The Repair of the Tobacco Specific Nitrosamine Derived Adduct O6

MGMT—a critical DNA repair gene target for chemotherapy resistance ... Mgmt deficiency alters the in vivo mutational spectrum of tissues exposed to ...
0 downloads 0 Views 231KB Size
424

Chem. Res. Toxicol. 2004, 17, 424-434

The Repair of the Tobacco Specific Nitrosamine Derived Adduct O6-[4-Oxo-4-(3-pyridyl)butyl]guanine by O6-Alkylguanine-DNA Alkyltransferase Variants Rene´e S. Mijal,† Nicole M. Thomson,† Nancy L. Fleischer,† Gary T. Pauly,‡ Robert C. Moschel,‡ Sreenivas Kanugula,§ Qingming Fang,§ Anthony E. Pegg,§ and Lisa A. Peterson*,† Division of Environmental and Occupational Health and Cancer Center, University of Minnesota, Minneapolis, Minnesota 55455, Laboratory of Comparative Carcinogenesis, National Cancer Institute at Frederick, Frederick, Maryland 21702, and Department of Cellular and Molecular Physiology, Milton S. Hershey Medical Center, Pennsylvania State University College of Medicine, Hershey, Pennsylvania 17033 Received November 21, 2003

The tobacco specific nitrosamine, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), a potent pulmonary carcinogen, both methylates and pyridyloxobutylates DNA. Both reaction pathways generate promutagenic O6-alkylguanine adducts. These adducts, O6-methylguanine (O6-mG) and O6-[4-oxo-4-(3-pyridyl)butyl]guanine (O6-pobG), are repaired by O6-alkylguanineDNA alkyltransferase (AGT). In this report, we demonstrate that pyridyloxobutyl DNA adducts are repaired by AGT in a reaction that results in pyridyloxobutyl transfer to the active site cysteine. Because minor changes within the binding pocket of AGT can alter the ability of this protein to repair bulky O6-alkylguanine adducts relative to O6-mG, we explored the ability of AGTs from different species as well as several human AGT variants and mutants to discriminate between O6-mG or O6-pobG adducts. We incubated proteins with equal molar amounts of oligodeoxynucleotides containing site specifically incorporated O6-mG or O6-pobG and measured repair. Bacterial AGTs poorly repaired O6-pobG. Mouse and rat AGT repaired both adducts at comparable rates. Wild-type human AGT, variant I143V/K178R, and mutant N157H repaired O6-mG approximately twice as fast as O6-pobG. Human variant G160R and mutants P140K, Y158H, G156A, and E166G did not repair O6-pobG until all of the O6-mG was removed. To understand the role of adduct structure on relative repair rates, the competition experiments were repeated with two other bulky O6-alkylguanine adducts, O6butylguanine (O6-buG) and O6-benzylguanine (O6-bzG). The proteins displayed similar repair preference of O6-mG relative to O6-buG as observed with O6-pobG. In contrast, all of the mammalian proteins, except the mutant P140K, preferentially repaired O6-bzG. These studies indicate that the rate of repair of O6-pobG is highly dependent on protein structure. Inefficient repair of O6-pobG by bacterial AGT explains the high mutagenic activity of this adduct in bacterial systems. In addition, differences observed in the repair of this adduct by mammalian proteins may translate into differences in sensitivity to the mutagenic and carcinogenic effects of NNK or other pyridyloxobutylating nitrosamines.

Introduction The tobacco specific nitrosamine, NNK,1 is carcinogenic in laboratory animals (1). NNK is primarily a lung carcinogen, which also induces liver and nasal tumors (2-4). It is present in tobacco products and smoke (1) and is a likely human carcinogen (5). NNK is metabolized to either a methylating agent or a pyridyloxobutylating * To whom correspondence should be addressed. Tel: 612-626-0164. Fax: 612-626-5135. E-mail: [email protected]. † University of Minnesota. ‡ National Cancer Institute at Frederick. § Pennsylvania State University College of Medicine. 1 Abbreviations: AGT, O6-alkylguanine-DNA alkyltransferase; hAGT, recombinant human wild-type AGT; mAGT, recombinant mouse AGT; rAGT, recombinant rat AGT; DTT, dithiothreitol; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NNKOAc, 4-(acetoxymethylnitrosamino)-1-(3-pyridyl)-1-butanone; O6-buG, O6-butylguanine; O6-bzG; O6-benzylguanine; O6-mG, O6-methylguanine; O6-pobG, O6-[4-oxo-4(3-pyridyl)butyl]guanine.

agent (Scheme 1). While the contribution of O6-mG to the carcinogenic properties of the methylation pathway of NNK has been investigated (6-11), the role of specific adducts in the mutagenic and carcinogenic properties of the pyridyloxobutylation pathway has not been fully explored. This pathway generates a variety of adducts, one of which is O6-pobG (12-14). This adduct is mutagenic, particularly in the absence of repair (15). O6-Alkylguanine adducts are repaired by the repair protein AGT in a reaction that involves transfer of the alkyl group from the O6-position of guanine to a cysteinyl residue at the active site of the protein (16). This transfer reaction renders the protein inactive. The alkylated protein is then rapidly degraded (17-19). Therefore, constitutive levels of AGT determine the initial repair capacity of a cell (16). Previous studies from our laboratory indicated that pyridyloxobutyl DNA adducts interfered with repair of

10.1021/tx0342417 CCC: $27.50 © 2004 American Chemical Society Published on Web 02/12/2004

Repair of O6-pobG by AGT Variants Scheme 1. Proposed Activation Pathway for NNK and NNKOAc

Chem. Res. Toxicol., Vol. 17, No. 3, 2004 425

interact with the active site to influence reaction rates, relative repair rates for O6-pobG were compared to rates calculated for O6-buG, a structurally similar adduct, and O6-bzG, a well-studied, bulky O6-alkyl adduct whose repair is known to be strongly influenced by binding pocket structure (Scheme 2) (25, 26).

Experimental Procedures

O6-mG by AGT (20, 21). This interference was consistent with the presence of AGT substrate adducts in the pyridyloxobutylated DNA. Subsequent studies led to the identification of O6-pobG, which is a substrate for AGT both in vitro (12, 13) and in vivo (13). Pyridyloxobutylated DNA may also contain other AGT reactive adducts since model pyridyloxobutylating agents, such as NNKOAc (Scheme 1), react with nucleophiles to generate both open chain and cyclic adducts (22, 23). Somewhat surprisingly, the bulky O6-pobG is efficiently repaired by human, mouse, and rat AGT (12, 13). In addition, the AGT reactive adducts in pyridyloxobutylated DNA compete quite well with O6-mG for reaction with AGT (20, 21). These observations were unanticipated since earlier studies with linear alkyl adducts suggested that repair rates declined with increased size of the O6-alkyl group (24). More recent studies have demonstrated that AGT amino acid structure affects substrate specificity since mammalian AGTs efficiently repair larger adducts (16). How adduct size and AGT binding site structure govern the repair of O6pobG is not well-understood. To investigate the influence of amino acid sequence on the repair of O6-pobG by AGT, we measured the relative rates of O6-pobG vs O6-mG repair by various AGT orthologs (Figure 1). Repair preferences of AGTs from organisms employed to model NNK carcinogenesis (mouse, rat, and Escherichia coli) were compared to the human AGT repair preference. The relative repair of O6-mG and O6-pobG by wild-type hAGT was compared to repair by two human AGT variants, G160R and I143V/K178R. A series of mutant human AGTs were also tested to evaluate the role of particular amino acid substitutions in adduct selection. To further understand how adducts

Materials. [5-3H]NNKOAc, O6-bzG, and [3H]methylated and [5-3H]pyridyloxobutylated calf thymus DNA were prepared as previously described (12, 20, 27, 28). Rat liver AGT was prepared as previously described (21). Recombinant his-tagged AGT proteins from human, mouse, rat, and E. coli and hAGT mutants and variants were expressed and purified as previously described (29-31). Sequencing grade trypsin and T4 polynucleotide kinase were purchased from Promega (Madison, WI) and Amersham Biosciences (Piscataway, NJ), respectively. All other chemicals were purchased from Sigma Chemical Co. (St. Louis, MO), BioRad (Hercules, CA), or Fisher Scientific (Fairlawn, NJ). Instrumental Analysis. HPLC analyses were carried out with either a Waters (Mildford, MA) HPLC system equipped with a Waters 600E pump, Waters 996 PDA detector, and a Waters 717 autosampler or a Waters 515 system linked to a Waters 2487 dual absorbance detector and β-Ram radioflow detector (IN/US Systems, Inc., Tampa, FL). NMR characterizations were acquired on a Varian VAC-300 (Varian, Inc., Palo Alto, CA). 5′-O-(4,4′-Dimethoxytrityl)-N2-(phenoxyacetyl)-O6-(nbutyl)-2′-deoxyguanosine-3′-O-(2-cyanoethyl)-N,N-di-isopropylamide-O-phosphite. This compound was prepared using slight modifications of literature procedures. Briefly, O6(n-butyl)-2′-deoxyguanosine was prepared by coupling n-butanol with 2′-deoxy-N2-isobutyrylguanosine 3′,5′-diisobutryate in a Mitsunobu reaction followed by removal of the three isobutyryl protecting groups in methanolic sodium hydroxide according to published procedures (12, 32). The N2-position of O6-(n-butyl)2′-deoxyguanosine was protected with a phenoxyacetyl group according to the method of Schulhof (33) as described by Wang et al. (34). The resulting crude product was reacted with 4,4′dimethoxytrityl chloride in anhydrous pyridine at 4 °C (34, 35). 5′-O-(4,4′-Dimethoxytrityl)-N2-(phenoxyacetyl)-O6-(n-butyl)-2′deoxyguanosine was purified by silica gel flash chromatography with 100:1:05 CH2Cl2/MeOH/Et3N as the mobile phase. This compound was reacted with cyanoethyl N,N-diisopropylchlorophophoramidite in dry CH2Cl2 containing N,N-diisopropylethylamine at 0 °C (34, 35). After 1 h at room temperature, the reaction mixture was concentrated under reduced pressure and the residue was dissolved in EtOAc. The organic solution was washed four times with brine, dried over MgSO4, filtered, and concentrated. The final product was purified by silica gel flash chromatography, eluting with 75:25:0.5 CH2Cl2/EtOAc/Et3N. 1H NMR (CDCl3): δ 8.6 (br s, 1 H, N2-H), 8.0 (2s, 1H, C8-H), 7.47.1 (m, 12H, ArH), 7.1-6.9 (m, 3H, ArH), 6.8-6.7 (m, 4H, ArH), 6.42 (m, 1H, 1′-H), 4.9-4.7 (m, 3H, 3′-H, phenyl-O-CH2CONH), 4.6 (t, 2H, O6-CH2), 4.3-4.2 (m, 1H, 4′-H), 3.9-3.5 (m, 10H, 2 ArOCH3, 5′-H, POCH2), 3.4-3.3 (m, 2H, Me2CHN), 2.9-2.65 (m, 2H, 2′-H), 2.6 (t, 2H, CH2CN), 1.9-1.8 (m, 2H, O6-CH2CH2), 1.6-1.4 (m, 2H, O6-CH2CH2CH2), 1.3-1.05 (m, 12 H, CH3), 0.95 (t, 2H, O6- CH2CH2CH2CH3). DNA Oligonucleotides. DNA oligonucleotides of the sequence 5′-AACAGCCATAT*gGCCC where *g is O6-pobG, O6-

Scheme 2. Structure of O6-Alkylguanine Adducts

426

Chem. Res. Toxicol., Vol. 17, No. 3, 2004

Mijal et al.

Figure 1. Alignment of the adduct binding regions of the AGT variant and mutant proteins. Genbank accession numbers for the various proteins are as follows: E. coli AdaC, 1SFE; E. coli Ogt, CAA68548; mouse, NP_502570; rat, CAA38648; and human, P_002403. Gray shading indicates sequence identity with human wild-type protein. The active site cysteine residue is shaded in black, and the amino acid changes from human wild-type in the variants and mutants are underlined. mG, O6-buG, or O6-bzG were prepared as previously described (34, 36, 37). Unmodified DNA oligonucleotides were purchased from Oligos, Etc. (Wilsonville, OR) and IDT-DNA (Coraville, IA). Reaction of AGT with Alkylated DNA. [5-3H]NNKOAcor [3H]MNU-treated calf thymus DNA (specific activity: [5-3H]NNKOAc, 2510 mCi/mmol; [3H]MNU, 23 500 mCi/mmol) containing 3.7 pmol of O6-pobG or O6-mG was combined with hAGT or crude rat liver AGT (2.7 pmol) in 50 mM Tris, 0.1 mM EDTA, and 1 mM DTT, pH 7.8 (total volume, 1.6 mL), for 30 min at 37 °C. Negative controls were performed in the absence of protein or with AGT that had been preincubated with 1 mM O6-bzG for 5 min prior to the addition of the DNA substrate. The reactions were stopped by freezing at -20 °C. Reaction mixtures were separated by SDS-PAGE (15% gel; 37.5:1 acrylamide/ N,N′-methylene bisacrylamide). Gels (73 mm × 83 mm × 1.5 mm) were run at constant voltage (200 V) for 35-45 min and then stained with GelCode blue stain reagent (Pierce, Rockford, IL). Prior to exposing the gels to film (Kodak BioMax MR1), they were soaked in Amplify fluorogenic reagent (Amersham Bioscience, Buckinghamshire, England). Dried gels were exposed to film in the presence of an intensifying screen for 1020 weeks at -80 °C. The developed film was imaged with Fluor-S Imager interfaced with BioRad’s Fluor-S image quantitation software. Mass Spectral Analysis of Tryptic Digests of hAGT following Reaction with O6-pobG in Oligonucleotides. Oligonucleotides [5′-d(AACAGCCATAT*gGCCC)-3′] containing O6-pobG (600 pmol) were reacted with recombinant hAGT (40 µg) at 37 °C in 50 mM Tris, pH 7.6. Negative controls were performed in the absence of oligonucleotide. After 30 min, CaCl2 was added to a final concentration of 1 mM. Trypsin (1.6 µg; ratio of AGT:trypsin, 25:1) was added, and the reaction was incubated overnight at 37 °C. The digestion was terminated with the addition of glacial acetic acid to achieve pH 3, and acetonitrile was added to a final concentration of 10%. Samples were stored at -20 °C. Prior to MALDI-TOF analysis, the peptide mixtures were desalted using Millipore C18 ZipTips (Bedford, MA). MALDI-TOF analyses were performed on a QSTAR Pulsar I (Applied Biosystems Inc., Foster City, CA) quadrapole timeof-flight mass spectrometer with an orthogonal MALDI source. Full scans of the peptide mixture from 600 to 3500 m/z and tandem mass spectral data of select ions were collected with dihydroxybenzoic acid (Agilent Technologies, Palo Alto, CA) as the matrix. The TOF region acceleration voltage was 4 kV, and the injection pulse repetition rate was 6.0 kHz. Laser pulses were generated with a nitrogen laser at 337 nm and 33 µJoules of laser energy at a laser repetition rate of 20 Hz. Mass spectra were the average of approximately 100 laser shots collected in positive mode. External calibration was performed with human angiotensin II (monoisotopic [MH+] ) m/z 1046.5) and adrenocorticotropin hormone fragment 18-39 (monoisotopic [MH+] ) m/z 2465.2). HPLC Analysis of O6-mG vs O6-pobG, O6-buG, and O6-bzG Repair Preference. Repair substrates consisted of

single-stranded oligonucleotides 5′-d(AACAGCCATAT*gGCCC)3′, where *g represents O6-mG, O6-pobG, O6-buG, or O6-bzG. Equal picomole amounts of O6-mG and either O6-pobG, O6-buG, or O6-bzG containing oligonucleotides (450-500 pmol each) were incubated with purified AGTs (0-100 µg) for 10 or 30 min at 37 °C in 50 mM Tris buffer, pH 7.8, containing 0.5 mM DTT and 0.1 mM EDTA. Reactions were stopped upon addition of SDS to a final concentration of 1%. Reaction mixtures were separated on either a Phenomenex Bondclone C18 column (4.6 mm × 300 mm; Torrance, CA) or a Keystone Scientific Hypersil ODS column (4.6 mm × 250 mm, Bellfonte, CA) with a linear gradient from 100 mM ammonium acetate containing 5% acetonitrile to 100 mM ammonium acetate containing 20% acetonitrile over 30 min (flow, 1 mL/min), monitoring at 260 nm. The repair preference was expressed as a ratio of O6-mG repair to X repair, where X refers to O6-pobG, O6-buG, or O6bzG. Peak areas of O6-mG, X, and unmodified oligonucleotides were summed to calculate the percentage of each oligonucleotide present after reaction with AGT. This percentage was subtracted from the percent found in the starting mixture (DNA only control), and the ratio of the percent change in O6-mG:X was calculated for each protein amount assayed. These ratios were averaged for reactions in which less than 50% of the total input oligonucleotides were repaired (n ) 3-7). PAGE Analysis of O6-pobG Repair Preference. A mixture containing equal amounts of oligonucleotides containing O6-mG and O6-pobG (250 pmol total) was [32P]end-labeled with [32P]ATP (250 uCi) and T4 kinase (25 units) in 50 mM Tris-HCl (pH 7.5) containing 10 mM MgCl2 and 10 mM 2-mercaptoethanol for 30 min at 37 °C. The DNA oligonucleotide was heated for 10 min at 65 °C to inactivate the kinase, and the DNA oligonucleotides were purified via C18 spin column chromatography. To produce double-stranded substrate, a portion of the [32P]end-labeled DNA oligonucleotides was combined with a 2-fold excess of the complimentary DNA oligonucleotide, heated at 65 °C, and slowly cooled to room temperature. Single-stranded and double-stranded [32P]end-labeled oligonucleotide mixtures (0.5 pmol of total oligonucleotide) were incubated with varying amounts of purified AGT (0-25 ng) for 30 min at 37 °C in 50 mM Tris buffer, pH 7.8, containing 1.0 mM DTT, 0.1 mM EDTA, and 0.1-0.5 mg/mL BSA (total volume 15 µL). Negative controls were performed without protein or with AGT that had been reacted with 5 mM O6-bzG for 5 min prior to the addition of the oligonucleotide substrates. All reactions were performed in duplicate, and each experiment was repeated 3-4 times. Reactions were stopped upon addition of 3 µL of formamide running buffer and electrophoresed on 20% denaturing polyacrylamide gels, pH 8.1 or 8.3, run at constant voltage (2500 V). The radioactivity associated with each gel band was determined with a Molecular Dynamics Storm 8400 phosphoimager. Repair preference was determined as follows. First, the radioactivity in each band area (O6-pobG, O6-mG, or unmodified oligonucleotide) was expressed as a percentage of

Repair of O6-pobG by AGT Variants

Figure 2. SDS-PAGE analysis of reactions between AGT and [5-3H]NNKOAc-treated DNA. (A) Rat liver or recombinant histagged hAGT were reacted with [5-3H]NNKOAc-treated DNA (specific activity, 2510 mCi/mmol). The rat protein runs faster than the human since it lacks the histidine tag. (B) hAGT was reacted with [5-3H]NNKOAc-treated DNA in the presence or absence of O6-bzG, as a free base. (C) hAGT was reacted with [3H]MNU-treated DNA (specific activity, 23 500 mCi/mmol) in the presence or absence of O6-bzG, as a free base. For B and C: lane 1, AGT + DNA + O6-bzG; lane 2, AGT + DNA; and lane 3, DNA alone. The arrow marks the AGT band. the total radioactivity in the substrate and product bands. Then, the values obtained for O6-pobG and O6-mG were plotted against the amount of protein in the reaction, and the slope of each line was calculated by regression (Microsoft Excel, 1998). Dividing the O6-mG slope by the O6-pobG slope yielded the repair preference ratio. Differences between repair ratios were assessed by two-tailed t-tests assuming nonequal variance (Microsoft Excel, 1998). Analyses were restricted to reactions in which there was less than 40% repair of the substrates.

Results To ensure that we were measuring competitive repair of pyridyloxobutyl DNA adducts and O6-mG, we wanted to confirm that the reaction between pyridyloxobutylated DNA and AGT resulted in the transfer of the pyridyloxobutyl group to the protein. To test this possibility, rat liver AGT or hAGT was incubated with [5-3H]NNKOActreated DNA for 30 min and the resultant protein mixtures were separated by SDS-PAGE. Radioactivity was associated with the AGT protein band (Figure 2A), demonstrating that the pyridyloxobutyl group had been transferred to the protein. Because the AGT inactivator, O6-bzG, blocks the transfer reaction (Figure 2B), it is likely that the alkyl group is covalently attached to active site cysteine. Similar results were obtained with [3H]MNU-treated DNA (Figure 2C). These data are consistent with an AGT inactivation mechanism involving alkyl group transfer to the protein. To confirm that the alkyl group is transferred from O6pobG to the active site of AGT, tryptic fragments of alkylated hAGT were analyzed by MALDI-TOF mass spectrometry. In these studies, the protein was reacted with an oligonucleotide containing O6-pobG (34), instead of pyridyloxobutylated DNA. We used this simpler model

Chem. Res. Toxicol., Vol. 17, No. 3, 2004 427

to achieve larger amounts of the alkylated protein, which were required for mass spectral analysis. MALDI-TOF MS/MS analysis of the tryptic digests of this protein indicated the presence of multiple peptides (Figure 3A,B). The tryptic fragment containing the active cysteine (GNPVPILIPCHR) produced an ion with m/z 1315.7 in the digest of the unmodified protein. In the alkylated protein, this fragment produces an ion at m/z 1462.7, consistent with the presence of a pyridyloxobutyl group on the peptide. This is the only peptide that experienced a molecular weight change. MS/MS analysis of this ion yielded daughter ions corresponding to b and y type ions expected for the peptide fragment containing a pyridyloxobutylated Cys145 residue (Table 1 and Figure 3D). These experiments established that O6-pobG competes with O6-mG predominantly through competitive alkyl group transfer. The next goal was to determine how different AGT proteins repaired O6-pobG. To determine the relative repair of this adduct by individual proteins, O6-pobG was competed with O6-mG for repair by the variant and mutant recombinant AGT proteins. Preliminary studies were conducted with [5-3H]NNKOAc- and [3H]methylnitrosourea-treated calf thymus DNA containing known levels of O6-pobG and O6-mG, respectively. However, these studies were complicated by the observation that the amount of rat liver AGT required to repair a known amount of O6-pobG in [5-3H]NNKOAc-treated DNA varied with each DNA preparation (Thomson, N. M., and Peterson, L. A. Unpublished observations). This phenomenon did not occur with methylated DNA, where O6-mG is the primary AGT substrate adduct. The reasons behind this variation are currently under investigation in our laboratory. One explanation is that there may be other unstable cyclic AGT reactive adducts present in [5-3H]NNKOAc-treated calf thymus DNA. To avoid these complexities, oligonucleotides containing site specifically incorporated O6-pobG and O6-mG were employed for the competition experiments. The relative repair of O6-mG vs O6-pobG was determined for a variety of AGT variants and mutants by incubating increasing amounts of recombinant AGT with equal molar amounts of single-stranded oligonucleotides containing O6-mG or O6-pobG. Repair was measured by HPLC analysis, monitoring the disappearance of the adducted oligonucleotides and the appearance of the unmodified strand. Representative traces are displayed in Figure 4. Relative rates of repair for all of the proteins are reported in Table 2. With the exception of the bacterial proteins, the amount of O6-pobG and O6-mG repair in the mixtures equaled the total amount of active AGT present in the reaction (Figure 5), indicating that the two adducts compete for reaction with AGT via competitive alkyl transfer. The ability of the AGT to repair O6-pobG relative to O6-mG was highly dependent on protein structure. The rodent proteins, rAGT and mAGT, reacted with both adducts at the same rate (Table 2). In contrast, the bacterial proteins, AdaC and Ogt, reacted poorly, if at all, with O6-pobG (Table 2 and Figure 5A) even in the absence of O6-mG (Table 3). Wild-type hAGT displayed an approximately 2-fold preference for the removal of O6-mG over O6-pobG (Table 2). While I143V/K178R was similar to wild-type hAGT in its relative ability to react with O6-pobG, G160R was less active toward the bulky adduct (Table 2). In competi-

428

Chem. Res. Toxicol., Vol. 17, No. 3, 2004

Mijal et al.

Figure 3. MALDI-TOF mass spectral analysis of tryptic digests of hAGT reacted with oligonucleotide of the sequence AACAGCCATAT*gGCCC where *g is G (A and C) or O6-pobG (B and D). (A,B) Full scans of the peptide mixtures. (C,D) Tandem mass spectral data of the active site peptide (C, m/z 1316; D, m/z 1463). Underlined C in panel C is Cys145; unlined B in panel D is pyridyloxobutylated Cys145. The m/z values for the b and y ions are listed in Table 1. The b-NH3 and y-NH3 are the ions in which the b or y ions have also lost ammonia. Table 1. Expected y and b Peptide Fragments for the Tryptic Fragment Containing Active Site CYS145 Alkylated with a Pyridyloxobutyl Groupa m/z

m/z

unmodified

alkylated

1140.7 1003.6 901.6 804.5 691.4 578.3 465.2 368.2 269.1 172.1 58

1288.7 1151.6 901.6 804.5 691.4 578.3 465.2 368.2 269.1 172.1 58

a

b11 b10 b9 b8 b7 b6 b5 b4 b3 b2 b1

b series

y series

GNPVPILIPBH GNPVPILIPB GNPVPILIP GNPVPILI GNPVPIL GNPVPI GNPVP GNPV GNP GN G

R HR BHR PBHR IPBHR LIPBHR ILIPBHR PILIPBHR VPILIPBHR PVPILIPBHR NPVPILIPBHR GNPVPILIPBHR

y1 y2 y3 y4 y5 y6 y7 y8 y9 y10 y11

alkylated

unmodified

175.1 312.2 562.3 659.3 772.4 885.5 998.6 1095.6 1194.7 1291.7 1405.8 1462.9

175.1 312.2 414.3 511.3 624.4 737.5 850.6 947.6 1046.7 1123.7 1257.8 1314.8

The bold B represents pyridyloxobutylated Cys145. The bold numbers are the observed ions.

tion experiments, there was no significant repair of O6pobG by G160R until almost all of the O6-mG had reacted (Figure 5B). However, G160R repaired O6-pobG in the absence of the O6-mG-containing oligonucleotide. The repair of O6-pobG was complete by 30 min, yielding the same pmol O6-alkylguanine repaired/µg protein as observed with O6-mG alone (O6-pobG, 24 ( 3 pmol/µg protein; O6-mG, 22 ( 2 pmol/µg protein). A series of mutant human proteins were also tested to determine how particular amino acid residues in the binding pocket affected the relative repair of O6-mG and O6-pobG. The human Asn157 amino acid residue is equivalent to rat Asn161 and mouse His161 (38). Changing His161 to Asn in the mouse protein generates a

mutant that is more sensitive to the free base O6-bzG (38). Substitutions of Asn157 in the human protein yield mutants that are more resistant to O6-bzG (39). Therefore, the expectation was that the human mutant, N157H, would be more resistant to O6-pobG than wildtype hAGT. However, N157H displayed a preference for O6-mG over O6-pobG like that observed with hAGT (Table 2). The other mutants tested were P140K, G156A, Y158H, and E166G. The first three of these mutants are all resistant to reaction with free base O6-bzG but for different reasons. P140K, like AdaC (40), lacks the proline at amino acid residue 140, which binds the benzyl group of O6-bzG; it is replaced by the charged amino acid

Repair of O6-pobG by AGT Variants

Chem. Res. Toxicol., Vol. 17, No. 3, 2004 429

Figure 4. Representative HPLC traces obtained from reactions of recombinant mouse, rat, or human his-tagged AGT with an equal amount of O6-mG and O6-pobG in oligonucleotides of the sequence AACAGCCATAT*gGCCC where *g is O6-pobG or O6-mG. Table 2. Single-Stranded Repair Preferences in 1:1 Competition Reactions with O6-mG vs O6-pobG, O6-buG, and O6-bzG as Measured by HPLC repair ratioa protein mAGT rAGT Ada Ogt hAGT I143V/K178R G160R N157H P140K G156A Y158H E166G

O6-mG/O6-pobG O6-mG/O6-buG O6-mG/O6-bzG 1.1 ( 0.1b 1.3 ( 0.2d >10g >10g 2.6 ( 0.5d 2.6 ( 0.5e ∼7.5 2.4 ( 0.7e >10g >10g >10g 6.0 ( 1.6e

0.9 ( 0.1e 0.9 ( 0.03d >10g >10g 2.4 ( 0.2d 3.0 ( 0.3c 3.5 ( 0.8d 2.0 ( 0.2e >10g 4.0 ( 0.2c >10g 3.0 ( 0.5d

0.15 ( 0.06d 0.10 ( 0.04d >10h 1h 0.04 ( 0.04c 0.09 ( 0.04e 0.27 ( 0.04f 0.15 ( 0.04e 5.8 ( 3.1e 5a >5a 1.9 ( 0.5b 1.8 ( 0.4c g2.8d 1.6 ( 0.2b

a No measurable repair of O6-pobG while O6-mG was still present. b n ) 3. c n ) 4. d Different from hAGT (p ) 0.04), I143V/ K178R (p ) 0.02), and E166G (p ) 0.01).

(42), we wanted to determine if the relative repair rates of O6-mG and O6-pobG were different when the substrates were double-stranded. Initial attempts to measure these rates with HPLC analysis were unsuccessful. A denaturing PAGE method was devised, which separated [32P]end-labeled oligonucleotides containing either O6pobG or O6-mG from one another as well as from the unmodified strand (Figure 6). The relative reactions rates were not significantly different when performed with double-stranded substrates (Table 4) except for E166G. This protein repaired O6-pobG more efficiently when double-stranded substrates were employed. To understand the role of adduct structure on relative repair rates, the competition experiments were repeated with two other bulky O6-alkylguanine adducts, O6-buG and O6-bzG. When oligonucleotides containing O6-buG, a structural analogue of O6-pobG, were co-incubated with those containing O6-mG in competition reactions, the proteins displayed similar preference as observed with O6-pobG (Table 2). The rodent proteins repaired O6-buG as efficiently as O6-mG. Wild-type and variant hAGT (G160R and I143V/K178R) differed less in their relative ability to repair the smaller O6-buG adduct but still

preferentially repaired O6-mG. The bacterial proteins, AdaC and Ogt, showed a strong preference for reaction with O6-mG over O6-buG (Table 2). However, Ogt repaired O6-buG when all of the O6-mG had been repaired (data not shown). As with O6-pobG, the human mutant, N157H, had the same preference for O6-buG as the wildtype protein (Table 2). While P140K and Y158H reacted poorly with O6-buG and O6-pobG, the mutants G156A and E166G repaired O6-buG more effectively than O6pobG. While O6-bzG is a bulky adduct, this adduct was preferentially repaired relative to O6-mG by most AGT proteins. These findings are in agreement with previous studies, which indicated that hAGT preferentially repairs O6-bzG over O6-mG in oligonucleotides (43). The repair of O6-bzG was much faster than O6-mG for all of the mammalian proteins. While G160R also exhibited a preference for O6-bzG over O6-mG, it was slower to react with O6-bzG than either wild-type hAGT or I143V/ K178R. Of the four mutant hAGT proteins examined, the only one that preferentially repaired O6-mG was P140K. Y158H displayed only a 2-fold preference for reaction with O6-bzG relative to O6-mG whereas N157H and E166G mutants were indistinguishable from wild-type hAGT. In similar competition experiments, G156A was previously shown to preferentially repair O6-bzG over O6mG (25). Previous studies with the bacterial proteins indicated that AdaC is resistant to reaction with O6-bzG in oligonucleotides and that Ogt reacts equally well with O6-mG and O6-bzG in oligonucleotides (43).

Discussion Previous studies from our laboratory indicated that pyridyloxobutyl adducts interfered with repair of O6-mG by AGT (20, 21). In this report, we demonstrate that pyridyloxobutyl DNA adducts are repaired by AGT in a reaction that results in pyridyloxobutyl transfer to the protein. Mass spectral data demonstrated that the alkyl group is transferred to the active site cysteine. Consistently, the loss of the alkyl groups from both of the O6alkylguanine-containing oligonucleotides coincides with the appearance of the unmodified strand. These data coupled with the observation that the total amount of O6alkylguanine repair equaled the amount of active AGT added to the reaction mixtures support competitive alkyl transfer as a primary mechanism for the previously observed O6-mG repair inhibition by pyridyloxobutylated DNA. Our studies indicate that the ability of AGT orthologs to repair O6-pobG is highly dependent on AGT protein structure. The bacterial AGTs, which have a small binding pocket, are ineffective in repairing O6-pobG. The human protein is better able to accommodate O6-pobG since it has a larger and more hydrophobic active site than AdaC (25). However, this protein is still more active

Figure 6. Representative gels obtained from reactions of AGT variants with equal amounts of O6-mG and O6-pobG in doublestranded oligonucleotides of the sequence AACAGCCATAT*gGCCC where *g is O6-pobG or O6-mG. D ) DNA only. BG ) AGT inactivated with O6-bzG. The number 1 denotes the initial amount of protein added as follows: rAGT ) 9 ng, hAGT ) 11 ng, G160R ) 12 ng, Ogt ) 8 ng, and AdaC ) 4 ng. Subsequent numbers represent fold increase in protein amount used in the reactions.

Repair of O6-pobG by AGT Variants

in repair of O6-mG relative to O6-pobG. Finally, rodent AGTs display no repair preference between O6-pobG and O6-mG. This ability to accommodate such large structural differences likely results from the additional amino acid residue (Gly166) in the binding pocket of the rodent proteins (38). Therefore, the steric constraints of an AGT’s active site will determine whether it can repair a bulky O6-alkylguanine adduct. Wild-type hAGT and the I143V/K178R variant are similar in the rate of bulky O6-alkylguanine repair (Tables 2 and 4). However, the G160R variant was less able to repair O6-pobG. This AGT variant repaired all three bulky O6-alkylguanine adducts less efficiently than the other human proteins. Its binding pocket has added steric bulk and an additional positive charge (44). However, G160R still preferentially reacted with O6-bzG relative to O6-mG when these adducts are in oligonucleotides, indicating that alkyl group transfer is facilitated by the benzyl group’s ability to delocalize a positively charged transition state. The poor repair of O6-pobG by AdaC is consistent with previous observations that the rate of O6-alkylguanine repair by this protein decreases as the size of the alkyl group increases (24, 25, 45, 46). Ogt more readily repairs bulky O6-alkylguanine adducts than AdaC (this study) (43); however, both bacterial proteins repaired O6-mG . O6-buG > O6-pobG. We observed that mammalian AGTs have broader substrate specificity than the bacterial proteins. This is consistent with previous reports (24, 25, 43). However, we were surprised to observe that the rodent proteins repaired O6-mG, O6-buG, and O6-pobG at comparable rates. This contrasts with a previous study in which the rate of repair of O6-alkylguanine adducts by rat AGT was measured in alkylated DNA (24). The relative rates of repair in this study were as follows: methyl > ethyl, n-propyl > n-butyl > isopropyl, isobutyl > 2-hydroxyethyl. There may be several reasons for this discrepancy. The earlier study by Morimoto et al. measured rates of O6alkylguanine repair in alkylated DNA substrates (24). Our studies were conducted with oligonucleotides, which contained a single DNA adduct in a defined sequence. Because alkylated DNA has adducts in addition to O6alkylguanine, the presence of these other adducts may retard repair of the target adduct. In the case of the propyl and butyl adducts, the alkylated DNA also contained both straight chain and branched O6-alkylguanine adducts within the same DNA substrate. These O6alkylguanine adducts will compete with each other for reaction with AGT so the repair rates observed for each individual adduct were likely affected by this competition. In addition, adduct density (pmol adduct per nucleotide) may affect repair rates since AGT activity can be retarded by the presence of unmodified DNA (47). In the Morimoto study, the levels of the bulky O6-alkylguanine adducts were significantly lower than those of O6-mG (24). In our studies, the adduct density is equivalent for both adducts. Finally, alkylated DNA has O6-alkylguanine adducts in a variety of sequence contexts. Therefore, the overall rate of repair is an average repair rate across these sequences. Several studies have demonstrated that the repair of O6-mG is affected by sequence context (48-50). In our studies, both adducts were in the same sequence context avoiding this potential problem.

Chem. Res. Toxicol., Vol. 17, No. 3, 2004 431

The results obtained with mutant hAGTs support the conclusions derived from the natural and species variants that binding pocket size determines the reactivity of AGT with larger O6-alkylguanine residues. Thus, the alterations P140K, Y158H, and G156A all reduce the space available in the binding pocket for the alkylated nucleoside to be accommodated. Although the former two mutants also have an additional positively charged side chain in this pocket, it appears that the steric effect is predominantly responsible for reducing repair of O6-pobG and O6-buG. The similarity between the N157H mutant and the wild-type hAGT support this hypothesis. It should be noted that other factors involving the binding of the AGT protein to DNA may also influence relative rates of repair of different adducts. Mutant E166G, which was supposed to serve as a negative control, was slow to repair O6-buG and O6-pobG when presented as a single-stranded context, indicating that changes in protein structure outside of the binding pocket have the potential to alter bulky O6-alkylguanine repair. However, when the experiment was repeated with doublestranded oligonucleotides, this mutant displayed the same repair preference as the wild-type protein. The relative rate of alkyl group transfer is also dependent on the alkyl group present in the O6-position. While the ability of the active site to accommodate large alkyl groups affects the speed of repair, rate is also determined in part by the ease of alkyl transfer to the protein. SN2 reaction rates are influenced by the structure of the alkyl group where the relative SN2 rates are benzyl > methyl > ethyl ) propyl. Therefore, it is not surprising that all of the mammalian proteins, which can readily accommodate O6-bzG, repaired it faster than O6-mG. However, the equal repair of O6-mG and the much larger adduct, O6-pobG, by rodent proteins was unanticipated. This cannot be explained by neighboring group participation of the carbonyl oxygen facilitating the transfer reaction since O6-buG, which has no potential for neighboring group participation, is repaired as efficiently or more efficiently than O6-pobG by all proteins. Therefore, the overall determinant of repair rate for this bulky adduct seems to be how well the protein’s binding site can accommodate the large pyridyloxobutyl group. The toxicological implications of these findings are significant. The inability of the bacterial proteins AdaC and Ogt to significantly repair O6-pobG may explain the increased sensitivity of bacterial strains to the mutagenic activity of pyridyloxobutylating agents relative to methylating agents (51, 52). Consistent with the poor repair of O6-pobG by bacterial AGTs, this adduct was highly mutagenic in E. coli when presented in gapped vectors (15). Nearly all of the resulting colonies were mutant; efficient repair of O6-pobG would result in a lower fraction of mutant colonies. The readily repaired O6-mG was substantially less mutagenic in this same system (53). These observations suggest that the inability to repair O6-pobG results in increased mutagenesis. The role of O6-pobG in the carcinogenic properties of NNK in laboratory animals requires further investigation. NNK will generate both O6-mG and O6-pobG in vivo (13). The relative contribution of these two adducts to the overall carcinogenic properties of NNK will depend on both the levels of each adduct formed and how they are repaired. Our studies indicate that rodent AGTs will repair O6-mG and O6-pobG at similar rates.

432

Chem. Res. Toxicol., Vol. 17, No. 3, 2004

In A/J mice, the formation and persistence of O6-mG are critical for NNK-induced lung tumorigenesis (6, 7, 11, 54). Consistently, the levels of O6-pobG in the lung 8 h following a carcinogenic dose of NNK are at least an order of magnitude lower than the levels of O6-mG in these animals (13). While both of these adducts are repaired by AGT in vivo (7, 8, 13), the available data indicate that there are insufficient levels of O6-pobG to significantly contribute to the carcinogenic activity of NNK or to influence the repair of O6-mG by AGT. The importance of each O6-alkylguanine adduct in NNK-mediated carcinogenesis has not been fully addressed in rats. In NNK-treated rats, O6-mG levels in Clara cells (55) and pyridyloxobutyl adducts in type II cells (56) correlate with lung tumor formation. Pyridyloxobutyl DNA adducts are also present at significant levels in Clara cells (56). Consequently, it is likely that both alkylation pathways contribute to the pulmonary carcinogenic properties of NNK in rats. The contribution of O6-pobG to the total pyridyloxobutyl adducts that accumulate in rat lung cells is unknown. Because AGT is depressed in both cell types following NNK exposure (10), it is possible that O6-pobG would persist in these cells unless other repair pathways, such as nucleotide excision repair, are also involved in its repair. Because repair of O6-alkylguanine adducts reduces the mutagenic activity of alkylating agents, individuals less able to repair this damage could be at increased risk when chronically exposed to alkylating carcinogens such as nitrosamines. Epidemiological data linking the I143V/ K178R hAGT variant with increased lung cancer risk are consistent with this hypothesis (57). Several potential functional differences between the variant and the wildtype proteins could explain the observed variation in risk. Because I143V/K178R and hAGT repair methylated DNA with similar efficiency (58), one could hypothesize that the proteins differ in rate of repair for other O6-alkylguanine adducts, as observed for the G160R variant (44). However, our data argue against this hypothesis; I143V/ K178R was indistinguishable from the wild-type protein in the repair of all three bulky adducts tested. While there may be a functional difference in the repair of yet uncharacterized pyridyloxobutyl or other DNA adducts, the variant is more likely to be in linkage with another gene associated with increased risk or expressed at different levels than wild-type AGT. Although the functional data presented in this report cannot support the epidemiological finding of Kaur and co-workers (57), our study clearly demonstrates that variation in bulky O6-alkylguanine repair capacity exists within the human population. G160R exhibited a markedly different ability to repair O6-pobG when O6-mG was present than either wild-type or I143V/K178R hAGT proteins. This functional difference, slow repair of O6pobG, could modify lung cancer risk. Unfortunately, the low prevalence of this variant (59, 60) has prevented the detection of an association between G160R and lung cancer risk (57). To evaluate the potential relationship between repair deficiencies and sensitivity to alkylating agents, bulky adduct repair phenotypes must be characterized and their prevalence within populations assessed. While many AGT phenotyping experiments have been completed measuring either total AGT activity (61, 62), rates of O6-mG repair (61-63), or inactivation by O6-bzG (64), none has assessed differences in bulky adduct repair.

Mijal et al.

This work underscores the potential use of O6-pobG contained in DNA, a representative bulky adduct substrate, as a tool for phenotyping bulky O6-alkylguanine adduct repair ability of human AGTs.

Acknowledgment. We thank Dr. LeeAnn Higgins for assistance with MALDI-TOF mass spectral analysis. They were performed at the Mass Spectrometry Consortium for the Life Sciences in the Department of Biochemistry, Molecular Biology and Biophysics at the University of Minnesota. R.S.M. is funded by NIH Training Grant ES-10956. This research was supported by CA-59887 (L.A.P.) and CA-18137 (A.E.P.) from the National Cancer Institute.

References (1) Hecht, S. S. (1998) Biochemistry, biology, and carcinogenicity of tobacco-specific N-nitrosamines. Chem. Res. Toxicol. 11, 560-603. (2) Hecht, S. S., Chen, C. B., Ohmori, T., and Hoffmann, D. (1980) Comparative carcinogenicity in F344 rats of the tobacco specific nitrosamines, N′-nitrosonornicotine, and 4-(N-methyl-N-nitrosamino)-1-(3-pyridyl)-1-butanone. Cancer Res. 40, 298-302. (3) Rivenson, A., Hoffmann, D., Prokopczyk, B., Amin, S., and Hecht, S. S. (1988) Induction of lung and exocrine pancreas tumors in F344 rats by tobacco-specific and areca-derived N-nitrosamines. Cancer Res. 48, 6912-6917. (4) Hecht, S. S., Morse, M. A., Amin, S. G., Stoner, G. D., Jordan, K. G., Choi, C. I., and Chung, F. L. (1989) Rapid single dose model for lung tumor induction in A/J mice by 4-(methylnitrosamino)1-(3-pyridyl)-1-butanone and the effect of diet. Carcinogenesis 10, 1901-1904. (5) Hecht, S. S. (1999) Tobacco smoke carcinogens and lung cancer. J. Natl. Cancer Inst. 91, 1194-1210. (6) Hecht, S. S., Jordan, K. G., Choi, C. I., and Trushin, N. (1990) Effects of deuterium substitution on the tumorigenicity of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone and 4-(methylnitrosamino)1-(3-pyridyl)-1-butanol in A/J mice. Carcinogenesis 11, 10171020. (7) Peterson, L. A., and Hecht, S. S. (1991) O6-Methylguanine is a critical determinant of 4-(methylnitrosamino)-1-(3-pyridyl)-1butanone tumorigenesis in A/J mouse lung. Cancer Res. 51, 55575564. (8) Peterson, L. A., Thomson, N. M., Crankshaw, D. L., Donaldson, E. E., and Kenney, P. J. (2001) Interactions between methylating and pyridyloxobutylating agents in A/J mouse lungs: implications for 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone-induced lung tumorigenesis. Cancer Res. 61, 5757-5763. (9) Devereux, T. R., Belinsky, S. A., Maronpot, R. R., White, C. M., Hegi, M. E., Patel, A. C., Foley, J. F., Greenwell, A., and Anderson, M. W. (1993) Comparison of pulmonary O6-methylguanine DNA adduct levels and Ki-ras activation in lung tumors from resistant and susceptible mouse strains. Mol. Carcinogenesis 8, 177-185. (10) Belinsky, S. A., Dolan, M. E., White, C. M., Maronpot, R. R., Pegg, A. E., and Anderson, M. W. (1988) Cell specific differences in O6methylguanine-DNA methyltransferase activity and removal of O6-methylguanine in rat pulmonary cells. Carcinogenesis 9, 20532058. (11) Ronai, Z. A., Gradia, S., Peterson, L. A., and Hecht, S. S. (1993) G to A transitions and G to T transversions in codon 12 of the Ki-ras oncogene isolated from mouse lung tumors induced by 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and related DNA methylating and pyridyloxobutylating agents. Carcinogenesis 14, 2419-2422. (12) Wang, L., Spratt, T. E., Liu, X. K., Hecht, S. S., Pegg, A. E., and Peterson, L. A. (1997) Pyridyloxobutyl adduct O6-[4-oxo-4-(3pyridyl)butyl]guanine is present in 4-(acetoxymethylnitrosamino)1-(3-pyridyl)-1-butanone-treated DNA and is a substrate for O6alkylguanine-DNA alkyltransferase. Chem. Res. Toxicol. 10, 562567. (13) Thomson, N. M., Kenney, P. M., and Peterson, L. A. (2003) The pyridyloxobutyl DNA adduct, O6-[4-oxo-4-(3-pyridyl)butyl]guanine, is detected in tissues from 4-(methylnitrosamino)-1-(3pyridyl)-1-butanone-treated A/J mice. Chem. Res. Toxicol. 16, 1-6. (14) Wang, M., Cheng, G., Sturla, S. J., McIntee, E. J., Villalta, P. W., Upadhyaya, P., and Hecht, S. S. (2003) Identification of adducts formed by pyridyloxobutylation of deoxyguanosine and DNA by 4-(acetoxymethylnitrosamino)-1-(3-pyridyl)-1-butanone, a chemically activated form of tobacco specific carcinogens. Chem. Res. Toxicol. 16, 616-626.

Repair of O6-pobG by AGT Variants (15) Pauly, G. T., Peterson, L. A., and Moschel, R. C. (2002) Mutagenesis by O6-[4-oxo-4-(3-pyridyl)butyl]guanine in Esherichi coli and human cells. Chem. Res. Toxicol. 15, 165-169. (16) Pegg, A. E. (2000) Repair of O6-alkylguanine by alkyltransferase. Mutat. Res. 462, 83-100. (17) Xu-Welliver, M., and Pegg, A. E. (2002) Degradation of the alkylated form of the DNA repair protein, O6-alkylguanine-DNA alkyltransferase. Carcinogenesis 23, 823-830. (18) Pegg, A. E., Wiest, L., Mummert, C., Stine, L., Moschel, R. C., and Dolan, M. E. (1991) Use of antibodies to human O6alkylguanine-DNA alkyltransferase to study the content of this protein in cells treated with O6-benzylguanine or N-methyl-N′nitroso-N-nitrosoguanidine. Carcinogenesis 12, 1679-1683. (19) Srivenugopal, K. S., Yuan, X. H., Friedman, H. S., and Ali-Osman, F. (1996) Ubiquitination-dependent proteolysis of O6-methylguanine-DNA methyltransferase in human and murine tumor cells following inactivation with O6-benzylguanine or 1,3-bis(2-chloroethyl)-1-nitrosourea. Biochemistry 35, 1328-1334. (20) Peterson, L. A., Liu, X. K., and Hecht, S. S. (1993) Pyridyloxobutyl DNA adducts inhibit the repair of O6-methylguanine. Cancer Res. 53, 2780-2785. (21) Liu, X. K., Spratt, T. E., Murphy, S. E., and Peterson, L. A. (1996) Pyridyoxobutylation of guanine residues by 4-(acetoxymethylnitrosamino)-1-(3-pyridyl)-1-butanone generates substrates of O6alkylguanine-DNA alkyltransferase. Chem. Res. Toxicol. 9, 949953. (22) Spratt, T. E., Peterson, L. A., Confer, W. L., and Hecht, S. S. (1990) Solvolysis of model compounds for R-hydroxylation of N′-nitrosonornicotine and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone: Evidence for a cyclic oxonium ion intermediate in the alkylation of nucleophiles. Chem. Res. Toxicol. 3, 350-356. (23) Carmella, S. G., Kagan, S. S., Spratt, T. E., and Hecht, S. S. (1990) Evaluation of cysteine adduct formation in rat hemoglobin by 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone and related compounds. Cancer Res. 50, 5453-5459. (24) Morimoto, K., Dolan, M. E., Scicchitano, D., and Pegg, A. E. (1985) Repair of O6-propylguanine and O6-butylguanine in DNA by O6alkylguanine-DNA alkyltransferases from rat liver and E. coli. Carcinogenesis 6, 1027-1031. (25) Pegg, A. E., Kanugula, S., Edara, S., Pauly, G. T., Moschel, R. C., and Goodtzova, K. (1998) Reaction of O6-benzylguanineresistant mutants of human O6-alkylguanine- DNA alkyltransferase with O6-benzylguanine in oligodeoxyribonucleotides. J. Biol. Chem. 273, 10863-10867. (26) Luu, K. X., Kanugula, S., Pegg, A. E., Pauly, G. T., and Moschel, R. C. (2002) Repair of oligodeoxyribonucleotides by O6-alkylguanine-DNA alkyltransferase. Biochemistry 41, 8687-8697. (27) Peterson, L. A., Spratt, T. E., Shan, W., Wang, L., Subotkowski, W., and Roth, R. (2001) An improved synthesis of radiolabeled 4-(acetoxymethylnitrosamino)-1-(3-pyridyl)-1-butanone. J. Labelled Compd. Radiopharm. 44, 445-450. (28) Dolan, M. E., Moschel, R. C., and Pegg, A. E. (1990) Depletion of mammalian O6-alkylguanine-DNA alkyltransferase activity by O6-benzylguanine provides a means to evaluate the role of this protein in protection against carcinogenic and therapeutic alkylating agents. Proc. Natl. Acad. Sci. U.S.A. 87, 5368-5372. (29) Pegg, A. E., Boosalis, M., Samson, L., Moschel, R. C., Byers, T. L., Swenn, K., and Dolan, M. E. (1993) Mechanisms of inactivation of human O6-alkylguanine-DNA alkyltransferase by O6-benzylguanine. Biochemistry 32, 11998-12006. (30) Crone, T. M., and Pegg, A. E. (1993) A single amino acid change in human O6-alkylguanine-DNA alkyltransferase decreasing sensitivity to inactivation by O6-benzylguanine. Cancer Res. 53, 4750-4753. (31) Crone, T. M., Goodtzova, K., Edara, S., and Pegg, A. E. (1994) Mutations in human O6-alkylguanine-DNA alkyltransferase imparting resistance to O6-benzylguanine. Cancer Res. 54, 62216227. (32) Mitsunobu, O. (1981) The use of diethyl azodicarboxylate and triphenylphosphine in synthesis and transformation of natural products. Synthesis 1-28. (33) Schulhof, J. C., Molko, D., and Teoule, R. (1987) The final deprotection step in oligonucleotide synthesis is reduced to a mild and rapid ammonia treatment by using labile base-protecting groups. Nucleic Acids Res. 15, 397-416. (34) Wang, L., Spratt, T. E., Pegg, A. E., and Peterson, L. A. (1999) Synthesis of DNA oligonucleotides containing site-specifically incorporated O6-[4-oxo-4-(3-pyridyl)butyl]guanine and their reaction with O6-alkylguanine-DNA alkyltransferase. Chem. Res. Toxicol. 12, 127-131. (35) Smith, C. A., Xu, Y. Z., and Swann, P. F. (1990) Solid-phase synthesis of oligodeoxynucleotides containing O6-alkylguanine. Carcinogenesis 11, 811-816.

Chem. Res. Toxicol., Vol. 17, No. 3, 2004 433 (36) Baumgart, P. M., Kliem, H. C., Gottfried-Anacker, J., Wiessler, M., and Schmeiser, H. H. (1993) Site-specific mutagenesis induced by single O6-alkylguanines (O6-n-propyl, O6-n-butyl, O6-n-octyl) in vivo. Nucleic Acids Res. 21, 3755-3760. (37) Pauly, G. T., Hughes, S. H., and Moschel, R. C. (1991) A sectored colony assay for monitoring mutagenesis by specific carcinogenDNA adducts in Escherichia coli. Biochemistry 30, 11700-11706. (38) Loktionova, N. A., and Pegg, A. E. (2002) Interaction of mammalian O6-alkylguanine-DNA alkyltransferases with O6-benzylguanine. Biochem. Pharmacol. 63, 1431-1442. (39) Xu-Welliver, M., and Pegg, A. E. (2000) Point mutations at multiple sites including highly conserved amino acids maintain activity, but render O6-alkylguanine-DNA alkyltransferase insensitive to O6-benzylguanine. Biochem. J. 347, 519-526. (40) Xu-Welliver, M., Kanugula, S., and Pegg, A. E. (1998) Isolation of human O6-alkylguanine-DNA alkyltransferase mutants highly resistant to inactivation by O6-benzylguanine. Cancer Res. 58, 1936-1945. (41) Mullapudi, S. R., Ali-Osman, F., Shou, J., and Srivenugopal, K. S. (2000) DNA repair protein O6-alkylguanine-DNA alkyltransferase is phosphorylated by two distinct and novel protein kinases in human brain tumour cells. Biochem. J. 351 (Pt. 2), 393-402. (42) Rasimas, J. J., Pegg, A. E., and Fried, M. G. (2003) DNA-binding mechanism of O6-alkylguanine-DNA alkyltransferase. Effects of protein and DNA alkylation on complex stability. J. Biol. Chem. 278, 7973-7980. (43) Goodtzova, K., Kanugula, S., Edara, S., Pauly, G. T., Moschel, R. C., and Pegg, A. E. (1997) Repair of O6-benzylguanine by the Escherichia coli Ada and Ogt and the human O6-alkylguanineDNA alkyltransferases. J. Biol. Chem. 272, 8332-8339. (44) Edara, S., Kanugula, S., Goodtzova, K., and Pegg, A. E. (1996) Resistance of the human O6-alkylguanine-DNA alkyltransferase containing arginine at codon 160 to inactivation by O6-benzylguanine. Cancer Res. 56, 5571-5575. (45) Graves, R. J., Li, B. F. L., and Swann, P. F. (1989) Repair of O6methylguanine, O6-ethylguanine, O6-isopropylguanine, O4-methylthymine in synthetic oligodeoxynucleotides by Escherichia coli ada gene O6-alkylguanine-DNA-alkyltransferase. Carcinogenesis 10, 661-666. (46) Liem, L. K., Lim, A., and Li, B. F. L. (1994) Specificities of human, rat and E. coli O6-methylguanine-DNA methyltransferases towards the repair of O6-methyl and O6-ethylguanine in DNA. Nucleic Acids Res. 22, 1613-1619. (47) Scicchitano, D., and Pegg, A. E. (1982) Kinetics of repair of O6methylguanine in DNA by O6-methylguanine-DNA methyltransferase in vitro and in vivo. Biochem. Biophys. Res. Commun. 109, 995-1001. (48) Georgiadis, P., Smith, C. A., and Swann, P. F. (1991) Nitrosamineinduced cancer: selective repair and conformational differences between O6-methylguanine residues in different positions in and around codon 12 of rat H-ras. Cancer Res. 51, 5843-5850. (49) Meyer, A. S., McCain, M. D., Fang, Q., Pegg, A. E., and Spratt, T. E. (2003) O6-alkylguanine-DNA alkyltransferases repair O6methylguanine in DNA with Michaelis-Menten-like kinetics. Chem. Res. Toxicol. 16, 1405-1409. (50) Delaney, J. C., and Essigmann, J. M. (2001) Effect of sequence context on O6-methylguanine repair and replication in vivo. Biochemistry 40, 14968-14975. (51) Hecht, S. S., Lin, D., and Castonguay, A. (1983) Effects of R-deuterium substitution on the mutagenicity of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK). Carcinogenesis 4, 305310. (52) Foiles, P. G., Peterson, L. A., Miglietta, L. M., and Ronai, Z. (1992) Analysis of mutagenic activity and ability to induce replication of polyoma DNA sequences by different model compounds of the carcinogenic tobacco-specific nitrosamine 4-(methylnitrosamino)1-(3-pyridyl)-1-butanone. Mutat. Res. 279, 91-101. (53) Pauly, G. T., Hughes, S. H., and Moschel, R. C. (1998) Comparison of mutagenesis by O6-methyl- and O6-ethylguanine and O4methylthymine in Escherichia coli using double-stranded and gapped plasmids. Carcinogenesis 19, 457-461. (54) Belinsky, S. A., Devereux, T. R., Maronpot, R. R., Stoner, G. D., and Anderson, M. W. (1989) Relationship between the formation of promutagenic adducts and the activation of the K-ras protooncogene in lung tumors from A/J mice treated with nitrosamines. Cancer Res. 49, 5305-5311. (55) Belinsky, S. A., Foley, J. A., White, C. M., Anderson, M. W., and Maronpot, R. R. (1990) Dose-response relationship between O6methylguanine formation in Clara cells and induction of pulmonary neoplasia in the rat by NNK. Cancer Res. 50, 3772-3780.

434

Chem. Res. Toxicol., Vol. 17, No. 3, 2004

(56) Staretz, M. E., Foiles, P. G., Miglietta, L. M., and Hecht, S. S. (1997) Evidence for an important role of DNA pyridyloxobutylation in rat lung carcinogenesis by 4-(methylnitrosamino)-1-(3pyridyl)-1-butanone: effects of dose and phenethyl isothiocyanate. Cancer Res. 57, 259-266. (57) Kaur, T. B., Travaline, J. M., Gaughan, J. P., Richie, J. P., Jr., Stellman, S. D., and Lazarus, P. (2000) Role of polymorphisms in codons 143 and 160 of the O6-alkylguanine DNA alkyltransferase gene in lung cancer risk. Cancer Epidemiol., Biomarkers Prev. 9, 339-342. (58) Ma, S., Egyhazi, S., Ueno, T., Lindholm, C., Kreklau, E. L., Stierner, U., Ringborg, U., and Hansson, J. (2003) O6-methylguanine-DNA-methyltransferase expression and gene polymorphisms in relation to chemotherapeutic response in metastatic melanoma. Br. J. Cancer 89, 1517-1523. (59) Imai, Y., Oda, H., Nakatsuru, Y., and Ishikawa, T. (1995) A polymorphism at codon 160 of human O6-methylguanine-DNA methyltransferase gene in young patients with adult type cancers and functional assay. Carcinogenesis 16, 2441-2445. (60) Deng, C., Xie, D., Capasso, H., Zhao, Y., Wang, L. D., and Hong, J. Y. (1999) Genetic polymorphism of human O6-alkylguanine-

Mijal et al.

(61) (62)

(63)

(64)

DNA alkyltransferase: identification of a missense variation in the active site region. Pharmacogenetics 9, 81-87. Grafstro¨m, R. C., Pegg, A. E., Trump, B. F., and Harris, C. C. (1984) O6-Alkylguanine-DNA alkyltransferase activity in normal human tissues and cells. Cancer Res. 44, 2855-2857. Myrnes, B., Giercksky, K. E., and Krokan, H. (1983) Interindividual variation in the activity of O6-methyl guanine-DNA methyltransferase adn uracil-DNA glycosylase in human organs. Carcinogenesis 4, 1565-1568. Strauss, B., Sagher, D., Schwartz, J., Karrison, T., and Larson, R. (1989) Heterogeneity in the O6-alkylguanine DNA alkyltransferase (AGT) activity of human peripheral blood lymphocytes (PBL’s). In DNA Repair Mechanisms and Their Biological Implications in Mammalian Cells (Lambert, M. W., and Laval, J., Eds.) p 618, Plenum Press, New York. Gerson, S. L., Schupp, J., Liu, L., Pegg, A. E., and Srinivasen, S. (1999) Leukocyte O6-alkylguanine-DNA alkyltransferase from human donors is uniformly sensitive to O6-benzylguanine. Clin. Cancer Res. 5, 521-524.

TX0342417