Characterization of Escherichia coli Endonuclease VIII
1997; Elsevier BV; Volume: 272; Issue: 51 Linguagem: Inglês
10.1074/jbc.272.51.32230
ISSN1083-351X
AutoresDongyan Jiang, Zafer Hatahet, Robert J. Melamede, Yoke W. Kow, Susan S. Wallace,
Tópico(s)Bacterial Genetics and Biotechnology
ResumoEscherichia coli endonuclease VIII (endo VIII) was identified as an enzyme that, like endonuclease III (endo III), removes radiolysis products of thymine including thymine glycol, dihydrothymine, β-ureidoisobutyric acid, and urea from double-stranded plasmid or phage DNA and cleaves the DNA strand at abasic (AP) sites (Melamede, R. J., Hatahet, Z., Kow, Y. W., Ide., H., and Wallace, S. S. (1994) Biochemistry 33, 1255–1264). Using apparently homogeneous endo VIII protein, we now show that endo VIII removes from double-stranded oligodeoxyribonucleotides the stable oxidative products of cytosine, 5-hydroxycytosine and 5-hydroxyuracil. Endo VIII cleaved the damage-containing DNA strand by β,δ-elimination as does formamidopyrimidine DNA glycosylase (Fpg). Like Fpg, endo VIII also excised the 5′-terminal deoxyribose phosphate from an endonuclease IV (endo IV) pre-incised AP site. Thus, in addition to amino acid sequence homology (Jiang, D., Hatahet, Z., Blaisdell, J., Melamede, R. J., and Wallace, S. S. (1997) J. Bacteriol. 179, 3773–3782), endo VIII shares a number of catalytic properties with Fpg. In addition, endo VIII specifically bound to oligodeoxynucleotides containing a reduced AP site with a stoichiometry of 1:1 for protein to DNA with an apparent equilibrium dissociation constant of 3.9 nm. Like Fpg and endo III, the DNase I footprint was small with contact sites primarily on the damage-containing strand; unlike Fpg and endo III, the DNA binding of endo VIII to DNA was asymmetric, 3′ to the reduced AP site. Escherichia coli endonuclease VIII (endo VIII) was identified as an enzyme that, like endonuclease III (endo III), removes radiolysis products of thymine including thymine glycol, dihydrothymine, β-ureidoisobutyric acid, and urea from double-stranded plasmid or phage DNA and cleaves the DNA strand at abasic (AP) sites (Melamede, R. J., Hatahet, Z., Kow, Y. W., Ide., H., and Wallace, S. S. (1994) Biochemistry 33, 1255–1264). Using apparently homogeneous endo VIII protein, we now show that endo VIII removes from double-stranded oligodeoxyribonucleotides the stable oxidative products of cytosine, 5-hydroxycytosine and 5-hydroxyuracil. Endo VIII cleaved the damage-containing DNA strand by β,δ-elimination as does formamidopyrimidine DNA glycosylase (Fpg). Like Fpg, endo VIII also excised the 5′-terminal deoxyribose phosphate from an endonuclease IV (endo IV) pre-incised AP site. Thus, in addition to amino acid sequence homology (Jiang, D., Hatahet, Z., Blaisdell, J., Melamede, R. J., and Wallace, S. S. (1997) J. Bacteriol. 179, 3773–3782), endo VIII shares a number of catalytic properties with Fpg. In addition, endo VIII specifically bound to oligodeoxynucleotides containing a reduced AP site with a stoichiometry of 1:1 for protein to DNA with an apparent equilibrium dissociation constant of 3.9 nm. Like Fpg and endo III, the DNase I footprint was small with contact sites primarily on the damage-containing strand; unlike Fpg and endo III, the DNA binding of endo VIII to DNA was asymmetric, 3′ to the reduced AP site. Free radicals can be produced during normal cellular metabolism and after exposure to ionizing radiation, near UV light (320–380 nm), and chemical oxidants (for reviews see Refs. 1Cadenas E. Ahmad S. Oxidative Stress and Antioxidant Defenses in Biology. Chapman & Hall, New York1995: 1-61Crossref Google Scholar and 2Hutchinson F. Prog. Nucleic Acid Res. Mol. Biol. 1985; 32: 115-154Crossref PubMed Scopus (450) Google Scholar). Free radical-mediated damages are believed to be the most frequently occurring DNA damages and include modifications to the purine and pyrimidine bases, the deoxyribose sugar, as well as breaks in the phosphodiester backbone (for reviews see Refs. 2Hutchinson F. Prog. Nucleic Acid Res. Mol. Biol. 1985; 32: 115-154Crossref PubMed Scopus (450) Google Scholar and 3Breen A.P. Murphy J.A. Free Radical Biol. & Med. 1995; 18: 1033-1077Crossref PubMed Scopus (910) Google Scholar). Base excision repair is the major pathway that processes oxidative DNA lesions with Escherichia coli defining the prototypic enzymes. Three E. coli DNA N-glycosylases containing an associated AP lyase activity, formamidopyrimidine DNA glycosylase (Fpg) 1The abbreviations used are: Fpg, formamidopyrimidine DNA glycosylase; endo, endonuclease; Tg, thymine glycol; DHT, dihydrothymine; AP, abasic; redAP, reduced AP; 5-OHC, 5-hydroxycytosine; 5-OHU, 5-hydroxyuracil; Pol I, polymerase I; dRp, deoxyribose phosphate; ds, double-stranded; FAPY, formamidopyrimidine; dTgMP, thymidine glycol 5′-monophosphate; dDHTMP, dihydrothymidine 5′-monophosphate; 5-OHdCHP, 5-hydroxycytidine 5′-monophosphate; 5-OHdVMP, 5-hydroxyuridine 5′-monophosphate. 1The abbreviations used are: Fpg, formamidopyrimidine DNA glycosylase; endo, endonuclease; Tg, thymine glycol; DHT, dihydrothymine; AP, abasic; redAP, reduced AP; 5-OHC, 5-hydroxycytosine; 5-OHU, 5-hydroxyuracil; Pol I, polymerase I; dRp, deoxyribose phosphate; ds, double-stranded; FAPY, formamidopyrimidine; dTgMP, thymidine glycol 5′-monophosphate; dDHTMP, dihydrothymidine 5′-monophosphate; 5-OHdCHP, 5-hydroxycytidine 5′-monophosphate; 5-OHdVMP, 5-hydroxyuridine 5′-monophosphate., endonuclease III (endo III), and endonuclease VIII (endo VIII), have been reported to recognize and remove oxidative base lesions and subsequently cleave the phosphodiester backbone, initiating the repair process (for reviews see Refs. 4Demple B. Harrison L. Annu. Rev. Biochem. 1994; 63: 915-948Crossref PubMed Scopus (1282) Google Scholar and 5Wallace S.S. Scandalios J.G. Oxidative Stress and the Molecular Biology of Antioxidant Defenses. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1997: 49-90Google Scholar). Fpg and endo III have been well studied. Fpg, encoded by the fpg or mutM gene (6Boiteux S. O'Connor T.R. Laval J. EMBO J. 1987; 6: 3177-3183Crossref PubMed Scopus (250) Google Scholar, 7Cabrera M. Nghiem Y. Miller J. J. Bacteriol. 1988; 170: 5405-5407Crossref PubMed Google Scholar), was initially identified by its ability to recognize and excise from DNA 2,6-diamino-4-hydroxy-5-N-methylformamidopyrimidine (FAPY-guanine), the imidazole ring open form of N 7-methylguanine (8Chetsanga C.J. Lindahl T. Nucleic Acids Res. 1979; 6: 3673-3684Crossref PubMed Scopus (229) Google Scholar). Fpg also releases 4,6-diamino-5-formamidopyrimidine (FAPY-adenine) (9Breimer L.H. Nucleic Acids Res. 1984; 12: 6359-6367Crossref PubMed Scopus (76) Google Scholar), 8-oxoguanine (10Tchou J. Kasai H. Shibutani S. Chung M.H. Laval J. Grollman A.P. Nishimura S. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 4690-4694Crossref PubMed Scopus (690) Google Scholar), and a number of pyrimidine products (11Hatahet Z. Kow Y.W. Purmal A.A. Cunningham R.P. Wallace S.S. J. Biol. Chem. 1994; 269: 18814-18820Abstract Full Text PDF PubMed Google Scholar). 8-Oxoguanine, an important biological substrate for Fpg, is a major stable oxidative product of guanine and can pair with both C and A in vitro(12Shibutani S. Takeshita M. Grollman A.P. Nature. 1991; 349: 431-434Crossref PubMed Scopus (2016) Google Scholar, 13Moriya M. Ou C. Bodepudi V. Johnson F. Takeshita M. Grollman A.P. Mutat. Res. 1991; 254: 281-288Crossref PubMed Scopus (325) Google Scholar, 14Chen K.C. Cahill D.S. Kasai H. Nishimura S. Loeb L.A. J. Biol. Chem. 1992; 267: 166-172Abstract Full Text PDF PubMed Google Scholar). If cellular 8-oxoguanine residues are not repaired, G → T tranversions result (13Moriya M. Ou C. Bodepudi V. Johnson F. Takeshita M. Grollman A.P. Mutat. Res. 1991; 254: 281-288Crossref PubMed Scopus (325) Google Scholar, 14Chen K.C. Cahill D.S. Kasai H. Nishimura S. Loeb L.A. J. Biol. Chem. 1992; 267: 166-172Abstract Full Text PDF PubMed Google Scholar, 15Wood M.L. Dizdaroglu M. Gajewski E. Essigmann J.M. Biochemistry. 1990; 29: 7024-7032Crossref PubMed Scopus (690) Google Scholar), and fpg null mutants are spontaneous mutators (16Michaels M.L. Pham L. Cruz C. Miller J.H. Nucleic Acids Res. 1991; 19: 3629-3632Crossref PubMed Scopus (198) Google Scholar). Endonuclease III, a protein with an iron-sulfur cluster (17Cunningham R.P. Asahara H. Bank J.F. Scholes C.P. Salerno J.C. Surerus K. Munck E. McCracken J. Peisach J. Emptage M.H. Biochemistry. 1988; 28: 4450-4455Crossref Scopus (200) Google Scholar), is encoded by the gene nth (18Cunningham R.P. Weiss B. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 474-478Crossref PubMed Scopus (143) Google Scholar). It was initially identified by its endonuclease activity on X-irradiated (19Strniste G.F. Wallace S.S. Proc. Natl. Acad. Sci. U. S. A. 1975; 72: 1997-2001Crossref PubMed Scopus (44) Google Scholar) and heavily UV-irradiated DNA (20Radman M. J. Biol. Chem. 1976; 251: 1438-1445Abstract Full Text PDF PubMed Google Scholar, 21Gates F.T.I. Linn S. J. Biol. Chem. 1977; 252: 2802-2807Abstract Full Text PDF PubMed Google Scholar). Endo III specifically removes radiolysis products of thymine including ring saturated, fragmented, or ring contracted lesions such as, 5,6-dihydrothymine (DHT), 6-hydroxy-5,6-dihydrothymine, cis- and trans-thymine glycol (Tg), 5-hydroxy-5-methylhydantoin, methyltartonyl urea, and urea (for reviews see Refs. 4Demple B. Harrison L. Annu. Rev. Biochem. 1994; 63: 915-948Crossref PubMed Scopus (1282) Google Scholar and 5Wallace S.S. Scandalios J.G. Oxidative Stress and the Molecular Biology of Antioxidant Defenses. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1997: 49-90Google Scholar). Endo III also excises 5-hydroxycytosine (5-OHC), 5-hydroxyuracil (5-OHU), uracil glycol, dihydrouracil (11Hatahet Z. Kow Y.W. Purmal A.A. Cunningham R.P. Wallace S.S. J. Biol. Chem. 1994; 269: 18814-18820Abstract Full Text PDF PubMed Google Scholar, 22Dizdaroglu M. Laval J. Boiteux S. Biochemistry. 1993; 32: 12105-12111Crossref PubMed Scopus (264) Google Scholar), and 6-hydroxyuracil (23Boorsteen R.J. Hilbert T.P. Cadet J. Cunningham R.P. Teebor G.W. Biochemistry. 1989; 28: 6164-6170Crossref PubMed Scopus (102) Google Scholar) which are stable oxidation products of cytosine (2Hutchinson F. Prog. Nucleic Acid Res. Mol. Biol. 1985; 32: 115-154Crossref PubMed Scopus (450) Google Scholar, 3Breen A.P. Murphy J.A. Free Radical Biol. & Med. 1995; 18: 1033-1077Crossref PubMed Scopus (910) Google Scholar, 24Wagner J.R. Hum C.C. Ames B.N. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 3380-3384Crossref PubMed Scopus (238) Google Scholar). A number of these lesions have been well studied with respect to their biological consequences. Both Tg and urea strongly inhibit DNA replication in vitro and are lethal in vivo (for reviews see Refs. 25Evans J. Maccabee M. Hatachet Z. Courcelle J. Bockrath J. Ide H. Wallace S.S. Mutat. Res. 1993; 175: 41-45Google Scholar and 26Wallace S.S. Int. J. Radiat. Biol. 1994; 66: 579-589Crossref PubMed Scopus (140) Google Scholar). The stable oxidative products of cytosine, uracil glycol, dihydrouracil, 5-OHU, and 5-OHC can pair with A during DNA synthesis in vitro(27Purmal A.A. Kow Y.W. Wallace S.S. Nucleic Acids Res. 1994; 22: 72-78Crossref PubMed Scopus (152) Google Scholar) 2A. A. Purmal, G. W. Lampman, Z. Hatahet, and S. S. Wallace, manuscript submitted for publication. 2A. A. Purmal, G. W. Lampman, Z. Hatahet, and S. S. Wallace, manuscript submitted for publication. and thus are potentially mutagenic. In fact, 5-OHC has been shown to be mutagenic in E. coli (29Feig D.I. Sowers L.C. Loeb L.A. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6609-6613Crossref PubMed Scopus (191) Google Scholar). Despite the potential deleterious effects of oxidative lesions recognized and removed by endo III, nthnull mutants are not hypersensitive to H2O2 or ionizing radiation which produces these lesions and exhibits only a minor spontaneous mutator phenotype (18Cunningham R.P. Weiss B. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 474-478Crossref PubMed Scopus (143) Google Scholar). One possibility that could account for the lack of a phenotype in nth mutants is the existence of a redundant activity for the repair of these damaged base substrates. Several years ago, we identified a third E. colioxidative DNA glycosylase, endo VIII, that had a substrate specificity similar to that of endo III (30Melamede R.J. Hatahet Z. Kow Y.W. Ide H. Wallace S.S. Biochemistry. 1994; 33: 1255-1264Crossref PubMed Scopus (159) Google Scholar). Endo VIII recognizes and removes from DNA Tg, DHT, β-ureidoisobutyric acid, and urea and cleaves DNA containing an AP site (30Melamede R.J. Hatahet Z. Kow Y.W. Ide H. Wallace S.S. Biochemistry. 1994; 33: 1255-1264Crossref PubMed Scopus (159) Google Scholar) suggesting it as a candidate for a redundant activity. The gene for endo VIII, nei, was recently cloned, and indeed, double mutants, nth nei, are hypersensitive to the cytotoxic effects of x-rays and hydrogen peroxide and exhibit a 20-fold increase in spontaneous mutation frequency (31Jiang D. Hatahet Z. Blaisdell J. Melamede R.J. Wallace S.S. J. Bacteriol. 1997; 179: 3773-3782Crossref PubMed Google Scholar). In this paper, we report the further characterization of endo VIII and show that it recognizes cytosine products in addition to thymine products and cleaves the DNA backbone via a β,δ-elimination. We also report that endo VIII possesses a 5′-deoxyribophosphodiesterase activity and show the specificity, affinity, and stoichiometry of endo VIII binding to DNA containing a reduced AP site. T4 DNA polynucleotide kinase,E. coli DNA polymerase I, terminal deoxynucleotidyltransferase, and shrimp alkaline phosphatase were purchased from U. S. Biochemical Corp. Uracil-DNA glycosylase was from Epicentre Technologies. E. coli endonuclease III was provided by Dr. Richard Cunningham, State University of New York at Albany. E. coli endonuclease IV and FAPY-DNA glycosylase were purified in this lab using a procedure similar to that for endo VIII (31Jiang D. Hatahet Z. Blaisdell J. Melamede R.J. Wallace S.S. J. Bacteriol. 1997; 179: 3773-3782Crossref PubMed Google Scholar). NAP-5 and Sephadex G-50 columns were from Pharmacia Biotech Inc. A premixed solution of acrylamide plus N,N′-methylenebisacrylamide was from National Diagnostics. [γ-32P]ATP and α-32P-cordycepin were from NEN Life Science Products. All other chemicals were obtained from either Sigma or J. T. Baker Inc. Oligodeoxynucleotides were synthesized in the Department of Microbiology and Molecular Genetics at the University of Vermont or purchased from Operon Technologies. The oligodeoxynucleotides were purified on a 20% denaturing polyacrylamide gel containing 8 m urea. For enzymatic studies of endo VIII, the oligodeoxynucleotides containing an AP site were prepared by incubating a uracil-containing oligodeoxynucleotide with E. coli uracil-DNA glycosylase at 37 °C for 1 h. The reduced AP (redAP) site containing oligodeoxynucleotides were prepared by incubating AP-containing substrates with 1–1.5 mNaBH4 (dissolved in 0.5 n NaOH), pH 13, at room temperature for 2 h, neutralized by 100 mm Tris-HCl, pH 7.5, and then filtered through a NAP-5 column. The preparation of oligodeoxynucleotides containing an internal modified pyrimidine, Tg, DHT, 5-OHC, or 5-OHU was described by Hatahet et al. (32Hatahet Z. Purmal A.A. Wallace S.S. Nucleic Acids Res. 1993; 21: 1563-1568Crossref PubMed Scopus (38) Google Scholar). Briefly, dTgMP, dDHTMP, 5-OHdCMP, or 5-OHdUMP was incorporated on to the 3′ end of an oligodeoxynucleotide with calf thymus terminal deoxynucleotidyltransferase and then the oligodeoxynucleotide containing a 3′-modified base was purified on a denaturing acrylamide gel. The preparation of double-stranded lesion-containing substrate was completed by ligation with T4 DNA ligase of the oligodeoxynucleotide containing a 3′-modified base with its downstream oligodeoxynucleotide after both were annealed to the template or by elongation with E. coli DNA polymerase I Klenow fragment of the oligodeoxynucleotide containing a 3′-modified base after annealing to a template. The oligodeoxynucleotides used for the various assays are listed in Fig.1. The lesion-containing oligodeoxynucleotides were either 5′-end-labeled by [γ-32P]ATP or 3′-end-labeled by α-32P-cordycepin before annealing to a complementary oligodeoxynucleotide to form a double-stranded substrate. The purification of endo VIII has been reported (31Jiang D. Hatahet Z. Blaisdell J. Melamede R.J. Wallace S.S. J. Bacteriol. 1997; 179: 3773-3782Crossref PubMed Google Scholar). Briefly, cell extracts were made from an E. coli nth xth double mutant, BW434. After nucleic acids were removed by precipitation with PEG 8000, the pooled crude extract was sequentially subjected to chromatography through S-Sepherose radial flow, Mono S, phenyl-Superose, and hydroxylapatite fast protein liquid chromatography, and an affinity column consisting of reduced AP-DNA-cellulose. Endo VIII was tracked by its ability to nick a Tg-containing double-stranded plasmid DNA in an alkaline fluorometric assay. The protein concentration of endo VIII fractions was measured following the instructions of the Bio-Rad Protein Assay kit using bovine serum albumin as a standard. In this paper, 1 unit of endo VIII, as well as endo III, endo IV, and Fpg, was defined as the amount of enzyme needed to cleave 1 pmol of double-stranded oligodeoxynucleotide containing an AP site per min at 37 °C. An apparently homogeneous endo VIII preparation, as determined by both SDS-polyacrylamide gel electrophoresis and protein N-terminal sequence analysis (31Jiang D. Hatahet Z. Blaisdell J. Melamede R.J. Wallace S.S. J. Bacteriol. 1997; 179: 3773-3782Crossref PubMed Google Scholar), was used in the biochemical characterization of the enzyme. 1–3 nmof a double-stranded (ds) substrate (Fig. 1) was incubated with 0.5 μl of the enzyme in 5 or 10 μl of buffer A (10 mmTris·HCl, pH 7.5, 50 or 100 mm NaCl) at 37 °C, and endo III, Fpg, and endo IV were used as controls (see details in figure legends). The reactions were stopped by adding 1 volume of the sample loading buffer (90% formamide, 1 mm EDTA, 0.1% xylene cyanol, and 0.1% bromphenol blue). Samples were loaded onto a 0.4-mm thick 10–20% denaturing polyacrylamide gel containing 8 murea with 19:1 of acrylamide to N,N′-methylenebisacrylamide in TBE buffer (90 mm Tris·HCl, 90 mm boric acid, pH 8.3, 2.5 mm EDTA). The gel was electrophoresed at 60–65 watts for 1.5–5 h, dried, and autoradiographed. For the DNA binding assay, 1 nm32P-labeled double-stranded DNA, redAP-, U-, or T-27-mer/23-mer (oligodeoxynucleotide 6 in Fig. 1) was incubated with increasing amounts of endo VIII in 10 μl of buffer A (containing 50 mm NaCl) plus 2 mm2-mercaptoethanol at room temperature (22 °C) for 10 min. For the binding competition assay, 10 nm labeled redAP-27-mer/23-mer was mixed with increasing amounts of the non-labeled redAP- or U-27-mer/23-mer before incubation with 20 nm endo VIII (see legend for Fig. 6). Immediately after adding glycerol to a 6% final concentration on ice, the samples were loaded onto a pre-run 0.75-mm thick 8% polyacrylamide gel with 80:1 of acrylamide to N,N′-methylenebisacrylamide and electrophoresed in TB buffer (90 mm Tris and 80 mm boric acid, pH 8.4–8.6) at 300 V/13 mA for 1 h 40 min at 4 °C. The gel was then dried and autoradiographed. The binding reactions of endo VIII to the redAP-27-mer/23-mer can be described by Equation 1. E8+redAPDNA⇌E8redAPDNAEquation 1 The equilibrium dissociation constant, K d, can be calculated from the concentrations of free endo VIII (E8), free DNA, and endo VIII·DNA complex at equilibrium from Equations 2 and 3. Kd=[E8][redAPDNA][E8redAPDNA]Equation 2 If f is the fraction of free DNA and [E8] is the concentration of free protein at equilibrium, the apparent Kd=[E8]f1−fEquation 3 If the total concentration of protein [E8]0 is close to that of the free protein concentration [E8], then the apparent K d = [E8]0 when 50% of the oligonucleotide is bound. To determine the apparent K d, the fraction of bound redAP-DNA was plotted against the total concentration of endo VIII. The Bio-Rad molecular imager package was used to quantitate the radioactive intensity of each band, and the image was analyzed using Molecular Analyst software. This assay was done following the method of Orchard and May (33Orchard K. May G.E. Nucleic Acids Res. 1993; 21: 3335-3336Crossref PubMed Scopus (66) Google Scholar). Briefly, after the DNA binding reaction described above, the sample was electrophoresed on 5–10% non-denaturing gels, side by side with the bromphenol blue marker and four standard molecular mass marker proteins (α-lactalbumin, chicken egg albumin, bovine albumin including the monomer and the dimer, and jack bean urease including the trimer and the hexamer). After electrophoresis, the gels were cut into two pieces, one with the lanes containing the binding reaction and the bromphenol blue marker and the other with the lanes containing the standard proteins and the bromphenol blue marker. The former was dried and autoradiographed; the latter was stained with Coomassie Blue and dried. The migration distance of each band was measured, and the relative mobility, R F, of the migration distances of the endo VIII·redAP·DNA complex and the standard proteins to that of the bromphenol blue were calculated and the data analyzed according to Ferguson (34Ferguson K.A. Metab. Clin. Exp. 1964; 13: 985-1002Abstract Full Text PDF PubMed Scopus (777) Google Scholar). The logarithms of the R F values were plotted against the percentages of the gel, and the slope of each plot was determined. The slopes for the standard proteins were plotted against their molecular masses, producing a molecular mass standard graph from which the molecular mass of endo VIII·redAP-DNA was deduced from its corresponding slope in the previous plot. Oligonucleotide 7 (Fig. 1) (2.5 nm) containing a reduced AP site at position 24 from the 5′ end was mixed with either endo VIII or Fpg protein, in the amounts indicated in the figure, in 20 mm Tris·HCl, pH 8.4, 2 mm MgCl2, 50 mm KCl, 20 μg/ml bovine serum albumin at room temperature. 0.1 unit of DNase I (Life Technologies, Inc., amplification grade) was then added and incubation continued at room temperature for 5 min. The reaction was stopped with 10 mm EDTA and 95% formamide, and the samples were analyzed by denaturing polyacrylamide gel electrophoresis. Endo VIII was previously shown to specifically cleave plasmid or phage DNA containing AP sites, urea, β-ureidoisobutyric acid, DHT, or Tg using a fluorometric nicking assay and to release Tg and DHT from Tg- and DHT-containing DNA by high pressure liquid chromatography analysis of the tritium-labeled products (30Melamede R.J. Hatahet Z. Kow Y.W. Ide H. Wallace S.S. Biochemistry. 1994; 33: 1255-1264Crossref PubMed Scopus (159) Google Scholar). These data suggested that during DNA repair, endo VIII specifically removes radiolysis products of thymine from DNA by cleavage of the N-glycosylic bond followed by cleavage of the DNA phosphodiester bond at the AP site created by the first cleavage reaction. Fig. 2 Acompares endo VIII with endo III cleavage of double-stranded oligodeoxynucleotides with a 5′-32P-labeled strand containing an AP site or Tg. Endo VIII (lanes 3 and 8) and endo III (lanes 2 and 5) cleaved the strand containing a Tg or AP site, and the 5′ products of the Tg- and AP-containing substrates generated by endo VIII (compare lanes 3 with 8) and endo III (compare lanes 2 with 5) had the same mobility. Similar cleavage of a DHT-containing strand by endo VIII was also observed (data not shown). In addition to thymine lesions, the specificity of endo VIII for two oxidative derivatives of cytosine was examined. Both 5-OHC and 5-OHU are stable oxidative products of cytosine (24Wagner J.R. Hum C.C. Ames B.N. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 3380-3384Crossref PubMed Scopus (238) Google Scholar) with potential mutagenic consequences (27Purmal A.A. Kow Y.W. Wallace S.S. Nucleic Acids Res. 1994; 22: 72-78Crossref PubMed Scopus (152) Google Scholar, 29Feig D.I. Sowers L.C. Loeb L.A. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6609-6613Crossref PubMed Scopus (191) Google Scholar). Fig. 2 B shows that endo VIII cleaved the strand containing 5-OHC (lanes 4 and 7) and 5-OHU (lanes 10 and 13) opposite both G and A as does endo III (lanes 3, 6, 9, and 12). Endo VIII also exhibited cleavage activity on substrates containing two other oxidative products of cytosine, uracil glycol, and dihydrouracil.2 We had previously reported that the nick produced in Tg-containing DNA by endo VIII was not a substrate for DNA polymerase I Klenow fragment suggesting that, like endo III, it acted as a lyase via a β-elimination (30Melamede R.J. Hatahet Z. Kow Y.W. Ide H. Wallace S.S. Biochemistry. 1994; 33: 1255-1264Crossref PubMed Scopus (159) Google Scholar). To examine further the mechanism of phosphodiester backbone cleavage by endo VIII, a double-stranded oligodeoxynucleotide containing an AP site was used as a substrate, and the mobilities of the 3′ and 5′ cleavage products produced by endo VIII were analyzed in a denaturing polyacrylamide gel. Endo IV, endo III, and Fpg were used as controls. As shown in Fig.3 A, the mobility of the 3′-labeled product from endo VIII incision of the AP-containing strand (lane 4) was identical to that of the 3′ product from endo III (lane 2) or Fpg (lane 3) incision, as well as that of the 3′-labeled 23-mer marker containing a 5′ phosphate (lanes 5). These data suggest that endo VIII nicked the phosphodiester backbone 3′ to the AP site, leaving a phosphate group at the 5′ side of the nick as do endo III and Fpg. The existence of the 5′ end phosphate group was confirmed by the results presented in Fig.4. After alkaline phosphatase treatment, the 3′ product from endo VIII incision at the AP site migrated as fast as the dephosphorylated 23-mer marker (compare lanes 10 with 11 in Fig. 4).Figure 4Removal of the 5′-terminal deoxyribose phosphate (dRp) from an endo IV-preincised substrate by endo VIII.ds AP-36-mer/42-mer (oligo 1) with a 3′-end-32P-labeled AP-containing strand was used as a substrate. 1 nm DNA was incubated with an excess of endo IV (2.4 × 10−2 unit, lane 4) in 9.5 μl buffer A at 37 °C for 20 min. The samples were then incubated with Fpg (∼1.5 × 10−2 unit) (lane 5) and increasing amounts of endo VIII (lane 6, ∼1.5 × 10−3;lane 7, ∼1.5 × 10−2; lane 8, ∼1.5 × 10−1 unit). The reaction mixtures were treated with 1.5–2 m NaBH4, neutralized with 100 mm Tris, pH 7.5, and filtered through NAP-5 column. A sample of the reaction with endo IV and then with ∼1.5 × 10−2 unit of endo VIII was further incubated with shrimp alkaline phosphatase (lane 9). As controls, the original AP-containing substrate was also incubated with ∼1.5 × 10−1 units of endo VIII (lane 10), followed by incubation with shrimp alkaline phosphatase (lane 11).Lanes 1 and 2 contained the p23-mer size marker (oligo 4) with and without incubation with phosphatase. The reaction samples were electrophoresed in a 20% denaturing polyacrylamide gel. AP* indicates that the AP-containing strand of the substrate was 3′-end-32P-labeled (61Aspinwall R. Rothwell D.G. Roldan-Arjona T. Anselmino C. Ward C.J. Cheadle J.P. Sampson J.R. Lindahl T. Harris P.C. Hickson I.D. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 109-114Crossref PubMed Scopus (252) Google Scholar).View Large Image Figure ViewerDownload Hi-res image Download (PPT) Endo III cleaves the DNA backbone by β-elimination, leaving an α,β-unsaturated aldehyde at the 3′ side of the nick (35Kim J. Linn S. Nucleic Acids Res. 1988; 16: 1135-1141Crossref PubMed Scopus (96) Google Scholar). Consistent with this mechanism, the 5′-labeled product from endo III incision migrated more slowly than the 5′ cleavage product from endo IV incision which contains a 3′ hydroxyl group produced by hydrolysis 5′ to the AP site (36Bailly V. Verly W.G. Biochem. J. 1989; 262: 581-589Crossref PubMed Scopus (132) Google Scholar) (compare lanes 9 with 7). The 5′ product generated by endo IV was the same as the 5′-labeled 12-mer marker (lane 16). As has been previously reported (37Mazumder A. Gerlt J.A. Absalon M.J. Stubbe J. Cunningham R.P. Withka J. Bolton P.H. Biochemistry. 1991; 30: 1119-1126Crossref PubMed Scopus (171) Google Scholar), the 5′ endo III cleavage product was a doublet (lane 7), the second band of which is thought to be the result of interaction with Tris in the buffer. Fpg cleaves the DNA backbone at the β and then the δ position through β,δ-elimination, leaving a phosphate at the 3′ end of the 5′ cleavage product (38Bailly V. Verly W.G. O'Connor T. Laval J. Biochem. J. 1989; 262: 581-589Crossref PubMed Scopus (141) Google Scholar). As expected, the mobility of the 5′-labeled product generated by Fpg incision was faster than that produced by endo IV incision because of the extra negative charge of the 5′ phosphate group (compare lanes 8 with 9 and 16). The mobility of the 5′ endo VIII cleavage product was the same as the β,δ-elimination product of Fpg (compare lanes 10 with 8). When the endo VIII reaction was followed by incubation with T4 polynucleotide kinase that contains a 3′-phosphatase activity, the mobility of the δ-elimination-like product produced by endo VIII now migrated as the 5′ product from endo IV incision which contains a 3′ hydroxyl group on the 5′ product (compare lanes 11 with 9). These data suggest that the 5′ product from endo VIII cleavage contained a 3′ phosphate resulting from a δ-elimination. When E. coli DNA polymerase I (Pol I) was added without or with four dNTPs following the endo VIII reaction, the δ-elimination-like product could not be cleaved by the 3′ to 5′ exonuclease activity of Pol I nor could it be elongated by the polymerase activity of Pol I (
Referência(s)