Pronounced Antitumor Effects and Tumor Radiosensitization of Double Suicide Gene Therapy1

The efficacy of HSV-1 thymidine kinase (TK) and Escherichia coil cytosine deaminase (CD) suicide gene therapies as cancer treatments are currently being examined in humans. We demonstrated previously that compared to single suicide gene therapy, greater levels of targeted cytotoxicity and radiosensitization can be achieved in vitro by genetically modifying tumor cells to express CD and HSV-1 TK concomitantly, as a fusion protein. In the present study, the efficacy of the combined double suicide gene therapy/radiotherapy approach was examined in vivo. Nude mice were injected either s.c. or i.m. with 9L gliosarcoma cells expressing an E. coil CD/HSV-1 TK fusion gene. Double suicide gene therapy using 5-fluorocytosine (500 mg/kg) and gancicbovir (30 mg/kg) proved to be markedly better at delaying tumor growth and achieving a tumor cure than single suicide gene therapy, which used 5-fluorocytosine or gancicbovir administered independently. Importantly, double suicide gene therapy was highly effective against barge experimental tumors (>2 cm3), reducing tumor volume an average of 99% and producing a 40% tumor cure. Moreover, double suicide gene therapy profoundly potentiated the antitumor effects of radiation. The results indicate that double suicide gene therapy, particularly when coupled with radiotherapy, may represent a highly effective means of eradicating tumors. INTRODUCTION A major impediment that limits the effectiveness of conventional cancer therapy is the lack of a significant differential response between malignant and normal tissue. Suicide gene therapy provides a means of circumventing this limitation through the selective introduction of genes encoding nonmamReceived 5/2/97; revised 7/1 8/97: accepted 7/24/97. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. I This work was supported by NIH Grant CA64323. 2 To whom requests for reprints should be addressed, at Molecular Biology Research, One Ford Place, Wing SD, Henry Ford Health System, Detroit, MI 48202-4689. Phone: (3 13) 876-195 1 ; Fax: (313) 876-1950. malian metabolic enzymes into tumor cells, rendering them susceptible to specific antimetabolites. Unlike mammalian TK,3 HSV-l TK phosphorylates a variety of nucleoside analogues, such as GCV, to their nucleoside monophosphate derivatives (1). Once phosphorylated, these anabogues are converted by endogenous cellular kinases into their corresponding nucleoside triphosphates, which inhibit DNA replication by chain termination (2). CD, an enzyme present in a variety of fungi and bacteria, can deaminate the innocuous prodrug S-FC, forming 5-FU, a widely used chemotherapeutic agent (3). The cytotoxic effect of 5-FU is largely exerted following its conversion to 5-fluoro-dUMP, which irreversibly inhibits thymidylate synthase, resulting in the inhibition of DNA synthesis (3). The potential of the HSV-l TK/GCV and Escherichia coli CD/S-FC enzyme/prodrug systems as cancer therapies has been demonstrated extensively in animal models (4-19). In most cases, marked antitumor effects have also been observed following in vivo gene transfer using both recombinant retroviruses and adenoviruses coupled with prodrug (GCV or 5-FC) therapy. These encouraging results have paved the way for the numerous ongoing clinical trials that will determine the efficacy of HSV1 TK/GCV and CD/5-FC therapies in humans. Despite the established efficacy of these therapies in animal models, however, limitations have been demonstrated. !n many studies, the HSV-1 TK/GCV (8, 17-19) or CD/5-FC (9, 11, 13, 14) enzyme/prodrug combination failed to result in complete tumor regression, and frequently, many tumors recurred once prodrug therapy ceased. Moreover, all of these studies began prodrug therapy when the tumors were relatively small ( 200 mm3), and the ability of suicide gene therapy to control large tumors has not been investigated. Thus, it is likely that significant improvements in the conventional suicide gene therapy approach will be needed before this modality will have any value in the clinic. Toward this end, we have been developing protocols that might improve the efficacy of the conventional suicide gene therapy approach. Because the cytotoxic effects of the CD/S-FC and HSV-l TK/GCV combinations are mediated through different mechanisms, we hypothesized that combining these two suicide gene systems might result in enhanced cell killing. Indeed, there is a biochemical basis for the fact that these two suicide gene systems, when combined, should result in a synergistic, and not an additive, enhancement of cytotoxicity. As stated previously, the cytotoxic effect of the CD/5-FC combination is mediated predominantly through the inhibition of thymidylate synthase by 5-fluoro-dUMP (3). This results in a 3 The abbreviations used are: TK, thymidinc kinase; GCV, ganciclovir; CD, cytosine deaminase; 5-FC. 5-fluorocytosine; S-EU, 5-fluorouracil: HSV, herpes simplex virus. Research. on October 31, 2017. © 1997 American Association for Cancer clincancerres.aacrjournals.org Downloaded from 2082 Antitumor Effects of Double Suicide Gene Therapy redistribution of cells in early S phase (20), as well as DNA strand breakage due to the disruption of deoxynucleotide pools (2 1 ). The cytotoxicity of the HSV1 TK/GCV system results from GCV-monophosphate inhibition of DNA polymerases, which leads to a disruption of DNA synthesis (2). It was therefore reasoned that when combined, the CD/5-FC and HSV-l TK/GCV systems may synergize, because CDIHSV-1 TK-expressing cells would accumulate in early S phase, at which point they are most sensitive to GCV. GCV might also inhibit the repair of DNA strand breaks caused by S-FC treatment. Moreover, because 5-FU (the product of the CD reaction) and certain antiviral agents are proven radiosensitizers (22), we hypothesized that HSV-l TK/GCV and CD/5-FC suicide gene therapies may potentiate the effect of radiotherapy. We previously examined these possibilities in vitro by coexpressing CD and HSV-l TK in 9L gliosarcoma cells as a fusion protein (23). Concomitant treatment of cells with 5-FC and GCV at suboptimal concentrations achieved slightly synergistic cytotoxicity that was approximately 2-fold greater than that expected from solely additive effects. These results raised the possibility that double suicide gene therapy might prove to be an improvement over conventional single suicide gene therapy in Vito, given that plasma prodrug concentrations much higher than those used in the in vitro studies can be achieved. Moreover, our investigations demonstrated that both the CD! 5-FC and HSV1 TK/GCV suicide gene systems radiosensitize tumor cells in vitro (24-27), and that even greater radiosensitization can be achieved when these two suicide gene/prodrug systems are combined (23). In the present study. we evaluated the in vito efficacy of double suicide gene therapy. alone and in combination with radiotherapy. An E. co/i C’D/HSV-I TK fusion gene (23) provided the means whereby the CD/5-FC and HSV-l TK/GCV suicide gene systems could function in the same tumor cell. The results demonstrate that dual prodrug therapy generates a significantly greater antitumor effect than conventional single prodrug therapy and is effective against large i.m. tumors. More importantly. double suicide gene therapy was observed to greatly potentiate the therapeutic effects of radiation. MATERIALS AND METHODS Cell Culture. The rat 9L gliosarcoma cell line expressing an E. (0/i CD/HSV-l TK fusion protein (9L-CDg1yTK) was described previously (23). The CDg1yTK fusion protein is produced from a stably integrated provirus that encodes a bicistronic mRNA. Cells were maintained in DMEM supplemented with 10% fetal bovine serum and 250 p.g!ml G418. No antibiotic or antifungal agents were used. In Vivo Prodrug Therapy. Female athymic nude mice [nu/nu (CD1 ); Charles River Laboratories] were used in all studies. Two X i0 ’ 9L-CDg1yTK cells in 0.9% NaC1 were implanted into the right hind leg either s.c. (50 il) above or i.m. (100 p.1) within the right gastrocnemius muscle. When tumors reached an average size of 200 mm3 (s.c. model) or 380 mm’ (i.m. model), animals were divided randomly into four treatment groups of four to five mice/group: no treatment, S-FC (500 mg/kg; Sigma Chemical Co.), GCV (30 mg/kg; Syntex), and S-FC (500 mg/kg) + GCV (30 mg/kg). Prodrugs were dissolved in 0.9% NaC1 and administered i.p. (1 ml) for 14 days beginning on day 0. During prodrug treatment, tumors were measured every other day and intermittently thereafter. Volumes of s.c. and i.m. leg tumors were estimated using the following equation (28): d’3 (0.6)2d’ volume (cm3), where d’ is the average tumor diameter (cm) and the product (0.6)2d’ is the correction factor for normal leg volume. Tumors were allowed to reach five times their initial volume (start of prodrug), at which time mice were euthanized. For large tumor studies, i.m. 9L-CDg1yTK tumors were allowed to attain an average volume of 2300 mm3. Mice subsequently received 21 daily i.p. injections of 5-FC (500 mg/kg) + GCV (30 mg/kg). Animals were euthanized when recurring tumors returned to their original volume (initiation of double prodrug treatment). Irradiation Procedure. Tumor cells were implanted s.c. over the right gastrocnemius muscle as detailed above. When tumors reached an average volume of 250 mm3, mice were divided into two groups (10 animals/group): no treatment and 5-FC + GCV. Prodrug-treated animals received four daily i.p. injections of 5-FC (500 mg/kg) + GCV (30 mg/kg) beginning on day 0. Shortly after the fourth prodrug injection (day 3), the two groups of animals were subdivided further into four groups (five animals/group) based on comparable tumor volumes: no treatment, 25 Gy of radiation, 5-FC + GCV, and 5-FC + GCV + 25 Gy of radiation. For irradiation, mice were anesthetized by i.p. injection of 60 mg/kg Nembutal. Anesthetized animals were secured in a compartmentalized plastic jig such that only the tumor-bearing legs protruded into the central cavity. A bobus of tissue-equivalent material of 2.0-cm thickness (which allows for a homogenous dose distribution throughout the treatment volume) was placed over the central cavity, and a lead alloy (cerrobend) donut coblimeter (three half-value layers) was positioned over the entire jig. This arrangement allowed for delivery of radiation to the tumored limb while simultaneously shielding the body ofthe animal. A 60Co Theratron 780 irradiator (Atomic Energy of Canada Ltd., Ottawa, Ontario, Canada) was used to deliver 25 Gy at a dose rate of I .54 Gy!min using a 62-cm source-to-tumor center distance. For 2 days postirradiation, mice within the prodrug treatment groups received i.p. doses of 5-FC + GCV (six doses total). Tumor volumes were determined as described above. Animals were euthanized when tumors reached 1000 mm3 (4 x initial tumor volume). For all tumor studies, animals were followed until death (euthanasia) from tumor burden or for at least 90 days after cessation of prodrug treatment. All animal procedures were performed with approved protocols and in accordance with published recommendations for the proper use and care of laboratory animals (29). RESULTS In Vivo Antitumor Activity of Double Suicide Gene Therapy. In view of the enhanced cytotoxicity attained via combining the CD/5-FC and HSV-l TKJGCV suicide gene systems in vitro (23), we examined the therapeutic efficacy of double suicide gene therapy in vivo. Rat 9L gliosarcoma cells expressing an E. coli CD/HSV-1 TK fusion protein (9LCDgIyTK) served as our model. A strength of our approach using the CDg1yTK fusion protein is that the efficacy of the Research. on October 31, 2017. © 1997 American Association for Cancer clincancerres.aacrjournals.org Downloaded from