AUGUST 2007 VOLUME 20, NUMBER 8 Copyright 2007 by the American Chemical Society
Communication Selective N1-Alkylation of 2′-Deoxyguanosine with a Quinolinyl Quinone Methide Qibing Zhou,* Ting Xu, and John B. Mangrum Department of Chemistry, Virginia Commonwealth UniVersity, 1001 West Main Street, Richmond, Virginia 23284-2006 ReceiVed May 11, 2007
Nucleobase modification by quinone methides (QMs) has been extensively studied in the past decade, and multiple QM adducts were observed. For 2′-deoxyguanosine (dG), the N2-dG alkylation adduct was favored under aqueous buffered conditions over other N1-dG, N7-dG, and N7-guanine adducts. We report in this communication that the N1-dG adduct was selectively formed with a quinolinyl QM in 30% aqueous DMF and 10 mM phosphate buffer (pH 7.0) as a favored dG alkylation product. The quinolinyl QM was formed through the fluoride-induced desilylation and elimination of acetate, and the structure of the N1-dG adduct was fully established by one- and two-dimensional NMR analyses. In addition, the concentration of salt played a significant role in N1-dG adduct formation. Further HPLC analysis indicated that the addition of salt decreased the rate of QM formation from the acetate intermediate, although an in-depth mechanistic study is needed. Introduction Quinone methides (QMs) 1 are reactive intermediates that have been widely used in the covalent modification of DNA by natural products and synthetic molecules to induce biological impacts (1–6). For example, formation of QMs through oxidation and isomerization is a major toxicological pathway in the metabolism of estrogens and related drugs (7–10). Also, reactive oxygen species may be produced as a potential oxidative stress in the formation of QM through a redox cycle as observed in terpenone QMs (11, 12). Recent synthetic examples of DNA * To whom correspondence should be addressed. Phone: (804) 828-3520. Fax: (804) 828-8599. E-mail:
[email protected]. 1 Abbreviations: COSY, correlation spectroscopy; ESI-MS, electrospray mass analysis; dG, 2′-deoxyguanosine; HMBC, heteronuclear multiple-bond correlation; HMQC, heteronuclear multiple-quantum correlation; MOPS, 3-(N-morpholino)propanesulfonic acid; QM, quinone methide; TBDMSCl, tert-butyldimethylsilyl chloride.
cross-linking by QMs include a target-promoted QM oligonucleotide adduct and photoinduced binol systems (13–15). Nucleobase modification by QMs has been extensively studied in the past decade (16–22). Multiple QM adducts have been found at different nitrogen sites of nucleobases except for thymine. The structures of these QM adducts and the percent distribution have also been fully established (16–19). Most interestingly, nucleobase alkylation by QMs involves a kinetic and thermodynamic controlled interconversion through the regeneration of QM from the initial nucleobase adducts such as N1-dA adduct (18, 20). For alkylation of 2′-deoxyguanosine (dG), stable N1-, N2-dG QM adducts were observed along with a reversible N-7 adduct, which is partially degraded to the N-7 guanine adduct (16, 19, 20). Under buffered conditions (pH 7.0), the N2-dG adduct was dominantly formed versus the N1-dG adduct at a ratio of 5 to 1 approximately (16, 19). In addition,
10.1021/tx700162d CCC: $37.00 2007 American Chemical Society Published on Web 07/13/2007
1070 Chem. Res. Toxicol., Vol. 20, No. 8, 2007 Scheme 1. Quinolinyl QM as a Substituted QM through Fluoride-Induced Formation
computational results were consistent with these experimental observations and indicated that the N2-dG adduct has a lower ∆G of -4.3, -2.0, and -1.6 kcal/mol than that of the N1-dG adduct in water, DMSO and acetonitrile, respectively (21). Furthermore, the reactivity and selectivity of QM toward nucleobases are influenced by the substituent groups on the QMs (22). In our laboratory, a quinolinyl QM was designed to be formed through the fluoride-induced QM generation by desilylation and elimination of acetate as the leaving group (Scheme 1) (16–19). The incorporation of the quinolinyl substitute was designed to enhance the potential interactions with DNA through possible partial intercalation and charges of the quinoline moiety and also for the investigation of the steric effect on the reactivity of QMs. We report in this communication that the N1-dG adduct was selectively formed with quinolinyl QM as a favored dG adduct in 30% aqueous DMF and 10 mM phosphate buffer (pH 7.0), while other dG adducts were not observed under the conditions that were studied.
Experimental Procedures All chemicals were purchased from Fisher Scientific (Pittsburgh, PA) or Sigma-Aldrich (Milwaukee, WI) and used without further purification. The NMR spectra were recorded with Variant NMR spectrometers. HPLC analysis was carried out on a Jasco (Easton, MD) 2000 series station with Microsorb columns from Varian (Walnut Creek, CA). Electrospray ionization mass spectroscopy (ESI-MS) analysis was carried out on a Thermo Finnigan (Waltham, MA) LCQ Deca XP Plus mass spectrometer. Water was purified with a Barnstead (Dubuque, IA) E-pure 4-Module Deionization System. Compound 1 was synthesized as reported previously (23). 2,2-Dimethyl-5-quinolin-3-ylbenzo[1,3]dioxin-4-one (2). To a solution of triisopropyl borate (2.8 mL, 12 mmol) and 3-bromoquinoline (2.08 g, 10 mmol) in dry THF (20 mL) was added n-butyllithium (7.5 mL of a 1.6 M solution in hexane, 12 mmol) dropwise via a syringe pump over 1 h under N2 at -78 °C. After 2 h, the acetone/dry ice bath was removed, and the reaction solution was allowed to warm to 0 °C. The reaction was then quenched with a 2 M HCl solution, and the pH was adjusted to 7 with a solution of 5 M NaOH. The resulting solution was extracted with diethyl ether (3 × 100 mL). The combined organic layers were dried with MgSO4 and concentrated to afford the corresponding boronic acid as an oil, which was directly used for the following reaction without further purification. Dry THF (30 mL) was added to a mixture of triflate 1 (456 mg, 1.40 mmol), the above boronic acid (290 mg, 1.68 mmol), Pd(PPh3)4 (81 mg, 0.07 mmol), KBr (167 mg, 1.40 mmol), and Cs2CO3 (682 mg, 2.10 mmol). The resulting mixture was stirred for 15 min, and then water (150 µL) was added. The reaction mixture was refluxed for 18 h. The reaction solution was extracted with diethyl ether (2 × 150 mL). The collected organic layers were washed with brine, dried with MgSO4, and concentrated. Flash chromatographic separation (30–45% EtOAc in hexanes) afforded 2 as a brown oil (312 mg) in 73% yield: 1H NMR (CDCl3, 300 MHz) δ 8.85 (s, 1H), 8.17–8.12 (m, 2H), 7.85 (d, J ) 8.12 Hz, 1H), 7.73 (t, J ) 8.43 Hz, 1H), 7.63–7.54 (m, 2H), 7.09–7.05 (m, 2H), 1.82 (s, 6H); 13C NMR (CDCl3, 75 MHz) δ 159.8, 157.6, 150.9, 147.2, 142.4, 135.8, 134.9, 133.6, 129.9, 129.4, 128.3, 127.9,
Communications 127.1, 126.5, 117.8, 112.2, 105.9, 25.9; ESI-MS calcd for C19H16NO3 (M + H+) 306.11, found 306.12. 3-[3-(tert-Butyldimethylsilanyloxy)-2-(tert-butyldimethylsilanyloxymethyl)phenyl]quinoline (3). To a solution of 2 (312 mg, 1.02 mmol) in dry THF (20 mL) was added lithium aluminum hydride (232 mg, 6.12 mmol) at 0 °C. The reaction mixture was stirred under N2 for 1 h at room temperature and then the reaction quenched with a 5% ammonium chloride solution at 0 °C. The reaction solution was diluted with water (50 mL), acidified to pH 5 with 1 M HCl, and then extracted with diethyl ether (2 × 60 mL). The organic layers were collected, washed with brine (100 mL), dried with MgSO4, and concentrated. Flash chromatographic separation (10% MeOH in CH2Cl2) afforded the bis-hydroxylated intermediate as a brown oil (225 mg) in 88% yield: 1H NMR (CD3OD, 300 MHz) δ 8.94 (s, 1H), 8.41 (s, 1H), 8.08 (d, J ) 8.30 Hz, 1H), 7.99 (d, J ) 8.17 Hz, 1H), 7.80 (t, J ) 8.44 Hz, 1H), 7.65 (t, J ) 8.09 Hz, 1H), 7.27 (t, J ) 7.84 Hz, 1H), 6.94 (d, J ) 8.15 Hz, 1H), 6.89 (d, J ) 7.59 Hz, 1H), 4.58 (s, 2H); 13C NMR (CD3OD, 75 MHz) δ 157.2, 151.0, 146.4, 139.8, 136.6, 134.6, 129.9, 129.0, 128.3, 128.1, 127.7, 127.3, 124.9, 121.4, 115.3, 56.8; ESI-MS calcd for C16H14NO2 (M + H+) 252.10, found 252.10. To a solution of the above bis-hydroxylated intermediate (200 mg, 0.80 mmol) in dry DMF (20 mL) were added TBDMSCl (720 mg, 4.78 mmol) and imidazole (325 mg, 4.78 mmol). The reaction solution was stirred for 16 h under N2 and then diluted with diethyl ether (300 mL). The organic layer was washed with 1 M HCl (150 mL), water (2 × 150 mL), and brine (150 mL), dried with MgSO4, and concentrated. Flash chromatographic separation (10–15% EtOAc in hexanes) afforded 3 as a colorless oil (332 mg) in 87% yield: 1H NMR (CDCl3, 300 MHz) δ 8.95 (s, 1H), 8.26 (s, 1H), 8.16 (d, J ) 8.52 Hz, 1H), 7.83 (d, J ) 8.43 Hz, 1H), 7.73 (t, J ) 8.45 Hz, 1H), 7.58 (t, J ) 8.07 Hz, 1H), 7.26 (t, J ) 7.89 Hz, 1H), 6.92 (d, J ) 7.90 Hz, 2H), 4.62 (s, 2H), 1.05 (s, 9H), 0.77 (s, 9H), 0.31 (s, 6H), –0.19 (s, 6H); 13C NMR (CDCl3, 75 MHz) δ 154.7, 151.6, 147.2, 141.6, 135.9, 134.8, 129.6, 129.5, 129.4, 128.7, 128.1, 127.8, 127.0, 123.7, 118.9, 57.3, 26.2, 26.1, 18.7, 18.4, –3.7, –5.4; ESI-MS calcd for C28H42NO2Si2 (M + H+) 480.28, found 480.28. Acetic Acid 2-(tert-Butyldimethylsilanyloxy)-6-quinolin3-ylbenzyl Ester (4). FeCl3·6H2O (204 mg, 0.76 mmol) was added to a stirred solution of 3 (242 mg, 0.50 mmol) in Ac2O (5 mL). After 15 min, the reaction mixture was diluted with diethyl ether (100 mL). The ether layer was washed with 1 M NaOH (2 × 75 mL) and brine (75 mL), dried with MgSO4, and concentrated. Flash chromatographic separation (5–20% EtOAc in hexanes) afforded 4 as a yellow oil (153 mg) in 74% yield: 1H NMR (CDCl3, 300 MHz) δ 8.94 (s, 1H), 8.19–8.14 (m, 2H), 7.83 (d, J ) 8.10 Hz, 1H), 7.76 (t, J ) 8.42 Hz, 1H), 7.60 (t, J ) 8.42 Hz, 1H), 7.36 (t, J ) 8.10 Hz, 1H), 7.01–6.95 (m, 2H), 4.99 (s, 2H), 2.04 (s, 3H), 1.01 (s, 9H), 0.32 (s, 6H); 13C NMR (CDCl3, 75 MHz) δ 170.9, 155.9, 151.0, 147.1, 142.0, 136.0, 133.8, 130.1, 130.1, 129.3, 128.2, 127.7, 127.5, 124.4, 123.4, 118.6, 60.1, 25.8, 21.2, 18.5, –3.9; ESIMS calcd for C24H30NO3Si (M + H+) 408.20, found 408.19. N1-dG Alkylation Study with Quinolinyl QM. N1-dG alkylation was carried out in a 10% aqueous DMF solution with or without 10 mM MOPS buffer (pH 7.5) and 25 mM KF. The alkylation was initiated with the addition of a KF solution. The final volume was 400 µL, and the final concentrations for 4 and dG were 1.25 and 2.5 mM, respectively. The progression of the alkylation at room temperature was monitored over a period of 72 h by reversed-phase HPLC analysis with a Microsorb MV C18 column (250 mm × 2.1 mm, 8 µm) at a flow rate of 1 mL/ min using a gradient condition (CH3CN in 50 mM triethylammonium acetate buffer at pH 5.0, 10% for 5 min, then from 10 to 70% over 30 min, and then from 70 to 100% over 5 min). All experiments were repeated at least three times independently. For the full characterization of the N1-dG adduct (7), to a solution of 4 (15 mg, 0.037 mmol) and dG (15 mg, 0.054 mmol) in DMF (5 mL) were added KF (11 mg, 0.18 mmol) and 500 µL of water. After being stirred for 12 h at room temperature, the reaction solution was diluted with water and filtered. Aliquots of the resulting solution were separated with a preparative Microsorb C4 column
Communications
Chem. Res. Toxicol., Vol. 20, No. 8, 2007 1071
Scheme 2. Synthesis of Quinolinyl QM Precursor 4
(250 mm × 10 mm, 10 µm) at a flow rate of 4 mL/min. HPLC fractions at 18.8 min were collected and lyophilized to afford N1-dG adduct 7 (10 mg) in 54% yield: 1H NMR (DMSO, 400 MHz) δ 8.71 (d, J ) 2.00 Hz, 1H), 8.04 (d, J ) 2.00 Hz, 1H), 7.96 (d, J ) 8.44 Hz, 1H), 7.79 (d, J ) 8.09 Hz, 1H), 7.64 (t, J ) 7.43 Hz, 2H), 7.63 (s, 1H), 7.55 (t, J ) 7.37 Hz, 1H), 7.15 (t, J ) 7.93 Hz, 1H), 6.84 (d, J ) 8.69 Hz, 1H), 6.61 (d, J ) 8.43 Hz, 1H), 5.93 (t, J ) 7.09 Hz, 1H), 5.17 (d, J ) 16.34 Hz, 1H), 5.04 (d, J ) 16.34 Hz, 1H), 4.28 (m, 1H), 3.75 (m, 1H), 2.43–2.33 (m, 1H), 2.12–2.06 (m, 1H); 13C NMR (DMSO, 75 MHz) δ 157.4, 157.0, 154.9, 151.3, 149.2, 147.0, 140.5, 135.6, 135.3, 134.5, 129.8, 129.2, 128.8, 128.4, 127.7, 127.2, 122.0, 121.3, 116.6, 116.1, 88.1, 82.7, 71.4, 62.4, 40.0, 40.0; ESI-MS calcd for C26H25N6O5 (M + H+) 501.19 and C26H24N6NaO5 (M + Na+) 523.17, found 500.86 and 522.82. Effects of Solvent and Salt on N1-dG Alkylation. All of the reactions were carried out at 37 °C similarly as described above in 10 or 30% aqueous DMF with 10, 20, or 40 mM phosphate buffer (pH 7.0) and 25 mM KF. The final concentrations for 4 and dG were 1.0 and 2.0 mM, respectively. The salt effects were studied with the addition of 10 or 30 mM NaCl to the reaction solution containing 10 mM phosphate buffer. For the kinetic assay, an internal standard (benzophenone, 0.2 mM) was included and has a retention time of 30 min under the HPLC analysis conditions that were used. All experiments were repeated at least three times independently.
control experiment without dG showed a dominant signal with a retention time of 23.6 min (Figure 1a,b), which was identified as the hydrolysis product 6 by ESI-MS analysis with a MS signal at m/z 252.18 (calcd for M + H+, 252.10). In the reaction with dG, two new signals were observed at 18.8 and 26.4 min in addition to the hydrolysis product at 3 h (Figure 1b,c). After 24 h, the magnitude of the HPLC signal at 26.4 min decreased significantly, while the magnitudes of the signals at 18.8 and 23.6 min increased simultaneously and remained unchanged even after 72 h (Figure 1c,d). The ESI-MS analysis confirmed the hydrolysis product 6 at 23.6 min and revealed that the signal at 26.4 min was acetate 5 with a MS signal at m/z 294.21 (calcd for M + H+, 294.11), and the signal at 18.8 min was a dG–quinolinyl QM adduct 7 with MS signals at m/z 500.86 and 522.82 (calcd for M + H+ and M + Na+, 501.19 and 523.17, respectively). Interestingly, when no buffer was added to the reaction solution, the dG–quinolinyl QM adduct 7 at 18.8 min was observed as the dominant product only after 3 h (Figure 1e). The structural identity of the dG–quinolinyl QM adduct 7 was then determined by NMR analysis, including 1H, 13C, 1 H–1H correlation spectroscopy (COSY), 1H–13C heteronuclear multiple-quantum correlation (HMQC), and 1H–13C heteronuclear multiple-bond correlation (HMBC) (19). A preparative
Results and Discussion Synthesis of Quinolinyl QM Precursor 4. The synthesis of quinolinyl QM precursor 4 was achieved in four steps starting with a Suzuki coupling of triflate 1 and 3-quinoline boronic acid (Scheme 2). Triflate 1 was obtained as reported previously (23), and 3-quinoline boronic acid was synthesized from the corresponding bromide (24). The Suzuki coupling afforded the desired quinoline 2 in 73% yield using Pd(PPh3)3 as the catalyst in wet THF (23, 25). The resulting quinone intermediate 2 was then reduced with lithium aluminum hydride, followed by silylation of the two hydroxyl groups. Finally, the benzylic silyl ether was selectively converted to an acetate group in acetic anhydride and 1.5 equiv of FeCl3 to quinoline QM precursor 4 in 74% yield (16–19). Selective N1-dG Alkylation with Quinolinyl QM. The alkylation of dG with quinolinyl QM precursor 4 in the presence of KF was first carried out in 10% aqueous DMF and MOPS buffer at room temperature (16–19). A high percentage of DMF is used to prevent precipitation of QM precursor 4 in the reaction solutions. The final reaction solution contained 2.5 mM dG, 1.25 mM QM precursor 4, 25 mM KF, and 10 mM MOPS (pH 7.5). The progress of the reaction was monitored by HPLC analysis with a reverse-phase C18 column over 72 h (Figure 1). The
Figure 1. HPLC analysis of dG alkylation with quinolinyl QM. The reactions were carried out in 10% aqueous DMF containing 1.25 mM 4 with or without 2.5 mM dG, 25 mM KF, or 10 mM MOPS buffer (pH 7.5) at room temperature: (a) precursor 4 in the absence of KF, (b) 72 h with KF in the absence of dG, (c) 3 h with KF and dG, (d) after 24 h with KF and dG, and (e) 3 h with KF and dG without MOPS buffer.
1072 Chem. Res. Toxicol., Vol. 20, No. 8, 2007
Communications
Figure 2. HMBC NMR analysis of the N1-dG–quinolinyl QM adduct. Long-range couplings of the benzylic protons (H17′′) were observed with five carbons (C2, C6, C11′′, C12′′, and C13).
synthetic scale of the dG–quinolinyl QM adduct 7 was achieved with the condition as described in 10% water with a HPLC separation. In addition, the isolated dG–quinolinyl QM adduct 7 was co-injected with the reaction solution in 10 mM MOPS buffer, which confirmed the same identity as a single peak with a retention time of 18.8 min. For the structural elucidation, the COSY NMR analysis established the coupling pattern of the protons, while the HMQC analysis provided the direct connectivity of protons to carbons. Most importantly, the HMBC analysis revealed five long-range couplings of the benzylic proton (H17′′) to five carbons, C2, C6, C11′′, C12′′, and C13′′ (Figure 2), which indicated that the HPLC signal at 18.8 min was the N1-dG–quinolinyl QM adduct. In addition, all of the chemical shifts observed in the NMR analysis were consistent with the N1-dG adduct structure (see the Supporting Information for the full assignment of protons and carbons), and similar to other reported N1-dG–QM adducts (16–19). The selective formation of the N1-dG adduct in 10% water without buffer was consistent with the alkylation of N1-dG with simple QMs (19). It has been reported that N1-dG alkylation preferably occurred under unbuffered conditions due to deprotonation while the N2-dG adduct was the major product in buffered solutions (16–19). However, selective N1-dG alkylation was observed in our study with quinolinyl QM under both conditions without or with 10 mM MOPS buffer (pH 7.5) in 10% aqueous DMF. Effects of Solvent and Salt on the N1-dG Alkylation with Quinolinyl QM. To confirm that the N1-dG adduct was not only selectively formed in MOPS buffer, the alkylation of dG with quinolinyl QM was investigated in a 10% aqueous DMF solution with a phosphate buffer (pH 7.0) at 37 °C under similar conditions. In addition, the percent of water in the reaction solution was increased to 30% with a series of phosphate buffer concentrations of 10, 20, and 40 mM. Using the same HPLC analysis method, we found that the N1-dG adduct 7 was also formed preferably in 10% aqueous DMF with 10 mM phosphate buffer (pH 7.0) and had a 2.4:1 ratio over hydrolysis product 6 after 12 h (Figure 3a). The higher ratio as compared to that in MOPS buffer was possibly due to a lower pH value (7.0 vs 7.5) and an increased temperature. In 30% aqueous DMF, the N1-dG adduct remained persistently as the only dG QM adduct even though the amount of hydrolysis product 6 was significantly increased due to the increased concentration of water (Figure 3b). Surprisingly, increasing the buffer concentration from 10 to 20 mM and from 10 to 40 mM in 30% aqueous DMF considerably decreased the extent of N1dG adduct formation and enhanced the amount of hydrolysis product 6 (Figure 3b–d), while no other significant signals were
Figure 3. HPLC analysis of the effects of solvent and buffer on the selective formation of N1-dG–quinolinyl QM adduct 7. The reactions were carried out in aqueous DMF containing 1.0 mM 4, 2.0 mM dG, and 25 mM KF in phosphate buffer (pH 7.0) at 37 °C for 12 h: (a) 10% water and 10 mM phosphate buffer, (b) 30% water and 10 mM phosphate buffer, (c) 30% water and 20 mM phosphate buffer, and (d) 30% water and 40 mM phosphate buffer. The ratios of N1-dG adduct 7 to hydrolysis product 6 were calculated on the basis of the area integration of the signals at 260 nm.
observed. The effect of increased buffer concentrations on the formation of the N1-dG adduct was unexpected because the concentrations of QM precursor 4 and dG were only 1.0 and 2.0 mM, respectively, and phosphate buffer at 10 mM should have sufficient capacity to maintain the pH of the reaction solution. To demonstrate that the negative effect on formation of the N1-dG adduct via increased concentrations of phosphate buffer was not due to the increased buffering capacity in aqueous DMF, we investigated the dG alkylation in 10 mM phosphate buffer (pH 7.0) containing 10 and 30 mM NaCl, respectively. The HPLC analysis revealed that the integration ratio of N1-dG adduct 7 to hydrolysis product 6 in 10 mM phosphate buffer with 10 mM NaCl was similar to that in 20 mM phosphate buffer (1:4.6). So did the ratio observed in the 10 mM phosphate buffer with 30 mM NaCl to that in 40 mM phosphate buffer (1:13). Therefore, we concluded that the decreased level of formation of the N1-dG adduct with increased buffer concentrations observed in Figure 3b–d was the salt effect in the reaction solution. To understand the role of the effect of salt on formation of the N1-dG adduct and the competitive hydrolysis, the progression of the dG alkylation with quinolinyl QM in 10 mM phosphate buffer and 0, 10, or 30 mM NaCl was monitored by HPLC analysis, respectively. HPLC analysis revealed that both N1-dG alkylation and QM hydrolysis occurred at a fast rate, and more than 50% of acetate 5 was consumed within 15 min
Communications
Figure 4. Kinetic profile of the effect of salt on acetate 5 by HPLC analysis. The reactions were carried out in 30% aqueous DMF containing 1.0 mM 4, 2.0 mM dG, and 25 mM KF in phosphate buffer (pH 7.0) at 37 °C with 0, 10, or 30 mM NaCl. Each datum represents the average of three independent experiments ( the standard deviation. The percent of acetate 5 was calculated on the basis of the area integration at 260 nm related to the combined amount of acetate 5 and hydrolysis product 6 at 5 min in 10 mM phosphate and 30 mM NaCl, assuming that the amount of the N1-dG adduct was negligible (< 2%) and acetate 5 and hydrolysis product 6 have similar extinction coefficients at 260 nm due to their structural similarity. In addition, the area integration of acetate 5 was further corrected with that of an internal standard (benzophenone) in each of the reactions.
under all three of the conditions that were studied (Figure 4). Therefore, the rates of N1-dG alkylation and QM hydrolysis could not be derived on the basis of the kinetic profiles obtained by HPLC analysis. Unfortunately, alternative kinetic analysis with an UV spectrometer (26, 27) was not feasible because the UV absorbance spectra of acetate 5, hydrolysis product 6, and N1-dG adduct 7 were similar to each other. However, the reaction profiles of the conversion of acetate 5 obtained by HPLC analysis did provide some insights into the mechanisms of the salt effects on the dG alkylation by quinolinyl QM (Figure 4). The percent of acetate 5 in all reaction conditions was obtained by calculations on the basis of the area integration obtained by HPLC analysis at 260 nm related to the combined amount of acetate 5 and hydrolysis product 6 at 5 min in 10 mM phosphate and 30 mM NaCl. This is based on the observation that the amount of the N1-dG adduct was negligible (