Kinetics of repair of O6-methylguanine in DNA by O6-methylguanine-DNA methyltransferase in, vitro and in, vivo

The demethylation of O⁶-methylguanine in double stranded DNA catalyzed by rat liver O⁶-methylguanine-DNA transmethylase was found to proceed much more rapidly when the DNA substrate was methylated to a high extent. When the content of O⁶-methylguanine in DNA was equal to 1 in 2000 guanines, the reaction was 90% complete within 2 min, but when the content was 1 in 500,000 it required 27 min at 37°C. These results suggest that the repair protein either moves along the DNA substrate or else has little selectivity for binding specifically to the sites containing O⁶-methylguanine rather than to the normal DNA. The repair of O⁶-methylguanine in rat liver $\text{in}\text{,}\text{vivo}$ occurred at rates comparable to those seen $\text{in}\text{,}\text{vitro}$ with the substrates alkylated to low extents and was virtually complete within 3 hours. These results provide strong evidence that this protein is the factor responsible for O⁶-methylguanine removal $\text{in}\text{,}\text{vivo}$ and explain the wide variation in time courses reported in the literature since substrates methylated to greatly different extents have been used for such experiments.     

Enzymatic removal of O6-methylguanine from DNA by mammalian cell extracts

Extracts from various rat tissues were incubated with [³H]methylated DNA or chromatin in order to compare their abilities to catalyze the removal of labeled $\text{O}{}^6$-methylguanine from acid precipitable DNA. Liver extracts had the greatest activity. Kidney extracts had about 35% of the activity in liver and extracts from lung, colon, small intestine and brain were much less active. The enzyme responsible for this reaction does not appear to be an $\text{N}$-glycosidase because no labeled $\text{O}{}^6$-methylguanine could be detected in the supernatant fraction even though more than 50% of this base was lost from the DNA. The released radioactivity was present as methanol which is consistent with the possibility that the reaction may involve a demethylase action on either the DNA substrate or an oligonucleotide derived from it.     

In vivo replication of carcinogen-modified rat liver DNA: increased susceptibility of O6-methylguanine compared to N-7-methylguanine in replicated DNA to S1-nuclease

In order to characterize rat liver DNA replicated $\text{in}\text{vivo}$ on a carcinogen-damaged template, the replicated DNA was treated with S₁-nuclease and the release of (¹⁴C)-dimethyl-nitrosamine induced 0⁶-methylguanine, a lesion associated with miscoding and N-7-methylguanine, a lesion that does not miscode were monitored. The results indicated that both the methylated guanines became susceptible to S₁-nuclease upon replication. However, a greater percentage of 0⁶-methylguanine (22% of the total 0⁶-methylguanine present in the DNA) compared to N-7-methylguanine (4% of the total N-7-methylguanine present in the DNA) was rendered acid soluble by S₁-nuclease. The preferential release of 0⁶-methylguanine compared to N-7-methylguanine from replicated DNA was interpreted to indicate its occurrence in local denatured regions probably generated as a result of misbase pairing.     

Methylation of DNA in rat liver and intestine by dimethylnitrosamine and N-methylnitrosourea

A chromatographic procedure for improved separation of deoxyribonucleosides and methylated deoxyribonucleosides is described. DNA was isolated from liver and small intestine of rats treated with [¹⁴C]dimethylnitrosamine ([¹⁴C]DMN) or $\text{N-[}{}^3\text{H}\text{]}\text{methyl}\text{-N-}\text{nitrosourea}$ ([³H]MNU), and the purified DNA was hydrolyzed enzymatically. The deoxyribonucleosides were chromatographed on an Aminex A-6 cation exchange column at 37°C with 0.4 M ammonium formate, pH 4.5, as eluant. In addition to showing the presence of the expected alkylated products, N⁷-methyldeoxyguanosine (determined as N⁷-methylguanine) and O⁶-methyldeoxyguanosine, several other minor methylated products were found in liver and intestinal DNA of rats treated with DMN or MNU. Two of these products are believed to be N³-methylthymidine and O⁴-methylthymidine.     

The accumulation of O6-methylguanine in the liver and kidney DNA of rats treated with dimethylnitrosamine for a short or a long period

A large dose of dimethylnitrosamine was administered to rats by two different dosing regimens, one being eleven intraperitoneal injections of 5 mg/kg body wt. at 12-h intervals (a dosing regimen strongly carcinogenic for the kidney but not the liver), and the other being a continuous dosing over several weeks by adding 8.5 mg of dimethylnitrosamine to each litre of drinking water giving an approximate daily dose of 0.7 mg/kg body wt. This treatment is known to be strongly carcinogenic for the liver but not the kidney. The accumulation in DNA of liver and kidney of the methylated purines 7-methylguanine and O⁶-methylguanine under each regimen were measured and compared. With the eleven-injection regimen there was a build up of O⁶-methylguanine in the DNA of the susceptible organ, the kidney, whilst in the liver virtually no accumulation was detected. Under the prolonged, low concentration regimen the liver, in spite of its susceptibility to the carcinogen did not accumulate O⁶-methylguanine. The results are discussed in terms of the hypothesis that production of O⁶-methylguanine and its persistence in the DNA of the target organ are responsible for the carcinogenic action of dimethylnitrosamine.     

Excision of O6-methylguanine from DNA of various mouse tissues following a single injection of N-methyl-N-nitrosourea

The persistence of O⁶-methylguanine produced by a single dose of N-methyl-N-nitrosourea (MNU) was determined in DNA of various murine tissues and compared with the location of tumours induced by MNU and related alkylating carcinogens in this species. A/J and C3HeB/FeJ mice received a single intravenous injection of MNU (10 mg/kg) and were killed at different time intervals ranging from 4 h to 7 days.The rate of loss of O⁶-methylguanine from brain DNA was considerably slower than from liver DNA; tumours have been found in both organs after administration of MNU and other alkylnitrosoureas. There was no difference in the rate of excision from cerebral DNA of A/J and C3HeB/FeJ mice, although these strains differ significantly in their susceptibility to the neurooncogenic effect of MNU and related carcinogens. Excision of O⁶-methylguanine from hepatic DNA was significantly slower in A/J than in C3HeB/FeJ mice; both strains have been found to develop hepatic carcinomas following MNU administration. Seven days after the injection of ³H-MNU, O⁶-methylguanine concentrations were highest in brain and lung DNA, lowest in the liver, and intermediate in kidney, spleen, small intestine and stomach. The lung is a principal target organ for tumour induction by MNU and other carcinogens in mice; however, neural tumours are usually induced at a low incidence.The results obtained do not contradict the hypothesis that O⁶-alkylation of guanine in DNA is a critical event in the initiation of tumour induction by alkylating agents. However, the location of tumours produced in mice does not seem to depend solely on the formation and persistence of O⁶-alkylguanine in DNA.     

Demethylation of O6-methylguanine in a synthetic DNA polymer by an inducible activity in Escherichiacoli

$\text{O}$⁶-Methyl[8-³H]deoxyguanosine in a synthetic DNA polymer, poly(dC, dG, m⁶dG), is demethylated by cell-free extracts of $\text{Escherichia}\text{coli}\text{B}\text{r}$ adapted by exposure to $\text{N}$-methyl-$\text{N}$′-nitro-$\text{N}$-nitrosoguanidine, as shown by the appearance of ³H-labeled deoxyguanosine in hydrolysates of the recovered DNA. The demethylating activity could not be detected in extracts of nonadapted $\text{E}\text{.}\text{coli}$. These results provide direct evidence that a previously described inducible repair activity in $\text{E}\text{.}\text{coli}$ acts by demethylating $\text{O}$⁶-methylguanine at the DNA level.     

Improved automated solid-phase microsequencing of peptides using DABITC

The methylated purines O⁶-methyl- and 7-methylguanine were isolated from mouse liver DNA hydrolysates by means of a column cleanup employing a Sep Pak C-18 reverse-phase cartridge. The purine bases were eluted from the cartridge with methanol, evaporated to dryness, and then dissolved in mobile phase for liquid chromatographic analysis by normalphase chromatography. The system consisted of a LiChrosorb Si 60 column with a watersaturated mobile phase of 20% methanol in chloroform containing 0.001% H₃PO₄. The two methylated bases eluted before adenine or guanine. For extremely low-level (<300 pmol) quantitation, the peaks corresponding to O⁶-methyl- and 7-methylguanine were collected and then analyzed by reverse-phase chromotography with a LiChrosorb RP-18 column and a mobile phase of 5% methanol in pH 7 phosphate buffer (for 7-methylguanine) or 9.5% methanol/buffer (for O⁶-methylguanine). Comparisons were made with fluorescence detection and with scintillation counting (in animal studies where [¹⁴C]dimethylnitrosamine was used). Minimum detectable levels at 254 nm were about 3 ng (3:1 signal to noise ratio) for each of the title compounds. As low as 10 pmol/mg of each could be detected in DNA hydrolysates. Recoveries of O⁶-methyl- and 7-methylguanine from DNA spiked at 750 pmol/mg were greater than 80%.     

MutS inhibits RecA-mediated strand transfer with methylated DNA substrates

DNA mismatch repair (MMR) sensitizes human and Escherichia coli dam cells to the cytotoxic action of N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) while abrogation of such repair results in drug resistance. In DNA methylated by MNNG, MMR action is the result of MutS recognition of O⁶-methylguanine base pairs. MutS and Ada methyltransferase compete for the MNNG-induced O⁶-methylguanine residues, and MMR-induced cytotoxicity is abrogated when Ada is present at higher concentrations than normal. To test the hypothesis that MMR sensitization is due to decreased recombinational repair, we used a RecA-mediated strand exchange assay between homologous phiX174 substrate molecules, one of which was methylated with MNNG. MutS inhibited strand transfer on such substrates in a concentration-dependent manner and its inhibitory effect was enhanced by MutL. There was no effect of these proteins on RecA activity with unmethylated substrates. We quantified the number of O⁶-methylguanine residues in methylated DNA by HPLC-MS/MS and 5–10 of these residues in phiX174 DNA (5386 bp) were sufficient to block the RecA reaction in the presence of MutS and MutL. These results are consistent with a model in which methylated DNA is perceived by the cell as homeologous and prevented from recombining with homologous DNA by the MMR system.     

Differential methylation of mitochondrial and nuclear DNA in cultured mouse, hamster and virus-transformed hamster cells. In vivo and in vitro methylation

Information has been lacking as to whether mitochondrial DNA of animal cells is methylated. The methylation patterns of mitochondrial and nuclear DNAs of several mammalian cell lines have therefore been compared by four methods: (1) in vivo transfer of the methyl group from [methyl-³H]methionine; (2) in vivo incorporation of [³²P]orthophosphate and a combination of (1) and (2); (3) in vivo incorporation of [³H]deoxycytidine; (4) in vitro methylation of DNAs with ³H-labeled S-adenosylmethionine as methyl donor and DNA methylase preparations from L cell nuclei. The cell lines were mouse L cells, $\text{BHK21}\text{C13}$, $\text{C13}\text{B4}$ (baby hamster kidney cells transformed by the Bryan strain of Rouse sarcoma virus), and PyY (BHK cells transformed by polyoma virus). DNA bases were separated chromatographically, using 5-methylcytosine, 6-methylaminopurine and, in some cases, 7-methylguanine as markers.Mitochondrial DNA was found to be significantly less methylated than nuclear DNA with respect to 5-methylcytosine in all cell types studied and by all methods used. The relative advantages and disadvantages of each method have been discussed. The level of 5-methylcytosine in mitochondrial DNA as compared with that in nuclear DNA was estimated as one-fourth to one-fourteenth in various cell lines. The estimated 5-methylcytosine content per circular mitochondrial DNA molecule (mol. wt 10 × 10⁶) was about 12 methylcytosine residues for L cells and 24, 30 and 36 methylcytosine residues for BHK, B4 and PyY cells, respectively. Relative to cytosine residues, the estimate was one 5-methylcytosine per 500 cytosine residues of mitochondrial DNA and one 5-methylcytosine per 36 cytosine residues of nuclear DNA from L-cells. The values for methylcytosine of mitochondrial DNA are presumed to be maximal. PyY cells as compared with other cells had the highest methylcytosine content of both mitochondrial and nuclear DNA as estimated by method (3). No methylation of nuclear DNA was observed in confluent L cells.Evidence for the presence of DNA methylase activity associated with mitochondrial fractions was obtained. This activity could be distinguished from other cellular DNA methylase activity by differential response to mercaptoethanol. Radioactivity from ³H-labeled S-adenosylmethionine was found only in 5-methyl-cytosine of DNA.     

Methylation of transfer RNA during transformation of human lymphocytes by phytohemagglutinin

Methylation of transfer RNA during phytohemagglutinin induced transformation of human lymphocytes was studied by labeling the tRNA $\text{in}\text{vivo}$ with (methyl-H³)-methionine and measuring the distribution of tritium in the methylated nucleotides. An alteration in the pattern of methylation occurred within hours after PHA-stimulation and this change was maintained through several cell generations. There was a 50 to 94% increase in the relative amount of methylated N²-methyl-guanine (1 to 3 hr) and a 40 to 59% decrease in the relative amount of 1-methyladenine (1 to 12 hr). Treatment of the stimulated cells with Actinomycin D prevented the subsequent methylation not tRNA. However, inhibition of protein synthesis by puromycin did not effect the early changes observed in the methylation of tRNA.     

DNA synthesis and the cell cycle in cultures of normal and mutant (mdg/mdg) embryonic mouse muscle cells.

The relationship between cell fusion, DNA synthesis and the cell cycle in cultured embryonic normal and dysgenic ($\text{mdg}\text{mdg}$) mouse muscle cells has been determined by autoradiography. The experimental evidence shows that the homozygous mutant myotubes form by a process of cell fusion and that nuclei within the myotubes do not synthesize DNA or undergo mitotic or amitotic division. The duration of the total cell cycle and its component phases was statistically the same in 2-day normal and mutant ($\text{mdg}\text{mdg}$) myogenic cultures with the approximate values: T, 21.5 hr; G₁, 10.5 hr; S, 7.5 hr; and G₂, 2.5 hr. In both kinds of cultures, labeled nuclei appeared in myotubes 15–16 hr after mononucleated cells were exposed to [³H]thymidine, and the rate of incorporation of labeled nuclei into multinucleated muscle cells was comparable in control and dysgenic cultures. Thus, homozygous $\text{mdg}\text{mdg}$ muscle cells in culture are similar to control cells with respect to their mechanism of myotube formation and the coordinate regulation of DNA synthesis and the cell cycle during myogenesis.     

Magnetic resonance study of 13C reductively methylated glycophorin and glycophorin glycopeptides

¹³C nuclear magnetic resonance (n.m.r.) spectral data for ¹³C reductively methylated N-terminal tryptic glycopeptides and for ¹³C reductively methylated N-terminal glyco-octapeptides derived from homozygous glycophorins A^M and A^N are presented. Their ¹³C chemical shift data are compared with the previously published ¹³C n.m.r. data for ¹³C reductively methylated homozygous glycophorins A^M and A^N in order to investigate the means of display of the MN blood determinants by these species. The pH dependence of the ¹³C resonances of $\text{N}{}^{\mathrm{\alpha }}\text{,N-[}{}^{13}\text{C}\text{]}\text{dimethyl}$ leucine of glyco-octapeptide A^N and of $\text{N}{}^{\mathrm{\alpha }}\text{,N-[}{}^{13}\text{C}\text{]}\text{dimethyl}$ serine of glyco-octapepti A^M indicated that only a slight structural perturbation occurs at the N-terminus when a large portion of the glycoprotein molecule is removed. However, one structural ‘state’ of ¹³C reductively methylated glycophorin A^M is lost when the glyco-octapeptide A^M is produced. The ¹³C resonance of $\text{N}{}^{\mathrm{\alpha }}\text{,N-[}{}^{13}\text{C}\text{]}\text{dimethyl}$ leucine of glycooctapeptide A^N titrated with a $\text{p}\text{K}{}_{\mathrm{<mtext>a</mtext>}}$ of 7.7 (Hill coefficient $\text{～ 1}$). The ¹³C resonance of $\text{N}{}^{\mathrm{\alpha }}\text{,N-[}{}^{13}\text{C}\text{]}\text{dimethyl}$ serine, on the other hand, exhibited an unusual pH dependence, indicating the existence of some possible steric constraints or hydrogen bonding in this molecule. In comparison to the data obtained for ¹³C-labelled glycooctapeptide A^M molecule, the pH dependence of the chemical shift of the ¹³C resonance of $\text{N}{}^{\mathrm{\alpha }}\text{,N-[}{}^{13}\text{C}\text{]}\text{dimethyl}$ serine of tripeptide tri-L-serine is also presented. Circular dichroism (c.d.) spectra indicated that the reductive methylation technique does not cause a large perturbation of the glycophorin A molecule.     

The nature of single-strand breaks in DNA following treatment of L-cells with methylating agents

DNA from untreated L-cells had a weight average molecular weight (Mw) of 5.7 ± 0.58·10⁸ daltons as measured by sedimentation in an alkaline sucrose gradient. This value was reduced by one half after the cells were treated for 1 h with 8 μg/ml of N-methyl-N-nitrosourea (MNUA), 34 μg/ml of methyl methanesulfonate (MMS) or 0.16 μg/ml of N-methyl-N′-nitro-N-nitrosoguanidine (MNNG). That dose of MNUA produced 52 methylations per 5.7·10⁸ daltons DNA. 20% of these were not purine derivatives and were assumed to contain some phosphotriesters. That dose of MMS (above) produced 290 methylations per 5.7·10⁸ daltons DNA and about 14% of these were not purine derivatives. The rates of loss of methylated purines from DNA were 2.3% per hour for 7-methylguanine (7-MeG), 7.4% per hour for 3-methyladenine (3-MeA) and no detectable loss of O⁶-methylguanine (O⁶-MeG) over a 12 h period. Since phosphotriesters are alkali-labile the single-strand breaks probably arose from this structure and did not form within the cell. This conclusion is supported by the following considerations. MNUA was more effective than MMS at reducing the molecular weight of DNA, as measured in alkaline medium. The greater SN₁ character of MNUA would cause a greater formation of phosphotriesters than would MMS.     

Incorporation of 3H-adenine into free cytokinins by cytokinin-autonomous tobacco callus tissue

Cytokinin-autonomous tobacco callus was incubated in defined mineral medium containing ³H-adenine for 60 minutes. Radioactivity was incorporated into the four predominant free cytokinins, ribosyl-$\text{trans}$-zeatin, $\text{trans}$-zeatin, N⁶-(Δ²-isopentenyl) adenosine and N⁶-(Δ²-isopentenyl) adenine. The bases were more abundant than their respective ribosides, N⁶-(Δ²-isopentenyl) adenine being the most abundant cytokinin. No discrete peaks of radioactivity could be detected on the HPLC column eluate corresponding to the elution volumes of $\text{cis}$-zeatin and ribosyl-$\text{cis}$-zeatin.     

Conversion of the Ca2+-ATPase from Rhodospirillum rubrum into a Mg2+-dependent enzyme by 1,N6-etheno ATP

Nucleoside triphosphate hydrolysis of $\text{R.}\text{rubrum}$ ATPase complexes can be changed from Ca²⁺-dependence to Mg²⁺-dependence by replacing ATP with 1,N⁶-etheno ATP. Four ATPase complexes which have been prepared by different procedures hydrolyze ATP and 1,N⁶-etheno ATP at different rates in dependence on the added metal ions. These differences allow an easy distinction of the various enzyme forms.     

Formation and subsequent removal of O6-methylguanine from deoxyribonucleic acid in rat liver and kidney after small doses of dimethylnitrosamine

1. The amounts of 7-methylguanine and O⁶-methylguanine present in the DNA of liver and kidney of rats 4h and 24h after administration of low doses of dimethylnitrosamine were measured. 2. O⁶-Methylguanine was rapidly removed from liver DNA so that less than 15% of the expected amount (on the basis of 7-methylguanine found) was present within 4h after doses of 0.25mg/kg body wt. or less. Within 24h of administration of dimethylnitrosamine at doses of 1mg/kg or below, more than 85% of the expected amount of O⁶-methylguanine was removed. Removal was most efficient (defined in terms of the percentage of the O⁶-methylguanine formed that was subsequently lost within 24h) after doses of 0.25–0.5mg/kg body wt. At doses greater or less than this the removal was less efficient, even though the absolute amount of O⁶-methylguanine lost during 24h increased with the dose of dimethylnitrosamine over the entire range of doses from 0.001 to 20mg/kg body wt. 3. Alkylation of kidney DNA after intraperitoneal injections of 1–50μg of dimethylnitrosamine/kg body wt. occurred at about one-tenth the extent of alkylation of liver DNA. Removal of O⁶-methylguanine from the DNA also took place in the kidney, but was slower than in the liver. 4. After oral administration of these doses of dimethylnitrosamine, the alkylation of kidney DNA was much less than after intraperitoneal administration and represented only 1–2% of that found in the liver. 5. Alkylation of liver and kidney DNA was readily detectable when measured 24h after the final injection in rats that received daily injections of 1μg of [³H]dimethylnitrosamine/kg for 2 or 3 weeks. After 3 weeks, O⁶-methylguanine contents in the liver DNA were about 1% of the 7-methylguanine contents. The amount of 7-methylguanine in the liver DNA was 10 times that in the kidney DNA, but liver O⁶-methylguanine contents were only twice those in the kidney. 6. Extracts able to catalyse the removal of O⁶-methylguanine from alkylated DNA in vitro were isolated from liver and kidney. These extracts did not lead to the loss of 7-methylguanine from DNA. 7. The possible relevance of the formation and removal of O⁶-methylguanine in DNA to the risk of tumour induction by exposure to low concentrations of dimethylnitrosamine is discussed.     

In vivo radiation-induced thymine residue release from E. coli DNA

Cells of $\text{E. coli}$ C thy^?321 are examined for thymine residue release from DNA following gamma-irradiation from 5 to 15 krad. Experimental conditions are designed to inhibit enzyme activity that might promote base residue release. Enzyme action is restricted in order to assess the physicochemical action of radiation on cellular DNA, and to this end irradiation is done under O₂, N₂, and N₂O saturating conditions. Both thymine and thymidine release from bacterial DNA are detected and quantitated, and three oxygen effects are noted in comparing yields of these products. No difference in effect is observed between N₂ and N₂O gassing conditions, suggesting that the hydroxyl radical has little effect on thymine or thymidine release from irradiated DNA $\text{in vivo}$.     

Evidence for DNA adducts in rat liver after adminstration of N-nitrosopyrrolidine

Liver DNA isolated from rats given N-nitrosopyrrolidine contained amounts of an unidentified fluorescent adduct which were dose dependent. The adduct formed gradually over the first 12 hr after administration and was slowly removed from the DNA $\text{in vivo}$. Fluorescence and chromatographic properties of the adduct suggested the compound was a substituted guanine; also, the putative adduct was readily removed from DNA by neutral thermal hydrolysis, as are 3-alkyladenines and 7-alkylguanines. Evidence was also obtained for the formation of 7-methylguanine in liver DNA of N-nitrosopyrrolidine-treated rats; however, N-nitrosopyrrolidine was not the methyl source for this alkylated guanine.