[CANCER RESEARCH 46, 6136-6138 December 1986] Increase in Human Exposure to Methylamine Precursors of 7V-Nitrosamines after Eating Fish1

Steven H. Zeisel2 and Kerry-Ann DaCosta

Nutrient Metabolism Laboratory, Nutrition-Pathology Unit, Departments of Pathology and Pediatrics, Boston University School of Medicine, Boston, Massachusetts 02118

ABSTRACT disease (8, 9). A mechanism for the effects of fish upon heart disease has not been proven, but it has been suggested that the Consumption of fish has been encouraged recently because it may effect is related to the o)3-polyunsaturated fatty acids present in prevent mortality due to heart disease. Fish contains methylamines, which are precursors of A'-nitrosamines. Nitrosamines can act as potent carcin fish oils (10). On the basis of available epidemiological and ogens in a wide variety of animal species, and there is no reason to experimental data, it has been recommended that humans in assume that humans are resistant. Human subjects (n = 5) ingested a crease their intake offish (11). diet of known methylamine content for 2 days. On Day 3, they ate fish Physicians need to consider the risk/benefit ratio when rec at the luncheon and dinner meals. On Day 4, they again ate the control ommending that humans change their eating habits. Exposure diet. A single portion of fish contained as many methylamines as were to increased amounts of nitrosatable methylamines and there normally excreted by the human in 2 days. Urinary excretion of mono- fore possible exposure to nitrosamines are a risk of eating more methylamine remained constant (1.3 to 1.5 /imol/24 h/kg of body weight) fish. For this reason we examined the effect that eating fish has throughout the study. Dimethylamine excretion increased more than 4- upon the exposure of humans to MMA, DMA, and TMA. fold after fish was eaten (from 5.6 to 24.1 /¿mol/24h/kgof body weight), while trimethylamine excretion increased more than 8-fold (from 0.2 to 1.6 ¿tmol/24h/kgof body weight). MATERIALS AND METHODS We conclude that the consumption of fish significantly increased exposure to methylamines, particularly to dimethylamine. Although there Experimental Protocol. Five healthy, normal subjects (1 male, 4 is potential for the in vivo conversion of dimethylamine to nitrosodi- females), ages 23 to 35 yr, weighing 52.3 to 72.7 kg, gave their informed methylamine, a carcinogen, we know of no studies that have determined consent for participation in this study. For 2 days they ingested a that the ingestion of fish increases the risk of cancer. This should be control diet of known methylamine content. On Day 3 of the study, carefully investigated prior to recommending that humans change their fish was substituted as the chief constituent of the luncheon and dinner eating habits. meals. On Day 4, the subjects returned to the control diet. The diets ingested by the subjects are described in Table 1. INTRODUCTION Sample Collection. Urine was collected by all subjects in 24-h blocks. Collection vessels contained 50 ml of 3 N hydrochloric acid (to inhibit DMA' and TMA are important because they are precursors bacterial growth and to trap methylamines). Urine was refrigerated of NDMA (1). /V-Nitrosamines can act as potent carcinogens during the collection period. Urine volumes were recorded, and aliquots were frozen at -90°Cuntil assayed. Collection and storage under these in a wide variety of animals species, and there is no reason to assume that humans are resistant (2, 3). The National Academy conditions did not affect methylamine concentrations when known of Sciences has recommended that exposure of humans to N- amounts of methylamines were spiked into the collection vessels. A sample of each food in the diet (100 mg) was added to 2 ml of 0.1 nitroso compounds should be reduced (3). Although a causal N hydrochloric acid and mixed thoroughly. This was allowed to stand relationship between nitrosamine exposure and cancer in hu on ice for 1 h. The mixture was centrifugea at 3000 x g for 10 min at mans has not been rigorously proved, epidemiological studies 4°C,and the supernatant was saved for methylamine analysis. linking nitrosamines in tobacco products to oral cancer in Analysis of Methylamines. MMA, DMA, and TMA were measured humans strongly suggest such an association (2). using a gas-liquid chromai ography method (1). An aliquot of the Fish, as consumed by humans, contains significant quantities acidified biological fluid (urine or tissue extract) was placed into a of TMA and DMA (4, 5). TMAO is an end-product of nitrogen sealed vial (Teflon-lined WISP septum; Waters Instruments, Milford, metabolism in fish (5, 6). It replaces urea as the major route MA). 2-Propanol was added into each vial, along with an internal for nitrogen excretion. TMAO is metabolized to form TMA by standard of isopropylamine. Potassium hydroxide (65%) was injected into the vial to bring the pH to 13. Vials were vigorously shaken, heated bacteria once fish have been killed (6). TMA gives fish its at 60°Cfor 30 min to volatilize the amines, and then subjected to characteristic "fishy" odor. DMA is formed by the action of centrifugation at 1000 x g for 5 min at -4°C. An aliquot was drawn endogenous tissue enzymes upon TMAO and occurs maximally at temperatures below the freezing point offish (-5 to -10°C) up into a syringe, followed by 1 p\ of air and 1 n\ of 30% ammonium hydroxide. This was injected into the glass-lined injection port (pre- (5, 7). treated with potassium hydroxide to inactivate any binding sites for Epidemiological evidence, chiefly from studies on Eskimos methylamines) of the gas Chromatograph (Sigma 2000; Perkin Elmer, of Greenland, Japanese on Okinawa, and Dutch men in the Norwalk, CT). We used a 2-m-long (2-mm inner diameter) glass column village of Zutphen, suggests that there is an inverse relationship packed with 60/80 CarbopackB/0.8% KOH on 4% Carbowax 20M (Supelco, Bellefonte, PA) and eluted methylamines with helium (25 between fish consumption and mortality from coronary heart ml/min) and a temperature gradient (75°Cfor 2 min, then rising to Received 5/27/86; revised 8/4/86; accepted 8/11/86. 200°Cat32°C/min).Amines were detected with a nitrogen-phosphorus The costs of publication of this article were defrayed in part by the payment detector (NPD; Perkin Elmer), and peaks were integrated using a of page charges. This article must therefore be hereby marked advertisement in computing integrator (LCI-100; Perkin Elmer). During each sample accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' This work was supported by a grant from the NIH (CA-26731). run, we injected 20 M'of 30% ammonium hydroxide onto the column 2To whom requests for reprints should be addressed, at Departments of when the temperature reached 150°C,and the temperature program Pathology and Pediatrics, Boston University School of Medicine, 85 E. Newton was then continued to 200°C.This markedly reduced "ghosting" of Street, Boston, MA 02118. 3The abbreviations used are: DMA, dimethylamine; TMA, trimethylamine; methylamines on subsequent runs. NDMA, nitrosodimethylamine; TMAO, trimethylamine oxide; MMA, mono- This method had a precision (SD/mean x 100) for MMA of 1.6%, methylamine. for DMA of 2.2%, and for TMA of 2.0%. The assay was linear for 6136 Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1986 American Association for Cancer Research. FISH AND METHYLAMINES IN HUMANS

Table 1 Diets ingested during the study 1.6-1 Fish diet Control diet (Day 3 only) 1.2- Breakfast 80 g bagel 80 g bagel 30 g cream cheese 30 g cream cheese 0.8 110 g orange juice 110 g orange juice 230 g whole milk 230 g whole milk 0 0.4- Lunch 300 g cheese pizza 270 g haddock (breaded)" 300 g ginger ale 300 g ginger ale 190 g apple 190 g apple 65 g cookies 65 g cookies Diet 30 g cole slaw c 90 g french fried potato o o U U 284 g cod* Dinner 100 g ham steak Fig. 1. Monomethylamine excretion by humans. Humans (n = 5) ingested a 100 g green beans 5 g butter diet of known methylamine content (Contrat) for 2 days. They then ingested a 100 g sweet potato 100 g rice diet containing fish at lunch and dinner (Fish) for 1 day. On the fourth day they 300 g ginger ale 30 g cole slaw ingested the control diet again. Diets are described in Table 1. Urine was collected 300 g ginger ale in 24-h blocks, and MMA was determined in duplicate samples using a gas 50 g ice cream Chromatographie technique. Data are expressed as mean ±SE. Snacks 600 g ginger ale 600 g ginger ale 70 g cookies 70 g cookies l 25" " Frozen haddock, baked prior to ingestion. * Fresh cod, broiled in butter. * 20-

i O) 15- Table 2 Methylamine-containing foods in diet '£ 10- When appropriate, all foods were cooked for use in the diet as described in Table 1. Data are expressed as the mean of triplicate determinations. 5- Mmol/lOOgfood FoodHam I l l steakHaddockCodMilkCheese Diet

o o U U pizzaGreen Fig. 2. Dimethylamine excretion by humans. Humans (n = 5) ingested control beansMMAND"NDND7.92.94.9DMA6.7197.278.2NDNDNDTMA40.8233.3114.9NDNDNDand fish-containing diets as described in Fig. 1. Urine was collected in 24-h ' ND, not detected. blocks, and DMA was determined in duplicate samples using a gas Chromato graphie technique. Data are expressed as mean ±SE. *, P < 0.05; **, P < 0.01 different from control by repeated measures of analysis of variance and Scheffe's amine concentrations of up to 1600 pmol, with a detection limit of less than 1 pmol. Recovery of added methylamines was 83.3% for MMA, 90.8% for DMA, and 98.7% for TMA. Ammonia, ethylamine, dieth- ylamine, propylamine, isopropylamine, benzylamine, creatinine, carni- 1.6 line, betaine, choline, ethanolamine, pyrrolidine, piperidine, and phos- phatidylcholine did not interfere with the assay when they were present 1.2- at concentrations 10 to 100 times those of the amines of interest. Statistics. Data were analyzed using repeated-measures analysis of 0.8- variance and the Schotte test (12). 0.4-

RESULTS Some of the foods ingested as part of the experimental diets Diet contained MMA (milk, pizza, and beans). Fish and ham were o the only foods which contained DMA and TMA (Table 2). On U Fig. 3. Trimethylamine excretion by humans. Humans (n = 5) ingested control control days, subjects ingested 31.6 ^mol of MMA, 6.7 ¿¿moland fish-containing diets as described in Fig. 1. Urine was collected in 24-h of DMA, and 40.8 ¿¿molofTMA per day. On the day fish was blocks, and TMA was determined in duplicate samples using a gas Chromato graphie technique. Data are expressed as mean ±SE. **,P< 0.01 different from eaten, subjects consumed 18 ^mol of MMA, 754 ¿¿molofDMA, control by repeated measures of analysis of variance and Scheffe's test. and 956 /¿molofTMA per day. MMA excretion in urine was similar on the day when fish was eaten as it was on the initial 2 control days (1.5 ±0.2 fish was eaten, TMA excretion remained elevated, being 5 times versus 1.3 ±0.1 /¿mol/24h/kg of body weight; mean ±SE; no greater than on the control days (0.86 ±0.14 //mol/24 h/kg of significant difference; Fig. 1). DMA excretion was more than 4 body weight; mean ±SE; Fig. 3). times greater on the day that fish was eaten than on the control days (24.1 ±3.5 versus 5.6 ±0.2 ^mol/24 h/kg of body weight; DISCUSSION mean ±SE; P < 0.01; Fig. 2). On the day after fish was eaten, DMA excretion remained elevated, being 2.8 times greater than Ingestion of fish increased exposure to methylamines many- on the control days (15.9 ±2.8 /^mol/24 h/kg of body weight; fold. A 100-g portion offish contained as many methylamines mean ±SE; P < 0.05; Fig. 2). TMA excretion was more than as were normally excreted by humans in a day. Fresh fish 9 times greater on the day that fish was eaten than on the contained fewer methylamines than did frozen fish. Humans control days (1.6 ±0.3 versus 0.17 ±0.03 ^mol/24 h/kg of excreted 866 ^mol of methylamines (MMA + DMA + TMA) body weight; mean ±SE; P < 0.01; Fig. 3). On the day after in their urine during the first 2 days when ingesting the control 6137 Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1986 American Association for Cancer Research. FISH AND METHYLAMINES IN HUMANS diet. Much of this was probably derived from endogenous tial risks should be thoroughly evaluated, so that optimal dietary (nondietary) sources of DMA (1). The increase in excretion of interventions can be designed. total methylamines (1811 ¿¿mol)duringDays 3 and 4 was almost identical to the methylamine content of the fish (1710 ACKNOWLEDGMENTS //mol) that was eaten. Much of the ingested TMA must have been converted to DMA within the human, as most of the We would like to thank Howard Cabrai, Boston University School of Public Health, for his help with the statistical analyses of data. increase in urine methylamine excretion was in the form of DMA. In rats, it has been established that TMA can be metab REFERENCES olized to form DMA (13). The methylamines in the fish must have been absorbed from 1. Zeisel, S. H.. DaCosta, K.. and Fox, J. G. Endogenous formation of dimeth- the intestine into blood in order to be excreted in urine. DMA ylamine. Biochem. J., 232: 403-408, 1985. 2. Bausch. H., and Montesano, R. Relevance of nitrosamines to human cancer. is readily transported from blood to gastric fluid (14). The other Carcinogenesis (Lond.), 5: 1381-1398, 1984. precursor needed for the formation of nitrosamine is nitrite, 3. Committee on Diet, Nutrition, and Cancer, National Research Council. Naturally occurring carcinogens. In: Diet, Nutrition, and Cancer, pp. 234- which is formed from dietary nitrate in the oral cavity and 276. Washington, DC; National Academic Press, 1982. stomach by reactions mediated by bacterial enzymes (15, 16). 4. Lundstrom, R. ( '., and Racicot, L. D. Gas Chromatographie determination People ingest from 10 to several hundred, mg of sodium nitrate of dimethylamine and trimethylamine in seafoods. J. Assoc. Off. Anal. per day in food and water (16-18). The acidic conditions of the Chem., 66: 1158-1163, 1983. 5. Singer, G. M., and Lijinsky, W. Naturally occurring nitrosatable compounds. stomach favor the formation of nitrous anhydride and nitrosyl 1. Secondary amines in foodstuffs. J. Agrie. Food Chem., 24:550-553, 1976. compounds which nitrosate amines to form nitrosamines (19). 6. Jebsen, J. W., and Riaz, M. Breakdown products of trimethylamine oxide in airdried stockfish. Means of enhancing the formation of formaldehyde and NDMA is rapidly formed when DMA and nitrite are mixed dimethylamine. Fish Dir. Skr., Ser. Ernoering., /: 145-153, 1977. with gastric juice in a test tube or in the stomach of the dog 7. Spinelli, J., and Koury, B. J. Some new observations on the pathways of formation of dimethylamine in fish muscle and liver. J. Agrie. Food Chem., (20, 21). The in vivo formation of nitrosamines has clinical 29:327-331, 1981. significance, as the same tumors are produced in experimental 8. Kromhout, D., Bosschieter, E. B., and Coulander, C. L. The inverse relation between fish consumption and 20-year mortality from coronary heart disease. animals that are fed aminéplus nitrite, as are formed in those N. Engl. J. Med., 312: 1205-1209, 1985. fed the corresponding nitrosamine (5). The formation of 9. Bang, H. O., Dyerber, J., and Sinclair, H. M. The composition of the Eskimo NDMA in vivo in humans has not been well documented food in north western Greenland. Am. J. Clin. Nutr., 33:2657-2661, 1980. 10. Philipson, B. E., Rothrock, D. W., Connor, W. E., Harris, W. S., and because this nitrosamine is rapidly metabolized. However, there Illingworth, D. R. Reduction of plasma lipids, lipoproteins, and apoproteins is direct evidence that the nonmetabolizable nitrosamine, nitro- by dietary fish oils in patients with hypertriglyceridemia. N. Engl. J. Med., soproline, is synthesized in humans following the ingestion of 312: 1210-1216, 1985. 11. Glomset, J. A. Fish, fatty acids, and human health. N. Engl. J. Med., 312: an amine (proline) and nitrate (22, 23). Although it appears 1253-1254, 1985. that most of the ingested methylamine was rapidly excreted by 12. Bruning, J. L., and Kim/. B. L. Computational Handbook of Statistics, Ed. 2. Glenview, IL: Scott, Foresman & Company, 1977. humans, it is possible that significant amounts of NDMA could 13. Asatoor, A. M., and Simenhoff, M. L. The origin of urinary dimethylamine. have been formed in vivo, as nitrosamines are of concern at very Biochim. Biophys. Acta, ///: 384-392, 1965. low concentrations (ppm) (5). 14. Zeisel, S. H., DaCosta, K., Edrise, B. M., and Fox, J. G. Transport of dimethylamine, a precursor of nitrosodimethylamine, into stomach of ferret The methylamine content of fish varies with its freshness, and dog. Carcinogenesis (Lond.). 7: 775-778, 1986. storage conditions, and species (5-7). DMA concentrations as 15. Tannenbaum, S. R., Weisman, M., and Fett, D. The effect of nitrate intake high as 1640 ^/mol/100 g (21 times greater than in our fish) on nitrite formation in human saliva. Food Cosmet. Toxicol., 14: 549-552, 1976. have been reported in frozen cod, while fresh spotted trout 16. White, J. W. Relative significance of dietary sources of nitrate and nitrite. J. contained only 15 /umol of DMA per 100 g (5). Fish also Agrie. Food Chem., 23: 886-891, 1975. 17. Kau alum. T., Oshima, H., Uibu, J., Nakamura, M., Matsui, M., and Ha- contains several other nitrosatable amines, such as morpholine mano, M. Occurrence, formation, and precursors of /V-nitroso compounds and di-/V-propylamine; however, these amines are present in in Japanese diet. In: E. C. Miller (ed.). Naturally Occurring Carcinogens— much smaller quantities than is DMA (5). Mutagens and Modulators of Carcinogenesis, pp. 195-209. Baltimore: Uni versity Press, 1979. We have documented a risk associated with increased inges 18. Hill, M. J., Hawksworth, G., and Tattersall, G. Bacteria, nitrosamines, and tion of fish in humans: the exposure to increased amounts of cancer of the stomach. Br. J. Cancer, 28: 562-567, 1973. 19. Mirvish. S. S. Formation of /V-nitroso compounds: chemistry, kinetics, and nitrosatable amines and therefore the possible exposure to in rivo occurrence. Toxicol. Appi. Pharmacol., 31: 325-351, 1975. nitrosamines. The relative importance of this risk, compared 20. Lane, R. P., and Bailey, M. E. The effect of pH on dimethylnitrosamine with the potential for benefits to the cardiovascular system, is formation in human gastric juice. Food Cosmet. Toxicol., //: 851-854,1973. 21. Lintas, C., Clark, A., Fox, J. G., Tannenbaum, S. R., and Newberne, P. M. not known. We known of no data which suggest that the In vivo stability of nitrite and nitrosamine formation in the dog stomach: ingestion of fish is associated with an increased incidence of effect of nitrite and amine concentration and of ascorbic acid. Carcinogenesis (Lond.), 3: 161-165, 1982. cancer. However, there are treatment options which might 22. Oshima, H., and limiseli. H. Quantitative estimation of endogenous nitro- entail less exposure to methylamines. For example, the benefi sation in humans by monitoring ¡V-nitrosoproline excreted in the urine. cial ingredients in fish may consist of certain fatty acids, and Cancer Res., 41: 3658-3662, 1981. 23. Wagner, D. A., Shuker, D. E. G., Hasic, G., and Tannenbaum, S. R. these could be administered in purified form. Before recom Endogenous nitrosoproline synthesis in humans. Nitrosamines and human mendations are made for changes in the American diet, poten cancer. Banbury Rep., 12: 319-333, 1982.

6138 Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1986 American Association for Cancer Research. Increase in Human Exposure to Methylamine Precursors of N -Nitrosamines after Eating Fish

Steven H. Zeisel and Kerry-Ann DaCosta

Cancer Res 1986;46:6136-6138.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/46/12_Part_1/6136

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/46/12_Part_1/6136. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from cancerres.aacrjournals.org on September 23, 2021. © 1986 American Association for Cancer Research.