Transgene expression in the rhesus cervix mediated by an adenovirus expressing β-galactosidase

Transgene expression in the rhesus cervix mediated by an adenovirus expressing [3-galactosidase Michele Follen Mitchell, MD, MS," Katsuyuki Hamada, MD, Phi), b K. Jagannadha Sastry, PhD, * Asis Sarkar, Phi), ~ Guillermo Tortolero-Luna, MD, PhD,"J. Taylor Wharton, MD," and Jack A. Roth, MDb Houston and Bastrop, Texas

OBJECTIVES: More than 90% of cervical cancers are positive for human papillomavirus, which functionally represses p53 and pRb. The remainder have been found to contain p53 mutations. Gene therapy involves insertion of a functioning gene into a patient to correct a genetic abnormality. STUDY DESIGN; The ability of a 13-galactosidaseadenovirus to mediate transgene expression in the rhesus cervix was evaluated. Three different doses and two different entry techniques of virus were investigated. RESULTS: The ideal dose determined by X-galactosidase staining was 2 x 10~~plaque-forming units, and the injection method yielded better staining than did abrasion with topical application. Increased adenoviral-specific immunoglobulin G antibody response in the injected monkeys confirmed the results. CONCLUSION: High transduction efficiency by use of adenoviral vectors can be achieved in the cervix. Reversing the effects of human papillomavirus and p53 mutations with gene therapy may become a novel therapy for invasive and preinvasive cervical cancer. (AM J O~STETGYNECOL1996;174:1094-101 .)

Key words: Rhesus monkey, gene therapy, adenoviral vector, animal model, gene therapy delivery system

Cervical cancer is the second most common malignancy in women worldwide; it has an overall 5-year survival rate of 40%. ~In spite of aggressive screening with the painless and relatively inexpensive Papanicolaou smear, cervical cancer remains an important health problem for women, especially underserved women. The most important risk factor is human papillomavirus infection (HPV), which increases the relative risk for cervical intraepithelial neoplasia and cervical cancer by 11 to 50 times that of the patient without HPV. HPV types 16 and 18 have been identified in the majority of cervical cancers.'-' Virus-encoded oncoproteins are important cellular regulatory proteins by which viruses can transform ceils.

From the Departments oJ Gynecologte Oncology," Thoracic and Cardio vascular Surgery/' and ~}terinary Sciences,' IJ~ziversi(v of Texas M.D. Anderson Cancer Center. Supported in part by' the National Cancer Institute and National Institutes o] Health grant No. RO1 CA45187 ([A.R.), National Cancer Institute Training Grant No. CA 09611 ([.A.R.), the Division of Surgery from 7~nneco and Fxxon for the Core Laboratory Facility, M. D. Anderson Cancer Center Support Grant No. CA 16672, the Mathers Foundation, the University Cancer bbundation, the Business and Professional 14bmen's Fund, and a sponsored research agreement with Introgen Therapeutics. Charles Hunter Award, presented at the Fourteenth Annual Meeting of the American Gynecological and Obstetrical Society, Napa, California, September 7-9, 1995. Reprint requests: Michele Follen Mitchell, MD, MS, University of Texas M. D. Anderson Cancer Centeg, Department of Gynecologic Oncology, Box 67, 1515 Holcombe, Houston, TX 77030. Copyright 9 1996 by Mosby-Year Book, Inc. 0002-9378/96 $5.00+ 0 6/6/71144

1094

HPV 16 and 18 E6 and E7 proteins have been demonstrated to be necessary and sufficient for malignant transformation of mammalian cells. E7 forms complexes with the retinoblastoma protein (pRB) and E6 forms complexes with p53. Howley et al? have demonstrated that the functional inactivation ofpRB and p53 are important steps in cervical carcinogenesis. Cervical cancer serves as a model of vitally induced carcinogenesis and seems an ideal model for genetic intervention. Gene therapy refers to the insertion of a functioning gene into the cells of a patient to correct an inborn error of metabolism or to restore function to the cell lost by gene inactivation.4 In vitro gene transfer, although highly efficient, is suitable for tissues that can be removed from the body, genetically altered, and returned to the body (hematopoietic cells). In vivo gene transfer, although less efficient, is suitable for tissues that cannot be removed. Several methods of gene transfer exist, including direct gene transfer, receptor-mediated gene delivery, and recombinant virus vectors. 4 Adenoviruses are common deoxyribonucleic acid (DNA) viruses that produce infections of the upper respiratory tract. To improve their safety, they have been rendered replication deficient by deletion of the E1A and E1B regulatory genes. Adenoviruses have several advantages over other recombinant virus vectors. They, unlike retroviruses, deliver genes efficiently regardless of the proliferative state of the tissue and they can he produced at high titers. They can be delivered by aerosol or injec-

Volume 174, Number 4 AmJ ObstetGynecol

Mitchell et al. 1095

Table I. Characteristics of monkeys

Animal No. L571 L741 L577 L345 L87 J9 L163 L879 J47 L745

[

Date of birth

Weight (kg)

4-24-85 8-24-86 5-7-85 7-9-81 8-4-78 5-10-90 4-26-79 4-23-88 3-24-90 3-22-87

8.6 6.6 5.5 7.8 8.6 5.2 6.9 7.0 7.5 5.8

I

Progeny No. 2 0 1 5 10 2 9 2 1 2

tion and are not associated with malignant transformation. Adenoviral vectors, however, infect both dividing and nondividing cells, do not integrate, and may induce a specific antiviral immune response?" 5 Previous studies have shown that Ad5-CMV-p53 can suppress the growth of human lung, head and neck, and cervical cancer cell lines in vitro and in vivo.~-13As preparation for a clinical trial in patients with cervical cancer, a study of transgene expression in the cervix with the adenovirus vector expressing [3-galactosidase (Ad5CMV-~gal) was performed in the rhesus monkey (Macaca mulatta). The rhesus monkey cervix has a similar histopathologic appearance to that of the human and is approximately one third the size of the human cervix (Fig. 1). The ~galactosidase gene is not found endogenously in the rhesus, nor in man, and by use of Xgalactosidase staining can thus be used as a marker of cells infected by the virus. Three different doses of virus and two different viral entry techniques were investigated. Demonstrating high-transduction efficiency in the cervix with adenoviral vectors is prerequisite for a clinical trial in humans in which p53 or antisense-HPV genes could be inserted into the adenoviral vector.

Material and m e t h o d s Recombinant adenovirus preparation. The recombinant adenovirus (Ad5CMV-[~-gal) contains the cytomegalovirus promoter, SV40 polyadenylation signal, and ~-galactosidase (LacZ) gene in a minigene cassette inserted into the El-deleted region of modified Ad5. ~Viral stocks were propagated in 293 cells, which are derived from primary embryonal kidney cells transformed by introducing sheared fragments of Ad5 DNA. This cell line contains Ela and is thus highly permissive for the replication of the E1 replication-deficient adenovirus. Cells were harvested 36 to 40 hours after infection, pelleted, resuspended on phosphate-buffered saline solution, and lysed; cell debris was removed by subjecting the cells to cesium chloride gradient purification. Concentrated virus was dialyzed, aliquoted, and stored at -80 ~ C. The viral doses were determined by plaque assays. ~4 Animal study. After review and approval by both the

Fig. 1. Rhesus monkey cervix showing epithelium and stroma. (Original magnification xl0.)

Animal Care and Use Committee and the Biosafety Committees at the University of Texas M.D. Anderson Cancer Center, 10 reproductive-competent rhesus monkeys (aged 5 to 17 years) were selected (Table I). To assess the health status of each monkey, hematologic and serum biochemistry panels were performed to establish baseline values before the animals were inoculated. Five animals each were to undergo two planned techniques of viral entry: injection of Ad5CMV-~-gal and abrasion with topical application of Ad5CMV-[5-gal at three doses in 150 pl of 10 m m o l / L Tris-hydrochloric acid, 2 m m o l / L magnesium chloride, pH 7.5, 10% glycerol solution (one monkey at 2 x 10s pfu, two monkeys at 2 x 1 0 9 pfu, and two monkeys at 2 x 10 '~ pfu for each viral entry technique). Those doses have been demonstrated to be safe in mice and rats and have mediated tumor regression. H The rhesus monkeys were anesthetized by ketamine hydrochloride 1 m g / k g given intramuscularly for all procedures. The first five monkeys underwent injection with a tuberculin syringe of the cervix with Ad5CMV-~-gal in two divided doses in the anterior and posterior cervix. The second five monkeys underwent abrasion of the cervix with sandpaper until punctate bleeding points were achieved; Ad5CMV-~-gal was then placed on Gelfoam against the abraded cervix. A Gelfoam tampon was placed against the cervix in the vagina. Good application against the cervix was verified with visual inspection before the animals were awakened. All animals were singly caged and fed commercial 20% protein monkey diet for the duration of the experiment. Postoperatively all animals were given butorphanol (0.1 mg/kg) twice a day for analgesia. The monkeys were carefully monitored for 3 days for side effects; none were noted. Three days later all 10 monkeys underwent loop electrosurgical excision cone biopsy and cautery of the cervix. All the monkeys were monitored carefully for bleeding. One monkey returned to the operating room 5 minutes after the procedure for

1096

Mitchellet al.

April 1996 Am.] ()bstet Gvnecol

Fig. 2. A, Rhesus monkey cervix after injection: ~-galactosidasc stain is present in stroma. (Original magnification • B, Rhesus monkey cervix after abrasion: ~-galactosidase stain is not present in epithelium or stroma. (Original magnification •

additional cautery, which p r o c e e d e d uneventfully. All animals were given trimethoprim (20 m g / k g ) and sulfamethoxazole (100 m g / k g ) suspension orally twice a day for 7 days prophylactically. Postoperatively all animals were given butorphanol (0.1 m g / k g ) twice a day tbr analgesia. MI monkeys were carefully m o n i t o r e d for complications. All monkeys r e t u r n e d to the operating r o o m 22 days later for inspection. The cone specimens were carefully inked and marked with suture for orientation and sent for frozen section studies. Hematoxylin and eosin stains and X-galactosidase staining were later performed. Blood was drawn and vaginal secretions were stored at baseline, 3 days after exposure, and 22 days after exposure for enzyme-linked i m m u n o s o r b e n t assay to study adenovirus-specific irnmunoglobulin G (IgG) and i m m u n o g l o b u l i n A (IgA). X-galactosidase staining. X-galactosidase uses ~-galactosidase as a substrate, turning cells infected with virus a bright blue color in the tissue. Five micron frozen sections of rhesus cervix were fixed with 1.95% glutaraldehyde in phosphate-buffered saline solution on ice for 10 minutes and then stained overnight with X-galactosidase (Strategene, La Jolla, Calif.) as described by MacGregor et al/:'

Detection of antiadenovirus-specific IgG and IgA. Serum samples were collected from the rhesus monkeys before and after they were exposed to different amounts of the recombinant adenovirus [~-galactosidase. A standard enzyme-linked i m m u n o s o r b e n t assay procedure was followed to detect adenovirus-specific lgG in these serum samples by use of a cell-flee extract prepared by sonification and centrifugation of cell pellet from 6 x 107 293 cells transfected with Ad5CMV-[3-gal (at a multiplicity of infection of 10, 3 days' infection) in a total volume of 3 ml

of normal saline solution. The control antigen was prepared by processing 6 • 10 r untransfected cells. A 1 : 10 dilution of the antigen was coated onto individual wells of 96-well round-bottonl microtiter plates b} incuhating overnight at 37 ~ C. A 5% solution of nont:at dry milk was used as a blocking reagent, and each sample of rhesus monkey sernm was used at a final dilution of 1:20. The second antibody, peroxidase-conjugated goat antihuman IgG, and the substrate c~phenylenediamine dihrochloride were obtained fiom Sigma I m m u n o Chemicals (St. I.ouis) and were used according to the manufacturer's specifications. Detection of antiadenovirus reactMty was based on readings at an optical densiw of 495 nm (OD 495), and the values obtained with control cell extract were subtracted from those with adenovirus-transfected cell extract to obtain the adenovirus-specific reactMtv. Similar procedures were carried out for detection of lgA on the vaginal secretions (dilution 1:5) and serum samples (dilnted 1:20). Statistical analysis. The SPSS program (version 6.0) (SPSS, Chicago) was used for statistical analysis. The mean OD 495 values tor IgG measurements in the five injected monkeys and in the five abraded monkeys and their confidence intervals were calculated. The mean difference in OD 495 values for IgG measurements in the five injected monkeys and in the five abraded monkeys and their confidence intervals were calculated, baseline values were compared with a zero value, 3-day values were c o m p a r e d with baseline, and 22-day values were c o m p a r e d with g-day values. Paired t tests were used to assess the statistical significance of baseline values c o m p a r e d with 3- and 22-day values within each of the two groups (injected and abraded) and to compare baseline values, g-day values, and 22-day values between injected and abraded gronps.

Volume 174, Number 4 AmJ Obstet Gynecol

Mitchell et al.

1097

Fig. 3. A, Rhesus cervix after injection with 2 • 10~pfu; []-galactosidase stain is present in stroma staining roughly one third of cells. (Original magnification • B, Rhesus cervix after injection with 2 x lff' pfu; ~galactosidase stain is present in stroma staining roughly two thirds of cells. (Original magnification • C, Rhesus cervix after injection with 2 • 10 I~pfn; ~-galactosidase stain is present in stroma staining roughly all cells. (Original magnification •

Results Monkeys. All the monkeys were carefully m o n i t o r e d for complications. Persistent spotting was noted after loop electrosurgical excision procedure in one monkey, the oldest monkey. O n e m o n t h later the cervices were completely reepithelialized in 8 of 10 monkeys and partially epithelialized in 2 monkeys: the 15-year-old monkey that required additional cautery and the older (17 years old) m o n k e y that spotted. No infections, stenoses, vaginal thermal injuries, fistulas, or other deformities were noted. X-galaetosidase staining for 13-galactosidase. [3-galactosidase staining was noted in the monkeys that underwent injection a n d n o t in those that u n d e r w e n t abrasion a n d topical application (Fig. 2). The cervical epithelium is an effective barrier to adenovirus. Additionally, progressive increases in staining were noted with increasing plaque-forming units of virus (Fig. 3). In the one monkey, injected with 2 • 10 ~ pfu approximately one third of cells were stained, in the two monkeys injected with 2 x 10 ~ pfu approximately two thirds of cell were stained, a n d in the two monkeys injected with 2 x 10 ~'' pfu all cells were stained. Antiadenovirtts-specific IgG and IgA. The levels of an-

tiadenovirus-specific IgG in rhesus monkeys before a n d after exposure to Ad5CMV-[3 gal are shown in Fig. 4. Of note, all 10 monkeys had some baseline IgG present, probably from their contact with h u m a n s who care for them. Surprisingly, both groups of monkeys demonstrate an increase in IgG in spite of the negative 13-galactosidase staining in the monkeys that u n d e r w e n t abrasion and topical application. These responses indicate a systemic response to the virus in spite of the local injection or application. The five monkeys that u n d e r w e n t injection had significantly higher increases in IgG than did those with abrasion 22 days after exposure. Fig. 4 shows OD 495 values at the time of the three measurements. In Table II m e a n OD 495 values a n d their confidence intervals are reported. In Table III m e a n differences and their confidence intervals are presented. Statistically significant differences are noted in the injected monkeys' IgG values between baseline and day 3 values and between day 3 and day 22 values. These results indicate that injection will result in higher levels of gene expression than do abrasion a n d topical application; however, specific i m m u n o logic responses can be elicited with both techniques. No serum or vaginal IgA to adenovirus was noted in the baseline, day 3, or day 22 specimens. Fig. 5 compares

1098

April ]996

Mitchell et al.

.~n J Obstet Gynecol

0.4

0.3

84

[]

Pre-Exposure

[]

3d Post-Exposure

[]

22d Post-Exposure

* N

9,el"

.< ,<

0.2

0.1

0

==. % % x el

~< el

E

"

x r

-L= o= )< el

x el

l++li 0-= Oil

et

i

Injected

o

g t~1

el

=

Abraded

Monkey # and PFU Dose Fig. 4. Blood samples collected from monkeys at different dates after injection with recombinant adenovirus were subjected to enzyme-linked immunosorbent assay for determining adenovirus-specific cell extracts from adenovirus-transfected cells and are adjusted to control values obtained with control cell extract as antigen.

Table II. Means and 95% confidence intervals in OD values for adenovirus-specific IgG antibodies at baseline, 3 days, and 22 days by viral entry technique Time interval

Injected mean value and 95% confidence interval

Abraded mean value and 95% confidence interval

Baseline Day 3 Day 22

0.0298 (0.0023-0.0573) 0.0526 (0.0303-0.0749) 0.2204 (0.1095-0.3313)

0.0476 (-0.0159-0.1111) 0.0734 (0.0300-0.1168) 0.1014 (0.0255-0.1773)

mean values between the injected and abraded monkeys at each point in time and demonstrates significantly differences between groups for the day 22 values. Comment

In spite of the availability of the screening Papanicolaou smear, cervical cancer remains an important health p r o b l e m in women, especially in underserved women. A total of 437,300 w o m e n were diagnosed with cervical cancer in 1985, making it the second most c o m m o n malignancy in w o m e n worldwide? After declining rates for the last several decades in the United States and other countries in which screening programs have been established, incidence rates are currently increasing in U.S. women.~ The Surveillance, Epidemiology and End Results (SEER) database, a 10% population-based sample of U.S. cancer incidence, estimates in 1995 that 15,800 w o m e n will be diagnosed with invasive cervical cancer, an increase from 15,000 in 1994 and 13,500 in 1993. Similarly, 4800 w o m e n will die from invasive cervical cancer, an

increase from 4600 in 1994 and 4400 in 1993. Similar trends are n o t e d for the preinvasive counterpart: 65,000 w o m e n will be diagnosed with carcinoma in situ, an increase from 55,000 in 1994 and 50,000 in 1992. 2 The reasons for this sudden increase are unknown and must be viewed cautiously; some increase may be due to increases in population, increases in HPV infection in the cervix, and some may be due to h u m a n immunodeficiency virus, which increases the risk of cervical intraepithelial neoplasia and cervical cancer. An estimated 5,000,000 w o m e n will have abnormal Papanicolaou smears in the United States in 1994, of which half will have some degree of dysplastic lesion and the other half will have atypia. ~ The most important epidemiologic risk factor for cervical cancer is h u m a n papillomavirus. This association has recently been found to be consistent i n d e p e n d e n t of HPV assay m e t h o d and epidemiologic study design.l~ The most c o m m o n types of HPV are those classified as highrisk types (HPV 16, 18, 45, and 56), the intermediate-risk types (HPV 31, 33, 35, 51, 52, and 58) and the low-risk types (HPV 6 and 11).2 These high- and intermediate-risk types have recently been identified in 77% of high-grade cervical intraepithelial neoplasias or squamous intraepithelial lesions and in 84% of invasive lesions. Epidemiologic prospective cohort studies demonstrate increases in the relative risk for the development of high-grade cervical intraepithelial neoplasia ranging from 11 to 60 and for the d e v e l o p m e n t ofinvasive cancer 15 to 50 times that of controls without HPV infection. ~ HPV DNA, particularly from HPV 16 or 18, has been

Volume 174, Number 4 AmJ ObstetGynecol

Mitchell et al.

1099

0.25 0.2204

0.2

~.

0.15

P=.04

a "5

0.1014 ,r a 0

0.1 P=.27 P=.50 0.0476

0.05

0.0734

0.0526

0.0298

OD Baseline

OD at 22 days

OD at 3 days

Fig. 5. Comparison of mean OD 495 for adenovirus-specific IgG antibodies by viral entry technique at baseline, 3 days, and 22 days.

Table III. Mean differences and 95% confidence intervals in OD values for adenovirus-specific IgG antibodies at baseline, 3 days, and 22 days by viral entry technique [

Time interval

Injected mean difference and 95 % confidence interval

Significance

Abraded mean difference and 95 % confidence interval

Baseline vs day 3 Baseline vs day 22 Day 3 vs day 22

-0.0228 (-0,037-0.008) -0.1906 (-0,0301-0.080) -0.1678 (-0.282-0.54)

p = 0.012 p = 0.009 p = 0.015

-0.0258 (-0.063-0.011) -0.0538 (-0.127-0.19) -0.0280 (-0.098-0.42)

identified in close to 90% of invasive carcinomas of the uterine cervix and has been found to contribute in a significant way to the genesis of human genital cancer? Viral-encoded oncoproteins interacting with important cellular regulatory proteins is a mechanism by which DNA viruses can transform cells. HPV 16 and 18 encode two transforming genes or oncoproteins: E6 and E7. The E6 and E7 genes of HPV 16 are frequently coexpressed and are the most abundant viral transcripts in biopsy specimens from HPV 16-positive cervical carcinomasJ 7 There is a strong evidence that expression of both E6 and E7 open reading frames are necessary and sufficient for efficient malignant transformation of a variety of mammalian cells. Furthermore, the continued expression of the E6 and E7 region of the viral genome appears to be necessary for the maintenance of malignant phenotype, ~* E7 shares functional and structural features with the adenovirus E1A proteins. Like adenovirus E1A and the large T proteins of the polyomaviruses, E7 can complex with pRB. The E7 proteins of the "high-risk" HPVs associate with pRB with approximately 10-fold higher affinity than "low-risk" HPVs; this is thought to be due to aminoterminal sequences that include the pRB binding do-

J J

Significance p = 0.123 p = 0.110 p = 0.327

main. E6 oncoprotein shares functional and structural features with SV40 large T and adenovirus 5 EIB and can form complexes with p53. In vitro E6 promotes degradation of p53 by the ubiquitin-dependent protease system. This selective degradation of negative regulatory proteins provides a novel mechanism for dominant-acting oncoproteins. Howley et al. '~ have demonstrated the relevance of the inactivation of pRB and p53 in human cervical carcinogenesis through studies with HPV- and HPVonegative cervical cancer cell line experiments, in which HPVpositive cell lines expressed normal pRB and low levels of wild-type p53, whereas HPV-negative cell lines demonstrated mutations in pRB and p53. Also, studies by Howley" suggest that inactivation of the normal functions of the tumor-suppressor genes pRB and p53 are important steps in cervical carcinogenesis. p53 tumor suppressor gene mutation and deletions are believed to be responsible for >50% of h u m a n cancers, z~ Cervical cancer is the exception. HPV E6 has been shown to functionally repress the action of wild-type p53, and thus p53 mutations are unnecessary. ~ Studies of cervical cancer show that most cervical tumors (95%) are HPV positive and do not overexpress p53. z~ Similar findings

1100 Mitchellet al.

exist for cervical intraepithelial neoplasia lesions. Those tumors that are HPV negative are more likely to have p53 mutations but not exclusively so. Cervical cancer, whether HPV or p53-mutation driven, should be an excellent candidate for gene therapy with a p53 expression vector. A replication-defective and helper-independent recombinant p53 adenovirus generated by Zhang et al. ~ carries an expression cassette that contains human cytomegalovirus E1 promoter, human wild-type p53 complementary DNA, and SV40 early polyadenylation signal. Four human non-small-cell lung cancer cell lines representing differences in p53 configuration were used to evaluate the Ad5CMV-p53 virus in one study. In the H358 cell line, which has a homozygous deletion of p53, the p53 gene was transferred with 97% to 100% efficienc}, as detected by immunohistochemical analysis, when the cells were infected with Ad5CMV-p53 at a multiplicity of infection of 30 to 50 pfu per cell. Western blot analysis showed that the p53 protein was expressed at a high level. Growth of the AdSCMV-p53 virus-infected H358 cells was inhibited 79%, whereas that of noninfected cells or the cells infected with the control virus was not inhibited. Growth of cell line H322 which has a point mutation in p53, was inhibited 72% by Ad5CMV-p53, whereas that of cell line H460 containing wild-type p53 was less affected (28% inhibition). Tests in nude mice demonstrated that tumorigenicity of the Ad5CMVp53 treated H358 cells was greatly inhibited. In a mouse model of orthotropic human lung cancer, the tumorigenic H226Br cells, with a point mutation in p53, were inoculated intratracheally 3 days before the virus treatment. Intratracheal instillation of Ad5CMV-p53 prevented tumor formation. These resuits suggest that adenovirus is an efficient vector for mediating transfer and expression of tumor-suppressor genes in human cancer cells. 7 Cisplatin has been demonstrated to be synergistic in this model. ~ Similar observations were noted for head and neck cancer cell lines. Mutations of the p53 gene constitute one of the most frequent genetic alterations in squamous cell carcinoma of the head and neckJ' In this study, wildtype p53 was introduced into two separate cell lines by a recombinant adenoviral vector, Ad5CMVp53. Northern blotting showed that, after infection by Ad5CMV-p53, cells produced up to 10-fold higher levels of exogenous p53 messenger ribonucleic acid than did cells treated with vector only (without p53). Western blotting showed that the increased levels of p53 protein produced in the AdSCMV-p53-infected cells were a reflection of p53 messenger ribonucleic acid expression. In vitro growth assays revealed growth arrest and cell morphologic changes consistent with apoptosis after Ad5CMV-p53 intiection. In vivo studies in nude mice with established subcutaneous squamous cell carcinoma nodules showed that tumor volumes were significantly reduced in mice that received peritumoral infiltration of Ad5CMV-p53. These data suggest that Ad5ClVI~r-p53 may be further developed as a potential novel therapeutic agent for squamous cell car-

April 1996 AmJ Obstet Gynecol

cinoma of the head and neck because introduction of wild-type p53 into these cell lines attenuates their replication and tumor growth." In preparation for a clinical trial in the treatment of lung cancer, a study was undertaken to look for potential adverse effects of Ad5CMV-p53. ~1 In vitro studies included identifying a lack of infectious replication in HeLa and lung cancer cell lines infected with Ad5CMV-p53 and demonstrating that normal human bronchial epithelial cell growth was not affected by Ad5CMV-p53 infection. The virus was then given by intratracheal injection to four groups of mice. In none of the mice was the injection lethal, causing only peribronchiolar and perivascular inflammation. ~~No evidence of systemic toxicity has been noted with treatments of Ad5CMV p53 (exceeding 10 '~ pfu) injected subcutaneously, intraperitoneally, or by tail vein. ' Hamada et al. '~ ~3have demonstrated similar results in cervical cancer cell lines. Eight human cervical carcinoma cell lines--C33A, HT3, SiHa, CaSki, HeLa, C4I, MS751, and ME180--were used in this study. Two cell lines, C33A and HT-3, have p53 mutations and are HPV negative. The remaining six cell lines are infected with HPV 16 or i8 and do not have p53 mutations. The cell lines were chosen because most cervical cancers are believed to inactivate p53 by the HPV E6-mediated mechanism rather than by mutation. High transduction efficiency of AdSCMV-p53 was demonstrated, as was growth inhibition. Ex vivo and in vivo nude mouse models have similarly demonstrated apoptosis of cervical tumors) ~' ~:~ Animal studies are crucial for development of therapeutic and prophylactic reagents against human malignancies. Rhesus monkeys are a commonly adopted nonhuman primate animal model because of their close relatedness to humans. In particular, the rhesus monkey provides an excellent model for the human female genital tract. A rhesus papillomavirus that infects the rhesus cervix was identified and it has been demonstrated that rhesus monkeys have cervical dysplasia similar in histologic features to cervical intraepithelial neoplasia in humans. ~'-'We were able to visualize the rhesus cervix with a pediatric speculum. The surface area was large enough to perform injection and cone biopsy without complication. Although abrasion and topical application caused a specific immune response, no positive ~galactosidase staining could be detected in those cervixes. Injection seems a viable technique for viral entry in cervix. The dose of 2 x 10'" pfu caused viral infection of every cell and would thus be the ideal dose tor a trial. The issue of humoral immunity to recombinant viral vectors is of great importance. We observed significant increases in the level of antiadenovirus-specific IgG, but not IgA, after exposure to Aad5CMV-~gal. Studies of recombinant adenoviruses carrying the cystic fibrosis gene have suggested that antiadenovirus-specific cytotoxic T lymphocytes and neutralizing antibodies are preexisting or induced by primary injection and may block

Volume 174, Number 4 AmJ Obstet Gynecol

subsequent adenovirus-mediated g e n e transfer. 23 Although in o t h e r studies g e n e expression of not affected after repetitive administration, 24 the use of i m m u n o s u p pressive medications may allow repeat administration o f vector with i m p r o v e d tolerance. 25 In summary, HPV-negative and p53-mutation-induced cervical cancers provide an exciting opportunity for g e n e therapy with wild-type p53 or HPV-antisense genes. Ham a d a et al. have d e m o n s t r a t e d that Ad5CMV-p53 suppresses growth in cervical cancer cell lines and in ex vivo and in vivo cervical cancer models in the n u d e mouse. In this e x p e r i m e n t a preclinical trial of the ~-galactosidase vector Ad5CMV-~gal was p e r f o r m e d in the rhesus monkey, in which three different doses of virus and two different viral entry techniques were investigated. T h e ideal dose d e t e r m i n e d by ]3-galactosidase staining was 2 x 101~ pfu, and the injection m e t h o d yielded better results than abrasion and topical application; these results were verified by observing increased adenoviral-specific IgG antibody response in those monkeys injected over those that u n d e r w e n t abrasion and topical application. This is the first study d e m o n s t r a t i n g that high transduction efficiency with adenoviral vectors can be achieved in the cervix. Reversing the effects of HPV and p53 mutations with g e n e therapy may b e c o m e a novel therapy for invasive and preinvasive cervical cancer. We thank Wei Wei Zhang and R a m o n Alemany for the vector; J u d y Sandella, RN, NP, MSN, and Karen Rabel, RN, for their assistance with the procedures; Drs. Stephanie Buchl and William C. Satterfield, Mary Stevens, and Mary H e m p h i l l for assistance with animal care; Gerald Costello for assistance with tissue sectioning and staining; and David M. Gershenson, MD, Gary L. Clayman, MD, Albert B. Deisseroth, MD, PhD, and Sunita Patterson for helpful c o m m e n t s with the manuscript. REFERENCES

1. Parkin DM, Pisani P, Ferlay J. Estimates of the worldwide incidence of eighteen major cancers in 1985. Int J Cancer 1993;54:594-606. 2. Mithell ME Hittelman WN, Lotan R, et al. Chernoprevention trials and surrogate endpoint biomarkers in the cervix. Cancer Suppl 1995;76:1956-77. 3. Howley PM, Scheffner M, Huibregtse J, Munger K. Oncoproteins encoded by the cancer-associated human papillomaviruses target the products of the retinoblastoma and p53 tumor suppressor genes. Cold Spring Harb Symp Quant Biol 1991 ;56:149-55. 4. Culver KW. Gene therapy: a handbook for physicians. New York: Mary Ann Liebert Publications, 1994:15-21. 5. Graham FL. Transformation by oncogenicity of human adenoviruses. In: Ginsberg HS, ed. The adenoviruses. New York: Plenum Press, 1984:354-62. 6. Zhang WW, Fang X, Branch CD, Mazur W, French BA, Roth JA. Generation and identification of recombinant adenovirus by liposomemediated transfection and PCR analysis. Bittechniques 1993;15:868-72. 7. Zhang WW, Fang X, Mazur W, French BA, Georges RN, Roth JA. High-efficiency gene transfer and high-level expression of wild-type p53 in human lung cancer cells mediated by recombinant adenovims. Cancer Gene Ther 1994;1:5-13.

Mitchell et al.

1101

8. Fujiwara T, Grimm EA, Mukhopadhyay T, Zhang WW, OwenSchaub LB, Roth JA. Induction of chemosensitivity in human lung cancer cells in vivo by adenovirusmediated transfer of the wild-type p53 gene. Cancer Res 1994;54:2287-91. 9. Liu TJ, Zhang WW, Taylor DT, Roth JA, Goepfert H, Clayman GL. Growth suppression of human head and neck cancer cells by the introduction of a wild-type p53 gene via a recombinant adenovirus. Cancer Res 1994;54:3662-7. 10. Zhang WW, Alemany R, Wang J, Koch PE, Ordonez NG, Roth JA. Safety evaluation of Ad5CMV-p53 in vitro and in vivo. Hum Gene Ther 1995;6:155-64. 11. Clayman GL, E1-Naggar AK, RothJA, et al. In vivo molecular therapy with p53 adenovirus for microscopic residual head and neck squamous carcinoma. Cancer Res 1995;55:1-6. 12. Hamada K, Zhang W, Alemany R, Roth JA, Wolf J, Mitchell ME Growth inhibition of human cervical cancer cells by the recombinant adenovirus p53 [Abstract 2609]. In: Proceedings of the eighty-sixth annual meeting of the American Association for Cancer Research, Toronto, Ontario, Canada, March 18-22, 1995. Toronto: American Association for Cancer Research, 1995. 13. Hamada K, Zhang WW, Alemany R, RothJA, WolfJ, Mitchell ME Gene therapy of cervical cancer by adenovirus-medicated p53 gene transfer. Keystone Symp J Cell Biochem Suppl 1995;21A:421. 14. Graham FL, Prevec L. Manipulation of adenovirus vectors. In: Murray EJ, ed. Methods in molecular biology, gene transfer and expression protocols. Clifton, New Jersey: Humana Press, 1991:109-28. 15. MacGregor GR, Mogg AE, Burke JE Casky CT. Histochemical staining of clonal mammalian cell lines expressing E. coli beta galactosidase indicates heterogeneous expression of the bacterial gene. Somat Cell Mol Genet 1987;13:253-65. 16. Schiffman MH, Bauer HM, Hoover RN, et al. Epidemiologic evidence showing that human papillomavirus infection causes most cervical intraepithelial neoplasia. J Natl Cancer Inst 1993;85:958-64. 17. Seedorf K, Oltersdorf T, Krammer G, et al. Identification of early proteins of the human papillomaviruses type 16 (HPV 16) and type 18 (HPV 18) in cervical carcinoma cells. EMBO J 1987;6:139-44. 18. Munger K, Phelps WC, Bubb V, Howley PM, Schlegel R. The E6 and E7 genes of the human papillomavirus type 16 together are necessary and sufficient for transformation of primary human keratinocytes. J Virol 1989;63:4417-21. 19. Howley PM, Munger K, Romanczuk H, Scheffner M, HuibregtseJM. Cellular targets of the oncoproteins encoded by the cancer associated human papillomaviruses. In: Harris CC, Hirohashi S, Ito N, Pitot C, eds. Multistage carcinogenesis. Boca Raton: CRC Press, 1992:239-48. 20. Hollstein M, Sidransky D, Vogelstein B, Harris CC. p53 mutations in human cancers. Science 1991;253:49-53. 21. RothJA. Recombinant DNA advisory committee: minutes of meeting of September 14-15, 1992. Hum Gene Ther 1993;4: 365-89. 22. Ostrow RS, McGlennen RC, Shaver MK, Kloster BE, Houser D, Faras AJ. A rhesus monkey model for sexual transmission of a papillomavirus isolated from a squamous cell carcinoma. Proc Natl Acad Sci U S A 1990;87:8170-4. 23. Yang Y, Li O~ Ertl HC, Wilson JM. Cellular and humoral responses to viral antigens create barriers to lung-directed gene therapy with recombinant adenoviruses. J Virol 1995; 69:2004-15. 24. ZabnerJ, Petersen DM, Puga AP, et al. Safety and efficacy of repetitive adenovirus-mediated transfer of CFTR cDNA to airway epithelia of primates and cotton rates. Nature Genet 1994;6:75-83. 25. Dai Y, Schwarz EM, Gu D, Zhang W-W, Sarvetnick N, Verma IM. Cellular and hormonal immune responses to adenoviral vectors containing factor IX gene: tolerization of factor IX and vector antigens allows for long-term expression. Proc Natl Acad Sci U S A 1995;92:1401-5.