To study molecular aspects of cytotoxicity of the anticancer drug β-D-glucose-ifosfamide mustard we investigated the potential of the agent to induce apoptosis and DNA breakage. Since β-D-glucose-ifosfamide mustard generates DNA interstrand crosslinks, we used as an in vitro model system a pair of isogenic Chinese hamster V79 cells differing in their sensitivity to crosslinking agents. CL-V5B cells are dramatically more sensitive (30-fold based on D10 values) to the cytotoxic effects of β-D-glucose-ifosfamide mustard as compared to parental V79B cells. After 48 h of pulse-treatment with the agent, sensitive cells but not the resistant parental line undergo apoptosis and necrosis, with apoptosis being the predominant form of cell death (70 and 20% of apoptosis and necrosis, respectively). Apoptosis increased as a function of dose and was accompanied by induction of DNA double-strand breaks in the hypersensitive cells. Furthermore, a strong decline in the level of Bcl-2 protein and activation of caspases-3, -8 and -9 were observed. The resistant parental cells were refractory to all these parameters. Bcl-2 decline in the sensitive cells preceded apoptosis, and transfection-mediated overexpression of Bcl-2 protected at least in part from apoptosis. From the data we hypothesize that non-repaired crosslinks induced by β-D-glucose-ifosfamide mustard are transformed into double-strand breaks which trigger apoptosis via a Bcl-2 dependent pathway.

Cells deficient in DNA polymerase beta (beta-pol) are impaired in base excision repair (BER) and hypersensitive to various DNA damaging agents, including methylating mutagens. Hypersensitivity of beta-pol-deficient cells to methylating agents is because of induction of apoptosis (Ochs et al., Cancer Res., 59: 1544-1551, 1999), indicating incompletely repaired DNA damage to trigger the response. Here we show that defective BER in beta-pol-null cells results in an early and transient increase in the frequency of DNA single-strand breaks on treatment with methyl methanesulfonate. These breaks arising as repair intermediates are not likely to trigger apoptosis directly because they were repaired efficiently and generated both in resting and proliferating cells, whereas only proliferating cells underwent with high frequency apoptosis after methylation. Therefore, we propose that single-strand breaks are converted into another kind of critical apoptosis-triggering lesion during replication. These critical secondary DNA lesions are likely to be non-repaired DNA double-strand breaks (DSBs), which are formed at higher frequency in beta-pol-null than in wild-type cells. Apoptosis was a late response not detectable before 24 h after methylation and was preceded by DSBs formation, extensive chromosomal breakage, and decline in Bcl-2 level and caspase-9 and caspase-3 activation. Caspase-8 was not significantly activated. Transfection of beta-pol-null cells with bcl-2 protected against methylation-induced apoptosis, indicating Bcl-2 to be causally involved. Overall, the data demonstrate that in cells lacking beta-pol, defective BER results in incompletely repaired DNA damage, which triggers apoptosis in a replication-dependent way by activating the mitochondrial death pathway. It is suggested that DSBs act as a critical ultimate apoptosis-inducing lesion.

DNA double-strand breaks (DSBs) are induced by ionizing radiation (IR) and various radiomimetic agents directly, or indirectly as a consequence of DNA repair, recombination and replication of damaged DNA. They are ultimately involved in the generation of chromosomal aberrations and were reported to cause genomic instability, gene amplification and reproductive cell death. To address the question of whether DSBs act as a trigger of apoptosis, we induced DSBs by means of restriction enzyme electroporation and compared the effect with IR in mouse fibroblasts that differ in p53 status [wild-type (+/+) versus p53-deficient (-/-) cells]. We show that (i) electroporation of PVU:II is highly efficient in the induction of DSBs, (ii) electroporation of PVU:II increases the rate of apoptosis, but not of necrosis in p53-/- cells, (iii) treatment with gamma-rays induces both apoptosis and necrosis in p53-/- cells, (iv) the frequency of DSBs correlates with the yield of apoptosis and (v) both PVU:II and gamma-ray treatment reduce the level of anti-apoptotic Bcl-2 protein in p53-/- cells whereas the level of Bax remains unaltered. Cells expressing wild-type p53 were more resistant than p53-deficient cells as to the induction of apoptosis and did not show Bcl-2 decline upon treatment with PVU:II and gamma-rays. The data provide evidence that blunt-ended DSBs induced by restriction enzyme PVU:II act as a highly efficient trigger of apoptosis, but not of necrosis. This process is related to Bcl-2 decline and does not require p53.

Methylating carcinogens and cytostatic drugs induce different methylation products in DNA. In cells not expressing the repair protein MGMT or expressing it at a low level, O6-methylguanine is the major genotoxic, recombinogenic, and apoptotic lesion. Genotoxicity and apoptosis triggered by O6-methylguanine require mismatch repair (MMR). In cells expressing O6-methylguanine-DNA methyl transferase (MGMT) at a high level or for agents producing low amounts of O6-methylguanine, N-alkylations become the major genotoxic lesions. N-Alkylations are repaired by base excision repair (BER). In mammalian cells, naturally occurring mutants of BER have not been detected, which points to the importance of BER for viability. In order to ascertain the role of BER in cellular defense, BER was modulated either by transfection or mutational inactivation. It has been shown that overexpression of N-methylpurine-DNA glycosylase (MPG) does not protect, but rather sensitizes cells to SN2 agents. This has been interpreted in terms of an imbalance in BER. Regarding abasic site endonuclease (APE), transient but not stable overexpression of the enzyme was achieved upon transfection in CHO cells, which indicates that unphysiologic APE levels are not tolerated by the cell. Besides the repair function, APE (alias Ref-1) exerts redox capability by which the activity of various transcription factors is modulated. Therefore, it is possible that stable overexpression of mammalian APE impairs transcriptional regulation of genes, whereas transient overexpression may exert some protective effect. DNA polymerase beta (Pol beta) transfection was ineffective in conferring resistance to methylmethane sulfonate (MMS). On the other hand, Pol beta-deficient cells proved to be highly sensitive to methylation-induced chromosomal aberrations and reproductive cell death. The dramatic hypersensitivity in the killing response is largely due to induction of apoptosis. Obviously, nonrepaired BER intermediates are clastogenic and act as a strong trigger of the apoptotic pathway. The elements of this pathway are currently under investigation.

Methylation at the O(6)-position of guanine (O(6)-MeG) by alkylating agents is efficiently removed by O(6)-methylguanine-DNA methyltransferase (MGMT), preventing from cytotoxic, mutagenic, clastogenic and carcinogenic effects of O(6)-MeG-inducing agents. If O(6)-MeG is not removed from DNA prior to replication, thymine will be incorporated instead of cytosine opposite the O(6)-MeG lesion. This mismatch is recognized and processed by mismatch repair (MMR) proteins which are known to be involved in triggering the cytotoxic and genotoxic response of cells upon methylation. In this work we addressed three open questions. (1) Is MGMT able to repair O(6)-MeG mispaired with thymine (O(6)-MeG/T)? (2) Do MMR proteins interfere with the repair of O(6)-MeG/T by MGMT? (3) Does MGMT show a protective effect if it is expressed after replication of DNA containing O(6)-MeG? Using an in vitro assay we show that oligonucleotides containing O(6)-MeG/T mismatches are as efficient as oligonucleotides containing O(6)-MeG/C in competing for MGMT repair activity, indicating that O(6)-MeG mispaired with thymine is still subject to repair by MGMT. The addition of MMR proteins from nuclear extracts, or of recombinant MutSalpha, to the in vitro repair assay did not affect the repair of O(6)-MeG/T lesions by MGMT. This indicates that the presence of MutSalpha still allows access of MGMT to O(6)-MeG/T lesions. To elucidate the protective effect of MGMT in the first and second replication cycle after N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) treatment, MGMT transfected CHO cells were synchronized and MGMT was inactivated by pulse-treatment with O(6)-benzylguanine (O(6)-BG). Thereafter, the recovered cells were treated with MNNG and subjected to clonogenic survival assays. Cells which expressed MGMT in the first and second cell cycle were more resistant than cells which expressed MGMT only in the second (post-treatment) cell cycle. Cells which did not express MGMT in both cell cycles were most sensitive. This indicates that repair of O(6)-MeG can occur both in the first and second cell cycle after alkylation protecting cells from the killing effect of the lesion.