Adenovirus-mediated transfer of the p53 family genes, p73 ... - Nature

Gene Therapy (2001) 8, 1401–1408  2001 Nature Publishing Group All rights reserved 0969-7128/01 $15.00 www.nature.com/gt

RESEARCH ARTICLE

Adenovirus-mediated transfer of the p53 family genes, p73 and p51/p63 induces cell cycle arrest and apoptosis in colorectal cancer cell lines: potential application to gene therapy of colorectal cancer Y Sasaki1,2, I Morimoto1,2, S Ishida1, T Yamashita1,3, K Imai2 and T Tokino1 1

Department of Molecular Biology, Cancer Research Institute, 2First Department of Internal Medicine, and 3Department of Dermatology, Sapporo Medical University School of Medicine, S-1, W-17, Chuo-ku, Sapporo, 060-8556 Japan

p53 gene therapy is being tested clinically for the treatment of human cancer, however, some cancer models (in vivo and in vitro) are resistant to p53. To explore the potential use of two p53 homologues, p73 and p51/p63, in cancer gene therapy, we introduced p53, p73 and p51/p63 into colorectal cancer cell lines via adenoviral vectors, and compared their effects on cell growth. Among 10 cell lines tested, six cell lines displayed a similar response following transduction of p53, p73␤ or p51A/p63␥; two lines underwent cell-cycle arrest, three lines exhibited apoptosis and one line showed no-effect following transduction. The effect on cell-cycle progression was variable in the other four cell lines. Interestingly, three cell lines were resistant to p53-mediated

apoptosis, including two lines having endogenous wild-type p53 alleles, but underwent apoptosis after transduction of p73␤ or p51A/p63␥. Similar to p53, transduction of p51A/p63␥ induced extensive apoptosis when combined with adriamycin or X-radiation in SW480 cells, which are normally resistant to apoptosis. Transduction of p73␤ and p51A/p63␥ also reduced the tumorigenicity of two colorectal cancer cells in vivo. These results suggest that adenovirusmediated p73␤ and p51A/p63␥ transfer are potential novel approaches for the treatment of human cancers, particularly for tumors that are resistant to p53 gene therapy. Gene Therapy (2001) 8, 1401–1408.

Keywords: p53; p73; p51/p63; apoptosis; adenoviral vector; gene therapy

Introduction The p53 tumor suppressor gene is involved in cell cycle control, apoptosis, genome stability and angiogenesis, through the transactivation of p53-target genes.1–3 Mutations in the p53 gene are the most commonly detected in human cancers.4,5 Inactivation of p53 is involved not only in carcinogenesis, but also in conferring resistance to chemotherapy.6,7 Therefore, therapeutic replacement of the wild-type p53 gene has been pursued as a potential gene therapy strategy in a variety of cancer types.8 This strategy relies in part on the apoptosis induced by wild-type p53. High level expression of p53 causes some cancer cells to undergo apoptosis, whereas others simply undergo prolonged cell-cycle arrest.9 Additionally, in clinical trials, the restoration of the wildtype p53 gene function does not always lead to tumor regression or tumor growth inhibition, suggesting resistance of some tumors to exogenous p53.10–12 Therefore, to improve the efficacy of p53 gene therapy, it would be important to overcome the resistance to p53-mediated apoptosis. Correspondence: T Tokino, Department of Molecular Biology, Cancer Research Institute, Sapporo Medical University School of Medicine, S-1, W-17, Chuo-ku, Sapporo, 060-8556 Japan Received 21 February 2001; accepted 28 June 2001

p73 and p51/p63 were recently identified as members of the p53 gene family and encode proteins that share considerable structural homology with p53.13–16 p73 and p51/p63 can bind to the p53 responsive elements and upregulate some p53 target genes, which suggest that the p53 family member genes have a potential for functional overlap with p53 itself.17–20 However, in contrast to p53, p73 and p51/p63 are rarely mutated in human cancers.21 Li–Fraumeni syndrome patients have inherited mutations of the p53 gene and develop normally, but are predisposed to cancer,22 while heterozygous germline mutations in the p51/p63 gene are the cause of ectrodactyly, ectodermal dysplasia and facial clefts (EEC) syndrome.23 Strikingly different phenotypes between p73(or p51/p63-) deficient and p53-deficient mice were also reported.24–27 In contrast to p53-deficient mice, mice lacking the p73 or p51/p63 genes show no increased susceptibility to spontaneous tumorigenesis. Instead, p73deficient mice have neurological, pheromonal and inflammatory defects. p51/p63-deficient mice have major defects in their limbs and craniofacial development, as well as a striking absence of stratified epithelia, suggesting that p51/p63 is required for limb and epidermal morphogenesis. These studies demonstrate a marked divergence in the developmental roles of p73 and p51/p63, and further distinguished these p53 family genes from p53. Unlike p53, both p73 and p51/p63 proteins exist in

Role of the p53 family genes in apoptosis Y Sasaki et al

1402

multiple isoforms with different, often contradictory, biological activities, and production of p73 and p51/p63 appears to be restricted to certain tissues.16 When overexpressed, exogenous p73␤ and p51A/p63␥ can activate the same target genes as p53 and can induce apoptosis. However, the studies of p53 and p53 family members have used different cell types and have used different methods to induce gene expression.17–20 It has therefore been difficult to make direct comparisons between the apoptotic activity observed in independent studies. In the present study, to evaluate the anti-cancer activity of p73 and p51/p63 in comparison with that of p53, we sought to analyze growth arrest and apoptosis induced by the p53 family genes using adenoviral vectors in a panel of human colorectal cancer cell lines. We found that the apoptotic response to the p53 family genes was uniform in some cell lines and heterogenous in others. Our results suggest that adenovirus-mediated p73␤ and p51A/p63␥ gene transfer is a potential novel approach for the treatment of colorectal cancers, particularly for tumors that are resistant to wild-type p53 gene therapy.

Results Adenovirus-mediated gene transfer in colorectal cancer cells To examine the effect of expression of exogenous p53 family genes on proliferation of colorectal cancer cells, we used the replication-deficient adenoviral vector harboring human p53 (Ad-p53), p73␣ (Ad-p73␣), p73␤ (Adp73␤), p51A/p63␥ (Ad-p51A) and p51B/p63␣ (Ad-p51B) cDNAs. The relative efficiency of adenovirus-mediated gene transfer was assessed as the percentage of X-gal staining cells after infection with adenovirus containing the bacterial lacZ gene (Ad-lacZ) at a multiplicity of infection (MOI) ranging from 10 to 200. Of 13 cell lines tested, 10 cell lines showed highly efficient gene transfer with 80 to 100% of the cells staining for ␤-galactosidase activity at a MOI of 100, and these 10 cell lines were used in this study. p53, p73, and p51/p63 expression in adenovirusinfected cells According to the efficiency of gene transfer assessed by X-gal staining, the cells were infected with adenovirus at a MOI of 100. The expression of p53 was detected by immunoblot analysis. A high level of expression of p53 protein was achieved in cells infected with Ad-p53 (examples in Figure 1A). In Ad-lacZ-infected SW480 cells, the endogenous mutant p53 protein was seen at a low level. In contrast, the endogenous p53 protein in DLD1 and HCT116 cells infected with Ad-lacZ was not detectable. Similarly, infection with Ad-p73␤ and Ad-p51A into these cells resulted in expression of exogenous p73␤ and p51A/p63␥ proteins, respectively (examples in Figure 1A). These results indicate that the expression of the p53 family genes by the adenovirus-mediated transfer were highly efficient in human colorectal cancer cells tested here. We then compared the ability of p53, p73␤ and p51A/p63␥ to transactivate endogenous p21, a known p53-target gene,28 24 h after infection. Overexpression of p51A/p63␥, like p53, led to an increase in p21 protein in both DLD1 and SW480 cells, while p73␤ expression resulted in a slight increase in p21 protein only in DLD1 cells (Figure 1B). Gene Therapy

Figure 1 Immunoblot analysis of p53, p73␤, p51A and p21 proteins. (A) Expression of the p53 family gene products at 24 h after adenoviral infection in SW480, DLD1 and HCT116 cells. Immunoblot analysis was performed using cell lysates following infection with Ad-p53 (a), Ad-p73␤ (b) and Ad-p51A (c) at a MOI of 100 with the gene-specific antibody. A significant amount of p53, p73␤ and p51A were detected after infection with Ad-p53, Ad-p73␤ and Ad-p51A, respectively. Note that a faint band corresponding to p53 was detected in Ad-lacZ-infected SW480 cells, which express endogenous mutant p53. (B) Induction of endogenous p21 protein in SW480 and DLD1 cells at 24 h after adenoviral infection.

Effect of adenovirus-mediated transfer of the p53 family genes on cell cycle progression The effect of adenovirus-mediated transfer of the p53 family genes on cell cycle progression was assessed by flow cytometry. Cells were infected at a MOI of 100, harvested at 48 h following adenoviral infection, and then assayed for DNA content. We evaluated the response to each of the adenoviral gene transfers, reflected by flow cytometry patterns, as follows: apoptosis (over 25% of the cells in sub-G1 population) and cell cycle arrest (over 50% decrease of cells in S phase in the absence of apoptosis). A summary of flow cytometric analysis is shown in Figure 2. In the present study, only a few cell lines displayed apoptosis after adenovirus-mediated transfer of p73␣ and p51B/p63␣ (Figure 2), and the degree of apoptosis was lower than in cells infected with Ad-p73␤ and Adp51A/p63␥, respectively (examples in Figure 3). Similar results were previously reported in different cell types.14,20 These differences in response were not simply attributable to differences in the expression levels, as determined by immunoblot analysis.20 Therefore, we analyzed p73␤ and p51A/p63␥ isoforms in further detail. Cell lines that underwent cell cycle arrest after infection were accompanied by a significant increase in the proportion of cells in the G0–G1 phase. Among 10 cell lines, six showed similar responses to each of p53, p73␤ and p51A/p63␥. For example, SW480 and CHC-Y1 cells underwent G1 arrest in response to p53, p73␤ or p51A/p63␥. colo320, DLD1 and HCT15 cells also underwent apoptosis in response to p53, p73␤ or p51A/p63␥. In contrast, cell cycle progression in HT29 cells was unaffected. In the four other lines (LoVo,


1403

Figure 2 Summary of FACS analysis. The effect of adenovirus-mediated transfer of the p53 family genes on cell cycle progression is shown. The adenoviral infection resulted in apoptosis (over 25% of the cells in subG1 population at 48 h following infection) or cell cycle arrest (over 50% decrease of cells in S phase in the absence of apoptosis). Some cells showed neither apoptosis nor cell cycle arrest following adenoviral infection (empty boxes). The p53 status in the cell lines was reported previously (wt, wild type; mut, mutant).40–43 The gene transfer efficiency of the adenovirus was evaluated after infection with Ad-lacZ at a MOI of 100 using X-gal staining. Experiments were repeated three times, and the mean percentages are shown.

HCT116, WiDr and BM314), expression of the p53 family genes had variable effects on the flow cytometric patterns (Figure 2). A time-course study was performed on two cell lines that exhibited apoptotic (DLD1) or G1 arrest (SW480) responses. Both cell lines initially displayed cell cycle arrest, with cells blocked in the G1 phase 24 h after infection with either Ad-p53, Ad-p73␤, or Ad-p51A (Figure 3a and b). A significant sub-G1 peak was detected in DLD1 cells at 48 h after infection (Figure 3a). In striking contrast to DLD1 cells, SW480 cells displayed a little or no subG1 fraction up to 96 h after infection (Figure 3b). Even when SW480 cells were infected at a MOI of 200, sub-G1 population was not observed 48 h after infection (Figure 3b). p21 is a well-known mediator of p53-induced growth arrest. To evaluate the relationship between p21 and the growth arrest/apoptosis switch, we analyzed the effect of exogenous p21 overexpression on these two cell lines. The infection of Ad-p21 resulted in a significant G1 arrest in both cell lines. Considered together with the result of endogenous p21 expression after overexpression of the p53 family genes, these observations demonstrate that p21 was not responsible for the growth arrest/apoptosis switch. As both cell lines have mutant alleles of the p53 gene, the status of endogenous p53 does not correlate with the cell line response to overexpression of p53, p73␤ or p51A/p63␥. In HCT116 and LoVo cell lines, which have wild-type p53, Ad-p53 infection did not result in apoptosis. Interestingly, LoVo showed an increase of the sub-G1 population following transduction of p73␤ or p51A/p63␥, and HCT116 showed a significant increase of the sub-G1 population after Ad-p51A infection (Figure 4). Moreover, WiDr cells, which were also resistant to p53-mediated apoptosis, underwent apoptosis after Ad-p51A infection. These findings indicate that p51A/p63␥ and possibly

Figure 3 Representative histograms of DNA content following adenovirus-mediated transfer of lacZ, p53, p73␣ p73␤, p51A, p51B or p21. Cells were harvested 24, 48, 72, or 96 h following infection and analyzed by flow cytometry. The Ad-lacZ and Ad-p21-infected cells showed almost similar patterns at these time-points, and therefore, histograms at 72 h after infection are shown. Experiments were repeated three times, and representative results are shown: (a) DLD1; (b) SW480.

p73␤ gene transduction may be useful in the treatment of cancers resistant to p53 gene therapy. Effect of adenovirus-mediated transfer of the p53 family genes on cell viability To compare the effect of expression of exogenous p53 family genes on growth of DLD1 and SW480 cells, cells Gene Therapy


1404

Figure 4 Cell cycle distribution of cells harboring wild-type p53 (LoVo and HCT116) and WiDr cells. Cells were infected with either Ad-lacZ, Ad-p53, Ad-p73␤ or Ad-p51A for 48 h, trypsinized, and then analyzed by flow cytometry. Experiments were repeated three times, and representative results are shown.

were infected at MOIs between 10 and 200, and the number of the viable cells was counted on the 4th day after infection. Figure 5 shows the inhibition of growth of cells exposed to different concentrations of adenovirus. As expected, adenovirus-mediated transfer of p73␤ or p51A/p63␥, like p53, cause growth inhibition of DLD1 cells in a dose-dependent manner. For example, the reduction in viable cells was 30, 65 and 98% for Ad-p53infected DLD1 cells compared with Ad-lacZ-infected cells at the MOI of 20, 50 and 100, respectively (Figure 5). On the other hand, the p53 family genes had little effect on the cell viability of SW480 cells. These results suggest that growth suppression observed in DLD1 cells was caused by apoptosis.

p73␤ or p51A/p63␥ adenovirus did not induced apoptosis in SW480 cells. Twenty-four hours after adenoviral infection, adriamycin was added to the cultures, and cells were analyzed by flow cytometry 3 days after commencement of treatment. As shown in Figure 6a, AdlacZ-infected cells co-treated with adriamycin (0.2 ␮g/ml) displayed a significant G2/M accumulation but no apoptosis, whereas either Ad-p53- or Ad-p51Ainfected cells underwent apoptosis after addition of adriamycin. These results suggest that overexpression of p53 or p51A/p63␥ augmented apoptosis induced by adriamycin. In contrast, adriamycin treatment of Ad-p73␤and Ad-p21-infected cells, both of which initially showed a significant G1 arrest, did not induce apoptosis. Therefore, induction of apoptosis after combination therapy of Ad-p53 or Ad-p51A and adriamycin was not due to the ability of this drug to act on arrested cells, but due to combined effects with gene transduction on chemosensitivity. Moreover, Ad-lacZ-infected cells treated with 8 Gy of X-radiation displayed G1 and G2/M accumulation but no apoptosis, whereas Ad-p53- or Ad-p51A-infected cells underwent apoptosis after exposure of X-ray, suggesting that overexpression of p53 or p51A/p63␥ also augment apoptosis induced by X-radiation. We then examined the effect of the combination therapy on cell viability. As

Effect of combined therapy of adenovirus-mediated transfer of the p53 family gene and adriamycin in SW480 cells Since expression of wild-type p53 is known to accelerate apoptosis induced by DNA-damaging agents, we tested whether Ad-p73␤ or Ad-p51A enhance the chemosensitivity of SW480 cells to DNA damaging agents. In the absence of DNA damaging agents, transduction of p53,

Figure 5 Effect of adenovirus-mediated transfer of the p53 family genes on cell viability. The DLD1 and SW480 cells (2 × 105 cells) were infected at a MOI ranging from 10 to 200. Three days after infection, viable cells were counted by the trypan blue exclusion method using a hemocytometer. Experiments were examined in triplicate. Gene Therapy

Figure 6 Combined effect of the adenovirus-mediated transfer of the p53 family genes with DNA damaging agents on cell cycle progression and cell viability in SW480 cells, which are resistant to apoptosis by adenoviral infection alone. (a) SW480 cells (2 × 105) were infected at a MOI of 100 for 24 h, treated with 0.2 ␮g/ml of adriamycin or exposed to 8 Gy of Xradiation, and then analyzed by flow cytometry after 48 h. Three independent experiments were performed, and representative data from one experiment are shown. (b) SW480 cells (2 × 105) were infected at a MOI of 100 for 24 h, treated with different doses of adriamycin for 72 h, and then assessed for viability by the trypan blue exclusion method. Percent of growth inhibition by Ad-p53, Ad-p73␤, or Ad-p51A relative to AdlacZ is shown above.


shown in Figure 6b, the increased growth suppression by combination therapy was dependent on adriamycin concentration. Prior Ad-p73␤ therapy had a weaker enhancing effect on adriamycin-induced cytotoxicity compared with that of Ad-p53 or Ad-p51A (Figure 6b). A similar increase in growth suppression was also seen in the combination with X-radiation (data not shown). Tumorigenicity of adenovirus-infected colorectal cancer cells To further determine whether the observed inhibitory effects of the p53 family proteins on tumor cell proliferation in vitro could be demonstrated in an in vivo model, we tested the tumorigenic potential of adenovirusinfected colorectal cancer cells in nude mice (Table 1). Viable DLD1 cells (2 × 106/mouse) infected with either Ad-lacZ, Ad-p53, Ad-p73␤ or Ad-p51A for 16 h were subcutaneously injected into seven mice in each treatment group. Tumors started to form around 7 days after injection and were observed in all mice who received injections of Ad-lacZ-infected DLD1 cells. Once tumors were established, they grew aggressively, ultimately reaching an average volume of about 600–900 mm3 within 4 weeks after injection. In contrast, however, no tumors formed in any mice who received injections of DLD1 cells that had been infected with either Ad-p53, Ad-p73␤ or Ad-p51A within the same periods of time, indicating that the tumorigenicity of DLD1 cells was completely suppressed by infection with Ad-p53, Adp73␤ and Ad-p51A. When nude mice were inoculated with HCT116 cells, they grew more aggressively than DLD1 cells. The tumorigenicity of Ad-p51A-infected HCT116 cells was significantly reduced compared with Ad-lacZ-infected HCT116 cells. In contrast, Ad-p53 had a weak activity in reducing tumor size and Ad-p73␤ had a moderate activity (Table 1). This in vivo result correlates with the specificity observed in vitro.

Table 1 In vivo tumorigenicity of adenovirus-infected colorectal cancer cells Mean tumor volume (mm3) ± s.d.

Cell treatment

Day 14 DLD1 lacZ p53a p73␤a p51Aa

282 ± 94 0 0 0

HCT116 lacZ p53 p73␤ p51A

610 327 177 70

± ± ± ±

138 197 135 72

Day 21

Day 28

436 ± 165 0 0 0

659 ± 188 0 0 0

1156 679 353 131

± ± ± ±

138 381 255 71

1793 1148 658 337

± ± ± ±

506 546 334 188

Cells were infected with Ad-lacZ, Ad-p53, Ad-p73␤ and Ad-p51A, respectively. Sixteen hours after adenoviral infection, 2 × 106 cells were injected s.c. into each nude mouse. Mice were scored for tumor formation after the indicated number of days. Results are reported as the mean ± s.d. in seven mice for each treatment group. a No tumor was observed in these mice for an extended period of 35 days.

1405

Discussion p53 gene therapy has been tested clinically for several types of cancers. However, some in vitro and in vivo models are resistant to p53 gene therapy.10–12,29 In this study, we investigated the efficacy of adenovirus-mediated transfer of p73 and p51/p63 gene in comparison with that of p53 for colorectal cancer therapy. We tested a panel of widely studied colorectal cancer cell lines for sensitivity to apoptosis after adenovirus-mediated transfer of p53 family genes. The adenoviral vector used here has the advantage of being applicable for in vivo gene delivery because of its broad host range, high viral titer and high transduction efficiency. Additionally, the ability to produce large quantities of purified adenoviral vector with ease makes this method very attractive for clinical use. Our findings led to several important conclusions regarding the p73 and p51/p63 gene transduction as a useful approach for anti-cancer therapy: (1) the majority of cell lines we tested displayed similar responses following transduction of p53, p73␤ or p51A/p63␥, but less effect was observed following transduction of p73␣ or p51B/p63␣; (2) three cell lines resistant to p53-mediated apoptosis underwent apoptosis after transduction of p73␤ or p51A/p63␥; (3) Ad-p51A, like Ad-p53, in combination with DNA damaging-agents (adriamycin or Xradiation) showed synergistic effects in SW480 colorectal cancer cells, which are resistant to apoptosis when treated with p53 family gene transfer alone; and (4) in vivo tumorigenicity assay revealed that overexpression of p73␤ and p51A/p63␥ efficiently inhibited the tumorigenicity of colorectal cancer cells in nude mice. The majority of cell lines studied here responded similarly to overexpression of the p53 family genes, p53, p73␤ and p51A/p63␥ (Figure 2). These results suggest that a number of genes transcriptionally activated by p53 are also activated by the p53 family gene products, p73␤ and p51A/p63␥ isoforms. Transduction of p53, p73␤ and p51A/p63␥ induced similar growth inhibition in DLD1 and SW480 cells (Figures 3 and 4), however, Ad-p73␤ transfer, unlike Ad-p53 or Ad-p51A, had little effect on endogenous p21 protein in both cell lines (Figure 1B). The heterogeneity in transactivation suggests that some of the genes transcriptionally activated by the p53 family genes are codependent on different transcription factors. Introduction of wild-type p53 induces apoptosis in some cancer cell lines with deleted or mutated p53.30–32 Polyak et al,9 however, demonstrated that the status of endogenous p53 is not strictly associated with the response of colorectal cancer cells to overexpression of exogenous wild-type p53, consistent with our findings using Ad-p53 infection. We also found that the status of endogenous p53 did not correlate with the cell’s response to p73␤ or p51A/p63␥ overexpression (Figure 2). Furthermore, three cell lines, including both cell lines with endogenous wild-type p53 alleles, underwent apoptosis after transduction of p73␤ or p51A/p63␥, but not after transduction of p53. These observations suggest that differences in apoptotic effects of overexpression of the p53 family genes might reflect the genetic characteristics of the recipient cancer cells. Further investigation is required to clarify the mechanisms underlying different responses between the p53 family genes on LoVo, HCT116 and WiDr cells. It is possible that downstream

Gene Therapy


1406

components in the p53-specific signaling pathway may abrogate the apoptotic response. A recent report has demonstrated that p73␣ transduction induces more apoptosis than p53 in U-87MG glioma cells, which express endogenous wild-type p53.33 Since human cancer cells containing the wild-type p53 gene are generally resistant to p53 gene therapy both in vitro and in vivo models,30–32 our results indicate an advantage of p73␤ and p51A/p63␥ gene therapy over p53 gene therapy in tumors expressing wild-type p53. Previous studies have shown that the restoration of wild-type p53 may increase drug- and radiation-sensitivity.34–36 Infection of p53 family gene adenovirus alone did not induce apoptosis in SW480 cells. In contrast, here we performed adenoviral infection in combination with adriamycin or X-radiation and noted induction of apoptosis. We demonstrated that p51A/p63␥ transduction, like p53, increased the sensitivity to adriamycin and X-rays in apoptosis-resistant SW480 cells. The p53 family gene transfer might permit a reduction in the dose of chemotherapeutic agents or radiation and thereby minimize harmful side-effects. These findings justify the investigation of Ad-p51A in addition to Ad-p53 as a single agent or in combination with DNA damaging agents or radiation for the treatment of human cancers. Ad-p73␤ did not enhance the chemosensitivity of SW480 cells to DNA damaging agents (Figure 6), however, there is conflicting data of p73␣ on the resistance to treatment with DNA-damaging agents in human ovarian cancer cell line.37 Although these findings were obtained in one cell line only, therefore have to be considered with caution, specific function of ␣-type isoform of the p73 gene would be attributable to the carboxy-terminal SAM domain. In the present study, a few cell lines displayed apoptosis after infection with Ad-p73␣ or Ad-p51B (Figure 2), and the degree of apoptosis induced by p73␣ and p51B/p63␣ was also weaker than other isoforms, p73␤ and p51A/p63␥, respectively (examples in Figure 3). We previously demonstrated that two isoforms of p73 (p73␣ and p73␤) and p51 (p51A/p63␥ and p51B/p63␣) have distinct abilities to activate the transcription of a p53 responsive reporter gene and to suppress cell growth as assessed by G418 colony formation assay;20 p73␤ and p51A/p63␥ were similar to p53, while p73␣ and p51B/p63␣ had less effect. Similar observations have been reported independently.14,38 Yang et al14 also showed that p51A/p63␥ has strong transactivation activity in reporter systems using p53 response elements. In transient transfection systems, p51A/p63␥ induced cell death with kinetics and features similar to p53induced apoptosis, while p51B/p63␣ did not transactivate the p53-responsive reporter gene and did not induce cell death. Prabhu et al38 showed that p73␣ was a less potent suppressor of SW480 cells in a colony suppression assay. In conclusion, we propose p73 and p51/p63 gene therapy as possible alternatives to p53 gene therapy. The p53 gene is abnormal in more than half of all human cancers, and thus, restoration of the functional p53 protein into these tumors is a desirable for anti-cancer therapy. The effect of retrovirus- or adenovirus-mediated p53 gene transfer into tumors of patients with non-small-cell lung cancer was effective in a subset of patients, however, some cancers exhibited resistance to p53 gene therapy.10–12 Although it should be noted that ample data suggest

Gene Therapy

that one of the major weaknesses of adenovirus-mediated gene therapy is less that optimal delivery, it would be important to override the resistant mechanisms against Ad-p53 gene therapy. Our results indicate that Ad-p51A and probably Ad-p73␤ gene therapy is not only effective, but also a candidate approach to overcome this resistance in some human cancers.

Materials and methods Cell lines and cell culture Ten human colorectal cancer cell lines used in this study were purchased from American Type Culture Collection (Manassas, VA, USA) or the Japanese Collection of Research Bioresources (Osaka, Japan). All cells were cultured under conditions recommended by their respective depositors. SW480 cells were maintained in Leibovitz’s L-15 with 10% fetal bovine serum (FBS) (JRH Biosciences, Lenexa, KS, USA). LoVo cells were grown in Ham’s F-12 medium with 20% FBS. The other eight cell lines HCT116, colo320, DLD1, HCT15, WiDr, HT29, BM314, and CHCY1 were maintained in RPMI 1640 with 10% FBS. Recombinant adenovirus The recombinant adenoviral vectors expressing human p53, p73␣, p73␤, p51A/p63␥, p51B/p63␣ and p21 cDNA and the bacterial lacZ gene were constructed as previously described.20,39 Essentially, we used a replicationdeficient E1-deleted Ad5 vector pJM17 (Microbix Biosystem, Toronto, Ontario, Canada) as a backbone and the corresponding genes were regulated under the control of a human cytomegalovirus promoter/enhancer and a bovine growth hormone polyadenylation signal. Adenovirus titer in plaque forming units (p.f.u.) was determined by plaque formation assays following infection of 293 cells. The multiplicity of infection (MOI) was defined as the ratio of the total number of p.f.u. used in a particular infection to the total number of cells to be infected. To assess efficiency of the adenovirus-mediated gene transfer, cells were infected with Ad-lacZ at a MOI ranging from 10 to 200. After 48 h, Ad-lacZ-infected cells were fixed in 0.05% glutaraldehyde at 4°C for 10 min, washed with PBS, and stained with X-gal (5-bromo-4-chloro-3indolyl-b-d-galactopyranoside), resulting in a blue color in ␤-galactosidase-expressing cells. Immunoblot analysis Five micrograms of each protein lysate were separated by electrophoresis on a 10% SDS/acrylamide gel and transferred to Immobilon P membranes (Millipore, Bedford, MA, USA). After blocking nonspecific binding with 5% milk and 0.1% Tween-20 in PBS, membranes were incubated with either anti-human p53 (clone DO-7, Santa Cruz Biotechnology, Santa Cruz, CA, USA), antihuman p73 (clone ER-15, Oncogene Research, Cambridge, MA, USA), or anti-human p63 monoclonal antibodies (clone 4A4, Oncogene Research) as recommended by the supplier. After incubation with horseradish peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology), the membrane was then processed in ECL Western blotting detection reagents (Amersham Pharmacia Biotech, Buckinghamshire, UK) and exposed to X-ray film.


Flow cytometry and cell growth analysis Cells (2 × 105) were plated in T25 flasks. Twenty-four hours after plating, the cells were incubated with purified virus in 1 ml medium with 1% FBS with brief agitation every 10 min. For flow cytometry analysis, at various times after infection, cells were harvested by trypsinization and pelleted by centrifugation. Pelleted cells were then fixed in 90% cold ethanol, treated with RNase A (500 units/ml), and stained with propidium iodide (50 ␮g/ml). Samples were analyzed on a FACSCalibur flow cytometer (Becton-Dickinson, San Jose, CA, USA). Experiments were repeated at least three times, and 50 000 events were analyzed for each sample. Data were analyzed using the ModFIT model program (BectonDickinson). For cell growth analysis, cells were infected with adenovirus at the MOI of 20, 50, 100, or 200. The number of viable cells was counted using the trypan blue exclusion method and a hemocytometer on the third day after infection. Combination therapy Twenty-four hours after infection, cells were treated with either 0, 0.008, 0.04, or 0.2 ␮g/ml adriamycin, or were exposed to 8 Gy of X-radiation at a dose rate of 3.2 Gy/min using 15 MV X-rays generated by a linear accelerator at room temperature. At various times the cell growth assay and cell cycle analysis were performed as described above. In vivo tumorigenicity assay All animals were maintained, and all animal experiments were conducted according to the principles enunciated in the Guide for the Care and Use of Laboratory Animals prepared by the Office of the Prime Minister of Japan. For the tumorigenicity assay, DLD1 or HCT116 cells were infected with either Ad-lacZ, Ad-p53, Ad-p73␤ or Adp51A at a MOI of 100 for 16 h. Infected cells were harvested by trypsinization, washed in PBS and counted by trypan blue exclusion staining. Viable cells (2 × 106 cells in 200 ␮l) were then injected subcutaneously (s.c.) into the right flank of female BALB/c nude mice, 4–6 weeks of age. Seven mice were used for each treatment group. Tumor formation in mice was observed weekly up to 5 weeks. The tumor volume was calculated using the equation V (mm3) = a × b2/2, where a is the largest dimension and b is the perpendicular diameter.

Acknowledgements We thank Dr Joseph F Costello for critical comments about this manuscript. This research was supported in part by Grants-in-Aid for Cancer Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

References 1 el-Deiry WS. Regulation of p53 downstream genes. Semin Cancer Biol 1998; 8: 345–357. 2 Tokino T, Nakamura Y. The role of p53-target genes in human cancer. Crit Rev Oncol Hematol 2000; 33: 1–6. 3 Vogelstein B et al. Surfing the p53 network. Nature 2000; 408: 307–310. 4 Hollstein M et al. p53 mutations in human cancers. Science 1991; 253: 49–53.

5 Levine AJ et al. The p53 tumour suppressor gene. Nature 1991; 351: 453–456. 6 Clarke AR et al. p53 dependence of early apoptotic and proliferative responses within the mouse intestinal epithelium following gamma-irradiation. Oncogene 1994; 9: 1767–1773. 7 Merritt AJ et al. The role of p53 in spontaneous and radiationinduced apoptosis in the gastrointestinal tract of normal and p53-deficient mice. Cancer Res 1994; 54: 614–617. 8 Roemer K, Friedmann T. Mechanisms of action of the p53 tumor suppressor and prospects for cancer gene therapy by reconstitution of p53 function. Ann NY Acad Sci 1994; 716: 265–280; discussion 280–262. 9 Polyak K et al. Genetic determinants of p53-induced apoptosis and growth arrest. Genes Dev 1996; 10: 1945–1952. 10 Roth JA et al. Retrovirus-mediated wild-type p53 gene transfer to tumors of patients with lung cancer. Nat Med 1996; 2: 985–991. 11 Swisher SG et al. Adenovirus-mediated p53 gene transfer in advanced non-small-cell lung cancer. J Natl Cancer Inst 1999; 91: 763–771. 12 Nemunaitis J et al. Adenovirus-mediated p53 gene transfer in sequence with cisplatin to tumors of patients with non-smallcell lung cancer. J Clin Oncol 2000; 18: 609–622. 13 Kaghad M et al. Monoallelically expressed gene related to p53 at 1p36, a region frequently deleted in neuroblastoma and other human cancers. Cell 1997; 90: 809–819. 14 Yang A et al. p63, a p53 homolog at 3q27–29, encodes multiple products with transactivating, death-inducing, and dominantnegative activities. Mol Cell 1998; 2: 305–316. 15 Osada M et al. Cloning and functional analysis of human p51, which structurally and functionally resembles p53. Nat Med 1998; 4: 839–843. 16 Yang A, McKeon F. p63 and p73: p53 mimics, menaces and more. Nat Rev Mol Cell Biol 2000; 1: 199–207. 17 Jost CA et al. p73 is a human p53-related protein that can induce apoptosis. Nature 1997; 389: 191–194. 18 Zhu J et al. The potential tumor suppressor p73 differentially regulates cellular p53 target genes. Cancer Res 1998; 58: 5061– 5065. 19 Shimada A et al. The transcriptional activities of p53 and its homologue p51/p63: similarities and differences. Cancer Res 1999; 59: 2781–2786. 20 Ishida S et al. Adenovirus-mediated transfer of p53-related genes induces apoptosis of human cancer cells. Jpn J Cancer Res 2000; 91: 174–180. 21 Irwin MS, Kaelin WG Jr. Role of the newer p53 family proteins in malignancy. Apoptosis 2001; 6: 17–29. 22 Malkin D et al. Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science 1990; 250: 1233–1238. 23 Celli J et al. Heterozygous germline mutations in the p53 homolog p63 are the cause of EEC syndrome. Cell 1999; 99: 143–153. 24 Donehower LA et al. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature 1992; 356: 215–221. 25 Yang A et al. p63 is essential for regenerative proliferation in limb, craniofacial and epithelial development. Nature 1999; 398: 714–718. 26 Mills AA et al. p63 is a p53 homologue required for limb and epidermal morphogenesis. Nature 1999; 398: 708–713. 27 Yang A et al. p73-deficient mice have neurological, pheromonal and inflammatory defects but lack spontaneous tumours. Nature 2000; 404: 99–103. 28 el-Deiry WS et al. WAF1, a potential mediator of p53 tumor suppression. Cell 1993; 75: 817–825. 29 Vinyals A et al. Failure of wild-type p53 gene therapy in human cancer cells expressing a mutant p53 protein. Gene Therapy 1999; 6: 22–33. 30 Gomez-Manzano C et al. Adenovirus-mediated transfer of the p53 gene produces rapid and generalized death of human glioma cells via apoptosis. Cancer Res 1996; 56: 694–699.

1407

Gene Therapy


1408

31 Harris MP et al. Adenovirus-mediated p53 gene transfer inhibits growth of human tumor cells expressing mutant p53 protein. Cancer Gene Ther 1996; 3: 121–130. 32 Bookstein R et al. p53 gene therapy in vivo of herpatocellular and liver metastatic colorectal cancer. Semin Oncol 1996; 23: 66–77. 33 Shinoura N et al. Degree of apoptosis induced by adenovirusmediated transduction of p53 or p73␣ depends on the p53 status of glioma cells. Cancer Lett 2000; 160: 67–73. 34 Fujiwara T et al. Induction of chemosensitivity in human lung cancer cells in vivo by adenovirus-mediated transfer of the wildtype p53 gene. Cancer Res 1994; 54: 2287–2291. 35 Gallardo D et al. Adenovirus-based transfer of wild-type p53 gene increases ovarian tumor radiosensitivity. Cancer Res 1996; 56: 4891–4893. 36 Ogawa N et al. Novel combination therapy for human colorectal cancer with adenovirus-mediated wild-type p53 gene transfer and DNA-damaging chemotherapeutic agent. Int J Cancer 1997; 73: 367–370.

Gene Therapy

37 Vikhanskaya F et al. p73␣ overexpression is associated with resistance to treatment with DNA-damaging agents in a human ovarian cancer cell line. Cancer Res 2001; 61: 935–938 38 Prabhu NS et al. p73␤, unlike p53, suppresses growth and induces apoptosis of human papillomavirus E6-expressing cancer cells. Int J Oncol 1998; 13: 5–9. 39 Yamano S et al. Induction of transformation and p53-dependent apoptosis by adenovirus type 5 E4orf6/7 cDNA. J Virol 1999; 73: 10095–10103. 40 Nigro JM et al. Mutations in the p53 gene occur in diverse human tumour types. Nature 1989; 342: 705–708. 41 Rodrigues NR et al. p53 mutations in colorectal cancer. Proc Natl Acad Sci USA 1990; 87: 7555–7559. 42 Cottu PH et al. Inverse correlation between RER+ status and p53 mutation in colorectal cancer cell lines. Oncogene 1996; 13: 2727–2730. 43 Jia LQ et al. Screening the p53 status of human cell lines using a yeast functional assay. Mol Carcinog 1997; 19: 243–253.