Relative Efficiencies of the Bacterial, Yeast, and Human DNA

0 downloads 0 Views 642KB Size Report
Feb 15, 2016 - coli, human, and. Saccharomyces cerevisiae. DNA. MTases repair .... plasmid pDS427 was a gift from Diane Shevell and Graham Walker.
Vol. 266, No. 5 , Issue of February 15, pp. 2767-2771, 1991 Printed I n U.S.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY (CI1991 by

The American Society for Biochemistry and Molecular Biology, Inc.

Relative Efficienciesof the Bacterial, Yeast, and Human DNA Methyltransferases forthe Repair of 06-Methylguanineand 04-Methylthymine SUGGESTIVEEVIDENCEFOR04-METHYLTHYMINEREPAIR METHYLTRANSFERASES*

BY EUKARYOTIC

(Received for publication, July 23, 1990)

Mandana Sassanfar$$,Manjit K. DosanjhllII, John M. Essigmannv, and Leona Samson$** From the $Laboratory of Toxicology, Haruard School of Public Health, Boston, Massachusetts 02115 and the TDepartment of Chemistry and WhitakerCollege - of . Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139

The suicidal inactivation mechanism of DNA repair methyltransferases (MTases)was exploited to measure the relative efficiencies with which the Escherichia coli, human, and Saccharomycescerevisiae DNA MTases repair 0'-methylguanine (O'MeG) and 04methylthymine (04MeT), twoof the DNA lesions produced by mutagenic and carcinogenic alkylating agents. Using chemically synthesized double-stranded 25-base pair oligodeoxynucleotides containing a single O'MeG or a single 04MeT,the concentration of O'MeG or 04MeT substrate that produced 50% inactivation (ICGO) was determined for each of four MTases.The E. coli ogt gene product had a relatively high affinity for the O'MeG substrate (ICso 8.1 nM) buthadan even higher affinity for the 04MeTsubstrate (ICs0 3 nM). By contrast, the E. coli Ada MTase displayed a striking preference for O'MeG (ICs0 1.25 nM) as compared to 04MeT (ICso 27.5 nM). Both the human and the yeast DNA MTases were efficiently inactivated upon incubation with the O'MeG-containing oligomer (ICs0 values of 1.5 and 1.3 nM, respectively). Surprisingly, the human and yeast MTases were also inactivated by the 04MeT-containing oligomer albeit at ICs0 values of 29.5 and 44 nM, respectively. This result suggests that 04MeT lesions can be recognized in this substrate by eukaryotic DNAMTasesbut the exact biochemical mechanism of methyltransferase inactivation remains to be determined.

either of the two known DNA glycosylases encoded by the alk4 and tag genes (1-3). In contrast, lesions such as 06methylguanine (O'MeG)' and 04-methylthymine (04MeT) do not measurably inhibit DNA replication but frequently cause mutations owing to base mispairing (4-6). The observed in uiuo mutagenecity of O'MeG and 04MeT depends upon the frequency with which they are formed, the frequency of mispairing they induce, and the efficiency with which they are repaired. O'MeG is believed to be the most abundant alkylation-induced mutagenic lesion (1).It induces the GC to AT transition (7, 8 ) , which also is the principal genetic change observed in the mutational spectrum of most DNA alkylating agents. 04MeT is formed at much lower levels than O'MeG and induces AT toGC transitions in E. coli (9).In uitro assays using both prokaryotic and eukaryotic DNA polymerases suggest that O'MeG and 04MeT arehighly, and approximately equally mutagenic (10). There is some suggestion, however, that the frequency of mispairing induced by these lesions is modulated by the surrounding nucleotide sequence i l l ) , particularly the 3' neighboring base. In E. coli O'MeG and 04MeT arerepaired by DNA methyltransferases (MTases) (1) and, in the case of the former lesion,nucleotide excision repair enzymeshavealso been implicated (12). DNA MTases are suicide enzymes (13) that irreversibly transfer methyl groups from O'MeG lesions onto one of their own cysteine residues (1, 14). The irreversible binding of methyl groups has been successfully exploited to radiolabel MTases with tritium thus facilitating the identification and characterization of the MTases of different organisms. E. coli has two MTases. TheAda DNA MTase is a 39Alkylating agents introduce a variety of lethal and mutakDa protein encoded by the inducible ada gene (15, 16) and genic lesions in DNA (1).In Escherichia coli, lesions such as possesses two alkyl-acceptor cysteines (17, 18), one for the N'-methylpurinesand@-methylpyrimidinesinhibit DNA repair of 0"MeG or 04MeT lesions (19, 20) and a second for replication and are thus lethal if they are not removed by the repair of methylphosphotriester lesions (21, 22). Upon binding of methyl groups frommethylphosphotriester lesions, * This work was supported by American Cancer Society Research Grant NP448 (to L. S.), National Cancer Institute Grant CA52127 the Ada MTase becomes a transcriptional activator of genes (to J. M. E.), National Institute of Environmental Health Science whose products are involved in the repair of alkylated DNA Grant ES03926 (to L. S. and J. M. E.). The costs of publication of (20, 23, 24). These include the ada-alkB operon (23, 25, 26) this article were defrayed in part by the payment of page charges. as well as the alkA gene (19, 27). The induction of the ada This article must therefore be hereby marked "aduertisement" in and alkA DNA repair genes leads toincreased cellular resistaccordance with 18 U.S.C. Section 1734 solely to indicate thisfact. ance to both the mutagenic andkilling the effectsof alkylating ยง Present address: Dept. of Genetics, Harvard Medical School and to as the adaptive response to Dept. of Molecular Biology, Massachusetts General Hospital, Boston, agents, aresponsereferred alkylation damage (28).Ada is readily cleaved in cell extracts MA 02114. (1 Present address: Lawrence Berkeley Laboratory, Cell & Molecular Biology Division, University of California, Berkeley, CA 94720. ** Supported by an American Cancer Society Faculty Research Award. To whom correspondence should be sent. Tel.: 617-432-1085; Fax: 617-432-1780.

' The abbreviations used are: 0"MeG; @-methylguanine; MTase, methyltransferase; 04MeT, 04-methylthymine; 25-mer, 25-base pair DNA oligomer; HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid.

2767

2768

Methyltransferases Affinity of

for @MeG and 04MeT

induction of ada69,0.2 mM IPTG was added to themedium when the into a 19- and a 20-kDa peptide (16,29). The 19-kDa fragment retains thecapacity to repair O'MeG and 04MeTlesions and cell density reached 4 X lo7 cells/ml, and cells were harvested 3 h later. Yeast cells were grown in YPD medium (1% yeast extract, 2% the 20-kDa fragment retains the capacity to repair methylBacto-peptone, and 2% dextrose) at 30 "C to a final density of 5 X phosphotriester lesions; neither fragment,when alkylated, has lo7 cells/ml. HeLa S3 cells were culturedin Dulbecco's modified the ability to induce ada gene expression (23, 24). A second Eagle's medium supplemented with calf serum (lo%),penicillin (100 DNA MTase activity, DNA MTase 11, was found in E. coli units/ml), streptomycin (100 yglrnl), and L-glutamine (0.03%) and (30). MTase I1 is a constitutively expressed 19-kDa protein grown to subconfluence a t 37 "C in humidified 5%CO,. Cells were that repairs both O'MeG and 04MeT lesions in vitro (30). harvested, resuspended in MTase assay buffer (10% glycerol, 50 mM HEPES, pH 7.5, 1 mM EDTA, 10 mM dithiothreitol) containing 5 Rebeck and Samson' have shown that DNA MTase I1 is pg/ml of each of the protease inhibitors leupeptin, aprotinin, and encoded by the ogt gene first isolated by Potter et al. (31). For pepstatin A (United States Biochemical Corp.), and lysed by sonicasimplicity, DNA MTase I1 will hereafter be referred to asOgt. tion (HeLa S3 cells) or by one passage through a French pressured Eukaryotic organisms express an O'MeG-DNA MTase ac- cell (yeast and bacteria). The lysateswere centrifuged a t 20,000 X g for 45 min, and the clarified supernatants were tested for MTase tivity (32, 33), but the role of excision repair or any other mechanism in the repairof either O'MeG or 04MeT has not activity. Protein concentrations were determined by the Bradford been characterized.The 25-kDa human DNA MTase hasbeen assay. partially purified (34-36) and its gene was recently cloned DNA Substrates (37-39). The substratespecificity of the human MTasefor 0@ - f H / M e G DNA-Micrococcusluteus DNA (ICN Biomedicals, alkyl lesions has been extensively investigated in vitro, and Inc.) enriched for 0-alkyl lesions was prepared essentially as described the results suggest that it can repair O'MeG but not 04MeT (41)using [ 3 H ] m e t h y l n i t r ~ ~ ~ ufrom r e a Amersham Corp. (10 Ci/ (32). Indeed, O'MeG has been shown to have a much shorter mmol)andhad a specific activity of 115 cpm/pg; thissubstrate half life in vivo than 04MeT (32, 40). The different in vivo contained 40% O'MeG and about 7% O'MeT. Oligonucleotides-25-mers containing a single O'MeG lesion or a kinetics of O'MeG and 04MeT removal have been taken to indicate that these lesions are repaired via different mecha- single 0'MeT lesion of the structure 5'-CCGC(m4T)AGCGGGTAand 5'-CCGCTA(mGG)CGGGTACCGAGnisms and, more specifically, that 04MeT is notrepaired by CGGAGCTCGAAT-3' CTCGAAT-3', and the correspondingunmodified 25-mer were synthe human DNA MTase. As a result, relative to O'MeG, thesized on an Applied Biosystemsmodel 381A automated synthesizer 04MeT lesions accumulate in mammalian cells (40), and, in by the cyanoethylphosphoramidite method (42, 43). The O'MeG view of their high mutagenic potential, they may contribute modified monomer was purchasedfrom AppliedBiosystems and added manually.The O'-methylthymidine was synthesized from thymore to mutagenesis than O'MeG. We recently identified and partially characterized a DNA midine using the method of Singer et al. (44). O'-methylthymidine then protectedusing 4,4'-dimethoxytrityl chloride and converted repair MTase from the yeast Saccharomyces cerevisiae (33). was to cyanoethanolphosphoramidite as described previously (10). A few The yeast MTase is a 25-kDa protein that repairsO'MeG at modifications were needed forthe synthesisof the 04MeT-containing 25 "C in vitro. We were unable to detect a MTase repair nucleotide because of the lability of the methyl group of the 04MeT. activityfor 04MeT or methylphosphotriester lesions in S. The 04MeT-protected monomer was added manually and after its addition the capping reactions were discontinued. After synthesis, cerevisiae extracts and concluded that, as with the human the oligonucleotides were deprotected under non-aqueous conditions MTase, theyeast MTase can only repair O'MeG lesions. using 1,8-diazabicyclo[5-4-0]undec-7-ene in tetrahydrofuran and In the present study we have measuredthe relative affinities by methanol (45,46) andpurified by electrophoresis ona 20% polyacrylof these four DNA MTases for O'MeG and 04MeT lesions. amide- urea gel (40 X 0.15 cm). The oligomers were visualized by UV Double-stranded 25-base pair DNA oligomers (25-mers), con- shadowing, excised from the gel, elutedin water, desaltedon a lyophilized or inTE (10 mM taining either a single O'MeG or a single 04MeT lesion, were Sephadex G10 column, and stored either used to measure the relative efficiency with which the two E Tris, 1 mM EDTA) at -70 "C. A complementary, unmodified synthetic oligonucleotide was annealed tothethree oligonucleotides coli MTases, the yeast, and the human MTases repair O'MeG described above by incubating an equal amount of the two compleand 04MeT. Our results suggest that DNA MTase I1 may mentary strands a t 80 "C for 10 minfollowed by slow cooling to room play an important role in the repair of 04MeT in E. coli and, temperature. surprisingly, that the eukaryotic DNA MTases are able to DNA MTase ASSQ~S recognize 04MeT in the oligomer substrate used in thisstudy. All four MTases repaired O'MeG efficiently. Cellular extracts containing 0.15-0.3 pmol of active MTase were EXPERIMENTALPROCEDURES

Strains

S. cereuisiae DBY745 (adel-lOO,ura3-52,leu2-3,112)was a gift from Eric Eisenstadt(Office of Naval Research,Arlington, VA). BS23 is a derivative of E. coli B/r F26 and carries a deletion of the Q ~ Q alkB operon (15, 29). GWRlO9 is an ogt- derivative of BS23.' The plasmid pDS427 was a gift from Diane Shevell and Graham Walker (Massachusetts Institute of Technology) and was transformed into GW7109 (AB1157,Aad~-25,ogt::Kan,[F' proA proB,lacIQ] lacZAM15,TnlO). pDS427 carries a mutant ada gene (ada69) that hasbeen placed under the control of the lac promoter and encodes an Ada protein with a methionine instead of a cysteine at position 69.3 The human HeLa Mer' S3 cell line was obtained from the American Tissue Type Culture Collection.

Extract Preparation Bacteria were grown in LB medium (1%NaC1, 1%tryptone, 0.5% yeast extract) a t 37 "C to a concentration of 2 X 10' cells/ml. For G. W. Rebeck and L. Samson, submitted for publication. D. Shevell and G. Walker, unpublished experiments.

incubated with various amounts of double-stranded 25-mers for 15 min at 37 "C (for human and bacterial extracts) or 30 min a t 25 "C (for yeast extracts) in a reaction volume of 185 p1 (for human and bacterial extracts) or250 p1 (for yeast extracts) containing20 mg/ml bovine serum albumin. Eighty percentof the inhibitory effect of the derivatized oligomers on MTase activity was observedwithin the first 5 min of incubation (data not shown). To measure the remaining MTase activity, an excess of 06-[3H]MeGM. luteus DNA substrate (16.5 pg) was added and the reaction continued for another 15 min a t 37 "C (for human and bacterial extracts) or 30 min a t 25 "C (for yeast extracts); labeled DNA substrate was added a t about a 4-fold excess over the original level of MTase. This amount of high molecular weight 06-[3H]MeGDNA was sufficient to compete out the 0alkylated 25-mer substrates efficiently even when both DNA substrates were added simultaneously to the extracts (data not shown), presumably becausehighmolecular weight DNA is thepreferred substrate for DNA MTases. The reaction was stopped by the addition of perchloric acid to 1 M, and the transferof tritiated methyl groups to DNA MTases was measured as described previously (47). Briefly, the acid-precipitated proteins and DNA were incubated for 1 h a t 70 "C to hydrolyze DNA, and the remainingacid-precipitable radioactivity associated with DNA MTases was collected by centrifugation. The pellet was washed twicewith 1 M perchloric acid, resuspended in 10 mM NaOH, and the radioactivity coupled to the acid-insoluble

Affinity of Methyltransferases for 06MeG and material was quantitated by scintillation counting in aquasol (DuPont). In order to compare the relative affinity of the various MTases forOfiMeG and 04MeT lesions, we normalized the 100% MTase activity in all experiments to 0.11 pmol of MTase. RESULTS

The relative efficiencies of DNA MTases for the repairof O'MeG and 04MeTDNA lesions can be estimated from the efficiency with which an O'MeG- or 04MeT-containingDNA substrate inactivates the MTase. For thispurpose, the DNA substrates must contain only the O'MeG lesion or only the 04MeT lesion. In this study we used duplex 25-mers of the sequence 5'-CCGCTAsCGGGTACGGAGCTCGAAT-3' containing an 04MeT atposition 5 or an O'MeG at position 7. Using these oligomers we measured the relative efficiencies of the E. coli Ada and Ogt MTases, the S. cereuisiae MTase and the human MTasefor the repair of O'MeG and 04MeT. E. coli DNA MTases-To measure the efficiency with which the constitutive 19-kDa Ogt repairs O'MeG and 04MeT lesions, cellular extracts from the E. coli uda deletion mutant BS23 were incubated for 15min at 37 "C with various concentrations of 25-mers containingeitherthe O'MeG or the 04MeT lesion. The extent of Ogt inactivation was indirectly measured by the subsequent addition of excess 06-[3H]MeG M. luteus DNA substrate tolabel the unreactedOgt molecules. We found that Ogt is readily inactivated by the 04MeT oligomer, with an IC,, of 3 nM (Fig. lA). Moreover, the 04MeT oligomer was almost %fold more effective a t inhibiting Ogt than was the O'MeG oligomer, which had an ICs0 of 8.1 nM (Fig. lA). To compare the relativeaffinity of the Ada MTase for O'MeG and 04MeTwe used the purified 19-kDa fragment of the Ada protein, which contains the OGMeG/04MeTrecognition site but lacks the methylphosphotriester recognition

0

IO

5

IS

0

40

20

60

so

IO0

oligomer (ntl)

0

100

200

0

40

80

I20

oligomer (nM)

FIG. 1. A, cellular extract from BS23. B, purified 19-kDa Ada protein mixed with cellular extract from strain GWRlOS. C, cellular extracts from yeast strain DBY745. D, cellular extract from HeLa S3 cells. Extracts were incubated 15 min at 37 'C ( A , B, and D )or 30 min at 25 "C ( C )with the indicatedamounts of O'MeG 25-mer (closed symbols), 04MeT 25-mer (open symbols), or unmodified 25-mer (closed circles). The remaining MTase activity was measured by a rapid assay following the addition of tritium-labeled high molecular weight DNA and further incubation at 37 "C for 15 min or at 25 "C for 30 min ( C ) as described under "Experimental Procedures."

@MeT

2769

site (cysteine 69)present in thefull-sized 39-kDa Ada protein. The purified 19-kDa Ada fragment was mixed with cellular extract from the E. coli ogt uda double mutant GWRlO9, completely deficient in MTase activity, to provide the same experimental conditions as those used for Ogt. An O'MeG 25mer concentration of 1.25 nM inactivated 50% of the initial DNA MTase activity, whereas 27.5 nM of the 04MeT25-mer was required to reach the samelevel of inactivation (Fig. 1B). Thus, comparing theICsovalues obtained under these experimental conditions, the repair efficiency of the Ada MTase for O'MeG lesions is about 22-fold higher than for 04MeT. Previous experiments have shown that the purified 19-kDa Ada fragment and its purified 39-kDa precursor repair O-alkyl lesions at the same rate (48) suggest and that results obtained with the 19-kDaAda fragment can be extrapolated to the39kDa protein. To verify this, we measured the relative O'MeG and 04MeTaffinities in cell extracts containinga mutant 39kDa Ada protein inwhich the cysteine at position 69 has been replaced by a methionine, eliminating methylation of the protein by methylphosphotriester lesions. As for the 19-kDa Ada MTase, this 39-kDa MTase had a much higher affinity for O'MeG than 04MeT with ICs0 values of 3 and 48 nM, respectively. The addition of control unmodified 25-mer up to 100 nM did not inhibit any of these DNA MTases (data notshown),demonstratingthatthe oberved inhibition is specific to the presence of O'MeG and 04MeT. Eukaryotic MTases-To measure the efficiency with which the S. cereuisiue DNA MTase repairs O'MeG and 04MeT lesions, yeast cell extracts were incubated with various concentrations of the 25-mers a t 25 "C for 30 min. Incubations were at 25 "C, because the activity of the yeast DNA MTase is labile at 37 "C (33). The ICso values were 1.3 nM for the O'MeG substrate and 44 nM for the 04MeT substrate (Fig. IC). The ICs0values for the human MTasewere determined by using the same experimental conditions as with the E. coli MTases and were 1.4 and 27 nM for the O'MeG and 04MeT substrate, respectively (Fig. 1D).Incontrasttotheother three MTases,we observed a biphasic inhibition of the human MTase with the O'MeG substrate; there was very little inhibition at concentrations below 1.5 nM (less than 25% inactivation) but striking inhibition a t slightly higher concentrations ofO'MeG substrate (see Fig. 2C). The concentration threshold for O'MeG was still observed when the reaction time was increased or when the concentration of total oligomer was maintained constant by the addition of unmodified A biphasic inhibition of the human oligomer (data not shown). MTase was not observed for the 04MeT substrate over the concentration range used here. The inactivation of the eukaryotic MTases by the 04MeT containing oligomer was an unexpected observation in view of the existing literature (32). We wondered, therefore, whether the observed enzyme inhibition could be due simply to thehigh concentration of DNA used in thisassay. However the additionof up to 250 nM of control 25-mer to human and yeast extracts had no detectable effect on the activity of the MTases (Fig. 1, C and D).Thus, the inactivating effect of 04MeT on the human and yeast MTases is not likely to be an artifact due to high DNA oligomer concentrations. The factthat O'MeG repair by both eukaryotic MTases was inhibited almost totally by the 04MeT 25-mer suggests that as for Ada and Ogt, O'MeG and 04MeT arerepaired by the same cysteine residues. We note, however, that these observations alone do not directly demonstrate that the inactivation of the two eukaryotic DNA MTases by 04MeT results from the actual transferof methyl groups from the lesions to theproteins.

2770

Affinity of

Methyltransferases for

O'MeG and 0 4 M e T

for the OGMeG(Fig. 2C) and the 04MeT (Fig. 2 0 ) 25-mer substrates suggesting that they areequally efficient a t repairing OGMeGand 04MeTlesions in this in vitro system. However, it should be noted that these ICs0 values will not only reflect the efficiency of methyl transfer, but also the efficiency with which the MTase reacts with a 25-mer DNA substrate. Thus, strictly speaking, a comparison of repair efficiencies may only be drawn for the repairof O'MeG versus 04MeT by a particular MTase. The human MTase is poorly inactivated by low concentrations of O'MeG 25-mer (Fig. 2C, inset) and could result from the binding of a protein present in the cellular extract, possibly a nucleotide excision repair enzyme, to thelesion preventing it from interacting with the MTase. In only one studyhave mammalian cell extracts been shown to remove 04MeT from a poly(dA. [3H,C]dT) oligonucleotide 0 S 10 1s 0 20 40 60 BO 100 120 substrate (53), but the mechanism of repair was not deter06MeG oligomer (nM) 04MeT oligomer (nM) mined. Transfer of radiolabeled methyl groups from 04MeT FIG. 2. The data from Fig. 1, A , B, and D , were combined lesions in a roughly 12-mer poly( [3H3C]dT) .poly(dA)DNA t o compare the inhibitory effectof the O'MeG ( A and C) and substrate onto MTasefrom yeast or human cells has notbeen 04MeT( B and D ) 25-mers on bacterialOgt MTase (squares), Ada MTase (closed circles), and human MTase (open circles). detected (32, 33). Furthermore, Dolan et al. (51) failed to observe mammalian cell extract mediated repairof 04MeT in Note thetwo different ordinate andabcissa scales. a dodecamer that self-anneals to form a double-stranded 6mer. Perhaps the combination of a low affinity for 04MeT DISCUSSION and shortoligonucleotide substates precludes the detectionof Synthetic oligonucleotides contianing a single 0-alkyl le- 04MeT repair. Indeed, it appears thateven for O'MeG repair sion were used to measure the efficiency with which various very short oligomers are notgood substrates for DNA MTases (48). Moreover, competition experiments between the methDNA MTases repair O'MeG and 04MeT. Our results are based on the inhibitory effect of these oligonucleotides on ylated 25-mers and alkylatedhigh molecular weight M. luteus MTase activity rather than the appearance of their demeth- DNA indicated atwo to three order of magnitude greater ylated forms as described by others (48-51). We have calcu- preference of the MTasesfor the high molecular weight DNA lated the concentration of oligomer required to inhibit 50% substrate (data not shown). In this study we have shown that a DNA oligomer harboring of the initial MTase activity (ICso), for E. coli, yeast, and To facilitate a direct comparisonof the a single 04MeT lesion specifically inhibits yeast and human human DNA MTases. inhibitory effect of the 0-alkylated 25-mers, the E. coli and MTase activity, albeitwith low affinity. Inhibition is probably human MTase datahave been assembled intoone graph (Fig. caused by the actual transfer of methyl groups from 04MeT 2). This direct comparison could not be made between the to the MTase, but we have not yet ruled out the possibility yeast and the other MTasesbecause the reaction conditions that MTase simply binds to the 04MeT lesion and is thus O'MeG lesion. differed. Fig. 2.4 shows that the E. coli Ogt and Ada MTases preventedfromsubsequentlyrepairingan are bothefficiently inactivated by the O'MeG 25-mer (ICsoof Since the human(37-39) and yeast4 DNA MTasegenes were 1.25 nM for Ada and 8.1 nM for Ogt) but the constitutiveOgt recently cloned it shouldsoon be possible to confirm whether DNA MTase (Fig. 2B) is inactivated by much lower concen- these MTases do indeed transfer methyl groups from 04MeT trations of 04MeT lesions (ICso of 3 nM) than the inducible both in vitro and i n vivo. Ada MTase (ICso of 27.5 nM). The Ogt MTase is present in Acknowledgments-We thank D.Shevell, G. Walker, and E. EisenE. coli at about 30 molecules/cell (30,52), and the fact that it stadt for cell strains and plasmids and G. W.Rebeck for discussion. has ahigheraffinityfor 04MeT than O'MeG will clearly influence the repair of aklylated DNA continually produced Addendum-Purified human O'MeG DNA MTase was very reby endogenous alkylating agents. Since 0 4 M e T is formed at cently shown to repair 04MeTDNA lesions in vitro (54). much lower concentrations than O'MeG, the bias of Ogt for REFERENCES 0 4 M e Twould permit the repairof 04MeT as well as O'MeG. 1. Lindahl, T., Sedgwick,B.,Sekiguchi, M., andNakabeppu, Y. If, instead, Ogt were like the Ada MTase with a 20-fold higher (1988) Annu.. Reu. Biochern. 5 7 , 133-158 affinity for O'MeG than 04MeT, 04MeT would probably 2. Boiteux, S., Huisman, O., and Laval, J. (1984) EMBO J . 3,2569seldom be repaired becausethe MTasewould be preferentially 2573 inactivated by the continual flux of high O'MeG concentra3. Friedberg, E. C. (1985) in DNA Repair, W. H. Freeman Co., New tions (relative to 04MeT) in the genome. The Ada MTase, York 4. Eadie, J. S., Conrad, M., Toorchen, D., and Topal, M. D. (1984) normally present at aboutone molecule per cell (52), is Nature 308, 201-203 induced several thousand-fold in adapted E. coli cells; this 5. Saffhill, R., Margison, G. P., and O'Connor P. J. (1985) Biochim. great excess of the Ada MTase probably compensates for its Biophys. Acta 823, 111-145 lower affinity for O'MeT. Our measurements of the affinity 6. Richardson, K. K., Richardson, F. C., Crosby R. M., Swenberg, of the E. coli MTases for O'MeG and 04MeT are consistent J. A,, and Skopek, T. R. (1987) Proc. Natl. Acad. Sci. U. S. A. 84,344-348 with related reports of others. Recent kinetic studiesshowed 7. Loechler, E. L., Green, C. L., and Essigmann, J. M. (1984) Proc. that Ada removes 0 4 M e Tlesions a t a 10,000-fold slower rate Natl. Acad. Sci. U. S. A. 81, 6271-6275 than O'MeG lesions (48) and that Ogt repairs 04MeT 84 8. Ellison, K. S., Dogliotti, E., Connors, T. D., Basu, A. K., and times faster than the Ada MTase (50). However, ours is the Essigmann, J. M. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, first report to show that Ogt MTase repairs 04MeT lesions 8620-8624 more efficiently than it repairsO'MeG lesions. B. Derfler and L. Samson, unpublished experiments. The human and the Ada MTases show quite similar IC&

Affinity of Methyltransferases for 06MeG and @MeT 9. Preston, B. D., Singer, B., and Loeb,L. A. (1986) Proc. Natl. Acad. Sci. U. S. A. 83,8501-8505 10. Dosanjh, M. K., Essigmann, J. M., Goodman, M., and Singer, B. (1990) Biochemistry 219,4698-4703 11. Topal, M. D., Eadie, J. S., and Conrad, M. (1986) J. Biol. Chem. 26 1,9879-9885 12. Samson, L., Thomale, J., and Rajewsky, M. (1988) EMBO J. 7 , 2261-2267 13. Lindahl, T., Demple, B., and Robins, P. (1982) EMBO J. 1,13591363 14. Olsson, M., and Lindahl, T. (1980) J. Biol. Chem. 255, 1056910571 15. Sedgwick, B. (1983) Mol. 6Gen. Genet. 191,466-472 16. Teo, I., Sedgwick, B. B., Demple,B.,Li,B., and Lindahl, T. (1984) EMBO J . 3, 2151-2157 17. Margison, G. P., Cooper, D. P., and Brennand, J. (1985) Nucleic Acids Res. 13. 1939-1952 18. Sedgwick, B., Robins, P., Totty, N., and Lindahl, T. (1988) J. Biol. Chem. 263,4430-4433 19. McCarthy, T. V., Karran, P., and Lindahl, T. (1984) EMBO J. 3,545-550 20. Takano, K., Nakabeppu, Y., and Sekiguchi, M. (1988) J. Mol. Biol. 201,261-271 21. McCarthy, T. V., and Lindhal, T. (1985) Nucleic Acids Res. 13, 2683-2698 22. Weinfeld, M., Drake, A. F., Saunders, T. K., and Paterson, M. C. (1985) Nucleic Acids Res. 13, 7067-7077 23. Teo, I., Sedgwick, B., Kilpatrick, M.W., McCarthy, T. V., and Lindahl, T. (1986) Cett 45, 315-324 24. Yoshikai, T., Nakabeppu, Y., and Sekiguchi, M. (1988) J. Mol. Biol. 263, 19174-19180 25. Nakabeppu, Y., and Sekiguchi, M. (1986) Proc. Natl. Acad. Sci. U. S. A . 83,6297-6301 26. Kataoka, H., and Sekiguchi, M. (1985) Mol. & Gen. Genet. 198, 263-269 27. Evensen, G., and Seeberg, E. (1982) Nature 296,773-775 28. Samson, L., and Cairns., J. (1977) Nature 267,281-283 29. Teo, I. A. (1987) Mutat. Res. 183, 123-127 30. Rebeck, G. W.,Coons, S., Carroll, P., and Samson, L. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 3039-3043 31. Potter, P. M., Wilkinson, M. C., Fitton, J., Carr, F.J., Brennand, J., Cooper, D. P., and Margison, G. P. (1987) Nucleic Acids Res. 15,9177-9193 32. Brent, T. P., Dolan, M. E., Fraenkel-Conrat, H., Hall, J., Karran,

P., Laval, F., Margison, G. P., Montesano, R.,Pegg, A. E., Potter, P. M., Singer, B., Swenberg, J. A,, and Yarosh, D. B.

2771

(1988) Proc. Natl. Acad. Sci. U. S. A . 85, 1759-1762 33. Sassanfar, M., and Samson, L. (1990) J. Biol. Chem. 265,20-25 34. Brent, T. (1985) Pharmacol. & Ther. 31, 121-140 35. Pfeifer, G. P., Grunwald, S., Palitti, F., Kaul, S., Boehm, T. L. J., Hirth, H-P., and Drahovsky, D. (1985) J. Biol. Chem. 260, 13787-13793 36. Boulden, A. M., Foote, R. S., Fleming, G. S., and Mitra, S. (1987) J. Biosci. 11, 215-224 37. Tano, K., Shiota, S., Collier, J., Foote, R. S., and Mitra, S. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 686-690 38. Rydberg,B., Spurr, N., and Karran, P. (1990) J. Biol. Chem. 265,9563-9569 39. Hayakawa, H., Koike, G., and Sekiguchi, M. (1990) J. Mol. Biol. 213,739-747 40. Richardson, F. C., Dyroff, M. C.,Boucheron, J. A., and Swenberg, J. A. (1985) Carcinogenesis 6, 625-629 41. Karran, P., Lindahl, T., and Griffin, B. (1979) Nature 280, 7677 42. Gait, M. J . (ed) (1984) Oligonucleotide Synthesis:A Practical

Approach, IRL, Washington, D. C. 43. Sinha, N. D., Biernat, J., McManus, J., and Koster, H. (1984) Nucleic Acids Res. 12,4539-4557 44. Singer, B., Sagi, J., andKuimierek J. T. (1983) Proc. Natl. Acad. Sci. U. S. A . 80,4884-4888 45. Kuzmich, S., Marky, L. A,, and Jones, R. A. (1983) Nucleic Acids Res. 11,3393-3407 46. Singer, B., Chavez, F., Goodman, M. F., Essigmann, J. M., and Dosanjh, M. K. (1989) Proc. Natl. Acad. Sci.U. S. A. 86,82718274 47. Margison, G. P., Butler, J., and Hoey, B. (1985) Carcinogenesis 6, 1699-1702 48. Graves, R. J., Li, B. F. L., and Swann, P. F. (1989) Carcinogenesis 10,661-666 49. Scicchitano, D., Jones, R. A., Kuzmich, S., Gaffney, B., Lasko, D. D., Essigmann, J. M., and Pegg, A. E. (1986) Carcinogenesis 8,1383-1386 50. Wilkinson, M. C., Potter, P. M., Cawkwell, L., Georgiadis, P., Patel, D., Swann, P. F., and Margison, G. P. (1989) Nucleic Acids Research 17,8475-8484 51. Dolan, M. E., Oplinger, M., and Pegg, A. E. (1988) Mutation Res. 193, 131-137 52. Rebeck, G. W., Smith, C. M., Goad, D. L., and Samson, L. (1989) J. Bacteriol. 171, 4563-4568 53. Becker, R. A., and Montesano R. (1985) Carcinogenesis 6, 313317 54. Koike, G., Maki, H., Takeya, H., Hayakawa, H., and Sekiguchi, M. (1990) J. Biol. Chem. 265, 14754-14762