Kinetics of Nucleotide Incorporation Opposite Polycyclic Aromatic

Hong Zang, Goutam Chowdhury, Karen C. Angel, Thomas M. Harris, and F. Peter Guengerich. Chemical Research in Toxicology 2006 19 (6), 859-867...
0 downloads 0 Views 519KB Size
Chem. Res. Toxicol. 2005, 18, 389-400

389

Kinetics of Nucleotide Incorporation Opposite Polycyclic Aromatic Hydrocarbon-DNA Adducts by Processive Bacteriophage T7 DNA Polymerase Hong Zang,†,‡ Thomas M. Harris,§ and F. Peter Guengerich*,† Department of Biochemistry and Department of Chemistry and Center in Molecular Toxicology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232-0146 Received November 15, 2004

A series of six oligonucleotides with dihydrodiol epoxide metabolites of the polycyclic aromatic hydrocarbons (PAHs) benz[a]anthracene and benzo[a]pyrene attached to adenine N6 and guanine N2 atoms were prepared and studied with the processive bacteriophage DNA polymerase T7, exonuclease- (T7-). HIV-1 reverse transcriptase was much less efficient in polymerization than T7-. Benz[a]anthracene and benzo[a]pyrene adducts strongly blocked incorporation of dTTP and dCTP opposite the A and G derivatives, respectively. dATP was preferentially incorporated in all cases. Steady state kinetic analysis indicated that the low catalytic efficiency with adducted DNA was due to both increased Km and lowered kcat values. Some differences due to PAH stereochemistry were observed. Fluorescence estimates of Kd and presteady state kinetic measurements of koff showed no major decrease in the affinity of T7- with damaged DNA substrates or with dNTPs. Presteady state kinetics showed a lack of the normal burst kinetics for dNTP incorporation with all PAH-DNA derivatives. These results indicate that the rate-limiting step is at or before the step of phosphodiester bond formation; release of the oligonucleotide is no longer the slowest step. Thio elemental effects (substitution of R-oxygen with sulfur) were relatively small, in contrast to previous work with T7- and 8-oxo7,8-dihydroguanine. The effect of these bulky PAH adducts is either to attenuate rates of conformational changes or to introduce an additional conformation problem but not to alter the inherent affinity of the polymerase for DNA or dNTPs.

Introduction The formation of DNA lesions is the initial step for a series of biologically important events including mutation, DNA repair, and apoptosis (1-3). Covalent DNA adduct lesions can be generated by exogenous or endogenous reactive compounds. Polycyclic aromatic hydrocarbons (PAHs)1 are widespread environmental contaminants present in tobacco smoke, food, and automobile exhaust (4-8). The reactive 7,8-diol 9,10-epoxides of benzo[a]pyrene modify the exocyclic amino groups of purines in DNA to generate N2-guanine and N6-adenine adducts (2, 7, 9). Research has provided evidence that tobacco-associated lung cancer in humans may be related, at least in part, to mutation of the p53 tumor suppressor gene by benzo[a]pyrene products (10-12). DNA replication can be blocked by the bulky benzo[a]pyrene-DNA adducts. Recent studies suggest that translesion synthesis polymerases, most of which are Y-family polymerases, can bypass these bulky DNA adducts in both error-prone and error-free manners (13-18). Pol κ has been reported to be the most efficient polymerase to incorporate dNTPs opposite benzo[a]pyrene-adducted bases (19). Such DNA adduct bypass process requires the recruitment of the translesion polymerase. The mecha* To whom correspondence should be addressed. Tel: 615-322-2261. Fax: 615-322-3141. E-mail: [email protected]. † Department of Biochemistry and Center in Molecular Toxicology. ‡ Recipient of a fellowship from Merck Research Laboratories. § Department of Chemistry and Center in Molecular Toxicology.

nism of cooperation between a replicative polymerase and a translesion polymerase is not clear yet (15, 20). Because the first encounter of the replicative polymerase and the adducted DNA substrate is the initial step for this overall process, understanding the kinetics of the processive polymerase and the chemically altered DNA substrate is important. We are interested in defining which of these conclusions we have reached about the interaction of processive polymerases with other DNA adducts (21-28) are applicable to the larger, bulky PAHs. The availability of 1 Abbreviations: PAH, polycyclic aromatic hydrocarbon; dCTPRS, 2′-deoxycytidine 5′-O-(1-thiotriphosphate); dATPRS, 2′-deoxyadenosine 5′-O-(1-thiotriphosphate); dTTPRS, 2′-deoxythymidine 5′-O-(1-thiotriphosphate); Tris, tris(hydroxylmethyl)aminomethane; BSA, bovine serum albumin; DTT, dithiothreitol; FAM, 6-carboxyfluorescein; T7, bacteriophage T7 DNA polymerase (T7- indicates exonuclease deficient); RT, human immunodeficiency virus-1 reverse transcriptase; KF, Escherichia coli polymerase I Klenow fragment (KF- indicates exonuclease deficient); TBE, 90 mM Tris-borate buffer (pH 8.5) containing 2 mM EDTA; MS, mass spectrometry; MALDI-TOF, matrix-assisted laser desorption ionization mass spectrometry with time-of-flight analyzer; CGE, capillary gel electrophoresis; dA-ATBA-1S, anti-trans(1S,2R,3S,4S)-2′-deoxy-N6-(1,2,3,4-tetrahydro)-2,3,4-trihydroxybenz[a]anthracene-1-yl-adenosine; dA-ATBA-11S, anti-trans-(8R,9S,10R,11S)2′-deoxy-N6-(8,9,10,11-tetrahydro)-8,9,10-trihydroxybenzo[c]phenanthrene-11-yl-adenosine; dA-ACBP-10R, anti-cis-(7R,8S,9R,10R)-2′deoxy-N6-(7,8,9,10-tetrahydro)-7,8,9-trihydroxy-benzo[a]pyrene-10-yladenosine; dA-STBP-10S, syn-trans-(7S,8R,9R,10S)-2′-deoxy-N6-(7,8,9,10tetrahydro)-7,8,9-trihydroxy-benzo[a]pyrene-10-yl-adenosine; dG-ATBP10R, anti-trans-(7S,8R,9S,10R)-2′-deoxy-N2-(7,8,9,10-tetrahydro)-7,8,9trihydroxybenzo[a]pyrene-10-yl-guanosine; dG-ATBP-10S, anti-trans(7R,8S,9R,10S)-2′-deoxy-N2-(7,8,9,10-tetrahydro)-7,8,9-trihydroxybenzo[a]pyrene-10-yl-guanosine. The designations for nucleic acid bases and nucleosides are standard for this journal.

10.1021/tx049683c CCC: $30.25 © 2005 American Chemical Society Published on Web 02/02/2005

390

Chem. Res. Toxicol., Vol. 18, No. 2, 2005 Scheme 1. General Scheme of Polymerase Catalysis (Minimal Mechanism)

several of these adducts from a series of previous studies (29-37) provides an opportunity to study polymerization reactions with bacteriophage T7 DNA polymerase (T7, T7- indicates exonuclease deficient). Some work was done with human immunodeficiency virus-1 reverse transcriptase (RT), although the low catalytic efficiency with the substrates containing PAH adducts precludes many experiments. In this study, we address several questions in a series of systematic experiments with these modified oligonucleotides and the processive polymerase T7-: (i) What are the effects of the adducts on full-length extension and one-base insertion preferences? (ii) What are the effects of these adducts on steady state kinetics of one-base insertion and how do these effects compare with each other? (iii) Are kinetic bursts still seen; that is, does the rate-limiting step in one-base insertion still occur after phosphodiester bond formation (Scheme 1)? (iv) Are kinetic elemental effects (due to sulfur substitution for R-oxygen in dNTP) seen and, if so, can these be interpreted in the terms of rate-limiting steps? (v) Does the presence of a bulky adduct raise koff and Kd for the DNA or the dNTP? The work focus on studies with six PAHs (Scheme 2), two G and four A adducts.

Experimental Procedures Chemicals. Chemicals and biochemicals were obtained from the following sources: ampicillin, chloramphenicol, lysozyme, poly(ethyleneimine), yeast extract (Sigma, St. Louis, MO); isopropyl-β-D-thiogalactopyranoside (Anatrace, Maumee, OH); casein hydrolysate (ICN, Aurora, OH); 2′-deoxycytidine 5′-O(1-thiotriphosphate) (dCTPRS) and 2′-deoxyadenosine 5′-O-(1thiotriphosphate) (dATPRS) (Amersham, Piscataway, NJ); 2′deoxythymidine 5′-O-(1-thiotriphosphate) (dTTPRS) (IBA, Go¨ttingen, Germany); [γ-32P]ATP (NEN Life Science, Boston, MA). All oligonucleotides were purchased from Midland Certified Reagents (Midland, TX) or Operon Technologies (Alameda, CA), except the adducted 11-mer or 13-mer oligonucleotides, which were synthesized in the Vanderbilt facility using general phosphoramidite technology (38-42). These oligonucleotides were either purified by HPLC or denaturing gel electrophoresis before use. 6-Carboxyfluorescein (FAM)-tagged and 3′-substituted [3-(3′-hydroxypropyl) to prevent elongation] oligonucleotides were purchased from Midland. Polymerases. RT was expressed and purified in this laboratory as previously described (21) using stock plasmids provided by S. Hughes (Frederick Cancer Facility, Frederick, MD). Protein concentrations were estimated using an 280 value of 522 mM-1 cm-1 (23). The polymerase T7- was expressed and purified using a new procedure (28), based on earlier work by the laboratory of K. A. Johnson (43) using the plasmid encoding T7- DNA polymerase, pG5X (Amp). Protein concentrations were determined using 280

Zang et al. Scheme 2. Chemical Structures of the PAH-DNA Adducts

) 144 mM-1 cm-1 for T7-. The overall yield of purified T7protein was about 1 mg L-1. E. coli thioredoxin was expressed and purified as described previously (21). Ligation of Carcinogen-Adducted Oligonucleotides and Purification. Oligonucleotide mixtures (18-mer) and (11-mer, containing the adduct) were ligated using a uracil-containing scaffold (33-mer), with HPLC to separate the product, as described in detail elsewhere (28). The purified oligonucleotides were characterized by matrix-assisted laser desorption ionization mass spectrometry (MS) with time-of-flight analyzer (MALDI-TOF), capillary gel electrophoresis (CGE), and UV spectroscopy (27, 28) (see Supporting Information). All electrophoretograms indicated >96% purity, and most of the extra peaks are attributed to random noise. In no case do we feel that any impurities substantially affected the conclusions we reached. Oligonucleotide concentrations were estimated from UV spectra (in H2O), which were recorded using a Cary 14/OLIS spectrophotometer (On-Line Instrument Systems, Bogart, GA). The extinction coefficients for the oligonucleotides determined by the Borer method (44) were as follow (Scheme 3): oligo 1, 23-mer, 260 ) 199 mM-1 cm-1; oligo 2, 40-mer, 260 ) 379 mM-1 cm-1; oligo 3, 25-mer, 260 ) 224 mM-1 cm-1; oligo 4, 42-mer, 260 ) 383 mM-1 cm-1. Labeling and annealing were done as described elsewhere (27, 28). Polymerization Assays and Gel Electrophoresis. A 32Plabeled primer, annealed to either an unmodified or adducted template, was extended in the presence of all four dNTPs (Scheme 3). Each reaction was initiated by adding 2 µL of dNTP-Mg2+ solution (final concentrations of 200 µM of each dNTP and 5 mM MgCl2) to a preincubated E‚DNA complex (final concentrations of 50 mM tris(hydroxylmethyl)aminomethane (Tris)-HCl (pH 7.8), 100 nM or 200 nM DNA duplex, 1 mM dithiothreitol (DTT), 50 µg bovine serum albumin (BSA) mL-1, 50 mM NaCl, and 10% glycerol (v/v), with concentrations of polymerase indicated in each figure) at 25 °C, yielding a total reaction volume of 8 µL (27, 28). After 30 min, reactions were quenched with 50 µL of 20 mM EDTA (pH 9.0) in 95%

Polymerases and Polycyclic Hydrocarbon-DNA Adducts

Chem. Res. Toxicol., Vol. 18, No. 2, 2005 391

Scheme 3. Oligonucleotides Used in These Studies

formamide (v/v). Aliquots (2 µL) were separated by electrophoresis on a denaturing gel containing 8.0 M urea and 16% acrylamide (w/v) (from a 19:1 acrylamide:bisacrylamide solution, AccuGel, National Diagnostics, Atlanta, GA) with 90 mM Trisborate buffer (pH 8.5) containing 2 mM EDTA (TBE buffer). The gel was exposed to a phosphorimager screen (Imaging Screen K, Bio-Rad, Hercules, CA) overnight. The bands, representing extension of the primer, were visualized with a phosphorimaging system (Bio-Rad, Molecular Imager FX) using Quantity One Software, Version 4.3.0. Steady State Kinetics. Unless indicated otherwise, all T7(or RT) reactions were performed at 25 °C in 50 mM Tris-HCl buffer (pH 7.8) containing 10% glycerol (v/v), 1 mM DTT, 50 mM NaCl, and 50 µg BSA mL-1. For unmodified templates, the molar ratio of primer/template to enzyme was 40:1 or higher, and the reactions were done at six dNTP concentrations (for a reaction time of 2 min). For modified templates, the molar ratio of primer/template to enzyme was 10:1, and reactions were done at nine dNTP concentrations (for reaction times of 20-30 min). For each experiment, T7- was reconstituted with thioredoxin (20-fold molar excess) prior to use. Thioredoxin was prepared separately in the same buffer containing 5 mM DTT. Presteady State Kinetics. Presteady state kinetics were performed using a model RQF-3 KinTek chemical quench flow apparatus (KinTek Corp., Austin, TX) with 50 mM Tris-HCl (pH 7.8) buffer containing 1 mM MgCl2 in the drive syringes. The reactions were initiated by mixing the preequilibrated polymerase-DNA complex (containing 100 mM Tris-HCl, pH 7.8, 200 nM 32P-labeled DNA duplex, 2 mM DTT, 100 µg BSA mL-1, 50 nM T7-, 50 mM NaCl, and 5% glycerol, v/v) in sample syringe A (12.5 µL) with either dNTP-Mg2+ or (Sp)-dNTPRS-Mg2+ (40400 µM, depending on oligonucleotide sequence) and 5 mM MgCl2 in syringe B (10.9 µL) at 25 °C. The reactions were quenched with 0.6 M EDTA in syringe C (pH 9.0) after reaction times of 5 ms to 10 s. Reactions were combined with 500 µL of formamide-dye solution and the components were separated by electrophoresis (2 µL of the sample) on a denaturing gel, with analysis as described for the gel electrophoresis of polymerization reactions. The resulting plot (of product vs time) was fit to the burst equation y ) A(1 - e-kpt) + ksst, where A ) burst amplitude, kp ) presteady state rate of nucleotide incorporation, t ) time, and kss ) steady state rate of nucleotide incorporation and analyzed using GraphPad Prism version 3.0a (San Diego, CA). Estimation of Kd for Oligonucleotide- and dNTPPolymerase Binding. Ground state binding (Kd) of the DNA substrate to T7- was estimated by fluorescence titration. T7(2.5-320 nM final) was added to a solution of 100 nM 17-FAMmer/40-mer or 19-FAM-mer/42-mer in Tris-HCl buffer (50 mM, pH 7.8, 25 °C, 5 mM MgCl2 , 5% glycerol (v/v), 1 mM DTT, 50 mM NaCl, and 100 µg BSA mL-1). Fluorescence was monitored using a Varian SF-330 spectrofluorimeter (Varian, Walnut Creek, CA) with an excitation wavelength of 492 nm and an emission wavelength of 516 nm. After adjustment by dilution factors for enzyme addition, the data were fit to a fluorescence quadratic equation in Prism using the equation (E‚DNA) ) 0.5-

(Kd + Dt + Et) - [0.25 (Kd + Dt + Et)2 - (Dt × Et)]1/2 where Et ) total enzyme concentration, Kd ) DNA dissociation constant from E‚DNA, and Dt ) total DNA concentration (24, 45, 46). dNTP binding was analyzed in a similar way, except that the T7- (( oligonucleotide) with titrated with dNTP solutions and the intrinisic tryptophan fluorescence of the T7- was monitored (24). Estimation of Rates of DNA Dissociation from Enzyme. The DNA dissociation rate from E‚DNA (koff) was determined using a rapid quench flow apparatus and a previously described approach (26-28, 47, 48). Sample syringe A (12.5 µL) contained a preincubated solution of T7- (350 nM) in Tris-HCl buffer (50 mM, pH 7.8) along with 50 mM NaCl, 5% glycerol (v/v), 1 mM DTT, 100 µg BSA mL-1, 1 mM MgCl2, and unlabeled target DNA (50 nM). Sample from syringe A was mixed with 32P-labeled 24mer/42-G-mer (450 nM) from sample syringe B (10.9 µL) containing Tris-HCl buffer (50 mM, pH 7.8), 5% glycerol (v/v), 1 mM DTT, 100 µg BSA mL-1, and 10 mM MgCl2 at time intervals ranging from 0.05 to 30 s (25 °C). After the samples were mixed and incubated for varying times, polymerization was initiated by the addition of 200 µM dNTP and 5 mM MgCl2 from the central drive syringe, with a constant reaction time of 0.25 s. After the sample was expelled from the rapid quench apparatus, the reaction was mixed rapidly (using a vortex device) into a tube containing 500 µL of 0.3 M EDTA (in 50% formamide, v/v) to stop the reaction. The amount of the incorporated product was quantified by gel analysis and plotted against time. The graph was fit to a single-exponential equation in GraphPad Prism using the equation y ) Ef + E0(1 - e-kofft) where Ef ) free enzyme concentration, E0 ) DNA-bound enzyme concentration, and koff ) dissociation rate of DNA from E‚DNA.

Results Strategy. The replicative polymerases T7- and RT have been used as model polymerases in the study of polymerization reactions on damaged DNA. Both of these enzymes can be readily expressed and purified, and there is a considerable body of kinetic (21-26) and structural (49-53) information available for these two enzymes. The following PAH adducts were selected for use in this study (Scheme 2): anti-trans-(8R,9S,10R,11S)-2′-deoxy-N6(8,9,10,11-tetrahydro)-8,9,10-trihydroxybenzo[c]phenanthrene-11-yl-adenosine (dA-ATBA-11S) has one benzene ring less than the related benzo[a]pyrene-dA adduct syntrans-(7S,8R,9R,10S)-2′-deoxy-N6-(7,8,9,10-tetrahydro)7,8,9-trihydroxy-benzo[a]pyrene-10-yl-adenosine (dASTBP-10S); anti-trans-(1S,2R,3R,4S)-2′-deoxy-N6-(1,2,3,4tetrahydro)-2,3,4-trihydroxybenz[a]anthracene-1-yladenosine (dA-ATBA-1S) is a bay-region regioisomer compared to dA-ATBA-11S. dA-STBP-10S and anti-cis(7R,8S,9R,10R)-2′-deoxy-N6-(7,8,9,10-tetrahydro)-7,8,9trihydroxy-benzo[a]pyrene-10-yl-adenosine (dA-ACBP10R) provide adenine-N6 adducts as two stereoisomers

392

Chem. Res. Toxicol., Vol. 18, No. 2, 2005

Zang et al.

Figure 1. (A) Extension of a 32P-labeled primer opposite the PAH adduct by T7- in the presence of all four dNTPs. Reactions were done in the presence of the indicated polymerase concentrations with 200 nM primer/template and 200 µM dNTP for 30 min. (B) Extension of a 32P-labeled primer (200 nM) opposite the adduct by T7- (50 nM) in the presence of single dNTPs (200 µM). All studies were done with oligonucleotides 1 and 2 (Scheme 3), with the bold A of oligonucleotide 2 changed as indicated.

with different chirality at C-7, C-9, and C-10. syn-trans(7S,8R,9R,10S)-2′-Deoxy-N6-(7,8,9,10-tetrahydro)-7,8,9trihydroxy-benzo[a]pyrene-10-yl-adenosine (dG-ATBP10R) and anti-trans-(7R,8S,9R,10S)-2′-deoxy-N2-(7,8,9,10tetrahydro)-7,8,9-trihydroxybenzo[a]pyrene-10-yl-guanosine (dG-ATBP-10S) guanine-N2 adducts were chosen for comparison to dA-N6 adducts; these two PAH systems are diastereoiomeric with each other. Thus, these choices cover a scope of DNA damage, and these bulky adducts can also be used as probes to detect enzyme-DNA interactions. The oligonucleotide sequence contexts (Scheme 3) were chosen because the modified site is at codon 61 (for dA) or 12 (dG) of the N-ras gene. Carcinogenbound 11-mer and 13-mer oligonucleotides had been synthesized previously using phosphoramidite chemistry (38-42). However, short oligonucleotides are poor substrates for polymerases, particularly the processive polymerases T7- and RT (21, 43, 54). Therefore, extending the adducted DNA to lengths of 35-40 nucleotides was necessary for doing the polymerization reactions. Each purified oligonucleotide was characterized by CGE for purity and by MALDI-TOF MS for identity (see Supporting Information). Polymerization Assays in the Presence of All Four dNTPs. Processive polymerization of T7- and RT was evaluated in the presence of all four dNTPs, beginning at the site opposite the adduct (see sequence context in Scheme 3). In the case of T7- with the benz[a]anthracene-derived dA-N6 adducts, only one-base incorporation bands were observed even with high concentrations of polymerase (Figure 1A). There were no detectable bands for full-length products with any adduct. The

incorporated base was primarily A in all cases, although G incorporation was seen for the dA-ATBA-1S and dAATBA-11S adducts (Figure 1B). With high concentrations of dNTPs, mispaired bases (C, A, or G) were incorporated by T7- even in the unmodified template. In contrast to T7-, RT showed no single nucleotide incorporation or full-length product for the benzo[a]pyrene-derived dA adducts (Figure 2). With the dG-N2-benzo[a]pyrene-based adducts (dGATBP-10R and dG-ATBP-10S), T7- did not generate fulllength products but yielded only one-base incorporation products resulting from dATP incorporation (Figure 3A,B). RT was blocked by the benzo[a]pyrene-derived dG adducts, but less digestion of the oligonucleotides was observed than with the A adducts (Figure 4A,B). Steady State Kinetics of One-Base Incorporation by T7-. Steady state parameters were measured for onebase incorporation opposite the adducts. The kinetic analysis was done for incorporation of the correct base compared to the most efficient misincorporation, which was insertion of dATP opposite dA-N6 adducts (Table 1). For correct base pair incorporation (T), the Km was increased by 3-4 orders of magnitude for the dA-ATBA1S and dA-ACBP-10R adducts relative to the unmodified template; kcat was decreased by 2 orders of magnitude. There was no detectable incorporation of dTTP opposite the dA-ATBA-11S or the dA-STBP-10S adduct. The differences in incorporation efficiency (kcat/Km) were primarily due to changes in the Km values. The misinsertion frequency (f), defined as (kcat/Km)incorrect/(kcat/ Km)correct, reflects the preference for the incorrect nucle-

Polymerases and Polycyclic Hydrocarbon-DNA Adducts

Chem. Res. Toxicol., Vol. 18, No. 2, 2005 393

Figure 2. (A) Extension of a 32P-labeled primer opposite the PAH adduct by RT in the presence of all four dNTPs. Reactions were done in the presence of the indicated polymerase concentrations with 200 nM primer/template and 200 µM dNTP for 30 min. (B) Extension of a 32P-labeled primer (200 nM) opposite the adduct by RT (50 nM) in the presence of single dNTPs (200 µM). All studies were done with oligonucleotides 1 and 2 (Scheme 3), with the bold A of oligonucleotide 2 changed as indicated.

Figure 3. (A) Extension of a 32P-labeled primer opposite the adducts by T7- in the presence of all four dNTPs. Reactions were done in the presence of the indicated polymerase concentrations with 100 nM primer/template and 200 µM dNTP for 30 min. (B) Extension of a 32P-labeled primer (100 nM) opposite each adduct by T7- (25 nM) in the presence of single dNTPs (200 µM). All studies were done with oligonucleotides 3 and 4 (Scheme 3), with the bold G of oligonucleotide 4 changed as indicated.

otide incorporation with the adducts (55). For the dAATBA-11S and dA-STBP-10S adducts, the f values were >250, indicating that misincorporation of A is highly preferred opposite these adducts. The dA-ATBA-1S and dA-ACBP-10R adducts gave relatively small f values but still significant misincorporation. For the benzo[a]pyrene-derived dG adducts, similar trends were observed (Table 2). Incorporation of dCTP opposite the adducts was decreased dramatically due to both kcat and Km effects. Misincorporation of dATP was favored opposite the adducts. The misincorporation efficiency (kcat/Km) for the dG-ATBP-10R adduct was 14 times greater than for the dG-ATBP-10S adduct. Thus, dG-ATBP-10R adduct is a lesion that is apparently much

Figure 4. (A) Extension of a 32P-labeled primer opposite each adduct by RT in the presence of all four dNTPs. Reactions were done in the presence of the indicated polymerase concentrations with 100 nM primer/template and 100 µM dNTP for 20 min. (B) Extension of a 32P-labeled primer (100 nM) opposite each adduct by RT (25 nM) in the presence of single dNTPs (100 µM). All studies were done with oligonucleotides 3 and 4 (Scheme 3), with the bold G of oligonucleotide 4 changed as indicated.

easier for T7- to accommodate than the dG-ATBP-10S adduct. Misinsertion values (f) were 160 for the dGATBP-10R adduct and 2.4 for the dG-ATBP-10S adduct. Determination of DNA-T7- Dissociation Constants. Fluorescence titrations with T7- and solutions of FAM-oligonucleotide primer/template mixture were used to determine the ground state binding of DNA substrates to T7- because the fluorescent group (FAM) is sensitive to T7- binding of the DNA substrate. Only minor differences (e3-fold) were observed between Kd values of the unmodified and the adducted DNA substrates, using experiments in which the adducts (in the template) were not paired with any bases in the primer (Table 3). Thus, the presence of bulky adducts at the dAN6 or dG-N2 position did not cause major disfavored overall contacts with T7-. We also considered the possibility that the presence of a base (paired or mispaired) in the primer opposite a

394

Chem. Res. Toxicol., Vol. 18, No. 2, 2005

Zang et al.

Table 1. Steady State Kinetic Parameters for One-Base Incorporation template

Km (µM)

kcat (s-1) × 10-3

dA dA-ATBA-1S dA-ATBA-11S dA-STBP-10S dA-ACBP-10R dA dA-ATBA-1S dA-ATBA-11S dA-STBP-10S dA-ACBP-10R

0.074 ( 0.001 38 ( 9

80 ( 2 0.80 ( 0.05

350 ( 85 120 ( 37 75 ( 8 37 ( 18 9.4 ( 0.8 52 ( 9

0.42 ( 0.05 0.34 ( 0.03 0.68 ( 0.02 1.0 ( 0.1 0.46 ( 0.01 0.98 ( 0.08

dNTP inserted T

A

a

kcat/Km × 10-3 1.1 × 0.021 450 16

f (misinsertion frequency) ) (kcat/Km)incorrect/(kcat/Km)correct. Table 2. Steady State Kinetic Parameters for One-Base Incorporation dNTP inserted

template

Km (µM)

kcat (s-1) × 10-3

kcat/Km × 10-3

fa

C

dG dG-ATBP-10R dG-ATBP-10S dG dG-ATBP-10R dG-ATBP-10S

0.19 ( 0.06 61 ( 36

58 ( 6 0.0087 ( 0.0013

95 ( 14 104 ( 16 270 ( 66

2.2 ( 0.1 2.4 ( 0.2 0.44 ( 0.13

310 0.00014 2.4

A

a

f (misinsertion frequency) ) (kcat/Km)incorrect/(kcat/Km)correct. Table 3. T7- ‚ DNA Dissociation Constants

series

template

Kd (nM)

∆ fluorescence amplitudea

dA-BA, BP

dA dA-ATBA-1S dA-ATBA-11S dA-STBP-10S dA-ACBP-10R dG dG-ATBP-10R dG-ATBP-10S

14 ( 3 24 ( 10 14 ( 2 43 ( 13 25 ( 6 56 ( 7 57 ( 13 131 ( 46

16.5 19.1 15.0 11.1 11.2 16.5 21.1 22.9

dG-ATBP

a

Table 5. Rates of DNA Dissociation from DNA ‚ T7series

template

koff (s-1)

dA-BA, BP

dA dA-ATBA-1S dA-ATBA-11S dA-STBP-10S dA-ACBP-10R dG dG-ATBP-10R dG-ATBP-10S

0.78 ( 0.14 1.8 ( 0.2 0.48 ( 0.11 1.8 ( 0.2 1.4 ( 0.3 0.83 ( 0.16 0.70 ( 0.10 0.88 ( 0.12

dG-ATBP

Table 6. dNTP Dissociation Constants with DNA ‚ T7-

Extrapolated to saturating concentration.

Table 4. T7- ‚ DNA Dissociation Constants with Paired Adducts dNTP

adduct

Kd,dNTP,apparent (nM)

∆ fluorescence amplitude a

dCTP

dG dG-ATBP-10R dG-ATBP-10S dG dG-ATBP-10R dG-ATBP-10S

19.2 ( 4.1 5.6 ( 0.8 3.3 ( 0.1 8.9 ( 0.7 1.9 ( 0.1 1.2 ( 0.1

98 106 74 97 83 112

dATP

dG* paired with dC dG* paired with dA

a

adduct

Kd (nM)

∆ fluorescence amplitudea

dG dG-ATBP-10R dG-ATBP-10S dG dG-ATBP-10R dG-ATBP-10S

56 ( 7 183 ( 52 89 ( 7 100 ( 7 188 ( 72 133 ( 14

16.5 6.7 38.5 24.3 6.2 16.9

Extrapolated to saturating concentration.

bulky adduct in the template might reduce the affinity of the oligonucleotide for the T7-, in light of the proposals of others (56-58). Either C or A was placed opposite either a dG-ATBP-10R or -10S-modified oligonucleotide, and binding to T7- was examined by fluorescence titration methods (Table 4). Again, the maximum difference in apparent affinity was only ∼3-fold. Estimation of DNA Dissociation Rate, koff. The dissociation rate of E‚DNA complex was estimated using a rapid quench experiment, as outlined in the Experimental Procedures. In comparison to the unmodified DNA substrates, koff values were slightly different with

a

Extrapolated to saturating concentration.

the dA- and dG-adducted oligonucleotides (Table 5). These results indicate that the dissociation rate of the DNA-enzyme complex is not affected much by the bulky adducts and are consonant with the results of the experiments presented in Tables 3 and 4. Determination of DNA-T7- Dissociation Constants (Kd,dNTP,apparent). One potential explanation for the high Km values (for dNTP) seen in the one-base incorporation studies with the PAH adducts is that the polymerase-oligonucleotide complexes lose affinity for dNTPs upon modification. Fluorescence titrations were done with dNTPs and T7-‚primer/template complexes in which the extension of the primer was blocked with a 3-(3′hydroxypropyl) group (Table 6). (The 3′-(3-hydroxy)propyl modification is marketed by commercial oligonucleotide synthesis laboratories as a nucleoside terminator.) The apparent affinity was increased for both dCTP and dATP when the bulky PAH adducts (dG-ATBP-10R and -10S) were present on the oligonucleotides (up to 5-fold) (Table

Polymerases and Polycyclic Hydrocarbon-DNA Adducts

Chem. Res. Toxicol., Vol. 18, No. 2, 2005 395

Table 7. Presteady State Kinetics and Phosphorothioate Analysis

effect should reflect primarily the rate of phosphodiester bond formation. The initial linear portion of the one-base incorporation reaction was utilized for comparisons, and the elemental effect ranged from 1.3 to 6.8 for the dA adducts. In case of the dG-N2 adducts, the elemental effects were 2.6 for the unmodified DNA substrate, 4.1 for dGATBP-10R, and 8.2 for dG-ATBP-10S. Thus, both the dAN6 and the dG-N2 adducts showed relatively modest elemental effects, suggesting that rates of phosphodiester bond formation may not be strongly affected by the presence of bulky adducts.

Rapid Quench

dA (unmodified)

+ dTTP kp (s-1)

+ dTTPRS kp (s-1)

elemental effecta

34 ( 5

20 ( 1

1.7

Steady State

dA-ATBA-1S dA-ATBA-11S dA-STBP-10S dA-ACBP-10R

+ dATP k, × 10-3 (s-1)b

+ dATPRS k, × 10-3 (s-1)b

elemental effecta

0.24 1.1 0.56 6.8

0.18 0.31 0.12 1.0

1.3 3.6 4.5 6.8

Rapid Quench

dG (unmodified)

+ dTTP kp (s-1)

+ dTTPRS kp (s-1)

elemental effecta

10 ( 1

3.9 ( 0.9

2.6

Steady State

dG-ATBP-10R dG-ATBP-10S

+ dATP k, × 10-3 (s-1)b

+ dATPRS k, × 10-3 (s-1)b

elemental effecta

1.5 0.10

0.36 0.012

4.1 8.3

a Rate with dNTP divided by rate with dNTPRS. b Calculated from slope (Figures 5 and 6).

6). These values are similar to other reported previously from this laboratory (27, 28, 45)2 and probably reflect the equilibria of both steps 2 and 3 of Scheme 1; that is, Kd,dNTP,apparent ) (k-2/k2)/[k-3/(k3 + k-3)] (59). Presteady State Kinetics and Phosphorothioate Analysis of dNTP Incorporation. Steady state analysis is useful for evaluation of nucleotide incorporation efficiency, but detailed mechanistic information is buried in the complex polymerization cycle. Presteady state analysis was carried out using a rapid quench instrument, with a focus on the initial reaction cycle. Incorporation of dTTP into unmodified substrate by T7- gave a burst rate of kp ) 34 ( 5 s-1, as determined from the burst equation y ) A(1 - e-kpt) + ksst (Table 7). The burst phase was completely abolished with all dA-N6- and dGN2-adducted oligonucleotides used as substrates (Figures 5 and 6). This lack of a burst indicates that the ratelimiting step is an enzyme conformational change or phosphodiester bond formation, as opposed to oligonucleotide dissociation from the enzyme (43) (Scheme 1). The issue of whether the phosphodiester bond formation step is rate-limiting can be addressed using nucleotide phosphorothioates (dNTPRS) as substrates and comparison to normal dNTPs. The replacement of the R-oxygen by sulfur did not affect the dNTP binding affinity to an DNA-enzyme complex in other work (60), suggesting that differences in rates should reflect the elemental effect (increased bond strength due to the electronegativity difference) during the phosphodiester bond formation. The elemental effect on kpol was 1.7 for the unmodified dA substrate, consonant with earlier work (21, 26). With the adducts, misincorporation of dATPRS was examined to determine the thio (sulfur) elemental effect. Assuming that the DNA-enzyme dissociation rate is similar for each adduct (Table 5), the thio elemental 2 Because of a printing error, the values presented in Tables 3 and 5 of ref 27 were labeled “µM” instead of the intended “nM,” and a correction has been published [(2004) J. Biol. Chem. 279, 29870].

Discussion Six oligonucleotides modified with PAHs were prepared and included three sets of isomers. All of the adducts strongly blocked the polymerization reactions with both polymerases T7- and RT. Polymerization was generally much more efficient with T7- than RT for these adducts. In terms of misincorporation opposite the adduct, A was the preferred base rather than T or C, and the incorporation efficiency was decreased due to both elevated Km,dNTP and reduced kcat values (Tables 1 and 2). There were no significant increases in Kd,dNTP,apparent or Kd,DNA or koff for the adducted oligonucleotides as compared to the unmodified DNA substrates (Tables 3-5). This latter result contrasts with reports that PAH adducts reduce the affinity of polymerases for DNA (5658). Our results with the bulky PAH adducts are similar to our previous findings demonstrating that small DNA modifications do not dramatically alter affinity of polymerases for DNA or dNTPs (21, 24, 26, 27, 61) and also with other relatively bulky adducts, e.g., arylamines and cisplatin (27, 28, 62, 63). Some of the differences may be due to the use of E. coli polymerase I Klenow fragment (KF, KF- indicates exonuclease deficient) in other studies (57, 58), because this is not a very processive enzyme (64). However, at least one study (56) used a T7- preparation3 and concluded that attenuated affinity was a reason for poor catalytic activity with the bulky adducts. A potential problem in these studies (56-58) is that binding constants were estimated using electrophoretic separation methods, in which the DNA-protein complex is continuously forced to dissociate during the electrophoresis itself. That is, the equilibrium is distorted because of the rapid koff rates (Table 5). If the dissociated fraction is overlooked, the estimated Kd could be artificially high. The conclusion has been presented that the polymerase (T7-) dissociates after encountering a PAH adduct (56). However, a more appropriate explanation is that the rate of polymerization is retarded to the point that it is uncompetitive with what is a typical koff rate (∼1 s-1, Table 5, see Scheme 1) and the result is a lack of a kinetic burst (Figures 5 and 6). Both dA and dG PAH adducts have been studied with individual polymerases and in cellular mutagenesis systems. Several laboratories have used T7- in these studies.3 In an early study in this field, T7- was very sluggish in processing dG and dA adducts derived from 7-bromomethylbenz[a]anthracene, but dATP was preferentially inserted opposite the adduct (65). Christner et al. (66) reported that benzo[a]pyrene diol epoxide-derived 3 Many reports have used the enzyme commercially marketed as Sequenase, which is a different exonuclease- mutant of T7.

396

Chem. Res. Toxicol., Vol. 18, No. 2, 2005

Zang et al.

Figure 5. Time progress and elemental effect for dNTP incorporation. All studies were done with oligonucleotides 1 and 2 (Scheme 3), with the bold A of oligonucleotide 2 changed as indicated. (A) T7- (20 nM), preincubated with 100 nM unmodified oligonucleotide, was mixed with a solution of dTTP (20 µM) in the rapid quench instrument (9). For analysis of the elemental effect on phosphodiester bond formation, dTTPRS (20 µM) ([) was used instead of dTTP. The resulting plot (of product vs time) was fit to the burst equation y ) A(1 - e-kpt) + ksst, where A ) burst amplitude, kp ) presteady state rate of nucleotide incorporation, t ) time, and kss ) steady state rate of nucleotide incorporation, analyzed using GraphPad Prism version 3.0a. (B) T7- (25 nM), preincubated with 100 nM dA-ATBA-1S oligonucleotide, was mixed with a solution of dATP (250 µM) (2) or dATPRS (b) using conventional methods. (C) T7(25 nM), preincubated with 100 nM dA-ATBA-11S oligonucleotide, was mixed with a solution of dATP (250 µM) (2) or dATPRS (250 µM) (b) using conventional methods. (D) T7- (25 nM), preincubated with 100 nM dA-STBP-10S oligonucleotide, was mixed with a solution of dATP (250 µM) (2) or dATPRS (250 µM) (b) using conventional methods. (E) T7- (25 nM), preincubated with 100 nM dA-ACBP-10R oligonucleotide, was mixed with a solution of dATP (250 µM) (2) or dATPRS (250 µM) (b) using conventional methods. The elemental effects are listed in Table 7.

Figure 6. Time progress and elemental effects for dNTP incorporation. All studies were done with oligonucleotides 3 and 4 (Scheme 3), with the bold G of oligonucleotide 4 changed as indicated. (A) T7- (20 nM), preincubated with 100 nM unmodified oligonucleotide, was mixed with a solution of dCTP (20 µM) in the rapid quench apparatus ([). For analysis of the elemental effect on phosphodiester bond formation, dCTPRS (20 µM) (b) was used instead of dCTP. (B) T7- (50 nM), preincubated with 100 nM dG-ATBP-10R oligonucleotide, was mixed with a solution of dATP (250 µM) (2) or dATPRS (b) using conventional methods. (C) T7- (50 nM), preincubated with 100 nM dG-ATBP-10S oligonucleotide, was mixed with a solution of dATP (250 µM) (2) or dATPRS (250 µM) (b) using conventional methods. The elemental effects are listed in Table 7.

dA adducts were strong blocks but could be bypassed by polymerases in the order KF- > T7- > RT ∼ mammalian polymerase R. Lipinski et al. (67) reported that benzo[a]pyrene diol epoxide-derived N2-dG adducts blocked both KF- and T7-, although the point should be made that only 12-mer/16-mer duplexes were used, which are rather short for efficient processing by T7- (21, 43, 54, 68). Eukaryotic translesion synthesis polymerases have also been studied and are more proficient in bypassing

PAH adducts, including polymerases κ (19, 69, 70), η (71, 72), ι (73), and (yeast) ζ (74). The bacterial translesion synthesis polymerases II and IV and replicative polymerase III (E. coli) have been examined with DNA-PAH adducts and the conclusion was reached that the results are a function of the PAH adduct and the polymerase used (75). Chary and Lloyd (29) reported that KF- preferentially incorporated dCTP opposite the benzo[a]pyrene diol

Polymerases and Polycyclic Hydrocarbon-DNA Adducts

epoxide-dervied N6-dA adducts with the R stereochemistry at C-10 (dA-ATBP-10R) but inserted dATP opposite the adduct with the S stereochemistry at C-10 (dA-ATBP10S). A T7 preparation3 preferentially inserted dATP opposite both (29). In our work with T7-, we observed little insertion of dTTP opposite either adduct but preferential incorporation of dATP opposite the R adduct (Table 2). Chary et al. (30) also used KF- with the N6dA adduct derived from benzo[a]pyrene diol epoxide and observed more incorporation (of a mixture of all four dNTPs) opposite dA-ACBP-10R than dA-STBP-10S (no extension beyond the adduct site after insertion in either case). In our work (Table 1), we saw a slight difference (in the opposite order) for dATP incorporation, with little incorporation of dTTP opposite either N6-dA adduct. Chary et al. (29, 31) reported that the S configuration at the 10-position of dA-PAH derivatives led to more bypass with RT than observed for the dA adducts with the R configuration; these studies were directed toward the issue of extension more than one-base incorporation. Work by others (66) indicated that such 10S-dA adducts are highly blocking to several polymerases, but part of the problem with some of the polymerases might have been the short oligonucleotides used (vide supra). Shibutani et al. (76) found a preference for the R > S configuration in the bypass of dG-ATBP adducts by KF-, with insertion of dATP, which we also observed with T7(Table 2). In work in E. coli mutagenesis systems, Shukla et al. (77) concluded that the S stereochemistry at the C-10 position of dG-benzo[a]pyrene diol epoxide adducts led to more A incorporation (in the mutants) while the R stereochemistry yielded a mixture of substitutions. In the work of Zhuang et al. (78) with KF, dA incorporation was dominant opposite the dG-ATBP-10S adduct; however, when dT was incorporated opposite the adduct, the primer (now containing 3′-dT) was most readily extended. In our work with T7-, we saw only A insertion (with R > S) (Figure 3, Table 2). Others have also reported on the effects of stereochemistry of PAH adducts, mainly with other polymerases (than T7-) and cellular systems. The effects seem to vary not only with regard to stereochemistry but also sequence context (12). Another potential problem in the interpretation of cellular mutagenesis studies is the finding of base pair mutations at sites nearby but not directly opposite adducts (79-82). In terms of comparing stereochemical effects within our own set of results, some patterns can be seen. dA-ATBA1S and dA-ATBA-11S are regio- (and stereo-) isomers; the former lead to considerable T incorporation but the latter to A incorporation (Table 1). With the dA-STBP stereoisomeric pair, both the R and S configurations (at C-10) led to preferential incorporation of A > T, but the tendency was greater for the R configuration (Table 1). The two G adducts are enantiomeric with respect to the PAH moiety (Scheme 2). Both were very inefficient for directing the incorporation of C; the R isomer was considerably more efficient in incorporating A than was the S isomer (Tables 2 and 7). In summary, all of the PAH adducts, derived from both A and G, showed a strong tendency to direct the insertion of dATP opposite the adduct (Figures 1 and 3; Tables 1, 2, and 7) and then stop (Figures 1 and 3). Consideration of the results for binding of dNTPs and DNA did not show remarkable differences (