Effect of O6-methylguanine-DNA methyltransferase gene activity on repair in human cells transformed by a regulatable ada gene.

To assess the biological role of DNA methylation at the O6 position of guanine (O6MeG) a human cell line was created that contains a regulatable gene of the O6MeG-DNA methyltransferase (MT), a repair activity that removes O6MeG adducts from the DNA. MT-deficient HeLa MR cells were transformed with an SV40-based expression vector in which the bacterial MT gene ada was put under the control of a glucocorticoid-inducible MMTV promoter. In response to dexamethasone (Dex), pSV MTV ada cells actively accumulated MT protein to reach a constant level after 10-12 h of approximately 15,000 MT molecules per cell. Co-induction by Dex and 12-O-tetradecanoylphorbol-13-acetate (TPA) further accelerated this synthesis approximately 2-fold and, as a result, higher final MT levels were achieved. The inducers were added to exponentially growing cells either before or at the time of N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) exposure and the kinetics of MT synthesis was studied. MNNG affected in a dose-dependent manner (i) the loss of the pre-existing MT activity; (ii) the lag before newly synthesized MT appeared; (iii) the final level of MT accumulated by the cells; and (iv) to a lesser extent the rate of MT synthesis. In cells with a down-regulated MT gene (no inducer) even small MNNG doses lead to an irreversible loss of the pre-existing MT activity, i.e. to incomplete repair, whereas an up-regulated MT gene supported the restoration of a pool of active MT molecules in the cells, i.e. an O6MeG repair that has gone to completion. Hence, effective O6MeG repair relies not only on the pre-existing MT level, but depends to an even greater extent on the state of expression of the MT gene. The activity of the MT gene also correlated with cell survival, which confirms our earlier finding that O6MeG adducts are cytotoxic for the cell.

Suppression of human DNA alkylation-repair defects by Escherichia coli DNA-repair genes.

The ada-alkB operon protects Escherichia coli against the effects of many alkylating agents. We have subcloned it into the pSV2 mammalian expression vector to yield pSV2ada-alkB, and this plasmid has been introduced into Mer- HeLa S3 cells, which are extremely sensitive to killing and induction of sister chromatid exchange by alkylating agents. One transformant (the S3-9 cell line) has several integrated copies of pSV2ada-alkB and was found to express a very high level of the ada gene product, the 39-kDa O6-methylguanine-DNA methyltransferase. S3-9 cells were found to have become resistant to killing and induction of sister chromatid exchange by two alkylating agents, N-methyl-N'-nitro-N-nitrosoguanidine and N,N'-bis(2-chloroethyl)-N-nitro-sourea. This shows that bacterial DNA alkylation-repair genes are able to suppress the alkylation-repair defects in human Mer- cells.

Adaptive resynthesis of O6-methylguanine-accepting protein can explain the differences between mammalian cells proficient and deficient in methyl excision repair.

Mammalian cells have been classified as proficient (Mer(+)) or deficient (Mer(-)) in methyl excision repair in terms of their cytotoxic reactions to agents that form O(6)-alkylguanine and their abilities to reactivate alkylated adenoviruses. O(6)-Methylguanine (O(6)MeGua) is considered to be a lethal, mutagenic, and carcinogenic lesion. We measured the abilities of cell extracts to transfer the methyl group from an exogenous DNA containing O(6)MeGua to acceptor protein. The constitutive level of acceptor activity was independent of the Mer phenotype and was approximately 100,000 acceptor sites per cell. Treatment of cells with N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) results in a dose-dependent decrease in the acceptor activity in extracts because the rapid reaction between endogenous O(6)MeGua and acceptor protein makes the latter unavailable for further reaction. Treatment of cells with 1 muM MNNG for 15 min or 2 muM for approximately 2 min uses up >95% of the constitutive activity. However, Mer(+) cells, which are resistant to MNNG, rapidly resynthesize new acceptor protein, and the activity returns to the basal level in approximately 90 min. In Mer(-) tumor cells and Chinese hamster cells, which are sensitive to MNNG, resynthesis is not detectable in 90 min. Mer(-) simian virus 40-transformed fibroblasts, known to have an intermediate sensitivity to MNNG, have an intermediate resynthesis rate. Treatment of cells with multiple low doses of MNNG results in the enhanced production of O(6)MeGua-accepting protein in levels 2.5-fold above the constitutive values for Mer(+) tumor cells and to approximately 1.5-fold for Mer(+) fibroblasts or Mer(-) simian virus 40-transformed cells. Such treatments reduce the activities in Mer(-) tumor cells and Chinese hamster cells. We conclude: (i) estimates of O(6)MeGua in cellular DNA shortly after treatment may be seriously in error because of the rapid repair of this lesion, and (ii) the adaptive resynthesis of acceptor protein, not its constitutive level, is the important correlate of cell resistance to methylating agents.

Adaptive increase of O6-methylguanine-acceptor protein in HeLa cells following N-methyl-N'-nitro-N-nitrosoguanidine treatment.

We have assayed in extracts of HeLa cells the amount of acceptor protein that removes O6-methylguanine adducts from alkylated DNA. Cells were treated with single or multiple nontoxic doses of N-methyl-N'-nitrosoguanidine (MNNG) and the extracts were analyzed up to 32 h after the last exposure. The acceptor activity assayed immediately (1 h) after single exposures decreases linearly with dose indicating that the acceptor protein is used up by endogenous O6-methylguanine adducts in a stoichiometric reaction. Multiple exposures, assayed 8-24 h after the last exposure, increase the amount of acceptor protein in a dose dependent fashion followed by a decrease above a cumulative dose of 100 ng/ml. Under conditions of maximum induction, there are about 300,000 acceptor protein sites per cell, approximately 3 fold above the constitutive level. Both in adapted and unadapted cells the methyl group from O6-methylguanine adducts in the alkylated DNA is transferred to cysteine residues of the acceptor protein(s).

Abilities of extracts of human lymphocytes to remove O6-methylguanine from DNA.

O6 MeGua is a presumptive mutagenic and carcinogenic product in DNAs treated with methylating agents. The abilities of lymphocyte extracts from 34 apparently normal individuals to remove O6 MeGua from exogenous DNA have been measured. The activity in extracts is stable to freezing and so permits repeat determinations and hence high precision in the assays. The data on removal are consistent with the idea that the removal is accomplished by the transfer of a methyl group to a methyl-accepting protein and that the protein acts in a stoichiometric fashion. Extracts from lymphocytes stimulated with PHA show on the average more activity than from unstimulated ones, although some extracts show no increase as a result of PHA stimulation of cells. There are large variations in the abilities of human lymphocytes to remove O6 MeGua, but the differences are not correlated significantly with sex or age. Unstimulated lymphocytes show a bimodal distribution of removing activity, whereas stimulated ones show a predominant single peak of activity. Extracts of T lymphocytes are more proficient than those of B lymphocytes and of any other white cells. On the average the number of presumptive acceptor molecules per cell in unstimulated lymphocytes is between 14 000 and 110 000 and in stimulated lymphocytes between 40 000 and 140 000.

Extracts of chronic lymphocytic leukemia lymphocytes have a high level of DNA repair activity fo O6-methylguanine.

Extracts of peripheral lymphocytes from six individuals with chronic lymphocytic leukemia (CLL) were assayed for the ability to remove O6-methylguanine (O6MeGua) from exogenous DNA. The O6MeGua-removing activity in CLL lymphocytes, predominantly B cells, was approximately 7-fold higher than in B lymphocytes of normal individuals and about 2-fold higher than in the unstimulated T type cells of normal persons. The activity measured in extracts of lymphocytes from three blood relatives was in the upper range of the normal distribution. Over 80% of the removal of O6MeGua was accomplished by the transfer of the methyl group to cysteine moieties of acceptor proteins in a stoichiometric reaction. If one assumes one acceptor group per acceptor protein, the calculated number of acceptor molecules per CLL lymphocyte falls between 91,000 and 220,000. Thus CLL lymphocytes do not show lower O6MeGua-removing activity, in contrast to many tumor cell strains or transformed cell lines, which are reported to have a deficient methyl excision repair phenotype (Mer-). Instead, the CLL lymphocytes act as if they have a super-Mer+ phenotype.

Mechanism of the ATP effect in the DNA repair synthesis of gamma-irradiated Escherichia coli cells.

Gamma-irradiation of Escherichia coli cells made permeable to deoxynucleoside triphosphates (dNTP) by toluene induces a repair-type DNA synthesis. As previous studies have shown ATP stimulates this DNA synthesis; we studied the mechanism of the ATP effect by analyzing the kinetics of nucleotide incorporation at various dNTP concentrations. The V values of the DNA repair synthesis rise with increasing dose (0-50 Gy); nonirradiated cells showed a negligible nucleotide incorporation. The apparent Michaelis constant KM for dNTP in the assay was 83-143 microM and the value was much higher than for a DNA polymerase reaction in vitro. ATP stimulated the DNA synthesis with concomitant decrease of KM yet unchanged V values. Similar results were obtained with a rec BC strain. We propose that the ATP effect is due to a greater affinity of dNTPs to the DNA polymerase, possibly by a stabilisation of the structural integrity of the complex DNA with repair enzymes. Activation of exonucleases by ATP could be excluded. Addition of NAD to the reaction mixture inhibits the DNA synthesis possibly by activation of ligase which closes the nicks in the DNA strand.

UV-endonuclease from calf thymus with specificity toward pyrimidine dimers in DNA.

We describe the partial purification of an endonuclease from calf thymus that nicks phage PM2 DNA irradiated with UV doses producing only a few pyrimidine dimers per molecule. It has much less activity on DNA that has been subjected to enzymatic photoreactivation after UV irradiation. The calf thymus endonuclease is different from other mammalian UV-endonucleases so far described in that it seems to be dimer specific. The enzyme is stimulated by Mg2+ and is inactive in the presence of EDTA. It binds to UV-irradiated DNA-Sepharose from which it is released by low concentrations of KCl. Gel filtration data indicate that the endonuclease may belong to a high molecular weight protein or protein complex. The enzyme is very labile and freezing increases its lability.

Role of exonucleases V and VIII in adenosine 5'-triphosphate- and deoxynucleotide triphosphate-dependent strand break repair in toluenized Escherichia coli cells treated with X-rays.

The repair of X-ray-induced strand breaks was studied in permeabilized Escherichia coli recBC cells deficient for the adenosine 5'-triphosphate (ATP)-dependent exonuclease V and in recBC sbcA cells that possess the ATP-independent exonuclease VIII. It is shown that repair induced by additon of ATP does not take place in recBC and recBC sbcB cells and is limited in recBC sbcA cells. ATP-dependent repair is nevertheless observable if together with ATP a mixture of deoxynucleotide monophosphates is supplied to the cells. These data fit with the assumption that in wild-type cells ATP-dependent repair involves exonuclease V-induced deoxyribonucleic acid degradation and rephosphorylation of the degradation products which are reused for deoxyribonucleic acid polymerase I-dependent break closure. Repair in the presence of deoxynucleotide triphosphates rejoins a similar fraction of breaks in all strains tested irrespective of the amount of postirradiation degradation resulting from exonuclease V and exonuclease VIII activities. Thus, exonuclease V is dispensable for deoxynucleotide triphosphate-dependent repair, i.e., does not "clean" the ends of breaks produced by X-irradiation. ATP- and deoxynucleotide triphosphate-dependent repair are not additive and seem to repair the same population of deoxyribonucleic acid molecules damaged by X-irradiation.

Repair of x-ray-induced single strand breaks in toluenized Escherichia coli cells.

We have used sedimentation in alkali to estimate the repair of X-ray-induced single strand breaks in the DNA of irradiated toluenized Escherichia coli cells. Extensive repair requires no exogenous cofactors except ATP although other individual NTPs (except U) or dNTPs can substitute for ATP. There is no repair in polA or resA cells and since nicotinamide mononucleotide (NMN) inhibits repair in wild type cells we interpret the results as indicating that both ligase and polymerase I are needed for repair but that the amount of any gap filling is small and extensive repair replication is not necessary.

Role of ATP in excision repair of ultraviolet radiation damage in Escherichia coli.

The effect of ATP on the first step of excision repair of ultraviolet damage in DNA has been studied using toluene-treated E. coli. During postirradiation incubation, five to six times more single-strand breaks are formed in DNA in the presence of exogenous ATP than in its absence. The ATP-dependent as well as the ATP-independent endonucleolytic activities appear to be catalyzed by the same enzyme since both activities are almost completely absent in uvrA and uvrB mutants. An ATP-dependent endonucleolytic activity has been detected in nonirradiated toluene-treated E. coli. It is concluded that ATP is required in vivo for either the incision step of repair or an enzymatic reaction preceding it.