Chem. Res. Toxicol. 1989,2, 247-253
247
The Fenton Degradation as a Nonenzymatic Model for Microsomal Denitrosation of N-Nitrosodimethylamine Young-Hun
Heur,? A n t h o n y J. Streeter, R a y m o n d W. Nims, and L a r r y K. Keefer*
Chemistry Section, Laboratory of Comparative Carcinogenesis, National Cancer I n s t i t u t e , Frederick Cancer Research Facility, Frederick, Maryland 21 701 Received M a r c h 3, 1989
The microsomal metabolism of the carcinogen N-nitrosodimethylamine (NDMA) was suggested to be initiated by hydrogen atom abstraction to form an a-nitrosamino radical, which either oxidizes further to an a-hydroxy nitrosamine as the initial product of the activating dealkylation pathway or fragments to the nitric oxide radical and N-methylformaldimine as the first step of the presumably inactivating denitrosation route. To examine the chemistry of the a-nitrosamino radical in a nonenzymatic setting, we exposed NDMA to the Fenton reagent, which is known to be capable of abstracting hydrogen atoms from organic species. The products observed were those expected of a denitrosation model. Solutions containing 13 mM [14C]NDMA,15 mM FeS04, 15 mM H202,and 7.5 mM H2S04were kept a t 4-10 "C for 1 h and then basified t o yield methylamine (3.2 f 0.5 mM, mean f SD, n = 8),formaldehyde (3.1 f 0.9 mM), and unreacted nitrosamine (10.2 f 0.7 mM) as the only radioactive species detected, with total nitrate/nitrite also being found a t a level of 2.8 f 0.5 mM. N-Methylformaldiminium ion was identified as an intermediate. The parallels between these results and those seen in the microsomal reaction support the hypothesis that the a-nitrosamino radical is a common intermediate in enzymatic denitrosation versus dealkylation of NDMA. The shift from -85% dealkylation in the microsomal metabolism to 100% denitrosation in the Fenton reaction demonstrates that the two pathways are separable and suggests that it might be possible t o reduce the risk of exposure to NDMA if a means can be found to protect metabolically produced a-nitrosamino radicals from further oxidation and other activating reactions in vivo, forcing them to decompose unimolecularly along the denitrosation pathway.
Introduction NDMA' is a potent carcinogen that has been found in the urine of healthy people (1) and is thus a suspected causative agent in human cancer. Several decades of research have provided a vast volume of information concerning the dealkylation pathway of NDMA metabolism which leads to the formation of a putative ultimate carcinogen (2,3). Recently, a competing and apparently inactivating pathway, the so-called denitrosation route, has been characterized for this substrate (4-25). Deuterium isotope effect studies suggested that the same enzyme was involved in both dealkylation and denitrosation of NDMA by acetone-induced rat liver microsomes, and an a-nitrosamino radical was proposed as a common intermediate in the two metabolic routes (21). If this radical is produced, it should have one of the two fates shown in Figure 1. Further oxidation (e.g., by recombination with the elements of hydroxyl radical) would lead to an a-nitrosamino alcohol which by spontaneous decomposition would methylate ambient nucleophiles with concurrent formation of dinitrogen gas and formaldehyde. Alternatively, unimolecular fragmentation with loss of nitric oxide radical would produce N-methylformaldimine. This product should hydrolyze to formaldehyde and methylamine, both of which are observed in the in vitro metabolism of NDMA (20), and nitric oxide would be oxidized to the other observed product of enzymatic denitrosation, nitrite (17 ) . 'Present address: Department of Pathology, Medical College of Ohio, Toledo, OH 43699.
Enzymatically produced hydroxyl radicals have been implicated in the denitrosation pathway (12,14,22). We have tested the hypothesis that reaction of NDMA with the hydroxyl radicals produced in the Fenton reaction (26) might provide a convenient nonenzymatic route to the a-nitrosamino radical, permitting us to study its behavior in the absence of further oxidation by cytochrome P-450 and other scavenging reactions. The results, reported below, suggest that the Fenton degradation of NDMA parallels cytochrome P-450 induced denitrosation of this substrate in several important respects.
Experimental Section Chemicals. [14C]NDMA (40.3 and 49.6 mCi/mmol) was purchased from SRI International, Menlo Park, CA, and NEN Research Products, Boston, MA (respectively). [14C]Toluene(4.6 X lo5 dpm/mL), [14C]methanol (2.4 mCi/mmol), [14C]formaldehyde (10 and 15 mCi/mmol), sodium [14C]formate (53 mCi/mmol), [14C]methylamine (49.0and 7.2 mCi/mmol), and [I4C]dimethylamine (5.3mCi/mmol) were obtained from NEN Research Products. [13C]NDMA (13C incorporation > 99% by mass spectrometry) was kindly provided by Dr. Joseph A. Hrabie. Iron(I1) sulfate heptahydrate (ACS reagent, 99+%), 1,3,5-trimethylhexahydro-l,3,5-triazine, hydrogen peroxide (ACS reagent, 30%), 5,5-dimethyl-l,3-cyclohexanedione (95%), D 2 0 (99.8atom % D), NaOD (99atom % D, 40 wt % in D20),and D2S04(99.5 atom % D, 98 wt % in D20) were purchased from Aldrich Chemical Co., Milwaukee, WI. Sulfuric acid (96%), formaldehyde (nominally 37%; actual analysis 39.6%), and nitric acid (Ultrex) were purchased from J. T. Baker Chemical Co., Phillipsburg, NJ. Abbreviation: NDMA, N-nitrosodimethylamine.
This article not subject t o U.S. Copyright. Published 1989 by t h e American Chemical Society
248
Chem. Res. Toxicol., Vol. 2, No. 4 , 1989 CH3, Cytochrome P-450
CH,
,
,N-N=O NDMA
% ,
*OH
CH3
"Dealkylation" Pathwav
,N-N=O
Proposed Radical lnlermediale
H e w et al.
*
,N-N=O CH3
X-
HOCH,,
CH,X
+
N,
+ O=CH,
,
'NO
Oxidation
-NO;
Figure 1. Proposed comprehensive mechanism of NDMA metabolism (see ref 21 and citations therein). X- is an available nucleophile. Hydrogen peroxide (90%)was obtained from FMC Co., Buffalo, was NY. Sodium 3-(trimethylsily1)propanesulfonate-2,2,3,3-d4 purchased from MSD Isotopes, Montreal, Canada. Aminex A-7 (5-11 pm) and Aminex A-27 (13-17 pm) were purchased from Bio-Rad Laboratories, Richmond, CA. Reverse-phase C18packing (10-Mm) material was obtained from Supelco Chromatography Supplies, Bellefonte, PA. The Spherisorb S5 ODS2 (el8) column (4.6 mm X 25 cm) was purchased from Anspec Co., Ann Arbor, MI. HPLC. For the simultaneous analysis of the products of the Fenton degradation of NDMA, an HPLC system based on the method described earlier by Sohn et al. (27) was chosen as the starting point. By coupling the previously reported system with a cation-exchange column, conditions were found in which base-line separations of NDMA, formaldehyde, methylamine, formate, and methanol were routinely achieved. Fenton reaction aliquots of 10 pL each were injected into an HPLC assembly consisting of two Waters 510 pumps and an automated gradient controller (Waters Associates, Milford, MA). HPLC columns were connected serially in the order of 5 cm C18, 25 cm CIS,5 cm Aminex A-7, and 10 cm Aminex A-27, with 200 mM ammonium phosphate buffer (pH 3.0, 0.1% methanol) a t a flow rate of 1 mL/min as mobile phase. For detection a Model IC Flow-One\Beta radioactive flow detector (Radiomatic Instruments, Inc., Tampa, FL) was used with a scintillation cocktail flow rate of 2 mL/min. The scintillation cocktail used was Flo-Scint 11,purchased from the same company. Nitrate analyses were carried out on a Dionex AS-5 column (4.6 mm X 250 mm; ammonium form) using 0.2 mM carbonate/bicarbonate buffer (pH 10) as mobile phase (1.5 mL/min) with detection by conductivity. Nitrite was determined by the Griess reaction. Calibration. Calibration curves for formaldehyde, NDMA, and methylamine were obtained by injecting known amounts of radioactivity for each and plotting the results versus the integrated cpm areas from the chromatographic traces. Four different concentrations of each compound (radiochemicallypure by HPLC) were injected on the same days the Fenton reaction mixtures were analyzed to obtain the calibration data. Scintillation Counting. An LS 8OOO Series liquid scintillation system (Beckman Instruments, Inc., Fullerton, CA) was used for scintillation counting, with Aquassure liquid (NEN Research Products, Boston, MA) as cocktail. Fenton Reaction. The quantitative course of the Fenton reaction was determined by running eight replicate transformations performed as follows. Ferrous sulfate was dissolved in an argon-purged solution of sulfuric acid in deionized, distilled water. Final concentrations were 60 mM in ferrous sulfate and 30 mM in sulfuric acid. Protection from atmospheric oxygen was essential. A 200-pL aliquot of the ferrous sulfate solution was placed in a 5-mL glass reaction vial kept in an ice/water bath (4-10 "C) and combined with an equal volume of ["CINDMA (52 mM, 1.17 pCc final specific activity = 112.5 rCi/mmol; radiochemically pure by HPLC). While the resulting solution was stirred with a magnetic stirrer, 400 pL of hydrogen peroxide (30 mM), which was prepared by diluting 30% H202with argon-purged HzO, was
added dropwise. The vial was then closed, and the reaction mixture was kept for 1 h a t 4-10 O C under constant stirring. Finally, the reaction was quenched by adding 200 pL of an aqueous solution that was 60 mM in methanol and 180 mM in sodium hydroxide, and the iron hydroxide precipitate was removed by centrifugation. The resulting supernatant was subjected to analysis by HPLC. The measured pH of these reaction mixtures was 0.7. Slight modifications of this procedure were used in the qualitative confirmation of product identities as described below. Isotope Dilution Experiment. A 100-pL aliquot of a [14C]NDMAFenton reaction mixture (0.107 pCi) was mixed with 132 mg of cold 39.6% formaldehyde solution. T o the mixture was added a solution of 596 mg of 5,5-dimethyl-1,3-cyclohexanedione in 50 mL of 0.1 M sodium hydroxide. After stirring a t room temperature for 30 min, 1mL of glacial acetic acid was added to induce precipitation. The precipitate was filtered and recrystallized from 50% methanol to obtain constant-melting formaldehyde-dimedone adduct of mp 188.5-190 "C [lit. mp 187-188 "C (28)]. The crystals were dried over phosphorus pentoxide, and 159.2 mg was subjected to scintillation counting. Catalytic Hydrogenation of the Intermediate. NDMA (50 r L of a 48 mM solution) was mixed with 0.5 mL of 240 mM ferrous sulfate in sulfuric acid, and 0.5 mL of 240 mM HzOzwas added dropwise, keeping the reaction temperature a t 4-10 "C. These conditions led to consumption of all the NDMA, as confirmed by HPLC. To the cold reaction mixture, -0.03 g of 10% palladium on carbon was added. A rubber balloon was attached to one neck of the reaction flask, and a rubber septum was attached to the other neck. The system was purged several times with hydrogen, and the balloon was filled through the septum by using a needle. The reaction mixture was kept in an ice/water bath for 1 h and then a t room temperature for 1 h. Finally, 1mL of 180 mM sodium hydroxide was added to precipitate iron hydroxide, and the precipitate was removed by centrifugation. The resulting supernatant was analyzed by HPLC using a 400 IC cation column (4.6 X 5 cm, Vydac-the Separation Group, Hesperia, CA) with 2.5 mM nitric acid as mobile phase at a flow rate of 4 mL/min and detection with the radioactive flow detector described above. Generation of N-Methylformaldiminium Ion Standard. N-Methylformaldiminium ion standard was generated by dissolving 1,3,5-trimethylhexahydro-1,3,5-triazine (0.5 mL) in 1.5 mL of D2S04as previously described (291,except that the solution was not heated. NMR Studies. Proton and 13C spectra were taken with a Varian XL-200 instrument a t 5 "C. Spectra were locked to either the external standard, sodium 3-(trimethylsily1)propanesulfonate-2,2,3,3-d,, at 0 ppm or the solvent peak a t 6.8 ppm. The 'H NMR spectrum of N-methylformaldiminium ion standard was determined on a D20 solution that was 510 mM in the iminium ion and 1.3 M in D2SO4. The 'H NMR spectrum of the Fenton reaction mixture was obtained as follows. To 1mL of 48 mM NDMA in DzO, 0.5 mL of 120 mM ferrous sulfate in DzO containing 123 mM DzS04was added, whereupon 0.5 mL of 120 mM hydrogen peroxide (prepared
Denitrosation of N D M A by the Fenton Reagent
Chem. Res. Toxicol., Vol. 2, No. 4, 1989 249
Trace 01 of CH,O
NDMA
Unreacted
0
5
10
15
20
25
30
Time (min)
Figure 2. Analysis by HPLC of (a) the Fenton degradation of 0.85 pCi of [14C]NDMArun in 7.5 mM sulfuric acid and (b) reaction mixture a after basification. by mixing aqueous 90% Hz02with D20)was added under constant stirring at 4-10 "C. The reaction mixture was diluted with 50 pL of D2S04,and the 'H NMR spectrum was run. For the co-NMR study, this solution was mixed with 100 pL of the Nmethylformaldiminium ion standard solution (510 mM). Finally, 40 pL of deuteriated sodium hydroxide (40% solution in D20, 99% deuterium) was added to determine the lH NMR spectrum of the basified products. The 13C NMR spectrum of the N-methylformaldiminiumion standard was obtained on a 5.1 M solution in 98% D2S04. The Fenton reaction was studied by using 13C NMR as follows. To 1 mL of 48 mM [13C]NDMA in DzO,0.5 mL of 120 mM ferrous sulfate in D20containing 123 mM D2S04was added. A 120 mM hydrogen peroxide solution was prepared by mixing 90% H202 with D20,and 0.5 mL was added under constant stirring at 4-10 O C . After its spectrum was run, this solution was mixed with 50 pL of the 5.1 M N-methylformaldiminiumion standard solution and the 13C NMR spectrum was redetermined. MS Study. To 50 pL of a 48 mM NDMA solution was added 0.5 mL of ferrous sulfate (240 mM) in sulfuric acid, followed by 0.5 mL of hydrogen peroxide (240 mM). This condition was shown to consume all NDMA within the limit of detection. The resulting reaction mixture was analyzed by mass spectrometry. A positive ion fast atom bombardment (FAB) mass spectrum was obtained by using a VG-ZAB-2F double-focusing mass spectrometer (VG Analytical Ltd., Manchester, U.K.) employing dithioerythritol as a matrix. The FAB gun was operated at 8 keV and 1mA with Xe as the reactant gas.
Results Fate of NDMA on Exposure to Fenton Reagent. Reactions were carried out under strongly acidic conditions to avoid complications caused by precipitation of iron hydroxide. Solutions 13 mM in [I4C]NDMA, 15 mM in ferrous sulfate, and 7.5 mM in sulfuric acid were placed in an icelwater bath and treated dropwise with enough hydrogen peroxide to have made an ultimate theoretical concentration of 15 mM. Analysis 1 h later showed that substantial conversion to an intermediate that was not identical with any of the 14C-containingstandards available to us had occurred, as shown in Figure 2a. After the reaction mixture was basified to precipitate iron hydroxide, however, peaks coeluting with formaldehyde (3.1 f 0.9 mM) and methylamine (3.2 f 0.5 mM) were the only ones
5.0
4.0
3.0
PPM
Figure 3. Confirmation of methylamine and formaldehyde as the ultimate carbon-containing products of the Fenton degradation of NDMA by 'H NMR. Spectra of (a) 24 mM NDMA mixed with 60 mM FeSO, and 60 mM Hz02in 30 mM D2S04,which was then basified with a few drops of NaOD to give a pH value of 8.0, (b) reaction mixture a spiked with authentic methylamine, and (c) solution b spiked with authentic formaldehyde. detected besides that of the unreacted NDMA (10.2 f 0.7 mM), as shown in Figure 2b. The radioactivity found in the two product peaks was approximately equal, and the total activity eluted accounted for 100% of the NDMA initially subjected to the reaction. The identity and quantity of the formaldehyde produced were confirmed by an isotope dilution technique in which an aliquot of a separate reaction mixture was mixed with cold formaldehyde and converted to the dimedone derivative, which was recrystallized to constant melting point for scintillation counting (expected, 0.107 FCi; found by isotope dilution, 0.107 pCi). Methylamine identity was confirmed by NMR, as shown in Figure 3. To investigate the fate of the remaining (nitroso) nitrogen atom, the reaction mixture was also analyzed by both the Griess reaction and an ion chromatography system capable of cleanly separating nitrate and sulfate. Nitrite, the normal product of enzymatic denitrosation, was found at a concentration of only 0.33 f 0.27 mM, while nitrate was the major product at 2.5 f 0.3 mM. Thus total nitritelnitrate was found in an amount that appeared to
250
Heur et al.
Chem. Res. Toxicol., Vol. 2, No. 4, 1989
a)
I in D,SO,
b)
,
,
Unreacted
220
I/
1 1
8.0 7.5
7.0
6.5
6.0 5.5 5.0
1:
1
i
'
4.5
4.0
I' '
1
3.5 3.0
Figure 4. Identification of imine intermediate by lH NMR. Spectra of (a) authentic iminium ion produced on dissolving 1,3,5-trimethylhexahydro-1,3,5-triazine in 1.3 M D2SO4, (b) reaction of 24 mM NDMA with 60 mM FeS04 and 60 mM H202 in 30 mM D2S04,and (e) mixture of solutions a and b. be slightly less than equimolar to formaldehyde and methylamine. It is possible that some of the nitric oxide (a low-boiling gas) may have escaped from the reaction mixture before it could be completely oxidized to nitrite and subsequently to nitrate. Taken together, these data indicate that the overall stoichiometry of the Fenton degradation of NDMA is NDMA + Fe2+ + H+ + HzOz 2. basify (13 mM) (15 mM) (15 mM) (15 mM) CHzO CH3NHz + N03-/N02- + NDMA (unreac) (3.1 mM) (3.2 mM) (2.8 mM) (10.2 mM) 1. stir at 4-10 "C
+
Direct Observation of the Intermediate Imine. When the Fenton degradation was run in strong acid and analyzed without basification to remove the iron, the major product was found to be something other than formaldehyde or methylamine, as shown in Figure 2a. Since the Schiff base, CH3N=CH2, had been postulated to be an intermediate in both the microsomal denitrosation (Figure 1) and the Fenton degradation of NDMA, we compared the intermediate with authentic CH3N=CHz prepared by dissolving its cyclic trimer (1,3,5-trimethylhexahydro-l,3&triazine) in strong acid. The 'H NMR spectral comparison is shown in Figure 4. Despite con-
200
180
180
140
120
100
80
60
40
20
0 PPM
Figure 5. Confirmation of the imine intermediate in the Fenton degradation of NDMA by 13C NMR. Spectra of (a) authentic iminium ion produced on dissolving 1,3,5-trimethylhexahydro1,3,5-triazinein D2S04,(b) reaction of [I3C]NDMA(24 mM) with 60 mM FeS04and 60 mM H202in 30 mM D2S04,and (c) mixture of solutions a and b. siderable line broadening and uncertainties in peak positions caused by the presence of paramagnetic iron in the Fenton reaction mixtures, two major peaks of equal intensity due to unreacted NDMA were observed (Figure 4b). In addition, two smaller peaks appeared in an integral ratio of approximately 3:2 that were similar in chemical shift to those of the authentic Schiff base (Figure 4a) in acid. Consistent with the identification of the Fenton intermediate as N-methylformaldimine, mixing their respective 'H NMR solutions in approximately equimolar proportions (Figure 4c) added no new peaks to the spectrum. Similar results were obtained by 13C NMR when the comparison was repeated with isotopically enriched NDMA, as shown in Figure 5. Further confirmation was obtained by fast atom bombardment mass spectral examination of a reaction mixture prepared in such a way that all the NDMA was consumed. As shown in Figure 6, the base peak was seen at mfz 45, as expected for the imine in deuteriated acid. A similar reaction in protiated solvent run so as to destroy all the NDMA was subjected to catalytic hydrogenation. Dimethylamine was found as the only 14C-containing product, as predicted for reduction of the iminium ion.
Discussion These results show that the Fenton reagent, a known radical-generating system, cleanly converts NDMA to effectively equimolar amounts of methylamine, formaldehyde, and nitritefnitrate. This outcome is so similar to that of the microsomal denitrosation of NDMA (Figure
Chem. Res. Toxicol., Vol. 2, No. 4, 1989 251
Denitrosation of NDMA by the Fenton Reagent
0 CH2= N-CH3
D
80
30-
30i I 20 10-
0
"
I,
I
II
i I
50
iI
I ,
60
,I
70
mlz
Figure 6. Positive ion FAB mass spectral confirmation of the imine as an intermediate in the Fenton degradation of NDMA.
7) that it offers strong evidence for a close parallelism in mechanism between the two reactions. The final products differ only in that the strongly oxidizing conditions of the Fenton system further oxidize much of the nitrite to nitrate. The intermediates, however, appear to be identical. Haussmann and Werringloer (17) found evidence for nitric oxide ('NO) as an intermediate in the enzymatic pathway, but the other postulated product of nitrosamino radical fragmentation, N-methylformaldimine, has been too unstable to be observed under physiological conditions. By contrast, we were able to stabilize and identify the imine, while the nitric oxide defied direct confirmation in the Fenton reaction. All of these considerations are fully consistent with the hypothesis that the a-radical is the initial product of NDMA metabolism. Results with this nonenzymatic model for the denitrosation pathway suggest answers to several remaining questions about the biological reaction. Firstly, formaldehyde could not be confirmed as the byproduct of methylamine production in the cytochrome P-450 induced reaction because denitrosation normally represents only 10-20% of the overall metabolism and formaldehyde is the
product of the major pathway. The present data show that it is reasonable to assume formaldehyde to be the nonbasic organic product of enzymatic NDMA denitrosation. Secondly, direct evidence for the intermediacy of the postulated imine intermediate was obtained by running the reaction in acidic media to stabilize it, while this is not possible for the enzymatic reaction. Perhaps most importantly, these results prove that the denitrosation pathway can,in principle, be uncoupled from the dealkylation route (12, 24). In the microsomal metabolism of NDMA, denitrosation is usually a minor course, accounting for a relatively reproducible 10-20% of total metabolism. In the Fenton reaction, this normally minor pathway becomes the exclusive outcome. This suggests that the further oxidation of the a-nitrosamino radical, once it is formed at the active site of the P-450, is so rapid that nearly all of it is converted to the a-hydroxy compound, with only a small fraction escaping to the denitrosation mechanism. This further oxidation must be remarkably rapid to capture so much of the radical, which is so unstable in aqueous media that we have not been able to observe electron spin resonance signals for it, nor have spin trapping experiments as yet proven successful. These considerations suggest that the a-nitrosamino radicals might serve as excellent radical clocks (30-32) for probing the sequence and timing of steps in the action of the P-450 enzymes that generate them. If the lifetimes of a series of such radicals can be measured by independent means, information on the fraction of each that escapes further oxidation by fragmenting to denitrosation products' could be quite useful in gathering data on the rates of such oxidation. It is relevant to ask why the dealkylation pathway is not observed in the Fenton reaction if both *OH and the anitrosamino radical are produced. As shown in Figure 1, they could couple to form the a-hydroxy nitrosamine which would decompose to an electrophile capable of alkylating water to form methanol. Previous investigators who have examined the effects of enzyme-free radical-generating systems on nitrosamines (33-38) have sometimes reported success in observing hydroxylation, but we found no methanol in the Fenton degradation of NDMA. Apparently, the concentrations of the two unstable radicals are so small at any given moment that a bimolecular reaction between them is relatively improbable. Other variances between our results and those of the earlier reports may
Oxidation
NO;
CH2
Figure 7. Comparison of chemical changes postulated and observed in (a) enzymatic denitrosation of NDMA and (b) Fenton degradation of NDMA. Intermediates not yet directly confirmed in the respective transformations are enclosed in brackets.
252 Chem. Res. Toxicol., Vol. 2, No. 4, 1989
be due to differences in substrate choice and other experimental factors, plus the fact that amine products would have escaped detection by the analytical methods employed in many of the prior studies. The finding that dealkylation and denitrosation are potentially separable pathways may have important implications for cancer risk reduction (12,21, 24). It could be possible to protect the body from the consequences of exposure to a potent carcinogen by selectively inhibiting its activation, allowing some diversionary mechanism that does not result in genotoxicity to clear the xenobiotic from the system at the expense of the activation route. It is possible that enzymatic denitrosation may be such a diversionary mechanism for NDMA. Further investigations of this interesting pathway are in progress with the aim of devising strategies for mobilizing it to reduce the cancer risk of exposure to NDMA even when the carcinogen is formed endogenously.
Acknowledgment. We thank J. A. Hrabie for preparing the [13C]NDMA,C. J. Metral for performing the MS analysis, J. Klose for running the NMR spectra, D. Weiss-Mellini and N. Schultz for nitrite and nitrate analyses, and B. A. Mico and M. Y. K. Ho for important suggestions. Registry No. NDMA, 62-75-9; methylamine, 74-89-5; formaldehyde, 50-00-0; N-methylformaldiminium ion, 51943-18-1.
References (1) Garland, W. A., Kuenzig, W., Rubio, F., Kornychuk, H., Norkus, E. P., and Conney, A. H. (1986) Urinary excretion of nitrosodi-
methylamine and nitrosoproline in humans: interindividual and intraindividual differences and the effect of administered ascorbic acid and a-tocopherol. Cancer Res. 46, 5392-5400. (2) Preussmann, R., and Stewart, B. W. (1984) N-Nitroso carcinogens. In Chemical Carcinogens (Searle, C. E., Ed.) 2nd ed., pp 643-828, ACS Monograph 182, American Chemical Society, Washington, DC. (3) Dipple, A,, Michejda, C. J., and Weisburger, E. K. (1985) Metabolism of chemical carcinogens. Pharmacol. Ther. 27,265-296. (4) Grilli, S., and Prodi, G. (1975) Identification of dimethylnitrosoamine metabolites in vitro. Gann 66, 473-480. (5) Rowland, I. R., and Grasso, P. (1975) Degradation of N-nitrosamines by intestinal bacteria. Appl. Microbiol. 29, 7-12. (6) Appel, K. E., Ruf, H. H., Mahr, B., Schwarz, M., Rickart, R., and Kunz, W. (1979) Binding of nitrosamines to cytochrome P-450 of liver microsomes. &em.-Biol. Interact. 28, 17-33. (7) Schwarz, M., Appel, K. E., Schrenk, D., and Kunz, W. (1980) Effect of ethanol on microsomal metabolism of dimethylnitrosamine. J . Cancer Res. Clin. Oncol. 97, 233-240. (8) Schrenk, D., Schwarz, M., Tennekes, H. A., and Kunz, W. (1982) A novel pathway of nitrosamine metabolism in liver microsomes: denitrosation of nitrosamines by cytochrome P-450. Adu. Exp. Med. Biol. 136B, 1157-1164. (9) Lorr, N. A., Tu, Y. Y., and Yang, C. S.(1982) The nature of nitrosamine denitrosation bv rat liver microsomes. Carcinogenesis 3, 1039-1043. (10) Kuthan. H.. Haussmann. H.-J., and Werrinaloer, J. (1984) Microsomal denitrosation of N-nitrosodimethylamine. In Oxygen Radicals in Chemistry and Biology (Bors, W., Saran, M., and Tait, D., Eds.) pp 231-234, de Gruyter, Berlin. (11) Yang, C. S., Tu, Y. Y., Koop, D. R., and Coon, M. J. (1985) Metabolism of nitrosamines by purified rabbit liver cytochrome P-450 isozymes. Cancer Res. 45, 1140-1145. (12) Haussmann, H.-J., and Werringloer, J. (1985) Oxidative denitrosation and activation of N-nitrosodimethylamine. Biochem. Pharmacol. 34, 411-412. (13) Tu, Y. Y., and Yang, C. S. (1985) Demethylation and denitrosation of nitrosamines by cytochrome P-450 isozymes. Arch. Biochem. Biophys. 242, 32-40. (14) Appel, K. E., Wiessler, M., Schoepke, M., Ruhl, C. S., and Hildebrandt, A. G. (1986) Metabolic nitrite formation from N nitrosamines: are there other pathways than reductive denitro-
Heur et al. sation by cytochrome P-450? Carcinogenesis 7, 659-663. (15) Patten, C. J., Ning, S.M., Lu, A. Y. H., and Yang, C. S. (1986) Acetone-inducible cytochrome P-450: purification, catalytic activity, and interaction with cytochrome-b,. Arch. Biochem. Biophys. 251, 629-638. (16) Keefer, L. K., Anjo, T., Hew, Y.-H., Yang, C. S., and Mico, B. A. (1987) Potential for metabolic deactivation of carcinogenic N-nitrosodimethylamine in uiuo. In Releuance of N-Nitroso Compounds t o Human Cancer: Exposures and Mechanisms (Bartsch, H., O'Neill, I. K., and Schulte-Hermann, R., Eds.) pp 113-1 16, IARC Scientific Publications 84, International Agency for Research on Cancer, Lyon. (17) Haussmann, H.-J., and Werringloer, J. (1987) Mechanism and control of denitrosation of N-nitrosodimethylamine. In Relevance of N-Nitroso Compounds to Human Cancer: Exposures and Mechanisms (Bartsch, H., O'Neill, I. K., and Schulte-Hermann, R., Eds.) pp 109-112, IARC Scientific Publications 84, International Agency for Research on Cancer, Lyon. (18) Arillo, A., and Tosetti, F. (1987) Denitrosation of N-nitrosodimethylamine and N-nitrosomethylurea by liver microsomes from trout (Salmo gairdneri Rich.). Enuiron. Res. 42, 366-371. (19) Yoo, J.-S. H., Ning, S. M., Patten C. J., and Yang, C. S. (1987) Metabolism and activation of N-nitrosodimethylamine by hamster and rat microsomes: comparative study with weanling and adult animals. Cancer Res. 47, 992-998. (20) Keefer, L. K., Anjo, T., Wade, D., Wang, T., and Yang, C. S. (1987) Concurrent generation of methylamine and nitrite during denitrosation of N-nitrosodimethylamine by rat liver microsomes. Cancer Res. 47, 447-452. (21) Wade, D., Yang, C. S., Metral, C . J., Roman, J. M., Hrabie, J. A., Riggs, C. W., Anjo, T., Keefer, L. K., and Mico, B. A. (1987) Deuterium isotope effect on denitrosation and demethylation of N-nitrosodimethylamine by rat liver microsomes. Cancer Res. 47, 3373-3377. (22) Malaveille, C., Brun, G., Park, S. S., Gelboin, H. V., and Bartsch, H. (1987) A monoclonal antibody against cytochrome P-450 enhances mutagen activation of N-nitrosodimethylamine by mouse liver S9: studies on the mode of action. Carcinogenesis 8, 1775-1779. (23) Yoo, J.-S. H., Guengerich, F. P., and Yang, C. S. (1988) Me-
tabolism of N-nitrosodialkylamines by human liver microsomes. Cancer Res. 48, 1499-1504. (24) Amelizad, Z., Appel, K. E., Oesch, F., and Hildebrandt, A. G. (1988) Effect of antibodies against cytochrome P-450 on demethylation and denitrosation of N-nitrosodimethylamine and Nnitrosomethylaniline. J . Cancer Res. Clin. Oncol. 114, 380-384. (25) Magee, P. N., Harrington, G . W., Pylypiw, H., Dollard D., Kozeniauskas, R., Bevill, R. F., Thurmon, J., and Nelson, D. (1988) The metabolism of N-nitrosodimethylamine in the pig. Proc. Am. Assoc. Cancer Res. 29, 124 (Abstract 492). (26) Walling, C. (1975) Fenton's reagent revisited. Acc. Chem. Res. 8, 125-131. (27) Sohn, 0. S., Fiala, E. S., Puz, C., Hamilton, S. R., and Williams, G. M. (1987) Enhancement of rat liver microsomal metabolism of
azoxymethane to methylazoxymethanol by chronic ethanol administration: similarity to the microsomal metabolism of N nitrosodimethylamine. Cancer Res. 47, 3123-3129. (28) Vorlander, D., and Kalkow, F. (1899) Die Formaldehydverbindung des Hydroresorcins. Liebigs Ann. Chem. 309, 356-374. (29) Olah, G. A., Donovan, D. J., and Keefer, L. K. (1975) Carcinogen chemistry. I. Reactions of protonated dialkylnitrosamines leading to alkylating and aminoalkylating agents of potential metabolic significance. J . Natl. Cancer Inst. 54, 465-472. (30) Griller, D., and Ingold, K. U. (1980) Free-radical clocks. Acc. Chem. Res. 13, 317-323. (31) Mathew, L., and Warkentin, J. (1986) The cyclopropylmethyl free radical clock. Calibration for the range 30-89 "C. J . Am. Chem. SOC.108, 7981-7984. (32) Ortiz de Montellano, P. R., and Steams, R. A. (1987) Timing of the radical recombination step in cytochrome P-450 catalysis 109, 3415-3420. with ring-strained probes. J. Am. Chem. SOC. (33) Preussmann, R. (1964) Zum oxydativen Abbau von Nitrosaminen mit Enzym-freien Modell-Systemen. Arzneim.-Forsch. 14, 769-774. (34) Rayman, M. P., Challis, B. C., Cox, P. J., and Jarman, M. (1975)
Oxidation of N-nitrosopiperidine in the Udenfriend model system and its metabolism by rat-liver microsomes. Biochem. Pharmacol. 24, 621-626. (35) Hsieh, S.-T., Kraft, P. L., Archer, M. C., and Tannenbaum, S. R. (1976) Reaction of nitrosamines in the Udenfriend system:
Denitrosation of NDMA by t h e F e n t o n Reagent principal products and biological activity. Mutat. Res. 35,23-28. (36) Suzuki, E., Mochizuki, M., Shibuya, K., and Okada, M. (1983) N,N-Dibutylnitrosamine metabolism by rat liver S9 fraction and oxidation by chemical model systems. Gann 74, 41-50. (37) Jarman, M., and Manson, D. (1986) The metabolism of N -
Chem. Res. Toxicol., Vol. 2, No. 4, 1989 253 nitrosomorpholine by rat liver microsomes and its oxidation by the Fenton system. Carcinogenesis 7 , 559-565. (38) Loeppky, R. N., and Li, E. (1988) Diazonium ion derived products from the Ce(1V) oxidation of @-hydroxynitrosamines. Chem. Res. Toxicol. 1, 334-336.
