Dissecting the Broad Substrate Specificity of Human 3-Methyladenine-DNA Glycosylase
2004; Elsevier BV; Volume: 279; Issue: 11 Linguagem: Inglês
10.1074/jbc.m312232200
ISSN1083-351X
AutoresPatrick O’Brien, Tom Ellenberger,
Tópico(s)CRISPR and Genetic Engineering
ResumoHuman alkyladenine-DNA glycosylase (AAG) catalyzes the excision of a broad range of modified bases, protecting the genome from many types of alkylative and oxidative DNA damage. We have investigated how AAG discriminates against normal DNA bases, while accommodating a structurally diverse set of lesioned bases, by measuring the rates of AAG-catalyzed (kst) and spontaneous N-glycosidic bond hydrolysis (knon) for damaged and undamaged DNA oligonucleotides. The rate enhancements for excision of different bases reveal that AAG is most adept at excising the deaminated lesion hypoxanthine (kst/knon = 108), suggesting that enzymatic activity may have evolved in response to this lesion. Comparisons of the rate enhancements for excision of normal and modified purine nucleobases provide evidence that AAG excludes the normal purines via steric clashes with the exocyclic amino groups of adenine and guanine. However, methylated purines are more chemically labile, and only modest rate enhancements are required for their efficient excision. Base flipping also contributes to specificity as destabilized mismatched base pairs are better substrates than stable Watson-Crick pairs, and many of the lesions recognized by AAG are compromised in their ability to base pair. These findings suggest that AAG reconciles a broad substrate tolerance with the biological imperative to avoid normal DNA by excluding normal bases from the active site rather than by specifically recognizing each lesion. Human alkyladenine-DNA glycosylase (AAG) catalyzes the excision of a broad range of modified bases, protecting the genome from many types of alkylative and oxidative DNA damage. We have investigated how AAG discriminates against normal DNA bases, while accommodating a structurally diverse set of lesioned bases, by measuring the rates of AAG-catalyzed (kst) and spontaneous N-glycosidic bond hydrolysis (knon) for damaged and undamaged DNA oligonucleotides. The rate enhancements for excision of different bases reveal that AAG is most adept at excising the deaminated lesion hypoxanthine (kst/knon = 108), suggesting that enzymatic activity may have evolved in response to this lesion. Comparisons of the rate enhancements for excision of normal and modified purine nucleobases provide evidence that AAG excludes the normal purines via steric clashes with the exocyclic amino groups of adenine and guanine. However, methylated purines are more chemically labile, and only modest rate enhancements are required for their efficient excision. Base flipping also contributes to specificity as destabilized mismatched base pairs are better substrates than stable Watson-Crick pairs, and many of the lesions recognized by AAG are compromised in their ability to base pair. These findings suggest that AAG reconciles a broad substrate tolerance with the biological imperative to avoid normal DNA by excluding normal bases from the active site rather than by specifically recognizing each lesion. DNA bases undergo spontaneous deamination and can be alkylated by reactive cellular metabolites and environmental mutagens (1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar). These damaged bases can block DNA-templated activities such as replication and transcription and cause mutations during replication. Failure to repair these inevitable modifications of DNA bases could lead to cancer or cell death, and thus, it is not surprising that there is an elaborate cellular machinery for the repair of DNA damage. Most DNA modifications affecting single bases are repaired via the base excision repair pathway. This pathway is initiated by one of many lesion-specific DNA glycosylases that survey the genome for damage and catalyze the hydrolysis of the N-glycosidic bond to release the lesioned base and generate an abasic site. In human cells a single enzyme, alkyladenine-DNA glycosylase (AAG, 1The abbreviations used are: AAG, human alkyladenine-DNA glycosylase; P, purine; 2AP, 2-aminopurine; 3MeA, 3-methyladenine; 7MeG, 7-methylguanine; ∈A, 1,N6-ethenoadenine; Hx, hypoxanthine; MOPS, 4-morpholinepropanesulfonic acid. also known as methylpurine-DNA glycosylase), is responsible for the recognition and excision of a diverse group of alkylated purine bases, including 3-methyladenine (3MeA), 7-methylguanine (7MeG), and 1-N6-ethenoadenine (∈A) (2Gallagher P.E. Brent T.P. Biochem. J. 1982; 21: 6404-6409Crossref Scopus (24) Google Scholar, 3Singer B. Antoccia A. Basu A.K. Dosanjh M.K. Fraenkel-Conrat H. Gallagher P.E. Kusmierek J.T. Qiu Z.H. Rydberg B. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 9386-9390Crossref PubMed Scopus (103) Google Scholar, 4O'Connor T.R. Nucleic Acids Res. 1993; 21: 5561-5569Crossref PubMed Scopus (135) Google Scholar, 5Engelward B. Weeda G. Wyatt M.D. Broekhof J. Wit J.D. Donker I. Allan J. Gold B. Hoeijmakers J. Samson L. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 13087-13092Crossref PubMed Scopus (206) Google Scholar, 6Hang B. Singer B. Margison G.P. Elder R.H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12869-12874Crossref PubMed Scopus (127) Google Scholar). AAG also removes hypoxanthine (Hx) from deoxyinosine-containing DNA, which is formed by the oxidative deamination of adenosine (6Hang B. Singer B. Margison G.P. Elder R.H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12869-12874Crossref PubMed Scopus (127) Google Scholar, 7Miao F. Bouziane M. O'Connor T.R. Nucleic Acids Res. 1998; 26: 4034-4041Crossref PubMed Scopus (64) Google Scholar, 8Saparbaev M. Laval J. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 5873-5877Crossref PubMed Scopus (234) Google Scholar). This broad substrate specificity is unusual because most DNA glycosylases are specific for a single lesion (9Duncan B.K. Enzymes. 1981; 14: 565-586Crossref Scopus (33) Google Scholar). It is difficult to identify the common structural features of lesioned bases that are recognized by AAG (Fig. 1). The methylated bases 3MeA and 7MeG are positively charged, whereas ∈A and Hx are neutral substrates. ∈A has a larger aromatic surface that is snugly accommodated in the active site (10Lau A.Y. Wyatt M.D. Glassner B.J. Samson L.D. Ellenberger T. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 13573-13578Crossref PubMed Scopus (218) Google Scholar), but Hx is essentially the same size as normal purines that are poor substrates. Indeed each lesioned base has more in common with the normal purines than with the other lesioned bases. This prompts the following question: what prevents AAG from attacking normal DNA? Pyrimidine bases are smaller than purine bases and would be difficult to exclude from the purine-binding pocket. Pyrimidines can bind to AAG, but they are not cleaved (11Biswas T. Clos III, L.J. SantaLucia Jr., J. Mitra S. Roy R. J. Mol. Biol. 2002; 320: 503-513Crossref PubMed Scopus (25) Google Scholar, 12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar). This strong discrimination against the excision of pyrimidines at the hydrolysis step can be explained by an acid-catalyzed mechanism that is optimized for purine excision (12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar). Crystal structures of AAG bound to ∈A-containing DNA suggest molecular models for how lesioned bases might be bound by this enzyme, but paradoxes remain. For example, how are 3MeA and 7MeG selectively excised but not A or G? Functional data are required to evaluate the proposed interactions with these substrates (10Lau A.Y. Wyatt M.D. Glassner B.J. Samson L.D. Ellenberger T. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 13573-13578Crossref PubMed Scopus (218) Google Scholar). To understand how AAG achieves its broad substrate specificity and yet discriminates against the normal undamaged DNA bases, we have determined the rate enhancements for AAG-catalyzed single turnover reactions with a variety of modified purines and with the normal purines found in DNA. The largest rate enhancement observed is for the excision of Hx, suggesting a biological selection for activity on this oxidative lesion. Alkylated purines are removed with similar rate enhancements despite differences in their shapes and charges, indicating a surprisingly tolerant active site pocket. Nevertheless this pocket does discriminate against the normal purines G and A through unfavorable interactions with their exocyclic amino groups. This barrier is overcome for 3MeA and 7MeG, which retain the exocyclic amino groups but are intrinsically more reactive than unmethylated purines. Although AAG provides similar rate enhancements toward methylated and unmethylated purines, the 104-fold greater stability of unmethylated bases ensures that they are only rarely excised, whereas methylated bases are rapidly removed (13Berdal K.G. Johansen R.F. Seeberg E. EMBO J. 1998; 17: 363-367Crossref PubMed Scopus (156) Google Scholar). An additional layer of discrimination against normal Watson-Crick base pairs is provided at the base-flipping step. DNA substrates of the sequence 5′-CGATAGCATCCTXCCTTCTCTCCAT annealed to the complementary oligonucleotide 5′-ATGGAGAGAAGGYAGGATGCTATCG, in which lesion X is paired with base Y (X·Y), were prepared as described previously (12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar). The 7MeG-containing oligonucleotide (7MeG·C) was synthesized by primer extension using Klenow DNA polymerase, and other oligonucleotides were chemically synthesized using the appropriate phosphoramidites (Glen Research) and then purified by denaturing PAGE as described previously (12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar). The catalytic domain of human AAG (Δ80) in which the N-terminal 80 amino acids have been deleted was overexpressed in Escherichia coli and purified (12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar). This domain has glycosylase activity identical to the full-length protein (4O'Connor T.R. Nucleic Acids Res. 1993; 21: 5561-5569Crossref PubMed Scopus (135) Google Scholar, 12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar). Single-stranded oligonucleotides were 5′ labeled with polynucleotide kinase, annealed to the complementary oligonucleotide, and subjected to AAG-catalyzed or spontaneous depurination. The hydrolysis of the N-glycosidic bond was followed with 32P-labeled oligonucleotide substrates as described previously (12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar). Enzymatic reactions were quenched with NaOH (0.2 m final), and nonenzymatic reactions were quenched by freezing. Apurinic DNA sites were cleaved with NaOH (0.2 m, 70 °C, 10 min), diluted with formamide loading buffer, and resolved by denaturing PAGE, and substrate and product bands were quantified with a phosphorimaging system (Fuji BAS1000). Enzymatic rate constants were obtained from exponential fits to the data (F = 1–e–k × t) in which F is the fraction of substrate cleaved, t is time, and k is the observed rate constant. To ensure single turnover conditions, the concentration of AAG was kept in excess of the concentration of labeled DNA (∼1 nm). For Km and kcat/Km determinations, the enzyme concentration was varied over a wide range. Although we refer to the K½ for maximal activity as Km, this value could differ from the Km value for multiple turnover because the latter can be affected by product release. For the determination of kst, the concentration of enzyme was varied over a range of 10–100-fold above the Km to ensure saturation. The inhibition constant (Ki) was measured for various DNA duplexes by inhibiting the Hx excision reaction under kcat/Km conditions (i.e. the concentration of enzyme was below the Km for Hx excision) with varying concentration of unlabeled competitor DNA duplexes. AAG-catalyzed reactions were measured at 37 °C in 50 mm sodium acetate, pH 6.0, with 1 mm dithiothreitol, 1 mm EDTA, and 0.1 mg/ml bovine serum albumin and an ionic strength of 100 mm adjusted with NaCl. Although the pH optimum for the reaction of neutral substrates (∼pH 6.0) is below physiological pH, we previously determined that the alkaline limbs for the pH dependence of the AAG-catalyzed reactions are very similar to those of the uncatalyzed reactions (12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar). The rates of the enzymatic and nonenzymatic reactions decrease to a similar extent with increasing pH, and therefore the rate enhancements are predicted to be pH-independent above pH 6. The fastest nonenzymatic reactions (knon ≥ 1 × 10–4 min–1) were followed to completion and analyzed as described above. Rate constants for slower reactions (knon < 1 × 10–4 min–1) were obtained from the first 10% of the reaction using a linear fit (F = knon × t). The rate constants were measured at 37 °C whenever possible, but for the slower reactions it was necessary to measure the rate constants at higher temperatures and to extrapolate to 37 °C using the temperature dependence. To ensure that the 25-mer oligonucleotide substrates were in duplex form at 70 °C we used a relatively high ionic strength (0.5 m) and a high concentration (1–10 μm) of the complement oligonucleotide. Control reactions showed that the rates of spontaneous reactions were insensitive to the concentration or identity of the buffer, the ionic strength between 100 mm and 1 m, and the presence or absence of dithiothreitol (1 mm) or EDTA (1 mm). Control reactions in which single-stranded DNA was incubated under these reaction conditions revealed 10–30-fold greater rates of hydrolysis (data not shown). This observation is consistent with a previous study that utilized genomic DNA (14Lindahl T. Nyberg B. Biochemistry. 1972; 11: 3610-3618Crossref PubMed Scopus (1180) Google Scholar), and it suggests that these short oligonucleotide duplexes were stable under our experimental conditions. The energetic contributions of individual purine substituents were expressed as ΔΔG‡ in kcal/mol [ΔΔG‡ = –RTln(RE1/RE2) in which R is the gas constant, T is temperature (310 K), and RE is the rate enhancement. A negative value of ΔΔG‡ indicates that a given substituent contributes to catalytic recognition by AAG (Reactions 2, 3, 4, 5, 6, 7, 8).Reactions 5 and 6View Large Image Figure ViewerDownload Hi-res image Download (PPT)Reaction 7View Large Image Figure ViewerDownload Hi-res image Download (PPT)Reaction 8View Large Image Figure ViewerDownload Hi-res image Download (PPT) Glycosylase Activity of AAG toward Damaged Purines with Diverse Structures—To investigate how AAG can accommodate a wide variety of damaged purines differing in shape, charge, and hydrogen-bonding potential, we examined the specificity expressed in the enzyme·DNA complex for cleavage of the N-glycosidic bond. Glycosylase reactions were carried out under single turnover conditions with enzyme in excess over the DNA substrate (kobs = kst; Reaction 1). The single turnover reaction includes all of the steps subsequent to formation of the initial AAG·DNA complex up to and including N-glycosidic bond cleavage. The E·S complex, in which the base is not yet flipped out, is considered to be at equilibrium (Kflip) with the E·S′ complex in which the base is flipped out into the active site pocket. Although AAG is likely to also sample other base pairs in the DNA, only the E·S complex that is on the pathway to excision is shown. We compared activity toward substrates representative of the different structural classes of known AAG substrates: Hx is a small, neutral substrate; ∈A is a bulky but uncharged alkylated substrate; and 7MeG is an alkylated substrate that bears a positive charge (Fig. 1). As the pH dependences of AAG-catalyzed excision of positively charged and neutral lesions are different, the enzymatic rates of excision were measured at the pH optimum for each substrate (12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar). At the pH optima, similar maximal rates of excision were observed for Hx and 7MeG, whereas a lower maximal rate was observed for ∈A (Table I). However, these rates alone do not reveal how well particular substrates are recognized in the transition state because these compounds have very different intrinsic reactivities as described below.Table IEnzymatic and nonenzymatic depurination in oligonucleotide duplexesBase releasedEnzymaticNonenzymaticRate enhancementiRate enhancement is defined as kst/knon.Catalytic proficiencyjCatalytic proficiency is defined as (kcat/Km)/kw in which kw is the bimolecular nonenzymatic rate constant for hydrolysis (kw = knon/55 m).KmaKm values were measured by varying the concentration of enzyme (see “Experimental Procedure”). or KibKi values were measured by inhibiting the Hx excision reaction under kcat/Km conditions (see “Experimental Procedures”).kcat/Kmckcat/Km values were measured as described under “Experimental Procedures.”, dIn some cases the analogous value of kst/Kd is reported instead.(kcat/Km)releValues of (kcat/Km)rel were calculated by dividing the kcat/Km value by that of 3MeA(T).kstfSingle turnover rate constants with saturating enzyme were measured at the pH optimum for each substrate.krelgValues of krel for AAG-catalyzed excision were calculated by dividing the kst value by that of 7MeG·T.knonkrelhValues of krel for nonenzymatic depurination were obtained by dividing the value of knon by that for excision of A.ΔH‡nmm-1s-1min-1min-1kcal/mol7MeG (T)30 ± 10aKm values were measured by varying the concentration of enzyme (see “Experimental Procedure”).1.7 × 107 ckcat/Km values were measured as described under “Experimental Procedures.”0.7431kpH 8.0.(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)1.0 ± 0.2 × 10-363,000NDlNot determined.3.1 × 1045.6 × 10137MeG (C)31 ± 6aKm values were measured by varying the concentration of enzyme (see “Experimental Procedure”).6.0 × 105 ckcat/Km values were measured as described under “Experimental Procedures.”0.031.1kpH 8.0.,mThis value is the same as the kcat for removal of 7MeG from genomic DNA (6).0.041.6 ± 0.2 × 10-4 nA similar value of 1.1 × 10-4 min-1 was obtained for release of 7MeG from methylated genomic DNA (24).10,00024oFrom Ref. 24.6.9 × 1031.2 × 10133MeA (T)8pFrom the previously reported excision of 3MeA from methylated genomic DNA at pH 7.5 (6).2.3 × 107 pFrom the previously reported excision of 3MeA from methylated genomic DNA at pH 7.5 (6).(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)11pFrom the previously reported excision of 3MeA from methylated genomic DNA at pH 7.5 (6).0.353.2 × 10-4 oFrom Ref. 24.20,00023oFrom Ref. 24.3.4 × 1042.4 × 1014Hx (T)18 ± 3aKm values were measured by varying the concentration of enzyme (see “Experimental Procedure”).1.0 × 107 ckcat/Km values were measured as described under “Experimental Procedures.”0.4310.80.351.2 ± 0.2 × 10-77.5219.0 × 1072.8 × 1017∈A (T)2 ± 1aKm values were measured by varying the concentration of enzyme (see “Experimental Procedure”).1.7 × 106 ckcat/Km values were measured as described under “Experimental Procedures.”0.070.26.5 × 10-37.0 ± 1.8 × 10-744192.9 × 1058.0 × 1015P (C)3 ± 1bKi values were measured by inhibiting the Hx excision reaction under kcat/Km conditions (see “Experimental Procedures”).4.8 × 104 dIn some cases the analogous value of kst/Kd is reported instead.2.1 × 10-38.6 × 10-32.8 × 10-41.7 ± 1.1 × 10-711215.1 × 1049.3 × 1014G (T)13 ± 2bKi values were measured by inhibiting the Hx excision reaction under kcat/Km conditions (see “Experimental Procedures”).5.4 × 102 dIn some cases the analogous value of kst/Kd is reported instead.2.3 × 10-54.2 × 10-41.4 × 10-52.9 ± 1 × 10-81.824qFor genomic DNA values of 26 and 27 kcal/mol have been reported (14, 43).1.4 × 1046.2 × 10132AP (C)9 ± 2bKi values were measured by inhibiting the Hx excision reaction under kcat/Km conditions (see “Experimental Procedures”).2.8 × 102 dIn some cases the analogous value of kst/Kd is reported instead.1.2 × 10-51.5 × 10-44.8 × 10-64.3 ± 0.8 × 10-72716rSee footnote 4 in the text.3.5 × 1022.1 × 1012A (C)14 ± 3bKi values were measured by inhibiting the Hx excision reaction under kcat/Km conditions (see “Experimental Procedures”).2.4 × 101 dIn some cases the analogous value of kst/Kd is reported instead.1.0 × 10-62.0 × 10-56.5 × 10-71.6 ± 1 × 10-8(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)251.3 × 1035.0 × 1012a Km values were measured by varying the concentration of enzyme (see “Experimental Procedure”).b Ki values were measured by inhibiting the Hx excision reaction under kcat/Km conditions (see “Experimental Procedures”).c kcat/Km values were measured as described under “Experimental Procedures.”d In some cases the analogous value of kst/Kd is reported instead.e Values of (kcat/Km)rel were calculated by dividing the kcat/Km value by that of 3MeA(T).f Single turnover rate constants with saturating enzyme were measured at the pH optimum for each substrate.g Values of krel for AAG-catalyzed excision were calculated by dividing the kst value by that of 7MeG·T.h Values of krel for nonenzymatic depurination were obtained by dividing the value of knon by that for excision of A.i Rate enhancement is defined as kst/knon.j Catalytic proficiency is defined as (kcat/Km)/kw in which kw is the bimolecular nonenzymatic rate constant for hydrolysis (kw = knon/55 m).k pH 8.0.l Not determined.m This value is the same as the kcat for removal of 7MeG from genomic DNA (6Hang B. Singer B. Margison G.P. Elder R.H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12869-12874Crossref PubMed Scopus (127) Google Scholar).n A similar value of 1.1 × 10-4 min-1 was obtained for release of 7MeG from methylated genomic DNA (24Osborne M.R. Phillips D.H. Chem. Res. Toxicol. 2000; 13: 257-261Crossref PubMed Scopus (36) Google Scholar).o From Ref. 24Osborne M.R. Phillips D.H. Chem. Res. Toxicol. 2000; 13: 257-261Crossref PubMed Scopus (36) Google Scholar.p From the previously reported excision of 3MeA from methylated genomic DNA at pH 7.5 (6Hang B. Singer B. Margison G.P. Elder R.H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 12869-12874Crossref PubMed Scopus (127) Google Scholar).q For genomic DNA values of 26 and 27 kcal/mol have been reported (14Lindahl T. Nyberg B. Biochemistry. 1972; 11: 3610-3618Crossref PubMed Scopus (1180) Google Scholar, 43Bruskov V. Malakhova L. Masalimov Z. Chernikov A. Nucleic Acids Res. 2002; 30: 1354-1363Crossref PubMed Scopus (238) Google Scholar).r See footnote 4 in the text. Open table in a new tab To more fully explore the structure-activity relationships that define the catalytic specificity of AAG, we examined the extent to which other purine analogs are accepted as substrates. Purine (P) has no exocyclic substitutions, and 2-aminopurine (2AP) is a guanine analog lacking the C-6 keto moiety (Fig. 1). Both bases are excised by AAG when opposite cytosine in DNA albeit with greatly reduced rates in comparison to the excision of alkylated bases (Table I). For example, the half-lives for excision of P and 2AP are ∼1 and ∼80 h, respectively, compared with a half-life of ∼4 s for the excision of Hx, 3MeA, and 7MeG. The above comparisons of the reactivities of different modified purine substrates bound in the active site identify the specific contacts with the functional groups of the substrate that affect catalysis. However, the discrimination between substrates competing for a single active site is given by the specificity constant kcat/Km and not simply the kcat value. This is because a given substrate can better compete either by more efficient binding (enhanced Km) or by more efficient reaction once bound (kcat). Therefore we measured Km and kcat/Km values for a variety of purine analogs in DNA (Table I). These specificity constants include all of the steps in the enzymatic reaction up to and including the first irreversible step (Reaction 1). Intriguingly AAG does not exhibit strong discrimination in binding to different DNAs with Km values of ∼10 nm observed for both mismatched, lesioned bases and normal unmodified base pairs under these experimental conditions (Table I and data not shown). The only notable exceptions were ∈A and P, which have an order of magnitude lower Km values (Table I). The similar Km values for most DNAs is consistent with AAG being a nonspecific enzyme, and it indicates that much of the available binding energy is provided by nonspecific interactions with the DNA. AAG Is Sensitive to the Disruption of Base Pairing Interactions—DNA glycosylases gain access to their substrate bases by the process of base flipping whereby they distort the DNA duplex to extrude the target nucleoside into the active site (15Roberts R.J. Cheng X. Annu. Rev. Biochem. 1998; 67: 181-198Crossref PubMed Scopus (308) Google Scholar). We therefore considered the possibility that substrates might have different propensities for base flipping (Kflip; Reaction 1). AAG has been shown to catalyze the removal of unmodified guanine residues from DNA albeit at low rates (13Berdal K.G. Johansen R.F. Seeberg E. EMBO J. 1998; 17: 363-367Crossref PubMed Scopus (156) Google Scholar, 16Connor E.E. Wyatt M.D. Chem. Biol. 2002; 9: 1033-1041Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar). We detected very low levels of activity toward both G and A in normal Watson-Crick base pairs, 2Activity toward normal purines was very low, but control experiments provide strong evidence that AAG is the mismatch-specific glycosylase in our assays: the pH dependence for excision of unmodified purines is similar that of modified purines; ∈A-containing DNA was a strong inhibitor; and catalytically inactive AAG mutants, E125Q and E125A (10Lau A.Y. Wyatt M.D. Glassner B.J. Samson L.D. Ellenberger T. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 13573-13578Crossref PubMed Scopus (218) Google Scholar, 12O'Brien P.J. Ellenberger T. Biochemistry. 2003; 42: 12418-12429Crossref PubMed Scopus (95) Google Scholar, 20Martin F.H. Castro M.M. Aboul-ela F. Tinoco I. Nucleic Acids Res. 1985; 13: 8927-8938Crossref PubMed Scopus (282) Google Scholar), showed no detectable activity toward mismatched DNAs. whereas much greater activity was observed for purines in mismatched base pairs (Table II). Although the excision of G and A from mismatches is several orders of magnitude slower than the excision of damaged bases, glycosylase activity toward mismatches represents substantial catalysis over the nonenzymatic rates (Table I). These opposing base effects could indicate a specific recognition of the orphaned base; however, the relative rates of A and G excision opposite different bases are consistent with the known base pairing stabilities (Table II) and not the identity of the orphaned base with unstable pairings exhibiting correspondingly higher rates of reaction. The decreased activity toward G·G, relative to G·T, is consistent with the observation that a G·G base pair is relatively stable (17Peyret N. Seneviratne P.A. Allawi H.T. SantaLucia Jr., J. Biochemistry. 1999; 38: 3468-3477Crossref PubMed Scopus (490) Google Scholar). Only an upper limit could be obtained for the activity toward a Watson-Crick A·T base pair, but adenine was removed from the least stable mismatch (A·C) at least 10-fold more efficiently than from an A·T base pair (Table II).Table IIEffect of the opposing base on the AAG-catalyzed glycosylase reactionPairedGaSodium acetate, pH 6.0; 37 °C.AaSodium acetate, pH 6.0; 37 °C.7MeGbNaMOPS, pH 8.0; 4 °C.2APaSodium acetate, pH 6.0; 37 °C.PaSodium acetate, pH 6.0; 37 °C.∈AaSodium acetate, pH 6.0; 37 °C.HxcSodium acetate, pH 6.0; 21 °C.Thymine(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)≤0.1(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)0.270.260.56(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)Cytosine0.0210.80.036(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)0.940.4Guanine0.056NDdNot determined.0.360.580.410.610.14Adenine0.43(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)0.820.930.76(1Lindahl T. Nature. 1993; 362: 709-715Crossref PubMed Scopus (4378) Google Scholar)0.18a Sodium acetate, pH 6.0; 37 °C.b NaMOPS, pH 8.0; 4 °C.c Sodium acetate, pH 6.0; 21 °C.d Not determined. Open table in a new tab The magnitude of opposing base effects (up to 50-fold) on the rates of excision of normal bases by AAG indicates that base pairing markedly contributes to substrate selection (Table II). A disruption of base pairing at the site of a lesion could serve as a signal to aid in flipping out damaged bases and thereby contribute to specificity for damaged bases. AAG has been shown to bind specifically to distorted DNAs containing pyrimidine·pyrimidine mismatches (11Biswas T. Clos III, L.J. SantaLucia Jr., J. Mi
Referência(s)