Human Polymorphic Variants of the NEIL1 DNA Glycosylase

Abstract

In mammalian cells, the repair of DNA bases that have been damaged by reactive oxygen species is primarily initiated by a series of DNA glycosylases that include OGG1, NTH1, NEIL1, and NEIL2. To explore the functional significance of NEIL1, we recently reported that neil1 knock-out and heterozygotic mice develop the majority of symptoms of metabolic syndrome (Vartanian, V., Lowell, B., Minko, I. G., Wood, T. G., Ceci, J. D., George, S., Ballinger, S. W., Corless, C. L., McCullough, A. K., and Lloyd, R. S. (2006) Proc. Natl. Acad. Sci. U. S. A. 103, 1864-1869). To determine whether this phenotype could be causally related to human disease susceptibility, we have characterized four polymorphic variants of human NEIL1. Although three of the variants (S82C, G83D, and D252N) retained near wild type levels of nicking activity on abasic (AP) site-containing DNA, G83D did not catalyze the wild type β,δ-elimination reaction but primarily yielded the β-elimination product. The AP nicking activity of the C136R variant was significantly reduced. Glycosylase nicking activities were measured on both thymine glycol-containing oligonucleotides and γ-irradiated genomic DNA using gas chromatography/mass spectrometry. Two of the polymorphic variants (S82C and D252N) showed near wild type enzyme specificity and kinetics, whereas G83D was devoid of glycosylase activity. Although insufficient quantities of C136R could be obtained to carry out gas chromatography/mass spectrometry analyses, this variant was also devoid of the ability to incise thymine glycol-containing oligonucleotide, suggesting that it may also be glycosylase-deficient. Extrapolation of these data suggests that individuals who are heterozygous for these inactive variant neil1 alleles may be at increased risk for metabolic syndrome. In mammalian cells, the repair of DNA bases that have been damaged by reactive oxygen species is primarily initiated by a series of DNA glycosylases that include OGG1, NTH1, NEIL1, and NEIL2. To explore the functional significance of NEIL1, we recently reported that neil1 knock-out and heterozygotic mice develop the majority of symptoms of metabolic syndrome (Vartanian, V., Lowell, B., Minko, I. G., Wood, T. G., Ceci, J. D., George, S., Ballinger, S. W., Corless, C. L., McCullough, A. K., and Lloyd, R. S. (2006) Proc. Natl. Acad. Sci. U. S. A. 103, 1864-1869). To determine whether this phenotype could be causally related to human disease susceptibility, we have characterized four polymorphic variants of human NEIL1. Although three of the variants (S82C, G83D, and D252N) retained near wild type levels of nicking activity on abasic (AP) site-containing DNA, G83D did not catalyze the wild type β,δ-elimination reaction but primarily yielded the β-elimination product. The AP nicking activity of the C136R variant was significantly reduced. Glycosylase nicking activities were measured on both thymine glycol-containing oligonucleotides and γ-irradiated genomic DNA using gas chromatography/mass spectrometry. Two of the polymorphic variants (S82C and D252N) showed near wild type enzyme specificity and kinetics, whereas G83D was devoid of glycosylase activity. Although insufficient quantities of C136R could be obtained to carry out gas chromatography/mass spectrometry analyses, this variant was also devoid of the ability to incise thymine glycol-containing oligonucleotide, suggesting that it may also be glycosylase-deficient. Extrapolation of these data suggests that individuals who are heterozygous for these inactive variant neil1 alleles may be at increased risk for metabolic syndrome. A major source of DNA lesions in eukaryotic cells is the interaction of reactive oxygen species with DNA constituents. DNA bases are particularly susceptible to reactive oxygen species, with the major DNA damages being, among others, 8-hydroxyguanine (8-OH-Gua), 3The abbreviations and trivial names used are: 8-OH-Gua, 8-hydroxyguanine; FapyGua, 2,6-diamino-4-hydroxy-5-formamido-pyrimidine; FapyAde, 4,6-diamino-5-formamido-pyrimidine; AP, abasic; GC/MS, gas chromatography/mass spectrometry. 8-hydroxyadenine, 2,6-diamino-4-hydroxy-5-formamidopyrimidine (FapyGua), and 4,6-diamino-5-formamido pyrimidine (FapyAde) (1Evans M.D. Dizdaroglu M. Cooke M.S. Mutat. Res. 2004; 567: 1-61Crossref PubMed Scopus (1057) Google Scholar, 2Dizdaroglu M. Mutat. Res. 2003; 531: 109-126Crossref PubMed Scopus (75) Google Scholar). To reverse the potentially deleterious effects of oxidatively induced DNA base lesions, cells primarily utilize the base excision repair path-way to restore the DNA to its original state. This pathway is initiated by lesion-specific DNA glycosylases that hydrolyze the bond attaching the damaged base to the deoxyribose, and many of these enzymes also possess an activity that catalyzes a β-or β,δ-elimination reaction at the newly formed abasic (AP) site. These incision intermediates are further processed by an AP endonuclease to yield a free 3′-OH that serves as a primer for repair synthesis and ligation. To initiate repair of oxidatively induced DNA base lesions, mammalian cells primarily use the products of the ogg1, nth1, neil1, and neil2 genes (reviewed in Ref. 2Dizdaroglu M. Mutat. Res. 2003; 531: 109-126Crossref PubMed Scopus (75) Google Scholar). Human and mouse NEIL1 proteins have been shown to possess a strong substrate preference for FapyAde and FapyGua, with no specificity for 8-OH-Gua when genomic DNA containing multiple lesions was used as a substrate (3Hazra T.K. Izumi T. Boldogh I. Imhoff B. Kow Y.W. Jaruga P. Dizdaroglu M. Mitra S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 3523-3528Crossref PubMed Scopus (440) Google Scholar, 4Jaruga P. Birincioglu M. Rosenquist T.A. Dizdaroglu M. Biochemistry. 2004; 43: 15909-15914Crossref PubMed Scopus (80) Google Scholar). However, in experiments using oligonucleotides with a single lesion, NEIL1 has exhibited some specificity for methyl-FapyGua, urea, (5R)- and (5S)-thymine glycols (opposite Thy, Cyt, and Gua and to a much lesser extent Ade), 5,6-dihydrouracil (and tandem 5,6-dihydrouracil), 5-formyluracil, 5-(hydroxymethyl)uracil, 5-hydroxyuracil, 5,6-dihydrothymine, mismatched bases uracil:cytosine and thymine:cytosine, and 8-OH-Gua opposite Gua or Thy, and AP sites (reviewed in Ref. 2Dizdaroglu M. Mutat. Res. 2003; 531: 109-126Crossref PubMed Scopus (75) Google Scholar). In contrast to all other glycosylases, except uracil DNA glycosylase, the activity of NEIL1 is stimulated when lesions are present in single-stranded bubble structures that are hypothesized to resemble stalled transcription and replication forks (5Dou H. Mitra S. Hazra T.K. J. Biol. Chem. 2003; 278: 49679-49684Abstract Full Text Full Text PDF PubMed Scopus (297) Google Scholar). Additionally, NEIL1 has been shown to efficiently cleave DNA-containing oxidatively induced base lesions near single-strand break sites, whereas NTH1 and OGG1 are very inefficient in catalyzing these reactions (6Parsons J.L. Zharkov D.O. Dianov G.L. Nucleic Acids Res. 2005; 33: 4849-4856Crossref PubMed Scopus (61) Google Scholar). In analogy to its prokaryotic homolog, NEIL1 utilizes its N-terminal proline to catalyze sequential glycosylase, β- and δ-eliminations (3Hazra T.K. Izumi T. Boldogh I. Imhoff B. Kow Y.W. Jaruga P. Dizdaroglu M. Mitra S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 3523-3528Crossref PubMed Scopus (440) Google Scholar, 7Bandaru V. Sunkara S. Wallace S.S. Bond J.P. DNA Repair (Amst.). 2002; 1: 517-529Crossref PubMed Scopus (286) Google Scholar), and mutations at Pro-2 compromise enzyme activity. The structure of human NEIL1 was solved by x-ray crystallography (8Doublie S. Bandaru V. Bond J.P. Wallace S.S. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 10284-10289Crossref PubMed Scopus (121) Google Scholar). This structure not only confirmed the identity of residues involved in the catalytic mechanism, but also showed that, although NEIL1 does not contain a zinc-finger motif similar to its Escherichia coli counterparts, endonuclease VIII and formamidopyrimidine DNA glycosylase, it maintains a similar overall fold termed a “zinc-less finger” motif. The potential biological importance of NEIL1 has recently been demonstrated in studies showing a correlation of inactivating mutations in neil1 with human gastric cancer (9Shinmura K. Tao H. Goto M. Igarashi H. Taniguchi T. Maekawa M. Takezaki T. Sugimura H. Carcinogenesis. 2004; 25: 2311-2317Crossref PubMed Scopus (86) Google Scholar). In addition, RNA interference knockdown experiments in which an ∼80% reduction in the mRNA levels of neil1 was achieved significantly sensitized cells to the killing effects of ionizing radiation (10Rosenquist T.A. Zaika E. Fernandes A.S. Zharkov D.O. Miller H. Grollman A.P. DNA Repair (Amst.). 2003; 2: 581-591Crossref PubMed Scopus (163) Google Scholar). These data may indicate a critical role for NEIL1 in long term maintenance of genetic integrity. To determine what role NEIL1 might play in the overall repair of DNA containing oxidatively induced lesions, we have constructed mice in which the neil1 gene has been knocked out. In the absence of exogenous oxidative stress, neil1 knock-out (neil1-/-) and heterozygotic (neil1+/-) mice develop severe obesity, dyslipidemia, and fatty liver disease, and also have a tendency to develop hyperinsulinemia (11Vartanian V. Lowell B. Minko I.G. Wood T.G. Ceci J.D. George S. Ballinger S.W. Corless C.L. McCullough A.K. Lloyd R.S. Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 1864-1869Crossref PubMed Scopus (194) Google Scholar). In humans, this combination of clinical manifestations, including hypertension, is known as the metabolic syndrome and is estimated to affect more than 40 million people in the United States. These data suggest an important role for NEIL1 in the prevention of the diseases associated with the metabolic syndrome. In the present study, we report on the enzymatic characterization of four polymorphic variants of human NEIL1 and on their specificities and excision kinetics for removal of oxidatively induced base lesions neil1 was by to using and The DNA was with and and that been were in the at and the wild type was by using with were confirmed by and and used in in a of in E. were and in with and to an of at of was induced with at for were and the was on in containing a and for on were to and by were for at and proteins and the proteins the with to the for were using the and on with AP AP were as A.K. A. P. Lloyd R.S. Biochemistry. PubMed Scopus Google Scholar). a containing a uracil was with and to the containing an opposite the The was with uracil DNA glycosylase for to the AP site. was with for at in and analyses, of was with for the reaction or and were by the of a of and at for were by of the substrate and and and by DNA for glycosylase activity on containing a lesion, nicking were out using an containing a single thymine of of were that, when to the were or a mismatched bubble the containing the thymine The damaged was and with of the oligonucleotides The DNA was with of wild type and polymorphic variants of NEIL1 for at in and were by and by of DNA and of of DNA and their to ionizing radiation in a were as M. C. H. J. Biochemistry. PubMed Scopus Google Scholar). were as P. R. McCullough A.K. H. Dizdaroglu M. Lloyd R.S. 2002; PubMed Scopus Google Scholar). the of excision kinetics, DNA were γ-irradiated at and of of DNA bases as were to of The were in a Two of these with three were of the was in of the and with wild type NEIL1 at for The of the FapyAde and were and of were to the to the reaction and The were at for the were at for at DNA and were was in a were to for The other of was used to determine the levels of DNA bases in The were with of in and for at The were in and for of or of were and by using a gas to the P. Birincioglu M. H. Dizdaroglu M. Biochemistry. 2002; PubMed Scopus Google Scholar). it has been estimated that single in the human M.D. M. J.D. P. M. J. G.L. C. S. J. M. C. M. R. A. I. M. A. S. S. S. C. K. K. J. E. K. V. R. M. I. R. K. V. P. K. C. P. J. A.K. B. S. B. J. A. J. M. R. C. C. A. J. M. H. S. S. S. C. A. T. H. A. H. M. I. K. A. A. S. R. S. K. S. A. B. J. C. C. S. J. T. C. J. S. A. S. T. I. M. B. K. C. E. V. H. M. R. B. R. C. M. E. R. E. R. S. C. J. S. M. S. E. K. J. K. R. B. A. H. B. T. A. K. A. S. V. S. R. R. B. S. A. J. M. J. P. M. C. A. M. M. S. C. H. S. K. A. M. K. B. M. J. S. J. C. J. J. C. A. M. J. J. S. M. T. M. S. J. M. R. J. M. T. A. T. R. E. M. M. M. A. A. PubMed Scopus Google Scholar), it is that many DNA repair genes contain as alleles cancer risk Mutat. Res. PubMed Scopus Google Scholar, Mutat. Res. 2003; PubMed Scopus Google Scholar, T. 2004; PubMed Scopus Google Scholar). the many DNA repair gene that have been some have been shown to affect T. 2004; PubMed Scopus Google Scholar, R. Carcinogenesis. PubMed Scopus Google Scholar, T. K. M. M. Sugimura H. T. H. J. PubMed Scopus Google Scholar, K. Nucleic Acids Res. PubMed Scopus Google and be associated with an increased of cancer Mutat. Res. 2003; PubMed Scopus Google Scholar, J.D. 2002; Google Scholar). single neil1 have been The of at the of has the neil1 gene in a of individuals who are of the United in the These variants that in the of NEIL1 (S82C, G83D, and In a an was near the of these studies and has not been characterized for neil1 single that in an are variants at this The of these in the NEIL1 structure (8Doublie S. Bandaru V. Bond J.P. Wallace S.S. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 10284-10289Crossref PubMed Scopus (121) Google are shown in the eukaryotic of NEIL1 the identity of the residues at the sites of these polymorphic variants is a and of the residues and that are to this Human NEIL1 is and to the E. coli endonuclease VIII and formamidopyrimidine DNA glycosylase, (3Hazra T.K. Izumi T. Boldogh I. Imhoff B. Kow Y.W. Jaruga P. Dizdaroglu M. Mitra S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 3523-3528Crossref PubMed Scopus (440) Google Scholar, 7Bandaru V. Sunkara S. Wallace S.S. Bond J.P. DNA Repair (Amst.). 2002; 1: 517-529Crossref PubMed Scopus (286) Google Scholar). of human NEIL1 with these prokaryotic enzymes that, among the variant sites, only is Additionally, to this study, it is of that of the and G83D, in to the that has been in of the (8Doublie S. Bandaru V. Bond J.P. Wallace S.S. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 10284-10289Crossref PubMed Scopus (121) Google Scholar). The C136R variant to in a the major of the and could affect the overall of the enzyme and of the and of human neil1 gene was the that and enzyme sites were and to the neil1 and using the DNA of the wild type was used to the mutations and in the NEIL1 G83D, and genes variants of NEIL1 were confirmed by DNA the wild type and variant genes of neil1 were E. coli and proteins The enzymes were to with the of This enzyme could be at levels to that of the wild type enzyme and was E. as by no in the of the C136R variant in E. coli the measured in the was for in cells as no enzyme was in the and of cells no Additionally, and in contrast to wild type NEIL1 and the other only a very of the C136R to the However, when the proteins in the were to the C136R was efficiently These data suggest that the C136R may the of this the except of C136R that to the could be using oligonucleotides containing an AP or a thymine but could not be for the more on the that variant mutations might have on the sequential DNA glycosylase, and we to for the effects on the β,δ-elimination reaction using AP site-containing This reaction on the incision of AP site-containing DNA using the of the N-terminal Pro-2 as the (3Hazra T.K. Izumi T. Boldogh I. Imhoff B. Kow Y.W. Jaruga P. Dizdaroglu M. Mitra S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 3523-3528Crossref PubMed Scopus (440) Google Scholar, 7Bandaru V. Sunkara S. Wallace S.S. Bond J.P. DNA Repair (Amst.). 2002; 1: 517-529Crossref PubMed Scopus (286) Google Scholar). of Pro-2 or in the of catalytic activity (3Hazra T.K. Izumi T. Boldogh I. Imhoff B. Kow Y.W. Jaruga P. Dizdaroglu M. Mitra S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 3523-3528Crossref PubMed Scopus (440) Google Scholar, 7Bandaru V. Sunkara S. Wallace S.S. Bond J.P. DNA Repair (Amst.). 2002; 1: 517-529Crossref PubMed Scopus (286) Google Scholar). data the incision activity of wild G83D, and C136R on DNA containing an AP in the in which the AP was by the of uracil DNA glycosylase on a DNA was with wild type G83D and C136R by reaction with to AP and similar the However, although G83D the abasic DNA, the was that of the β-elimination a similar to that for AP that utilize an of or a of an N-terminal (reviewed in Ref. A.K. Lloyd R.S. PubMed Scopus Google Scholar). activity of C136R was significantly to of the other and similar to G83D, showed an of the β,δ-elimination However, as this variant to and it be whether the in activity was to or an in catalytic the abasic DNA shown in was in for an of and with these data that no of nicking in and the products in are to be the of the catalytic activities of the the DNA was not with to analyses, the in a near to the β-elimination not of AP incision were with the wild G83D, and NEIL1 proteins a nicking in which it is that the wild and NEIL1 proteins all the whereas the G83D NEIL1 yielded the β-elimination product. both data and only in the catalytic of these variants to the wild type NEIL1. of NEIL1 and on major biological role of NEIL1 is to be and of repair at oxidatively damaged have shown that, using oligonucleotides containing thymine lesions, both mouse and human NEIL1 catalyze and To the NEIL1 polymorphic variants for this wild type and variant were using a thymine at the in which the was a bubble or shown in the wild type and variant activity on both and and the nicking reactions similar to that of wild type NEIL1 and In and although G83D DNA with it was devoid of glycosylase nicking and to the of AP the C136R variant was to initiate incision and To that these proteins were not out the glycosylase reaction the β- and the DNA reaction products were with to cleave AP These data no both that the G83D and C136R variants were inactive for bond and that the wild and carry out the and reactions not of Human NEIL1 and of bases DNA by NEIL1 and its polymorphic variants was further using DNA that been to ionizing radiation in and DNA were at a of and with wild type or NEIL1 polymorphic or or no enzyme as a of enzyme or and of DNA were and by with excision of FapyAde and DNA by wild and proteins was an the of excision by wild type NEIL1 of FapyAde and on the enzyme Additionally, the of was measured using and with the excision with it a not excision of was also but it was significantly when with the excision of FapyAde and not with the of incision of thymine glycol-containing G83D exhibited no activity for of these In all other bases including 8-OH-Gua were not significantly The enzymes no activity. determine the for the wild type and polymorphic were DNA were at and to levels of bases to the of excision on substrate of bases in were used for the of the of FapyAde and in DNA with the enzyme were and kinetics A to and Scholar). for the excision of FapyAde and by wild type NEIL1 are in The of the and were achieved using a with the of the The of the excision of FapyAde and are in The excision of DNA did not an of its excision In the of all three the specificity for excision of FapyAde was significantly than that for excision of FapyGua, a preference of these enzymes for the the for excision of FapyAde and by human NEIL1, and DNA γ-irradiated in with type the The enzyme was of FapyAde and were and the The enzyme was of FapyAde and were and in a Human to both and exogenous of reactive oxygen species are hypothesized to be in the of a of diseases M. J. M. J. J. Biol. PubMed Scopus Google Scholar, Mutat. Res. PubMed Scopus Google Scholar), including but not to hypertension, obesity, fatty liver disease, and However, the by which these reactive that in disease are not The products reaction of reactive oxygen species with and are with DNA base including at M. Mutat. Res. 2005; PubMed Scopus Google Scholar). repair of these lesions are critical to the maintenance of the of these To the of damaged bases to their original possess a base excision repair In humans, the repair of oxidatively induced lesions be initiated by at four NEIL1, OGG1, and have the of inactivating or inactivating the gene NEIL1 in in which symptoms with metabolic syndrome (11Vartanian V. Lowell B. Minko I.G. Wood T.G. Ceci J.D. George S. Ballinger S.W. Corless C.L. McCullough A.K. Lloyd R.S. Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 1864-1869Crossref PubMed Scopus (194) Google Scholar). The that neil1 also develop disease to whether are human polymorphic variants of NEIL1 that possess catalytic to initiate a was the that an reduction in the neil1 in a in ionizing radiation (10Rosenquist T.A. Zaika E. Fernandes A.S. Zharkov D.O. Miller H. Grollman A.P. DNA Repair (Amst.). 2003; 2: 581-591Crossref PubMed Scopus (163) Google Scholar). of these data to may suggest that similar in the of NEIL1 could individuals to a of diseases that are associated with the metabolic syndrome. this the present that human wild type NEIL1 and its polymorphic variants and possess an activity to FapyAde and DNA containing multiple three enzymes a preference for FapyAde FapyGua, with some activity FapyAde and are among the major and products in DNA (reviewed in Ref. M.D. Dizdaroglu M. Cooke M.S. Mutat. Res. 2004; 567: 1-61Crossref PubMed Scopus (1057) Google Scholar). these have been shown to be lesions, with more than 8-OH-Gua in mammalian cells K. S. M. A.K. Nucleic Acids Res. 2006; PubMed Scopus Google Scholar, J. Chem. 2002; PubMed Scopus (86) Google Scholar), to their importance in the biological effects of oxidative DNA This that the G83D and potentially the C136R variants of human NEIL1 were devoid of DNA glycosylase activity. In the of the G83D we that it was it was and AP that activity was a wild type activity to catalyzing a β-elimination These data suggest that to the AP the of this variant was not but that the of the base and the of the to the This suggest that the of the G83D variant was more than the wild type or that the of an at the of the the of the of the The for the of a to a β-elimination enzyme a single has been reported for the that DNA-containing 8-OH-Gua lesions by a similar reaction V. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). of the structure of human NEIL1 a for the of glycosylase activity for the G83D of the structures of glycosylases sites are formed at the that these are by not The of this to a to an may be particularly deleterious to the activity of the whereas other at might not in of glycosylase activity. This is by the that an did not compromise the catalytic of this variant to these data suggest that, are other individuals the G83D polymorphic are to with levels of NEIL1. of the of catalytic activity for the C136R variant is more to the of of the to all of the enzyme of the enzyme could be and for activity In the enzyme was This suggests that C136R of the the The activity suggest that this variant is in both the glycosylase and AP the that of activity may be to be out. a in both and enzymatic activity also be on the C136R to be in a that the and it may be to be critical in of the enzyme to in the and of the that might be to the be to affect However, we it that in this and may not the a human In to the polymorphic variants that were in this study, have variants in the neil1 gene (9Shinmura K. Tao H. Goto M. Igarashi H. Taniguchi T. Maekawa M. Takezaki T. Sugimura H. Carcinogenesis. 2004; 25: 2311-2317Crossref PubMed Scopus (86) Google Scholar, P. T. J. S. I. R.S. 2006; PubMed Scopus Google Scholar). (9Shinmura K. Tao H. Goto M. Igarashi H. Taniguchi T. Maekawa M. Takezaki T. Sugimura H. Carcinogenesis. 2004; 25: 2311-2317Crossref PubMed Scopus (86) Google neil1 variants associated with gastric were a gastric cancer and The and the variants have catalytic whereas the of has activity similar to a of The gastric in a variant that of NEIL1, whereas the other was an in a of cancer for nth1, neil1, and no neil1 mutations were P. T. J. S. I. R.S. 2006; PubMed Scopus Google Scholar). This only three mutations in and The only variant in neil1 was at the of the site. is whether this variant has functional data suggest that at and of the four polymorphic variants of NEIL1 have or data obtained in the mouse and in be to the human we that individuals this may be at a risk of to metabolic we that in NEIL1 the of oxidative that is to a to disease