Enhanced Apoptotic Activity of a p53 Variant in Tumors Resistant to Wild-Type p53 Treatment

doi:10.1006/mthe.2001.0416, available online at http://www.idealibrary.com on IDEAL

ARTICLE

Enhanced Apoptotic Activity of a p53 Variant in Tumors Resistant to Wild-Type p53 Treatment Isabella A. Atencio,1,* Jenny B. Avanzini,1 Duane Johnson,1 Saskia Neuteboom,2 Mei-Ting Vaillancourt,1 Loretta L. Nielsen,3 Gerry Hajian,3 Suganto Sutjipto,1 Barry J. Sugarman,1 Jennifer Philopena,1 Diane L. McAllister,1 Josefina C. Beltran,1 Margarita Nodelman,1 Murali Ramachandra,1 and Ken N. Wills1 1Canji,

Inc., 3525 John Hopkins Court, La Jolla, California 92121, USA 2NewBiotics, San Diego, California 92121, USA 3Schering-Plough Research Institute, Kenilworth, New Jersey 07033, USA *To whom correspondence and reprint requests should be addressed. Fax: (858) 623-2032. E-mail: [email protected].

TP53 is the most commonly altered tumor-suppressor gene in cancer and is currently being tested in Phase II/III gene replacement trials. Many tumors contain wild-type TP53 sequence with elevated MDM2 protein levels, targeting p53 for degradation. These tumors are more refractory to treatment with exogenous wild-type p53. Here we generate a recombinant adenovirus expressing a p53 variant, rAd-p53 (d 13–19), that is deleted for the amino acid sequence necessary for MDM2 binding (amino acids 13–19). We compared the apoptotic activity of rAd-p53 (d 13–19) with that of a recombinant adenovirus expressing wild-type p53 (rAd-p53) in cell lines that differ in endogenous p53 status. rAd-p53 (d 13–19) caused higher levels of apoptosis in p53 wild-type tumor lines compared with wild-type p53 treatment, as measured by annexin V-FITC staining. In p53-altered tumor lines, rAd-p53 (d 13–19) showed apoptotic activity similar to that seen with wild-type p53 treatment. In normal cells, no increase in cytopathicity was detected with rAd-p53 (d 13–19) compared with wild-type p53 treatment. This variant protein displayed synergy with chemotherapeutic agents to inhibit proliferation of ovarian and breast cell lines. The p53 variant showed greater antitumor activity in an established p53 wild-type tumor compared with treatment with wild-type p53. The p53 variant represents a means of expanding TP53 gene therapy to tumors that are resistant to p53 treatment due to the cellular responses to wild-type p53. KEY WORDS: adenovirus, p53, variant, MDM2, apoptosis, antitumor, efficacy, gene therapy, chemotherapeutics

INTRODUCTION The protein p53 is involved in many biochemical pathways leading to human carcinogenesis. p53 regulates transcription of genes whose protein products regulate cell-cycle progression and apoptosis [1–5]. Loss of p53mediated apoptosis is an important event in tumor progression in a number of systems [reviewed in 6]. The synergistic action of p53 and chemotherapy in the induction of apoptosis [7–9] has raised the possibility that the reactivation of p53 in transformed cells can be an effective tumor therapy [10–12]. The use of a replication-deficient recombinant adenovirus encoding wild-type p53 (rAdp53) has been evaluated in clinical trials for a variety of cancer [8,13–16]. Previous reports suggested that the p53 status of tumor cells may influence the response to gene transduction MOLECULAR THERAPY Vol. 4, No. 1, July 2001 Copyright © The American Society of Gene Therapy

with adenovirus delivery of wild-type p53 [17,18]. It has been proposed that tumor cells expressing endogenous wild-type p53 are less responsive to cell-cycle arrest and apoptosis, and are less efficiently tranduced by adenovirus [19]. Increasing data, however, support the view that the stability of p53 is controlled through binding of this protein to the human homologue of the mouse double minute-2 protein, MDM2 [20–22]. Targeting p53 for degradation has been shown to be the most effective way for MDM2 to inhibit the apoptotic function of p53 [22]. The binding of p53 to MDM2 promotes the rapid degradation of p53; binding leads to ubiquitination [23] and subsequent proteolytic degradation of p53 by proteasomes [24]. The human gene MDM2 has been shown to contain a p53 DNA-binding site and its expression can be regulated by the level of wild-type p53 protein.

5

ARTICLE

A

doi:10.1006/mthe.2001.0416, available online at http://www.idealibrary.com on IDEAL

B

FIG. 1. Analysis of MDM2 association and p53 protein levels. (A) Immunoprecipitation for p53 protein and western blot analysis for MDM2 protein in tumor cells infected with virus constructs. PC-3 cells were infected with 1  109 P/ml rAd-p53, p53 (d 13–19), or rAd-control for 24 h. Immunoprecipitation for p53 protein was carried out on cell lysates, followed by western blot analysis for MDM2. (B) Western blot analysis for p53 was done on PC-3 cells infected for 24, 48, or 72 h with 1  109 P/ml rAdp53, rAd-p53 (d 13–19), or rAd-control.

The MDM2 protein, in turn, complexes with p53 and decreases its ability to act as a positive transcription factor. In this way, MDM2 is autoregulated. The p53 protein regulates MDM2 at the level of transcription and the MDM2 protein regulates p53 at the level of its activity. This creates a feedback loop that regulates both the activity of p53 and the expression of MDM2. A wide variety of human tumors overexpress the human analogue of the mouse MDM2 protein, human MDM2 [25–29]. Human tumors with endogenous wild-type p53 overexpress human MDM2 by enhanced translation [30] or amplification of the gene [31], preventing p53-mediated apoptosis [32]. High MDM2 levels may, therefore, provide an alternative explanation for the resistance to exogenous p53-mediated apoptosis. Recombinant adenoviruses expressing p53 variants that lack the MDM2-binding domain were shown to have enhanced apoptotic [33] or antiproliferative activity [34] in tumor cells with endogenous wild-type p53. These modified p53 proteins, however, were less effective then wild type in inducing apoptosis in tumor cells with altered p53 phenotypes [33] or have not been adequately assessed for apoptotic activity [34]. Here we report on a recombinant adenovirus containing a modified p53 cDNA with deletion of nucleotide sequences encoding amino acids 13–19 ((rAdp53 (d 13–19)). Previous publications have shown abolished MDM2 binding with this modified protein [20]. Treatment with rAd-p53 (d 13–19) resulted in an increased sensitivity to apoptosis in cell lines and in established tumors with endogenous wild-type p53 that were more resistant to rAd-p53 treatment. Unlike previous results, rAd-p53 (d 13–19) retained equivalent apoptotic activity compared with rAd-p53 in cell lines with altered p53 status. Our results indicate that endogenous p53 status can predict the response of tumor cells to adenoviral delivery of exogenous p53. They also demonstrate the

6

FIG. 2. Induction of apoptosis in cell lines with wild-type p53 status. Cell lines with endogenous p53 wild-type status (SK-Hep1, A549, and U87) were uninfected (UI), or infected with 1  109 P/ml rAd-p53, rAd-p53 (d 13–19), or rAd-control. Cells collected at 48 h post-infection were monitored for apoptosis by labeling with annexin V-FITC, followed by flow cytometric analysis. The response (mean + standard deviation) is depicted as the percentage of cells positive for annexin V. Error bars represent standard deviations from three independent experiments.

potential use of p53 variants to expand TP53 gene therapy to include tumors resistant to treatment with wildtype p53.

RESULTS p53 (d 13–19) Lacks MDM2 Binding and Has Increased p53 Levels We first determined whether the recombinant mutant p53 protein expressed in cells infected with rAd-53 (d 13–19) retained the capacity to bind MDM2. The TP53null cell line, PC-3, was infected with rAd-p53 (d 13–19), rAd-p53, or control virus expressing no transgene (rAdcontrol). At 24 hours post-infection, we assayed cell lysates for binding of p53 to MDM2 by immunoprecipitation (Fig. 1A). Lysates from PC-3 infected with rAd-p53 showed co-precipitation of MDM2 with p53, whereas cells infected with rAd-p53 (d 13–19) showed no co-precipitation of MDM2. Lysates from uninfected cells and cells infected with a recombinant adenovirus expressing no transgene (rAd-control) showed no co-precipitation of MDM2. Ribonuclease protection assays showed equivalent message for the p53 transgene from both constructs, and a western blot analysis for MDM2 protein confirmed that cells infected with rAd-p53 (d 13–19) upregulated MDM2 (data not shown). In lieu of these results, we assessed the steady-state level of recombinant p53 in lysates using either a western blot assay or a p53 ELISA. We saw elevated levels of p53 (d 13–19) protein over time compared with wild-type p53 (Fig. 1B). After 24 hours, p53 was detected in cells infected with either rAd-p53 (d 13–19) or rAd-p53. Levels of p53 remained elevated at 48 hours in cells infected with rAd-p53 (d 13–19), but declined in cells infected with rAd-p53. At 72

MOLECULAR THERAPY Vol. 4, No. 1, July 2001 Copyright © The American Society of Gene Therapy

ARTICLE

doi:10.1006/mthe.2001.0416, available online at http://www.idealibrary.com on IDEAL

FIG. 3. Induction of apoptosis in cell lines with altered p53 status. Cell lines with altered p53 status (HLF, HLE, Hep3B, and H358) were uninfected (UI), infected with 1  108 P/ml rAdp53, rAd-p53 (d 13–19) or rAd-control (A), or 1  109 P/ml rAd-p53, rAd-p53 (d 13–19) or rAd-control (B). Cells were collected 48 h post-infection and apoptosis was monitored by labeling cells with annexin V-FITC. The percentage of annexin V positive cells in the population are plotted on the y axis. Error bars represent standard deviations from three independent experiments.

A

B

hours post-infection, we detected lower p53 protein levels in lysates from cells infected with either rAd-p53 (d 13–19) or rAd-p53, although levels were substantially higher in rAd-p53 (d 13–19) infected lysates. We carried out an ELISA test to quantify higher levels of p53 (d 13–19) in the TP53-null cell line H358. There was an approximately twofold increase in p53 levels (17.9 ng/ml for p53 (d 13–19); 9.9 ng/ml for wild-type p53).

p53, we infected various p53-altered cell lines. Apoptosis was as efficiently induced by rAd-p53 (d 13–19) as by rAd-p53 (Fig. 3), regardless of endogenous p53 status in all cell lines assessed at two concentrations of virus: 1  108 P/ml and 1  109 P/ml. Similar results were obtained with the glioblastoma lines U251 and G112, and the colorectal DLD-1, all of which have endogenous mutant p53.

rAd-p53 (d 13–19) Enhances Apoptosis in Cell Lines with Endogenous Wild-Type p53 It has been proposed that cell lines expressing endogenous wild-type p53 may be more resistant to apoptosis mediated by rAd-p53 because they are less efficiently transduced [19]. To ascertain if such cells are more resistant to exogenous p53-mediated apoptosis, we infected different cell lines expressing endogenous wild-type p53 with rAd-p53 (d 13–19), rAd-p53, or a control virus, and assayed for apoptosis 48 hours post-infection. All cell lines tested showed a greater percentage of cells undergoing apoptosis in response to rAd-p53 (d 13–19) (Fig. 2), as detected by an increase in the percentage of cells binding annexin V-FITC. Other cell lines showing enhanced sensitivity to apoptosis in response to rAd-p53 (d 13–19) included the colorectal cell line RKO and the hepatocellular carcinoma cell line Hep G2. For RKO cells, 55 + 4% of the cells were apoptotic in response to rAd-p53 (d 13–19), whereas 22 + 3% of the cells were apoptotic in response to rAd-p53. For HepG2 infected with rAd-p53 (d 13–19), 45 + 2% of the cells were apoptotic. Infection of these cells with rAd-p53 resulted in 20 + 3% of the cells being apoptotic (data not shown).

Activation of Caspase-9 in Cell Lines That Differ in Endogenous p53 Status in Response to rAd-p53 (d 13–19) To show that the apoptotic pathway activated by rAdp53 (d 13–19) was similar to p53-induced apoptotic pathways, we assessed for caspase-9 activation [35] in SK-Hep1, HLF, HLE, and H358 cells infected with rAdp53 (d 13–19) or rAd-p53. After 48 hours, apoptotic cells were detected by annexin V-FITC staining (data not shown). We analyzed cell lysates prepared from these infected populations by incubating with a specific caspase-9 substrate. Infection with either rAd-p53 (d 13–19) or rAd-p53 resulted in a three- to sevenfold increase in activity of caspase-9 (Fig. 4). Control virus elicited no increase in caspase-9 activity compared with uninfected cells.

rAd-p53 (d 13–19) Induces Apoptosis in Cell Lines That Differ in Endogenous p53 Status Another adenovirus containing p53 variant has also been shown to abolish MDM2 binding [33], but it was less effective than wild-type p53 at inducing apoptosis in tumor cells with either an endogenous p53 mutant or a null phenotype. To assess whether the rAd-p53 (d 13–19) described here induced apoptosis as efficiently as rAd-

MOLECULAR THERAPY Vol. 4, No. 1, July 2001 Copyright © The American Society of Gene Therapy

No Increase in rAd-p53 (d 13–19) Mediated Death of Primary Cells We next determined if primary cells undergo enhanced cell killing in response to infection with rAd-p53 (d 13–19). Four different primary cell types were grown in vitro and infected with rAd-p53, rAd-p53 (d 13–19), or control virus. Mammary and bronchial epithelial cells showed no enhanced cell death after treatment with rAdp53 (d 13–19) infection compared with rAd-p53 at lower virus concentrations (Fig. 5). At only the highest concentrations (≥ 3E+08 P/ml) of virus was rAd-p53 (d 13–19) more effective in reducing the cell viability, but the difference in concentration of virus needed to induce apoptosis in 50% of the cells was less then twofold. Treatment of primary lung microvascular endothelial

7

ARTICLE

doi:10.1006/mthe.2001.0416, available online at http://www.idealibrary.com on IDEAL

to rAd-p53, and 60 ± 1 of control in response to rAd-p53 (d 13–19). At 1  109 P/ml, fibroblast viability was 30 ± 1% of control in response to rAd-p53, and 25 ± 1% of control in response to rAd-p53 (d 13–19). Again the difference in concentration of viruses needed to induce apoptosis in 50% of the cells was less then twofold. Treatment of cells with the control virus did not result in reduction of cell viability.

FIG. 4. Caspase-9 activation in response to rAd-p53 (d 13–19). Cells with wild-type (SK-Hep1) or altered (HLF, HLE, H358) p53 were uninfected (UI), or infected with 1  109 P/ml rAd-p53, rAd-p53 (d 13–19), or rAd-control. Infected cells were then analyzed for caspase-9 activation 48 h post-infection. Relative fluorescence units are plotted on the y axis. Error bars represent standard deviations from three independent experiments.

cells and lung fibroblasts with the highest doses of virus (3  108 to 1  109 P/ml) revealed no enhanced killing when treated with rAd-p53 (d 13–19) compared with rAd-p53. Endothelial cell viability, expressed as the percentage of uninfected cell control, was reduced to 45 ± 3% by rAd-p53 (d 13–19) and 65 ± 25% by rAd-p53 at a dose of 3  108 P/ml. At 1  109 P/ml, cell viability was 20 ± 2% of untreated cell control for rAd-p53 (d 13–19) and 30 ± 25% of control for rAd-p53. Lung fibroblast cells at 3  108 P/ml showed 75 ± 1% of control in response

Tumors Injected with Recombinant Adenovirus Show Enhanced Levels of p53 (d 13–19) To determine relative levels of protein expression in vivo, we treated immunosuppressed mice with preestablished xenograft tumors in their flanks with five intra-tumoral injections of rAd-p53 (d 13–19), rAd-p53, or the control virus expressing the green fluorescent protein transgene (rAd-GFP). We harvested tumors and assayed for p53 protein 48 hours after the last injection. Immunohistochemistry (Fig. 6) showed higher levels of p53 after treatment with rAd-p53 (d 13–19) compared with tumors treated with rAd-p53. Tumor sections from mice injected with control virus showed background levels of p53 staining. Enhanced Sensitivity of Tumors to rAd-p53 (d 13–19) Mediated Inhibition of Tumor Growth To compare the effects of rAd-p53 and rAd-p53 (d 13–19) in tumors with endogenous wild-type p53 status, we carried out in vivo tumor efficacy studies. Nude mice bearing preestablished flank SK-Hep1 tumors were injected with recombinant adenovirus and tumor measurements were

FIG. 5. Effect of rAd-p53 and rAd-p53 (d 13–19) on viability of primary human cells. Dilutions of adenoviral constructs were added to mammary and bronchial epithelial, and lung endothelial and fibroblast primary cells. At the end of the 3-d incubation period, cell viability was assessed using the MTS reagent. The response at each concentration (and for each type of virus) was determined and is depicted as a percentage of the uninfected control. Triangles, rAd-control; circles, rAd-p53; squares, rAd-p53 (d 13–19). Error bars represent standard deviations from experiments performed in triplicate.

8

MOLECULAR THERAPY Vol. 4, No. 1, July 2001 Copyright © The American Society of Gene Therapy

doi:10.1006/mthe.2001.0416, available online at http://www.idealibrary.com on IDEAL

ARTICLE

FIG. 6. Immunohistochemistry for p53 detection in flank tumors. SK-Hep1 flank tumors in Balb/c nude mice were injected with rAd-p53, rAd-p53 (d 13–19), or control virus rAd-GFP, each given by intra-tumoral infection daily for 5 d. Tumors from two animals per group were removed 48 h after the last virus injection and assayed for p53 protein expression by immunohistochemistry. The pink color is representative of p53 expression, whereas the blue is a DAPI nuclear counter stain. Representative images are at ×100 and ×400.

taken weekly over seven weeks. The greatest difference between the two p53-containing constructs in antitumor efficacy occurred at 7 weeks after the last virus injection at concentrations of 5  108 and 1  109 P/ml (Fig. 7A). At the higher doses of virus (5  109 and 1  1010 particles; data not shown), there were no statistical differences in anti-tumor efficacy between the constructs. rAd-p53 (d 13–19) treatment resulted in statistically greater anti-tumor efficacy at these lower doses (P < .01) compared with tumors treated with rAd-p53 (Fig. 7B).

rAd-p53 (d 13–19) Is Also Synergistic with Paclitaxel and Cisplatin at Inhibiting Proliferation in Vitro To determine if rAd-p53 (d 13–19) retained activity similar to that of rAd-p53 in response to chemotherapeutics, two ovarian cell lines were treated with either cisplatin or paclitaxel in combination with recombinant adenovirus. rAdp53 (d 13–19) synergized with paclitaxel in both cell lines tested (Table 1). It also synergized with cisplatin in SK-BR3 cells. rAd-p53 synergized with cisplatin, but was only additive with paclitaxel in SK-BR-3 cells (Table 1). We previously showed synergy between rAd-p53 and paclitaxel in six of seven tumor cell lines tested, independent of p53 status [7]. Table 1 Effects of chemotherapeutic agents in combination with rAd-p53 and rAd-p53 (d 13–19) on tumor cell proliferation DISCUSSION Origin p53 Status Agent 1 Agent 2 Isobole Analysis Reports suggest that endogenous human ovary wild-type rAd-p53 (d 13-19) paclitaxel synergy (p=0.0658) p53 status can predict a cell’s human breast mutant rAd-p53 (d 13-19) paclitaxel synergy (p=0.0004) response to a recombinant adenhuman breast mutant rAd-p53 (d 13-19) cisplatin synergy (p=0.0279) ovirus expressing a wild-type p53 transgene [36,37]. We have human breast mutant rAd-p53 paclitaxel additive (p=0.2243) shown that a recombinant adenhuman breast mutant rAd-p53 cisplatin synergy (p=0.0239) ovirus expressing a p53 variant PA-1 and SK-BR-3 tumor cells were incubated with varying concentrations of rAd-p53 or rAdp53 (d 13–19) and/or paclithat is deleted for the sequences taxel or cisplatin for 3 d. Multiple dose-response curves were generated for each agent alone and in combination, from zero to maximal response, for each individual cell line. Cell proliferation was quantitated using the MTT assay. Data necessary for MDM2 binding can were analyzed using the Thin Plate Spline methodology. induce apoptosis more efficiently

MOLECULAR THERAPY Vol. 4, No. 1, July 2001 Copyright © The American Society of Gene Therapy

9

ARTICLE

A

doi:10.1006/mthe.2001.0416, available online at http://www.idealibrary.com on IDEAL

B

This study supports the premise that the cellular responses to p53 overexpression, rather than differing transduction efficiencies of cell lines [19], lead to higher resistance to rAd-p53 mediated effects in cells with endogenous wild-type p53 status. These results, as well as others [33], showing enhanced apoptosis in response to p53 variants lacking MDM2 binding indicate that endogenous p53 status can be predictive of the response via MDM2 regulation. FIG. 7. In vivo anti-tumor activity of rAd-p53 (d 13–19). (A) Anti-tumor activity over time of rAd-p53 and rAd- As both of these constructs were p53 (d 13–19). Subcutaneous SK-Hep1 tumors were established in nude mice. A cycle of 5 intra-tumoral injecmade in identical adenoviral tions with rAd-p53, rAd-p53 (d 13–19), or control virus rAd-GFP were given every other day. The amounts of virus injected ranged from 5E+08 (5  108) to 1E+10 (1  1010) particles. A high dose of the control virus, con- backbones and propagated in taining the transgene coding for green fluorescent protein (rAd-GFP) at 1E+10 particles was used. An addition- the same type of packaging cell al five animals received vehicle injections (V) of phosphate-buffered saline (PBS). Tumor sizes (y axis) were deter- line (293), we anticipate that mined over a period of seven weeks from measurements taken weekly following the initial virus injections. X, replication competent aden1E+10 particles rAd-GFP; –, PBS; open squares, 5E+08 particles rAd-p53; filled squares, 5E+08 particles rAd-p53 9 (d 13–19); open triangles, 1E+09 (1  10 ) particles rAd-p53; filled triangles, 1E+09 particles rAd-p53 (d ovirus (RCA) would occur at the 13–19); open circles, 5E+09 (5  109) particles rAd-p53; filled circles, 5E+09 particles rAd-p53 (d 13–19). Error same frequency for the two conbars represent standard deviations generated from five animals per dose per group. (B) Anti-tumor activity of structs, and would not account rAd-p53 and rAd-p53 (d 13–19). Tumor sizes (y axis) are from measurements taken 7 weeks following the ini- for the enhanced apoptosis tial virus injections. P values (*) were determined from a two tailed t test analysis assuming equal variance. observed with rAd-p53 (d 13–19). Substantial RCA contamination in the rAd-p53 (d 13–19) virus preparation would, in addition, lead to differences than a vector expressing wild-type p53 in cells with wildin transgene expression from these two constructs, type 53 status. A variety of cell lines expressing endogewhich did not occur (data not shown). nous wild-type p53 were more susceptible to rAd-p53 (d Our results also indicate that a variety of human 13–19) mediated apoptosis compared with infection by tumor cells with endogenous wild-type p53 status rAd-p53. Unlike the p53 variant CTS1 [33], the p53 (d express the inhibitor of p53, human MDM2, which may 13–19) variant retained similar apoptotic activity in cell lead to the inactivation of exogenous p53. lines with altered p53 status. Approximately 50% of human tumors have an endogep53 (d 13–19) was shown to lack MDM2 binding and nous wild-type TP53 genotype/phenotype [reviewed in had increased p53 protein levels over wild-type p53. We 39] and may not be successfully targeted with gene therhave shown by activation of caspase-9 that rAd-53 (d apy using wild-type TP53 [37]. Our report indicates that 13–19) infection of cells leads to p53-mediated apoptothe patient population targeted for TP53 gene therapy sis [38]. In vivo in mice, xenograft tumors expressing may be expanded to include those with tumors refractoendogenous wild-type p53 were treated with rAd-p53 (d ry to wild-type TP53 treatment due to cellular responses 13–19) or rAd-p53. Enhanced levels of p53 (d 13–19) to the wild-type protein. Our results showed no appreprotein compared with wild-type p53 were seen by ciable increase in cytopathicity with this p53 (d 13–19) immunohistochemistry. These tumors showed greater variant in a variety of primary cells, as well as the ability sensitivity to rAd-p53 (d 13–19) mediated inhibition of of this variant to retain synergism with paclitaxel and cisgrowth compared with rAd-p53. rAd-p53 (d 13–19) platin in the induction of cell cycle arrest in tumor cell treatment resulted in statistically greater (P < .01) antilines. Furthermore, our results also imply that the dose of tumor efficacy at the lower doses of virus compared with virus needed to inhibit tumor growth may be reduced. tumors treated with rAd-p53. At the higher doses of viruses (5  109 and 1  1010 particles), there were no statistical differences in anti-tumor efficacy between the MATERIALS AND METHODS constructs. The greatest difference in anti-tumor efficacy occurred at seven weeks following the last virus injection, supporting the premise that accumulation of p53 (d 13–19) occurs over time, leading to p53-mediated anti-tumor effects.

10

rAd-p53 (d 13–19) vector construction and purification. All adenovirus vectors were constructed in the same backbone. A recombinant adenoviral vector encoding wild-type p53 (rAd-p53) has been described [40]. We constructed a rAd-p53 with deletion of nucleic acid sequences encoding amino acids 13–19. The p53 (d 13–19) fragment was generated by inverse

MOLECULAR THERAPY Vol. 4, No. 1, July 2001 Copyright © The American Society of Gene Therapy

doi:10.1006/mthe.2001.0416, available online at http://www.idealibrary.com on IDEAL

PCR on wild-type TP53 sequence using primers that excluded the region encoding amino acids 13–19. In addition, silent mutations were incorporated at amino acids 24 and 25 to create a HindIII recognition site used for screening purposes. Gel isolation and subsequent insertion of this amplicon into a cloning vector by blunt end ligation was done using standard techniques. A fragment encoding p53 (d 13–19) was removed by digesting with XbaI and ClaI and inserted into the adenovirus transfer plasmid pACN [40] digested with the same enzymes. This plasmid was linearized using XmnI and NgoMI and co-transformed with a ClaI-digested plasmid designated pTG4213 (Transgene) into the Escherichia coli strain BJ 5183 [41]. Plasmid pTG4213 encodes the entire Ad 5 sequence, except an E1 deletion of ~2.9 kb and an E3 deletion of ~1.9 kb [41]. Homologous recombination resulted in a plasmid encoding the entire Ad sequence [40]. The Ad DNA was modified by an E1 deletion where the Ad sequence was replaced by an expression cassette encoding the p53 (d 13–19) sequence driven by the human CMV promoter/enhancer and Ad 2 TPL from plasmid ACN and the E3 deletion described above. This plasmid was then used to transfect 293 cells to produce viable rAd-p53 (d 13–19). Identity was confirmed by sequencing and restriction mapping of the entire viral genome. Construction of a replication-deficient adenovirus vector encoding green fluorescent protein (rAd-GFP) has been described [42]. Viruses were propagated in the 293 cell line, purified by column chromatography, and quantitated based on infectious virus particles per milliliter (P/ml), as described [38,43]. All virus infections performed were based on P/ml, rather than multiplicity of infection, based on previous publications [38,43], and the Food and Drug Administration guidance for dosing of adenovirus (Guidance for Human Somatic Cell Therapy and Gene Therapy, Center for Biologics Evaluation and Research, March 1998). We assessed the ratio of infectious to non-infectious particles and determined that there was no difference between the two constructs. Primary cells and cell lines. Primary human mammary epithelial cells and lung fibroblasts were purchased from Clonetics (San Diego, CA) and cultured in specific growth medium (Clonetics). Tumor cell lines SK-Hep1 (#HTB-52), Hep3B (Hep3 B2.1-7, #HB-8064), U87 MG (#HTB-14), A549 (#CCL 185), H358 (#CRL 5807), PA-1 (#CRL 1572), and Hep G2 (#HB 8065) were obtained from the American Type Culture Collection (Manassas, VA). HLE and HLF cell lines were obtained from the Japanese Cancer Research Resource Bank (Tokyo, Japan). The RKO line was obtained from M. Brattain, Medical College Hospital (Toledo, OH). These cells were cultured in Dulbecco’s modified Eagle’s medium (DME) supplemented with sodium pyruvate and 10% fetal bovine serum. PC-3 (#CRL 1435) cells were grown in Ham’s F12/DME supplemented with sodium pyruvate and 10% fetal bovine serum (FBS). The PA-1 cells were cultured in Eagle’s minimal essential media with non-essential amino acids and Earle’s BSS plus 10% fetal calf serum (FCS) at 37°C and 5% CO2. The SK-BR-3 human breast tumor cells were cultured in Isocove’s MEM with 15% FCS at 37°C and 5% CO2. Apoptosis assay. Cell lines were infected with 1  109 particles/ml (P/ml) [38] purified recombinant adenovirus for 1 h, at which time virus was removed and fresh medium was added. Cells were collected 48 h postinfection. Apoptosis was monitored visually by observing blebbed nuclei characteristic of apoptosis, by propidium iodide staining (Molecular Probes) followed by flow cytometric analysis to quantitate populations of cells with subgenomic DNA, and by labeling cells with annexin V-FITC (Boehringer Mannheim), followed by flow cytometric analysis [44]. Cell viability assay. Primary human mammary epithelial cells and lung fibroblast cells were seeded at 5  103 cells per well in 96-well plates (Nunc) and incubated until 50–75% confluent. Dilutions of adenoviral constructs ranging from 1  109 to 3  106 P/ml were added to the cells. Plates were incubated at 37°C in 7% CO2 for 1 h, after which the viruses were removed. Cells were then fed with 200 µl growth medium and reincubated at 37°C. Three days later, 40 µl MTS reagent (Promega) was added to each well and the plates were incubated for an additional 1–4 h at 37°C. The absorbance was read using a SpectraMax 250 plate reader (Molecular Devices, Sunnyvale, CA) at 490 nm. The values of the treated wells were calculated as a percentage of untreated control. Western blot assays. PC-3 cells were infected for 1 h with 1  109 P/ml of virus. Virus was removed and the cells were fed with fresh growth media. At

MOLECULAR THERAPY Vol. 4, No. 1, July 2001 Copyright © The American Society of Gene Therapy

ARTICLE

24, 48, and 72 h post-infection, cells were lysed in 50 mM Tris, 250 mM NaCl, 50 mM NaF, 5 mM EDTA, 0.1% NP-40, and 1 mM PMSF. Approximately 10 µg of each lysate was loaded and resolved on a 4–20% polyacrylamide gel and transferred onto nitrocellulose. These membranes were subjected to western blot analysis using antibodies specific for p53 and β-actin (Calbiochem, p53 antibody Ab-6, mouse monoclonal to β-actin). Blots were incubated for 1 h with HRP-conjugated secondary antibodies. The enhanced chemiluminescence detection system was used for development. p53 ELISA. H358 cells were infected with 1  109 P/ml with either rAd-p53 or rAd-p53 (d 13–19) for 1 h, at which time virus was removed and fresh medium was added. Cells were collected by scraping 24 h post-infection, washed in PBS, then pelleted at 660g for 2 min. Cell pellets were lysed, and the ELISA was performed according to the manufacturer’s instructions (Oncogene Research). Sample ODs were assessed by measuring the ratio of ODs at 450/540 nm for p53 concentration by the use of Spectromax software to calculate mean optical density values for sample replicates by interpolation from a standard curve. Immunoprecipitation assay. PC-3 cells were infected with recombinant adenovirus at 1  109 P/ml. At 24 h post-infection, cells were lysed in 50 mM Tris, 250 mM NaCl, 50 mM NaF, 5 mM EDTA, 0.1% NP-40, and 1 mM PMSF. Total protein (500 µg) was added to 10 µg anti-p53 antibody for 2 h at 4°C. 40 µl of Sephrose G beads (Santa Cruz) were added, and incubation was done for 2 h at 4°C. Lysates were resolved on a 4–20% SDS gel, transferred onto nitrocellulose and probed with a mouse monoclonal anti-MDM2 antibody. Blots were incubated for 1 h with an HRP-conjugated secondary antibody. Blots were developed using the enhanced chemiluminescence detection system. Caspase activation assay. HLF, HLE, SK-Hep1, and H358 cell lines were treated with 1  109 P/ml rAd-p53 for 1 h, then washed with fresh media and incubated for 48 h. Approximately 1  106 cells per assay were lysed in 50 µl cell lysis buffer (Clontech). To these lysates, 20 mM substrate (Enzyme Systems) for caspase-9 (Ac-LEHD-AFC) was added and incubated at 37°C for 1 h. Arbitrary relative fluorescence units were determined using a Cytofluor fluorescence multi-well plate reader (PerSeptive Biosystems) using a 400-nm excitation filter and 505-nm emission filter. Paclitaxel and cisplatin treatment. For in vitro drug interaction studies, cells were aliquoted into culture wells and allowed to attach for 3 h. The cells were then incubated with varying concentrations of rAd-p53 (d 13–19) and/or drug (paclitaxel (Taxol, Calbiochem, La Jolla, CA) or cisplatin (Sigma Chemical Co., St. Louis, MO)) for 3 d. Multiple dose-response curves were generated for each agent alone or in combination, from zero response to maximal response, for each individual cell line. Cell proliferation was quantitated using the MTT assay [19]. Briefly, 25 µl of 5 mg/ml MTT vital dye (3-(4,5 dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) was added to each well and allowed to incubate for 3–4 h at 37°C and 5% CO2. Then, 100 µl of 10% SDS was added to each well and the incubation was continued overnight. Fluorescence in each well was quantitated using a Molecular Devices microtiter plate reader. Data were analyzed using the Thin Plate Spline methodology [45]. Tumor mouse model and treatment. SK-Hep1 cells were inoculated subcutaneously into the flanks of female Balb/c athymic nude mice (5  106 cells/200 µl). Ten days later, animals were randomized by tumor size (approximately 70 mm3). Next a cycle of five intra-tumoral injections was administered every other day (100 µl volume per injection). Particle number of viruses used ranged from 5  108 to 1  1010. Only a high dose of the control virus containing the transgene coding for green fluorescent protein (rAd-GFP) was used. Five animals were used per group. An additional five animals received PBS control injections as a control. Tumor dimensions (length, width, and height) were taken weekly over a 7-week period of time following the initial virus injections. Tumor volumes were estimated for each animal, assuming a spherical geometry with radius equal to one-half the average of the measured tumor dimensions. Immunohistochemistry for p53 detection. SK-Hep1 cells were inoculated into Balb/c nude mice as mentioned above. Two animals per group, treated with rAd-p53, rAd-p53 (d 13–19), or control virus rAd-GFP, each dosed at

11

ARTICLE

doi:10.1006/mthe.2001.0416, available online at http://www.idealibrary.com on IDEAL

5  109 P/injection on five consecutive days, were sacrificed 48 h after the last virus injection. Tumors were assayed for p53 protein by immunohistochemical staining using formalin-fixed, paraffin-embedded tissues. Tissues were sectioned (5 µm) onto Superfrost Plus slides (Fisher Scientific, Pittsburgh, PA). Slides were de-paraffined in Propar clearant (Anatech Ltd., Battle Creek, MI) and re-hydrated through decreasing alcohol steps to water. Antigen retrieval was performed in a microwave using Zymed citrate buffer, pH 6.0 (Zymed Laboratories Inc., San Francisco, CA), in deionized water. Tissues were blocked with Zymed CAS Block. The primary antibody used was anti-p53 CM1 rabbit polyclonal (Novocastra Laboratories Ltd, UK), and was applied at a 1:1000 dilution in PBS and incubated in a humidity chamber for 60 min and washed with PBS. A Texas red conjugated goat anti-rabbit secondary antibody (Molecular Probes, Eugene, OR) was applied at a dilution of 1:200 in 1% goat serum in PBS, and subsequently incubated for 30 min in a humidity chamber in the dark. Slides were washed and mounted with Prolong Anti-fade (Molecular Probes) containing a 1:1000 dilution of DAPI nuclear counter stain (Molecular Probes). Fluorescent signal was visualized using single band filter blocks B-2A (Long Pass filter) for Texas red and DAPI (Band Pass filter). Images were captured using a Hammamatsu 3CCD analog camera and controller.

ACKNOWLEDGMENTS We thank Robert Ralston for advice; Gaby Terracina and Bin Shi for technical assistance; Doug Antelman for construction of the p53 plasmids; Canji Cell Core for tissue culture work; and the Departments of Molecular Biology and Process Sciences for use of recombinant adenoviruses. RECEIVED FOR PUBLICATION FEBRUARY 27, 2001; ACCEPTED MAY 11, 2001.

REFERENCES 1. Lane, D. P., Lu, X., Hupp, T., and Hall, P. A. (1994). The role of the p53 protein in the apoptotic response. Philos. Trans. R. Soc. Lond. B Biol. Sci. 345: 277–280. 2. Bellamy, C. O. p53 and apoptosis. (1997). Br. Med. Bull. 53: 522–538. 3. Wiman, K. G. (1997). p53: emergency brake and target for cancer therapy. Exp. Cell Res. 237: 14–18. 4. Wang, X. W., and Harris, C. C. (1996). TP53 tumour suppressor gene: clues to molecular carcinogenesis and cancer therapy. Cancer Surv. 28: 169–196. 5. el-Deiry, W. S. (1998). Regulation of p53 downstream genes. Semin. Cancer Biol. 8: 345–357. 6. Peller, S. (1998). Clinical implications of p53: effect on prognosis, tumor progression and chemotherapy response. Semin. Cancer Biol. 8: 379–387. 7. Nielsen, L. L., Lipari, P., Dell, J., Gurnani, M., and Hajian, G. (1998). Adenovirus-mediated p53 gene therapy and paclitaxel have synergistic efficacy in models of human head and neck, ovarian, prostate, and breast cancer. Clin. Cancer Res. 4: 835–846. 8. Roth, J. A., et al. (1998). Gene therapy for non-small cell lung cancer: a preliminary report of a phase I trial of adenoviral p53 gene replacement. Semin. Oncol. 25 (suppl. 8): 33–37. 9. Horio, Y., et al. (2000). Synergistic effects of adenovirus expressing wild-type p53 on chemosensitivity of non-small cell lung cancer cells. Cancer Gene Ther. 7: 537–544. 10. Nielsen, L. L., et al. (1997). Efficacy of p53 adenovirus-mediated gene therapy against human breast cancer xenografts. Cancer Gene Ther. 4: 129–138. 11. Lowe, S. W. (1995). Cancer therapy and p53. Curr. Opin. Oncol. 7: 547–553. 12. Anklesaria, P. (2000). Gene therapy: a molecular approach to cancer treatment. Curr. Opin. Mol. Ther. 2: 426–432. 13. Schuler, M., et al. (2001). Adenovirus-mediated wild-type p53 gene transfer in patients receiving chemotherapy for advanced non-small-cell lung cancer: results of a multicenter phase II study. J. Clin. Oncol. 19: 1750–1758. 14. Weill, D., et al. (2000). Adenoviral-mediated p53 gene transfer to non-small cell lung cancer through endobronchial injection. Chest 118: 966–970. 15. Dummer, R., et al. (2000). Biological activity and safety of adenoviral vector-expressed wild-type p53 after intratumoral injection in melanoma and breast cancer patients with p53-overexpressing tumors. Cancer Gene Ther. 7: 1069–1076.

12

16. Swisher, S. G., et al. (1999). Adenovirus-mediated p53 gene transfer in advanced nonsmall-cell lung cancer. J. Natl. Cancer Inst. 91: 763–771. 17. Harris, M. P., et al. (1996). Adenovirus-mediated p53 gene transfer inhibits growth of human tumor cells expressing mutant p53 protein. Cancer Gene Ther. 3: 121–130. 18. Kock, H., et al. (1996). Adenovirus-mediated p53 gene transfer suppresses growth of human glioblastoma cells in vitro and in vivo. Int. J. Cancer 67: 808–815. 19. Wolf, J. K., et al. (1999). Adenovirus-mediated p53 growth inhibition of ovarian cancer cells is independent of endogenous p53 status. Gynecol. Oncol. 75: 261–261. 20. Kubbutat, M. H., Jones, S. N., and Vousden, K. H. (1997). Regulation of p53 stability by Mdm2. Nature 387: 299–303. 21. Midgley, C. A., and Lane, D. P. (1997). p53 protein stability in tumour cells is not determined by mutation but is dependent on Mdm2 binding. Oncogene 15: 1179–1189. 22. Yap, D. B., Hsieh, J. K., and Lu, X. (2000). Mdm2 inhibits the apoptotic function of p53 mainly by targeting It for degradation. J. Biol. Chem. 275: 37296–37302. 23. Maki, C. G., Huibregtse, J. M., and Howley, P. M. (1996). In vivo ubiquitination and proteasome-mediated degradation of p53(1). Cancer Res. 56: 2649–2654. 24. Haas, M., et al. (1998). Effect of proteasome inhibitors on monocytic IκB-α and -β depletion, NF-κB activation, and cytokine production. J. Leukoc. Biol. 63: 395–404. 25. Suzuki, S. O., and Iwaki, T. (2000). Amplification and overexpression of mdm2 gene in ependymomas. Mod. Pathol. 13: 548–553. 26. Florenes, V. A., et al. (1994). MDM2 gene amplification and transcript levels in human sarcomas: relationship to TP53 gene status. J. Natl. Cancer Inst. 86: 1297–1302. 27. Toi, M., Saji, S., Suzuki, A., Yamamoto, Y., and Tominaga, T. (1997). MDM2 in breast cancer. Breast Cancer 4: 264–268. 28. Gorgoulis, V. G., et al. (2000). Altered expression of the cell cycle regulatory molecules pRb, p53 and MDM2 exert a synergetic effect on tumor growth and chromosomal instability in non-small cell lung carcinomas (NSCLCs). Mol. Med. 6: 208–237. 29. Watanabe, T., Ichikawa, A., Saito, H., and Hotta, T. (1996). Overexpression of the MDM2 oncogene in leukemia and lymphoma. Leuk. Lymphoma 21: 391–397. 30. Landers, J. E., Haines, D. S., Strauss, J. F., and George, D. L. (1994). Enhanced translation: a novel mechanism of mdm2 oncogene overexpression identified in human tumor cells. Oncogene 9: 2745–2750. 31. Momand, J., Jung, D., Wilczynski, S., and Niland, J. (1998). The MDM2 gene amplification database. Nucleic Acids Res. 26: 3453–3459. 32. Blaydes, J. P., Gire, V., Rowson, J. M., and Wynford-Thomas, D. (1997). Tolerance of high levels of wild-type p53 in transformed epithelial cells dependent on auto-regulation by mdm-2. Oncogene 14: 1859–1868. 33. Bougeret, C., et al. (2000). Cancer gene therapy mediated by CTS1, a p53 derivative: advantage over wild-type p53 in growth inhibition of human tumors overexpressing MDM2. Cancer Gene Ther. 7: 789–798. 34. Lin, J., et al. (2000). A modified p53 overcomes mdm2-mediated oncogenic transformation: a potential cancer therapeutic agent. Cancer Res. 60: 5895–5901. 35. Soengas, M. S., et al. (1999). Apaf-1 and caspase-9 in p53-dependent apoptosis and tumor inhibition. Science 284: 156–159. 36. St. John, L. S., Sauter, E. R., Herlyn, M., Litwin, S., and Adler-Storthz, K. (2000). Endogenous p53 gene status predicts the response of human squamous cell carcinomas to wild-type p53. Cancer Gene Ther. 7: 749–756. 37. Gomez-Manzano, C., Fueyo, J., Alameda, F., Kyritsis, A. P., and Yung, W.K. (1999). Gene therapy for gliomas: p53 and E2F-1 proteins and the target of apoptosis. Int. J. Mol. Med. 3: 81–85. 38. Shabram, P. W., et al. (1997). Analytical anion-exchange HPLC of recombinant type-5 adenoviral particles. Hum. Gene Ther. 8: 453–465. 39. Soussi, T. (2000). The p53 tumor suppressor gene: from molecular biology to clinical investigation. Ann. N Y Acad. Sci. 910: 121–139. 40. Wills, K. N., et al. (1994). Development and characterization of recombinant adenoviruses encoding human p53 for gene therapy of cancer. Hum. Gene Ther. 5: 1079–1088. 41. Chartier, C., et al. (1996). Efficient generation of recombinant adenovirus vectors by homologous recombination in Escherichia coli. J. Virol. 70: 4805–4810. 42. Cheney, I. W., et al. (1998). Suppression of tumorigenicity of glioblastoma cells by adenovirus-mediated MMAC1/PTEN gene transfer. Cancer Res. 58: 2331–2334. 43. Huyghe, B. G., et al. (1995). Purification of a type 5 recombinant adenovirus encoding human p53 by column chromatography. Hum. Gene Ther. 6: 1403–1416. 44. Zhang, G., Gurtu, V., Kain, S. R., and Yan, G. (1997). Early detection of apoptosis using a fluorescent conjugate of annexin V. Biotechniques 23: 525–531. 45. Wolfinger, R. D. (1997). An example of using mixed models and PROC MIXED for longitudinal data. J. Biopharm. Stat. 7: 481–500.

MOLECULAR THERAPY Vol. 4, No. 1, July 2001 Copyright  The American Society of Gene Therapy