1038
Chem. Res. Toxicol. 2005, 18, 1038-1047
Nitric Oxide-Mediated Nitrosation of 2-Amino-3,8-dimethylimidazo[4,5-f]quinoxaline Potentiated by Hemin and Myeloperoxidase Vijaya M. Lakshmi,† Fong Fu Hsu,‡ and Terry V. Zenser*,†,§ Division of Geriatric Medicine, VA Medical Center, and Department of Biochemistry and Molecular Biology, St. Louis University School of Medicine, St. Louis, Missouri 63125, and Department of Medicine, Washington University, St. Louis, Missouri 63110 Received January 18, 2005
N-Nitrosamines formed by nitrosation of heterocyclic amines might initiate colon cancer in individuals consuming well-done red meat diets and with inflammatory conditions in their colon. This study investigates nitric oxide (NO)-mediated nitrosation of 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx) and the influence of dietary (hemin) and inflammatory [NO, myeloperoxidase (MPO), and H2O2] components on nitrosation. Using the NO donor spermine NONOate (1.2 µM NO/min) at pH 7.4 with 0.005 mM MeIQx, a product due to NO autoxidation was at the limit of detection (1% of total radioactivity recovered by HPLC). Product formation increased 13- or 16-fold in the presence of 10 µM hemin or 85 nM MPO, respectively, with an in situ system for generating H2O2 (glucose oxidase/glucose). The nitrosation product and its chloro derivative were analyzed by electrospray ionization mass spectrometry, and the product was determined to be 2-nitrosoamino-3,8-dimethylimidazo[4,5-f]quinoxaline (N-NO-MeIQx). Nitrosation by NO autoxidation was only detected at g1.2 µM NO/min and was not affected by H2O2. Investigations with hemin determined minimum effective components necessary for potentiation: 1 µM hemin, 1 µM H2O2/min, and 0.012 µM NO/min. The reactive nitrogen oxygen species (RNOS) produced by hemin and MPO had a 4- and 3-fold, respectively, greater affinity for MeIQx than those produced by NO autoxidation. Test agents were used to characterize nitrosation. Results with catalase, SOD, azide, and NADH are consistent with multiple RNOS, the lack of peroxynitrite involvement in nitrosation, and peroxidatic potentiation by oxidative nitrosylation rather than nitrosation. Using phorbol ester stimulated human neutrophils, the formation of N-NO-MeIQx and its modification by test agents was consistent with MPO and not peroxynitrite. Thus, nitrosation of MeIQx and its potentiation by hemin and MPO provide a mechanism by which well-done red meat consumption and inflammation can generate N-nitroso compounds and initiate colon cancer under inflammatory conditions, such as colitis.
Introduction High temperature cooking of proteinaceous foods can result in the formation of a group of structurally related heterocyclic amines that are potent genotoxic carcinogens (1). These amines are formed by condensation of creatine or creatinine with amino acids and sugars or their thermal decomposition products. Human exposure to heterocyclic amines is dependent upon food preparation methods, portion size, and consumption frequency, resulting in potential doses of micrograms per day. There is a strong association between well-done red meat intake and colorectal cancer risk (2). The presence of heterocyclic amines in well-done red meat (1, 3) and the initiation of colon cancer by heterocyclic amines (4) suggest that heterocyclic amines present in well-done red meat may be responsible for some of the increased risk of colon cancer associated with this dietary component. * To whom correspondence should be addressed. Tel: 314-894-6510. Fax: 314-894-6614. E-mail:
[email protected]. † VA Medical Center. ‡ Washington University. § St. Louis University School of Medicine.
2-Amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx)1 is a heterocyclic amine produced by high temperature cooking of red meat (1). It is a strong mutagen in the Ames assay (5) and at high doses can induce tumors at multiple sites in rodents (6). In vivo mutagenicity of low dose MeIQx in transgenic rats has been demonstrated in the liver, Zymal gland, and colon but not in eight other tissues examined (7). This is consistent with macroscopically, aberrant crypt foci in the colon of MeIQx-treated rats (8). Cytochrome P450s oxidize heterocyclic amines to their N-hydroxy derivatives. Esterification of the latter with acetate or sulfate is believed to produce unstable intermediates that decomposes to a nitrenium ion, which binds DNA forming adducts such as N-(deoxyguanosin8-yl)-MeIQx that are thought to initiate carcinogenesis 1 Abbreviations: MeIQx, 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline; IQ, 2-amino-3-methylimidazo[4,5-f]quinoline; DETAPAC, diethylenetriaminepentaacetic acid; DMPO, 5,5-dimethyl-1-pyrroline Noxide; iNOS, inducible nitric oxide synthase; MPO, myeloperoxidase; NO, nitric oxide; N-NO-MeIQx, 2-nitrosoamino-3,8-dimethylimidazo[4,5-f]quinoxaline; N-NO-IQ, 2-nitrosoamino-3-methylimidazo[4,5-f]quinoline; PMA, β-phorbol 12-myristate 13-acetate; PMNs, polymorphonuclear neutrophils; RNOS, reactive nitrogen oxygen species; SpN, spermine NONOate; SOD, superoxide dismutase.
10.1021/tx0500070 CCC: $30.25 © 2005 American Chemical Society Published on Web 05/07/2005
Nitrosation of MeIQx
(9, 10). Some clinic-based control studies of colorectal adenomas patients have focused on the relationship between these tumors and heterocyclic amines and their activation. When the intake of three major heterocyclic amines (MeIQx, 2-amino-3,4,8-trimethylimidazo[4,5f]quinoxaline, and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine) commonly found in well-done meat was examined, only MeIQx was shown to be a risk factor for colorectal adenomas (11). A 6-fold increase in adenoma risk was observed among rapid N-acetyltransferase 1 acetylators who consumed more than 27 ng of MeIQx per day (12). In contrast, N-acetyltransferase 2 genotype and cytochrome P450 1A2 and N-acetyltransferase 2 activity were not associated with adenoma risk whether or not MeIQx intake was analyzed as a variable. While the latter is in general consistent with a study by Chen et al. (13), other studies have shown a relationship between combined rapid cytochrome P450 1A2 and rapid Nacetyltransferase 2 phenotypes and an increased risk of colorectal cancer (14). Oxidation of heterocyclic amines will likely be reduced by the significant decrease in cytochrome P450 levels that occur during infection/ inflammation (15). Thus, despite considerable research, the biotransformation pathways involved in heterocyclic amine initiation of colon cancer need further study. Chronic inflammation/infection and injury play an important role in colon cancer (16). The incidence of colorectal carcinoma in patients with inflammatory bowel disease, a chronic inflammatory disease, is 20-fold higher and has an average age of onset 20 years younger than the general population (17). High levels of nitric oxide (NO) are observed during inflammation due to upregulation of inducible nitric oxide synthase (iNOS) (18). By reacting with superoxide to form peroxynitrite anions or with oxygen (autoxidation), NO produces reactive nitrogen oxygen species (RNOS) such as nitrogen dioxide radicals (NO2•) and dinitrogen trioxide (N2O3). NO inhibits DNA repair (19, 20) and can nitrosatively deaminate DNA (21). Inflammatory bowel disease is associated with marked mucosal infiltration of macrophages, lymphocytes, and neutrophils in which high levels of iNOS and 3-nitro-tyrosine, a marker of RNOS, are detected (22-24). Noncancerous colon tissue from ulcerative colitis patients and primary colon tumors exhibit a statistically significant positive correlation between their iNOS activity and the frequency of G:C to A:T transitions at CpG sites in the p53 tumor suppressor gene (22, 25). iNOS -/- knockout mice reduce the adenomas observed in a rodent colitis model for colon cancer (26). Thus, under certain conditions, NO and its RNOS might promote colon cancer. We have demonstrated that during NO autoxidation 2-amino-3-methylimidazo[4,5-f]quinoline (IQ) undergoes nitrosation to form 2-nitrosoamino-3-methylimidazo[4,5f]quinoline (N-NO-IQ) (27). However, NO autoxidation is slow and considered unlikely to account for significant amounts of S-, N-, and heme-nitros(yl)ation detected in vivo (28, 29). In addition to providing increased levels of NO, conditions that mediate the inflammatory response also have the potential to promote nitrosation of heterocyclic amines. During an inflammatory response, infiltrating neutrophils produce H2O2 (30), which can be used by myeloperoxidase (MPO) to oxidize NO to RNOS (31) with the potential to nitrosate heterocyclic amines. Furthermore, the component of red meat that gives it its color, heme, has pseudoperoxidase activity (32) and
Chem. Res. Toxicol., Vol. 18, No. 6, 2005 1039
could also oxidize NO to nitrosating RNOS. The consumption of 250 g of beef per day for 3 days results in a recovery of 1.07 mg of hemoglobin/g feces (70 µM heme) on day 3 (33). Heme can induce the formation of peroxyl radicals in fats (34), and in diets containing increased concentrations of heme, lipid peroxidation in fecal water correlated with dietary heme content (35). Nitrosamines/ amides are known to cause tumors in a variety of organs, including gastrointestinal tumors (36). N-NO-IQ is a potent mutagen,2 which binds DNA (27, 37) and forms N-(deoxyguanosin-8-yl)-IQ.2 In addition, inconsistent results are observed in epidemiology studies evaluating cytochrome P450/N-acetyltransferase activation pathways for heterocyclic amines (12-14), and cytochrome P450s are suppressed during inflammation (15). Therefore, we propose N-nitroso heterocyclic amines as an alternative to N-hydroxy derivatives initiating colon cancer in individuals with colitis. These heterocyclic nitrosamines would be formed by components of the inflammatory response (NO, MPO, and H2O2) and diet (heterocyclic amines and heme). This study evaluates N-nitrosation of MeIQx by NO in the presence and absence of hemin or MPO. Hemin is the ferric porphyrin component of hemoglobin and often used in place of heme (35), as is the case in the present study. While some epidemiology studies evaluating the risk of well-done meat in colorectal adenomas have excluded individuals with Crohn’s disease and ulcerative colitis (38), our results suggest that such an evaluation is warranted. Nitrosation of heterocyclic amines may be a more general phenomenon than previously reported (27).
Experimental Procedures Caution: 2-Amino-3-methylimidazo[4,5-f]quinoline is carcinogenic to rodents and should be handled in accordance with NIH Guidelines for the Laboratory Use of Chemical Carcinogens (39). Materials. [2-14C]-MeIQx (50 mCi/mmol, >98% radiochemical purity) was purchased from Toronto Research Chemicals (Toronto, ON). H2O2, potassium ferrocyanide, catalase (bovine liver), ascorbic acid, diethylenetriaminepentaacetic acid (DETAPAC), NaN3, β-phorbol 12-myristate 13-acetate (PMA), NADH, 5,5-dimethyl-1-pyrroline N-oxide (DMPO), Dglucose, glucose oxidase (Aspergillus niger; type X-S), and superoxide dismutase (SOD) (bovine erythrocytes) were purchased from Sigma-Aldrich Corporation (St. Louis, MO). Spermine NONOate (SpN), a NO donor, and MPO from human polymorphonuclear leukocytes [180-200 units/mg protein, g95% pure by SDS-PAGE, an RZ of 0.85 (A430/A280)] were purchased from Calbiochem (San Diego, CA). SpN decomposes to release free NO into solution with a half-life of 42 min at pH 7.4 and 37 °C (40). Stock solutions of SpN (10 mM) were prepared in 10 mM NaOH and were stored at -70 °C. Concentrations of SpN were determined from absorbance values at 250 nm (250 ) 8000 M-1; ref 40) directly prior to use. The rate of SpN decomposition was unaltered in reaction mixtures demonstrating potentiation of 2-nitrosoamino-3,8-dimethylimidazo[4,5-f]quinoxaline (N-NOMeIQx) formation by hemin or MPO (Figure 1, middle and lower panels). Rather than adding a bolus of H2O2, it was generated in situ, using d-glucose and glucose oxidase. H2O2 was measured by MPO and 3,5,3′,5′-tetramethylbenzidine (652 ) 3.9 × 104 M-1; ref 41). Ultima-Flo AP was purchased from Perkin-Elmer LAS (Shelton, CT). Transformation of MeIQx by RNOS. To assess NOmediated nitrosation, 14C-MeIQx (0.005 mM) was incubated in 2 Zenser, T. V., Lakshmi, V. M., Zhou, H.-j., Josephy, P. D., and Schut, H. A. J. Abstract published as follows: (2004) Proc. Am. Assoc. Cancer Res. 45, 4863.
1040
Chem. Res. Toxicol., Vol. 18, No. 6, 2005
Figure 1. MeIQx nitrosation. Illustrated is the transformation of 14C-MeIQx (0.005 mM) following a 60 min incubation with NO (1.2 µM NO/min, 0.05 mM SpN) (upper panel); NO, hemin (10 µM), and H2O2 (10 µM/min) (middle panel); NO, MPO (85 nM), and H2O2 (10 µM/min) (lower panel). 14C-N-NO-MeIQx formation was analyzed by HPLC. 100 mM sodium phosphate buffer, pH 7.4, containing 0-0.05 mM SpN, 0.1 mM DETAPAC, and 0-30 µM hemin or 85 nM MPO in a total volume of 0.1 mL for 60 min at 37 °C. H2O2 was generated in situ with 1 mM glucose and 0-15 mU/ml glucose oxidase. Reactions were started by the addition of SpN, and tubes were capped. Blank values were obtained in the absence of SpN. The reactions were stopped by adding 10 mM ascorbic acid in dimethylformamide (0.025 mL). Samples were immediately frozen and kept at -70 °C for analysis by HPLC (27). HPLC Analysis of Products. Products were assessed using a Beckman HPLC with System Gold software and a 5 µm, 4.6 mm × 150 mm C-18 ultrasphere column attached to a guard column. The mobile phase contained 20 mM ammonium formate, pH 3.1, in 8% acetonitrile, 0-2 min; 8-16% acetonitrile, 2-10 min; 16-21% acetonitrile, 13-18 min; 21-35% acetonitrile, 18-23 min; 35-8% acetonitrile, 30-35 min; flow rate 1 mL/min. Radioactivity in HPLC eluents was measured using a FLO-ONE (Perkin-Elmer) radioactive flow detector. Data are expressed as a % of total radioactivity or pmol recovered by HPLC. Preparation of Human Polymorphonuclear Neutrophils (PMNs). Human blood from healthy donors was mixed with EDTA (0.2% final concentration) and immediately layered
Lakshmi et al. over an equal volume of neutrophils isolation medium from Robins Scientific Corp. (Sunnyvale, CA). Neutrophils were isolated by centrifugation using the manufacturer’s specifications. Red blood cell contamination was eliminated by hypotonic lysis at 4 °C. Cells were resuspended at 2 × 106 cells/mL in Hank’s balanced salt solution with calcium, magnesium, and bicarbonate. All protocols were in accordance with institutional guidelines on research involving human subjects at the St. Louis VA Medical Center. All participants gave written informed consent. Metabolism of MeIQx by Neutrophils. Neutrophils (0.3 × 106 cells in 0.3 mL) were incubated with 0.005 mM 14C-MeIQx in 12 mm × 75 mm polypropylene tubes at 37 °C for 36 min in Hank’s balanced salt solution containing calcium, magnesium, and bicarbonate without phenol red in capped tubes. Where indicated, 0.054 mM PMA, 1.2 µM NO/min (0.05 mM SpN), 1 mM NaN3, and 0.1 mM NADH were added at 5, 6, 8, and 8 min, respectively, while 66 µg/mL catalase and 66 µg/mL SOD were present at zero time. Blank values were obtained in the absence of SpN. The reaction was stopped by placing on ice and freezing at -70 °C. Samples were sonicated for 15 s three times, and 0.3 mL of dimethylformamide containing 2 mM ascorbic acid was added. Samples were spun at 1500g for 10 min. The supernatant was evaporated, and the residue was dissolved in 0.1 mL of methanol:dimethylformamide (8:2) and analyzed by HPLC. To evaluate the metabolism of N-NO-MeIQx by neutrophils, cells were incubated under the same conditions as described above with 0.01 mM 14C-N-NO-MeIQx. Incubations were stopped by placing on ice and freezing at -70 °C. Blanks had N-NOMeIQx added at the end of the incubation. Supernatants from sonicated samples were analyzed for total recovery of radioactivity and by HPLC. The oxidant burst response was measured for each PMN preparation (42). Superoxide generation was activated by addition of PMA. Superoxide specific reduction of cytochrome c was determined spectrophotometrically (550 ) 21.1 mM-1 cm-1) and was inhibited by SOD (10 µg/mL). Values observed with cells in the absence of PMA were considered as blanks. To assess the authenticity of neutrophil-derived N-NOMeIQx, selected supernatants were pooled and the azido derivative was prepared as previously described for N-NO-IQ (37). Pooled samples were evaporated to dryness and dissolved in pH 2.0 ammonium formate buffer containing 10 mM NaN3. After 60 min at 37 °C, the reaction was stopped by adjusting the pH to 7.0-7.4. Authentic N-NO-MeIQx (98% pure) was treated in a manner similar to prepare the standard. Mass Spectral Analysis. Electrospray ionization (ESI) MS analyses were performed on a Finnigan TSQ-7000 triple stage quadrupole mass spectrometer (San Jose, CA) equipped with a Finnigan ESI source and controlled by Finnigan ICIS software operated on a DEC alpha workstation. Samples were loop injected onto the ESI source with a Harvard syringe pump, which is continuously infused with methanol at a flow rate of 5 µL/min. The skimmer was at ground potential, and the electrospray needle was at 4.5 kV. The heated capillary temperature was 250 °C. To obtain collisionally activated dissociation (CAD) tandem mass spectra, the collision energy was set at 22 eV, and argon (2.3 mTorr) was used as the target gas. The product ion spectra were acquired in the profile mode at the scan rate of one scan per 3 s. Statistical Analysis. Data are expressed as a means ( SE, and significant differences were evaluated using a Student’s paired t test with p < 0.05.
Results NO Nitrosation of MeIQx and Product Identification. The reaction of NO (0.05 mM SpN, 1.2 µM NO/min) with 0.005 mM MeIQx for 60 min at 37 °C is depicted in Figure 1, top panel. A small amount of product (N-NOMeIQx) representing 1% of the total radioactivity recov-
Nitrosation of MeIQx
Figure 2. ESI CAD tandem mass spectrometric analysis of N-NO-MeIQx. Illustrated is the positive ion mode with [M + Na]+ ions at m/z 265 (upper panel) and the negative ion mode with [M - H]- ions at m/z 241 (lower panel).
ered by HPLC eluted at 14.3 min. In the presence of 10 µM hemin + 10 µM H2O2/min (Figure 1, middle panel), product formation increased to 13%. Product formation was linear for at least 60 min. With 85 nM MPO + 10 µM H2O2/min (Figure 1, bottom panel), a similar increase in product formation (16%) was observed. Glucose oxidase/glucose was used to generate H2O2 in situ. H2O2 did not influence product formation by NO alone (Figure 1, top panel). However, the potentiation of nitrosation observed with hemin or MPO required H2O2. In the absence of NO, nitrosamine formation was not detected with either hemin or MPO. Thus, nitrosation of MeIQx was dependent upon NO and could be potentiated by hemin or MPO in the presence of H2O2. The 14.3 min peak was purified by HPLC and analyzed by ESI/MS in both positive and negative ion modes. In the positive ion mode, the product exhibited an [M + Na]+ ion at m/z 265 and gave ions at m/z 235 and 220, arising from losses of NO and CH3, respectively (Figure 2, upper panel). Fragmentation of the quinoxaline ring yielded m/z 197, which produced ions at m/z 170 and 143 due to consecutive losses of HCN. In negative ion mode, the product yielded deprotonated [M - H]- molecular ions at m/z 241 (Figure 2, lower panel). The latter gave rise to prominent ions at m/z 213 and 198, representing the loss of N2 and fragmentation of the quinoxaline ring, respectively. These results are consistent with the product being a nitroso derivative of MeIQx. To provide additional support for our structural assignment of the MeIQx nitrosation product, a chloro derivative was prepared by incubating N-NO-MeIQx in acetonitrile with 0.1 N HCl for 30 min at 37 °C and was analyzed by ESI/MS. The new product gave the [M + H]+
Chem. Res. Toxicol., Vol. 18, No. 6, 2005 1041
Figure 3. Positive ion ESI mass spectrometric analysis of a chloro derivative of N-NO-MeIQx. The upper panel illustrates a mass spectral scan from m/z 225 to 240. The middle and lower panels illustrate the CAD product ion spectra of m/z 233 and 235, respectively.
ions at m/z 233 and 235 with a ratio of approximately 3:1, indicating the presence of one Cl (Figure 3, top panel). The product ion spectrum of m/z 233 (Figure 3, middle panel) contains ions at m/z 218 and 191, probably arising from consecutive losses of a CH3 and HCN, respectively. The ion at m/z 206 arises from loss of HCN and gives rise to m/z 191 via subsequent loss of CH3. Ions at m/z 198 and 197 arise from elimination of 35Cl and H35Cl, respectively. These fragmentation pathways are further confirmed by the product ion spectrum of m/z 235 (Figure 3, lower panel), which contains the similar ions and shows the expected mass shift of 2 Da for the fragment ions possessing a 37Cl. The results are consistent with the structure being 2-chloro-3,8-dimethylimidazo[4,5-f]quinoxaline. The results also suggest that the nitroso substitution occurred at the 2-amino group rather than N-1 on the imidazole ring and the NO nitrosation product of MeIQx is N-NO-MeIQx. Conditions for Hemin Potentiation of MeIQx Nitrosation. The hemin concentration-dependent formation of N-NO-MeIQx was evaluated with 1.2 µM NO/ min and 10 µM H2O2/min (Figure 4). A significant increase in N-NO-MeIQx formation was observed at 1 µM hemin (p < 0.05). A linear increase in product formation was observed from 0 to 10 µM hemin with a maximum increase at 20 µM. Using 10 µM hemin and 10 µM H2O2/min, NO-dependent formation of N-NO-MeIQx was determined (Figure 5). Significant (p < 0.05) N-NO-MeIQx formation was detected at 0.012 µM NO/min with linear increases observed from 0 to 0.12 µM NO/min. Maximum increases
1042
Chem. Res. Toxicol., Vol. 18, No. 6, 2005
Figure 4. Effect of hemin concentration on MeIQx nitrosation. 14C-MeIQx (0.005 mM), NO (1.2 µM/min), and H O (10 µM/ 2 2 min) were incubated with increasing concentrations of hemin for 60 min. In all subsequent figures, 14C-N-NO-MeIQx was determined by HPLC and results represent means ( SEM for triplicate determinations.
Figure 5. Effect of varying fluxes of NO on hemin potentiation of MeIQx nitrosation. 14C-MeIQx (0.005 mM), hemin (10 µM), and H2O2 (10 µM/min) were incubated with increasing fluxes of NO for 60 min.
occurred at 0.48 µM NO/min. At this concentration of NO, N-NO-MeIQx formation in the absence of hemin is at the limit of detection (