Cytotoxic and clastogenic effects of a DNA minor groove ... - Nature

Leukemia (2000) 14, 1451–1459  2000 Macmillan Publishers Ltd All rights reserved 0887-6924/00 $15.00 www.nature.com/leu

Cytotoxic and clastogenic effects of a DNA minor groove binding methyl sulfonate ester in mismatch repair deficient leukemic cells L Tentori1, P Vernole2, PM Lacal3, R Madaio1, I Portarena1, L Levati3, A Balduzzi1, M Turriziani1, P Dande4, B Gold4, E Bonmassar1,3 and G Graziani1 1

Department of Neurosciences, Section of Pharmacology and Medical Oncology, and 2Department of Public Health and Cell Biology University of Rome Tor Vergata, Rome; 3‘Istituto Dermopatico dell’Immacolata’ (IDI-IRCCS), Rome, Italy; and 4Eppley Institute for Research in Cancer and Allied Diseases and Department of Pharmaceutical Sciences, University of Nebraska Medical Center, Omaha, Nebraska, USA

Mismatch repair deficiency contributes to tumor cell resistance to O6-guanine methylating compounds and to other antineoplastic agents. Here we demonstrate that MeOSO2(CH2)2lexitropsin (Me-Lex), a DNA minor groove alkylating compound which generates mainly N3-methyladenine, has cytotoxic and clastogenic effects in mismatch repair-deficient leukemic cells. Moreover, MT-1 cells, which express p53 upon drug treatment and possess low levels of 3-methylpurine DNA glycosylase activity, are more susceptible to cytotoxicity induced by MeLex, with respect to p53-null and 3-methylpurine DNA glycosylase-proficient Jurkat cells. In both cell lines, the poly(ADPribose) polymerase inhibitor 3-aminobenzamide, which inhibits base excision repair capable of removing N-methylpurines, increases cytotoxicity and clastogenicity induced by Me-Lex or by temozolomide, which generates low levels of N3-methyl adducts. The enhancing effect is more evident at low Me-Lex concentrations, which induce a level of DNA damage that presumably does not saturate the repair ability of the cells. Nuclear fragmentation induced by Me-Lex + 3-aminobenzamide occurs earlier than in cells treated with the single agent. Treatment with Me-Lex and 3-aminobenzamide results in augmented expression of p53 protein and of the X-ray repair crosscomplementing 1 transcript (a component of base excision repair). These results indicate that N3-methyladenine inducing agents, alone or combined with poly(ADP-ribose) polymerase inhibitors, could open up novel chemotherapeutic strategies to overcome drug resistance in mismatch repair-deficient leukemic cells. Leukemia (2000) 14, 1451–1459. Keywords: base excision repair; N3-methyladenine; mismatch repair; MeOSO2(CH2)2-lexitropsin; poly(ADP-ribose) polymerase

Introduction High levels of O6-alkylguanine-DNA alkyltransferase (OGAT) or loss of DNA-mismatch repair (MR) activity are associated with resistance to methylating agents.1,2 In the latter case, resistance is due to reduced MR-dependent activation of signal transduction pathways which ultimately lead to apoptosis. In fact, O6-methylguanine pairs either with thymine or cytosine to form mismatched substrates that trigger the intervention of MR. Futile attempts to find the complementary base for O6-methylguanine eventually lead to the formation of strand breaks and apoptosis.3 In MR-proficient cells, resistance to methylating agents, due to high levels of OGAT, can be abrogated by depletion of OGAT activity using O6-benzylguanine, that is presently undergoing phase I studies.4,5 To date, there are no therapeutic strategies that can reverse resistance induced by functional loss of MR. Actually, MR deficiency is also associated with low-level resistance to other antineoplastic agents

Correspondence: G Graziani, Department of Neurosciences, University of Rome Tor Vergata, Via di Tor Vergata 135, 00133 Rome, Italy; Fax: 39-6-72596323 Received 5 January 2000; accepted 27 March 2000

which do not methylate DNA, such as busulfan, platinum compounds, doxorubicin, etoposide and 6-thioguanine.6 A functional MR is required for the maintenance of the genetic stability of DNA. Loss of its function leads to a mutator phenotype, characterized by a high rate of mutations throughout the genome. A consequence of the mutator phenotype is the generation of multiple replication errors in repetitive DNA sequences, resulting in microsatellite instability (MSI). Thus, the presence of MSI in tumor cells has been considered an indirect evidence of MR gene disruption. In hematological malignancies MSI and MR gene mutations are not as common as in solid tumors. Nevertheless, they have been described in lymphoid tumors and in acute myeloid leukemias secondary to previous chemotherapy or arising in elderly patients.7–9 We recently demonstrated that MR-deficient leukemic cells can be rendered sensitive to the methylating agent temozolomide (TZM), when this agent is combined with poly(ADPribose) polymerase (PARP) inhibitors.10,11 Since PARP is a component of the base excision repair system (BER), we suggested that cytotoxic and apoptotic effects of TZM and PARP inhibitors in cells tolerant to O6-methylguanine, might be the consequence of reduced repair of other methyl lesions generated from TZM. N3-methyladenine (N3-MeA) and N3 and N7-methylguanines are among the modified DNA bases, that induce the single-nucleotide replacement mediated by BER. The first step of the BER process is the excision of the modified base by damage-specific DNA glycosylases. This is followed by incision of the phosphodiester bond 5⬘ to the lesion by an apurinic/apyrimidinic endonuclease.12 The removal of the abasic residue generates a nucleotide gap, which is finally filled by the co-ordinate intervention of PARP, DNA polymerase ␤, X-ray repair cross-complementing 1 (XRCC1), and ligase I and III. Although methylating agents react to some extent with DNA at all four bases, the adduct profile for each class of compounds is often different. For instance, in the case of TZM at least 70% of the methyl adducts are located at N7-guanine, 9% at N3-adenine and 5% at O6-guanine.13 Moreover, the steady-state level of DNA adducts formed is dependent not only upon the rate of alkylation but also upon the efficiency of repair systems. Even though N7-methylguanine is the most abundant adduct generated by methylating agents, it appears to have little or no lethal effects.14 In contrast, unrepaired N3-MeA is an extremely toxic DNA lesion, that is capable of inhibiting DNA replication and of inducing apoptosis and chromosome aberrations.15 In fact, 3-methylpurine DNA glycosylase (MPG)-deficient murine cells are highly sensitive to MeOSO2(CH2)2-lexitropsin (Me-Lex).15,16 Moreover, the cytotoxicity of this agent is greatly enhanced in bacterial that cannot repair N3-MeA.17 Me-Lex is a methyl sulfonate ester tethered to N-methylpyrrolecarboxamide dipeptide, which

N3-methyladenine-induced toxicity in mismatch repair-deficient leukemic cells L Tentori et al

1452

targets the minor groove of AT-rich DNA sequences. It almost exclusively produces N3-MeA (⬎99%), as demonstrated in previous studies in which DNA adducts from Me-Lex have been quantitated by treating genomic DNA or mammalian cells.18,19 In contrast to Me-Lex, the level of N3-MeA generated by TZM is extremely low and, consequently, in the presence of a functional BER system, this damaged base is rapidly repaired. Inhibition of PARP is likely to hamper the completion of BER process, which occurs after removal of the methylated base by MPG. The aim of the present study is to investigate whether MRdeficient leukemic cells are susceptible to the cytotoxic effects of N3-MeA and whether PARP inhibition might increase chemosensitivity to this lesion. MR-deficient cells and Me-Lex provide an ideal model to clarify better the biological consequences deriving from the incomplete repair of N3-MeA, which would occur when a PARP inhibitor is combined either with Me-Lex or TZM. Furthermore, this model allows exploration of new pharmacological modalities to overcome tumor cell resistance to O6-methylating agents and possibly to other antineoplastic drugs. Materials and methods

Cell line and culture conditions The T lymphoblastic leukemia cell line Jurkat and the human colon cancer cell line HT-29 were purchased from the American Type Culture Collection (ATCC, Rockville, MD, USA). The human lymphoblastoid cell line MT-120 was a generous gift from WG Thilly (Massachusetts Institute of Technology, Cambridge, MA, USA). Both MT-1 and Jurkat are MR-deficient cell lines.20,21 Jurkat cells are p53-null since they harbor in their genome a mutated form of p53 gene,22 whereas MT-1 are capable of inducing p53 upon treatment with DNA damaging agents.3 Cell lines were cultured in RPMI-1640 (Gibco, Paisley, UK) supplemented with 10% fetal calf serum (Gibco), 2 mM L-glutamine, 100 units/ml penicillin and 100 ␮g/ml streptomycin (Flow Laboratories, McLean, VA, USA), at 37°C in a 5% CO2 humidified atmosphere.

Drugs Me-Lex was prepared as previously described.23 The PARP inhibitor 3-aminobenzamide (3-AB) was purchased from Sigma (St Louis, MO, USA). TZM was kindly provided by Shering Plough Research Institute (Kenilworth, NJ, USA). Drug stock solutions were prepared by dissolving Me-Lex in 95% ethanol (10 mM), TZM in dimethyl sulfoxide (100 mM) and 3AB in RPMI-1640 (16 mM). The final concentration of dimethyl sulfoxide or ethanol in drug-treated cultures was always less than 0.5% (v/v) and did not contribute to toxicity (data not shown).

Drug treatment and cell growth evaluation Cells were cultured in flasks (Falcon, Becton Dickinson Labware, Oxnard, CA, USA) at the concentration of 3 x 105 cells/ml. Inhibition of PARP was obtained by treating the cells with 4 mM 3-AB, a concentration which has been described Leukemia

as completely inhibiting PARP activity.24 Me-Lex or TZM were added to cell cultures either alone or immediately after 3-AB addition, and it was used at concentrations ranging from 1.5 to 50 ␮M and 62.5 to 250 ␮M, respectively. Cells were then incubated at 37°C for 3 days and cell growth was evaluated, every 24 h, by counting viable cells in quadruplicate. Cell viability was determined by trypan blue dye exclusion. Cell line chemosensitivity to Me-Lex was evaluated in terms of IC50, ie the concentration of the drug expressed in ␮M, capable of inhibiting cell growth by 50%. The IC50 was calculated on the regression line in which the number of viable cells was plotted vs the drug concentration.

Assessment of apoptosis by flow cytometry analysis Cells were harvested, washed with phosphate-buffered saline and fixed in 70% ethanol at −20°C for 18 h. After centrifugation, cells were resuspended in 1 ml of hypotonic solution containing 50 ␮g/ml propidium iodide (PI; Sigma), 0.1% sodium citrate, 0.1% Triton-X, and 10 ␮g/ml RNase (Boerhinger Mannheim, Milan, Italy), and incubated at 37°C in the dark, for 15 min. Data collection was gated utilizing forward light scatter and side light scatter to exclude cell debris and aggregates. PI fluorescence was measured on a linear scale using a FACSscan flow cytometer (Becton Dickinson, San Jose, CA, USA). Apoptotic cells are represented by a broad hypodiploid peak, which is easily distinguishable from the narrow peak of cells with diploid DNA content in the red fluorescence channel.

Analysis of chromosome aberrations For chromosomal analysis, cell cultures were incubated for 2 h with 0.8 ␮g/ml of colchicine (Sigma) before harvesting. Metaphase preparations were obtained according to standard techniques, treating cells first with a hypotonic solution (KCl 0.075 M) and then with a fixative mixture methanol/acetic acid (3/1). Air-dried slides were stained with 5% Giemsa in phosphate buffer. Chromatid and chromosome breaks were evaluated by analyzing one hundred metaphases for each culture. Chromatid exchanges (tri- and tetra-radials) were counted as two chromatid breaks and chromosome exchanges (dicentric and translocations) as two chromosome breaks. The number of cells with aberrations was also calculated. For each slide the mitotic indexes (MI) (ie the percentage of dividing cells on 1000 nuclei examined), and the percentages of nuclei that appeared fragmented were also evaluated. Statistical analysis was performed using the normal standardized deviate as described.25

Western blotting Cell lysates and samples for electrophoresis were prepared as previously described.11 Eighty ␮g of protein per sample were electrophoresed in 12% SDS-polyacrylamide mini-gels. Afterwards, proteins were transferred to nitrocellulose membranes (Schleicher & Schuell, Keene, NH, USA). Equal protein loading was visualized by Ponceau S staining. Filters were blocked with blocking buffer (Boehringer Mannheim) and incubated with the monoclonal antibody directed against p53 (PAb1801) (Oncogene, Cambridge, MA, USA). Immune-complexes were


visualized using a chemiluminescence kit (Amersham International, Amersham, UK), according to the manufacturer’s instructions. Filters were exposed to X-OMAT AR autoradiographic films (Kodak) for 10–45 s, depending on the intensity of the signal. Bidimensional densitometry of the blots was performed using an Imaging densitometer, GS-670 (BioRad, Richmond, CA, USA).

Measurement of MPG activity MPG activity was assayed as previously described.26 Tumor cells (107) were sonicated at 4°C in 0.5 ml buffer I (50 mM Tris-HCl, 3 mM dithiothreitol, and 2 mM EDTA, pH 8.3), with freshly added 1 mM 4-(2-aminoethyl)-benzene-sulfonyl fluoride hydrochloride. After removal of cell debris by centrifugation, supernatants were immediately tested for MPG activity. Various amounts of tumor cell extracts were incubated with 10 ␮g (10 000 c.p.m.) of freshly dissolved calf thymus DNA methylated by N-3H-methyl-N-nitrosourea (19 Ci/mmol; Amersham), in a total volume of 100 ␮l of buffer II (20 mM Tris-HCl, 1 mM dithiothreitol, 60 mM NaCl, and 1 mM EDTA pH 8). After 1 h at 37°C, the reaction was stopped on ice by the addition of 30 ␮l of 2 m NaCl containing 0.5 mg/ml calf thymus DNA and 1 mg/ml bovine serum albumin. DNA was ethanol-precipitated and samples were centrifuged at 10 000 g for 15 min. Three hundred ␮l of the supernatants were transferred to a scintillation tube and counted. The MPG activity was determined for protein and time-limiting conditions and expressed as fmol of methyl groups released per 1 × 106 cells per hour.

Northern blot analysis Total cellular RNA was extracted using the TriPure isolation reagent (Boehringer Mannheim). RNAs (15 ␮g) were fractionated by electrophoresis on a formaldehyde-containing 1.2% agarose gel. The integrity of RNA was confirmed by RNA visualization, adding ethidium bromide to the RNA gel loading buffer. RNA was transferred to nylon membrane (Gene Screen Plus; NEN Life Science Products, Boston, MA, USA) and hybridized at 68°C for 1 h with a 32P-labeled probe, using the QuickHyb hybridization solution (Stratagene, Cambridge, UK). Washing of the blots was performed according to the manufacturer’s instructions. The blotted membrane was exposed to X-ray film at −80°C. The DNA probe used for the detection of XRCC1 was obtained by polymerase chain reaction as previously described.11 The glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probe, a 0.9 kb EcoRI fragment of the human GAPDH gene, was a generous gift from Dr R Dalla Favera (Department of Pathology, Columbia University, New York, NY, USA). Bidimensional densitometry of the blots was performed as described for Western blot analysis. Results

Cytotoxic effects of Me-Lex, alone or combined with PARP inhibitor, in MR-deficient cells In order to investigate whether MR-deficient leukemic cells are susceptible to cytotoxicity induced by N3-MeA, Jurkat or MT-1 cells were exposed to graded concentrations of Me-Lex

and cell growth was evaluated, at daily intervals, during 72 h of culture. Moreover, it was also explored whether the PARP inhibitor 3-AB might be capable of potentiating the cytotoxic effects of Me-Lex. Treatment with Me-Lex alone inhibited the growth of Jurkat or MT-1 cells at different extents depending on the drug concentrations used (Figure 1). Actually, Me-Lex induced a more pronounced growth inhibitory effect in MT-1 cells with respect to Jurkat cells. The IC50s of the drug for each cell line, calculated with the data of three independent experiments, were as follows: 7.6 ± 0.3 ␮M (MT-1) and 13.8 ± 0.3 ␮M (Jurkat) at 48 h, or 4 ± 0.1 ␮M (MT-1) and 8.2 ± 0.2 ␮M (Jurkat) at 72 h of culture. When the drug was combined with the PARP inhibitor 3AB, an increase of Me-Lex-mediated cytotoxicity was observed (Figure 2). This effect was more evident in both cell lines at 24 h after drug treatment and with the lowest drug concentrations (1.5 and 3.1 ␮M). Treatment with TZM alone (62.5–250 ␮M) did not substantially affect the growth of Jurkat or MT-1 cells. In contrast, when the drug was associated with 3-AB a significant (P ⬍ 0.01) reduction of cell growth with respect to controls, was observed at all the drug concentrations tested, in both cell lines (Ref. 11, and data not shown).

1453

PARP inhibition enhances chromosome aberrations induced by Me-Lex or TZM Previous studies indicated that Me-Lex induces chromosome aberrations in murine embryonic stem (ES) cells.15 Studies were performed to explore whether this agent also damaged chromosomes in human leukemic cell lines. Therefore, Jurkat and MT-1 cells were treated with graded concentrations of Me-Lex and metaphase cells were analyzed for the presence of chromosome aberrations. MT-1 cultures appeared to be more susceptible than Jurkat cells to the cytotoxic effect of the N3-methylating agent. The mitotic indexes of MT-1 cells were heavily depressed and to a higher extent than those of Jurkat cells. Therefore, in the case of MT-1 only cultures exposed to Me-Lex concentrations lower (ie 3.1 and 6.2 ␮M) than those used for Jurkat cells could be analyzed. In both cell lines all drug treatments, with only the exception of 3.1 ␮M Me-Lex alone in MT-1 cells at 24 h, induced a significant (P ⬍ 0.01) number of aberrations (Figure 3), compared with values found in untreated or in 3-AB-treated cultures. Moreover, chromosome damage was evident as early as 6 h after drug exposure (Figure 3). Addition of 3-AB to Me-Lex increased the clastogenic effect of the drug (Figure 3). In fact, a significantly higher number of aberrations was found in the groups exposed to the drug combination with respect to that detected in groups treated with the single agent (Figure 3). Actually, in the case of MT1 cells statistically significant differences were observed only at 24 h. Moreover, in cells treated with 3-AB and Me-Lex the number of aberrant cells was significantly higher with respect to that of cells treated with Me-Lex alone (data not shown). TZM treatment produced, in Jurkat cells, significant clastogenic effects only when associated with the PARP inhibitor (data not shown). No chromosome aberrations higher than those detectable in untreated controls, were found, instead, when Jurkat cells were exposed to 3-AB alone (data not shown). Leukemia


1454

Figure 1 Cell growth analysis of Jurkat or MT-1 cells treated with Me-Lex. Cells were exposed to graded concentrations of Me-Lex and the numbers of viable cells were counted in quadruplicate, at daily intervals, during 72 h of culture after drug treatment. The results are expressed as percentage of viable cells with respect to untreated control cultures and are representative of one out of three independent experiments with similar results. Cell viability was determined by trypan blue dye exclusion. The standard error (s.e.) of the mean value of quadruplicate counts never exceeded 5% of the mean for each sample.

Figure 2 Growth inhibition of Jurkat or MT-1 cells treated with Me-Lex, as single agent or combined with PARP inhibitor. Cells were treated with Me-Lex (solid column) or with 3-AB + Me-Lex (striped column) and cell growth was evaluated at 24 h of culture. The results are represented as percentages of growth inhibition calculated comparing cell growth of cultures treated with the methylating agent, alone or combined with 3-AB, with that of controls. Data represent the mean values of three independent experiments. The indicated means and the relative s.e. (bars) were calculated following angular transformation of the percentage of inhibition values. Differences between cells treated with the drug combination (Me-Lex + 3-AB) and cells treated with the single methylating agent (Me-Lex) were statistically significant in all cases (P ⬍ 0.01).

Analysis of apoptosis and p53 induction mediated by Me-Lex, alone or in combination with PARP inhibitor In order to test whether Me-Lex might induce apoptosis in human MR-deficient cell lines, Jurkat and MT-1 cells were exposed to graded concentrations of Me-Lex, ranging from 3.1 ␮M to 12.5 ␮M. Apoptosis was evaluated, 24 h after drug exposure, measuring sub-G1 DNA content by flow cytometry after staining of the cells with PI (Figure 4). The results indicate that at 24 h Me-Lex induced apoptosis only in a limited percentage of Jurkat cells, but it was slightly more effective in MT-1 (Figure 4). Moreover, treatment with 3-AB always increased the apoptotic effects of the drug (Figure 4). Morphological analysis of the nuclear apoptotic modifications occurring in treated cells was also performed. Fragmented nuclei were counted and the results were expressed as the fragmented nuclei ratio between drug-treated and untreated or 3-AB-treated groups. The results indicate that Jurkat cells treated with elevated Me-Lex concentrations (25 ␮M and 50 ␮M) in combination with 3-AB showed substantially higher levels of nuclear fragmentation compared to cells treated with Me-Lex alone, either at 6 h or at 24 h after treatment (Figure 5). DNA damage resulting from the presence of unrepaired N3MeA or from MR-activation by O6-methylguanine is known to induce p53 expression.3,15 In the present model system we have investigated whether the failure of BER to complete the Leukemia

repair process due to PARP inhibition might induce p53 (Figure 6). MT-1 cells, endowed with a wild-type p53, were exposed to Me-Lex (3.1 ␮M and 6.2 ␮M), alone or in combination with 3-AB, and levels of p53 protein were analyzed by Western blot. The results, illustrated in Figure 6, indicated that Me-Lex either used as single agent or in combination induced p53 as early as 6 h after drug exposure. At the concentrations of 3.1 ␮M, the expression of p53 was higher in cells treated with the drug combination than in cells treated with Me-Lex alone.

Analysis of MPG activity in Jurkat and MT-1 cells We have investigated whether the higher susceptibility of MT1 cells with respect to Jurkat cells might be at least in part due to low levels of MPG. The MPG activity in MT-1 cells was barely detectable (limit of detection is 5 fmol/106 cells/h), whereas Jurkat cell extracts showed a higher level of activity (49 fmol/106 cells/h). In these experimental conditions Jurkat cells showed a 1.6-fold lower activity than HT-29 cells (80 fmol/106/h), in accordance with a previous report.27 HT-29 cells were used as positive controls since they express high levels of MPG activity.28


1455

Figure 3 Chromosome aberrations observed after treatment of Jurkat or MT-1 cells with Me-Lex, alone or in combination with 3-AB. Shown are the number of chromosome aberrations per 100 cells detected in untreated or Me-Lex treated (solid column) and in 3-AB or 3-AB treated + Me-Lex (striped column) cells. Statistically significant differences between groups treated with the drug combination and Me-Lex treated groups are indicated (**, P ⬍ 0.01).

Figure 4 Effect of 3-AB treatment on apoptosis induced by Me-Lex in Jurkat and MT-1 cells. Cells were treated with Me-Lex (solid column) or with 3-AB plus Me-Lex (striped column) and apoptosis was measured by flow cytometry analysis of the subG1 DNA content at 24 h posttreatment. Data are expressed as percentage of apoptotic cells and represent the mean of three independent experiments, ± s.e., calculated as described in Figure 2.

Figure 5 Nuclear fragmentation in Jurkat cells treated with Me-Lex, alone or combined with 3-AB. Data were represented as treated/control ratio, defined by the percentage of fragmented nuclei detected in Me-Lex (쐌) or in 3-AB + Me-Lex (앩) treated cultures divided by the percentage of fragmented nuclei observed in untreated or 3-AB treated controls, respectively. Leukemia


1456

Figure 6 Induction of p53 in MT-1 cells treated with Me-Lex, alone or combined with 3-AB. Western blot analysis of p53 protein levels was performed 6 h after exposure to Me-Lex or Me-Lex combined with PARP inhibitor. The ratio between the optical densities of groups treated with the drug combination and those treated with the single agent were: 1.1 and 1.96, at 6.2 ␮M and 3.1 ␮M, respectively. The data are representative of one out of two independent experiments with similar results.

Me-Lex treatment modulates XRCC1 gene transcript We have previously demonstrated that treatment of Jurkat cells with TZM associated with 3-AB increased the expression of XRCC1 gene transcript, a component of BER complex.11 Therefore, it was hypothesized that in cells tolerant to O6methylguanine, this could be due to unrepaired strand breaks derived from the processing of methyl-adducts different from O6-methylguanine. In this study, we investigated whether modulation of XRCC1 gene transcript might occur after treatment with PARP inhibitor and a methylating agent which mainly generates N3-MeA adducts. Expression of XRCC1 after drug treatment of Jurkat cells changed upon drug treatment. In particular, XRCC1 was found to be augmented in cells either exposed to the drug combination or, albeit less markedly with the single agent at the highest concentration (Figure 7). Discussion Methylation of O6-guanine has been considered as the main DNA lesion responsible for the cytotoxicity, mutagenicity and clastogenicity of methylating compounds in cells endowed with a functional MR.29–31 Loss of MR function results not only in resistance to these agents and to other antineoplastic drugs, but is also associated with genetic instability. Thus, drug treatment itself might contribute in enriching genetically unstable cells, which rapidly acquire resistance to different antitumor agents. Considering this scenario it was of interest to analyze the possible biological consequences of DNA damage other than that derived from the processing of O6-methylguanine by MR. Actually, PARP inhibitors are capable of increasing the susceptibility of MR-deficient leukemic cells to TZM, presumably by affecting repair of methyl adducts different from O6methylguanine such as N3-MeA. Therefore, we have focused our attention on the effect on cell growth, apoptosis, chromosome aberrations of a selective N3-MeA inducing agent, and the modulation of this effect by PARP impairment. Moreover, the results obtained with Me-Lex were compared with those obtained with TZM, which methylates N3-adenine to a minor extent. Leukemia

Figure 7 Expression of XRCC1 gene in Jurkat cells treated with Me-Lex, in the presence or absence of 3-AB. Expression of XRCC1 was evaluated by Northern blot analysis. The 2.1 kb and 1.2 kb transcripts of the XRCC1 and GAPDH genes, respectively, are indicated. The hybridization signals were quantified by densitometric scanning of the autoradiograms and normalized in relation to GAPDH. Histograms indicate the value of the ratio between the optical densities of Me-Lex or 3-AB + Me-Lex-treated groups and those of untreated or 3AB-treated controls. The results are representative of one out of three different experiments.

The findings of the present study indicate that Me-Lex is cytotoxic for MR-deficient Jurkat and MT-1 leukemic cell lines and that PARP inhibition increases toxicity of the N3-methylating agent. The kinetics of appearance of the growth inhibitory effect was different from that resulting from the processing of O6-methylguanine in MR-proficient and OGAT-deficient cells. In fact, cytotoxicity induced by Me-Lex or by the drug combination was already evident 24 h after drug exposure. The killing effect, instead, deriving from O6-methylguanine appeared only at 48 h,3 suggesting that permanent DNA damage would occur during the second cycle of DNA replication. In this case the replication fork encounters MR-induced strand interruptions and generates double-strand breaks.2 In contrast, N3MeA behaves as a replication blocking lesion, since the presence of the methyl group would alter the ability of the N3 atom to stabilize the contact between DNA polymerase and DNA template.15 This might explain the early appearance of cytotoxicity already evidenced within the first round of DNA replication. The enhancing effect on Me-Lex cytotoxicity by PARP inhibitor was detectable in both lines although it was more evident at low drug concentrations that slightly affected cell growth at 24 h. This phenomenon might be due to the interruption of the BER repair process, after the removal of N3MeA by MPG, leading to strand break generation. Similar enhancement of cytotoxicity was previously observed when leukemic cells were exposed to TZM combined with PARP inhibitors.11 MR-deficient Jurkat cells express high levels of OGAT activity and are capable of repairing the low percent-


1457

Figure 8 Biological consequences deriving from the presence of N3-MeA or from interruption of its repair by PARP inhibition. Short patch BER pathway involving single nucleotide replacement is illustrated. DNA damage deriving either from persistence of N3-MeA (left gray arrow) or from incomplete repair of N3-MeA due to PARP inhibition (right gray arrow) induce cytotoxic, apoptotic and clastogenic effects in MRdeficient leukemic cells.

age of O6-methylguanine generated by TZM.32 Therefore, it is reasonable to hypothesize that PARP inhibition might impair the repair process of adducts generated by TZM at DNA site positions different from O6-guanine. It has been recently demonstrated that the presence of N3MeA in the DNA is capable of inducing chromosome aberrations.15 The clastogenic effect of N3-methylguanine would derive from a block of the replication fork, leaving short unreplicated regions which are prone to be converted into double-strand breaks. In the present study, we show that in human MR-deficient leukemic cells Me-Lex is capable of inducing clastogenic effects and that PARP inhibition enhanced the clastogenic potential of the agent. It seems reasonable to hypothesize that DNA strand interruptions, deriving from the incomplete repair of N3-MeA, might be converted in double-strand breaks that would give rise to chromosome aberrations. Methylating triazene compounds have been previously shown to have clastogenic effects.33,34 In the case of TZM, OGAT-deficient cells, or cells pretreated with the OGAT inactivator O6-benzylguanine, are extremely sensitive to chromosome aberrations induced by treatment with methylating agents.33 However, OGAT overexpression did not completely abolish the appearance of the clastogenic effect of methylating agents, thus indicating that DNA adducts different from O6methylguanine might also contribute to the clastogenicity of these compounds. In our experimental model, where a MRdeficient cell line with high levels of OGAT activity (ie Jurkat) has been used, TZM alone did not induce a significant chromosome damage. Impairment of PARP activity increased this effect, suggesting that repair inhibition of DNA adducts different from O6-methylguanine is responsible for the clastogenic properties of TZM. The apoptotic effect of Me-Lex alone was more pronounced in MT-1 than in Jurkat cells at 24 h (Figure 3). Moreover, 3-AB significantly increased Me-Lex-induced apoptosis. Apoptosis

consequent to Me-Lex treatment along with interruption of BER process due to 3-AB-induced PARP inhibition, appears at an earlier time than that induced by the single agent. Consistent with this observation, fragmented nuclei were detected in Jurkat cells exposed to Me-Lex + 3-AB, but not to Me-Lex alone, as early as 6 h after treatment. The present results clearly demonstrate that MT-1 cells were more susceptible to the cytotoxic effects of the N3-adenine methylating agent than Jurkat cells. Consistent with the direct role of N3-MeA in cytotoxicity, MT-1 cells expressed very low MPG activity in vitro, with respect to that of Jurkat cells. Moreover, it cannot be excluded that the activation of p53 by drug treatment might contribute to the marked apoptotic effect induced by the drug combination in MT-1 cells. There are conflicting data regarding the role of MPG in resistance to alkylating agents. A clear relationship between MPG expression and resistance to these drugs has not been established.35,36 However, MPG-null ES murine cells were found to be more susceptible to Me-Lex and to methyl methanesulphonate, with respect to MPG-positive ES cells.15,16 In our model system, it is possible to hypothesize that the low levels of MPG activity present in MT-1 cells might be sufficient to repair the limited number of N3-methyl adducts induced by extremely low concentrations of Me-Lex. In this case the inhibition of the BER repair process, that follows the intervention of MPG could result in enhanced cytotoxicity. However, at high Me-Lex concentrations, the amount of active enzyme present in the cells becomes a rate limiting factor during the repair process. In these conditions, 3-ABinduced BER impairment appears to play a less important role. Treatment with Me-Lex increased expression of XRCC1 transcript, which was more pronounced when Me-Lex was combined with 3-AB. XRCC1 is a member of the BER multiprotein complex which is capable of interacting with PARP, DNA polymerase ␤ and DNA ligase III.37,38 Increased XRCC1 expression was also detected in Jurkat cells treated with TZM Leukemia


1458

and 3-AB, thus suggesting that unrepaired strand breaks might trigger the synthesis of this molecule.11 The results obtained with Me-Lex alone or combined with 3-AB reinforce this hypothesis. The increase of XRCC1 transcript induced by MeLex alone, could be due to the renewed synthesis of the molecule consumed during the very active repair process. In conclusion, the present work shows that Me-Lex induces cytotoxic, apoptotic and clastogenic effects in MR-deficient leukemic cells and that PARP inhibition potentiates cytotoxicity of this compound (see model illustrated in Figure 8). MRdeficient cells, which are resistant to the toxicity induced by O6-methylguanine, and possibly to various antineoplastic agents, remain susceptible to methylating agents capable of inducing N3-MeA adducts. In this case, high levels of MPG activity might induce resistance to these compounds, especially to those which generate low amounts of N3-MeA. However, association of the methylating agent to a PARP inhibitor would result in enhanced cytotoxicity in tumor cells with either high or low levels of MPG activity. These results appear to be of potential clinical interest, because the blasts obtained from more than 90% of the patients with therapy-related leukemia present microsatellite instability, which is a hallmark of the genomic instability consequent to the loss of MR.8 Therefore, therapeutic strategies based on the selective methylation of N3-adenine and/or inhibition of its repair might be of benefit for the treatment of MR-deficient tumors, which might be resistant to different antineoplastic agents.

Acknowledgements This study was supported by a grant from the Italian Association for Cancer Research (AIRC) and, partly, by MURST funds to E Bonmassar (Molecular bases for the pharmacological control of neoplastic diseases). The study on chromosome aberrations was supported by MURST funds to P Vernole.

References 1 Pegg AE, Dolan ME, Moschel RC. Structure, function and inhibition of O6-alkylguanine-DNA alkyltransferase. Prog Nucleic Acid Res Mol Biol 1995; 51: 167–223. 2 Jiricny J. Eukaryotic mismatch repair: an update. Mutat Res 1998; 409: 107–121. 3 D’Atri S, Tentori L, Lacal PM, Graziani G, Pagani E, Benincasa E, Zambruno G, Bonmassar E, Jiricny J. Involvement of the mismatch repair system in temozolomide-induced apoptosis. Mol Pharmacol 1998; 54: 334–341. 4 Dolan ME, Mitchell RB, Mummert C, Moschel RC, Pegg AE. Effect of O6-benzylguanine analogues on sensitivity of human tumour cells to the cytotoxic effects of alkylating agents. Cancer Res 1991; 51: 3367–3372. 5 Friedman HS, Kokkinakis DM, Pluda J, Friedman AH, Cokgor I, Haglund MM, Ashley DM, Rich J, Dolan ME, Pegg AE, Moschel RC, McLendon RE, Kerby T, Herndon JE, Bigner DD, Schold SC Jr. Phase I trial of O6-benzylguanine for patients undergoing surgery for malignant glioma. J Clin Oncol 1998; 16: 3570–3575. 6 Fink D, Aebi S, Howell SB. The role of DNA mismatch repair in drug resistance. Clin Cancer Res 1998; 4: 1–6. 7 Ben-Yehuda D, Krichevsky S, Caspi O, Rund D, Polliack A, Abeliovich D, Zelig O, Yahalom V, Paltiel O, Or R, Peretz T, BenNeriah S, Yehuda O, Rachmilewitz EA. Microsatellite instability and p53 mutations in therapy-related leukemia suggest mutator phenotype. Blood 1996; 88: 4296–4303. 8 Zhu Y-M, Das-Gupta EP, Russell NH. Microsatellite instability and p53 mutations are associated with abnormal expression of the MSH2 gene in adult acute leukemia. Blood 1999; 94: 733–740. Leukemia

9 Indraccolo S, Minuzzo S, Nicoletti L, Cretella E, Simon M, Papakonstantinou G, Hehlmann R, Mion M, Bertorelle R, Roganovich J, Chieco-Bianchi L. Mutator phenotype in human hematopoietic neoplasms and its association with deletions disabling DNA repair genes and bcl-2 rearrangements. Blood 1999; 94: 2424– 2432. 10 Tentori L, Lacal PM, Benincasa E, Franco D, Faraoni I, Bonmassar E, Graziani G. Role of wild-type p53 on the antineoplastic activity of temozolomide alone or combined with an inhibitor of poly(ADP-ribose) polymerase. J Pharmacol Exp Ther 1998; 285: 884–893. 11 Tentori L, Turriziani M, Franco D, Serafino A, Levati L, Roy R, Bonmassar E, Graziani G. Treatment with temozolomide and poly(ADP-ribose) polymerase inhibits induces early apoptosis and increases base excision repair gene transcripts in leukemic cells resistant to triazene compounds. Leukemia 1999; 13: 901–909. 12 Wilson DM III, Thompson LH. Life without DNA repair. Proc Natl Acad Sci USA 1997; 94: 12754–12757. 13 Newlands ES, Stevens MFG, Wedge SR, Wheelhouse RT, Brock C. Temozolomide: a review of its discovery, chemical properties, pre-clinical development and clinical trials. Cancer Treat Rev 1997; 23: 35–61. 14 Lawley PD, Phillips DH. DNA adducts from chemotherapeutic agents. Mutat Res 1996; 355: 13–40. 15 Engelward BP, Allan JM, Dreslin AJ, Kelly JD, Wu MM, Gold B, Samson LD. A chemical and genetic approach together define the biological consequences of 3-methyladenine lesions in the mammalian genome. J Biol Chem 1998; 273: 5412–5418. 16 Engelward BP, Dreslin A, Christensen J, Huszar D, Kurahara C, Samson LD. Repair-deficient 3-methyladenine DNA glycosylase homozygous mutant mouse cells have increased sensitivity to alkylation-induced chromosome damage and cell killing. EMBO J 1996; 15: 945–952. 17 Dinglay S, Gold B, Sedgwick B. Repair in Escherichia coli alkB mutants of abasic sites and 3-methyladenine residues in DNA. Mutat Res 1998; 407: 109–116. 18 Encell L, Shuker, DE, Foiles PG, Gold B. The in vitro methylation of DNA by a minor groove binding methyl sulfonate ester. Chem Res Toxicol 1996; 9: 563–567. 19 Kelly J, Inga A, Chen F-X, Dande P, Shah D, Monti P, Aprile A, Burns PA, Scott G, Abbondandolo A, Gold B, Fronza G. Relationship between DNA methylation and mutational patterns induced by a sequence selective minor groove methylating agent. J Biol Chem 1999; 274: 18327–18334. 20 Kat A, Thilly WG, Fang WH, Longley MJ, Li GM, Modrich P. An alkylation-tolerant, mutator human cell line is deficient in strandspecific mismatch repair. Proc Natl Acad Sci USA 1993; 90: 6424–6428. 21 Levati L, Marra G, Lettieri T, D’Atri S, Vernole P, Tentori L, Lacal PM, Pagani E, Bonmassar E, Jiricny J, Graziani G. Mutation of the mismatch repair gene hMSH2 and hMSH6 in a human T-cell leukemic line tolerant to methylating agents. Genes Chromosome Cancer 1998; 3: 159–166. 22 Iwamoto KS, Mizuno T, Ito T, Tsuyama N, Kyozumi S, Seyama T. Gain of function p53 mutations enhance alteration of the T-cell receptor following X-irradiation, independently of the cell cycle and cell survival. Cancer Res 1996; 56: 3862–3865. 23 Zhang Y, Chen F-X, Metha P, Gold B. The design of groove and sequence selective alkylation of DNA by sulfonate esters tethered to lexitropsins. Biochemistry 1993; 32: 7954–7965. 24 Boulton S, Pemberton LC, Porteous JK, Curtin NJ, Griffin RJ, Golding BT, Durkacz BW. Potentiation of temozolomide induced cytotoxicity: a comparative study of the biological effects of poly(ADPribose) polymerase inhibitors. Br J Cancer 1995; 72: 849–856. 25 Vernole P, Tedeschi B, Caporossi B, Maccarone M, Melino G, Annichiarico-Petruzelli M. Induction of apoptosis by bleomycin in resting and cycling human lymphocytes. Mutagenesis 1998; 13: 209–215. 26 Vassal G, Boland I, Terrier-Lacombe MJ, Watson AJ, Margison GP, Ve´nuat AM, Morizet J, Parker F, Lacroix C, Lellouch-Tubiana A, Pierre-Kahn A, Poullain MG, Gouyette A. Activity of fotemustine in medulloblastoma and malignant glioma xenografts in relation to O6-alkylguanine-DNA alkyltransferase and alkylpurine-DNA Nglycosylase activity. Clin Cancer Res 1998; 4: 463–468. 27 Damia G, Imperatori L, Citti L, Mariani L, D’Incalci M. 3-methyla-


28

29 30

31

32

denine-DNA glycosylase and O6-alkylguanine-DNA alkyltransferase activities and sensitivity to alkylating agents in human cancer cell lines. Br J Cancer 1996; 73: 861–865. Vickers MA, Vyas P, Harris PC, Simmons DL, Higgs DR. Structure of the human 3-methyladenine DNA glycosylase gene and localization close to the 16p telomere. Proc Natl Acad Sci USA 1993; 90: 3437–3441. Mitra G, Pauly GT, Kumar R. Molecular analysis of O6-substituted guanine-induced mutagenesis of ras oncogene. Proc Natl Acad Sci USA 1989; 86: 8650–8654. Kaina B, Ziouta A, Ochs K, Coquerelle T. Chromosomal instability, reproductive cell death and apoptosis induced by O6methylguanine in Mex−, Mex+ and methylation-tolerant mismatch repair compromised cells: facts and models. Mutat Res 1997; 56: 227–241. Friedman HS, Johnson SP, Dong Q, Schold SC, Rasheed BKA, Bigner SH, Ali-Osman F, Dolan E, Colvin OM, Houghton P, Germain G, Drummond JT, Keir S, Marcelli S, Bigner DD, Modrich P. Methylator resistance mediated by mismatch repair deficiency in a glioblastoma multiforme xenograft. Cancer Res 1997; 57: 2933–2936. Tentori L, Graziani G, Gilberti S, Lacal PM, Bonmassar E, D’Atri S. Triazene compounds induce apoptosis in O6-alkylguanine-DNA alkyltransferase deficient leukemia cell lines. Leukemia 1995; 9: 1888–1895.

33 Chinnasamy N, Rafferty JA, Hickson I, Ashby J, Tinwell H, Margison GP, Dexter TM, Fairbairn, LJ. O6-benzylguanine potentiates the in vivo toxicity and clastogenicity of temozolomide and BCNU in mouse bone marrow. Blood 1997; 89: 1566–1573. 34 Vernole P, Caporossi B, Tedeschi B, Porfirio B, Melino G, Bonmassar E, Nicoletti B. Cytogenetic effects of 1-p-(3methyltriazeno)benzoic acid potassium salt on human lymphocytes in vitro. Mutat Res 1987; 189: 349–356. 35 Grombacher T, Kaina B. Constitutive expression and inducibility of O6-methylguanine-DNA methyltransferase and N-methylpurine-DNA glycosylase in rat liver cells exhibiting different status of differentiation. Biochim Biophys Acta 1995; 17: 63–72. 36 Yen L, Woo A, Christopoulopoulos G, Batist G, Panasci L, Roy R, Mitra S, Alaoui-Jamali MA. Enhanced host cell reactivation capacity and expression of DNA repair genes in human breast cancer cells resistant to bi-functional alkylating agents. Mutat Res 1995; 337: 179–189. 37 Dantzer F, Schreiber V, Niedergang C, Trucco C, Flatter E, De La Rubia G, Oliver J, Rolli V, Me´nissier-de Murcia J, de Murcia G. Involvement of poly(ADP-ribose) polymerase in base excision repair. Biochimie 1999; 81: 69–75. 38 Rice PA. Holding damaged DNA together. Nature Struct Biol 1999; 6: 805–806.

1459

Leukemia