Therapy against Carcinoembryonic Antigen ...

Application of the Cre Recombinase/loxP System Further Enhances Antitumor Effects in Cell Type-specific Gene Therapy against Carcinoembryonic Antigen-producing Cancer Takashi Kijima, Tadashi Osaki, Kazumi Nishino, et al. Cancer Res 1999;59:4906-4911.

Updated version

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/59/19/4906

Cited Articles

This article cites by 30 articles, 10 of which you can access for free at: http://cancerres.aacrjournals.org/content/59/19/4906.full.html#ref-list-1

Citing articles

This article has been cited by 11 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/59/19/4906.full.html#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on June 6, 2013. © 1999 American Association for Cancer Research.

[CANCER RESEARCH 59, 4906 – 4911, October 1, 1999]

Application of the Cre Recombinase/loxP System Further Enhances Antitumor Effects in Cell Type-specific Gene Therapy against Carcinoembryonic Antigen-producing Cancer1 Takashi Kijima, Tadashi Osaki,2 Kazumi Nishino, Toru Kumagai, Toshiki Funakoshi, Hiroyuki Goto, Isao Tachibana, Yoshiro Tanio, and Tadamitsu Kishimoto Department of Molecular Medicine, Osaka University Medical School, Osaka 565-0871 [T. Kij., T. O., K. N., T. Ku., T. F., H. G., I. T.]; Department of Internal Medicine, Osaka Prefectural General Hospital, Osaka 558-0056 [Y. T.]; and Osaka University, Osaka 565-0871 [T. Kis.], Japan

ABSTRACT A considerable number of studies of cancer have shown that the cell type-specific promoter is an effective tool for selective expression of foreign genes in tumor cells. However, few reports have demonstrated significant in vivo antitumor effects using this strategy thus far, possibly because the low activity of such a promoter results in insufficient expression of genes in cancer cells as well as in insignificant antitumor effects, even when the cells are infected by highly efficient gene transfer methods. To overcome this problem, we used the Cre/loxP system for the cell type-specific gene therapy against carcinoembryonic antigen (CEA)-producing cancer. We constructed a pair of recombinant Ads. One expresses the Cre recombinase (Cre) gene under the control of the CEA promoter (Ad.CEA-Cre). The other contains the herpes simplex virus thymidine kinase (HSV-TK) gene separated from the strong CAG promoter by insertion of the neomycin resistance (neo) gene (Ad.lox-TK). The HSV-TK gene of the latter Ad is designed to be activated through excisional deletion of the neo gene by Cre enzyme released from the former one only when CEA-producing cells are infected simultaneously with these Ads. Coinfection by these Ads rendered a human CEA-producing cancer cell line 8.4-fold more sensitive to ganciclovir (GCV) compared with infection by Ad.CEA-TK alone, the HSV-TK gene of which is directly regulated by the CEA promoter. On the other hand, coinfection with these Ads did not significantly change the GCV sensitivity of non-CEA-producing cells. Intratumoral injection of Ad.CEA-Cre combined with Ad.lox-TK followed by GCV treatment almost completely eradicated CEA-producing tumors established in the subcutis of athymic mice, whereas intratumoral injection of Ad.CEA-TK with GCV administration at most retarded the growth of inoculated tumors. These results suggest distinct advantages of the Cre/loxP system applied in the conventional cell type-specific gene therapy against cancer.

INTRODUCTION CEA3 is an oncofetal protein that is often detected at high levels in the serum of cancer patients, including those with cancers of the lung, colon, stomach, and pancreas (1– 4). Many cases of these cancers are found to be already advanced at initial diagnosis, are refractory to conventional chemoradiotherapy, and show poor prognosis (3, 5, 6). Recently, a strategy called virus-directed enzyme/prodrug therapy has been used as an alternative treatment for localized brain tumors in clinical trials and has demonstrated promising results (7, 8). When so-called “suicide” genes, such as the Escherichia coli CD gene and Received 2/18/99; accepted 8/6/99. 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. 1 Supported in part by Grants-in-Aid for Cancer Research from the Ministry of Education, Science and Culture in Japan. 2 To whom requests for reprints should be addressed, at Department of Molecular Medicine, Osaka University Medical School, 2-2, Yamadaoka, Suita, Osaka 565-0871, Japan. 3 The abbreviations used are: CEA, carcinoembryonic antigen; CD, cytosine deaminase; HSV-TK, herpes simplex virus thymidine kinase; GCV, ganciclovir; Ad, adenoviral vector; CAT, chloramphenicol acetyltransferase; TK, thymidine kinase; NLS, nuclear localization signal; moi, multiplicity of infection; pfu, plaque-forming unit(s); MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; i.t., intratumoral(ly).

the HSV-TK gene, are expressed in cancer cells, they convert the nontoxic prodrugs 5-fluorocytosine and GCV, respectively, to toxic compounds and induce apoptosis of cancer cells (9). In applying virus-directed enzyme/prodrug therapy to refractory CEA-producing cancer, disseminated cancer cells must be controlled without affecting normal cells. Cell type-specific promoters have been shown to be effective tools for targeting tumor cells (10 –17). They induce tumor cell-specific death by selective expression of the CD or the HSV-TK genes in tumor cells. However, few reports have demonstrated sufficient antitumor effects, even when tumor cells were introduced with specific promoter-suicide chimera genes in vivo by an Ad, the most efficient vector for introducing foreign genes into target cells (13, 16). It is conceivable that the activity of cell type-specific promoters, including the CEA promoter, is too low to induce sufficient drug sensitivity in vivo. Richards et al. (18) reported that multimerization of a part of the CEA promoter alone or ligated with its specific enhancer sequences resulted in copy number-related increases in promoter activity. We also analyzed by the CAT assay whether the activity of the CEA promoter was augmented by ligation of the enhancer sequence which extends from 26.1 to 24.0 kb upstream regions from the transcriptional start of the CEA gene or by tandem ligation of four copies of the 420-bp fragment of the CEA promoter sequence. These fragments, however, did not contribute to a significant increase in the CAT activity in CEA-producing LoVo cells.4 Cre recombinase (Cre) derived from bacteriophage P1 is a Mr 38,000 protein and mediates the site-specific excisional deletion of a DNA sequence that is flanked by a pair of loxP sites and is composed of 34 nucleotides (19). In a-fetoprotein-producing hepatocellular carcinoma cells, the Cre/loxP system contributed to an ;100-fold increase in the magnitude of the LacZ gene expression compared with that obtained by infection with an Ad expressing the LacZ gene directly driven by the a-fetoprotein promoter (20). To obtain enhanced expression of the HSV-TK gene exclusively in tumor cells and subsequent significant in vivo antitumor effects, we applied the Cre/ loxP system to HSV-TK/GCV therapy for CEA-producing cancer. We constructed a Cre-producing Ad driven by the CEA promoter (Ad.CEA-Cre) and another Ad designed for inducible expression of the HSV-TK gene by Cre (Ad.lox-TK). Coinfection with these Ads induced selective GCV sensitivity at a higher level in vitro and augmented in vivo antitumor effects in human CEA-producing tumors, as compared with those obtained by infection with Ad.CEA-TK alone, an Ad that expresses the HSV-TK gene under the control of the CEA promoter. Here, we demonstrate that the Cre/loxP system significantly enhances antitumor effects of the cell type-specific gene therapy against CEA-producing cancer.

4

T. Kijima, unpublished observations.

4906


GENE THERAPY WITH THE Cre/loxP SYSTEM

MATERIALS AND METHODS

Hoffmann-La Roche, Basel, Switzerland). After cultivation for an additional 6 days, MTT assay was carried out as described previously (11). The 50% growth inhibitory concentration (IC50) of GCV for each cell line was calculated using a curve-fitting parameter, and the results are represented as the means 6 SD from three independent experiments. Student’s t test was used in the statistical analysis of IC50 in individual conditions. The difference was considered statistically significant at P , 0.05. Animal Experiments. On day 0, CEA-producing LoVo cells were inoculated s.c. into the flank of male BALB/c nu/nu mice at 5–7 weeks of age (Clea Japan, Tokyo, Japan). Mice were divided into eight groups: group I, Ad.CEACre and Ad.lox-TK injection with GCV treatment; group II, Ad.CEA-Cre and Ad.lox-TK injection with 0.9% NaCl (saline) treatment; group III, Ad.CEATK injection with GCV treatment; group IV, Ad.CEA-TK injection with saline treatment; group V, PBS injection with GCV treatment; group VI, PBS injection with saline treatment; group VII, Ad.CEA-Cre and Ad.lox-LacZ injection with GCV treatment; and group VIII, Ad.CEA-Cre and Ad.lox-LacZ injection with saline treatment. Groups VII and VIII were excluded in the experiment shown in Fig. 4B. When the average area of the inoculated tumors reached 20 mm2, Ads were injected i.t.. Each recombinant Ad was injected at a ratio of 1:1 in the coinfection groups (groups I, II, VII, and VIII). From the day after the first viral injection, mice received i.p. administration of 50 mg/kg body weight of GCV or saline twice a day for 2 weeks. The tumor size was measured every 3– 4 days and expressed as the product of the perpendicular diameters of individual tumors. The treatment schedule including the number of inoculated cells and the dosage of Ads is specified in each experiment. The number of mice in each group is five in Fig. 4A and seven in Fig. 4B. Each animal experiment was repeated twice. Results are presented as the means 6 SD. Student’s t test was used in the statistical analysis of the size of tumors in individual groups. The difference was considered statistically significant at P , 0.05.

Cell Lines. Two human adenocarcinoma cell lines of the lung, VMRCLCD and A549, and a human colon cancer cell line, LoVo, were all obtained from Japan Cancer Research Resources (Tokyo, Japan). A human embryonal kidney cell line, 293, was purchased from American Type Culture Collection (Manassas, VA). Three cancer cell lines were cultured in RPMI 1640 supplemented with 10% heat-inactivated FBS and penicillin/streptomycin. 293 cells were grown in DMEM containing 10% FBS and penicillin/streptomycin. CEA Production by Tumor Cells. Cells grown in culture medium were washed, suspended in PBS, and homogenized with a sonicator at 50 W for 2 min. The CEA content in the cell lysate was measured by solid-phase RIA. The lower limit of the detection of CEA was 0.5 ng/ml The total protein concentration was assayed by Lowry’s method (Special Reference Laboratory, Tokyo, Japan). Each value represents the former concentration divided by the latter one (ng/mg protein). CAT Assay. The CAT assay was carried out as described previously (11). In brief, 5 mg of a plasmid with the SV40 promoter-CAT gene (pSV2CAT) or a plasmid with the CEA promoter-CAT gene (pCEACAT) was cotransfected with 10 mg of pSV-b-galactosidase, a plasmid with the LacZ gene driven by the SV40 promoter as an internal control for transfection, into each cell line. 14 C-labeled chloramphenicol and its acetyl form were quantitated in duplicate by liquid scintillation counting in Aquasol (DuPont-NEN, Boston, MA). The ratio of the count for the acetyl form to total count was calculated as the CAT activity. The CAT activity of pCEACAT was expressed as a percentage of that of pSV2CAT, a positive control plasmid. Construction of Ads. The recombinant replication-defective Ads were constructed by the cosmid-adenoviral DNA terminal protein complex method (21). The CEATK fragment prepared as the BglII-HindIII fragment from pCEATK (11) was ligated into the SwaI site of pAxcw cassette cosmid and designated pAxosCEATK. pAxCAosTK and pAxCALNLosTK were constructed by subcloning the TK fragment from the BglII-PvuII site of pTK4 (11) into the SwaI site of RESULTS pAxCAwt and pAxCALNLw cassettes (20), respectively. The BglII-HindIII CEA promoter fragment from pCEACAT (11) was also inserted into the SwaI site of CEA Production and CEA-specific Promoter Activity in HupAxAwNCre cassette to produce pAxAosCEANCre (20). Cassette cosmids man Cancer Cell Lines. The amount of CEA protein in cell homopAxcw, pAxCAwt, pAxCALNLw, and pAxAwNCre were all gifts from Dr. I. genate was measured (Table 1). LoVo and VMRC-LCD cells proSaito (Tokyo University, Tokyo, Japan). Each expression cosmid and adenoviral duced significant amounts of CEA protein (77.2 6 12.7 and DNA terminal protein complex were cotransfected into 293 cells by the calcium 204.1 6 43.5 ng/mg protein, respectively), whereas secretion from phosphate precipitation method to produce Ad through homologous recombinaA549 cells was below the detectable level. The activity of the CEA tion, and each Ad was isolated from a single plaque. As shown in Fig. 1A, promoter was also analyzed using CAT assay (Table 1). The CAT AxCAosTK (designated Ad.CAG-TK here), an Ad that contains the HSV-TK gene activities of pCEACAT in LoVo and VMRC-LCD were 98.2 6 0.7 directly driven by the CAG promoter (consisting of the cytomegalovirus immeand 61.7 6 3.4% of that of pSV2CAT, respectively, whereas it was diate early enhancer and a modified chicken b-actin promoter; Ref. 22), strongly expresses the HSV-TK gene in all types of infected cells, regardless of whether very low in A549 cells (4.2 6 0.4%). A549 produced a significant they produce CEA. The Ads AxosCEATK (Ad.CEA-TK) and AxAosCEANCre amount of CEA in our previous experiments (11), but it became a (Ad.CEA-Cre) are designed to express the HSV-TK and the Cre gene, respectively, non-CEA-producing cell line during a long-term cultivation. These exclusively in CEA-producing cells under the control of the CEA promoter. results suggest that the CEA promoter shows selective activity in Ad.CEA-Cre possesses a SV40 poly(A) signal inserted upstream of the CEA CEA-producing cancer cells, although the activity is not so strong. promoter for prevention of nonspecific read-through transcription of the Cre gene Mechanisms Activating the HSV-TK Gene Specifically in CEAand a NLS for keeping Cre recombinase inside the nucleus. AxCALNLosTK producing Cancer Cells by the Cre/loxP System. Fig. 1B illustrates (Ad.lox-TK) is designed to express the HSV-TK gene under the control of the the mechanisms activating the HSV-TK gene specifically in CEAstrong CAG promoter only after Cre excises the neomycin resistance gene in a loxP site-specific manner. AxCANLacZ (Ad.CAG-LacZ), with NLS-tagged b-ga- producing cancer cells. When CEA-producing cells are infected simultaneously with Ad.CEA-Cre and Ad.lox-TK, the former Ad prolactosidase (NLacZ) gene driven by the CAG promoter, and AxCALNLNZ (Adduces Cre recombinase under the control of the CEA promoter. Then .lox-LacZ), identical to Ad.lox-TK, except that the HSV-TK gene is replaced by the NLacZ gene, were also supplied by Dr. I. Saito (20, 23). Solutions of these Ads were prepared as described previously (24) and stored at 280°C until use. b-Galactosidase Assay. One hundred thousand LoVo or A549 cells were Table 1 CEA production and CEA-specific promoter activity in human cancer cell lines coinfected with Ad.CEA-Cre and Ad.lox-LacZ at various ratios of moi, the number of pfu of viral vectors used for infection per cell. Three days after CEA productiona CEA promoter activityb infection, cell lysates were prepared, and the b-galactosidase activity was Cell lines (ng/mg protein) (% of SV40 promoter) measured using o-nitrophenyl-b-D-galactopyranoside as the substrate (25). LoVo 77.2 6 12.7 98.2 6 0.7 In Vitro GCV Sensitivity of Tumor Cells Infected with Recombinant VMRC-LCD 204.1 6 43.5 61.7 6 3.4 A549 Not detectable 4.2 6 0.4 Ads. One thousand A549 or 3 3 103 LoVo or VMRC-LCD cells were seeded a in triplicate onto 96-well culture plates. Twenty-four h later, the culture Level of CEA production was calculated as the concentration of CEA protein divided medium was removed, and suspensions of Ad.CEA-TK alone, Ad.CAG-TK by that of the total protein in cell lysates. Results represent the means 6 SD of three measurements. alone, or Ad.lox-TK in combination with Ad.CEA-Cre were distributed onto isolated b The CAT activity of pCEACAT is expressed as a percentage of that of pSV2CAT. cell monolayers at a total moi of 30. After another 24 h of incubation, the Results represent the means 6 SD of three independent experiments. Each experiment medium was replaced with that containing various concentrations of GCV (F. was performed in duplicate. 4907



strongest b-galactosidase activity (7.1 6 0.3 munits/cell) was obtained with Ads at a ratio of 1:1. No nonspecific b-galactosidase activity was detected on infection with Ad.lox-LacZ alone. Moreover, in nonCEA-producing A549 cells, the b-galactosidase activity remained at very low levels, regardless of the ratios of the paired Ads. The b-galactosidase activity at each ratio increased in proportion to the total moi and reached the maximum level at a total moi of 30 (data not shown). Significant cell toxicity caused by infection with Ads was observed at a moi higher than 30 (data not shown). These results show that the Cre/loxP system allows specific expression of the LacZ gene in CEA-producing cells and confers the highest level of expression when these coupled Ads are used at a ratio of 1:1. In Vitro Enhanced GCV Sensitivity of CEA-producing Cancer Cells by Coinfection with Ad.CEA-Cre and Ad.lox-TK. To yield maximum expression of the HSV-TK gene in CEA-producing cancer cells using the Cre/loxP system, we coinfected LoVo, VMRC-LCD, and A549 cells with Ad.CEA-Cre and Ad.lox-TK at various ratios. There was no significant difference in transduction efficiency of Ads among these three cell lines, as determined by infection with Ad.CAG-LacZ (data not shown). GCV sensitivity was evaluated using MTT assay (Fig. 3). The lowest IC50 was obtained in both LoVo and VMRC-LCD cells when coinfection with the paired Ads was performed at a ratio of 1:1. These results were compatible with those obtained by infection with Ad.CEA-Cre and Ad.lox-LacZ (Fig. 2). The IC50s of each cell line (0.17 6 0.04 and 1.40 6 0.50 mM, respectively) were 8.4- and 3.3-fold lower than that obtained by infection with Ad.CEA-TK alone (1.43 6 0.21 and 4.63 6 0.71 mM, respectively; P , 0.01) and 2488- and 316-fold lower than the values for nontransduced cell lines (423.4 6 40.4 and 443.3 6 55.1 mM, respectively). The IC50s of these CEA-producing cells conferred by the Cre/loxP system were comparable with those obtained by infection with Ad.CAG-TK, which rendered all of these cell lines highly sensitive to GCV, regardless of CEA production. In LoVo and VMRC-LCD cells, the lowest IC50s were only 2.1- and 1.5-fold higher than the value for those infected with Ad.CAG-TK (0.17 6 0.04 and 0.08 6 0.02 mM, respectively). Infection with Ad.lox-TK alone did not significantly change GCV sensitivity in CEA-producing cells. Moreover, coinfection with Ad.CEA-Cre and Ad.lox-TK did not induce significant change in sensitivity of nonCEA-producing A549 cells. These results show that coinfection with Fig. 1. A, structure of recombinant Ads. Ad.CAG-TK expresses the HSV-TK gene driven by the CAG promoter. Ad.CEA-TK and Ad.CEA-Cre express the HSV-TK gene and the Cre gene, respectively, under the control of the CEA promoter. Ad.lox-TK is designed to express the HSV-TK gene under the CAG promoter only after Cre recombinase excises the neomycin resistance (neo) gene. OO, the adenoviral genome; E, adenoviral DNA terminal protein complexes; Œ, deletions of adenoviral early regions. CAG, CAG promoter; CEA, CEA promoter; NCre, NLS-tagged Cre gene; SpA, SV40 early poly(A) signal; GpA, rabbit b-globin poly(A) signal; p, loxP site. B, the Cre/loxP system mediates high level expression of the HSV-TK gene specifically in CEA-producing cancer cells. Cre recombinase expressed by Ad.CEA-Cre excises the neo gene with SpA at the loxP sites out of Ad.lox-TK and activates the HSV-TK gene driven by the strong CAG promoter.

the produced Cre excises the neomycin resistance gene between a pair of loxP sites out of the latter Ad and activates the HSV-TK gene driven by the strong CAG promoter. Therefore, administered GCV is phosphorylated and induces cell death exclusively in CEA-producing cells. Determination of the Ideal Ratio of the Paired Ads to Achieve Maximum Expression of the LacZ Gene Using the Cre/loxP System. To maximize the expression of the LacZ gene using the Cre/loxP system in vitro, we analyzed the ideal ratio of Ad.CEA-Cre to Ad.loxLacZ. Cre released from Ad.CEA-Cre activates the LacZ gene driven by the CAG promoter. Human adenocarcinoma cell lines, LoVo and A549, were infected simultaneously with these coupled Ads at various ratios at a total moi of 30. The LacZ gene expression was quantified by b-galactosidase assay (Fig. 2). In CEA-producing LoVo cells, the

Fig. 2. Coinfection with Ad.CEA-Cre and Ad.lox-LacZ at a ratio of 1:1 maximizes selective expression of the LacZ gene in CEA-producing cell lines. One hundred thousand LoVo (f) or A549 (M) cells were coinfected with these Ads at various ratios at a total moi of 30. Three days after infection, cell lysates were prepared and subjected to b-galactosidase assay. Columns, mean b-galactosidase activity; bars, SD. Numbers below columns indicate the moi of Ads used for infection. The same experiment was repeated twice more with similar results.

4908



we injected tumors of LoVo cells established in the subcutis of athymic mice i.t. with Ads, followed by GCV administration when the average tumor area reached 20 mm2 (Fig. 4). When a low dose of Ads (23108 pfu per injection) was injected (Fig. 4A), significant regression of tumor growth was observed only in the mice treated with Ad.CEA-Cre and Ad.lox-TK followed by GCV (group I; group I versus groups III-VIII, P , 0.005). In three of five mice (60%) in this group, tumors were completely eradicated, and two other mice

Fig. 3. Coinfection with Ad.CEA-Cre and Ad.lox-TK remarkably augments in vitro GCV sensitivity of CEA-producing cancer cells. Three thousand LoVo or VMRC-LCD or 1 3 103 A549 cells were infected with Ad.CAG-TK alone (p), Ad.CEA-TK alone (o), or Ad.lox-TK in combination with Ad.CEA-Cre at various ratios (f) at a total moi of 30. Cells without viral infection (M) were also studied. Sensitivity to GCV was analyzed by MTT assay. Columns, mean IC50 of GCV from three independent experiments; bars, SD. Each experiment was performed in triplicate. The lowest IC50 values were obtained in both CEA-producing LoVo and VMRC-LCD cells (0.17 6 0.04 and 1.40 6 0.50 mM, respectively) when coinfection with the paired Ads was performed at a ratio of 1:1, 8.4and 3.3-fold lower than for infection with Ad.CEA-TK alone (1.43 6 0.21 and 4.63 6 0.71 mM, respectively). Moreover, coinfection with these coupled Ads did not change GCV sensitivity in non-CEA-producing A549 cells.

Ad.CEA-Cre and Ad.lox-TK remarkably augments selective GCV sensitivity in CEA-producing cancer cells compared with infection by Ad.CEA-TK alone and that the magnitude of enhancement in GCV sensitivity correlates with the activity of the CEA promoter rather than the amount of CEA protein produced (Table 1). Immediate Rejection of Established CEA-producing Tumors by Coinfection with Ad.CEA-Cre and Ad.lox-TK Followed by GCV Treatment. To evaluate in vivo antitumor effects on CEAproducing tumors by coinfection with Ad.CEA-Cre and Ad.lox-TK,

Fig. 4. Coinfection with Ad.CEA-Cre and Ad.lox-TK followed by GCV treatment resulted in the rejection of CEA-producing tumors. Nude mice were inoculated s.c. with CEAproducing LoVo cells on day 0. Treatment with Ad (arrows) and GCV (rectangles) was initiated when the average area of the inoculated tumors reached 20 mm2. f, group I, Ad.CEA-Cre and Ad.lox-TK injection with GCV treatment; M, group II, Ad.CEA-Cre and Ad.lox-TK injection with saline treatment; F, group III, Ad.CEA-TK injection with GCV treatment; E, group IV, Ad.CEA-TK injection with saline treatment; Œ, group V, PBS injection with GCV treatment; ‚, group VI, PBS injection with saline treatment; r, group VII, Ad.CEA-Cre and Ad.lox-LacZ injection with GCV treatment; e, group VIII, Ad.CEA-Cre and Ad.lox-LacZ injection with saline treatment. Each Ad was injected at a ratio of 1:1 in the coinfection groups (groups I, II, VII, and VIII). Data points, tumor area, shown as the mean of the products of the perpendicular diameters of each tumor; bars, SD. The difference was considered statistically significant at P , 0.05. N.S., not significant. The numbers of inoculated LoVo cells were 2 3 106 (A) and 3 3 106 (B). Ad was i.t. injected at 2 3 108 pfu on days 8, 15, and 18 (A) and 5 3 108 pfu on days 3, 8, and 13 (B). GCV was i.p. administered at 50 mg/kg twice a day on days 9 –22 (A) and on days 4 –17 (B). The numbers of mice in each group were n 5 5 (A) and n 5 7 (B). Groups VII and VIII were excluded in B. Ad.CEA-Cre and Ad.lox-TK/GCV treatment (group I) completely eradicated tumors in three of five (A) and six of seven (B) mice, but Ad.CEA-TK/GCV treatment (group III) at most retarded the growth of inoculated tumors (B).

4909



showed transient regression of tumor during GCV treatment. Nevertheless, the other treatment including Ad.CEA-TK/GCV (group III) did not have any significant effects on tumor growth (group III versus group VII, P . 0.05). To further augment antitumor effects, we administered an increased dose of Ads to the mice (5 3 108 pfu per injection; Fig. 4B). All mice treated with Ad.CEA-Cre and Ad.lox-TK followed by GCV injection (group I) showed immediate regression of inoculated tumors. Six of seven mice (86%) became free of palpable tumor by the end of the treatment schedule, and the tumor in the rest of the mice in this group remained a small nodule. On the other hand, Ad.CEA-TK/GCV (group III) as well as Ad.CEA-Cre and Ad.loxTK/saline treatment (group II) slightly retarded the growth of tumors but did not eradicate them (group III versus group IV, P , 0.05, from day 24 to day 44, and group II versus group VI, P , 0.01, from day 24 to day 44, respectively). Thus, antitumor effects observed in the Ad.CEA-Cre and Ad.lox-TK/GCV treatment group (group I) were more remarkable compared with those of the Ad.CEA-TK/GCV treatment group (group III; group I versus group III, P , 0.0001). We did not observe any apparent side effects including body weight loss or death in these treatment cycles. These results suggest that the Cre/loxP system remarkably enhances antitumor effects on CEA-producing tumors in this cancer gene therapy model. DISCUSSION This is the first report demonstrating that application of the Cre/lox P system remarkably enhances selective expression of the HSV-TK gene in CEA-producing tumor cells and subsequent in vivo antitumor effects. The activity of the CEA promoter in CEA-producing LoVo cells was comparable with that of the SV40 promoter determined by CAT assay (Table 1). Ad.CEA-TK rendered LoVo cells 296-fold more sensitive to GCV in vitro in comparison with uninfected cells (Fig. 3). However, repeated i.t. injection of Ad.CEA-TK followed by administration of GCV at most retarded the growth of the inoculated tumors (Fig. 4). On the other hand, i.t. injection of Ad.CAG-TK, which has the strong CAG promoter, with GCV treatment completely eradicated s.c. LoVo tumors in all animals.4 It is conceivable that expression of the HSV-TK gene at a higher level is necessary for complete rejection of tumors. To augment selective expression of the HSV-TK gene in CEA-producing cells, we applied the Cre/loxP system. Coinfection with Ad.CEA-Cre and Ad.lox-TK significantly enhanced the sensitivity of CEA-producing LoVo and VMRC-LCD cells to GCV. The IC50s of these cells were only 2.1- and 1.5-times as high as those of Ad.CAG-TK-transfected LoVo and VMRC-LCD cells, respectively. Moreover, infection with Ad.lox-TK alone did not induce nonspecific GCV sensitivity in CEA-producing cells, nor did coinfection with these Ads significantly change GCV sensitivity in Non-CEA-producing cells (Fig. 3). GCV treatment following i.t. injection of Ad.CEACre combined with Ad.lox-TK induced immediate regression of established CEA-producing LoVo tumors and resulted in complete tumor rejection in six of seven animals (Fig. 4B). This treatment significantly suppressed the growth of the other CEA-producing VMRC-LCD tumors as well in all mice but did not show any antitumor effects on non-CEA-producing A549 tumors (data not shown). The following reasons might account for the enhanced antitumor effects. (a) A limited amount of induced Cre recombinase effectively processes a large number of molecules on the target Ad genome (23, 26). (b) The activity of the CAG promoter is far stronger than that of the CEA promoter. In fact, Ad.CAG-TK made LoVo cells 18-fold more sensitive to GCV compared with Ad.CEA-TK (Fig. 3). (c) Cre is efficiently activated inside the nucleus by NLS with an effective initiator codon of Ad.CEA-Cre (23).

Marginal retardation of tumor growth was observed in mice injected i.t. with a higher dose of Ad.CEA-Cre combined with Ad.loxTK then administered saline (Fig. 4B). When LoVo cells were coinfected with Ad.CEA-Cre and Ad.lox-TK or infected with Ad.CAG-TK alone in vitro at a total moi of 30, the viability of the infected cells decreased to 70 or 50% of that of uninfected cells, respectively (data not shown). Other reports also showed that i.t. injection of Ads containing the CAG promoter significantly retarded the growth of s.c. tumors (16). We speculate that infection by a higher dose of Ads with the CAG promoter interferes with the transcription of genes necessary for cell survival and induces apoptosis (12, 16). Death related to severe hepatic toxicity is reported when animals receive administration of Ad expressing the HSV-TK gene driven by the cytomegalovirus promoter followed by GCV treatment (27, 28). In another report, reverse transcriptase-PCR analysis detected expression of transduced genes predominantly in the liver, the kidneys, and the lung, and histological examination revealed zonal necrosis in the periportal area of the liver when an Ad expressing the CD gene under the control of the CAG promoter was i.p. administered then 5-fluorocytosine given (29). We did not observe any side effects including elevation of transaminase or creatinine in the serum of animals treated with i.p. injection of Ad.CEA-Cre and Ad.lox-TK followed by GCV administration (data not shown). Thus, application of the Cre/loxP system is expected to not only enhance antitumor effects at a reduced dosage of Ads but also avoid damage to normal tissues. Another advantage in application of this system is that it might be no longer needed to specify the enhancer or silencer sequences in the genome for each cell type-specific promoter. Our ultimate goal is to control metastasized tumor cells without affecting normal cells. We are now planning to evaluate the efficacy and side effects of this system in the peritonitis carcinomatosa model in mice, mimicking pleuritis for lung cancer and peritonitis for gastrointestinal cancers in future clinical applications to human beings. Recently, a combination of HSV-TK/GCV gene therapy with cytokines, including interleukin 2 (30, 31), interleukin 4 (32), and granulocyte-macrophage colony-stimulating factor (31), has been shown to enhance antitumor effects. Modified Ad.lox-TK in which the neomycin resistance gene is replaced by these cytokine genes would enhance antitumor effects. When these Ads are used simultaneously with Ad.CEA-Cre, not only is the HSV-TK gene expressed specifically in CEA-producing tumor cells, but the cytokines released nonspecifically from surrounding normal cells also induce antitumor immunity. In conclusion, the Cre/loxP system augments possible clinical application of gene therapy specific for CEA-producing cancer. ACKNOWLEDGMENTS We are grateful to Dr. Izumu Saito for providing materials used for viral constructs and for helpful technical support. We also thank Dr. Jun-ichi Miyazaki for the CAG promoter and Yoko Habe for technical and secretarial assistance.

REFERENCES 1. Shively, J. E., and Beatty, J. D. CEA-related antigens: molecular biology and clinical significance. CRC Crit. Rev. Oncol. Hematol., 2: 355–399, 1985. 2. Shinkai, T., Saijo, N., Tominaga, K., Eguchi, K., Shimizu, E., Sasaki, Y., Fujita, J., Futami, H., Ohkura, H., and Suemasu, K. Serial plasma carcinoembryonic antigen measurement for monitoring patients with advanced lung cancer during chemotherapy. Cancer (Phila.), 57: 1318 –1323, 1986. 3. Wanebo, H. J., Kennedy, B. J., Chmiel, J., Steele, G., Jr., Winchester, D., and Osteen, R. Cancer of the stomach: a patient care study by the American College of Surgeons. Ann. Surg., 218: 583–592, 1993. 4. Podolsky, D. K. Serologic markers in the diagnosis and management of pancreatic carcinoma. World J. Surg., 8: 822– 830, 1984.

4910



5. Belani, C. P. Multimodality management of regionally advanced non-small-cell lung cancer. Semin. Oncol., 20: 302–314, 1993. 6. Warshaw, A. L., and Fernandez, C. C. Pancreatic carcinoma. N. Engl. J. Med., 326: 455– 465, 1992. 7. Eck, S. L., Alavi, J. B., Alavi, A., Davis, A., Hackney, D., Judy, K., Mollman, J., Phillips, P. C., Wheeldon, E. B., and Wilson, J. M. Treatment of advanced CNS malignancies with the recombinant adenovirus H5.010RSVTK: a Phase I trial. Hum. Gene Ther., 7: 1465–1482, 1996. 8. Ram, Z., Culver, K. W., Oshiro, E. M., Viola, J. J., DeVroom, H. L., Otto, E., Long, Z., Chiang, Y., McGarrity, G. J., Muul, L. M., Katz, D., Blaese, R. M., and Oldfield, E. H. Therapy of malignant brain tumors by intratumoral implantation of retroviral vector-producing cells. Nat. Med., 3: 1354 –1361, 1997. 9. Freeman, S. M., Abboud, C. N., Whartenby, K. A., Packman, C. H., Koeplin, D. S., Moolten, F. L., and Abraham, G. N. The “bystander effect”: tumor regression when a fraction of the tumor mass is genetically modified. Cancer Res., 53: 5274 –5283, 1993. 10. Vile, R. G., Nelson, J. A., Castleden, S., Chong, H., and Hart, I. R. Systemic gene therapy of murine melanoma using tissue specific expression of the HSVtk gene involves an immune component. Cancer Res., 54: 6228 – 6234, 1994. 11. Osaki, T., Tanio, Y., Tachibana, I., Hosoe, S., Kumagai, T., Kawase, I., Oikawa, S., and Kishimoto, T. Gene therapy for carcinoembryonic antigen-producing human lung cancer cells by cell type-specific expression of herpes simplex virus thymidine kinase gene. Cancer Res., 54: 5258 –5261, 1994. 12. Lan, K. H., Kanai, F., Shiratori, Y., Okabe, S., Yoshida, Y., Wakimoto, H., Hamada, H., Tanaka, T., Ohashi, M., and Omata, M. Tumor-specific gene expression in carcinoembryonic antigen-producing gastric cancer cells using adenovirus vectors. Gastroenterology, 111: 1241–1251, 1996. 13. Tanaka, T., Kanai, F., Lan, K. H., Ohashi, M., Shiratori, Y., Yoshida, Y., Hamada, H., and Omata, M. Adenovirus-mediated gene therapy of gastric carcinoma using cancerspecific gene expression in vivo. Biochem. Biophys. Res. Commun., 231: 775–779, 1997. 14. Kumagai, T., Tanio, Y., Osaki, T., Hosoe, S., Tachibana, I., Ueno, K., Kijima, T., Horai, T., and Kishimoto, T. Eradication of Myc-overexpressing small cell lung cancer cells transfected with herpes simplex virus thymidine kinase gene containing Myc-Max response elements. Cancer Res., 56: 354 –358, 1996. 15. Chen, L., Chen, D., Manome, Y., Dong, Y., Fine, H. A., and Kufe, D. W. Breast cancer selective gene expression and therapy mediated by recombinant adenoviruses containing the DF3/MUC1 promoter. J. Clin. Invest., 96: 2775–2782, 1995. 16. Kanai, F., Lan, K. H., Shiratori, Y., Tanaka, T., Ohashi, M., Okudaira, T., Yoshida, Y., Wakimoto, H., Hamada, H., Nakabayashi, H., Tamaoki, T., and Omata, M. In vivo gene therapy for a-fetoprotein-producing hepatocellular carcinoma by adenovirusmediated transfer of cytosine deaminase gene. Cancer Res., 57: 461– 465, 1997. 17. Braiden, V., Nagayama, Y., Iitaka, M., Namba, H., Niwa, M., and Yamashita, S. Retrovirus-mediated suicide gene/prodrug therapy targeting thyroid carcinoma using a thyroid-specific promoter. Endocrinology, 139: 3996 –3999, 1998. 18. Richards, C. A., Austin, E. A., and Huber, B. E. Transcriptional regulatory sequences of carcinoembryonic antigen: identification and use with cytosine deaminase for tumor-specific gene therapy. Hum. Gene Ther., 6: 881– 893, 1995.

19. Sternberg, N., and Hamilton, D. Bacteriophage P1 site-specific recombination I. Recombination between loxP sites. J. Mol. Biol., 150: 467– 486, 1981. 20. Sato, Y., Tanaka, K., Lee, G., Kanegae, Y., Sakai, Y., Kaneko, S., Nakabayashi, H., Tamaoki, T., and Saito, I. Enhanced and specific gene expression via tissue-specific production of Cre recombinase using adenovirus vector. Biochem. Biophys. Res. Commun., 244: 455– 462, 1998. 21. Miyake, S., Makimura, M., Kanegae, Y., Harada, S., Sato, Y., Takamori, K., Tokuda, C., and Saito, I. Efficient generation of recombinant adenoviruses using adenovirus DNA-terminal protein complex and a cosmid bearing the full-length virus genome. Proc. Natl. Acad. Sci. USA, 93: 1320 –1324, 1996. 22. Niwa, H., Yamamura, K., and Miyazaki, J. Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene, 108: 193–199, 1991. 23. Kanegae, Y., Lee, G., Sato, Y., Tanaka, M., Nakai, M., Sakaki, T., Sugano, S., and Saito, I. Efficient gene activation in mammalian cells by using recombinant adenovirus expressing site-specific Cre recombinase. Nucleic Acids Res., 23: 3816 –3821, 1995. 24. Kanegae, Y., Makimura, M., and Saito, I. A simple and efficient method for purification of infectious recombinant adenovirus. Jpn. J. Med. Sci. Biol., 47: 157–166, 1994. 25. Sambrook, J., Fritsch, E. F., and Maniatis, T. Molecular Cloning: A Laboratory Manual, Ed. 2, pp. 16.66 –16.67. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1989. 26. Kanegae, Y., Takamori, K., Sato, Y., Lee, G., Nakai, M., and Saito, I. Efficient gene activation system on mammalian cell chromosomes using recombinant adenovirus producing Cre recombinase. Gene, 181: 207–212, 1996. 27. Brand, K., Arnold, W., Bartels, T., Lieber, A., Kay, M. A., Strauss, M., and Dorken, B. Liver-associated toxicity of the HSV-tk/GCV approach and adenoviral vectors. Cancer Gene Ther., 4: 9 –16, 1997. 28. van der Eb, M. M., Cramer, S. J., Vergouwe, Y., Schagen, F. H., van Krieken, J. H., van der Eb, A. J., Rinkes, I. H., van de Velde, C. J., and Hoeben, R. C. Severe hepatic dysfunction after adenovirus-mediated transfer of the herpes simplex virus thymidine kinase gene and ganciclovir administration. Gene Ther., 5: 451– 458, 1998. 29. Lan, K. H., Kanai, F., Shiratori, Y., Ohashi, M., Tanaka, T., Okudaira, T., Yoshida, Y., Hamada, H., and Omata, M. In vivo selective gene expression and therapy mediated by adenoviral vectors for human carcinoembryonic antigen-producing gastric carcinoma. Cancer Res., 57: 4279 – 4284, 1997. 30. Kwong, Y. L., Chen, S. H., Kosai, K., Finegold, M., and Woo, S. L. Combination therapy with suicide and cytokine genes for hepatic metastases of lung cancer. Chest, 112: 1332–1337, 1997. 31. Chen, S. H., Kosai, K., Xu, B., Pham-Nguyen, K., Contant, C., Finegold, M. J., and Woo, S. L. Combination suicide and cytokine gene therapy for hepatic metastases of colon carcinoma: sustained antitumor immunity prolongs animal survival. Cancer Res., 56: 3758 –3762, 1996. 32. Benedetti, S., Dimeco, F., Pollo, B., Cirenei, N., Colombo, B. M., Bruzzone, M. G., Cattaneo, E., Vescovi, A., Didonato, S., Colombo, M. P., and Finocchiaro, G. Limited efficacy of the HSV-TK/GCV system for gene therapy of malignant gliomas and perspectives for the combined transduction of the interleukin-4 gene. Hum. Gene Ther., 8: 1345–1353, 1997.

4911
