DNA Adduct Formation of the Food Carcinogen 2-Amino-3

File failed to load: https://cdn.mathjax.org/mathjax/contrib/a11y/accessibility-menu.js .... The C8-2′-Deoxyguanosine Adduct of 2-Amino-3-methylimid...
0 downloads 0 Views 3MB Size
Download PDF
Chem. Res. Toxicol. 1994, 7, 752-761

752

DNA Adduct Formation of the Food Carcinogen 2-Amino-3-methylimidazo[4,5-flquinoline at the C-8 and N2Atoms of Guanine Robert J. Turesky*and Jovanka Markovic Nestec Ltd., Research Centre, Vers-chez-les-Blanc,1000 Lausanne 26, Switzerland Received May 26, 1994@

DNA adduct formation of 2-amino-3-methylimidazo[4,5-flquinoline (IQ) has been investigated by 32P-postlabeling. Similar adduct profiles were observed from calf thymus DNA modified in vitro with the putative carcinogenic metabolite W-acetoxyamino-3-methylimidazo[4,5-flquinoline (N-acetoxy-IQ) and from hepatic DNA of rats treated with I&. N-(Deoxyguanosin8-yl)-2-amino-3-methylimidazo[4,5-flquinoline (dG-C8-IQ)accounted for approximately 90% of the total adducts observed in calf thymus DNA under postlabeling conditions where ATP was limiting; however, 5-(deoxyguanosin-W-yl)-2-amino-3-methylimidazo[4,5-flq~noline (dG-N2IQ) was detected only when DNA was labeled with excess ATP. Under these labeling conditions, dG-CS-IQ and dG-N2-IQ accounted for approximately 75% and 7% of the total adducts, respectively. Five other spots accounted for the remaining radioactivity. Comparable results were obtained from rat liver DNA. Following DNA adduct enrichment by solid phase extraction, dG-C8-IQ and dG-N2-IQaccounted for 60-76% and 10-13%, respectively, of the total adducts in rat liver. The adduct profiles obtained from reaction of 2'-deoxyguanosine 3'-monophosphate (dG-3'-P04-) with the photoactivated azide derivative of I&, 2-azido-3-methylimidazo[4,5-f]quinoline (&-1Q), were qualitatively similar to those obtained by reaction with N-acetoxy-IQ. The C-8 and N2 adducts were the only reaction products detected. The reactivity and sites of adduct substitution were dependent upon solvent conditions and pH, with increasing adduct formation under alkaline pH. The chemical reactivity of photoactivated N3-IQ with dG-3'P04- was significantly greater than that of N-acetoxy-IQ when reactions were conducted in water, in citrate buffer (pH 5.01,or in phosphate buffer (pH 7.4). Increased reactivity was attributed to increased levels of dG-C8-IQ adduct formation, except for reactions conducted in citrate buffer (pH 5.01,where there was a proportional increase in both C-8 and N2 guanine adducts. However, the chemical reactivity of these two IQ derivatives and their sites of dG substitution were identical when the reactions were conducted in phosphate buffer (pH 9.0). The ratio of the dG-N2-IQadduct to the total adducts increased a t alkaline pH in reactions involving N3-IQ,but the ratio was not affected by a change in the pH of the medium for reactions with N-acetoxy-IQ. The ratio of the dG-N2-IQadduct to the total adducts also increased as a function of phosphate concentration for reactions involving both N-acetoxy-IQ and N3-IQ. Formation of the ring substituted dG-N2-IQadduct indicates that nitrenium and carbenium ion formation occurred for both N-acetoxy-IQ and photoactivated N d Q . The reactivity and the effects of pH and solvent on sites of adduct substitution are distinct for these two chemically reactive derivatives of IQ and suggest that the mechanism of adduct formation is not identical.

Introduction Heterocyclic aromatic amines (HAAS)' are potent bacterial mutagens and carcinogens which are formed in cooked meats and fish at the low part per billion level (1-5). 2-Amino-3-methylimidazo[4,5-flquinoline (IQ) is structurally representative of this class of genotoxins and has recently been shown to be a potent hepatocarcinogen in nonhuman primates (6). Thus, despite the presence of HAAs in low amounts, their appearance in a wide variety of daily staples suggests that human exposure is significant and that HAAs may be involved in the etiology of human cancers (4, 5). The covalent binding of genotoxic carcinogens such as HAAs t o DNA is regarded as a critical event in cancer initiation (7). As is the case with many genotoxins, HAAs *Please address correspondence to this author at the Nest16 Research Centre, Vers-chez-les-Blanc, CH-1000 Lausanne 26, Switzerland. 021-785-8833 (phone), 021-785-8553 (fax). e Abstract published in Advance ACS Abstracts, October 1, 1994.

0893-228X/94/2707-0752$04.50l0

require metabolism to bind to DNA and exert their genotoxic effects. Metabolic activation is catalyzed primarily by cytochrome P450 1A2 to form the arylhydroxylamine, which can bind t o DNA (8-12) or may undergo Abbreviations: dG, a'-deoxyguanosine; dG-3'-P04-, 2'-deoxyguanosine 3'-monophosphate, 3'dNp, 2'-deoxyribonucleoside 3'-monophosphate; HAAs,heterocyclic aromatic amines; IQ, 2-amino-l-methylimidazo[4,5-flquinoline;nitro-I&, 2-nitro-3-methylimidazo[4,5-flquinoline; NHOH-IQ, 2-(hydroxyamino)-3-methylimidazo[4,5-flquinoline; N-acetoxy-IQ, An-a,~xyamino-3-methylimidazo[4,5-flquine; Na-IQ, 2-azido-3-methylimidazo[4,5-flquinoline; dGN2-IQ, 5-(deoxyguanosin-Wyl)-2-amino-3-methylimidazo[4,5-flquinoline; dG-C8-IQ,N-(deoxyguanosin-8-yl)-2-amino-3-methylimidazo[4,5-flquinoline; MeIQx, 2-amino3,8-dimethylimidazo[4,5-flquinoxaline; dGfl-MeIQx, 5-(deoxguanosinhn-yl)-2-amino-3,8-dimethylimidazo[4,5-flquinoxaline; dG-C8-MeIQx, N-(deoxyguanosin-8-yl)-2-amino-3,8-dimethylimidazo[ 4,5-flquinoxaline; PhIP, 2-amino-l-methyl-6-phenylimidazo[4,5-blpyridine; AAF, 2-(acetylamino)fluorene;dGC8-AAF, N-(deoxyguanosin-8-y1)-2-(acetylaminofluorene; dGP-AAF, 3-(deoxyguanosin-An-yl)-AAF; 2-AF',2-aminofluorene; dGC8-M,N-(deoxyguanosin-8-yl)-2-aminofluorene; DTT, dithiothreitol; DMSO, dimethyl sulfoxide; DMF, dimethylformamide; MOPS, (3-(N-morpholino)propanesulfonicacid; PNK, polynucleotide kinase; MNSPD, micrococcal nuclease spleen phosphodiesterase; SPD, spleen phosphodiesterase; RAL, relative adduct labeling; TLC, thinlayer chromatography.

0 1994 American Chemical Society

Chem.Res. Toxicol., Vol. 7, No. 6, 1994 753

C-8a n d N2 Guanine A d d u c t s of IQ

OH I K(deoxyguanosin-8-yl)-2-amino-3-methylimidazo[4,5-tjquinoline (dG-C8-IQ)

AH 5-(deoxyguanosln-N2-yl)-2-amino-3-methylimidazo[4,5-flquinoline (dGN2-IQ)

Figure 1. Chemical structures of I&, dG-C8-IQ, and dG-N2-IQ.The C8 adduct is drawn in the syn conformation, and the N2 adduct is drawn in the anti form.

further activation by esterification to produce highly unstable derivatives which readily react with DNA (1113). Direct 0-acetylation of the N-hydroxy metabolites by acetyl coenzyme A-dependent 0-acetyltransferases is a major pathway for HAA mutagenesis and carcinogenesis (11-13). Sulfotransferase and L-prolyl-tRNA synthetase also have been reported to activate several N-hydroxy HAA derivatives (11). For all the HAAs investigated to date, the major DNA adduct formed in vitro by reaction of the N-hydroxy derivatives or the putative N-acetoxy intermediates with DNA is a C-8 guanine adduct with covalent attachment occurring at the exocyclic amino group of the HAA heteronucleus (13-19). DNA adduct formation also occurs through the respective azide derivatives of HAAs where metabolic activation is replaced by photolytic activation (20-22). The pattern of DNA adducts obtained from photoactivated aryl azides and HAA azides is similar to that obtained with the respective metabolically activated arylamines, nitroarenes, or synthetic N-hydroxy derivatives and suggests that the nitrenel nitrenium ion is the common reactive species (21, 22). Analysis by the 32P-postlabelingtechnique (23)also has shown that the C-8 guanine adduct of I&, as well as those of the structurally related HAAs, 2-amino-l-methyl-6phenylimidazo[4,5-blpyridine (PhIP) and 2-amino-3,8dimethylimidazo[4,5-flquinoxaline(MeIQx), are major adducts formed in rodents and nonhuman primates (17, 24-27), which suggests that these in vitro model systems are representative of DNA adduct formation in vivo. Recently, we reported the synthesis of a second guanine adduct of I& and MeIQx, 5-(deoxyguanosin-W-yl)IQ (dG-N2-IQ)and 5-(deoxyguanosin-W-yl)-MeIQx (dGN2-MeIQx)where adduction occurs at the HAA C-5 atom (Figure 1)(18).The dG-N2adducts were formed at 5-10fold lower levels than the respective dG-CS adducts from reaction of N-acetoxy derivatives with dG or DNA (18). The formation and detection of dG-N2 adducts of IQ and MeIQx have not been reported in vivo. Therefore, the objective of this study was to develop analytical methods to measure the 3,5’-bisphosphate adducts N-(deoxyguanosin-8-yl)-2-amino-3-methylimidazo[4,5-~quinoline(dGC8-IQ) and dG-N2-IQin DNA using both TLC and HPLC following 32P-postlabeling. The chemical reactivities of W-acetoxyamino-3-methylimidazo[4,5-flquinoline (N-ac-

etoxy-IQ) and photoactivated 2-azido-3-methylimidazo[4,5-flquinoline (N3-IQ)with 2’-deoxyguanosine 3’-monophosphate (dG-3’-P04-) and DNA were compared, and the influence of solvent and pH on sites of adduct substitution was investigated in order to increase the quantity of the dG-N2-IQ adduct which is formed in low amounts relative to dG-CS-IQ.

Materials and Methods Chemicals. Caution: ZQ and several of its derivatives are carcinogenic to rodents and should be handled carefully. The following chemicals were obtained from Sigma Chemical Co. (St. Louis, MO): calf thymus DNA, 2’-deoxyguanosine (dG), 2-deoxyguanosine %-phosphate, sodium azide, 3-(N-morpholino)propanesulfonic acid (MOPS), ATP, bicine, calcium chloride, dithiothreitol (DTT), urea, sodium succinate, spermidine, proteinase K, micrococcal nuclease, nuclease P1, and spleen phosphodiesterase (type I). IQ and [2-l4CIIQ, sp act. 10 mCi/mmol, radiochemical purity >98%, were obtained from Toronto Research Chemical (Downsview, Ontario, Canada). Cloned T4 polynucleotide kinase was obtained from New England Biolabs (Beverly,MA). [32P]ATP(7000 CUmmol) was obtained from ICN Chemicals (Imine, CA). PEI-cellulose thin-layer plates and an ET 250/8/4 Nucleosil5;um CISHPLC column were obtained from Machery-Nagel (Duren, Germany). Qiagen tip-2500 columns were purchased through Kontron Instruments (Lausanne, Switzerland). Bond-Elut CIS cartridges (100 mg) were purchased from Analytichem International (ICT AG, Basel, Switzerland). All other chemicals were reagent grade unless specified. Chemical Syntheses. The nitro derivative of IQ was synthesized by the method of Grivas (28)with minor modifications (29). N3-IQ was prepared by reacting 2.4 mg of 2-nitro3-methylimidazo[4,5-flquinoline (nitro-IQ) with a 10-fold mol excess of sodium azide in 0.5 mL of dimethylformamide (DMF) for 18 h at room temperature in the dark. The mixture was then diluted with 10 mL of water and purified with a BondElut CIS cartridge. After thorough washing with water, the azide was eluted with methanol, evaporated to dryness, and resuspended in DMSO/ethanol (4:l). The reaction was quantitative. Electron impact mass spectroscopy (29) confirmed the structure (molecular M+ ion a t m / z 224.08117, calculated 224.08104, with a major fragment ion observed a t m / z 196 [M - N#). 2-(Hydroxyamino)-3-methylimidazo[4,5-flquinoline (NHOH-IQ) was prepared from nitro-IQ as previously described (12)and stored in DMSO/ethanol(4:1) under argon and kept in liquid nitrogen until use. The synthesis of the C8-IQ and N2IQ adducts of dG-Y-PO4- was performed by reacting dG%-P04-

Turesky and Markovic

754 Chem. Res. Toxicol., Vol. 7, No. 6, 1994 (5 mg) with 2 pmol of NHOH-IQ in 5 mL of argon purged 50 mM potassium phosphate containing 0.25 mM EDTA (pH 7.4), which had been preincubated at 37 "C. Immediately after addition of NHOH-IQ, a 10-fold mol excess of acetic anhydride was added. The reaction was allowed to proceed a t 37 "C for 30 min. The reaction mixture was then exhaustively extracted with chloroform. The aqueous phase was applied to a BondElut CIScartridge and purified as described below for the DNA adduct digests. The respective W spectra were in excellent agreement to those observed for dG-C8-IQ and dG-Nz-IQ (I@, and treatment with nuclease P l ( 3 0 ) resulted in disappearance of these adducts and formation of dG-C8-IQ and dG-W-IQ. The 3',5'-bisphopho-dG adducts were prepared enyzmatically by incubating between 0.5 and 3 nmol of 3'-phospho-dGC8-IQ and 3'-phospho-dG-N2-IQ in 0.5 mL of polynucleotide kinase (PNK) buffer (30 mM bicine, 10 mM MgC12,2 mM spermidine, and 10 mM DTT, pH 9.6) containing 50 nmol of ATP and 20 units of PNK for 1h a t 37 "C. The bisphosphate adducts were collected by HPLC under the conditions described below. The conversion to the bisphosphate adducts was quantitative. Reactions with DNA and dG3'-P04-. Reactions of NHOHIQ and N3-IQ with DNA and dG-Y-PO4- were performed in either 1 mL of water, 0.05 M sodium citrate buffer containing 0.25 mM EDTA (pH 5.01, 0.05 M KH2P04, or 0.5 M KH2P04 containing 0.25 mM EDTA and potassium hydroxide to arrive at the final desired pH (pH 5.0, 7.4, or 9.0). Prior to reactions, the buffers were equilibrated a t room temperature and purged with argon. The IQ derivatives (50 nmol) were reacted with either 1 mg of calf thymus DNA or an approximate mole equivalent of dG-Y-PO4- (0.265 mg, 760 nmol). Acetic anhydride (10pL) was diluted with 1mL of ethanol, and 15 pL (a 30-fold mol excess to NHOH-IQ) was immediaetly added t o the solutions containing NHOH-IQ and the reactions were allowed to proceed for 30 min. In the case of N3-IQ, photoactivation was done for 30 min at 366 nm with a UV lamp held 3 cm above the stirred incubation mixture. HPLC analysis revealed that the photoactivation of N3-IQ was complete (29). Decomposition products were removed by exhaustive extraction with chloroform and a precipitation step with NaCVethanol was performed for DNA. Adduct formation was measured by 32P-postlabeling under standard labeling conditions using 0.17 pg of DNA or dG-

3'-Po4-. Animal Experiments. Male Sprague Dawley rats (200250 g) were obtained from Iffa Credo (L'Arbresle, France) and given tap water and chow ad libitum. IQ was administered as its hydrochloride salt in 1 mL of water at a dose of 10 mg/kg. Animals were sacrificed after 6 h by anesthesia with sodium pentobarbital (60 mg/kg ip). Following complete anesthesia, the abdomen was opened and blood was removed. The liver tissue was homogenized in 3 volumes of 10 mM Tris, 140 mM KC1,lO mM EDTA, and 1mM D l l ' (pH 7.4), and the nuclear pellet was obtained by centrifugation of the homogenate a t lOOOOg for 30 min a t 4 "C. A 1g equivalent of tissue was diluted in 10 mL of 50 mM MOPS (pH 7.0) and then digested with 0.5 mL of proteinase K (20 mg/mL) for 30 min at 37 "C. The homogenate was then digested with 4 mL of RNase A (100puglmL), and DNA was purified with a Qiagen column according to specifications provided by Qiagen. The recovery of DNA was approximately 500-700 ,ug/gram of liver tissue. DNA Digestion and Adduct Enrichment. DNA (3.75 pg) was digested in 50 pL of 2 mM sodium succinate, 1 mM CaC12, and 1 mM Tris-HC1, (pH 8.251, containing 3.0 unitd0.38 unit of micrococcal nuclease/spleen phosphodiesterase (MNSPD) in microcentrifuge tubes (note the final pH was 6.3). Incubations were done a t 37 "C for varying times up to the optimal conditions (4 h in vivo, 24 h in vitro). In some instances, digestion was also done in 20 mM sodium succinate and 10 mM CaC12, pH 6.0 (23,30). Samples analyzed directly by postlabeling were lyophilized to dryness. For DNA adduct enrichment, the digested 2'-deoxyribonucleoside 3'-monophosphate (Y-dNps) were diluted with 1 mL of 1mM Na2HP04 acidified to pH 3.0 with acetic acid and applied to a Bond-Elut CIS cartridges (100 mg) preequilibrated in methanol, followed by acidic phosphate

buffer. The cartridge was washed with 3 mL of phosphate buffer (pH 3.0) followed by 3 mL of water. The adducts were then eluted in 2.5 mL of methanol and lyophilized by vacuum centrifugation in the presence of 1pL of 0.5 M Tris-HC1 (pH 8.0). s2P-Postlabeling. The 32P-postlabelingand relative adduct labeling (RAL) calculations were performed a s described by Randerath et al. (23). Analyses were performed under both intensification (ATP deficient) (31)and standard (ATP excess) labeling conditions containing 10 units of PNK in a volume of 20 pL. Labeling was for 60 min a t 37 "C. In the intensification assay, 3.75 pg of DNA was labeled with 55 pmol of [32PlA"F' (7000 Ci/mmol), and in the standard assay 0.17 pg of DNA was labeled with 700 pmol of [32PUTP (400-600 Ci/mmol). Adducts were resolved with the following TLC solvents: D1,l.O M NaH2PO4, (pH 5.8); D2, omitted; D3, 3.6 M lithium formate, 8.5 M urea (pH 3.5); D4, 1.0 M LiCl, 0.5 M Tris-HC1, 8.5 M urea (pH 8.0); D5, 1.7 M NaH2P04, pH 6.0. In some analyses a second D4 was performed (0.7 M NaH2P04, 8.5 M urea, pH 8.0) to increase the resolution between the C-8 and N2 guanine adducts.

Recovery of DNA Adducts, Quantification, and HPLC Analysis. Excised adducts were quantitated by Cerenkov counting with an LKB 1219 Rackbeta scintillation counter. The counting efficiency was estimated at 44%. DNA adducts were recovered by extracting the excised TLC spots with 1 mL of methanol42 N NI-LOH (1:l) for 15 min. The excised adducts were washed a second time and the combined methanoWH4OH eluents were passed through a 0.45 pm filter to remove particulates. The eluent was concentrated to near dryness by vacuum centrifugation. The extracts were resuspended in 100 pL of water conatining ca. 0.25 nmol of 3',5'-bisphospho-dG-C8IQ and 3',5'-bisphospho-dG-W-IQ,which served as W markers for HPLC analysis. The recovery of radioactivity for both bisphosphate adducts was approximately 70%. The analysis of 3'-phospho-dGIQ and 3,5'-bisphospho-dG-IQ adducts was done by HPLC using either a Hewlett Packard 1090M system with a photodiode array detector system or a Varian 5500 system. Radioactivity measurements were performed on-line with a Berthold LB 506 C-1 radioactivity monitor. Resolution of adducts was done with a Machery Nagel ET 250/8/4 Nucleosil 5 pm CIScolumn at a flow rate of 0.5 mumin. The A solvent contained 95% 1 mM Na2HP04 acidified to pH 3.0 with acetic acid and 5% acetonitrile. The B solvent contained 90% acetonitrile and 10%A solvent. The entire run was for 60 min: 0-10 min a t 100% A, followed by a linear gradient to 45% B a t 60 min.

Results Calf thymus DNA modified in vitro with N-acetoxy-

I&was analyzed by the standard and adduct intensification 32P-postlabelingmethods. Representative chromatograms of the DNA digest and the dG-C8-IQ and dG-N2I& adduct standards are shown in Figure 2. It was necessary to use 8.5 M urea in both D3 and D4 solvents to resolve dGN2-I& (adduct 1)from dG-C8-IQ (adduct 2). After 1h of MNSPD hydrolysis, there are nummerous spots and a complex adduct pattern is observed (Figure 2A); however, the adduct pattern becomes simplified over 24 h, and dG-C8-IQ is the predominant adduct (Figure 2B). Under labeling conditions with limiting ATP, the dG-C8-IQ adduct accounted for greater than 90% of the total W s ; however, dGN2-I&was not detected (Figure 2C). As shown in Figure 3, the optimal enzymatic hydrolysis time for extensively modified DNA (2.83 nmol of [l4C1IQboundmg, 9.33 adducts/104 3'dNp, assuming 1 mg of DNA = 3.03 pmol of 3'dNp) was 24 h. The unmodified nucleotides and the dG-N2-IQadduct, which accounted for approximately 5% of the bound IQ, were completely recovered after 1h of hydrolysis. In contrast, recovery of dG-Cs-IQ increased over time with a con-

C-8 and W Guanine Adducts of I&

Chem. Res. Toxicol., Vol. 7, No. 6, 1994 755

Figure 2. 32P-Postlabeling analysis of DNA modified in vitro with N-acetoxy-IQ (2.83 nmol of IQ bound/mg of DNA, 9.33 adducts/ lo4 3’dNp). (A) Labeling with excess [32PlATPafter 1 h MNSPD hydrolysis. (B) Labeling with excess [32P]ATPafter 24 h MNSPD hydrolysis. (C) Labeling with limiting [32PlATPafter 24 h MNSPD hydrolysis. (D) 500 fmol of 3’,5’-bisphospho-dG-N2-IQ (adduct 1) and 3’,5’-bisphospho-dG-C8-IQ (adduct 2).

comitant decrease in the amount of oligomeric/unknown adducts. The percentage of the dG-C&IQ adduct reached a maximum after 24 h where it accounted for 70-75% of the total adducts. The amount of IQ-modified DNA estimated by 32P-postlabelingwas 9.4 f 1.6 adducts in lo4 3’dNp (N = 6) and is in excellent agreement to our 14C based measurement of 9.3 adducts in lo4 3’dNp (Figure 3B). Following 24 h MNSPD hydrolysis, in addition to dG-C8-IQ and dG-N2-I&, there were five additional spots which accounted for approximately 2025% of the total RAL. Prolonged hydrolysis time did not increase the RAL values for either IQ-guanine adduct at the expense of these presumed oligomeric/unknown adducts, and the total RAL value declined after 24 h hydrolysis time.2 The 32P-postlabelingchromatogram of rat liver DNA from rats treated with IQ is shown in Figure 4. Under labeling conditions with limiting ATP, the dG-C8-IQ adduct accounted for 80-90% of the total RALs. As had been observed with calf thymus DNA modified with IQ in vitro, dG-N2-IQ adduct formation was not detected with limiting amounts of ATP. Due to the low levels of DNA adduct formation, adducts were not detected under standard labeling conditions. Therefore, a n adduct enrichment procedure was developed to eliminate unmodified nucleotides and increase the senstivity of DNA adduct detection. Solid phase extraction of digested DNA with Bond-Elut C18 cartridges effectively removed > 97% of the unmodified nucleotides. Based on postlabeling analysis of IQ-modified calf thymus DNA with excess ATP (2.83 nmol of I& bound/mg of DNA), the percent R. J. Turesky, unpublished observations.

recovery of dG-C8-IQ and dG-N2-I& adducts was 88 f 23% and 95 f9.7%, respectively (N = 3). The oligomeric/ unknown adducts were recovered at levels exceeding 50%. When this adduct enrichment procedure was applied to rat liver DNA samples, both dG-C8 and dGN2-IQ adducts were readily detected. Adduct detection following solid phase extraction was 20-fold greater than in the intensification method based upon the relative amount of dG-C8-IQ adduct postlabeled following adduct enrichment versus the adduct labeled directly under intensification conditions. The DNA adducts measured in rat liver of two animals is presented in Table 1. The identities of these IQ-guanine adducts from rat liver were corroborated by HPLC analysis. Approximately 70% of the 32Pradioactivity attributed to 3’,5’bisphospho-dG-C&IQ and 3’,5’-bisphospho-dG-N2-IQ was recovered from the TLC spots by elution with methanol/ 12 N NH40H (1:l). Both bisphosphate adducts were stable to this alkaline elution solvent, and 90% of the radioactivity coeluted with each adduct by HPLC. An analysis of the recovered adducts measured simultaneously is shown in Figure 5, which shows that the UV markers of the unlabeled 3’,5’-bisphopho-dG derivatives of dG-C8-IQ and dG-N2-IQcoelute with the 32Pradioactivity. dG-N2-IQ and dG-C8-IQ adduct formation from reaction of dG-3’-P04- and DNA with N-acetoxy-IQ and N3IQ was investigated. The results are summarized in Table 2, and representative chromatograms are depicted in Figure 6. Both dG-N2-IQ and dG-C8-IQ are formed by reaction of dG-3’-P04- with either N-acetoxy-IQ or N3I&; no other adduction products were detected. Adduct stabilities in the various buffers were determined, and

Turesky and Markovic

756 Chem. Res. Toxicol., Vol. 7, No. 6, 1994 100% 80% --

60% -40% --

20% -r

B

I I

1 1 - 1 I

T

TIooi

was equivalent for both reactive IQ derivatives. The contribution of dG-N2-IQ to total adduct formation increased under alkaline pH for the photoactivated azide, but the ratio remained relatively constant in reactions involving N-acetoxy-IQ. Adduct formation and sites of substitution were comparable for N-acetoxy-IQ and N3IQ when the reaction was conducted in alkaline phosphate buffer (pH 9.0). Increasing the phosphate concentration from 50 to 500 mM (pH 7.4 or 9.0) resulted in an overall decrease in adduct formation for both IQ derivatives; however, phosphate influenced the site of adduct substitution, and the percent contribution of the dG-N2 adduct to the total adducts increased by as much as 2-3fold. Qualitatively similar adduct profiles were observed in DNA that was treated with either N-acetoxy-IQ or N3I&. The level of adduction was 6-fold greater in reactions with N3-IQ-modified DNA, which was also reflected by elevated levels of dG-C8-IQ and dG-N2-I&. The ratio of the dG-N2-IQ adduct to total adduct formation was slightly greater in DNA modified with N-acetoxy-IQ, but the increase was not statistically significant.

Discussion

hydrolysis time

Figure 3. Recovery of dG-C8-IQ and dG-N2-IQ from calf thymus DNA (2.83 nmol of IQ bound/mg of DNA, 9.33 adducts/ lo43’dNp) as a function of time. (A) Percentage of total adducts present as dG-N2-IQ, dG-C8-IQ, and oligomeridunknown adducts. (B)Total RAL and CPM recovery as a function of time.

the percent recovery is reported in Table 3. Adduct recovery was greater for both adducts at pH 7.4 and 9.0 than at pH 5.0. In large scale syntheses analyzed by HPLC, some hydrolysis of the 3’-phosphate ester of the dG-3’-P04- adducts was observed at pH 5.0;3 however, there was no evidence of adduct decomposition at any of the pH ranges examined. The decrease in adduct recovery measured by postlabeling also may be attributed in part to decreased labeling efficiency of the kinase in the presence of the various salts and buffers. The small variations in adduct stablity in the various solvents and pH ranges investigated cannot explain the large differences in reactivity of N-acetoxy-IQ and N3-IQ with dG3’-Po4- and the sites of adduct substitution. The reactivity and sites of adduct substitution were dependent upon solvent conditions, and adduction increased as a function of pH. The chemical reactivity of photoactivated N3-IQ was significantly greater than the reactivity of N-acetoxy-IQ when reactions were conducted in water, in 50 mM citrate buffer (pH LO), or in 50 mM potassium phosphate buffer (pH 7.4) and was attributed to significantly elevated levels of dG-C8-IQ. Comparable reactivity was observed when reactions were conducted in 50 mM potassium phosphate buffer (pH 5.0). However, in all of these reaction conditions, except for the citrate buffer, the contribution of dG-N2-IQ to total adduct formation was 2-&fold greater in the reactions involving N-acetoxy-IQ. In citrate buffer, the percent contribution of dG-N2-IQ to the total adduct formation R. J. Turesky, unpublished observations.

A number of studies have been reported on HAA-DNA adduct formation and 32P-postlabelinganalysis (17, 19, 22,24-27,32-35). There is considerable interlaboratory variation in the DNA adduct profiles, which appear quite complicated in many instances. A part of the variation in results is attributed to differences in the conditions used for enzymatic digestion of adducted DNA and to the chromatographic procedures employed for adduct resolution. We observed that enzymatic digestion of calf thymus DNA extensively modified with IQ required 24 h, although digestion of DNA from animals containing low levels of adduction was complete within 4 h. Nonmodified nucleotides and the dG-N2-IQ adduct were completely recovered from IQ-bound calf thymus DNA after 1h enzyme hydrolysis, which indicates that acetylation of DNA by acetic anhydride during the in situ formation of N-acetoxy-IQ did not significantly affect the efficiency of MNSPD hydrolysis. However, dG-C8-IQ was a relatively poor substrate for MNSPD and required prolonged hydrolysis time. The optimal pH of spleen phosphodiesterase (SPD) is between pH 6 and 7, and the pH optimum of MN is between pH 9 and 10, depending on the substrate and the calcium concentration which also may have inhibitory effects (36, 37). In our hands, a final CaC12 concentration of 1 mM in the MNSPD hydrolysis (final pH 6.3) was superior to the conventionally used 10 mM CaC12 at pH 6.0 (23) and resulted in a simpler DNA adduct profile with fewer oligomeric adducts of 3’,5’-bisphospho-dG-C8-IQ. RAL values of dGC8-IQ were 30-50% greater when MNSPD hydrolysis was performed in 1 mM CaC12 than in 10 mM CaC12, although values for dG-N2-IQ and the total RAL were ~ n c h a n g e d . Under ~ optimal MNSPD digestion conditions, dG-C8-IQ accounted for 70-75% and dG-N2-IQ accounted for approximately 3-7% of the total adducts when postlabeled with excess ATP (Figure 3). These estimates closely agree to our data obtained when the DNA was digested to deoxynucleosides followed by quantification with HPLC (18).It is not known whether the remaining five spots detected by 32P-postlabeling (Figure 2) are unique adducts or MNSPD resistant R. J. Turesky, unpublished observations.

C-8 and N2 Guanine Adducts of IQ

Chem. Res. Toxicol., Vol. 7, No. 6,1994 757

e

Figure 4. 32P-Postlabeling analysis of rat liver DNA. (A) DNA from an untreated rat. (B) DNA from a rat 6 h after treatment with IQ (10 mgkg). (C) DNA from an untreated rat enriched by solid phase extraction. (D) DNA from a rat enriched by solid phase extraction 6 h aRer treatment with I&.All samples (3.75 ,ug of DNA or equivalent after solid phase extraction) were postlabeled with 55 pmol of [32PlATP(7000 Cdmmol).

Table 1. Analysis of dG-W-IQand dG-Cs-IQAdduct Formation in Rat Liver after Solid Phase Extractiona totalRAL x rat 1 rat 2

total

dG-C8-IQ

dG-N2-IQ

5.42 f 1.66 5.48 f 0.79

3.34 1.18 4.19 f 0.69

0.71 f 0.29 0.67 f 0.10

gJ

h

Y

a 32PPostlabeling analysis of liver of rats treated with IQ (10 mgkg) and sacrificed after 6 h (average f SD, N = 3). RAL values were based upon the normal nucleotide estimate of 3.75 pug of DNA = 11.4 nmol of 3’dNps.

oligomeric adducts of dG-C8-IQ. We also noticed several adduction products accounting for 20-25% of the IQbound DNA following enzymatic hydrolysis to deoxynucleosides (18). Recovery of both dG-C8-IQ and dG-N2I& adducts was low when IQ-modified calf thymus DNA was digested with MNSPD a t pH 9.0, and the adduct pattern was complex with many spots. Performing the MNSPD hydroylsis for 4 h at pH 6.3 followed by a 20 h incubation at pH 9.0 also led to significantly lower recovery of dG-C8-IQ than a 24 h hydrolysis at pH 6.3.5 These results suggest that SPD is the critical enzyme in the complete hydrolysis of IQ-DNA adducts to monomers. The major DNA adduct formed in vitro by reaction of N-acetoxy-IQ with dG and DNA is dG-C8-IQ (13,18). This is the major adduct detected in bacteria, rodents, and nonhuman primates undergoing carcinogen bioassays with IQ when analyzed by 32P-postlabeling(24-27). Our data confirm that dG-C8-IQ is the principal adduct R. J. Turesky, unpublished observations.

0

10

20

30

40

50

60

Time (min)

Figure 5. HPLC analysis of 3’,5’-bisphospho-dG-C8-IQ and 3’,5’-bisphospho-dG-N2-IQ rat liver DNA adducts excised from PEI-cellulose TLC plates. 32P-Radiolabeledadducts were coinjected with approximately 0.25 nmol of unlabeled bisphosphate adducts, which served as W markers.

in rat liver DNA. In addition, we have shown that the second major adduct observed in rat liver is dG-N2-IQ (Figure 4, Table 1). The resolution of 3’,5’-bisphosphodG-N2-IQfrom 3’,5’-bisphospho-dG-C8-IQ by TLC with commercially available plates required 8.5 M urea. Lower concentrations of urea commonly used in many HAA-DNA adduct studies resulted in the comigration of these two adducts in our laboratory. DNA adduct analysis by 32P-postlabeling is often performed under conditions where the amount of ATP is limiting and PNK preferentially labels aromatic DNA adducts over unmodified nucleotides and, thus, results

Turesky and Markovic

758 Chem. Res. Toxicol., Vol. 7, No. 6,1994

Table 2. dG-W-IQand dG-Cs-IQAdduct Formation from Reaction of N-Acetoxy-IQand Ns-IQ as a Function of Solvent and pHa total RAL x reaction with dG-X-PO4substrate

solvent

N-acetoxy-IQ N3-IQ N-acetoxy-IQ N3-IQ N-acetoxy-IQ N3-IQ N-acetoxy-IQ N3-IQ N-acetoxy-IQ N3-IQ N-acetoxy-IQ N3-IQ N-acetoxy-IQ N3-IQ

water water sodium citrate, pH 5.0 sodium citrate, pH 5.0 50 mM m o 4 - , pH 5.0 50 mM -04-, pH 5.0 50 mM I(po4-, pH 7.4 50 mM m o d - , pH 7.4 500 mM -04-, pH 7.4 500 mM -04-, pH 7.4 50 mM mod-, pH 9.0 50 mM -04-, pH 9.0 500 mM KPo4-, pH 9.0 500 mM -04-, pH 9.0

total 2.0 f 0.9 11.3 f 0.4* 1.4 f 0.4 5.0 f 0.7* 8.0 f 4.8 16.3 f 6.3 81 f 9 195 f 35* 42 f 13 115 f 30* 426 f 183 315 f 80 74 f 16 208 f 103**

dG-C8-IQ 1.8 f 0.9 11.1f 0.4*

1.2 f 0.3 4.4 f 0.9* 7.3 f 4.3 15.6 f 6.2 71f8 189 f 33* 33 f 10 102 f 18* 384 f 170 281 f 75 61 f 15 168 f 82**

dG-N2-I& 0.13 f 0.08 0.19 f 0.06 0.14 f 0.03 0.46 f 0.07” 0.7 f 0.4 0.8 f 1.2 llfl 9 f 2 9 f 2 20 f 4* 42 f 10 34 f 7 13 f 2 39 f 21**

% dG-N2-IQ/total adducts

7.7 f 3.2 1.7 f 0.5* 9.9 f 2.1 8.7 f 1.1 8.1 f 0.3 5.0 f 1.1* 13.3 f 0.3 4.7 f 0.3” 21.1 f 1.2 16.3 f 2.1* 10.2 f 1.7 10.8 f 1.4 18.8 f 0.9 18.0 f 3.6

total RAL x

reaction with DNA substrate

solvent

total

dGC8-IQ

dG-N2-IQ

% dGN2-IQ/total adducts

N-acetoxy-IQ N3-IQ

50 mM KP04-, pH 7.4 50 mM KP04-, pH 7.4

4.1 f 2.1 24.5 f 8.9**

3.0 f 1.5 18.0 f 7.0**

0.3 f 0.2 1.2 f 0.3**

7.5 f 1.8 4.9 f 1.2

+

a Reaction conditions: 50 nmol of NHOH-IQ 1500 nmol of acetic anhydride or 50 nmol of N3-IQ with 760 nmol of dG-3’-P04- or 1 mg/mL DNA (ca. 760 nmol of dG-3’-P04-) in water, 50 mM sodium citrate (pH 5.0), 50 or 500 mM KH2P04- containing 0.25 mM EDTA (pH 5.0,7.4, or 9.0). It is assumed that the recovery of dG-C8-IQ and dG-N2-IQfrom calf thymus DNA is 100%. The percent of oligomeric/ unknown postlabeled material with DNA was 18.3 f 11.5% and 22.4 f 2.7% for DNA modified with N-acetoxy-IQ and Ns-IQ, respectively. Values are the average f SD of 3 or 4 experiments. Unpaired t-test: *P < 0.005, **P< 0.05, of respective adduct formation of NHOH-IQ The data for % dG-N2-IQ/total adducts was tested for statistical significance after logistical transformation. Adduct recoveries vs from Table 3 have not been taken into account.

Figure 6. 32P-Postlabelinganalysis of dG-3’-P04- and DNA adduction products with N-acetoxy-IQ and photoactivated N3-IQ reacted in 50 mM KH2P04- buffer pH 7.4. (A) Reaction of dG-3’-P04- with N-acetoxy-IQ. (B)Reaction of dG-3’-P04- with N3-IQ. (C) Reaction of DNA with N-acetoxy-IQ. (D) Reaction of DNA with N3-IQ photoactivated. All postlabeling analyses were conducted with excess [32P]ATP.

in increased sensitivity of adduct detection (31). The intensification labeling method enhances the sensitivity of dG-C8-IQ adduct analysis (24-26);however, the dG-

N2-IQadduct is a relatively poor substrate for PNK, and it is not labeled under limiting ATP conditions. Conserequires quently, analysis of 3’,5’-bisphospho-dG-N2-IQ

C-8 and 1v2 Guanine Adducts of I&

Chem. Res. Toxicol., Vol. 7, No. 6, 1994 759

Table 3. Percent Recovery of dG-W-IQ-3-PO4-and dG-CB-IQ-3-PO4-as a Function of pH and Buffer by 32P Postlabeline % recovery of adducts

buffer

dG-N2-IQ

dG-C&IQ

water 50 mM citrate, pH 5.0 50 mM I(po4, pH 5.0 50 mM I(po4,pH 7.4 50 mM KP04, pH 9.0

74-93 55-59 39-41 60-69 58-64

96-100 63-73 61-66 67-82 68-71

a Values obtained from 2 independent experiments. 43 pmol of dG-C8-IQ-X-P04- and 7 pmol of dG-N2-IQ-3’-P04- were incubated in 0.2 mL of respective buffers for 30 min a t room temperature followed by dilution with water to 1 mL. Then, 20 pL was removed, neutralized with 1 p L of 0.5 M Tris (pH 8.01, lyophilized to dryness, and postlabeled. The percent recovery of adducts is based upon the theoretical amount of 860 fmol for dGC8-IQ and 140 fmol for dG-N2-I&.

postlabeling of modified DNA with excess [32PlATP. The dG-C8-IQ adduct was readily detected in rat liver DNA postlabeled with limiting ATP; however, adduction levels were too low to measure with standard labeling conditions, and it was necessary to enrich the DNA adducts. Butanol extraction is an effective means for enrichment of various classes of DNA adducts (38), but 3’-phospho-dG-C&IQand 3’-phospho-dG-N2-IQare poorly extracted with this solvent. Nuclease P1 also has been reported to enhance the sensitivity of diverse classes of carcinogen DNA adducts where the modified nucleotide is resistant to the 3’-dephosphorylating action of the enzyme (30). This method has recently been used to detect DNA adducts of the structurally related HAA mutagen MeIQx (39). In our laboratory nuclease P1 hydrolyzed all of the 3’-phospho-dGN2-IQ and about 90% of 3‘-phospho-dG-C8-IQ in IQ-modified calf thymus.6 Therefore, we opted for DNA adduct enrichment by solid phase extraction with Bond-Elut cartridges, which increased the sensitivity over the intensification method by 20-fold for the dG-C8-IQ adduct and also enabled the detection of dG-N2-I&. Since there are no internal standards in this adduct enrichment procedure, the RAI, values of the adducts are estimates. The dG-C8-IQ and dG-N2-IQ adducts accounted for 60-76% and 10-13%, respectively, of the total adducts recovered in these two rat liver samples. The ratio of dG-N2-IQ/dG-C8-IQ formation is comparable to the ratio detected in calf thymus DNA modified in vitro with N-acetoxy-IQ. The chemical reactivity of N-acetoxy-IQ and photoactivated N3-IQ was investigated with dG-3’-P04- and DNA. The differences in adduct formation and sites of substitution as a function of pH indicate that the chemically reactive intermediate(s) and the mechanism of N-acetoxy-IQ and N3-IQ adduction with dG-3’-P04- are not identical. Reactivity and adduct substitution sites were dependent upon the solvent conditions and pH; both C-8 and N2 guanine adduct formation increased as a function of pH (Table 2). Adduct formation was considerably greater for N3-IQ than for N-acetoxy-IQ when the reaction was conducted in water or phosphate buffer (pH 7.4) and was attributed to significantly higher levels of dG-C8-IQ formation. However, the ratio of the dG-N2IQ/total adducts was greater for reactions involving N-acetoxy-IQ. The pattern of adduct substitution was also dependent upon the solvent. When reactions were conducted in citrate buffer (pH LO), the reactivity of N3R. J. Turesky, unpublished observations.

IQ with dG-3’-P04- was once again greater than that of N-acetoxy-IQ; however, the ratio of dG-N2-IQ/total adducts was the same in both reactions. When reactions were performed under alkaline conditions (phosphate buffer pH, 9.01, the chemical reactivity and the pattern of dG substitution sites were identical for these two activated derivatives of IQ. The ratio of dG-N2-IQ/total adducts increased under alkaline pH in reactions involving N3-IQ7which suggests increased reactivity of the N2 group at alkaline pH. However, the ratio remained relatively constant over the pH range investigated for reactions involving N-acetoxy-IQ. The concentration of phosphate also appeared to influence chemical reactivity and adduct substitution sites. With elevated concentrations of phosphate (500 mM), it was possible to increase the percentage of dG-N2-IQto total adducts by as much as 2-3-fold, which suggests that phosphate may be interacting with the reactive nitrenelnitrenium ion. Increased reactivity of the N2 group of dG a t alkaline pH has also been reported by Pei and Moschel (40) where W-(benz[alanthracen-7-ylmethyl)-2’-deoxyguanosine formation from reaction of 7-(bromomethyl)benz[alanthracene with dG was higher in alkaline solution than in water. Kriek (41) also reported that the ratio of 3-(deoxyguanosin-W-yl)-AAF(dG-N2-AAF)to N-(deoxyguanosin-8-yl)-2-(acetylamino)fluorene(dG-C8-AAF)adduct formation increased as a function of pH in the reaction of N-acetoxy-AAF with DNA, however, the ratio of adduct substitution products of the N-sulfate of N-hydroxy-4-(acetylamino)biphenyl was not affected by a change in pH of the reaction medium. The formation of the ring substituted dG-N2-IQadduct by both N-acetoxy-IQ and N3-IQindicates that nitreniud carbenium ion formation occurred by heterolytic cleavage of N-acetoxy-IQ to produce the nitreniudcarbenium cation-acetate anion pair (42)and, by protonation of the photoactivated product of N3-IQ,the nitrene (20-22). The distinct changes in adduct substitution sites for Nacetoxy-IQ and N3-IQ with dG-3’-P04- as a function of solvent and pH reveal that there are indeed differences in the reactive intermediate species. A similar ratio of dG-N2-IQ/total adduct formation was observed in calf thymus DNA modified with photoactivated N3-IQ and with N-acetoxy-IQ under conditions where differences in the ratio of dG-N2-IQ/total adduct formation were observed for the respective reactions with dG-3’-P04- (Table 2, Figure 6). This suggests that macromolecular DNA structure, in addition to the relative contribution of the nitrenium ion and carbocation resonance forms, influences the site of adduct substitution. Further investigations on the mechanisms involved in IQ-DNA adduct formation are warranted. Although dG-N2-IQ adduction is less prominent than dG-C8-IQ formation, the relative repair rates and persistence of these two adducts in vivo is presently not known. ‘H-NMR spectroscopy revealed that the anti form is the preferred conformation of dG-N2-IQ,while the syn form is preferred for dG-Cs-IQ (18). Differences in DNA adduct conformation also have been reported for the C-8 guanyl adducts of 2-AF and AAF,which preferentially exist in the anti and syn forms, respectively (43). These conformational differences also appear in DNA and affect the biological properties and relative persistence of these adducts in vivo (44-46). The dG-C&AAF adduct appears to induce a greater distortion of the DNA helix than the dG-C&AF adduct, which results in a more rapid removal of dG-Ct3-M in vivo and a greater persistence

760 Chem. Res. Toxicol., Vol. 7, No. 6, 1994 of dG-C8-AF (47-50). Biological data also suggest that the dG-N2-AAFadduct does not cause major distortions of the native DNA structure, and this adduct also persists in rodents (44-47).The preferential conformation of d G C8-IQ and dG-N2-IQ in DNA and their effect on the biological properties of the adducts require further investigation. It is important to assess the relative contribution that these two IQ-guanine adducts have in animals which have been chronically exposed to this carcinogen to better evaulate their respective roles in tumorigenesis (51,52).

Acknowledgment. We would like to thank Dr. F. Kadlubar and Dr. K. Kaderlik of the National Center of Toxicological Research, Jefferson, AR,for helpful discussions; Dr. s. Rossi, Nestec Ltd., for introduction of the 32P-postlabelingmethod; and Dr. J. M. Aeschlimann for in statistical analyses. References Felton, J. S., Knize, M. G., Shen, N. H., Andresen, B. D., Bjeldanes, L. F., and Hatch, F. T. (1986) Identification of the mutagens in cooked beef. Environ. Health Perspect. 67, 17-24. Felton, J. S., and Knize, M. G. (1990) Heterocyclic amine mutagendcarcinogens in foods. In Handbook of Experimental Pharmacology (Cooper, C. S., and Grover, P. L., Eds.) Vol. 94iI, pp 471-502, Springer-Verlag, Berlin and Heidelberg. Sugimura, T. (1986) Studies on environmental chemical carcinogenesis in Japan. Science 233, 312-318. Sugimura, T. (1988) Successful use of short-term tests for academic purposes: their use in identification of new environmental carcinogens with possible risk for humans. Mutat. Res. 206,33-39. Weisburger, J. H. (1987)Mechanisms of nutritional carcinogenesis associated with specific human cancers. ISI Atlas Sci.: Pharmacol. 1, 162-167. Adamson, R. H., Thorgeirsson, U. P., Snyderwine, E. G., Reeves, J., Dalgard, D. W., Takayama, S., and Sugimura, T. (1990) Carcinogenicity of 2-amino-3-methylimidazo[4,5-flquinoline in nonhuman primates: induction of tumors in 3 macaques. Jpn. J. Cancer Res. 81, 10-14. Miller, J. A. (1970) Carcinogenesis by chemicals: An overview. Cancer Res. 30, 559-576. Yamazoe, Y., Shimada, M. Shinohara, A., Saito,K. Kamataki, T., and Kato, R. (1983) Microsomal activation of 2-amino-3methylimidazo[4,5-flquinoline, a pyrolysate of sardine and beef extracts, to a mutagenic intermediate. Cancer Res. 43, 57685774. Shimada, T., Iwasaki, M., Martin, M. V., and Guengerich, F. P. (1989) Human liver microsomal cytochrome P-450 enzymes involved in the bioactivation of procarcinogens detected by umu gene response in Salmonella typhimurium TA1535/pSk1002. Cancer Res. 49, 3218-3228. Butler, M. A., Iwasaki, M. Guengerich, F. P., and Kadlubar, F. F. (1989)Human cytochromeP-45OPA(P-450IA2),the phenacetin 0-deethylase, is primarily responsible for the hepatic 3-demethylation of caffeine and N-oxidation of carcinogenicarylamines. Pm. Natl. Acad. Sci. U S A . 86, 7696-7700. Kato, R. (1986) Metabolic activation of mutagenic heterocyclic amines from protein pyrolysates. CRC Crit. Rev. Toxicol. 16,308347. Turesky, R. J., Lang, N. P., Butler, M. A., Teitel, C. H., and Kadlubar, F. F. (1991) Metabolic activation of carcinogenic heterocyclic aromatic amines by human liver and colon. Carcinogenesis 12, 1839-1845. Snyderwine, E. G., Roller, P. P., Adamson, R. H., Sato, S.,and Thorgeirsson, S. S. (1988) Reaction of the N-hydroxylamine and N-acetoxy derivatives of 2-amino-3-methylimidazo[4,5-flquinoline with DNA. Synthesis and identification of N-(deoxyguanosin-8yl)-IQ. Carcinogenesis 9, 1061-1065. Hashimoto, Y., Shudo, K. and Okamoto, T. (1982) Modification of DNA with potent mutacarcinogenic 2-amino-6-methyldipy~ido[1,2-a:3’,2’-dlimidazole isolated from a glutamic acid pyrolysate: structure of the modified nucleic acid base and initial chemical event caused by the mutagen. J. Am. Chem. Soc. 104,7635-7640. Hashimoto, Y., Shudo, K., and Okamoto, T. (1980) Activation of

Turesky and Markouic a mutagen, 3-amino-l-methyl-5H-pyrido[4,5-b]indole. Identification of 3-hydroxyamino-l-methyl-5H-pyrido[4,3-blindole and its reaction with DNA. Biochem. Biophys. Res. Commun. 96, 355362. (16) Frandsen, H., Grivas. S., Andersson, R., Dragsted, L., and Larsen, J. C. (1992) Reaction of the Nz-acetoxy derivative of 2-amino-lmethyl-6-phenylimidazo[4,5-blpyridine (PhIP) with 2’-deoxyguanosine and DNA. Synthesis and identification of NW-deoxyguanosin-8-yl)-PhIP. Carcinogenesis 13, 629-635. (17) Lin, D., Kaderlik, K. R., Turesky, R. J., Miller, D. W., Lay, J. O., Jr., and Kadlubar, F. F. (1992)Identification ofN-(deoxyguanosin8-yl)-2-amino-l-methyl-6-phenylimidazo[4,5-blpyridine as the major adduct formed by the food-bomecarcinogen 2-amino-l-methyl6-phenylimidazo[4,5-blpyridine,with DNA. Chem. Res. Toxicol. 5,691-697. (18) Turesky, R. J., Rossi, S. C., Welti, D. H., Lay, J. O., Jr., and Kadlubar, F. F. (1992) Characterization of DNA adducts formed in vitro by reaction of N-hydroxy-2-amino-3-methylimidazo[4,5flquinoline and N-hydroxy-2-amino-3,8-dimethylimidazo[4,5-flauinoxaline at the C-8 and N2 atoms of euanine. Chem. Res. ” h ~ i c o l5, . 479-490. (19) Naeaoka. H.. Wakabavashi. K.. , Kim.. S.-B.. Kim. I.-S.. Tanaka. Y.,-bchiai, M., Tada, k , N&aya, H., Sugimura,T.,and Nagao; M. (1992)Adduct formation at C-8 of guanine on in vitro reaction of the ultimate form of 2-amino-l-methyl-6-phenylimidazo[4,5blpyridine with 2’-deoxyguanosine and its phosphate esters. Jpn. J . Cancer Res. 83, 1025-1029. (20) Sarrif, A. M., White, W. E., and DiVito, N. (1978) Photolysis of 2-azidofluorene in situ as a probe in chemical carcinogenesis; bypass of requirement for metabolic activation. Biochem. Biophys. Res. Commun. 83,506-512. (21) Wild, D., Fasshauer, I., and Henschler, D. (1989) Photolysis of arylazides and generation of highly electrophilic DNA-bindingand mutagenic intermediates. Carcinogenesis 10, 335-341. (22) Kerdar, R. S., Fabhauer, I., Probst, M., Blum, M., Meyer, U. A., and Wild, D. (1993)32P-Postlabellingstudies on the DNA adducts of the food mutagendcarcinogens IQ and PhIP adduct formation in a chemical system, and by rat and human metabolism. In Postlabelling Methods for Detection of DNA Adducts (Phillips, D. H., Castegnaro, M., and Bartsch, H., Eds.) pp 173-179, Intemational Agency for Research on Cancer, Lyon. (23) Randerath, K., E. Randerath, Agrawal, H. P., Gupta, R. C., Schurdak, M. E., and Reddy, M. V. (1985) Postlabeling methods for carcinogen-DNA adduct analysis. Envrion. Health Perspect. 62, 57-65. (24) Schut, H. A. J., Putman, K., and Randerath, K. (1988) DNA adduct formation of the carcinogen 2-amino-3-methylimidazo[4,5flquinoline in target tissues of the F-344 rat. Cancer Lett.41,345352. (25) Zu,H.-X., and Schut, H. A. J. (1991) Formation and persistence of DNA adducts of 2-amino-3-methylimidazo[4,5-flquinoline in Male Fischer-344 rats. Cancer Res. 51, 5636-5641. (26) Schut, H. A. J., Snyderwine, E. G., Zu,H.-X., and Thorgeirsson, S.S.(1991)Similar patterns of DNA adduct formation of 2-amino3-methylimidazo[4,5-flquinolinein the Fischer 344 rat, CDFl mouse, cynomolgus monkey and Salmonella typhimurium. Carcinogenesis 12, 931-934. (27) Snyderwine, E. G., Davis, C. D., Nouso, K., Roller, P. P., and Schut, H. A. J. (1993)32P-Postlabelinganalysis of IQ, MeIQx and PhIP adducts formed in vitro in DNA and polynucleotides and found in vivo in hepatic DNA from IQ-, MeIQx- and PhIP-treated monkeys. Carcinogenesis 14, 1389-1395. (28) Grivas, S. (1988) Synthesis of 3,8-dimethyl-2-nitro-3H-imidazo[4,5-flquinoxaline, the 2-nitro analogue of the food carcinogen MeIQx. J. Chem. Res. (S), 84. (29) Turesky, R. J., Bracco-Hammer, I., Markovic, J., Richli, U., Kappeler, A. M., and Welti, D. H. (1990) The contribution of N-oxidation to the metabolism of the food-borne carcinogen 2-amino-3,8-dimethylimidazo[4,5-flquinox&nein rat h e p a e s . Chem. Res. Toxicol. 3, 524-535. (30) Reddy, M. V., and Randerath, K. (1986) Nuclease P1-mediated enhancement of sensitivity of 32P-postlabelingtest for structurally diverse DNA adducts. Carcinogenesis 7, 1543-1551. (31) Randerath, E., Agrawal, H. P., Weaver, J. A., Bordelon, C. B., and Randerath, K. (1985) 32P-Postlabeling analysis of DNA adducts persisting for up to 42 weeks in the skin, epidermis and dermis of mice treated topically with 7,12-dimethylbenz[alanthracene. Carcinogenesis 6, 1117-1126. (32) Hall, M., She, M. N., Wild, D., Fasshauer, I., Hewer, A., and Phillips, D. H. (1990)Tissue distribution of DNA adducts in CDFI mice fed 2--~3-methylimidazo[4,5-flquinoline (IQ)and 2-amine 3,4-dimethylimidazo[4,5-flquinoline (MeIQ). Carcinogenesis 11, 1005-1011. (33) Pfau, W., O’Hare, M. J., Grover, P. L., and Phillips, D. H. (1992) Metabolic activation of the food mutagens 2-amino-3-methylimi-

C-8 and N2 Guanine Adducts of I&

Chem. Res. Toxicol., Vol. 7, No. 6, 1994 761

dazo[4,5-flquinoline (IQ) and 2-amino-3,4-dimethylimidazo[4,5flquinoline (MeIQ) to DNA binding species in human mammary epithelial cells. Carcinogenesis 13, 907-909. (34) Peluso, M., Castegnaro, M., Malaveille, C., Talaska, G., Vineis, P., Kadlubar, F., and Bartsch, H. (1990)32P-Postlabelinganalysis of DNA adducted with urinary mutagens from smokers of black tobacco. Carcinogenesis 11, 1307-1311. (35) Peluso, M., Castegnaro, M., Malaveille, C., Friesen, M., Garren, L., Hautefeuille, A., Vineis, P., Kadlubar, F., and Bartsch, H. (1991)32Postlabelinganalysis of urinary mutagens from smokers of black tobacco implicates 2-amino-1-methyl-6-phenylimidazo[4,5-blpyridine (PhIP) as a major DNA-damaging agent. Curcinogenesis 12, 713-717. (36) Bernardi, G. (1971) Spleen acid deoxyribonuclease. In The Enzymes (Boyer, P. D., Ed.) Vol. IV,pp 271-287, Academic Press, New York and London. (37) Anfinsen, C. B., Cuatrecasas, P., and Taniuchi, H. (1971) Staphylococcal nuclease, chemical properties and catalysis. In The Enzvmes (Bover. P. D.. Ed.) Vol. IV.DD 177-204. Academic Press. Ne; York and London. (38) Gupta, R. C. (1985) Enhanced senstivitv of 32P-~ostlabelinp analysis of aromatic carcinogen DNA adducts. Cancer Res. 4 5 5656-5662. (39) Ochiai, M., Nagaoka, H., Wakabayashi, K., Tanaka, Y., Kim, S.B., Tada, A,, Nukaya, H., Sugimura, T., and Nagao, M. (1993) Identification of N2-(dexoyguanosin-8-yl)-2-amino-3,8-dimethylimidazo[4,5-flquinoxaline 3’,5’-diphosphate, a major DNA adduct, detected by nuclease P1 modification of the 32P-postlabeling method, in the liver of rats fed MeIQx. Carcinogenesis 14,21652170. (40) Pei, G. K., and Moschel., R. C. (1990) Arylalkylation of 2’-deoxyguanosine: Medium effects on sites of reaction with 74bromomethyl)benz[a]anthracene. Chem. Res. Toxicol. 3,292-295. (41) Kriek, E. (1979) Effect of pH on the ratio of substitution products in DNA after reaction with the carcinogen N-acetoxy-2-acetylaminofluorene. Cancer Lett. 7, 141-146. (42) Kadlubar, F. F., and Beland, F. A. (1985) Chemical properties of ultimate carcinogenic metabolites of arylamines and arylamides. In Polycyclic Hydrocarbons and Carcinogenesis (Harvey, R. G., I

_

_

Ed.) ACS Symposium Series 283, pp 331-370, American Chemical Society, Washington, D.C. (43) Evans, F. E., Miller, D. W., and Beland, F. A. (1980) Sensitivity of the conformation of deoxyguanosine to binding at the C-8 position by N-acetylated and unacetylated 2-aminofluorene. Curcinogenesis l , 955-959. (44) Westra, J. G., Kriek, E., and Hittenhausen, H. (1976) Identification of the persistently bound form of the carcinogen N-acetyl-2aminofluorene to rat liver DNA in vivo. Chem.-Biol. Interact. 16, 149-164. (45) Beland, F. A., Dooley, K. L., and Jackson, C. D. (1982)Persistence of DNA adducts in rat liver and kidney after multiple doses of the carcinogen N-hydoxy-2-acetylaminofluorene(1982). Cancer Res. 42, 1348-1354. (46) Gupta, R. C. (1988) 32P-Adductassay: Short- and long-term persistence of 2-acetylaminofluorene-DNA adducts and other applications of the assay. Cell Biol. Toxicol. 4, 467-474. (47) Yamasaki, H., Pulkrabek, P., Grunberger, D., and Weinstein, I. B. (1977) Differential excision from DNA of the C-8 and N2 guanosine adducts of N-acetyl-2-aminofluorene by single strandspecific endonucleases. Cancer Res. 37,3756-3760. (48) Moore, P. D., Rabkin, S. D., Osbom, A. L., King, C. M., and Straws, B. S. (1982) Effect of acetylated and deacetylated 2-aminofluorene adducts on in vitro DNA svnthesis. Proc. Natl. Acad. Sci. U S A . 79,7166-7170. (49) Bichara. M.. and Fuchs. R. P. P. (19851DNA bindinn and mutation spectra of the carcinogen N-2-aminofluorene in Escherichia coli. A correlation between conformation of the premutagenic lesion and the mutation specificity.J. Mol. Biol. 183, 341-351. (50) Pierce, J. R., Case, R., and Tang, M.-S. (1989) Recognition and repair of 2-aminofluorene- and 24acetylamino)fluorene-DNA adducts by UVRABC nuclease. Biochemistry 28,5821-5826. (51) Beland, F. A., and Kadlubar, F. F. (1985) Formation and persistence of arylamine DNA adducts in uiuo. Enuiron. Health Perspect. 62, 19-30. (52) Poirier, M. C., and Beland, F. A. (1992) DNA adduct measurements and tumor incidence during chronic carcinogen exposure in animal models: Implications for DNA adduct-based human cancer risk assessment. Chem. Res. Toxicol. 5, 749-755.