Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Inactivation of p53 is a frequent event in development of human malignancies (Hollestein et al., 1991, 1994). Wild-type p53 is in fact a critical element in suppressing cell proliferation in response to DNA-damaging agents, either inducing G1 arrest or activating programmed cell death (reviewed in Hansen and Oren, 1997). Loss of p53 function leads to genomic instability and tumor cell resistance to anticancer therapy, including chemotherapeutic drugs and radiation (Livingstone et al., 1992; Yin et al., 1992; Lowe et al., 1994).

Wild-type p53 is also involved in activating DNA repair either directly or through transactivation of other genes (reviewed in Ko and Prives, 1996; Marx, 1994). In particular, p53 appears to affect the rate and efficiency of excision repair system (Wang et al., 1995a). In contrast, it has been recently demonstrated that wild-type p53 suppresses transcription of OGAT (Harris et al., 1996). This repair enzyme removes the alkyl adducts on DNA at the O⁶ position of guanine (for a review see Pegg et al., 1995). Alkyl adducts are generated by monofunctional methylating and chloroethylating agents. Killing by cloroethylating agents involves the formation of DNA interstrand cross-links, which are produced by intramolecular rearrangement of O⁶-chloroethylguanine adducts (Tong et al., 1982; Jlang et al., 1989). Conversely, the mechanism of cytotoxicity by methylating compounds is likely to involve an inappropriate processing of O⁶-methylguanine during DNA replication by the mismatch repair system, which generates DNA strand breaks (Karran and Bignami, 1992). Thus, high levels of OGAT and the loss of components of mismatch repair system, that involves long patch DNA repair, are the main factors responsible of tumor cell resistance to O⁶-methylating agents (Liu et al., 1996; Karran and Bignami, 1994; Kat et al., 1993).

The PADPRP enzyme is activated in response to agents that cause DNA strand breaks (Lindahl et al., 1995). Its function has been implicated in a variety of biological processes, including DNA repair and cell survival after DNA damage. Interaction of PADPRP with DNA might allow access of the repair enzymes to the damaged sites on DNA. This enzyme has been defined as one of the substrates cleaved by the ICE-like proteases during the process of apoptosis (Tewari et al., 1995).

Programed cell death is characterized by different types of DNA fragmentation, such as degradation of DNA into oligonucleosome sized or large DNA fragments (50-300 Kb), and cleavage due to single strand breaks (for a review see Kumar and Lavin, 1996). Therefore, taking into account the role of PADPRP in DNA repair, it is conceivable to hypothesize that proteolysis of PADPRP during apoptosis would avoid the activation of this enzyme by DNA fragmentation, which would lead to repair under conditions where DNA should be degraded.

Recently, it has been shown that PADPRP inhibitors have the potential to act as resistance modifiers when used in combination with chemotherapeutic agents (Griffin et al., 1995). A previous study indicated that combined treatment with TZM and BZ, an inhibitor of PADPRP (Griffin et al., 1995), results in overall increase of apoptosis in either OGAT-proficient or OGAT-deficient leukemia cells (Tentori et al., 1997). Furthermore, it has been demonstrated that PADPRP inhibitors might enhance methylating agent activity also in tumor cells devoid of a functional mismatch repair system (Wedge et al., 1996). Thus, it was speculated that PADPRP might be involved in coordinating the repair of DNA adducts, different from O⁶-methylguanine, induced by TZM. In fact, interaction of methylating compounds with DNA leads to the formation of adducts also at the N⁷ of guanine, N³ of adenine and other DNA base positions that are repaired by the base excision repair pathway (Barnes et al., 1993). The different types of modified bases are cleaved by specific glycosylases. Once the base is removed the further repair of the apurinic site requires the coordinate intervention of endonucleases, phosphodiesterases, DNA polymerases and ligases. In the absence of an efficient repair, depurination of 7-methylguanine or 3-methylguanine is followed by the generation of DNA strand breaks.

Methylating compounds include chemotherapeutic drugs such as dacarbazine, and its in vitro active derivative TZM, which is currently under phase II clinical study (Woll et al., 1995; Bleehen et al., 1995). Dacarbazine is active against melanomas, sarcomas and lymphomas and has shown a therapeutic potential for the treatment of acute nonlymphoid leukemias relapsed or refractory to conventional treatment (Franchi et al., 1992; D'Atri et al., 1995). Differently from dacarbazine, TZM does not require metabolic activation, has limited bone marrow toxicity and has demonstrated promising clinical activity in the treatment of gliomas and melanomas (Newlands et al., 1992). Recently, we have shown that TZM treatment of leukemia cell lines (i.e., U937 and K562) with low basal levels of OGAT is followed by the induction of apoptosis (Tentori et al., 1995). Moreover, transfection of these cell lines with the human OGAT cDNA reduced tumor cell sensitivity to the cytotoxic and apoptotic effects of this compound (Tentori et al., 1997). Down-modulation of OGAT by BG of OGAT-transfected lines or of HL60 cells, which express high levels of OGAT activity, restored leukemia cell sensitivity to apoptosis induced by TZM (Tentori et al., 1995 and 1997). In this model the contribution of a functional p53 to apoptosis induced by TZM could not be explored, since both U937 and K562 cells possess a mutated p53 gene (Dou and An, 1995; Bedi et al., 1995) and it has recently been demonstrated that p53 mutants might interfere differently with normal p53 function, depending on the type of mutations (Pocard et al., 1996).

The aim of our study was to investigate the influence of wild-type p53 on apoptosis induced by TZM. Moreover, the contribution to TZM-mediated cytotoxicity of DNA lesions different from those generated by the inappropriate processing of O⁶-methylguanine, has been explored. The human p53 cDNA was transfected in p53 null HL60 cells (Wolf and Rotter, 1985) resulting in a stable expression of low levels of p53 protein. The data indicate that the presence of wild-type p53 increases apoptosis by TZM even in cells naturally resistant to O⁶-methylating agents, due to high levels of OGAT activity. Inhibition of PADPRP potentiated apoptosis induced by TZM in p53+ and p53- cells with high levels of OGAT activity and in leukemia cells characterized by a mutated form of p53 gene and tolerant to O⁶-methylguanine adducts.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Cell lines. Three human leukemia cell lines were used: HL60, a promyelocitic cell line (ATCC CCL 240); Daudi, a Burkitt lymphoma cell line (ATCC CCL 213); Jurkat, a T-cell leukemia line (ATCC CRL 8136).

Reagents. TZM was kindly provided by Schering-Plough Research Institute (Kenilworth, NJ). [³H]-methylated-DNA, to be used as substrate for determination of OGAT activity, was prepared using [³H]-methyl-N-nitrosourea (specific activity 19.2 Ci/mmol, Amersham International Plc, Amersham, UK) (Margison et al., 1985). BG was synthesized and kindly gifted by dr. L. Lassiani (Institute of Pharmacological Chemistry, University of Trieste, Trieste, Italy). Etoposide (VP16), BZ and ABZ were purchased from Sigma Chemical Co. (St. Louis, MO). TZM, BZ and ABZ stock solutions were prepared by dissolving the drugs in RPMI-1640, whereas BG was dissolved in ethanol.

Transfection and transduction of p53 cDNA. A HindIII fragment, encompassing the human p53 cDNA (Lamb and Crawford, 1986) was cloned into the retroviral vector pLNSX (Miller and Rosman, 1989). After DNA sequencing by the chain termination method using Sequenase version 2.0 (US Biochemical, Cleveland, OH), these constructs or the vector pLNSX were transfected into the PA317 packaging cell line, using the calcium phosphate precipitation procedure. Stable virus producing cell lines were generated following selection of the mass culture with 1 mg/ml G418. The apparent virus titers were determined on murine NIH3T3 cells by G418 selection and ranged between 10⁴ and 4 × 10⁴ colony forming units/ml. HL60.12 clone, in the exponential phase of growth, was transduced by replacing culture medium with a 1:1 mixture of p53 recombinant, or control virus producing PA317 cells and fresh CM additioned with polybrene (8 µg/ml). Two days after transduction, cells were G418 selected (0.5 mg/ml). Control or not transduced cells died at this antibiotic concentration.

Drug treatment and cell growth evaluation. Cell suspensions containing 1 × 10⁶ cells/ml in CM were placed in 50-ml tubes (Falcon, Becton Dickinson Labware, Oxnard, CA) and TZM was added to each tube at final concentrations ranging between 50 and 200 µM. Cell growth was evaluated in terms of viable cell count every 24 hr for 3 days. Cells were manually counted using a haemocytometer and cell viability was determined by trypan blue exclusion test or by flow cytometry analysis of cells stained with PI. All determinations were made in quadruplicate. Depletion of OGAT activity was obtained by treating tumor cells (1 × 10⁶/ml in CM) with 1 µM BG for 2 hr before TZM exposure. BG was then added again every 24 hr. Inhibition of PADPRP was obtained by treating tumor cells with BZ (5 mM) or with ABZ (5 mM).

Cell irradiation. Exponentially growing cells were seeded at 0.5 × 10⁶ cells/ml in CM, before irradiating in a RAD GIL irradiator (Gilardoni, Milan, Italy) at -.5, or 1.0 Gy. After 8 and 21 hr of incubation at 37°C in a CO₂ incubator, aliquots of 10⁶ irradiated or control cells were collected, washed with phosphate-buffered saline and fixed in 70% ethanol at -20°C, until the time of cell cycle analysis.

Reverse-transcriptase polymerase chain reaction (RT-PCR). Total RNA was isolated by extraction according to the guanidinium thiocyanate method (Chomczynski and Sacchi, 1987), and treated with DNase to eliminate the possibility of genomic DNA contamination. cDNA was synthesized by incubating 2 µg of RNA with 25 U of Moloney murine leukemia virus reverse transcriptase (New England Biolabs, Inc., Beverly, MA) at 37°C for 1 hr in the presence of 50 pmol of oligo(dT)15. After heat inactivation of the enzyme at 95°C for 5 min, aliquots of the synthesized first strand cDNA, for each sample, were used as template for polymerase chain reaction amplification. Human p53, or GAPDH amplification was performed in DNA thermal cycler for 30 cycles (Perkin Elmer Cetus, Norwalk, CT; denaturation at 95°C for 1 min, annealing at 56°C for 1 min and extension at 72°C for 2 min, for each cycle). The oligonucleotide primer pairs used for p53 amplification were 5'-ATGGAGGAGCCGCAGTCA-3' and 5'-AGGGGCCAGACCATCGCTATC-3'. These primers allow amplification of a 572-bp DNA fragment, spanning exon 2, which is the first expressed exon, to the exon 5/6 boundary (Lamb and Crawford, 1986). The primers used for GADPH (5'-TGGTATCGTGGAAGGACTCATGAC-3' and 5'-ATGCCAGTGAGCTTCCCGTTCAGC-3') amplified a 190 bp product.

Northern blot analysis. Aliquots (15 µg) of total cellular RNA, obtained from each cell line, were fractionated by electrophoresis on a formaldehyde-containing 1.2% agarose gel. The integrity of RNA was confirmed by RNA visualization, after staining the gel with ethidium bromide. RNA was transferred to nylon membrane (Gene screen plus, New England Nuclear, Boston, MA) and hybridized at 42°C for 24 hr with [³²P]-labeled either p53 or OGAT probe. The p53 probe was a 572-bp cDNA fragment obtained by RT-PCR using total RNA derived from Daudi cells and the primers described above. The OGAT probe was a PCR-derived cDNA probe obtained after reverse transcription of the RNA from Molt-4 cells (Tentori et al., 1995). After washing with 0.1 x SSC (10 mM sodium chloride, 1.5 mM sodium citrate) at room temperature for 30 min, the blotted membranes were exposed to x-ray films (Kodak, Rochester, NY) at -80°C. Bidimensional densitometry of the immunoblot was performed using a BioRad (Richmond, CA) scanning apparatus (Imaging densitometer, GS-670).

Assay for OGAT activity. This assay was performed as previously described (Morten and Margison, 1988), with minor modifications. Briefly, cells (2 × 10⁶) were lysed in 0.5 ml of lysis buffer containing 0.5% CHAPS, 50 mM Tris-HCl pH 8.0, 1 mM EDTA, 3 mM dithiothreitol, 100 mM NaCl, 10% glycerol, 2 µg/ml leupeptin and freshly added 1 mM PMSF. The lysates were then microfuged at 15,000 rpm at 4°C for 10 min. to pellet cell debris and supernatants were used for the assay. OGAT activity was determined by measuring the transfer of ³H-labeled methyl groups from a calf thymus DNA, which had been previously methylated by reaction with N-[³H]methyl-N-nitrosourea. OGAT activity was expressed in terms of fmol/10⁶ cells.

Flow cytometry analysis. Cells from cultures were washed with phosphate-buffered saline and fixed in 70% ethanol at -20°C for 18 hr. The centrifuged pellets were resuspended in 1 ml of hypotonic solution containing PI (50 µg/ml), 0.1% sodium citrate, 0.1% Triton-X and RNase (10 µg/ml). Cells were incubated in the dark, at 37°C for 30 min. Data collection was gated using forward light scatter and side light scatter to exclude cell debris and cell aggregates. The PI fluorescence was measured on a linear scale using a FACSscan flow cytometer (Becton Dickinson, San Josè, CA). Apoptotic cells are represented by a broad hypodiploid peak, which is easily discriminable from the narrow peak of cells with diploid DNA content in the red fluorescence channel (Nicoletti et al., 1991). All data were recorded and analyzed using Lysis II software (Becton Dickinson). Cell-Fit software (Becton Dickinson) was used for cell cycle analysis.

Western blotting. Cell pellets were resuspended in hypotonic buffer [10 mM Tris-HCl, 1 mM EDTA, pH 7.4 and freshly added 1 mM phenyl-methyl-sulfonyl-fluoride (PMSF), an aliquot saved for protein concentration determination (using BioRad protein assay solution and bovine serum albumin as standard), and the rest immediately boiled in sodium dodecyl sulfate (SDS) sample buffer (50 mM Tris-HCl pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol). One hundred µg of proteins per sample were electrophoresed on 10% SDS-polyacrylamide gels. Afterwards, proteins were transferred on Hybond-C membrane (Amersham). For p53, bcl-2 and bax, immunodetections were carried out using Boehringer-Mannheim chemiluminescence western blotting kit, according to manufacturer's instructions. Filters were exposed to X-OMAT AR autoradiographic films (Kodak) for periods of time ranging between 5 sec and 20 min. For PADPRP immunodetection, filters were blocked with 3% non-fat dry milk in TBST (20 mM Tris-HCl pH 8.0, NaCl 0.9%, 0.03% Tween 20) and then incubated with PADPRP monoclonal antibody (5 µg/ml) in the same blocking buffer for 1 hr. After washing with TBST, immunocomplexes were visualized using an alkaline phosphatase-coupled anti-mouse IgG antibody and the Protoblot color development system (Promega Biotech, Madison, WI). Densitometric analysis of the immunoblots was performed as described above.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Establishment of HL60 cells expressing low levels of wild-type p53. The role of wild-type p53 protein in modulating apoptosis by TZM has been studied by transfecting p53 cDNA in HL60 cell line, which contains gross deletions in both alleles and therefore it can be considered null for p53. Amphotropic PA317 cells were transfected either with the construct pLNSX/p53 or with the pLNSX control vector. Supernatants from virus producer PA317 cells were used to infect a clone obtained from HL60 cell line. To determine p53 expression in transfected cells, G418 selected cells were analyzed for the presence of p53 transcript by RT-PCR, as described in "Materials and Methods." The p53 transcript was detected in HL60 cells transduced with the human p53 cDNA (fig. 1A, lane 1) as well as in Daudi cells that express a mutated form of p53 (fig. 1A, lane 3) (Gaidano et al., 1991), whereas it was absent in cells transduced with the pLNSX vector (fig. 1A, lane 2). Amplification of the GAPDH cDNA was obtained in all samples (fig. 1A). The presence of the p53 transcript was further confirmed by the results of northern blot analysis which indicated the presence of virus-derived p53 transcript in p53 transduced HL60 cells (fig. 1B, lane 1).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Expression of wild-type p53 in HL60 transduced cells visualized by: RT-PCR (A), Northern (B) and Western blot analysis (C). A, RT-PCR of the p53 transcript. For the detection of the p53 (upper band) and GAPDH transcripts (lower band), RT-PCR, was performed from total RNA (2 µg) using oligo(dT)15. Lane 1, p53+ cells; lane 2, p53- cells; lane 3, Daudi; lane 4, 1 Kb DNA ladder (GIBCO-BRL, Milan, Italy). B, Northern blot analysis of p53 mRNA transcript. Equal amounts of total RNA samples (15 µg of RNA) from each cell line (lane 1, p53+ cells; lane 2, p53- cells), were separated by gel electrophoresis and transferred to nylon membrane as described in "Material and Methods." The p53 transcript was identified using a 572 bp [32P]-labeled p53 cDNA probe. The ethidium bromide staining of the gel is also represented. C, Western blot analysis of p53 protein expression. Immunodetection was performed using anti-p53 (Ab-2) monoclonal antibody. Lane 1, p53- cells; lane 2, p53+ cells. Filter was exposed to autoradiographic film for 20 min.

Wild-type p53 transduction increases apoptosis induced by TZM in OGAT-proficient HL60 cells. Loss of p53 function is generally associated with inability of the cell to undergo apoptosis when cells are exposed to serum deprivation or DNA damaging agents. In fact, in our model, wild-type p53+ cells were more sensitive than p53- cells to apoptosis that follows serum deprivation of cell culture or treatment with the anticancer agent VP16 (data not shown).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Expression of wild-type p53 increases apoptosis by TZM. Left panels, Flow cytomety analysis of DNA content. Flow cytometry profiles of p53- (top) or p53+ (bottom) cells untreated or exposed to 100 and 200 µM TZM. Cells were processed for FACS analysis 72 hr after drug treatment. Percentages of cells with hypodiploid DNA content are indicated. Right panel, Western blot analysis of PADPRP protein cleavage. Lane 1, untreated p53+ cells; lane 2, p53+ cells treated with 100 µM TZM; lane 3, p53+ cells treated with 200 µM TZM; lane 4, p53- cells treated with 200 µM TZM; lane 5, p53- cells treated with 100 µM TZM; lane 6, untreated p53- cells. The presence of the 25-kDa cleaved product, detected only in p53+ cells treated with TZM is indicated by an arrow.

Low levels of wild-type p53 expression do not affect OGAT gene transcription or activity. A previous study described that p53 overexpression can downregulate the OGAT promoter (Harris et al., 1996). Therefore, to verify whether increased apoptotic effects of temozolomide in p53+ OGAT-proficient cells might be due to the reduction of OGAT transcription, Northern, Western blot analysis and evaluation of OGAT activity were performed. The results, illustrated in figure 3, indicate that p53 transduced or control HL60 cells expressed comparable levels of OGAT transcript and protein (fig 3A and B). Both cell lines showed similar basal levels of OGAT activity (fig. 3). Wild-type p53+ and p53- cells were also assayed for OGAT activity before and 24, 48 or 72 hr after treatment with increasing concentrations of TZM. After 24 hr, TZM treatment induced a parallel decrease of activity in both cell lines (fig. 3C). Moreover, resynthesis of new OGAT protein molecules occurred in a similar fashion in both lines, as indicated by the values of OGAT activity obtained at 48 and 72 hr (data not shown).

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Low levels of wild-type p53 do not affect OGAT gene expression. A, Northern blot analysis of OGAT transcript. Equal amounts of total RNA samples (15 µg of RNA) from each cell line (lane 1, p53+ cells; lane 2, p53- cells), were separated by gel electrophoresis and transferred to nylon membrane. The OGAT transcript was identified using a 600 bp [32P]-labeled OGAT cDNA probe. O.D., expressed in arbitrary units, were as follows: p53+ 4.2 O.D.; P53-: 3.9. O.D. B, Western blot analysis of OGAT protein expression. Immunodetection was performed using anti-OGAT polyclonal antiserum. Lane 1, p53+ cells; lane 2, p53- cells. Filter was exposed to autoradiographic film for 30 sec. The results of densitometric analysis were the following: p53+, 3.1 O.D.; p53-, 2.9 O.D. Basal OGAT activities of p53+ and p53- HL60 cells were 62 ± 4 and 56 ± 4 fmol/106 cells, respectively. C, Evaluation of OGAT activity in p53+ () or p53- cells () untreated or exposed to 50, 100 and 200 µM TZM for 24 hr. Data represent the mean of three determinations. Bars: S.E.

Effect of BG on HL60 transduced cells treated with TZM. Cells were deprived of OGAT activity before treatment with TZM using BG. This specific OGAT inhibitor acts as a substrate of the OGAT protein that, upon interaction with BG, becomes inactivated and rapidly degraded (Dolan et al., 1990). OGAT-depleted cells were subsequently exposed to increasing concentrations of TZM (50-200 µM) and cultured in the presence of BG to avoid recovery of OGAT activity, due to resynthesis of new molecules of the enzyme. Treatment of the cells with BG reduced OGAT activity to undetectable levels (data not shown). The results illustrated in figure 4 show that depletion of OGAT activity in p53- cells augmented cell susceptibility to cytotoxicity induced by TZM. Conversely, inhibition of OGAT activity in p53+ cells did not increase the cytotoxic effects of the agent. Moreover, it is noteworthy that TZM treatment inhibited cell growth more efficiently in p53+ cells than in either OGAT-proficient or OGAT depleted p53- cells (fig. 4). Similar results were obtained when apoptosis was evaluated (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Growth inhibition of TZM treated p53+ () and p53- () cells before and after OGAT depletion by BG (1 µM). A, Cells treated with 50, 100 and 200 µM TZM. B, Cells pretreated with BG and then exposed to 50, 100 and 200 µM TZM. Percentages of growth inhibition were calculated comparing cell growth of cells treated with TZM, or pretreated with BG and then exposed to TZM, with that of untreated or BG-treated controls, respectively. The results are representative of one of three repeated experiments. Statistical analysis was calculated according to Student's t test analysis. The differences between p53+ and p53- cells were statistically significant at all drug concentrations (P < .01); no significant differences were observed between p53+ cells treated with TZM and p53+ cells treated with BG and TZM.

Effect of wild-type p53 transfection on bax and bcl-2 protein expression. To investigate whether wild-type p53 protein might affect the expression of gene products that are involved in apoptosis downstream of p53, we evaluated the relative expression of bax and bcl-2 in p53 or pLNSX transduced HL60 cells. The results of Western blot analysis showed that p53+ cells were characterized by basal levels of the bcl-2 protein lower than those detected in control p53- cells (O.D. values: p53+, 6; p53-, 10). However, both cell lines expressed equal levels of bax protein (data not shown).

Treatment with PADPRP inhibitors increases susceptibility to apoptosis induced by TZM in p53- or p53+ cells and in O⁶-methylguanine tolerant cells. In a previous study we demonstrated that treatment of either OGAT-proficient or -deficient cells with the PADPRP inhibitor BZ and TZM resulted in increased apoptosis (Tentori et al., 1997). This result suggested that PADPRP might be involved in coordinating the repair of DNA lesions induced by TZM different from the methylation of the O⁶ position of guanine. In order to test whether a wild-type p53 might potentiate the cytotoxic and apoptotic effects of the combined treatment with TZM and a PADPRP inhibitor, HL60 transduced cells were exposed to a BZ concentration devoid of cytotoxic or apoptotic activity in parental HL60 cells (fig. 5A). Treatment with BZ (5 mM) induced apoptosis only in cells transfected with wild-type p53 (fig. 5A). Coexposure of transduced cells to TZM and BZ caused a marked increase in the percentage of apoptotic cells compared to treatment with TZM alone in both lines (fig. 5A). Similar results were obtained with not transduced HL60 bulk population (fig. 5A). Similar results were obtained when TZM was associated with the PADPRP inhibitor ABZ (data not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Effect of combined treatment with BZ and TZM in leukemia cells. A, p53 null (parental , p53- ) or p53+, () HL60 cells. Untreated or BZ- (5 mM) treated cells were exposed to 200 µM TZM and processed for FACS analysis 72 hr after drug treatment. The figure shows the percentage of cells which exhibit the hypodiploid peak typical of apoptosis. Data represent the mean of two independent experiments. Bars: S.E. B, Jurkat cells, untreated or BZ (5 mM) treated, were exposed to 200 and 400 µM TZM and processed for FACS analysis 72 hr after drug-treatment. Data represent the mean of two independent experiments. Bars: S.E.

Perturbation of cell cycle progression in transduced cells after treatment with ABZ and TZM. Transduced cells were tested for cell cycle arrest at various times (24, 48 and 72 hr) after treatment with ABZ (5 mM) or TZM (200 µM) alone, or with the drug combination. The time course of the percentage of cells in G1 and in S phase are shown in figure 6. Both p53 or pLNSX transduced HL60 cells, untreated or treated with ABZ, showed a similar pattern of cell cycle distribution (fig. 6). The G1 percentage of wild-type p53 transduced cells exposed to the drug combination did not change during the first 48 hr and increased at 72 hr. In contrast, the S phase dramatically decreased during the period of observation, thus indicating that cells that are in G1 do not continue to enter S phase. On the contrary, in pLNSX transduced HL60 cells an initial drop in G1 percentage, was observed with no substantial changes in the percentage of cells in S phase, with respect to untreated or ABZ-treated cells. Both p53+ or p53- cells that were in S phase appeared to continue to progress to G2/M (data not shown). Exposure of wild-type p53 transduced cells to TZM induced a progressive decrease in the percentage of cells in G1 and only a slight increase in the percentage of cells in S (fig. 6) and G2/M phase (data not shown). TZM treatment of pLNSX transduced HL60 cells did not affect the growth rate and cell cycle distribution.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Time course of cell cycle changes in HL60 transduced cell lines. Curves represent the evolution of the percentages of cells in G1 (left panels) and S-phase (right panels). Wild-type p53+ () and p53- cells () were treated with ABZ (5 mM), TZM (200 µM) or the combination of the two drugs. Untreated cells are also represented (CTR). Each time point represents the average of two independent experiments. Bars: S.E.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Positive regulators of apoptosis include p53 protein, which participates in genome surveillance and DNA repair. One of the functions of p53 consists in cell cycle arrest at the G1/S boundary to allow repair of damaged DNA before DNA replication or induction of apoptosis, when DNA damage is too severe. Expression of wild-type p53 in p53-deficient tumor cell lines renders them more susceptible to induction of apoptosis by radiation or DNA-damaging chemotherapeutic drugs (Banerjee et al., 1995).

To test whether the presence of wild-type p53 might modulate apoptosis induced by TZM, a clone from p53 null HL60 cells was infected with an amphotropic virus containing the human p53 cDNA or the control vector. A stable p53+ transfectant was obtained, which presented clearly detectable levels of virus-derived p53 transcript. However, the level of expression of the corresponding translational product was extremely low. Nevertheless, p53 expression in HL60 cells induced growth arrest, as indicated by the marked decrease of the percentage of cells in the S phase of the cell cycle, 20 hr after exposure to low doses of x-rays (1 Gy). A role for p53 protein in G1 arrest is consistent with previously published experiments demonstrating that transfection of wild-type p53 gene into various cell lines induces G1 arrest (Baker et al., 1990; Diller et al., 1990, Wang et al., 1995b).

Restoration of p53 function in p53-null cells increased susceptibility also to apoptosis by TZM. Actually, HL60 cell line is characterized by high levels of OGAT and therefore it is resistant to the cytotoxic and apoptotic effects of TZM concentrations with a clinical relevance (50 µM) (Newlands et al., 1992). Transfection of wild-type p53 in these cells rendered them susceptible to TZM concentrations that did not affect (50 µM) or slightly impaired (100-200 µM) the growth of cells transduced with the control vector.

Pretreatment of transduced lines with the specific OGAT inhibitor, BG, increased cell killing by TZM mainly in p53- cells. It could be hypothesized that differential sensitivity to TZM, could be due to differences in the basal levels of OGAT activity between p53- and p53+ cells. In fact, it has been demonstrated that overexpression of wild-type p53 can suppress transcription of the human OGAT gene (Harris et al., 1996). However, in our study the low levels of p53 protein did not affect OGAT protein expression and activity. Because repair of methylated DNA by OGAT is followed by inactivation of the enzyme, methylating agents indirectly decrease OGAT levels by generating O⁶-methylguanine DNA adducts that are repaired by the alkyltransferase (Lacal et al., 1996). Therefore, OGAT activity was also tested in p53+ or p53- cells after exposure to TZM. Treatment of both transduced lines with TZM equally down-regulated OGAT activity. Moreover, resynthesis of new OGAT protein molecules at 48 and 72 hr occurred in a similar fashion in both lines. These data suggested that the increase of TZM-mediated apoptosis in p53+ cells did not involve inhibition of the repair of O⁶-methylguanine adducts.

It has been reported that p53 activates transcription of death genes (bax) or repress the transcription of survival genes (bcl-2) (Miyashita et al., 1994; Miyashita and Reed, 1995). These two gene products have been implicated as mediators of p53-induced apoptosis (for a review see White, 1996). The results illustrated in our report indicated that the basal levels of bcl-2 protein were lower than those detected in control p53- cells. This raises the possibility that p53 transfection might have down-modulated bcl-2 expression. Thus, the resulting inbalance between bax and bcl-2 proteins might contribute to increase cell susceptibility to apoptotic signals.

A variety of studies indicate that DNA strand-breaks activate PADPRP, which in turn binds strand interruptions and undergoes autoribosylation (Lindahl et al., 1995). After ribosylation, the PADPRP enzyme rapidly dissociates from DNA and allows repair of DNA damage (Satoh and Lindahl, 1992). In vitro activation of PADPRP by methylating compounds, such as TZM, has been previously described (Tisdale, 1985). In addition, it has been demonstrated that PADPRP inhibitors potentiate cytotoxicity of anticancer agents, including TZM (Griffin et al., 1995; Boulton et al., 1995; Tentori et al., 1997). Specifically, we demonstrated that BZ treatment of OGAT-proficient and OGAT-deficient cell lines resulted in a marked enhancement of apoptosis by TZM. In this report, we have shown that treatment of OGAT proficient either p53- or p53+ cells with BZ or ABZ and TZM resulted in increased apoptosis. Furthermore, PADPRP inhibition enhanced TZM susceptibility to apoptosis of Jurkat leukemia cells, which are endowed with high OGAT activity and possess a mutated p53 gene. Thus, treatment with PADPRP inhibitors allowed to increase tumor cell chemosensitivity to methylating compounds even in the presence of mutated p53, which is generally associated with reduced susceptibility to anticancer agents. Moreover, the possibility to enhance cytotoxicity induced by TZM in OGAT-proficient cells appears to be relevant especially when downmodulation of OGAT by BG cannot be achieved (Edara et al., 1996).

When tumor cells are mismatch repair deficient OGAT depletion by BG is ineffective in potentiating methylating agents cytotoxicity (Wedge et al., 1996). We have previously demonstrated that treatment of Jurkat cells with BG inhibited their OGAT activity without increasing their sensitivity to the cytotoxic and apoptotic effects of TZM (Tentori et al., 1995). Thus, these cells show a methylation-tolerant phenotype that presumably derives from a deficiency in the mismatch repair pathways. Therefore it is conceivable to hypothesize that the BZ-mediated increase of apoptosis induced by TZM might be related to an inefficient repair of methyl adducts at N⁷ and N³ positions of guanine.

The results of the analysis of cell cycle distribution in HL60 cells treated with the combination ABZ + TZM showed a G1 arrest only in wild-type p53 transduced cells, as evidenced by the marked and progressive decrease of S phase. Conversely, exposure of p53+ cells to TZM was associated with a dramatic drop in the percentage of cells in G1 phase and a progressive increase in the percentage of cells in S and G2/M phase. These data indicate that the DNA lesions involved in the cytotoxicity associated with the drug combination of TZM and PADPRP inhibitors provoked cell cycle perturbations different from those induced by the exposure of cells to TZM alone. Therefore it is possible to speculate that PADPRP inhibitors might affect the repair of DNA adducts which are processed differently from O⁶-methylguanine, thus inducing a different pattern of cell cycle distribution.

Further studies are required to investigate whether the presence of a functional p53 in mismatch deficient cells might further improve the role of BZ in potentiating temozolomide activity.

In conclusion, these data indicate that the presence of wild-type p53 increases tumor cell killing by TZM in OGAT-proficient cells. Furthermore, the results suggest a possible role of PADPRP inhibitors in potentiating TZM activity against acute leukemias, independently on OGAT or mismatch repair systems and even in the presence of a mutated p53 gene.