Gastric Carcinogenesis: 2-Chloro-4 ... - ACS Publications

58

Chem. Res. Toxicol. 1996, 9, 58-66

Gastric Carcinogenesis: 2-Chloro-4-methylthiobutanoic Acid, a Novel Mutagen in Salted, Pickled Sanma Hiraki Fish, or Similarly Treated Methionine† Wei Chen, John H. Weisburger,* Emerich S. Fiala, Thomas E. Spratt, Steven G. Carmella, Di Chen, and Stephen S. Hecht American Health Foundation, Valhalla, New York 10595 Received April 11, 1995X

The customary salting and pickling of fish in high risk gastric cancer regions were modeled to explore the relevant causative chemicals. The fish Sanma hiraki was treated with sodium chloride and sodium nitrite at pH 3. Previously, it had been found that an extract of the treated fish was mutagenic in Salmonella typhimurium TA 1535 without S9 and also that it induced glandular stomach cancer upon gavage to rats. We now demonstrate that the mutagenicity was enhanced by preincubation of the raw meat for several days before salt-nitrite treatment. HPLC techniques showed that three mutagens were present in the fish extract. One of the mutagens was found to be stable over the pH range of 1.0-9.0. This mutagen was purified by silica gel solid phase extraction, followed by a series of reverse phase HPLC steps, and was characterized by low and high resolution MS, NMR, and FT-IR. While N-nitroso compounds were generally believed to be associated with gastric carcinogenesis, it was unexpectedly found that the mutagen has the novel structure 2-chloro-4-methylthiobutanoic acid (CMBA). Based on the structure, it seemed likely that methionine might be the precursor, and this was, indeed, proven. Both salt and nitrite are essential factors for forming this mutagen. The yield of CMBA was linear for chloride concentrations from 0 to 800 mM NaCl. Of 20 amino acids reacted with nitrite and chloride at pH 3, only methionine generated a mutagen for S. typhimurium TA 1535. Tryptophan gave a product mutagenic in S. typhimurium TA 100 and TA 98, but not TA 1535, and in the case of tyrosine, the mutagen was active only for TA 100. These results suggest an important role for salt in gastric carcinogenesis and provide new approaches for exploring the formation of mutagens/carcinogens for specific target organs.

Introduction Gastric cancer remains one of the most common causes of cancer deaths throughout the world, although its incidence is declining, especially in the U.S. Epidemiological studies indicate that a risk of gastric cancer is linked with high consumption of salted, pickled food, or elevated intake of salt and nitrate, the main common components in salted, pickled foods (2-8). Laboratory results and clinical observations revealed that salt influences stages of gastric carcinogenesis, such as causing mucosal damage, inducing atrophic gastritis, and enhancing the action of carcinogens. Salt also augments mucosal damage associated with Helicobacter pylori infection, detected in high risk populations (9-15). Nitrate in water or foods can be reduced to nitrite by the bacterial flora in the oral cavity, or in the stomachs of individuals with salt-induced atrophic gastritis. Nitrite can nitrosate constituents in food to potentially mutagenic/carcinogenic N-nitroso compounds that are genotoxic carcinogens (16-22). However, the chemical relationship between salted, pickled food intake and gastric cancer induction is far from clear. Our approach is based on the concept that gastric carcinogens form during traditional salting and pickling of specific foods and thus † A brief “correspondence” summarizing the results published in detail here has appeared (1). A poster was presented at the 1995 annual meeting of the American Association for Cancer Research, Proceedings 36, 116 (Abstract 688). * To whom correspondence should be addressed. Phone: 914-7897141; Fax: 914-592-6317; E-mail: [email protected]. X Abstract published in Advance ACS Abstracts, November 15, 1995.

0893-228x/96/2709-0058$12.00/0

are present in foods so treated, in distinction to the production of gastric carcinogens in the stomach. We undertook studies to elucidate the chemistry of formation of carcinogens during salting-pickling. Earlier approaches were based on the pioneering finding by Sugimura in 1966 that the nitrosamide N-methyl-N′nitro-N-nitrosoguanidine (MNNG)1 was the chemical of choice for inducing glandular gastric cancer in rats, identical to the human disease (23). MNNG is a direct acting mutagen in the Salmonella typhimurium system of Ames. It alkylates DNA at the N-7and O-6 positions of guanine, as well as at other sites. Since the 1970s, there have been numerous studies on MNNG-induced gastric carcinogenesis (cf. ref 5). Sanma hiraki2 is a fish consumed extensively in Northern Japan, a high risk region for gastric cancer. We demonstrated that an extract of the fish homogenate pickled with a 2% NaCl/73 mM nitrite solution at pH 3 yielded a direct-acting mutagen in S. typhimurium TA 1535 (24). The mutagen formed rapidly over a broad concentration range of nitrite. The fish mutagen sample was stable at pH 5-7 for several days without loss of mutagenicity and so was essentially different from MNNG, which decomposes rapidly at neutral pH. Furthermore, the mutagen could be extracted into NaOH 1 Abbreviations: MNNG, N-methyl-N′-nitro-N-nitrosoguanidine; DMSO, dimethyl sulfoxide; EI-MS, electron impact mass spectrometry, CMBA, 2-chloro-4-methylthiobutanoic acid; HMBA, 2-hydroxy-4methylthiobutanoic acid. 2 Sanma hiraki is a processed, slightly salted fish, also called Pacific mackerel, Scomber japonicus.

© 1996 American Chemical Society

2-Chloro-4-methylthiobutanoic Acid, Fish Mutagen

solution (pH 13) from organic solvents and then returned back to organic solvents after adjusting the aqueous phase to pH 3. Clearly, the mutagen in the salt-nitritetreated fish did not display the characteristics of an N-nitrosamide. Even more importantly, glandular stomach cancer and precancerous lesions were successfully induced in Wistar strain rats by gavage feeding of the extract of the mutagenic product over 6 months, followed by a further 9 month period of observation (25). The mutagenic extract, therefore, contained a gastric carcinogen. Thus, it was important to develop procedures to isolate and identify the fish mutagen. The extent of purification was followed by assays of intermediate fractions for mutagenicity in S. typhimurium TA 1535. Chromatography involved preliminary solid phase extraction, purification through multiple HPLC steps, and, finally, determination of chemical structure by appropriate methods. It was found that three mutagens are actually present in the fish extract. One of these was effectively purified to homogeneity. An unexpected discovery was made that this mutagen is not an N-nitroso compound, but rather a new, not heretofore described chemical, 2-chloro-4-methylthiobutanoic acid (CMBA). The same product was obtained by using methionine (Met) in the reaction sequence, and the concentration of chloride played a key role in its formation. Mutagenic products, similar to that obtained with Met, were not seen with 19 other amino acids treated in the same way, although tryptophan yielded mutagenic activity for S. typhimurium TA 100 and TA 98, and tyrosine for TA 100 only. These experiments and supporting conclusions are presented in this paper.

Chem. Res. Toxicol., Vol. 9, No. 1, 1996 59 Chart 1. Isolation of Mutagen 3 (CMBA) from NaCl-NaNO2-Treated Fish or Met

Experimental Procedures Fish. Sanma hiraki, a lightly salted and semidried frozen Japanese fish, was imported from Shimoya Heijiroh Shooten (2192 Uyematsu-cho, Chohshi-shi, Chiba, Japan), purchased from a local Japanese supermarket, and kept at -20 °C until use. Chemicals. ACS-grade sodium chloride, sodium nitrite, sodium hydroxide, acetic acid, and ammonium sulfamate were purchased from Aldrich (Milwaukee, WI), as was CD3OD (99.8% D). Twenty L-amino acids were acquired from Sigma (St. Louis, MO), among which alanine, asparagine, cysteine, methionine, phenylalanine, proline, threonine, tryptophan, tyrosine, and valine were SigmaUltra grade. HPLC-grade ethyl acetate, methylene chloride, hexane, methanol, and acetonitrile were obtained from Scientific Products/Baxter (Edison, NJ). Concentrated H3PO4 (85%) was from Fisher (Springfield, NJ). 2-Hydroxy-4-methylthiobutanoic acid (HMBA) was a gift of Dr. W. D. Shermer, Monsanto Co. (Chesterfield, MO). Instrumentation. A Savant Model SC100 Speed Vac (Savant Instruments, Inc., Farmingdale, NY) was used to concentrate samples. HPLC was carried out on a Waters Model 600 multisolvent delivery system (Millipore Corp., Milford, MA) with a Hewlett Packard (HP, Avondale, PA) Model 1040 Series II HPLC detection system. UV spectra were determined with a Beckman DU 640 spectrometer (Beckman Instruments, Inc., Fullerton, CA). GC/MS was performed using a HP Model 5971A mass-selective detector, interfaced to a HP Model 5890 Series II gas chromatograph with a HP Model 7673 automatic sampler. NMR spectra were obtained with a Bruker AM360 WB spectrometer using TMS as an internal standard. FT-IR spectra were acquired on a Polaris spectrometer (Mattson Instruments, Inc., Madison, WI). The samples were directly applied to Type 61 disposable IR cards (3M, St. Paul, MN) for analysis, and spectra were acquired with a nitrogen purge through the sample compartment.

Mutagenicity Assays. Direct-acting mutagenic activity (without liver S9) was determined in the American Health Foundation In Vitro Molecular Analysis Facility, using the method described by Maron and Ames (26) and Levin et al. (27), in S. typhimurium strains TA 1535, TA 100, TA 98, and TA 102. Samples for the Ames tests were concentrated to dryness and redissolved in H2O with 10-20% DMSO, except for the samples tested for stability as a function of pH, which were redissolved in small volumes of different pH buffers. For each test, the assays were done in triplicate at two concentrations of substrate. Preparation of Fish Extract. Sanma hiraki fillet was minced and incubated at room temperature (23 °C) for 0-7 days (Chart 1). The preincubated fish muscle was homogenized with twice the volume (w/v) of deionized H2O in a blender, and then solid NaCl was added to a concentration of 2%. After stirring for 1 h at room temperature, the homogenate was centrifuged at 21000g for 30 min. The sediment was discarded, and the supernatant was adjusted to pH 3 by adding 12 N HCl. The volume was reduced to one-fourth in vacuo on a rotary evaporator. Following another centrifugation under the same conditions, the clear supernatant was incubated with NaNO2 (5 mg/g fish) at 23 °C in the dark for 1 h, and the pH was kept at 3 with 12 N HCl. The reaction was terminated by adding an equimolar amount of ammonium sulfamate. A three step extraction was then carried out. First, the solution was extracted three times with equal volumes of ethyl acetate. Second, the combined ethyl acetate fractions were extracted twice with 0.25 volume of 0.4 N NaOH, and the aqueous layers were quickly readjusted to pH 3 with 12 N HCl.

60

Chem. Res. Toxicol., Vol. 9, No. 1, 1996

Third, the aqueous fraction was extracted three times with 0.8 volume of methylene chloride. The collected extracts were evaporated to dryness, and the residue, dissolved in a small volume of MeOH (20%)-H2O water (80%), was used for further study. Controls were done (1) with 100 mL of supernatant obtained in the same way, but without nitrite, and (2) with 100 mL of H2O, instead of fish supernatant, treated with salt and nitrite at pH 3. Treatment of Methionine. NaCl was added to 10 mL of Met (1 mmol) in H2O to give a concentration of 2%. The solution was incubated with 0.73 mmol of NaNO2 at 23 °C for 1 h after adjustment to pH 3 with 0.5 N HCl. Excess nitrite was quenched by the addition of an equimolar amount of ammonium sulfamate. The mixture was extracted three times with equal volumes of ethyl acetate, and the combined organic phases were evaporated in vacuo. The residue was dissolved in MeOH and diluted with H2O to give a 1:4 (v/v) MeOH-H2O solution for determination of mutagenicity, or purification by HPLC. Three different controls, (1) without Met, (2) without NaCl and with H2SO4 replacing HCl, and (3) without NaNO2, were prepared, using the same procedure. Chromatographic Separation and Purification. The fish extract was separated into crude mutagenic fractions using solid phase silica gel extraction. After the extract was dissolved in a small volume of hexane-CH2Cl2 (1:2), it was loaded onto a 900 mg silica gel Maxi-Clean cartridge (Alltech 20988). The cartridge was eluted sequentially with a series of organic solvents: (1) 100% hexane; and then hexane-CH2Cl2-ethyl acetate mixtures: (2) 70:25:5; (3) 60:30:10; and (4) 50:30:20. Each fraction was evaporated to dryness, and the residues were dissolved in MeOH-H2O (1:9) for mutagenicity assays and further separation by HPLC. A three-step reverse phase HPLC was developed to isolate and purify the mutagen from the crude fraction. For step 1, a 10 × 250 mm Ultrasphere ODS column with a 4.6 × 45 mm Ultrasphere ODS guard column was employed. The mobile phase A was 0.34 mM H3PO4 in 10% MeOH, and B was 0.5 mM H3PO4 in 100% MeOH at a flow rate of 1.5 mL/min. The gradient program was an initial 8 min at 100% A, then 30 min to 50% B (linear), and a further 50 min to 100% B. For step 2, a 4.6 × 250 mm Ultrasphere ODS column with a 4.6 × 45 mm same type guard column was used. The eluant was H2O (A)acetonitrile (B) at a flow rate of 1 mL/min, using a linear gradient from 1% B to 50% B in 25 min, and to 95% B in 35 min. For step 3, the same columns were used as step 2, but the mobile phase was 0.1% acetic acid in 10% MeOH (pH 3.5) (A)-acetonitrile (B). The gradient was linear from 0% to 50% B over 20 min at a flow rate of 1 mL/min. The fractions eluting from each step of the chromatographic procedures were collected, were evaporated to dryness, and were used for mutagenicity assays, further purification, or instrumental analysis. Except for the silica gel solid phase extraction, which was not required in this instance, the same HPLC methods were applied for purification of the mutagen from NaCl-NaNO2treated Met. MS Analysis. For GC/MS, the purified mutagenic compound was dissolved in acetonitrile. A HP Ultra-2 capillary column (5% phenylmethylsiloxane), 12 m × 0.2 mm i.d., 0.11-µm film thickness, was used. The injector temperature was 175 °C, the detector temperature was 280 °C, and helium was the carrier gas. The initial temperature was 40 °C for 3 min, then increased to 200 °C at 10 min at a rate of 30 °C/min. The spectral properties of mutagen 3 (F3) from fish or Met were as follows: UV (H2O), λmax less than 190 nm, 205 (sh); 1HNMR (CD3OD) δ 4.57 (d/d, 1H, 2-CH, J ) 2.0, 1.2 Hz), 2.68 (m, 2H, 4-CH2), 2.34 (m, 1H, 3-CH), 2.20 (m, 1H, 3-CH), 2.17 (s, 1H, CH3); 1H-NMR (CD3OD + NaOD) δ 4.37 (d/d, 1H, 2-CH), 2.67 (m, 2H, 4-CH2), 2.30 (m, 1H, 3-CH), 2.15 (m, 1H, 3-CH), 2.13 (s, 1H, CH3); 13C-NMR (CD3OD) δ 173.0 (s, CO), 57.5 (d, J ) 156 Hz), 31.2 (t, J ) 137 Hz), 35.5 (t, J ) 131 Hz), 15.0 (q, J ) 138 Hz); low resolution GC/EI-MS, 6.83 min, M, m/z 168 and M + 2, m/z 170 (33.3% of M); high resolution FT-MS, calcd for C5H9ClO2S 168.000628, found 167.996597.

Chen et al. Effect of NaCl Concentration in the Met-NaCl-NaNO2 Reaction. Different amounts of NaCl were added to 0.5 mmol of Met in 10 mL of H2O to final concentrations of 0, 12.5, 50, 100, 200, 400, 800, or 1600 mM. At room temperature, each solution, adjusted to pH 3 with 0.5 N H2SO4, was reacted with 0.5 mmol of NaNO2. After 1 h, the reaction was stopped with 0.5 mmol of ammonium sulfamate. The solutions were extracted three times with equal volumes of ethyl acetate. After removal of the organic solvent, the residue was dissolved in 500 µL of 0.3 mM H3PO4-20% MeOH for analysis. Every experiment was done in duplicate. Quantitation of CMBA Produced from the Reaction of Met with Nitrite and Different Concentrations of NaCl. Reverse phase HPLC employing an Ultrasphere ODS 5 µm 4.6 × 250 mm column with a 4.6 × 45 mm guard column was used to quantitate CMBA. The mobile phase A was 0.85 mM H3PO4 in 10% MeOH (pH 3); B was 0.85 mM H3PO4 in 100% MeOH, and C was acetonitrile. A linear gradient program started from 100% A to 50% B in 20 min and reached 60% B and 30% C in 25 more min. The homogeneity of each CMBA sample was determined by monitoring UV spectra with diode array detection and by MS. Standard curves of peak area, as a function of µmol of CMBA, as well as of HMBA were prepared. Also, a synthetic sample of CMBA, obtained by three-step synthesis, beginning with HMBA (DL-methionine hydroxy analog, Sigma M-0126) (esterification to benzyl ester, conversion to CMBA benzyl ester with thionyl chloride, and hydrolysis of the ester), was available through the courtesy of Drs. S. Amin and Dr. J. Kreminski, American Health Foundation Organic Synthesis Facility. Effect of Salt-Nitrite at pH 3 on Amino Acids. Solutions of 1 mmol each of Ala, Arg, Asn, Asp, Cys, Gln, Glu, Gly, His, Ile, Leu, Lys, Met, Phe, Pro, Ser, Thr, or Val, or of 0.4 mmol of Trp, were prepared in 10 mL of H2O. Because of limited solubility, 0.5 mmol Tyr was dissolved in 50 mL H2O. NaCl was added to each of the solutions to give a concentration of 2% NaCl. These solutions were incubated at 23 °C with 0.73 mmol of NaNO2 at pH 3 in dark for 1 h, and an equal concentration of ammonium sulfamate was added. The treated solutions were then extracted with 3 equal volumes of ethyl acetate. The organic solvent was evaporated, and the residue of each sample was dissolved in 200 µL of MeOH and diluted with H2O for determination of mutagenicity.

Results Direct-Acting Mutagenicity. The extract of Sanma hiraki treated with salt-nitrite displayed mutagenic activity in S. typhimurium TA 1535 without S9. Deletion of either fish or nitrite from the incubation, with analogous workup, failed to yield mutagenic activity. The mutagenicity was not significantly affected by exposure to light or by incubation of the treated fish extract at 50 °C. The mutagenicity was further enhanced by preincubation of the fish fillet at room temperature for 1-7 days, with a maximum activity observed at day 2 (Figure 1). Extracts of Met that had been reacted with salt and nitrite under identical conditions also yielded mutagenicity. Mutagenicity was not found when Met, NaCl or NaNO2 were omitted (Table 1). Preliminary Separation of Mutagens. The initial chromatography of the fish extract resulted in three separately eluting fractions containing mutagenic activity (Figure 2A), suggesting the presence of three individual mutagens, designated mutagen 1, 2, and 3 (F1, F2, and F3). Mutagens 1 and 2 were relatively polar, eluting between 23 and 30 min. Mutagen 3 was relatively less polar, eluting between 46 and 47 min and containing about 26% of the total mutagenicity. Mutagen 3 could also be separated from the other two by adsorption onto a silica solid phase extraction cartridge, from which it


Figure 1. Mutagenicity in S. typhimurium TA 1535 as a function of the amount of the extracts from salt-nitrite-treated fish fillets (g) that had been preincubated at 23 °C for various times. Each treatment was performed in duplicate. Some of the highest levels of fish that had been preincubated for 2 and 7 days were toxic. Error bars show standard deviation of 4 determinations, and * indicates significant differences between 1-7 days and 0 day preincubation. Table 1. Direct-Acting Mutagenicity in S. typhimurium TA 1535 Induced by NaCl-NaNO2-Treated Fish Extract or Met and the Effect of Temperature and Light on the Mutagenicity of the Fish Extract rev/plate ( SD treatment H2O + Cl- + NO2fish + Clfish + Cl- + NO223 °C + darkc 50 °C + dark 4 °C + light Met + ClMet + NO2Met + Cl- + NO2-

time (min)

1 ga or 8 µmolb

2 ga or 16 µmolb

150 30 150

37 ( 2 27 ( 4 360 ( 16 362 ( 35 354 ( 6 379 ( 4 26 ( 2 39 ( 3 297 ( 11

25 ( 4 32 ( 2 520 ( 18 503 ( 18 505 ( 6 534 ( 10 34 ( 8 44 ( 3 462 ( 17

Chem. Res. Toxicol., Vol. 9, No. 1, 1996 61

Figure 2. Reverse phase HPLC of the extracts of NaCl-NaNO2 -treated fish on a 10 × 250 mm Ultrasphere ODS column. The upper curve (-‚-) shows the solvent gradient program used; 0.34 mM H3PO4 in 10% MeOH (pH 3.4) was mobile phase A and 0.34 mM H3PO4 in 100% MeOH was mobile phase B at a flow rate of 1.5 mL/min. Three mutagenic fractions (F1, F2, and F3) were detected in the crude fish extract (boxed area, A). The fraction containing mutagen 3 (F3), eluting from 47 to 48 min, was separated on a silica gel solid phase cartridge followed by HPLC (B). Part C shows that the mutagenic compound from the extract of NaCl-NaNO2-treated Met eluted with the same retention time as F3.

a Extract equivalent to wet weight grams of fish fillets. b Extract equivalent to amounts of Met. c Samples of extracts from saltnitrite-treated fish were maintained as shown for the times listed.

could be eluted with a mixture of 60% hexane, 30% CH2Cl2, and 10% ethyl acetate. Isolation and Purification of Mutagen 3. After preliminary separation of mutagen 3 by solid phase extraction, followed by the step-1 HPLC system, mutagenicity was contained in a fraction eluting at 47-48 min (Figure 2B). This fraction was collected, concentrated, and reinjected into the step-2 HPLC system. Upon elution with H2O-acetonitrile, a peak eluting at 26 min contained most of the mutagenic activity (Figure 3A). Since this compound had weak absorption at 254 nm, the detecting wavelength was adjusted to 220 nm. The final chromatography yielded a single mutagenic peak eluting at 16 min (Figure 4A). Formation of Mutagen 3 from Met. The reaction of salt-nitrite at pH 3 with Met gave only a single mutagenic peak with an HPLC mobility identical to that of mutagen 3 from the treated fish extract (Figures 2C, 3B, and 4B). Stability of Mutagen 3 at Different pH Values and Storage Time. After storage for 60 min over a broad

Figure 3. Step-2 HPLC. (A) After the mutagenic fraction at 47-48 min from the fish extract was concentrated to a small volume, it was further separated by a 4.6 × 250 mm Ultrasphere ODS column with H2O (phase A) and acetonitrile (phase B) at a flow rate of 1.0 mL/min. The fraction eluting from 25 to 26 min contained most of the mutagenic activity and was collected for further HPLC purification. (B) The Met-derived product displays an identical elution pattern.

pH range (1-9), purified mutagen 3 showed no significant change in its mutagenic activity. It was stable in phosphoric acid solution, pH 3, for 26 days. After 48 days, the mutagen retained more than 60% of its activity (Figure 5). Sensitivity of Different S. typhimurium Strains to Mutagen 3. Purified mutagen 3 exhibited a higher

62


Figure 4. Final purification step (base line adjusted). The same column was used as in step 2, and the mobile phases were 0.1% acetic acid in 10% MeOH (A) and acetonitrile (B), with a 1 mL/ min flow rate. The main peak, eluting at 17 min and showing mutagenicity in S. typhimurium TA 1535, was collected and evaporated to dryness for spectral analysis. (A) Compound from fish and (B) compound from Met.

Chen et al.

Figure 6. Mutagenicity in S. typhimurium TA 1535 and TA 100 induced by purified mutagen 3.

Figure 5. Mutagenicity induced by mutagen 3 as a function of pH or storage time. Determination of the effects of storage time was done with the sample kept in H3PO4 solution, pH 3.4.

mutagenic activity in the base-substitution strain S. typhimurium TA 1535 than in TA 100 at equal test levels (Figure 6). Identification of Mutagen 3. The UV spectrum of mutagen 3 (F3) in H2O showed a shoulder at 205 nm ( 1600) with a λmax of less than 190 nm, indicating absence of aromatic or conjugated structure. Analysis by GC/EIMS showed a major peak eluting at 6.83 min with M+ m/z 168 and 170, the latter at one-third intensity, suggesting the presence of Cl (Figure 7A). High resolution FT-MS gave a molecular weight of 167.996597, with the possible formula C5H9ClO2S (calculated mol wt 168.000628). The FT-IR spectrum showed a strong absorption band at 1726 cm-1, providing evidence for the presence of a carbonyl group. The position of this band is consistent with that expected of an R-chloro-substituted carboxylic acid. The 1H-NMR spectrum of mutagen 3 was similar to that of Met, with a methyl resonance at 2.13 ppm, a methine proton at 4.57 ppm, two protons at 2.68 ppm, and individual protons at 2.34 and 2.20 ppm (Table 2, Figure 8B). Decoupling and COSY experiments showed

Figure 7. GC/EI-MS analysis of the mutagen, CMBA, purified from treated fish (A) or Met (B). The relative abundance of fragment M + 2, m/z 170, is 35% of that of M, m/z 168, providing evidence for the presence of a chlorine atom. Table 2. NMR Data for CMBA Obtained from NaCl-NaNO2-Treated Fish or L-Met δ 1Hb δ 1Ha (CD3OD + NaOD) (CD3OD) CO 2-CH,d 1H, d/d 3-CH, 1H, m 3-CH, 1H, m 4-CH2, 2H, m CH3, 1H, s

4.57 2.34 2.20 2.68 2.17

4.37 2.30 2.15 2.67 2.13

δ 13Cc (CD3OD) 173.0 s 57.5 d (J ) 156 Hz) 31.2 t (J ) 137 Hz) 35.5 t (J ) 131 Hz) 15.0 q (J ) 138 Hz)

a 1H-NMR spectra of CMBA obtained from either treated fish or Met. b This spectrum of CMBA was obtained from treated fish. c 13C-NMR spectrum of CMBA from Met. d J ) 2.0, 1.2 Hz.

that the signals at 2.34 and 2.20 ppm were coupled to each other and to the signals at 4.57 and 2.68 ppm. The coupling pattern of the 1H coupled 13C-NMR spectra



Figure 9. Typical reverse phase HPLC (see text for method) of standard CMBA (A) and Met treated with salt-nitrite (B), giving CMBA and HMBA.

Figure 8. 1H-NMR spectrum of CMBA (B: from treated fish, C: from Met). Part A shows that the signal at 4.57 ppm was shifted by adding NaOD, which indicates that the chlorine is in the R-position to the carboxylic group (see text).

confirmed that there was one CH, two CH2, and one CH3 groups. The chemical shift of the methyl group in the 1H-NMR spectra suggested that it could be attached to S or to a carbonyl. FT-IR spectroscopy and NMR spectrometry with NaOD, as described below, indicate that the carbonyl in the compound is a carboxylic acid. The above evidence is consistent with a structure of XYCHCH2CH2Z, in which X,Y, and Z are COOH, Cl, and CH3S. To identify the position of the carboxylate, the 1H-NMR spectrum in CD3OD was compared with the one obtained upon NaOD addition. It would be expected that the protons closest to the COOH group show the largest decrease in chemical shift. The signal at 4.75 ppm, which was assigned to the methine proton, was shifted upfield 0.2 ppm, while the shifts of the other signals were less than 0.05 ppm (Figure 8A). Therefore, X (or Y) was determined to be a carboxy group. The chemical shift of the terminal methylene is consistent with Z as the thiomethyl but not the chloro moiety. Similarly, the chemical shift of the methine proton is consistent with carboxylate and chloro but not carboxylate and thiomethyl substituents. Evaluation of the combined analytical data indicated that mutagen 3 is 2-chloro-4methylthiobutanoic acid (CMBA), a new compound. Mutagen 3 from Met. In all of the analytical systems used, the Met-derived mutagen displayed the same characteristics as mutagen 3 from fish. GC/EI-MS data were essentially identical (Figure 7B), as were the 1HNMR data (Figure 8C). FT-IR, a strong absorption band at 1726 cm-1, was characteristic of a CdO group. Thus, these data entirely support the structure of mutagen 3 as CMBA, produced independently from Met. Formation of CMBA from Met Treated with Nitrite and Different Concentrations of Sodium Chloride. Under the HPLC conditions described, standard CMBA eluted at 27 min as a single peak (Figure 9A). A linear relationship (r ) 0.996) was established for peak area versus µmol of CMBA injected. Using the same HPLC system, CMBA from Met treated with nitrite and different concentrations of NaCl eluted at the same time as the standard (Figure 9B). This peak was absent when Met was treated only with nitrite (no NaCl). Under those conditions, HMBA was formed, which eluted as a

Figure 10. Effect of NaCl concentration on CMBA formation, quantitated by reverse phase HPLC. Data ( SD with 3-5 assays.

single peak at 14 min (B, Figure 9). This elution pattern was identical to that of an authentic sample of HMBA obtained from Dr. W. D. Shermer. The structure was confirmed by GC/EI-MS and 1H-NMR. The yield of CMBA was directly proportional to the concentration of sodium chloride, up to 800 mM, but it leveled off when excess chloride (1600 mM) was added (Figure 10). The quantity of the HMBA decreased with increasing concentrations of NaCl. In addition to CMBA and the HMBA, HPLC of Met treated with NaCl-NaNO2 showed other minor products of unknown structure, but these had no mutagenicity. Mutagenicity of 20 Amino Acids Treated with NaCl-NaNO2 at pH 3. Following reaction with salt and nitrite at pH 3, only Met yielded direct-acting mutagenicity in S. typhimurium TA 1535. The NaCl-NaNO2treated Met also showed direct-acting mutagenicity in TA 100, but the response was lower compared to that obtained with TA 1535. Mutagenicity was not evident in S. typhimurium strains TA 102 or TA 98. A doseresponse of mutagenicity in TA 100 and TA 98 was observed only with NaCl-NaNO2-treated Trp and was positive only in TA 100 with Tyr. None of the other amino acids tested gave significant mutagenic products (Tables 3 and 4).

Discussion For decades, N-nitroso compounds have drawn most attention as the carcinogens associated with gastric cancer (5, 16-23). Indeed, such chemicals are broadly acting carcinogens and are among the most potent animal carcinogens. Since many nonmutagenic/noncarcinogenic precursor compounds become mutagenic/carcinogenic

64


Chen et al.

Table 3. Direct-Acting Mutagenicity in S. typhimurium TA 1535, TA 102, TA 100, and TA 98 Induced by Amino Acids Treated with Salt-Nitrite, pH 3.0 rev/platea TA 1535 sample DMSOb controlsc Met + Cl- + NO2Phe + Cl- + NO2Pro + Cl- + NO2Ser + Cl- + NO2Trp + Cl- + NO2Tyr + Cl- + NO2-

8 µmol

TA 102

16 µmol

8 µmol

26 ( 7 1058 ( 116 297 ( 11 462 ( 17 24 ( 3 43 ( 12 41 ( 14 36 ( 7 48 ( 13 54 ( 11 38 ( 6 50 ( 10 28 ( 6 25 ( 6

TA 100

16 µmol

372 ( 34 1111 ( 158 389 ( 11 381 ( 16 426 ( 2 100 ( 23 321 ( 8 349 ( 17 520 ( 7 584 ( 27 539 ( 9 660 ( 13 225 ( 66 26 ( 1d

8 µmol

16 µmol

128 ( 7 882 ( 4 310 ( 8 358 ( 6 198 ( 21 31 ( 10 145 ( 8 146 ( 4 188 ( 3 141 ( 5 347 ( 1 492 ( 6 263 ( 28 21 ( 1d

TA 98 8 µmol

16 µmol

32 ( 7 391 ( 41 34 ( 6 35 ( 1 40 ( 2 39 ( 1 43 ( 1 43 ( 12 39 ( 3 32 ( 1 120 ( 28 74 ( 6 53 ( 7 22 ( 4

a Every sample was tested in triplicate at 8 and 16 µmol of amino acid, except Trp and Tyr, which were 4 and 8 µmol. b DMSO, 100 µL/plate. c Controls: Sodium azide, 5 µg/plate, for TA 1535 and TA 100; cumene hydroperoxide, 150 µg/plate, for TA 102; 2-nitrofluorene, 5 µg/plate, for TA 98. d Toxic.

Table 4. Direct-Acting Mutagenicity in S. typhimurium TA 1535 of Amino Acids Treated with Salt-Nitrite at pH 3.0 rev/platea sample

8 mmol

DMSOb NaN3b Ala + Cl- + NO2Arg + Cl- + NO2Asn + Cl- + NO2Asp + Cl- + NO2Cys + Cl- + NO2Gln + Cl- + NO2Glu + Cl- + NO2Gly + Cl- + NO2His + Cl- + NO2Ile + Cl- + NO2Leu + Cl- + NO2Lys + Cl- + NO2Thr + Cl- + NO2Val + Cl- + NO2-

31 ( 4 40 ( 2 35 ( 3 29 ( 3 43 ( 2 28 ( 1 21 ( 4 27 ( 7 23 ( 1 25 ( 1 30 ( 4 29 ( 1

16 mmol 26 ( 7 1058 ( 116 25 ( 5 29 ( 1 30 ( 1 24 ( 1 28 ( 1 33 ( 5 42 ( 1 26 ( 1 26 ( 2 22 ( 3 40 ( 5 23 ( 1 40 ( 5 26 ( 4

a Every sample was tested in duplicate at 8 and 16 µmol of treated amino acid. b DMSO solvent (100 µL/plate) and positive control NaN3 (5 µg/plate).

upon treatment with nitrite at low pH, it has been hypothesized that gastric cancer in humans may result from the exposure to N-nitroso compounds formed when nitrite reacts with specific substrates in food during traditional methods of preservation with salt and nitrite (5, 22). Nitrosation, therefore, is extensively used as a method which imitates the traditional food preservation process for generating mutagenic/carcinogenic compounds. Through this reaction, some constituents in different food mixtures were identified as precursors to potential mutagens/carcinogens. 4-Chloro-6-methoxyindole in fava beans yielded 4-chloro-6-methoxy-2-hydroxy1-nitrosoindolin-3-one oxime, a direct-acting mutagen to S. typhimurium TM 677 (28). Indole-3-acetonitrile from Chinese cabbage formed a mutagen to S. typhimurium TA 100 and TA 98 without S9, 1-nitrosoindole-3-acetonitrile (29, 30). Thus, N-nitroso compounds have been considered as the most probable candidates for gastric carcinogens. The identification of the mutagen in Sanma hiraki extract as CMBA opens a new, distinct area of investigation in gastric cancer and markers associated therewith. Basically, two levels of nitrite are generally used in studies on factors bearing on gastric cancer. One is nitrite in the micromolar range, simulating “realistic” conditions in the gastric environment with formation of the carcinogen in the stomach (18). In contrast, nitrite in the millimolar range, used in the present approach,

is designed to reproduce traditional pickling methods (22, 24, 25). Recipes reviewed by Binkerd and Kolari (22) and by Lauer (31) called for the use as much as 7.5% crude salt and as much as 2% crude sodium nitrate/sodium nitrite. The separation of mutagens in the salt-nitrite-treated fish extract yielded three fractions with mutagenicity in the first HPLC step. The first two mutagenic compounds, designated mutagen 1 (F1) and mutagen 2 (F2), were present in the earlier-eluting more polar fractions, and mutagen 3 (F3) was contained in a later-eluting fraction. This paper is the first detailed report of a unique kind of mutagen, F3, formed under laboratory conditions of traditional “salting-pickling” of food. The procedure reproduced the usual custom in the Far East of storing fish, usually with crude salt, at room temperature for a period of time to enhance its flavor. This technique increased the mutagenicity significantly when the raw fish fillets were kept at 23 °C for 1-7 days. Based on the chemical structure of mutagen 3, CMBA, it was likely that Met derived from the fish protein is the precursor. One of the possible causes of the increased mutagenicity on standing is that more free precursors such as Met were released by the hydrolysis of protein, a reaction possibly affected by the salt added to the fish as part of the preservation process (32). Thus, the role of Met in forming the mutagen was, indeed, confirmed. The reaction of nitrite with the aliphatic amino group is expected to generate a reactive diazonium ion. In the presence of H2O, this ion would yield the corresponding hydroxy acid, but in the presence of sufficient amounts of chloride ion, the R-chlorocarboxylic acid is formed:

Clearly, Cl- is essential for the production of the mutagen, and the formation of CMBA was linear for Clfrom 0 to 800 mM. The effect of salt on gastric carcinogenesis has been studied in terms of a potent enhancing


action through several mechanisms. The mutagenicity of nitrosated black beans was nearly doubled by adding salt to this staple food in Costa Rica, a region with a high gastric cancer rate (33). Our results provide strong evidence of a chemical relationship between NaCl and nitrite under acidic pickling conditions which thus influences gastric carcinogenesis. Wang et al. (34), in an abstract, reported the formation of reactive products, suggested to be diazo compounds, that could interact with 3,4-dichlorothiophenol, upon treating amino acids with nitrite at low pH. Of the 20 amino acids studied in this research, only Met yielded a mutagen in the Ames tester strain S. typhimurium TA 1535. The product, CMBA, was more active in this strain than in strain TA 100. It displayed no activity in the frame shift sensitive strain S. typhimurium TA 98 and also was negative in a strain sensitive to active oxygen compounds, TA 102 (27). It was noted that the product of the reaction of nitrite and NaCl with Trp and Tyr displayed activity in S. typhimurium strains TA 100 and TA 98, but not in strain TA 1535. Thus, the products from Trp and Tyr have different mutagenic attributes from those of CMBA. In previous research, nitrosoindoles or phenolic diazotates were identified, but sodium chloride was not used (3537). In summary, we have developed a systematic approach to investigate the direct-acting mutagens to S. typhimurium TA 1535, and likely gastric carcinogens in rats, from an extract of salt-nitrite-treated Sanma hiraki fish. Instead of the expected N-nitroso compound, a novel mutagen, CMBA, was isolated and identified. According to the chemical structure analysis, Met from the fish protein is proposed as its precursor, and the identical product was, indeed, obtained from Met. Importantly, both nitrite and Cl- are essential to its formation during the salting-pickling process. The formation of a mutagenic R-chloro acid as a product appears to be unique to Met. Yet, the simplest analog, R-chloroacetic acid, is not mutagenic or carcinogenic (38). Thus, the entire CMBA molecule is required for direct-acting mutagenicity and probable gastric carcinogenicity. Dr. C. Furihata, who pioneered effective in vivo rapid bioassays for gastric carcinogens (39), has observed a positive effect with CMBA. Of the 20 amino acids studied, no other amino acids yielded any mutagenic products to the Ames tester strain S. typhimurium TA 1535. The carcinogenic and genotoxic effect of CMBA is being studying in this laboratory.

Acknowledgment. This research was supported, in part, by U.S. Public Health Service Grants CA-29602 and CA-47520, and Cancer Center Support Grant CA-17613, from the National Cancer Institute. In the past, visiting scientists in our laboratory, Dr. Hildegard Marquardt, now in Hamburg, Germany, Dr. Howard Mower, University of Hawaii, on sabbatical leave, and Ms. Graczina Prokopcyzk, had been active in this research. We are most grateful for the effective contributions of R. Sodum, S. Coleman, D. Pullo, J.-M. Lin, F.-Q Lou, and B. McKinney for expert and dedicated assistance in various aspects of this investigation. We thank Dr. Sanford L. Shew, Millipore Co., for performing high resolution MS, James E. Gagnon, 3M (St. Paul, MN), for providing Disposable IR cards and information on their use, and Dr. William D. Shermer for supplying a sample of HMBA.


References (1) Chen, W., Weisburger, J. H., Fiala, E. S., Carmella, S. G., Chen, D., Spratt, T. E., and Hecht, S. S. (1995) Unexpected mutagen in fish. Nature 374, 599. (2) Boeing, H. (1991) Epidemiological research in stomach cancer: progress over the last ten years. J. Cancer Res. Clin. Oncol. 117, 133-143. (3) Correa, P. (1992) Human gastric carcinogenesis: a multistep and multifactorial process. First American Cancer Society Award Lecture on Cancer Epidemiology and Prevention. Cancer Res. 52, 6735-6740. (4) Parkin, D. M., Pisani, P., and Ferlay, J. (1993) Estimates of the worldwide incidence of eighteen major cancers in 1985. Int. J. Cancer 54, 594-606. (5) O’Neill, I. K., Chen, J., and Bartsch, H., Eds. (1991) Relevance to human cancer of N-nitroso compounds, tobacco and mycotoxins. IARC Scientific Publications No. 105, International Agency for Research on Cancer, Lyon, France. (6) Howson, C. P., Hiyama, T., and Wynder E. L. (1986) The decline in gastric cancer: Epidemiology of an unplanned triumph. Epidemiol. Rev. 8, 1-27. (7) Kato, I., Tominaga, S., Ito, Y., Kobayashi, S., Yoshii, Y., Matsuura, A., Kameya, A., Kano, T., and Ikari, A. (1992) A prospective study of atrophic gastritis and stomach cancer risk. Jpn. J. Cancer Res. 83, 1137-1142. (8) Kneller, R. W., Guo, W.-D., Hsing, A. W., Chen, J.-S., Blot, W. J., Li J.-Y., Forman, D., and Fraumeni J. F. Jr. (1992) Risk factors for stomach cancer in sixty-five Chinese Counties. Cancer Epidemiol., Biomarkers Prev. 1, 113-118. (9) Hwang, H., Dwyer, J., and Russell, R. M. (1994) Diet, Helicobacter pylori infection, food preservation and gastric cancer risk: are there new roles for preventative factors? Nutr. Rev. 52, 75-83. (10) Hirono, I., Funahashi, M., Kaneko, C., Ogino, H., Ito, M., and Yoshida, A. (1990) Gastric lesions in rats fed salted food materials commonly eaten by Japanese. Nutr. Cancer 14, 127-132. (11) Honjo, S., Kono, S., and Yamaguchi, M. (1994) Salt and geographic variation in stomach cancer mortality in Japan. Cancer, Causes Control 5, 285-286. (12) Kokubo, M. T., Furukawa, F., Kurokawa, Y., Tatematsu, M., and Hayashi, Y. (1983) Effect of high salt diet on rat gastric carcinogenesis induced by N-methyl-N′-nitro-N-nitrosoguanidine. Gann 74, 28-34. (13) Watanabe, H., Takahashi, T., Okamoto, T., Ogundigie, P., and Ito, A. (1992) Effects of sodium chloride and ethanol on stomach tumorigenesis in ACI rats treated with N-methyl-N′-nitroN-nitrosoguanidine: A quantitative morphometric approach. Jpn. J. Cancer Res. 83, 588-593. (14) Mun˜oz, N. (1994) Is Helicobacter pylori a cause of gastric cancer? An appraisal of the seroepidemiological evidence. Cancer Epidemiol., Biomarkers Prev. 3, 445-451. (15) Tsugane, S., Tei, Y., Takahashi, T., Watanabe, S., and Sugano, K. (1994) Salty food intake and risk of Helicobacter pylori infection. Jpn. J. Cancer Res. 85, 474-478. (16) Forman, D., and Shuker, D. E. G., Eds. (1989) Nitrate, nitrite and nitroso compounds in human cancer. Cancer Surv. 8, 205490. (17) Mirvish, S. S. (1994) Experimental evidence for inhibition of N-nitroso compound formation as a factor in the negative correlation between vitamin C consumption and the incidence of certain cancers. Cancer Res. 54, 1948-1951. (18) Correa, P., Haenszel, W., Cuello, C., Tannenbaum, S., and Archer, M. (1975) A model for gastric cancer epidemiology. Lancet 2, 5860. (19) Wishnok, J. S., Tannenbaum, S. R., Stillwell, W. G., Glogowski, J. A., and Leaf, C. D. (1993) Urinary markers for exposures to alkylating or nitrosating agents. Environ. Health Perspect. 99, 155-159. (20) Bos, P. M. J., Van Den Brandt, P. A., Wedel, M., and Ockhuizen, T. H. (1988) The reproducibility of the conversion of nitrate to nitrite in human saliva after a nitrate load. Food Cosmet. Toxicol. 26, 93-97. (21) Spiegelhalder, B., Eisenbrand, G., and Preussmann, R. (1976) Influence of dietary nitrate on nitrite content of human saliva: possible relevance to In Vivo formation of N-nitroso compounds. Food Cosmet. Toxicol. 14, 545-548. (22) Binkerd, E. F., and Kolari, O. E. (1975) The history and use of nitrate and nitrite in the curing of meat. Food Cosmet. Toxicol. 13, 655-661. (23) Sugimura, T., and Kawachi, T. (1973) Experimental stomach cancer. Methods Cancer Res. 7, 245-308. (24) Marquardt, H., Rufino, F., and Weisburger, J. H. (1977) On the aetiology of gastric cancer: Mutagenicity of food extracts after incubation with nitrite. Food Cosmet. Toxicol. 15, 97-100.

66


(25) Weisburger, J. H., Marquardt, H., Hirota, N., Mori, H., and Williams, G. M. (1980) Induction of cancer of the glandular stomach in rats by an extract of nitrite-treated fish. JNCI 64, 163-167. (26) Maron, D. B., and Ames, B. N. (1983) Revised methods for the Salmonella mutagenicity tests. Mutat. Res. 113, 173-215. (27) Levin, D. E., Hollstein, M., Christman, M. F., Schweirs, E. A., and Ames, B. N. (1982) A new Salmonella tester strain (TA102) with A-T base pairs at the site of mutation detects oxidative mutagens. Proc. Natl. Acad. Sci. U.S.A. 79, 7445-7449. (28) Yang, D., Tannenbaum, S. R., Buchi, G., and Lee, G. C. M. (1984) 4-Chloro-6-methoxyindole is the precursor of a potent mutagen (4-chloro-6-methoxy-2-hydroxy-1-nitrosoindolin-3-one oxime) that forms during nitrosation of the fava bean (Vicia faba). Carcinogenesis 5, 1219-1224. (29) Wakabayashi, K., Nagao, M., Tahira, T., Saito, H., Katayama, M., Marumo, S., and Sugimura, T. (1985) 1-Nitrosoindole-3acetonitrile, a mutagen produced by nitrite treatment of indole3-acetonitrile. Proc. Jpn. Acad. 61, 190-192. (30) Wakabayashi, K., Nagao, M., Chung, T. H., Yin, M.-Q., Karai, I., Ochiai, M., Tahira, T., and Sugimura, T. (1984) Appearance of direct-acting mutagenicity of various foodstuffs produced in Japan and Southeast Asia on nitrite treatment. Mutat. Res. 158, 119124. (31) Lauer, K. (1991) The history of nitrite in human nutrition: a contribution from German cookery books. J. Clin. Epidemiol. 44, 261-264.

Chen et al. (32) Alais, C., and Linden, G. (1991) Meat and blood products. In Food Biochemistry (Morton, I. D., Ed.) pp 174-192, Ellis Horwood, New York. (33) Rojas-Campos, N., Sigaran, M. F., Bravo, A. V., Jimenez-Ulate, F., and Correa, P. (1990) Salt enhances the mutagenicity of nitrosated black beans. Nutr. Cancer 14, 1-3. (34) Wang, X. J., Chen, S. C., Perini, F., Gould, B. I., and Mirvish, S. S. (1995) Carboxylalkylating agents produced by nitrosation of R-amino acids. Proc. Am. Assoc. Cancer Res. 36, 116 (Abstract). (35) Ohara, A., Mizuono, M., Danno, G., Kanazawa, K., Yoshioka, T., and Natake, M. (1988) Mutagen formed from tryptophan reacted with sodium nitrite in acidic solution. Mutat. Res. 206, 65-71. (36) Kato, T., and Kikugawa, K. (1992) Proteins and amino acids as scavengers of nitrite: inhibitory effect on the formation of nitrosodimethylamine and diazoquinone. Food Cosmet. Toxicol. 30, 617-626. (37) Ohta, T., and Suzuki, S. (1983) Formation of mutagens by the reaction of amino acids with nitrite in acidic solution. Agric. Biol. Chem. 47, 405-406. (38) Ashby, J., and Tennant, R. W. (1994) Prediction of rodent carcinogenicity for 44 chemicals: results. Mutagenesis 9, 7-15. (39) Furihata, C., and Matsushima, T. (1995) In vivo short-term assays for tumor initiation and promotion in the glandular stomach of Fischer rats. Mutat. Res. 339, 15-35.

TX9500585

Gastric Carcinogenesis: 2-Chloro-4 ... - ACS Publications

Recommend Documents