Abstract
We have exploited the ability of wild-type (wt) p53 to repress gene expression and produce tumor-selective cytotoxicity using viral-directed enzyme prodrug therapy. Vectors containing either the cytomegalovirus or Rous sarcoma virus promoter regulating transcription of a rabbit liver carboxylesterase (CE) have been constructed. Upon transfection of these plasmids into cells expressing either wt or mutant p53, differential expression of the CE has been observed, resulting in sensitization of the cells expressing the latter protein to the anticancer prodrug irinotecan, 7-ethyl-10-[4-(1-piperidino)-1-piperidino] carb- onyloxycamptothecin (CPT-11). Transduction of isogenic cell lines with adenovirus containing CE under control of the Rous sarcoma virus promoter confirmed the decreased sensitization of cells expressing wtp53 to CPT-11. These studies indicate that the inactivation of wtp53 by mutant p53 in human tumor cells may be sufficient enough to generate a therapeutic window for enhanced cytotoxicity with CPT-11.
p53 is a 53-kDa nuclear phosphoprotein involved in control of cell growth and apoptosis (Levine, 1997). Inactivation of the p53 tumor suppressor gene, either by mutation or deletion, has been observed in a variety of human malignancies (Levine et al., 1991). The p53 protein can positively regulate transcription of a number of genes, includingGADD 45, p21, bax, mdm2, and cyclin G. Transcriptional activation is mediated by protein binding to a p53 consensus motif, consisting of two copies of a 10-bp sequence, rrrCwwGyyy, that can be separated from each other by up to 13 bp (el-Deiry et al., 1992). p53 can also repress expression of a number of genes, including those involved in the response to chemotherapeutic drugs, such as topoisomerase II, multidrug resistance (MDR1), multidrug-associated protein (MRP1), andO6-methylguanine-DNA methyltransferase (MGMT; Nguyen et al., 1994; Harris et al., 1996; Thottassery et al., 1997; Wang and Beck, 1998). In addition, down-regulation of expression of a variety of viral promoters has been observed, e.g., Rous sarcoma virus (RSV), SV40, and Herpes simplex virus thymidine kinase (Subler et al., 1992; Zambetti and Levine, 1993). This probably occurs as a result of p53 interacting with Sin3a (Murphy et al., 1999).
We have exploited the ability of p53 to regulate transcription by designing vectors that contain different gene promoters controlling expression of a rabbit liver carboxylesterase (CE) that can convert the camptothecin analog irinotecan, 7-ethyl-10-[4-(1-piperidino)-1-piperidino]carbonyloxycamptothecin (CPT-11) into the potent topoisomerase I poison 7-ethyl-10-hydroxycamptothecin (SN-38) (Potter et al., 1998a). Expression of this CE in mammalian cells confers sensitivity to CPT-11 both in culture and when grown as human tumor xenografts in immune-deprived mice (Danks et al., 1998, 1999; Potter et al., 1998a). Hence, selective expression of the rabbit CE in tumor cells and subsequent treatment with CPT-11 may provide a novel method of cancer therapy. To determine whether the expression of mutant or wtp53 could influence exogenous gene transcription for use in viral-directed enzyme prodrug therapy (VDEPT), we assessed the ability of different forms of p53 to regulate CE expression from both the cytomegalovirus (CMV) or RSV promoters. Our results indicate that the described method may prove useful in the selective treatment of human tumors expressing mutant p53.
Materials and Methods
Cell Lines, Plasmids, Adenovirus, and CPT-11.
The osteosarcoma cell line Saos-2, was obtained from the American Type Culture Collection (Manassas, VA) and grown in Dulbecco's modified Eagle's medium containing 10% fetal calf serum at 37°C under an atmosphere of 10% CO2. This cell line lacks the p53 gene. The cell lines LS180neo and LS180tdn have been described previously (Thottassery et al., 1997). Briefly, they were derived by transfection of LS180 cells with plasmids containing the CMV promoter regulating expression of either a transdominant negative (tdn) p53, or no cDNA. They demonstrate similar growth rates with doubling times of approximately 24 h.
The plasmids used in this study are detailed in Table1. Briefly, pCIneo was obtained from Promega (Madison, WI), and plasmids containing both mutant and wild-type p53 (pSN3 and pCMV281G) were obtained from Dr. G. Zambetti (St. Jude Children's Research Hospital, Memphis, TN). pCMV281G differs from pSN3 by a single point mutation at position 281 in the p53 protein. Replication-deficient adenovirus containing a cDNA encoding a secreted form of the rabbit liver CE under control of the RSV promoter (AdRSVrCES) was generated by homologous recombination, and its construction and partial characterization have been described in detail elsewhere (Wierdl et al., 2001).
CPT-11 was a generous gift from Dr. J. P. McGovren (Pharmacia and Upjohn, Kalamazoo, MI).
Transient Transfections.
Cells (1 × 107) were electroporated in 200 μl of phosphate-buffered saline using an electroporator (Bio-Rad, Hercules, CA) and a capacitance extender. Optimal conditions for transfection were achieved using 200 V and 960 μF for Saos-2 cells and 260 V and 960 μF for LS180 cells. All transfections contained the same total amount of plasmid DNA (25 μg) and cells were harvested after 48 h for CE activity analysis, Western analysis, or growth inhibition assays.
Carboxylesterase Assays.
CE assays were performed as described previously (Beaufay et al., 1974; Potter et al., 1998). Briefly, cells were sonicated in minimal volumes of 50 mM HEPES, pH 7.4, on ice and CE activity determined using a spectrophotometric assay with o-nitrophenol acetate as a substrate. Protein concentrations were determined using Bio-Rad protein reagent and bovine serum albumin as a standard. Activity was expressed as micromoles of substrate converted per minute per milligram of protein.
Growth Inhibition Assays.
Growth inhibition assays were performed as described previously (Danks et al., 1998; Potter et al., 1998a). Routinely, plasmid-transfected cells were harvested by trypsinization, plated at 5 × 104 cells per well in 35-mm six-well plates, and allowed to attach overnight. The following day, CPT-11 was diluted in fresh media and applied to cells. All drug concentrations were repeated in triplicate. After a time equivalent to three cell doublings, the media were aspirated and cells counted using a Multisizer II (Beckman Coulter, Inc., Fullerton, CA). IC50 values were determined by computer analysis of the data using the Prism program (GraphPad Software Inc., San Diego, CA).
Adenoviral Transduction.
Cells were plated at a density of 3 × 106 cells per T162 flasks and allowed to adhere overnight. Adenovirus at a multiplicity of infection (moi) of 20 was applied to the cells for 24 h. The following day, the media were removed and replaced, and cells allowed to grow for up to 7 days. At 24-hour intervals, a small aliquot of media was recovered and CE activity was monitored. For growth inhibition studies, media were removed at day 5 and applied to untransduced LS180neo and LS180tdn cells plated at a density of 1 × 105 cells per well of a six-well plate. This approach was used because adenoviral transduction can alter the growth rates of cells. The IC50 values for each cell line with CPT-11 were then determined as described previously (Wierdl et al., 2001).
Statistical Analysis of Data.
Analysis of data were performed using Instat (GraphPad Software Inc.). For CE activities from transient transfection studies, one-way analysis of variance analyses were used to determine the statistical significance of all results, compared with each other. For the levels of secreted CE produced from adenovirus-transduced LS180 cell lines, Tukey-Kramer multiple comparison tests were performed to yield p values from data derived from each time point.
Results
Construction of Plasmids Containing the Rabbit Liver Carboxylesterase under Control of the Human mdr1 Promoter.
To assess the ability of p53 to regulate expression of rCE from the human mdr1 gene, we constructed plasmids containing the CE cDNA under control of the minimal mdr1promoter. Figure 1 indicates the manipulations involved. The plasmid pCIneo was digested withBglII and I-PpoI to remove the CMV promoter and after isolation of the 4.6-kb fragment containing the “backbone”, a double-stranded oligonucleotide was ligated to yield pCIBgl. The oligonucleotide contained both 5′ and 3′ overhangs to allow facile ligation to the 4.6-kb fragment and was designed to recreate theBglII site. After digestion of the ligation reaction withSacI to remove any parental pCIneo from the preparation, the sample was transformed into Escherichia coli Top10 cells, and the integrity of clones determined by DNA sequencing. The resulting plasmid, pCIBgl, was restricted with BglII and ligated to a 160-bp BglII fragment containing the human mdr1promoter (Thottassery et al., 1997). This fragment contains DNA sequence from position −137 to +30 relative to the initiation of transcription from the promoter (GenBank accession no. L07624). Orientation of the transcriptional control region was confirmed by DNA sequencing, and after large-scale preparation, a 1.7-kbEcoRI containing the entire rCE open reading frame (Potter et al., 1998a) was ligated into the unique EcoRI site in the multiple cloning site. The integrity of the resulting plasmid, pmdrRabfl, was determined by restriction endonuclease analysis, and after large-scale preparation using endotoxin-free reagents (QIAGEN, Valencia, CA), was used for transfection studies.
p53-Mediated Regulation of Carboxylesterase Expression Using themdr1 Promoter.
Because mutant p53 can specifically up-regulate gene transcription from mammalian promoters, including the human mdr1 gene, we assessed expression of rCE from pmdrRabfl. This plasmid was cotransfected into Saos-2 cells with vectors containing different p53 cDNAs. All experiments contained 25 μg of total DNA consisting of 20 μg of pmdrRabfl, 5 μg of mutant p53 plasmid (pCMV281G), or 1 μg of wtp53 plasmid (pSN3) + 4 μg of pCIneo. Forty-eight hours after transfection, CE activity was measured in the cell extracts. As indicated in Table2, very high levels of CE activity were observed in Saos-2 cells after transfection with plasmid containing the CE under control of the CMV promoter. However, very little activity was seen after transfection with pmdrRabfl, even after cotransfection with mutant p53 cDNAs. We presumed that the mdr1 promoter was relatively weak in comparison with CMV and that up-regulation of CE expression sufficient to change the sensitivity of cells to CPT-11 would be unlikely.
Regulation of Carboxylesterase Expression from the CMV Promoter.
Because we were unable to use the human mdr1promoter to regulate CE expression, we adopted a different strategy that relied upon the repression of gene transcription by wtp53. To determine whether the CMV promoter could regulate CE expression in a p53-dependent manner, we cotransfected 20 μg of pCIRabfl DNA into Saos-2 cells with plasmids containing either mutant (5 μg) or wtp53 (1 μg) cDNAs. Table 3 and Fig.2 indicate the CE activities and a cell survival curve for a representative experiment. As indicated in Table 3, cells cotransfected with pCIRabfl and either the parent plasmid pCIneo or mutant p53 demonstrated high levels of enzyme activity. In contrast, cotransfection with plasmid expressing wtp53 resulted in marked suppression of CE expression, resulting in approximately 50% less enzyme activity. This is consistent with specific down-regulation of gene expression from the CMV promoter by wtp53. The reduction in CE activity translated into a 2.7-fold change in the IC50 value for CPT-11 with Saos-2 cells transfected with pCIneo compared with those transfected with wtp53 (Fig. 2).
Western analysis of extracts derived from these transfections indicated that approximately 5-fold more mutant p53 protein was present than wtp53 in the Saos2 cells (data not shown). Because pCMV281G demonstrates no ability to repress gene expression (Thottassery et al., 1997; Sampath et al., 2001), the increased amount of this protein will not influence levels of rCE after transfection. If similar levels of wt and mutant p53 were present, we would expect an even more marked reduction in CE activity from the CMV promoter. These results indicate that the repression in CE activity by wtp53 may provide a viable means to alter sensitivity of cells to CPT-11.
p53-Mediated Regulation of Carboxylesterase Expression from Adenoviral Vectors.
Because we observed regulation of CE expression from the CMV promoter by wtp53 after transient transfection of p53 null cells, we wished to determine whether the endogenous levels of wtp53 would be sufficient to regulate transcription from an exogenous viral promoter. Because work by Subler et al. (1992)indicated that expression from the RSV promoter was down-regulated 28-fold by wtp53, nearly 20-fold more than that observed with CMV regulatory sequences, we constructed adenoviral vectors expressing a secreted form of the rabbit liver CE under control of the RSV promoter (Wierdl et al., 2001).
After transduction of LS180neo and LS180tdn cells with AdRSVrCES at an moi of 50 for 24 h, the media were replaced, and CE activity in the culture media was monitored for up to 8 days. Figure3 demonstrates that LS180tdn cells transduced with AdRSVrCES produce approximately 3 times more CE activity than LS180neo cells. These data are consistent with the hypothesis that expression from the RSV promoter is regulated by the endogenous p53 protein.
To confirm that these levels of CE expression could confer sensitivity to CPT-11, we performed growth inhibition assays with media derived from LS180neo and LS180tdn cells, 140 h after AdRSVrCES transduction at an moi of 20. Figure 4indicates the survival of these cell lines to CPT-11, with the IC50 values for LS180neo and LS180tdn cells being 0.97 and 0.35 μM, respectively.
Discussion
The p53 gene is considered to be the most frequently mutated gene in human tumors. Mutations typically arise in regions of the protein that affects its binding to DNA, resulting in the formation of molecules that demonstrate aberrant transactivation or repression of gene expression (Levine et al., 1991; Wang et al., 1994). We reasoned that promoters that are regulated by p53 might be suitable candidates for use in VDEPT in combination with an enzyme that can selectively activate a drug. Based on this premise, we initially designed mammalian expression vectors containing the rabbit liver CE under control of the human mdr1 promoter. In reporter assays, this promoter is up-regulated when cotransfected with plasmids expressing mutant p53 (Sampath et al., 2001). However, the level of promoter activity was not sufficient to increase the levels of CE expression such that sensitization to CPT-11 was observed.
Reports have indicated, however, that the degree of repression of viral promoters by wtp53 may be sufficient to reduce gene expression to levels that would affect drug sensitivity. Because we had already constructed plasmid vectors containing the rCE cDNA under control of the CMV promoter (pCIRabfl; Potter et al., 1998a), we determined whether p53 could modulate expression of CE from this transcriptional regulatory element. As indicated in Table 3, a significant reduction in CE activity occurred when these plasmids were cotransfected with wtp53 expression vectors. To determine whether such repression was sufficient to elicit a response after transfection with a drug-activating enzyme, we assessed the sensitivity of cells to CPT-11 after cotransfection with pCIRabfl and the p53 expression plasmids. We have previously demonstrated that expression of this CE in mammalian cells confers sensitivity to the camptothecin analog CPT-11 (Danks et al., 1998;Potter et al., 1998a,b). After transfection of pCIRabfl into the p53-null cell line Saos-2 in combination with p53 expression vectors, we observed selective down-regulation of CE expression by the wtp53 protein (from pSN3; Table 3). Additionally, this abrogated the sensitivity of cells to CPT-11 conferred by the pCIRabfl vector. This was not observed in cells cotransfected with mutant p53, indicating that this form of the protein was abrogating the function of the endogenous wtp53. These data indicate that a VDEPT approach using the regulated expression of the rabbit liver CE in combination with CPT-11 may be viable therapy for human tumors.
However, before initiating large-scale virus experiments to assess the applicability of the CMV promoter to regulate CE expression, a search of the literature was then performed to determine whether these were the optimum transcriptional control sequences that could be regulated by the tumor suppressor gene. This search identified an article presented by Subler et al. (1992) that indicated that the greatest level of repression of gene transcription by wtp53 was observed with the RSV promoter. Levels of repression of gene expression using chloramphenicol acetyltransferase reporter constructs indicated that although the CMV promoter resulted in 8-fold less activity than control plasmids, the RSV promoter was down-regulated 28-fold by wtp53. Because we had already detailed the construction of adenovirus-expressing rCE under control of the RSV promoter (AdRSVrCE and AdRSVrCES; Wierdl et al., 2001), we opted to use these viruses to assess p53-mediated regulation of expression in the LS180 cell lines.
To provide a more biologically relevant system to assess the regulation of gene expression compared with the somewhat artificial approach using cotransfection of plasmid vectors, we opted to use adenovirus to transduce isogenic cell lines expressing different forms of p53. AdRSVrCES, which contains rCE under control of the RSV promoter, was used to transduce paired cell lines that contained either the endogenous wtp53 gene or had been transfected with a tdn p53. The tdn p53 protein forms tetramers with wtp53, resulting in inactivation of the protein (Wang et al., 1994), similar to what has been observed in human tumors. We opted for this approach with isogenic cell lines because it would reduce any variables such as the susceptibility to cells to adenoviral transduction or the inherent differences in the levels of transcription factors in unrelated cell types. After adenoviral transduction, we observed less CE expression in cells containing functionally active wtp53 (LS180neo) compared with tdn p53. Consistent with our hypothesis, cells expressing tdn p53 did not down-regulate expression of the CE and hence were relatively sensitive to CPT-11. Although the changes in the IC50values were modest (2.8-fold), it is generally considered that relatively small changes in drug resistance and sensitivity may have significant impact in chemotherapeutic treatment of human tumors. Hence, this level of CE expression and reduction in the IC50 value may be sufficient to elicit a biological response in vivo. To this end, we are testing the potential use of these adenoviral vectors to selectively sensitize human tumor xenografts containing different p53 mutations to CPT-11 in immune-deprived mice.
Although adenovirus are exceptionally useful delivery vehicles for recombinant gene expression, in studies with these vectors expressing rCE, we have observed two mechanisms that alter the growth rates of cells, and hence affect sensitivity or apparent sensitivity, to anticancer agents. First, high-level expression of CE can actually stop cell growth, with cells being arrested in the G2 phase of the cell cycle. This essentially makes cells resistant to the toxic effects of CPT-11. Second, even transduction with adenovirus that does not contain recombinant DNA for high-level gene expression can significantly affect cell growth rates (McPake et al., 1999). To avoid these potentially complicating issues, we opted to use adenovirus encoding the secreted rCE, and apply the media onto nontransduced cells. Any effects caused by the adenovirus on the response of the parental cells to CPT-11 would therefore be eliminated. This also allowed the direct measurement of CE activity in the media from an identical population of cells during the time course of protein expression. Hence, any variability in levels of CE between different samples would be minimized.
Several investigators have indicated that gene therapy with virus expressing wtp53 might be a viable approach in the treatment of human cancer and there are clinical trials currently underway using adenovirus containing this gene (Habib et al., 1999; Nemunaitis et al., 2000). However, for such approach to effective, it would be predicted that virtually all tumor cells would need to be transduced. With the currently available viral vector systems, this seems unlikely. The method described in this article differs in that with an enzyme that can elicit a bystander effect with CPT-11, i.e., the secreted rabbit CE (Wierdl et al., 2001), only a subpopulation of the tumor cells would need to be transduced. Neighboring tumor cells would be killed by the SN-38 produced in the extracellular fluid by the secreted CE. As indicated above, the effectiveness of these approaches is being determined in animal model systems.
In conclusion, we have developed a VDEPT approach using the rabbit liver CE and CPT-11 where cytotoxicity is dependent upon the endogenous p53 status of the cell. Our results indicate that selective toxicity can be achieved in cells overexpressing mutant p53 and that a protective effect elicited by the repression of CE expression in cells expressing wtp53 can be generated. These studies demonstrate a novel approach to regulation of gene expression with a potential therapeutic application.
Acknowledgments
We thank Dr. J. P. McGovren for the gift of CPT-11, Dr. G. Zambetti for the plasmids pSN3 and pCMV281G, the St. Jude Hartwell Center for Biotechnology for expert DNA synthesis and sequencing, and Dr. Peter Houghton for critical review of this manuscript.
Footnotes
-
This work was supported in part by National Institutes of Health Grants CA66124, CA76202, CA79763, and ES05851; the Cancer Center Core Grant CA21765, and the American Lebanese Syrian Associated Charities.
-
DOI: 10.1124/jpet.102.044149
- Abbreviations:
- bp
- base pair(s)
- mdr
- multidrug resistance
- RSV
- Rous sarcoma virus
- SV40
- simian virus 40
- CE
- carboxylesterase
- wtp53
- wild-type p53
- VDEPT
- viral-directed enzyme prodrug therapy
- CMV
- cytomegalovirus
- tdn
- transdominant
- rCE
- rabbit liver carboxylesterase
- kb
- kilobase(s)
- Received September 6, 2002.
- Accepted October 25, 2002.
- The American Society for Pharmacology and Experimental Therapeutics