660
Chem. Res. Toxicol. 1997, 10, 660-666
Accurate and Sensitive Quantitation of N7-Methyldeoxyguanosine-3′-monophosphate by 32P-Postlabeling and Storage-Phosphor Imaging Kemal Haque,† Donald P. Cooper,† Joost H. M. van Delft,‡ Siow M. Lee,§ and Andrew C. Povey*,†,| Cancer Research Campaign Departments of Carcinogenesis and Medical Oncology, Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Manchester M20 9BX, U.K., Toxicology Division, TNO Nutrition and Food Research Institute, Utrechtseweg 48, P.O. Box 360, 3700 AJ Zeist, The Netherlands, and School of Epidemiology and Health Sciences, Medical School, University of Manchester, Oxford Road, Manchester M13 9PT, U.K. Received October 8, 1996X
As N7-methyldeoxyguanosine-3′-monophosphate (N7-MedGp) is the major, persistent DNA lesion generated by methylating agents, a combined HPLC/32P-postlabeling assay has been developed to quantitate this adduct in human DNA. N7-MedGp was purified from normal nucleotides by anion-exchange chromatography followed by reverse-phase HPLC procedures. The adduct was then 32P-postlabeled and resolved by two-dimensional TLC for detection and quantitation by storage-phosphor imaging. The effect of conditions used for DNA purification and digestion on the recovery of N7-MedGp has been investigated. Extended, raised temperature incubations normally employed during DNA purification were demonstrated to result in considerable loss of adduct through depurination after 22 h at 65 and 37 °C (82% and 20% loss, respectively), but depurination was reduced to 5% if the incubation was performed at either 4 or 22 °C. Similarly, close to optical recovery (83%) of N7-MedGp was achieved after DNA digestion by incubating at 4 °C, pH 7.4, for 18 h in the presence of micrococcal nuclease and calf spleen phosphodiesterase from Sigma and Boehringer Mannheim, respectively. Overall, the recovery of N7-MedGp was 40%, resulting in a detection limit of 1.3 fmol which is equivalent to 0.16 µmol of adduct/mol of 2′-deoxyguanosine-3′-monophosphate (dGp) when analyzing 10 µg of DNA. The N7-MedGp content of DNA that had been methylated in vitro using 0, 16, and 80 µM N-methyl-N-nitrosourea (NMU) was determined by 32P-postlabeling to be 12, 112, and 671 µmol of N7-MedGp/mol of dGp. Electrochemical detection of N7-methylguanine (N7-MeG) after HPLC purification measured approximately 2-fold higher levels, i.e., 25, 225, and 1080 µmol of N7-MeG/mol of Gua, at each NMU concentration, respectively. The levels of N7-MedGp in the white blood cell (WBC) DNA of patients receiving a single dose of 5-(3,3-dimethyl-1-triazeno)imidazole-4-carboxamide (DTIC) chemotherapy were determined by 32P-postlabeling. Maximum levels were found 4-6 h after treatment, and in two out of four individuals adduct levels were decreased by 21 h. Prior to treatment, N7MedGp was detectable in WBC DNA in two out of the four individuals indicating that nontherapeutic exposure to methylating agents had occurred.
Introduction The formation of DNA adducts is believed to be the earliest indicator of the biologically effective dose arising from exposure to chemical carcinogens (1, 2). As a consequence, techniques have been developed and applied to the detection and quantitation of DNA adducts formed by a wide variety of chemical carcinogens (3, 4). The DNA adduct N7-methylguanine (N7-MeG1) is potentially the most useful marker of exposure to methylating agents as it is quantitatively the major adduct formed by these compounds (5) and is relatively slowly repaired (6). A wide variety of techniques has been applied to the quantitation of N7-MeG in human DNA, e.g., 32P-postlabeling (7-11), fluorescence detection (12), enzymelinked immunoassay (13, 14), and electrochemical detec* To whom correspondence should be addressed. † CRC Department of Carcinogenesis. ‡ TNO Nutrition and Food Research Institute. § CRC Department of Medical Oncology. | University of Manchester. X Abstract published in Advance ACS Abstracts, May 15, 1997.
S0893-228x(96)00172-5 CCC: $14.00
tion (15). The 32P-postlabeling-based assays possess the greatest sensitivity and are generally considered to be the most appropriate approach for quantitating low levels of N7-MeG arising from environmental exposures. However, using this technique we were previously unable to detect N7-MeG in human DNA containing quantifiable levels of the related adduct O6-methylguanine (16). In this present study we describe an improved 32Ppostlabeling assay for N7-methyldeoxyguanosine-3′1 Abbreviations: AE-HPLC, anion-exchange HPLC; CSPD, calf spleen phosphodiesterase; CT-DNA, calf thymus DNA; D1, first dimension; D2, second dimension; dAp, 2′-deoxyadenosine-3′-monophosphate; dCp, 2′-deoxycytidine-3′-monophosphate; dGp, 2′-deoxyguanosine-3′-monophosphate; DMSO, dimethyl sulfoxide; dNp, 2′-deoxynucleoside-3′-monophosphate; DTIC, 5-(3,3-dimethyl-1-triazeno)imidazole-4-carboxamide; dTp, thymidine-3′-monophosphate; HPLCECD, HPLC/electrochemical detection; MN, micrococcal nuclease; N7MedGp, N7-methyldeoxyguanosine-3′-monophosphate; N7-MeG, N7methylguanine; N7-MepdG, N7-methyldeoxyguanosine-5′-monophosphate; NMU, N-methyl-N-nitrosourea; pdG, 2′-deoxyguanosine5′-monophosphate; PEI-cellulose, poly(ethylenimine)-cellulose; RPHPLC, reverse-phase HPLC; T4 PNK, T4 polynucleotide kinase; TEAA, triethylammonium acetate; [3H]Me-CT-DNA, [3H]methyl calf thymus DNA; WBC, white blood cell.
© 1997 American Chemical Society
Accurate and Sensitive Quantitation of N7-MedGp
Figure 1. MedGp.
32P-Postlabeling
scheme for the quantitation of N7-
monophosphate (N7-MedGp) (see Figure 1) which has incorporated two consecutive HPLC separations (8), careful optimization of DNA purification and digestion procedures, and storage-phosphor-based quantitation of radioactivity. The resultant increases in assay sensitivity and specificity have permitted the quantitation of N7MedGp in human DNA.
Materials and Methods Enzymes and Chemicals. T4 polynucleotide kinase (T4 PNK) was purchased from Boehringer Mannheim; [γ-32P]ATP was obtained from NEN (6000 Ci/mmol). PEI-cellulose thinlayer chromatography (TLC) plates were from Macherey-Nagel or Schleicher and Schuell. Micrococcal nuclease (MN) and calf spleen phosphodiesterase (CSPD) were purchased from Worthington Biochemical Corp. or from Sigma and Boehringer Mannheim, respectively. N7-MedGp standard was prepared and purified as previously described (16). [3H]Methyl calf thymus DNA ([3H]Me-CT-DNA) was produced by methylating CT-DNA with [3H]-N-methyl-N-nitrosourea ([3H]NMU) and was provided by Dr. G. Margison (17). All other reagents were analytical grade or higher; double-distilled water was used. 32P-Postlabeling Conditions. N7-MedGp and 2 pmol of internal standard dGp (2′-deoxyguanosine-3′-monophosphate) were combined in 0.5 mL Eppendorf tubes and dried in vacuo. Reaction buffer (5 µL; containing 30 mM bicine, pH 8.6, 30 mM MgCl2, 30 mM dithiothreitol, 3 mM spermidine), [γ-32P]ATP (2 µL, 20 µCi), T4 PNK (0.2 µL, 2.0 units), and 7.8 µL of H2O were added to each sample. After incubation at 37 °C for 60 min, 32P-postlabeled 3′,5′-bisphosphates were converted to 32P-postlabeled 5′-monophosphates by incubating the labeling reaction with ZnSO4 (1 µL, 1 mM), CH3COONa (1 µL, 1 M, pH 5), and nuclease-P1 (2 µL, 0.6 unit) for a further 30 min at 37 °C. Two-Dimensional TLC and Storage-Phosphor Adduct Quantitation. The TLC plates were initially washed by developing with methanol. 32P-Postlabeled, nuclease-P1-digested samples (10 µL) were then applied to PEI-cellulose TLC plates and chromatographed with 1 M ammonium acetate (pH 8)/propan-2-ol (90:10, v/v) as the first dimension (D1) and
Chem. Res. Toxicol., Vol. 10, No. 6, 1997 661 saturated sodium citrate/saturated ammonium sulfate/propan2-ol (50:5:1, v/v/v) as the second dimension (D2). The solvents were run to the edge of the plates (without the use of wicks) and allowed to dry in air. TLC plates were covered in Saran Wrap prior to adduct quantitation. Storage-phosphor quantitation was performed using a Molecular Dynamics 425S phosphorimager and ImageQuant software. TLC plates were exposed to storage-phosphor screens for 1-4 h at room temperature whereupon the screens were scanned at a resolution of 176 µm and the resultant digital image was analyzed. The signals generated by the [32P]-N7MepdG and [32P]-2′-deoxyguanosine-5′-monophosphate ([32P]pdG) spots were quantitated using the ImageQuant software using the average signal along the perimeter of the [32P]-N7MepdG and [32P]pdG regions to correct for background levels of radioactivity in these regions. N7-MedGp levels were then determined as (2 pmol of dGp) × (storage-phosphor signal of [32P]-N7-MepdG/storage phosphor signal of [32P]pdG) (21). High-Performance Liquid Chromatographic Purification of N7-MedGp. All HPLC separations were performed using a Waters 600E multisolvent delivery system and a Waters 484 tunable absorbance detector operated at 254 nm. Column fractions were collected using either an LKB 2211 Superrac or a Frac 100 fraction collector (Pharmacia). N7-MedGp was initially purified by weak anion-exchange HPLC (AE-HPLC) using a SynChropak AX300 column (5 µm, 250 mm × 4.6 mm) (8) eluted isocratically with 1 mL/min triethylamine acetate (0.5 M TEAA, pH 7). Before application of subsequent samples, the column was flushed with 0.1 M TEAA containing 20% (v/v) acetonitrile. The amount of dGp released by enzymic digestion was quantitated by the area under the appropriate HPLC peak (A254nm). The HPLC column fractions corresponding to N7MedGp were collected, pooled, dried in vacuo (-40 °C), and dissolved in H2O prior to reverse-phase HPLC (RP-HPLC). RPHPLC was performed using a Chromex Hypersil ODS 5 column (5 µm, 250 mm × 4.6 mm) eluted isocratically with 0.1 M TEAA (pH 7) containing 1% (v/v) acetonitrile (8, 16). After 15 min, the acetonitrile content of the buffer was increased to 20% in order to elute remaining bound nucleotides from the column. HPLC column fractions corresponding to N7-MedGp were collected, pooled, and dried in vacuo prior to 32P-postlabeling. Enzymic Digestion of DNA. DNA was digested to nucleoside-3′-monophosphates (dNp) with MN and CSPD. A typical digestion comprised 10 µg of DNA (in 10 µL of H2O), digestion buffer (5 µL; 50 mM Tris-HCl, 20 mM CaCl2, pH 7.4), MN (5 µL; 2 units from Worthington or 0.2 unit from Sigma), and CSPD (5 µL; 1.7 × 10-3 unit from Worthington or 2 µg from Boehringer Mannheim). Reactions were incubated at 4 °C for 18 h. Any variations in buffer pH or incubation temperature are specified in the text. Preparation of Methylated DNA and Quantitation of N7-Methylguanine by Combined HPLC/Electrochemical Detection. Salmon sperm DNA (1 mg/mL) in 50 mM sodium cacodylate buffer, pH 7.4, was methylated for 60 min at 37 °C with 16 or 80 mM NMU (Sigma, Poole, Dorset, U.K.) which had been dissolved in glacial acetic acid containing 1% DMSO. The DNA was then ethanol precipitated and redissolved in 10 mM KH2PO4‚KOH, pH 7.0, to give a final concentration of 1-1.5 mg/mL. The N7-MeG content of the methylated DNA was determined using an established HPLC-ECD modified procedure (13, 19) with minor modifications. DNA (200-300 µg in 200 µL) was incubated at 70 °C for 70 min following the addition of 20 µL of 1 M HCl. The hydrolyzed bases were separated on a Partisil 10 SCX HPLC column (10 µm, 250 mm × 4.6 mm; Whatman, Maidstone, U.K.) using solvent A (0.01 M ammonium formate, pH 4.0, containing 12% (v/v) methanol) and solvent B (0.2 M ammonium formate, pH 4.0, containing 12% (v/v) methanol) at a flow rate of 1.5 mL/min under the following gradient conditions: 0-10% B over 14 min, then 10-100% over 4 min. The column fractions corresponding to N7-MeG were collected, pooled, dried in vacuo, and dissolved in H2O. The N7MeG present in the pooled fractions was determined by electrochemical detection (oxidation potential 1250 mV; Antec
662 Chem. Res. Toxicol., Vol. 10, No. 6, 1997
Figure 2. Phosphorimages of two-dimensional TLC analyses of 32P-postlabeled N7-MedGp. Samples containing 2 pmol of dGp and (A) 2 fmol or (B) 200 fmol of N7-MedGp were 32P-postlabeled, nuclease-P1 digested, and separated by PEIcellulose TLC as described in the text. Samples were exposed for approximately 2-3 h. X ) unknown contaminants. Detector, Leiden, The Netherlands) following HPLC using a Chromspher C-18 column (5 µm, 100 mm × 3 mm; Chrompack, Vlissingen, The Netherlands) eluted with 25 mM H3PO4‚KOH, pH 6.0, containing 4% (v/v) methanol (flow rate was 0.5 mL/ min; retention time of N7-MeG was 7.5 min). The recovery of N7-MeG ranged from 85% to 95%; the coefficient of variation of repeat analyses was
