Supercritical fluid extraction in the determination of tobacco-specific N

chewing tobacco and snuff including the induction of tu- mors of the oral cavity (2,5-7). The most abundant, strong carcinogens in smokeless tobacco a...
0 downloads 0 Views 705KB Size
Chem. Res. Toxicol. 1992,5, 336-340

336

Art i c 1es Supercritical Fluid Extraction in the Determination of Tobacco-Specific N-Nitrosamines in Smokeless Tobacco Bogdan Prokopczyk,* Dietrich Hoffmann, Jonathan E. Cox, Mirjana V. Djordjevic, and Klaus D. Brunnemann Naylor Dana Institute for Disease Prevention, American Health Foundation, Valhalla, New York 10595 Received October 31, 1991 A new approach to the analysis of the carcinogenic, tobacco-specific N-nitrosamines (TSNA) in moist snuff tobacco is based on the extraction of tobacco with methanol-modified supercritical carbon dioxide. Extracted TSNA are trapped across a glass cartridge filled with Tenax GR, from which they are subsequently released by thermal desorption and analyzed by capillary gas chromatography with a thermal energy analyzer. The analytical recoveries for the major TSNA range from 83 t o 98%; the detection limits are below 2 ng/g. T h e methodology is fast, reproducible, highly selective, and sensitive. The supercritical fluid extraction (SFE) releases up to 7 times more of the highly carcinogenic 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone (NNK) from tobacco than has been determined after conventional solvent extraction. Studies have confirmed that this is not an artifact. In contrast, the cyclic N-nitrosamines, N’-nitrosonornicotine, N’-nitrosoanabasine, and N’-nitrosoanatabine, showed no significant quantitative differences whether determined by the SFE method or the conventional solvent extraction method.

Introduction Epidemiologicalstudies have demonstrated that the oral use of snuff is causally associated with Cancer of the mouth and pharynx. Snuff dipping has also been linked with an increased risk for cancer of the nasal cavity, esophagus, pancreas, and urinary bladder (1-4). Bioassays in laboratory animals have shown carcinogenic activity for chewing tobacco and snuff including the induction of tumors of the oral cavity (2,5-7). The most abundant, strong carcinogens in smokeless tobacco are the nicotine-derived tobacco-specificN-nitrosamines (TSNA),l N’-nitrosonornicotine (NNN) and 4-(methy1nitrosamino)-1-(3pyridy1)-1-butanone (NNK; Figure 1). NNN and NNK induce benign and malignant tumors in mice, rats, and hamsters in the lung, upper aerodigestive tract, liver, and/or pancreas (8). A mixture of NNN and NNK induces tumors of the oral cavity in rats (5). Upon metabolic activation, NNN and NNK bind to DNA and hemoglobin (8). In snuff dippers and smokers, the resulting hemoglobin adduct has been utilized as a biomarker for the exposure to carcinogenic TSNA. The concentration of the hemoglobin adduct in snuff dippers was far higher than could be expected on the basis of analytical data for TSNA in snuff as determined by conventional methods (9,lO). Therefore, it was hypothesized either that significant amounta of NNK and/or NNN were formed endogenously in snuff dippers from nicotine, nornicotine, and nitrite or that the analytical data did not reflect the total concentrations of TSNA in snuff. Considering the latter possibility we developed a more efficient, and yet simplified, analytical method for the TSNA determination in snuff

* T o whom correspondence should be addressed.

tobacco. The key step in this method is the application of supercritical COz as an extracting solvent. Supercritical fluid extraction (SFE)has been used for many years to selectively remove certain compounds in food and chemical processing industries. Typical industrial applications include caffeine extraction from coffee beans (1l ) , nicotine from tobacco (12), and fat and oil extraction from plant and animal tissue. In recent years SFE has found wide applications as an analytical tool. A comprehensive review of different supercritical fluid extraction methods has been published by S. B. Hawthorne (13). It is evident that SFE can be a very useful analytical tool for identification and quantitation of a variety of organic compounds, widely varying in polarity. However, according to our knowledge no reports have been published on the application of SFE to the quantitation of the highly carcinogenic tobacco-specific N-nitrosamines. In this paper we describe a new method for the analysis of TSNA in snuff tobaccos. This method isolates significantly more NNK than was recovered through conventional solvent extraction-gas chromatography (GC) methods (9,10,14,15).

Experimental Procedures Materials. Leading US.moist snuff brands (A and C) were obtained from retailers in Weetchester County, N e w York. T h e Abbreviations: GC, gas chromatography; HPLC, high-performance liquid chromatography; MS, mass spectrometry; MSD, mass-selective detector; NAB, N’-nitrosoanabasine; NAT, N’-nitrosoanatabine; NG, N-nitrosoguvacoline;NNK, 4-(methylnitro9amino)-l-(3-pyridyl)-l-butanone; NNN, N’-nitrosonornicotine; SFE,supercritical fluid extraction; SIM, selective-ion monitoring; TD, thermal desorption; TDU, thermal desorption unit; TEA, thermal energy analyzer; TSNA, tobacco-specific N-nitrosamines.

0893-228~ 192J 27Q5-0336$Q3.QQ/Q0 1992 American Chemical S o c i e t y

Chem. Res. Toxicol., Vol. 5, No. 3, 1992 337

N-Nitrosamines in Smokeless Tobacco

NO NNN

NNK

N'-Nitrosonornicotine

4-1Methylnitrosamino)~l-(3-pyridyl).l-butanone

@Y

n N=O

NAT

NAB

N'.Nilrosoanalabine

N'-Nitrosoanabasine

Figure 1. Tobacco-specific N-nitrosamines.

Pi-Q

l:l

I s 1

Gas Outlel

0 0 Regulator

a

I

Collector

I

Megabore Column

T a n k

U

I

Chamber

U

Compressor

L

#2

,

t GC Oven

Figure 2. Schematic of the supercritical extraction apparatus. third snuff brand (B) was bought in Gothenburg, Sweden. The samples were stored in a cold room (4 "C). Carbon dioxide (99.99%) was purchased from Liquid Carbonic Specialty Gas Corp., Chicago, IL, and Tenax TA (20/35 mesh) was procured from Chrompack, Raritan, NJ. Tenax GR (20/35 mesh) was obtained from Aldrich, Milwaukee, WI. Carbotrap (20/40 mesh) was supplied by Supelco Inc., Bellafonte, PA. The reference compounds NNN, NNK, N'-nitrosoanatabine (NAT), and N'nitrosoanabasine (NAB) as well as the internal standard N-nitrosoguvacoline [methyl l-nitroso-1,2,5,6-tetrahydronicotinate (NG)] were synthesized according to previously published methods (16-18). The purity of the synthesized compounds, verified by capillary GC and high-performance liquid chromatography (HPLC), was greater than 99%. TSNA are established animal carcinogens; therefore, they were handled according to the NIH Guidelines for the Laboratory Use of Chemical Carcinogens (NIH Publication No. 81-2385). Instrumentation. The extractions were performed in a simple, self-assembled apparatus (Figure 2). The required pressure was generated by a motor-driven diaphragm compressor (New Port Scientific, Inc., Jessup, MD, Model 46-13411-2) capable of producing pressures up to 670 atm. The extraction chamber was made of a stainless steel threaded tube (4 in. X 3/8-in. i.d.; Autoclave Engineers Inc., Erie, PA). This chamber was heated to the required temperature in a GC oven (Hewlett-Packard Model 5710A). The pressure of the system was controlled by two medium-pressure metering valves (valve 1, Autoclave Engineers, no. 20SC4071; valve 2, Nupro fine metering valve, Nupro Co., Willoughby, OH). Fittings and connectors were either Autoclave Engineers medium-pressure components or Swagelok fittings (Swagelok, Solon, OH). To prevent accidental overpressurization, the system was equipped with a safety valve, 20-RPV 9072 (Autoclave Engineers), set a t 1000 atm. The extracts were collected after depressurization on glass cartridges (Chrompack, Raritan, NJ, no. 0162251), packed with 200 mg of Tenax GR (20-35 mesh). Prior to their use the cartridges were preconditioned by washing with 2 mL of methanol and 2 mL of n-hexane and were kept overnight a t 250 "C. The cartridge and fine metering valve 2 were connected by a 1-m uncoated silica megabore column (J&W Scientific, Folsom, CA). The Chrompack thermal desorption cold trap injector (Chrompack, Raritan, NJ, Model 16230) was used to introduce samples for GC analysis. The TSNA were separated and quantitated on a HewlettPackard Model 5890 gas chromatograph interfaced with a Model 502 thermal energy analyzer (Thermal Electron Corp., Waltham, MA) and HP Model 3393A integrator (Hewlett-Packard, Paramus,

NJ). The GC/mass-selective detector (GC/MSD) system consisted of a Hewlett-Packard Model 5890A gas chromatograph interfaced with an H P Model 5970 mass-selective detector, and an H P 5970 GC/MS workstation run on an H P 9000 Series 300 computer. Sample Preparation. The ground, freeze-dried snuff (5-20 mg), containing 4.2-4.5% water, was placed in the extraction chamber and was held in place by glass wool plugs. Methanol (0.25 mL) was added to the extraction chamber. Air was removed by flushing carbon dioxide through the chamber. The system was heated to 60 "C with both valves closed, then pressurized to 435 atm, and maintained under these conditions for 30 min. Then, the extraction was run for an additional 45 min in the dynamic mode (both valves opened, constant flow of exiting gaseous carbon dioxide at 200 mL/min, 60 "C, 435 a b ) . The sample was collected across the glass cartridge filled with Tenax GR. The system was then depressurized, and another 0.25 mL of methanol was added directly to the extraction chamber before the extraction process was repeated. The "loaded" glass cartridge was placed in the thermal desorption unit (TDU) and heated to 220 "C over 12 min. During this time, the temperature of the cryofocusing zone of the TDU was maintained a t 0 "C. During this operation the flow rate of helium was kept a t 10 mL/min. Afterward, the gas flow was changed to 1mL/min and the sample was transferred to the first part of the capillary column by ballistically heating the cryofocusing part of the TDU (15 "C/s) to 220 "C, while the temperature of the GC oven was maintained at 40 "C. In order to quantify portions of the extracted TSNA trapped in the extraction device, the fine metering valve 2 and the megabore column were washed with 3 mL of methanol. The organic solvent was evaporated to dryness. Samples were dissolved in 100 p L of methanol containing 8 pg NG/mL, and 2-pL aliquots were analyzed by GC-TEA. The levels of TSNA determined by this analysis were combined with those established by the thermal desorption of the trap. GC Conditions. TSNA were analyzed on a DB-5 fused silica capillary column (30 m X 0.32 mm, 0.25-pm film thickness, J&W Scientific, Folsom, CA), using the following temperature program: the initial temperature of 40 OC was kept for 10 min, then increased a t 20 "C/min to 140 "C, maintained a t this level for 5 min, and then increased again a t 1 OC/min to 145 "C. This temperature was kept for 5 min, and finally raised to 220 "C at 20 OC/min, and held for an additional 5 min. The injection port was maintained at 220 "C. Under these chromatographic conditions NG (internal standard) was eluted at 9.20 min, and NNN, NAT, NAB, and NNK were eluted at 14.95,15.95,16.24,and 17.64 min, respectively. GCIMSD Conditions. The structure and levels of NNK in snuff were also verified by GC/MS. The Tenax-filled glass cartridge was eluted with 5 mL of methanol. Valve 2 and the megabore column were washed with 3 mL of methanol. The organic extracts were combined, concentrated by evaporation in vacuo to approximately 1 mL, and further concentrated to 100 pL by a gentle stream of nitrogen. Two-microliter aliquots were analyzed using the GC conditions described above. The GC/MSD system consisted of a Hewlett-Packard Model 58906 gas chromatograph, interfaced with an H P Model 5970 mass-selective detector, and H P 5970 GC/MS workstation software run on an H P 9000 Series 300 computer. The GC/MSD ion source and transfer line were maintained a t 200 and 280 "C, respectively. The source pressure was approximately 5 X Torr, and the emission current and ionization energy were the standard 300 pA and 70 eV. For the identification of NNK the total ion chromatogram was scanned, while for quantitative analyses the MSD was used in the single-ion monitoring (SIM) mode. The ions of m / z 78,106, 146, and 177, characteristic for NNK, were selected for quantitation. Stock Solutions. Stock solutions of TSNA were prepared in acetonitrile. Standard solution 1 contained (per milliliter) 9.3 pg of NNN, 7.6 pg of NAT, 0.72 pg of NAB, and 4.3 pg of NNK, standard solution 2 contained 27.9 pg of NNN, 22.8 pg of NAT, 2.16 pg of NAB, and 12.9 pg of NNK, and standard solution 3 contained 46.5 pg of NNN, 38.0 pg of NAT, 3.6 pg of NAB, and 21.5 pg of NNK. These solutions were stored a t 4 "C.

Prokopczyk et al.

338 Chem. Res. Toxicol., Vol. 5, No. 3, 1992 Table I. Comparison of Extraction Efficiencies Obtained Using COz and COz Modified with either Methanol, n -Hexane, or Ethyl Acetate" % recovery NNN NAT NNK 15.9 14.4 16.2 COP 47.9 41.3 49.1 COz + methanol 6.1 6.1 9.8 COz + n-hexane 15.9 13.4 29.3 COz + ethyl acetate "The extractions were performed at 60 "C and 435 atm for 30 min in a static mode followed by 45-min extraction in a dynamic mode. N-Nitrosamines were extracted from spiked, analyte-free snuff tobacco.

The linearity of each standard was confirmed by plotting the nitrosamine concentration versus the respective peak area. Testing for Artifacts. Filter paper squares (3 X 3 cm, 100 mg) or analyte-free smokeless tobacco were spiked with 0.5, 1.0, and 2.0 mg of nicotine, respectively; or with 0.5, 1.0, and 2.0 mg of nornicotine; and with 0.25 pg of nitrite and 0.25 mg of nitrate. SFE was then performed with 100, 267, and 535 atm at 40, 60, and 90 "C, either with or without methanol or water. No artifactual formation of TSNA was observed. Conventional Analysis of TSNA. To provide comparative analytical data, TSNA were extracted and quantified in snuff tobaccos using a previously published method (IO). After solvent extraction, samples were generally analyzed by GC-TEA using the column and conditions described in the Experimental Section for analyses of TSNA in supercritical fluid extracts.

-

Results and Discussion Method Development. Several experimental parameters were investigated. Initially, the extractions were performed at 60 "C in a static mode at 300,435, and 533 atm for 30 min. Only trace amounts of TSNA were extracted at 300 atm. The efficiency of extraction was slightly improved at 435 atm. At this pressure, approximately 6% NNN, 5% NAT, and 7% NNK were recovered from the fiiter paper. Further improvement was observed when the pressure was raised to 535 atm. Under these conditions 8.2% NNN, 6.5% NAT, and 9.1% NNK were extracted. The extraction time of 30 min was found sufficient for the equilibrium to occur. On the other hand, when the extractions were performed for 15 min, approximately 20% less TSNA was detected by GC analysis. Extending the extraction time to 45 min did not result in an improvement in TSNA recovery. Other attempts were also made to improve the efficiency of the supercritical extraction step. First, the static extraction was followed by the extraction in a dynamic mode. This resulted in further improvement in TSNA recovery. However, the most dramatic increase in the amount of TSNA extracted was observed when methanol was added as a modifier directly into the extraction chamber. The effect of modifiers on the extraction efficiency of TSNA is illustrated in Table I. Additional improvements in the efficiency of the supercritical method were achieved when the entire process of static plus dynamic extractions was repeated twice. Choice of Trapping Agent. Carbon (Carbotrap, 20/40 mesh), Amberlite, silica gel, Tenax TA, and Tenax GR were tested. There was considerable break-through with single cartridges (200 mg) filled with either silica gel or Amberlite. Therefore, two cartridges in series were needed to provide sufficient trapping capacity. In addition, these trapping materials contributed to an unacceptable background noise in the GC analysis. The thermal release from carbon did not yield reproducible results. The trapping capacity of Tenax GR for TSNA was found to be ap-

proximately 30% better than that of Tenax TA. No break-through was observed. Choice of Thermal Desorption Conditions. Several temperature programs were tested for optimal desorption of TSNA from the trap. Generally, NNK was most affected by changes in these conditions. The maximum response was obtained when the glass cartridge was heated from room temperature to 220 "C over a 12-min period while the temperature of the cryofocusing zone was maintained at 0 "C. This was followed by heating the cryofocusing zone to 220 "C for 10 min. When the desorption temperature was lowered to 200 "C, only 60% of NNK transferred to the capillary column. The peak areas of NNN, NAT, and NAB were not affected by this change in temperature. The low desorption of NNK was likely due to its relatively lower volatility and higher polarity compared to that of other TSNA. A similar effect was observed when the time of thermal desorption was shortened. On the other hand, when the desorption temperature was increased to 240 "C, thermal decomposition of NNN and NAT was observed and resulted in peak splitting. Accuracy and Precision. The analytical recovery and precision of this method at three different concentrations were assessed. Filter papers (3 X 3 cm, weighing approximately 100 mg) were spiked with 0.5 mL of standard solution 1, standard solution 2, and standard solution 3, respectively. Recoveries of TSNA analyzed in this study as a mean of replicate runs (n = 5) at each spiking concentration were 93.9-95.5% for NNN, 83.2-83.8% for NAT, 92.9-93.3% for NAB, and 94.6-98.7% for NNK, respectively. Low standard deviations were observed (f1.5-2.6% for NNN, f0.5-0.8% for NAT, fl.l-2.0% for NAB, and f2.9-5.8% for NNK). The recovery of TSNA from snuff tobacco was also determined by the addition of known quantities of nitrosamines to analyte-free matrix. The analyte-free matrix was obtained in the following experiment: Freeze-dried snuff tobacco (200 mg),wrapped in filter paper, was placed in the extraction chamber, and 0.25 mL of methanol was added. The extraction chamber was first flushed with COz, then heated to 60 "C, and pressurized to 435 atm. After 30 min of equilibration, the extraction was performed in the dynamic mode at 60 "C and 435 atm with the flow of exiting COz set at 200 mL/min. The system was then depressurized and cooled to room temperature. The entire process was repeated three times. Afterward, the tobacco sample was removed and stored in the desiccator over PzO3 This sample tested negative when analyzed for TSNA by the SFE-thermal desorption method. The recoveries of TSNA from the analyte-free matrix were in the same range as those determined by spiking the filter paper. For comparison, the recovery studies were also performed using the conventional extraction method. Recoveries of all TSNA analyzed by this method, as a mean of five repeat runs, ranged from 84.4% for NNK to 95.0% for NAT. We observed low standard deviations for all TSNA (f1.18% for NNN, f4.16% for NAT, h4.6870 for NAB, and *10.46% for NNK). The findings of this study demonstrate that carbon dioxide in its supercritical stage, and when modified with methanol, is a suitable solvent for extraction of TSNA from tobacco. However, because of their relatively low volatility, TSNA rapidly precipitate when carbon dioxide is depressurized. Their transfer to the trap cannot be enhanced by heating both the valve and the transfer line. Experiments with the valve and transfer line heated to 90 " C did

Chem. Res. Toxicol., Vol. 5, No. 3, 1992 339

N-Nitrosamines in Smokeless Tobacco

Table 11. Comparison of the Conventional Solvent Extraction Method with a New Methodology Utilizing Supercritical C 0 2 as a n Extracting Solvent concentrations of TSNA” in snuff tobaccos (fig/g dry weight) conventional supercritical snuff brand NNN NAT NAB NNK NNN NAT NAB NNK A mean 8.47 5.91 0.56 1.69 6.71 6.86 0.59 7.23 SD 1.49 0.14 0.04 0.06 0.71 0.40 0.10 0.65 B mean 5.67 3.49 0.26 1.56 4.40 2.03 0.30 10.5 SD 0.01 0.25 0.56 0.44 0.35 0.04 3.06 C mean 3.33 4.28 0.34 1.89 4.29 4.58 0.51 5.21 SD 0.14 0.27 0.02 0.18 0.60 0.39 0.20 0.25 a

Concentrations of TSNA were calculated from replicate determinations ( n = 5).

not demonstrate any significant increase of the TSNA in the Tenax trap. Consequently, only 60-7570 of the TSNA were recovered from the Tenax trap. The remaining TSNA had to be recovered by washing the fine metering valve and the transfer line (30-cm megabore column) that connects this valve with the trap. Thus, to calculate the total recovery, the amounts of TSNA precipitated in the valve and those in the megabore column were quantified and added to those obtained from the thermal release of the Tenax trap. Alternatively, to avoid two chromatographic runs, the entire system (the valve, transfer line, and the Tenax trap) should be washed with methanol (5 mL). After evaporation of the solvent, samples can be analyzed as described for the GC/MS analysis. Comparison of Methods. The new method of TSNA analysis was compared with the conventiona! method consisting of tobacco extraction with water (to which 20% of ammonium sulfamate in 3.6 N H2S04was added to prevent artifactual formation of nitrosamines), partition at three different pH levels with ethyl acetate, and GCTEA quantification. Comparisons of the results of both methods are shown in Table 11. The levels of NAT and NAB as determined by the two methods are almost the same. The difference for NNN is not significant and most likely reflects the lack of uniformity of snuff tobacco. However, the levels of NNK are strikingly different. According to previously published reports, the NNN:NNK ratios in snuff tobaccos ranged between 3:l and 5:l. On the other hand, the new method has shown that the levels of NNK exceed those of NNN. Since the recovery experiments do not support the reversal of this ratio and since there was no evidence for artifactual formation of nitrosamines, there are only two possible explanations. Either there is additional NNK present in tobacco, which is not released during conventional solvent extraction, or there is another nitrosamine, which coelutes with NNK under the chromatographic conditions used for TSNA analysis. The latter was excluded by GC/MSD analysis. The supercritical fluid extraction resulted in significantly higher concentrations of NNK than any of the conventional extraction methods employed thus far (IO,14,15). Since NNK is the most potent carcinogen known to occur in tobacco (8)) knowledge about its actual levels in tobacco is critical. Equally important is an answer to the question of how NNK is bound. Is it bound in the laminae, or in the stems? The presence of an open chain and carbonyl group distinguishes NNK from the other TSNA (Figure 1) which do not have polar, functinal groups and/or an open chain. Is there a covalent bond being formed in the processed tobacco? Even more important is whether this “bound” NNK is available to the snuff dipper. Can it be released during chewing by enzymes present in saliva? A series of assays are currently underway to answer these questions which are of major significance for estimating

the carcinogenic potential of smokeless tobacco for chewers and snuff dippers (8).

Conclusions On the basis of the results presented here, it can be concluded that the combination of supercritical fluid extraction, thermal desorption, and capillary GC with TEA detection provides a favorable method for the determination of TSNA in tobacco. The key step in this methodology lies in the application of supercritical carbon dioxide as an extracting solvent which releases additional amounts of NNK that are not amenable to conventional solvent extraction. Acknowledgment. This work was supported by Grant CA-29580 from the US.National Cancer Institute. Registry No. NNK, 64091-91-4; NNN, 16543-55-8; NAT, 71267-22-6; NAB, 1133-64-8.

References (1) Winn, D. M., Blot, W. J., Shy, C. M., Pickle, L. W., Toledo, A.,

and Fraumeni, J. F., Jr. (1981) Snuff dipping and oral cancer among women in the southern United States. New Engl. J. Med. 304, 745-749. (2) International Agency for Research on Cancer (1985) Tobacco Habits Other Then Smoking; Betal Quid and Areca-nut Chewing; and Some Nitroso Compounds. ZARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans, Vol. 37, IARC, Lyon. (3) U.S.Surgeon General (1986) The Health Consequences of Using Smokeless Tobacco. NIH Publication No. 86-2874. (4) Kabat, G. C., Dieck, G. S., and Wynder, E. L. (1986) Bladder cancer in non-smokers. Cancer 57, 362-367. (5) Hecht, S. S., Rivenson, A,, Braley, J., DiBello, J., Adams, J. D., and Hoffmann, D. (1986) Induction of oral cavity tumors in F344 rats by tobacco-specific nitrosamines and snuff. Cancer Res. 46, 4162-4166. (6) Park, N. H., Sapp, J. P., and Herbosa, E. G. (1986) Oral cancer induced in hamsters with herpes simplex infection and simulated snuff-dipping. Oral Surg., Oral Med., Oral Pathol. 62, 164-168. (7) Johansson, S. L., Hirsch, J. H., Larson, P. A., Sadi, J., and Oesterdahl, B. G. (1989) Snuff-induced carcinogenesis: Effect of snuff in rats initiated with 4-nitroquinoline-N-oxide. Cancer Res. 49, 3063-3069. (8) Hecht, S. S., and Hoffmann, D. (1989) The relevance of tobacco-specific nitrosamines to human cancer. Cancer Surveys 8, 273-294. (9) Carmella, S. G., Kagan, S. S., Kagan, M., Foiles, P. G., Palladino, G., Quart, A. M., Quart, E., and Hecht, S. S. (1990) Mass-spectrometric analysis of tobacco-specific nitrosamine hemoglobin adducts in snuff dippers, smokers, and non-smokers. Cancer Res. 50, 5438-5445. (10) Djordjevic, M. V., Brunnemann, K. D., and Hoffmann, D. (1989) Identification and analysis of a nicotine-derived N-nitrosamino acid and other nitrosamino acids in tobacco. Carcinogenesis 10, 1725-1731. (11) Swivama. K.. Saito. M.. Hondo. T.. and Senda. M. (1985) New doubikkage separation analysis method. Directly coupled laboratory-scale supercritical fluid extraction-supercritical fluid chromatography, monitored with a multiwavelength ultraviolet detector. J. Chromatogr. 332, 107-116.

340

Chem. Res. Toxicol. 1992, 5, 340-345

(12) Roselius, W., Vitzthum, O., and Hubert, P. (1979) Process for

the extraction of nicotine from tobacco. US.Patent 4,153,063, 9 PP. (13) Hawthorne, S. B. (1990) Analytical-scale supercritical fluid extraction. Anal. Chem. 62, 633A-642A. (14) Tricker, A. R., Haubner, R., Spiegelhalder,B., and Preussmann, R. (1988) The occurrence of tobacco-specificnitrosamines in oral tobacco products and their potential formation under simulated gastric conditions. Food Chem. Toxicol. 26, 861-865. (15) Chortyk, 0. T., and Chamberlain, W. J. (1991) The application of solid phase extraction to the analysis of tobacco-specific ni-

trosamines (TSNA). J. Chromatogr. Sci. 29, 522-527. (16) Hecht, S. S., Chen, C. B., Dong, M., Ornaf, R. M., and Hoff-

mann, D. (1977) Chemical studies on tobacco smoke. LI. Studies on non-volatile nitrosamines in tobacco. Beitr. Tabakforsch. 9, 1-6. (17) Christensen, S. B., and Krogsgaard-Lamen, P. (1977) Preparation of deuterium labeled guvacine and isoguvacine. J. Labelled Compd. 17, 191-202. (18) Hu, W., Bondinell, W. E., and Hoffman, D. (1974) Synthesis of carbon-14 labelled myosmine, nornicotine, and "-nitrosonornicotine. J. Labelled Compd. 10, 79-88.

The Oxidation of 4-Aminobiphenyl by Horseradish Peroxidase Michael F. Hughes,? Bill J. Smith,$ and Thomas E. Eling* Laboratory of Molecular Biophysics, National Institute of Environmental Health Sciences, National Institutes of Health, Post Office Box 12233, Research Triangle Park, North Carolina 27709 Received December 3, 1991 The oxidation of the carcinogen 4-aminobiphenyl(I-ABP)catalyzed by the model peroxidase enzyme horseradish peroxidase (HRP) was investigated. 4-ABP served as a reducing cosubstrate for HRP during the enzyme-catalyzed reduction of the synthetic hydroperoxide, 5-phenyl-4penten-1-yl hydroperoxide, to its corresponding alcohol. Spectral analysis during the incubation of HRP, 4-ABP, and H 2 0 2showed an increase in absorbance a t 230 and 325 nm and decrease at 270 nm, suggesting metabolite formation. Oxygen consumption was not detected in incubations of HRP, 4-ABP, and H202. However, oxygen uptake was observed after the addition of glutathione, which indicated that a free radical metabolite of 4-ABP was formed by the peroxidase. The 4-ABP free radical reacted with glutathione forming a glutathionyl radical which, in turn, reacted with and consumed oxygen. HPLC analysis of organic extracts of incubations with HRP, [3H]-4-ABP,and H202showed the formation of one major peak identified by mass spectroscopy as 4,4'-azobis(biphenyl). The addition of glutathione to the incubations decreased the formation of 4-ABP metabolites, suggesting a reduction of the 4-ABP free radical and/or the formation of glutathione conjugates. Subsequent HPLC analysis of incubations including [35S]glutathione indicated formation of several unidentified 4-ABP-glutathione conjugates as well as recovery of parent compound. These studies suggest that H R P metabolizes 4-ABP by a one-electron oxidation mechanism, resulting in formation of a free radical. This radical can either react with a second radical to form azobis(biphenyl), be reduced by glutathione back to parent, or react with glutathione to form glutathione conjugates.

Introduction Aromatic amines such as 4-aminobiphenyl (4-ABP)' and benzidine have been intensely studied for a number of years to determine the role metabolism has in their activation and subsequent carcinogenic effect (1). The generally accepted hypothesis is that the metabolism of many chemical carcinogens, including aromatic amines, results in the formation of electrophilic intermediates which react covalently with critical nucleophiles in the cell and initiate the process of carcinogenesis (2,3). Cytochrome P-450 is an important catalyst in the activation of many aromatic amines ( 4 ) , with the pivotal step being N-hydroxylation of these chemicals (5). Aromatic amines induce tumors in many different tissues and organs (6-8). 4-ABP induces tumors in several organs such as the bladder of dogs (7,9) and man (10) and the intestine and liver of rats (11). Several extrahepatic organs, such as the bladder, that are susceptible to aro*To whom correspondence should be addressed at NIEHS, MD 19-04, P.O. Box 12233, Research Triangle Park, NC 27709. +Presentaddress: ManTech Environmental Technology Inc., P.O. Box 12313, Research Triangle Park, NC 27709. *Present address: The Procter & Gamble Company, Health & Personal Care Technology Division, Miami Valley Laboratories, P.O. Box 398707, Cincinnati, OH 45239-8707.

matic amine induced tumorigenesis are low in cytochrome P-450 activity, yet contain appreciable amounts of the mammalian peroxidase, prostaglandin H synthase (PHS) (12, 13). Several aromatic amines are good peroxidase substrates (14). Studies have shown that microsomal PHS has a role in the metabolism of 4-ABP and its structural analogue benzidine (12, 13, 15). The involvement of peroxidases in the metabolism/activation of benzidine has been extensively studied (12,13,15-20). However, limited studies have been conducted on the peroxidase-mediated metabolism of 4-ABP. In this study, we report on the peroxidative oxidation of 4-ABP using horseradish peroxidase (HRP) as a model enzyme.

Experlmentai Procedures Materials. Ascorbic acid, benzidine (caution: a potential carcinogen), glutathione, and HRP type VI were purchased from Sigma Chemical Co. (St. Louis, MO). 5-Phenyl-4-penten-1-yl hydroperoxide (PPHP) was purchased from Oxford Biomedical Research (Oxford, MI). Unlabeled 4-ABP (caution: a potential carcinogen) was purchased from Aldrich Chemical Co. (Mill Abbreviations: 4-ABP, 4-aminobiphenyl; HRP, horseradish peroxidase; PHS, prostaglandin H synthase;PPHP, 5-phenyl-4-penten-1-yl hydroperoxide;PPA, 5-phenyl-4-pentenylalcohol; UV/vis, ultraviolet/ visible; TLC,thin-layer chromatography.

This article not subject to US.Copyright. Published 1992 by the American Chemical Society