaque purification on CEF cells in the presence of mycophenolic acid (25g/mL).19 Subsequent virus stocks were also prepared in CEF cells. Analysis of Viral DNA Viral DNA was isolated by digestion of virus-infected cells with proteinase K (50 g/mL) in TE buffer (10 mM Tris HCl, pH ⫽ 8.0, 1 mM ethylenediaminetetraacetic acid), during 8 hours at 37 °C followed by extraction with phenol/chloroform. DNA was precipitated with ethanol and resuspended in water. Viral DNA (3 g) was PCR-amplified using 40 cycles of 1 minute at 94 °C, 2 minutes at 45 °C, and 2 minutes at 72 °C. The primers used for amplification were GS82 (5⬘-GGT TGT TGA TGG ATC TGT GAT GCA TGC GAT AGC TGA-3⬘) and GS83 (5⬘-GAA TGC ACA TAC ATA AGT ACC GGC ATC TCT AGC AGT-3⬘). These primers are complementary to sequences located within the flank 1 and flank 2 regions adjacent to deletion III within the Hind III A fragment of the vaccinia genome, respectively (Fig. 1). PCR products were cut with the restriction enzymes BamHI and AscI, and analyzed by electrophoresis on 1% agarose gels. Papilloma E2 Gene in Tumor Therapy/Valadez et al. Reverse Transcription-PCR Analysis of RNA Total RNA (3 g) from infected BS-C-1 cells (for the E2 gene) or from infected 3T3 F4 cells (for the ␤-galactosidase gene) was isolated as described30 and transcribed with 20 units of avian myeloma virus reverse transcriptase (Boehringer Mannheim, Indianapolis, IN), using 0.5 g of oligo(dt)12–18 in a total volume of 50 L of buffer (50 mM Tris-HCl, pH 8; 70 mM KCl; 10 mM MgCl2; 4 mM dithiothreitol; and 1 mM each of the 4 deoxyribonucleoside triphosphates) at 42 °C for 90 minutes. The cDNA product was purified by phenolchloroform extraction, precipitated with ethanol, and resuspended in 50 L of water. One L of cDNA was amplified by 40 cycles of PCR as described above. Primers used for amplification of the E2 gene were GS82 and GS83, and for amplification of a 294 bp within the ␤-galactosidase gene were BGR-10 (5⬘- TCG CGG TGA TGG TGC TGC GTT GG-3⬘) and BGR-11 (5⬘-GTT ACG ATG CGC CCA TCT ACA CCA A-3⬘). Detection of the Recombinant Papilloma E2 Protein BS-C-1 cells grown in 12-well plates were infected with the MVA or the MVA E2 recombinant virus at a multiplicity of 15 plaque formation unit (pfu) per cell. Virus were adsorbed on the cells for 1 hour at 37 °C in DMEM with 2% FCS. Cultures were then supplemented with DMEM containing 10% FCS and incubated at 37 °C in a 5% CO2 atmosphere. Six hours after infection, the medium was removed, and the cells were washed once with 1 mL of methionine free DMEM. To each well, 0.2 mL of methionine free DMEM supplemented with 50 Ci of [35S]-methionine (Dupont, Boston, MA) was added and incubated for 30 minutes at 37 °C. Cytoplasmic extracts of infected cells were then prepared, after removal of the medium, by incubating the cells in each well with 0.2 mL of lysis buffer (0.5% Nonidet P-40; 10 mM Tris-HCl, pH 8; 1 mM EDTA) for 10 minutes at 37 °C. Cell extracts were incubated with an anti-E2 polyclonal serum for 1 hour, and protein-antibody complexes were immunoprecipitated with 50 L of Protein A-Sepharose (Pharmacia Biotech., Uppsala, Sweden). Immunoprecipitates were then washed with PBS, resuspended in 50 L of Laemmli sample buffer, and separated by 10% sodium dodecyl sulfate (SDS)–polyacrilamide gel electrophoresis (PAGE). Resolved proteins were analyzed by autoradiography. 1653 sidase specific activity32 were determined as previously described. Maximum activity of ␤-galactosidase (100%) was taken as 104 units/mg of protein from uninfected cells. Analysis of Tumor Growth One hundred fifty nude mice were inoculated subcutaneously with 2 ⫻ 106 HeLa cells in 200 L of PBS. Two or 3 weeks later, when tumor size was between 0.1 and 0.5 cm2, the animals (50 for each group) were injected directly into the tumor with 5 ⫻ 107 units of MVA or MVA E2 once a week for 3 weeks. Control animals received only PBS. The area of tumor growth was evaluated every week. The relative tumor growth was calculated by measuring the tumor size every week and subtracting the size of the tumor in the previous week. Because tumor growth in these animals followed a sigmoidal curve, and mathematical studies have shown that tumor growth in animals can be described with a logistic differential equation, we decided to use the logistic differential equation dG共t兲/dt ⫽ rG关1 ⫺ G/k兴, which describes the tumor growth in the absence of an immune response. With this equation we can calculate the behavior of tumor growth when different amounts of MVA E2 virus are used for the tumor treatment. We can also determine how the tumor changes over time with or without an external agent like the MVA E2 virus. In this equation, r is the intrinsic rate of tumor growth and k is the maximum size of the tumor.33 The solution for the logistic differential equation is as follows: G共t兲 ⫽ Gok/关Go ⫹ 共k ⫺ Go兲exp共 ⫺ rt兲兴, where G(o) is the initial size of the tumor, r and k have the same definition as above, and t is time. Fitting of the experimental data (area of tumor size estimated every week) to the last equation was made using a nonlinear least-square estimation following a Marquardt procedure.34 Survival Analysis Survival data were analyzed by the standard Kaplan– Meier survival curves, using the computer program GraphPad Prism.33 Identification of Apoptosis in Tumor Cells ␤-Galactosidase Assay Confluent 3T3 F4 cells were infected with different amounts of MVA or MVA E2 viruses and incubated at 37 °C. Twelve hours later, cytoplasmic extracts were prepared, and protein concentration31 and ␤-galacto- Human tumors grown in nude mice were injected with MVA or MVA E2 recombinant vaccinia viruses. After 24 hours of virus infection, tumors were taken from the animals, fixed in 4% buffered formaldehyde, and embedded in paraffin. Sections of 6 m adhered 1654 CANCER April 1, 2000 / Volume 88 / Number 7 to slides treated with 0.05% poly-L-Lysine (300,000 daltons of molecular weight) (Sigma, St. Louis, MO) were deparaffinized in xylol, dehydrated, and incubated with 20 g/mL proteinase K. After washing in PBS, the slides were treated according to the instructions of the In Situ Cell Death Detection Kit, AP (Boehringer Mannheim Co., Indianapolis, IN). Briefly, sections were incubated with TUNEL reaction mixture for 1 hour at 37 °C, washed in PBS, and analyzed by fluorescence microscopy (Nikon Inc., Melville, NY). Positive and negative controls were included for each experiment. For inducing DNA strand breaks, sections were treated with 1 g/mL DNAase (Sigma, St. Louis, MO) for 30 minutes at 37 °C. For negative controls, sections were incubated in Label solution without terminal deoxynucleotidyl transferase. RESULTS Construction and Isolation of the MVA Recombinant Virus Expressing the Bovine Papillomavirus E2 Gene Previous work has shown that MVA can serve as an efficient expression vector and also as a highly attenuated smallpox vaccine.19,23,26,35,36 This virus was used to construct a recombinant virus carrying a papillomavirus regulatory gene (the E2 gene). The MVA plasmid pIIIgpt dsP (Fig. 1) contains two copies of the strong synthetic vaccinia virus early/late promoter and multiple cloning sites. The bovine papillomavirus E2 gene was inserted into pIIIgpt dsP, as described in “Methods,” to form the recombinant plasmid pIIIgpt dsP E2 (Fig. 1). The plasmid pIIIgpt E2 was transfected into CEF cells. These cells were then infected with MVA virus to obtain the new MVA E2 recombinant virus. Homologous recombination occurs between the flank 1 and flank 2 sequences of the plasmid with the same corresponding sequences in the vaccinia virus genome in order to generate a recombinant vaccinia virus carrying the E2 gene (Fig. 1). Serial dilutions of the infected cell lysates were plated on CEF cells in the presence of mycophenolic acid, to select for recombinants. Characterization of MVA E2 Recombinant Virus Infected BS-C-1 cells were used to characterize the MVA E2 DNA. Integration of the E2 gene into the viral genome was demonstrated by PCR and restriction enzyme analysis. Viral DNA was isolated from infected cells and two sets of primers (GS82 and GS83) were used to amplify a DNA fragment covering the region between flank 1 and flank 2 of the vaccinia virus sequences (Fig. 1). The amplified DNA fragment was cut with the restriction enzymes BamHI and AscI in order to release the E2 gene. The DNA fragment obtained corresponded in size to the E2 gene, and it was also inserted with the right orientation into the MVA genome (data not shown). In order to confirm that the MVA E2 recombinant virus induced the expression of the E2 gene, BS-C-1 cells were infected with this virus, and the presence of E2 RNA and E2 protein was then analyzed. First-strand reverse transcription products, derived from purified total RNA from MVA E2–infected and uninfected BSC-1 cells, were amplified by PCR with the primers GS78 and GS79. The PCR product derived from the E2 gene is a 942 bp fragment. Twenty-four hours after MVA E2 infection, papilloma E2 RNA could be easily detected in BS-C-1 cells (Fig. 2A). E2 RNA also augmented considerably by 72 hours postinfection (Fig. 2B). This result confirmed that MVA E2 was in fact directing the expression of the E2 gene in infected cells and also suggested that there was an accumulation of E2 RNA during infection. Synthesis of the E2 protein was confirmed by metabolic labeling with [35S]-methionine of BS-C-1 cells infected with MVA or MVA E2 recombinant virus. In each case the labeled polypeptides were immunoprecipitated with an anti-E2 polyclonal serum. These antibodies reacted specifically with a 48 kDa protein, which corresponded in size to the E2 gene product (Fig. 2C). This result showed that a complete E2 protein was being produced inside MVA E2–infected cells. MVA E2 Recombinant Virus Represses Specifically the Human Papillomavirus P105 Promoter In order to determine the efficacy of the MVA E2 recombinant virus to repress the papillomavirus P105 promoter, which negatively regulates the E6 and E7 oncogenes, a stable 3T3 F4 cell line containing the ␤-galactosidase gene controlled by the P105 promoter of papillomavirus type 18 was infected with MVA E2. After infection with either MVA or MVA E2 recombinant viruses, total RNA from these 3T3 F4 –infected cells was isolated, and the specific RNA for ␤-galactosidase analyzed by reverse transcription–PCR. The DNA fragment of 295 bp corresponding to an internal portion of the ␤-galactosidase gene could be easily detected in these cells. The MVA E2 recombinant virus reduced the transcription activity from the P105 promoter by about 50% as indicated by a reduction in the amount of ␤-galactosidase RNA (Fig. 3A). The enzymatic activity of ␤-galactosidase was also reduced in a dose-dependent manner by MVA E2, up to 60% from the activity of uninfected cells (Fig 3B). In contrast, MVA infection did not cause any reduction in ␤-galactosidase activity (Fig. 3B). Papilloma E2 Gene in Tumor Therapy/Valadez et al. 1655 FIGURE 2. Expression of the E2 gene and production of the E2 protein from MVA E2. (A) Amplification of E2 RNA. Cellular RNA from BS-C-1 cells infected with either with MVA (lanes 1–3) or MVA E2 (lanes 4 – 6) was purified 24 hours after infection. The E2 RNA was amplified by RT-PCR. The quantities, in L, of the PCR product are indicated above each lane. The band representing the E2 gene transcript is marked by an arrow. M represents DNA size markers in bp. (B) E2 RNA accumulation. Cellular RNA from BS-C-1 cells was isolated at different times postinfection and processed as described in (A). Three L of the PCR product were analyzed at 24, 48, and 72 hours postinfection (hpi) on an agarose gel. (C) Immunoprecipitation analysis of cytoplasmic extracts from BS-C-1 cells infected with MVA or MVA E2. At 6 hours after infection, cells were labeled with [35S]-methionine for 60 or 120 minutes. Cell lysates were prepared and 50 L were immunoprecipitated with an anti-E2 polyclonal antibody. Immunoprecipitates were resolved by SDS-PAGE and autoradiographed. Numbers on the left indicate the position and molecular mass in kilodaltons (kDa) of protein standards. The band representing the E2 protein (48 kDa) is marked by an arrowhead. 1656 CANCER April 1, 2000 / Volume 88 / Number 7 FIGURE 3. Repression of human papillomavirus P105 promoter by MVA E2 recombinant virus. (A) RNA from MVA- or MVA E2–infected 3T3-F4 cells was used to amplify the internal fragment of the ␤-galactosidase gene product (arrow) by RT-PCR. Quantities, in L, of the PCR product are indicated above each lane. (B). ␤-galactosidase activity. 3T3 F4 cells were infected for 12 hours with different amounts (1, 5, or 10 pfu) of MVA or MVA E2 recombinant virus. Cytoplasmic extracts were prepared, and ␤-galactosidase activity was determined. Maximum activity (100%) is the activity of ␤-galactosidase from noninfected cells. Results are mean ⫾ standard error of two experiments done in triplicate. MVA E2 Recombinant Virus Reduced Human Tumor Growth in Nude Mice Because the E2 protein can repress the expression of E6 and E7 oncogenes and also promote apoptosis in some cancer cells,37 we decided to evaluate the effect of the MVA E2 recombinant virus on human tumors grown in nude mice. Groups of 50 tumor-bearing nude mice were injected with MVA or MVA E2 virus using 5 ⫻ 107 infectious units per injection, or just with PBS directly into the tumor. This dose of virus has been shown to be adequate for good infection and stimulation of the immune system in normal animals.26,35 Mice were treated once a week for 3 weeks. A significant reduction in tumor size was observed in MVA E2–treated animals (Fig. 4). In contrast, mice injected with MVA showed only a very slight reduction in tumor growth. No untreated or PBS-treated animals showed any reduction in tumor size (Fig. 4). When tumors in control mice reached about 4 –5 cm2 in size, most animals needed to be sacrificed. These results clearly showed that the MVA E2 virus is capable of stopping human tumor growth. In order to obtain additional information about the specific effects of MVA and MVA E2 recombinant viruses on human tumor growth in nude mice, we decided to use a logistic differential equation. This type of mathematical model has been shown to be useful for describing tumor growth in the absence of an immune response.33 Fitting the solution of the logistic model to our experimental data resulted in a remarkably good match (continuous lines, Fig. 4). This means that the treatments with MVA or MVA E2 did not affect the underlying mechanisms of tumor growth, but rather reduced the maximum size of the tumor. The intrinsic rate of tumor growth was slightly reduced when MVA (r ⫽ 0.14) or MVA E2 ( r ⫽ 0.13) were used in comparison with the value obtained for PBS (r ⫽ 0.16). However, the maximum size of tumor in MVA E2–treated animals was much smaller than in MVA- or PBS-treated mice. These results taken together indicated that the MVA E2 recombinant virus and, most likely, the E2 protein are the main factors determining the reduction of the maximum size of the tumor. Papilloma E2 Gene in Tumor Therapy/Valadez et al. 1657 creasing the life expectancy of tumor-bearing animals almost fourfold (Fig. 6). These data indicated that MVA E2 treatment was indeed the responsible factor for tumor arrest and that it also helped the animals to live longer. DISCUSSION FIGURE 4. Growth rate analysis of HeLa cell tumors in nude mice after MVA or MVA E2 treatment. Human tumors grown in nude mice were injected with MVA, MVA E2, or just phosphate-buffered saline (PBS). Average (from all treated animals) tumor size per week is shown (symbols). Equations describing tumor growth for each treatment are as follows: G ⫽ 3.57/(1 ⫹ exp[(3.0624) ⫺ (0.163624)*t]), where k ⫽ 3.57, r ⫽ 0.163624, Go ⫽ 0.2 for PBS; G ⫽ 2.72/(1 ⫹ exp[(2.27365) ⫺ (0.145432)*t]), where k ⫽ 2.72, r ⫽ 0.145432, Go ⫽ 0.2 for MVA; G ⫽ 1.3/(1 ⫹ exp[(2.35396) ⫺ (0.135547)*t]), where k ⫽ 1.3, r ⫽ 0.135547, Go ⫽ 0.2 for MVA E2. In each case, t ⫽ time (days), Go ⫽ initial tumor size (cm2), and r ⫽ intrinsic growth rate (cm2/day). Equations are fitted to data of tumor growth for each treatment. Results of a nonlinear least-square estimation of the parameters following the procedure of Maquardt are also shown (continuous lines). Arrows indicate times when virus injections were made. MVA E2 Infection Induces Apoptosis in Tumor Cells In Vivo Human HeLa tumors in nude mice were infected with MVA or MVA E2 recombinant virus. After 24 hours of infection, tumors were excised and processed for detection of apoptotic bodies. MVA E2 induced the formation of clusters of apoptotic cells in various parts of the tumor. Apoptotic cells were indicated by the presence of high fluorescent nuclei (Fig. 5B and C). In these regions of the tumor, nuclei appeared to be surrounded by unstained cytoplasm, and none of the peripheral cells were stained. These results clearly indicated that apoptotic bodies were present in discrete tumor regions where the E2 protein was very likely present. In uninfected or MVA-infected tumors, very few, if any, apoptotic bodies were detected (Fig. 5). These results clearly showed that infection with MVA E2 induced tumor death not by simple necrosis, but by inducing (most likely through the E2 protein) programmed cell death. The MVA E2 Recombinant Virus Increased the Life Expectancy of Mice Bearing Human Tumors It was observed that mice treated with MVA E2 recombinant virus survived longer than MVA- or PBS-treated mice. This indicated that MVA E2 is capable of in- A new recombinant virus carrying the papilloma E2 gene (MVA E2) was constructed and its therapeutic properties for human tumors analyzed. In this study, we found that expression of the E2 protein in human cancer cells through MVA E2 infection was able to stop tumor growth. Cervical carcinoma is a very serious health problem affecting thousands of women all over the world. Papillomaviruses are the main infectious agents that cause this disease. The papillomavirus proteins E6 and E7 are the molecules responsible for cell transformation.38 These proteins achieve their effects by interacting with two cellular proteins. E6 binds to the protein p53 and promotes its degradation. p53 is a tumor suppressor gene that plays a central role in controlling the cell cycle. So, when E6 binds to p53, the papilloma protein eliminates the tumor suppressor activity of p53. Similarly, E7 binds to the retinoblastoma (Rb) protein, provoking its inactivation. The Rb gene product forms a complex with the transcription factor E2F, which is required for the transcription of cellular genes that drive the cell into the S-phase of the cell cycle. Inactivation of Rb results in the release of the E2F transcription and, in turn, stimulation of cellular DNA replication. Expression of the viral E6 and E7 oncoproteins is negatively regulated by the binding of the E2 gene product to their promoters. This has been clearly shown for the P105 promoter of HPV type 18 in vivo and in vitro.16,39 – 41 When the viral DNA gets integrated into the cell genome, the E2 gene is disrupted or inactivated. The lack of E2 protein then results in activation of transcription of the E6 and E7 genes.16,28,39,41– 43 This event probably represents the most critical step in progression to invasive carcinoma.9,16,28 Moreover, the E2 protein has been reported to be the main factor promoting cell growth arrest and apoptosis in some cancer cells.27,37 Based on this information, we reasoned that it could be possible to suppress tumor growth if the E2 protein could be introduced into tumor cells. Because vaccinia viruses are excellent vehicles for introduction of foreign genes into cells, we used the highly attenuated vaccinia virus strain MVA to deliver the E2 protein to tumor cells. We also decided to use the MVA vaccinia virus as a recombinant vector because it is very efficient in expressing many foreign 1658 CANCER April 1, 2000 / Volume 88 / Number 7 FIGURE 5. MVA E2–induced apoptosis in tumor cells. Nude mice bearing tumors were injected with MVA or MVA E2 virus directly into the tumor. Twentyfour hours later, infected tumors were isolated and processed for detection of apoptosis. (A) MVA-treated tumor. (B) MVA E2–treated tumor. (C) MVA E2– treated tumor at 10-fold higher magnification. (*) indicates large clusters of apoptotic cells with fluorescent nucleus. Different stages of programmed cell death are shown: 1) nucleus with perinuclear chromatin; 2) pignotic nuclei, and 3) apoptotic bodies. Scale bar is 50 m. genes, leading to the production of large quantities of protein in infected cells.23,24 In addition, MVA vaccinia virus is a safe virus because it possesses a high degree of attenuation due to its loss of 30,000 base pairs of genetic material, and because its use in humans was extensively documented during the worldwide eradication of smallpox.19 –21,36 MVA has also a great vaccine potential due to its capacity for stimulating the immune system.26 In the particular case of papillomavirus tumors, it has already been reported that recombinant vaccinia viruses expressing the E6 and E7 oncoproteins are able to protect 70 – 80% of the immunized rats from a challenge with tumor cells. These animals did not present tumor development after virus treatment.44 In the current study, infection of cells with our MVA E2 recombinant virus resulted in very good expression of the protein E2 inside these cells, as demonstrated by the presence of both the mRNA and the actual protein (Fig. 2). Considering that one of the major functions of E2 is the repression of transcription from the P105 promoter, it was essential to confirm that E2 was really functioning inside cells. Repression of the promoter would result in lower levels of the protein E7. Direct measurements of this protein are, however, very difficult to perform. It has been esti- Papilloma E2 Gene in Tumor Therapy/Valadez et al. 1659 FIGURE 5. (Continued) mated that in HeLa cells, E7 has a very short life—just 13.5 minutes.45 So detecting changes in synthesis, catabolism, or immunoreactivity of E7, even in response to external stimuli, turns out to be very hard. Although an estimation of E7 can be obtained when determinations are made in tissue culture cells, where it is possible to control the number of cells used in each experiment, analyzing changes in the E7 protein concentration inside the tumor is even more difficult because it is complicated to determine the number of cells in the tumor mass. To avoid these difficulties, we looked at the level of transcription of a reporter gene expressed by the P105 papillomavirus promoter. Previous work from our group has shown that the E2 protein of bovine papillomavirus is able to repress different promoters, including the P105 promoter, of genital and cutaneous HPV by directly binding to HPV promoters.14 ␤-galactosidase gene is under control of the P105 promoter in 3T3 F4 cells. When these cells were infected with MVA E2, a dose-dependent inhibition of transcription from the P105 promoter was observed (Fig. 3). These results confirmed that E2 is correctly expressed in infected cells and that it is a functional protein with the capacity of repressing transcription from the P105 promoter. These results are in agreement with previous reports showing that it is possible to repress 100% of the transcriptional activity of different papillomavirus promoters by expressing the E2 protein of bovine papillomavirus inside cells.16,40,41 Because MVA E2 was successfully inducing the production of E2 protein in cells, we decided to infect tumor cells that were growing in an animal, to see whether the presence of the E2 protein in them could inhibit tumor formation. Nude mice, which have a deficient immune system, were inoculated subcutaneously with 2 ⫻ 106 HeLa cells in 200 L of PBS. Two or 3 weeks later, when tumors had grown to about 0.2 cm2, the animals’ tumors were injected directly with 5 ⫻ 107 pfu of MVA or MVA E2. This dose was chosen because it provides a good response in immunocompetent animals26,35 and because it is the amount normally used in humans.46,47 We also wanted to evaluate whether this amount of virus would induce negative side effects in the animals. Infection of human cancer cells with MVA E2 dramatically inhibited tumor growth in nude mice (Fig. 4). In contrast, animals treated with the parental MVA virus or just PBS had tumors that continued growing. The only difference between MVA and MVA E2 is the presence of the E2 gene and therefore the expression of this protein in infected cells. Our observations in nude mice support the hypothesis that by introducing the E2 protein into cancer cells, the transcription of the E6 and E7 oncogenes will stop. Nude mice may not reflect what will be observed with tumors in a human system. However, we wanted to investigate the direct effects of MVA E2 on tumors, without the participation of a complete immune system, to understand better the direct actions of MVA E2 on cancer cells within a tumor. Currently, experiments are underway to assess whether the efficacy of MVA E2 against tumors in 1660 CANCER April 1, 2000 / Volume 88 / Number 7 FIGURE 6. MVA E2 increased the survival of tumor-bearing mice. Nude mice bearing human tumors of different sizes were inoculated with phosphate-buffered saline (PBS), control MVA virus, or MVA E2 virus directly into the tumor, with 5 ⫻ 107 infectious units per injection once a week for 3 weeks. The number of living animals in each group was determined each week for up to 4 months. Inoculation with MVA E2 (open squares) significantly prolonged the survival of mice compared with MVA or PBS treatment (solid symbols). This experiment was done with 50 animals for each treatment using different virus preparations. nude mice is maintained in immunocompetent animals. The efficient way in which MVA E2 infection was able to reduce tumor growth in mice (Fig. 4) suggested that the E2 protein, produced inside infected cells, could also get inside the surrounding cells and then stop the transcription of the E6 and E7 genes, and maybe also promote apoptosis in these cells as well. Ninety days after the MVA E2 recombinant vaccinia virus treatment was finished, mice again showed tumor growth. It is likely that the E2 protein that appears inside tumor cells is capable of not only repressing transcription from the P105 promoter inside tumor cells, but also inducing apoptosis of all E2-expressing cells. However, tumor cells that did not acquire the E2 protein would not be eliminated, and they could continue dividing until new tumors appeared again. Macroscopic observation of tumor lesions indicated that tumor necrosis was present in most animals treated with the MVA E2 virus, whereas in MVAtreated animals only a small spot of necrosis was visible after 3 weeks of treatment. We reasoned that dying cells in MVA E2–treated tumors were dying due to apoptosis triggered by the E2 protein. When tumor sections were analyzed for immunohistochemical detection of DNA strand breaks, a large number of apoptotic cells were observed in MVA E2–infected tumors. Very few apoptotic bodies were seen in control tumors (Fig. 5). These observations also supported the idea that the MVA E2 recombinant virus is capable of stopping tumor growth by inducing apoptosis, most likely through expression of the E2 protein. Mathematical analyses of our results with the formulas describing tumor growth for each treatment (Fig. 4) suggested that it may be possible not only to reduce the maximum size of the tumor, but also, given the logistic pattern of growth, to reduce tumor growth significantly if there is sufficient E2 protein inside all tumor cells. This conclusion is in good agreement with the fact that the E2 protein can induce suppression of growth and cell cycle arrest leading to apoptosis.15,27,37 In conclusion, it is possible to stop tumor growth by introducing the E2 protein inside all cancer cells. It is also probable that a complete tumor disappearance could be achieved if the E2 protein were present inside every cancer cell. Tumor growth inhibition by MVA E2 treatment also resulted in a significant increase in life expectancy for these animals (Fig. 6). Mice survived three or four times longer than animals treated with MVA or PBS. This means that although our recombinant virus did not completely eliminate all tumors in mice, it can efficiently reduce the tumor burden in these animals and therefore be a very good candidate for a new therapeutic agent. Other experiments, similar to ours, underline the protective properties of other MVA recombinant viruses as tools for anticancer therapy. One of these viruses, which carries the hemagglutinin (HA) and nucleoprotein (NP) genes of the influenza virus, was able to protect immunized mice against a lethal influenza virus challenge.26 Another virus, a trivalent simian immunodeficiency virus (SIV)–recombinant MVA virus carrying the gap, pol, and env genes, appeared to affect the extent and pattern of SIV replication following challenge.35 Other types of viruses, such bovine enteroviruses, have been shown to prolong the lifespans of mice bearing human carcinoma cells.22,48 Related experiments have been performed with recombinant vaccinia virus vectors using the wild-type virus expressing the large-T (LT), middle-T (MT), and small-T (ST) antigens of polyoma virus. Immunization of animals with these recombinant viruses could prevent the proliferation of cognate tumors.22,48 The use of BPV-1 proteins cloned into vaccinia virus vectors Papilloma E2 Gene in Tumor Therapy/Valadez et al. also resulted in reduced proliferation of tumors.49 By using other types of viruses, such as bovine enterovirus, as immunizing agents, it was possible to prolong the life-spans of rabbits in which a T cell like leukaemia was induced by injecting F-647a cells (F-647 is an HTLV-1 transformed T-cell line).22 The E2 protein may have other antitumor properties besides its direct effects on malignant cells, as suggested by recent reports in which immunization of rabbits with E1 and E2 proteins (of cottontail rabbit papillomavirus, or CRPV) could prevent the formation of a new focus of papillomas. This treatment also produced strong regression of preexisting papillomas.50 Taken together, these observations strongly support the idea that by using the E2 protein it is possible to induce papilloma tumor regression. Therefore, the usefulness of our new recombinant virus for the treatment of cervical carcinoma is evident. These reports, all taken together, show that viruses are effective tools for inducing expression of particular proteins in cells and that they are capable of inducing, in this manner, a protective response in the treated animals. The current report shows that the introduction of the E2 gene product into human cancer cells by means of our recombinant vaccinia virus (MVA E2) is able to stop human tumor growth. MVA E2 virus was also able to prolong the life expectancy of animals harboring tumors without showing any side effects. The MVA E2 recombinant virus is in fact a very good therapeutic agent for human papilloma tumors present in immunosuppressed animals. Our results suggest, then, that MVA E2 virus is safe biologic agent and that it is possible to use this new recombinant vaccinia virus for the treatment of cervical carcinoma. Experiments are now in progress to explore further the role of the E2 protein in the arrest of different types of tumors and also in immunocompetent animals. REFERENCES 1. 2. 3. 4. 5. 6. Beaudenon S, Kremsdorf D, Croissant O, Jablonska S, WainHobson S, Orth G. A novel type of papillomavirus associated with genital neoplasias. Nature 1986;321:246 –9. Beaudenon S, Kremdorf D, Obalek S, Jablonska S, PehauArnaudet G, Croissant O, et al. Plurality of genital human papillomaviruses: characterization of two new types with distinct biological properties. Virology 1987;161:374 – 84. Lörincz AT, Quinn AP, Lancaster WD, Temple GF. A new type of papillomavirus associated with cancer of the uterine cervix. Virology 1987;159:187–90. Cuzick J, Terry G, Ho L. Human papillomavirus type 16 DNA in cervical smears as predictor of high-grade cervical cancer. Lancet 1992;339:959 – 60. Bosch FX, Muñoz N, de Sanjosé S. Human papillomavirus and other risk factors for cervical cancer. Biomed Pharmacother 1997;51:268 –75. Muñoz N, Kato I, Bosch FX, Eluf Neto J, De Sanjosé S, 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 1661 Ascunce N, et al. Risk factors for HPV DNA detection in middle-aged women. Sex Transm Dis 1996;23:504 –10. Muñoz N, Bosch FX. Cervical cancer and human papillomavirus: epidemiological evidence and perspectives for prevention. Salud Publica Mex 1997;39:274 – 82. Nuovo GJ, Friedman D, Richart RM. In situ hybridization analysis of human papillomavirus DNA segregation patterns in lesions of the female genital tract. Gynecol Oncol 1990;36: 256 – 62. Cullen AP, Reid R, Campion M, Lörincz A. Analysis of the physical state of different human papillomavirus DNAs in intraepithelial and invasive cervical neoplasm. J Virol 1991; 65:606 –12. Dürst M, Gissmann L, Ikenberg H, zur Hausen HA. 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. zur Hausen H, Schneider A. The role of papillomavirus in human anogenital cancer. In: Salzman NP, Mltowley P. The papillomaviruses. New York: Plenum Press, 1987:245– 63. De Villiers EM. Human pathogenic papillomavirus types: an update. Curr Top Microbiol Immunol 1994;186:1–12. Matsukura T, Koi S, Sugase M. Both episomal and integrated forms of human papillomavirus type 16 are involved in invasive cervical cancers. Virology 1989;172:63–72. Guido MC, Zamorano R, Garrido-Guerrero E, Gariglio P, Garcı́a-Carranca A. Early promoters of genital and cutaneous human papillomaviruses are differentially regulated by the bovine papillomavirus type 1 E2 gene product. J Gen Virol 1992;73:1395– 400. Hwang ES, Riese DJ, Settleman J, Nilson CA, Honig J, Flynn S, et al. Inhibition of cervical carcinoma cell line proliferation by the introduction of a bovine papillomavirus regulatory gene. J Virol 1993;67:3720 –9. Thierry F, Howley PM. Functional analysis of E2-mediated repression of the HPV18 P105 promoter. New Biol 1991;3: 90 –100. Berumen J, Casas L, Segura E, Amezcua J, Garcı́a-Carranca A. Genome amplification of human papillomavirus types 16 and 18 in cervical carcinomas in related to retention of E1/E2 genes. Int J Cancer 1994;56:640 –5. Fuchs P, Giardi F, Pfister H. Human papillomavirus 16 DNA in cervical cancers and in lymph nodes of cervical cancer patients: a diagnostic marker for early metastases? Int J Cancer 1989;43:41– 4. Fenner F, Henderson DA, Arita I, Jezek Z, Ladnyi ID. Smallpox and its erradication. Geneva: World Health Organization, 1988. Mayr A, Hochstein-Mintzel V, Stickl H. Infection 1975;3:6 – 14. Mayr A, Stickl H, Multer HK, Danner K, Singer H. Zbl Bakt Hyg I Abt Orig B 1978;167:375–90. Shingu M, Chinami M, Taguch C, Shingu MJ. Therapeutic effects of bovine enterovirus infection on rabbits with experimentally induced adult T cell leukaemia. J Gen Virol 1991;72:2031– 4. Sutter G, Moss B. Nonreplicating vaccinia vector efficiently expresses recombinant genes. Proc Natl Acad Sci U S A 1992;89:10847–1085. Sutter G, Ohlmann M, Erfle V. Non-replicating vector efficiently expresses bacteriophage T7 polymerase. Febs Lett 1995;371:9 –12. 1662 CANCER April 1, 2000 / Volume 88 / Number 7 25. Lathe R, Kieny MP, Gerlinger P, Clertant P, Guzani I, Cuzin F, et al. Tumor prevention and rejection with recombinant vaccinia. Nature 1987;326:878 – 80. 26. Sutter G, Wyatt LS, Foley PL, Bennick JR, Moss B. A recombinant vector derived from the host range–restricted and highly attenuated MVA strain of vaccinia virus stimulates protective immunity in mice to influenza virus. Vaccine 1994;12:53–9. 27. Dowhanick JJ, McBride AA, Howley PM. Supression of cellular proliferation by the papillomavirus E2 protein. J Virol 1995;69:7791–9. 28. Romanczuk H, Thierry F, Howley PM. Mutational analysis of cis elements involved in E2 modulation of human papillomavirus type 16 P97 and type P105 promoters. J Virol 1990; 64:2849 –59. 29. Yang Y-C, Okayama H, Howley PM. Bovine papillomavirus contains multiple transforming genes. Proc Natl Acad Sci U S A 1985;82:1030 – 4. 30. Chirgwin JM, Przybyla AE, McDonald RJ, Rutter WJ. Isolation of biologically active ribonucleic acid form sources enriched in ribonuclease. Biochemistry 1979;18:5294 –7. 31. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye. Anal Biochem 1976;72:248 –54. 32. Miller JH. Experiments in molecular genetics. New York: Cold Spring Harbor Laboratory, 1972. 33. Kuznetsov VA, Makalkin IA, Taylor MA, Perelson AS. Nonlinear dynamics of immunogenic tumors: parameter estimation and global bifurcation analysis. Bull Math Biol 1994; 56:259 –321. 34. Draper NR, Smith H. Applied regression analysis. New York: John Wiley & Sons, Inc., 1981. 35. Hirsch VM, Fuerst TR, Sutter G, Carroll MW, Yang LC, Goldstein S, et al. Patterns of viral replication correlate with outcome in simian immunodeficiency virus (SIV)–infected macaques. Effect of prior immunization with a trivalent SIV vaccine in modified vaccinia virus Ankara. J Virol 1996;70: 3741–52. 36. Stickl H, Hochstein-Mintzel V, Mayr A, Huber HC, Schsfer H, Holzner A. MVA—Stufenimpfung gegen Pocken. Dtsch Med Wochenschr 1974;99:2386 –92. 37. Desaintes C, Demeret C, Goyat S, Yaniv M, Thierry F. Expression of the papillomavirus E2 protein in HeLa cells leads to apoptosis. EMBO J 1997;16:504 –14. 38. Lörincz AT, Temple GF, Kurman RJ, Jenson AB, Lancaster WD. Oncogenic association of specific human papillomavirus type with cervical neoplasia. J Natl Cancer Inst 1987;79: 671–7. 39. Baker CC, Phelps WC, Lingren V, Braun MJ, Gonda MA, Howley PM. Structural and transcriptional analysis of human papillomavirus type 16 sequences in cervical carcinoma cell lines. J Virol 1987;64:962–71. 40. Dostatni C, Lambert PF, Sousa R, Ham J, Howley PM, Yaniv M. The functional BPV-1 E2 trans-activating protein can act as a repressor by preventing formation of the initiation complex. Genes Dev 1991;69:1657–71. 41. Thierry F, Yaniv M. The BPV1-E2 trans-acting protein can be either an activator or a repressor of the HPV18 regulatory region. EMBO J 1987;6:3391–7. 42. Cripe TP, Haugen TH, Turk JP, Tabatabai F, Schmid P, Dürst M, et al. Transcriptional regulation of the human papillomavirus E6 –E7 promoter by a keratinocyte-dependent enhancer, and by viral E2 trans-activator and repressor gene product: implications for cervical carcinogenesis. EMBO J 1987;6:3745–53. 43. 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. 44. Meneguzzi G, Cerni C, Kieny MP, Lathe R. Immunization against human papillomavirus type 16 tumor cells with recombinant vaccinia virus expressing E6 and E7. Virology 1991;181:62–9. 45. Selvey LA, Dunn LA, Tindle RW, Park DS, Frazer IH. Human papillomavirus (HPV) type 18 E7 protein is a short-lived steroid-inducible phosphoprotein in HPV-transformed cell lines. J Gen Virol 1994;75:1647–53. 46. Borysiewicz LK, Fiander A, Nimako M, Man S, Wilkinson GWG, Westmoreland D, et al. A recombinant vaccinia virus encoding human papillomavirus types 16 and 18, E6 and E7 proteins as immunotherapy for cervical cancer. Lancet 1996; 347:1523–7. 47. Binns MM, Smith GL. Recombinant poxvirus. Boca Raton, Florida: ACR Press, 1993. 48. Fujimaru T. Virus as an aid to cancer therapy: selective method of more effective oncolytic bovine enterovirus. J Kurum Med Assoc 1978;41:15–32. 49. Meneguzzi G, Kieny MP, Lecocq JP, Chambon P, Cuzin F, Lathe R. Vaccinia recombinants expressing early bovine papillomavirus (BPV1) proteins: retardation of BPV1 tumor development. Vaccine 1990;8:199 –206. 50. Selvakumar R, Borenstein LA, Lin YL, Ahmed R, Wettstein FO. Immunization with nonstructural proteins E1 and E2 of cottontail rabbit papillomavirus stimulates regression of virus-induced papillomas. J Virol 1994;69:602–5.