ABSTRACT In this study, it was attempted to identify the mutagen produced by chlorination of insecticide fenitrothion using a quadrupole GC-MS and a high-resolution GC-MS. From the mass spectra obtained by the GC-MS, the chemical structures of two unidentified compounds were extrapolated: one was identified as 2-chloro-5-methyl-4-nitrophenol by comparison to the synthesized standard compound, while another one was identified as trichloro-3-methylphenol by high-resolution GC-MS. The mutagenicity of these identified chlorination by-products (CBPs) of fenitrothion were evaluated by the Ames assay (preincubation method) using Salmonella typhimurium TA100 without exogenous activation by S9 mix (TA100-S9). However, none were found to be mutagenic. Further studies are needed to identify the mutagenic CBP(s) of fenitrothion
Journal of Water and Environment Technology, Vol. 8, No.3, 2010 Address correspondence to Hirokazu Takanashi, Graduate School of Science and Engineering, Kagoshima University, Email: takanashi@be.kagoshima-u.ac.jp Received March 25, 2010, Accepted July 6, 2010. - 185 - Production of Trichloromethylphenol from Organo- phosphorus Pesticide Fenitrothion by Chlorination Misako KISHIDA*, Yusuke KATO*, Hirokazu TAKANASHI*, Tsunenori NAKAJIMA*, Akira OHKI*, Yuichi MIYAKE** and Takashi KAMEYA** * Graduate School of Science and Engineering Kagoshima University, Kagoshima 890-0065 JAPAN ** Graduate School of Environmental and Information Sciences, Yokohama National University, 79-7 Tokiwadai, Hodogaya-ku, Yokohama 240-8501, JAPAN ABSTRACT In this study, it was attempted to identify the mutagen produced by chlorination of insecticide fenitrothion using a quadrupole GC-MS and a high-resolution GC-MS. From the mass spectra obtained by the GC-MS, the chemical structures of two unidentified compounds were extrapolated: one was identified as 2-chloro-5-methyl-4-nitrophenol by comparison to the synthesized standard compound, while another one was identified as trichloro-3-methylphenol by high-resolution GC-MS. The mutagenicity of these identified chlorination by-products (CBPs) of fenitrothion were evaluated by the Ames assay (preincubation method) using Salmonella typhimurium TA100 without exogenous activation by S9 mix (TA100-S9). However, none were found to be mutagenic. Further studies are needed to identify the mutagenic CBP(s) of fenitrothion. Keywords: Ames Salmonella assay, chlorination by-products, mutagenicity INTRODUCTION The production, sale, and usage of pesticides in Japan are regulated by Japan’s Agricultural Chemicals Regulation Law. Before pesticides are permitted to be registered, they must undergo various toxicity tests, including the Ames assay. According to the legislation, the major derivatives of pesticides, such as hydrolysates or metabolic products produced by vegetation, are also required to undergo the tests. However, substances produced through the chlorination process are exempt from the legislation. On the other hand, the pesticides and their degradation products are actually found in natural water (Hladik et al., 2005). The natural water including these pesticides and their degradation products will be chlorinated at water purification plant, and distributed as tap water. These facts aroused our interests on whether pesticides produce mutagens when they react with chlorine. Fenitrothion [O, O-dimethyl O-(3-methyl-4-nitrophenyl) phosphorothioate] is an organo-phosphorus insecticide which has been widely used around the world. Fenitrothion has been found frequently in natural water (Tanabe et al., 2001; Iwafune et al., 2010) because of large amount of usage. Therefore, there are many studies on by-products which are produced from fenitrothion in a natural environment. Regarding chlorination by-products (CBPs), however, only two studies can be found according to our knowledge, one reported by Onodera et al., and the other reported by the authors (Kishida et al., 2010). Onodera et al. (1976) identified the fenitrothion oxidative product “fenitrooxon,” and the authors (Kishida et al., 2010) identified 3-methyl-4-nitrophenol (3M4NP), 4-nitrophenol, 2, 6-dichloro-4-nitrophenol and 2, 4, - 186 - 6-trichlorophenol, but none were found to be mutagenic. Due to the fact that fenitrothion can produce mutagen(s) through chemical reactions with chlorine (Takanashi et al., 2007), the authors attempted to identify the mutagenic CBP(s) from the chlorinated fenitrothion sample and some related samples by using GC/MS in the present study. MATERIALS AND METHODS Chemicals and Sample Preparation Fenitrothion and fenitrooxon were purchased in purity grade of Standards for Pesticide Residue Analysis from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). The chemical reagent 3-methyl-4-nitrophenol (3M4NP) was obtained from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan) while 2-chloro-5-methyl-4-nitrophenol (2C5M4NP) was obtained by custom order from Chemical Soft R&D Inc. (Kyoto, Japan). 20 mg of each of the chemicals is dissolved into 2 L of distilled water and 4 mg-Cl/mg-C of sodium hypochlorite was added at pH 7.0±0.2. After 24 hours, more than 0.1 mg/L of residual chlorine was confirmed. The CBPs were concentrated 1,000 times by a Sep-Pak Plus CSP-800 cartridge (Takanashi et al., 2001) and used for the Ames assay. To perform qualitative analyses with GC/MS, the CBPs were concentrated with a Sep-Pak Plus PS-2 cartridge (Kishida et al., 2010). GC/MS Analysis To identify and quantify compounds produced by chlorination, a quadrupole GC-MS (Shimadzu GC-17A, QP-5000, Kyoto, Japan) and a high-resolution GC-MS (Agilent HP6890, CA, USA; JEOL JMS-700, Tokyo, Japan) with an Agilent J&W DB-5 ms capillary column were used. The oven temperature was kept at 70°C for 2 min and then programmed to reach 250°C at a rate of increase of 10°C/min and was kept at 25 °C for 12 min. Helium carrier gas flow rate was set at 0.8 mL/min. The MS operated in electron-ionization mode with an ionization voltage of 70 eV, and a scan mass range of m/z 45-350. Fractionization with TLC and Ames Salmonella Mutagenicity Assay To obtain pure compounds from the chlorinated samples which contain various kinds of CBPs, the chlorinated samples were spotted on a thin-layer chromatogram (TLC) plate. Aqueous solution of 3M4NP (100 mg/L) was chlorinated as described above. The chlorinated solution was passed through three serially connected PS-2 cartridges at 15 mL/min. Residual water was removed by an aspirator and 20 mL of dichloromethane was passed through each cartridge to desorb the CBPs. Three samples of eluates from three cartridges were combined into one sample and dichloromethane was evaporated to 1 mL using a rotary evaporator. Chromatograms were developed with n-hexane – acetone (4:1, v/v) on the TLC Silica gel 60 F 245 plates (20 cm × 20 cm) (E. Merck, Darmstadt, Germany). The fractions were detected with UV light at λ = 245 nm. Each fraction was scraped from the plate and the substance was eluted from the silica gel by suspending the gel three times in methanol. The eluates were evaporated to dryness under a nitrogen stream and the solvent was exchanged with dichloromethane and DMSO to use for GC/MS analysis and the Ames assay, respectively. The preincubation method of the Ames assay was performed with the method using - 187 - Salmonella typhimurium TA100 strains, without exogenous activation (S9). The mutagenicity of samples was evaluated as net revertant per mg (net rev./mg), which was calculated from the slope of dose-response lines (Takanashi and Urano, 1998; Kishida et al., 2010). RESULTS AND DISCUSSION Qualitative and Quantitative Information Using the quadrupole GC-MS, productions of two CBPs from fenitrothion were studied. As shown in Fig. 1(A), productions of fenitrooxon (peak a) and 3M4NP (peak b) in the sample of chlorinated fenitrothion were confirmed by GC retention time (RT) and mass spectrum analyses of GC/MS results (data not shown) (Kishida et al., 2010). Comparing the mass spectrum of peak b’ and that of the standard compound (data not shown), peak b’ was identified as a 3M4NP derivative. The production of 3M4NP derivative would be due to contamination of a derivatizing agent in the instrument. The mass spectrum shown in Fig. 1(B) was obtained from peak c (RT = 19.3 min). Because m/z 210 ion was observed in the mass spectrum, production of trichloro-3-methylphenol (TC3MP) in the sample of chlorinated fenitrothion was inferred. To confirm the production of TC3MP, 3M4NP was chlorinated because mutagenic CBP(s) in chlorinated fenitrothion are expected to be found in chlorinated 3M4NP (Kishida et al., 2010). As shown in Fig. 2(A), peak c, having a mass spectrum shown in Fig. 2(B), was observed (RT = 19.3 min). This mass spectrum was clearer than that obtained in the sample of chlorinated fenitrothion (Fig.1(B)). Since TC3MP and its isomers were not commercially available, identification through a high-resolution GC-MS was attempted. A high-resolution GC-MS analysis of a chlorinated 3M4NP sample revealed that the accurate mass of compound c was 209.9420 (data not shown), and the molecular formula of compound c was estimated to be C 7 H 5 Cl 3 O (mass error, 0.0014 u). From the result of the GC-MS analysis, compound c was extrapolated to be TC3MP. Fig. 1 - A chromatogram of the chlorinated sample of fenitrothion (A) and a mass spectrum of peak c (B). Time (min) a b b’ c Relative Intensity (%) m/z Relative Intensity (%) (B) (A) - 188 - From the mass spectrum analyses for all of the peaks in Fig. 2(A), only compound d (RT = 19.6 min) was found as a substance which has a clear mass spectrum to be analyzed. Because a molecular ion at m/z 187 was observed in the mass spectrum, production of chloro-3-methyl-4-nitrophenol (C3M4NP) in the sample of chlorinated 3M4NP was inferred. C3M4NP and its isomers were not commercially available, so that 2C5M4NP, which is a monochloro substitution of 3M4NP and had a literature (Flaugh et al., 1978) for synthesis, was obtained by custom order. The NMR analytical results of 2C5M4NP was as follows: 1H-NMR (60 MHz, CDCl 3 ) δ 8.13 (s, 1H, aromatic H), δ 6.93 (s, 1H, aromatic H), δ 6.07 (s, 1H, OH), δ 2.65 (s, 3H, CH 3 ). The GC retention times of the synthesized 2C5M4NP peak and peak d were identical (19.6 min) and their mass spectra were also identical (data not shown). Therefore, it was concluded that 2C5M4NP was produced in the sample of chlorinated 3M4NP. Fig. 2 - A chromatogram of the chlorinated sample of 3M4NP (A) and mass spectra of peak c (B) and d (C). Time (min) Relative Intensity (%) c d Time (min) Relative Intensity (%) c d m/z Relative Intensity (%)Relative Intensity (%) (C) (A) (B) - 189 - Fig. 3 - Proposed fenitrothion chlorination pathway. Consequently, fenitrooxon (a), 3M4NP (b) and 2C5M4NP (d) were confirmed to be produced, and TC3MP (c) was suggested to be produced from fenitrothion by chlorination (Fig. 3). The measurable compounds, fenitrothion, fenitrooxon, 3M4NP and 2C5M4NP were quantified but fenitrothion and 2C5M4NP were not detected. Yield of fenitrooxon and 3M4NP were 25 mol% and 23 mol%, respectively. The rest of the unknown CBPs were calculated to be 52%. Fig. 4 - A chromatogram of a fraction containing TC3MP (c) and its mass spectrum. Mutagenicity of Chlorination By-Products of Fenitrothion In order to obtain TC3MP (c), compounds produced in the sample of chlorinated 3M4NP were separated and purified by TLC, and six fractions were obtained. A result of a quadrupole GC-MS analysis of the closest fraction to solvent front is shown in Fig. 3. From the GC retention time (RT = 19.3 min) and the mass spectrum (Fig.4(B)), TC3MP (c) was concluded to be isolated by TLC. c Time (min) Relative Intensity (%) c Time (min) Relative Intensity (%) m/z Relative Intensity (%) m/z Relative Intensity (%) (B) (A) NO 2 CH 3 HO NO 2 CH 3 HO NO 2 Cl HO CH 3 NO 2 Cl HO CH 3 (b) (d) NO 2 CH 3 O P CH 3 O CH 3 O S NO 2 CH 3 O P CH 3 O CH 3 O S Fenitrothion (a) NO 2 CH 3 O P CH 3 O CH 3 O O NO 2 CH 3 O P CH 3 O CH 3 O O 3Cl HO CH 3 3Cl HO CH 3 (c) - 190 - Mutagenicity of the isolated TC3MP (c) and the synthesized 2C5M4NP (d) was examined. Before conducting the Ames assays, the mutagenicity of these compounds was investigated by CCRIS (Chemical Carcinogenesis Research Information System) (U.S. National Library Medicine, 2009), which is a database of peer-reviewed carcinogenicity and mutagenicity test results. However, no information on mutagenicity of these compounds can be found in CCRIS. Hence, the Ames assays were conducted for these compounds and resulted in negative response for both the compounds (data not shown). Consequently, it was clarified that the two compounds identified are not mutagens or not particularly strong mutagens. There are some studies on mutagens produced by chlorination of pesticides. Onodera et al. (1995) chlorinated several organothiophosphorus (P=S type) pesticides and reported the mutagenicity of their oxons (P=O type). Kodama et al. (1997) chlorinated a carbamate herbicide thiobencarb and reported the mutagenicity of its sulfoxide. These mutagens are produced in a one-step chemical reaction between their parent pesticides and chlorine. On the other hand, Inoue et al. (1995) chlorinated a fungicide isoprothiolane and reported that a mutagenic CBP was produced through several steps of chemical reactions. Furthermore, Kamoshita et al. (2007, 2010) chlorinated an organophosphorus herbicide butamifos and reported similar results. It is obvious that identifying CBPs produced through several steps of chemical reactions is more difficult than that produced in just one step. Mutagenic CBP(s) of fenitrothion must be produced through several steps, which indicates that it is difficult to identify the mutagenic CBP(s) from a sample of chlorinated fenitrothion. CONCLUSIONS Chlorination by-products (CBPs) in chlorinated fenitrothion were investigated by a quadrupole GC-MS and a high-resolution GC-MS. Trichloro-3-methylphenol and 2-chloro-5-methyl-4-nitrophenol were identified as the CBPs of fenitrothion, but none were found to be mutagenic. The compounds identified in our previous study were also not the mutagens in the chlorinated sample of fenitrothion. Further studies are needed to identify the mutagenic CBP(s) of fenitrothion. REFERENCES Flaugh M. E., Crowell T. A., Clemens J. A. and Sawyer B. D. (1978). Synthesis and evaluation of the antiovulatory activity of a variety of melatonin analogues. J. Med. Chem., 22(1), 63-69. Hladik M. L., Hsiao J. J. and Roberts A. L. (2005). Are natural chloroacetamide herbicide degradates of potential environmental concern? Analysis and occurrence in the upper Chesapeake Bay. Environ. Sci. Technol., 39(17), 6561-6574. Inoue T., Aizawa T., Goto S. and Magara M. (1995). Identification of the degradation products of isoprothiolane by chlorination, and their mutagenicity and activated carbon adsorbability, J. Jap. Water Works Assoc., 64(11), 19-27. Iwafune T., Horio T., Iwasaki N., Yokoyama A. and Nagai T. (2010). Behavior of paddy pesticides and major metabolites in the Sakura River, Ibaraki, Japan. J. Pestic. Sci., 35(2), 114-123. Kamoshita M., Kosaka K., Endo O., Asami M. and Aizawa T. (2007). Reactivities of - 191 - organophosphorus pesticides in relation to their chemical classifications and their transformation to oxons by chlorination, J. Jap. Soc. Water Environ., 30(3), 145-154. Kamoshita M., Kosaka K., Endo O., Asami M. and Aizawa T. (2010). Mutagenic activities of a chlorination by-product of butamifos, its structural isomer, and their related compounds, Chemosphere, 78(4), 482-487. Kishida M., Kumabe T., Takanashi H., Nakajima T., Ohki A., Miyake Y. and Kameya T. (2010). Chlorination by-products of fenitrothion. Water Sci. Technol., 62(1), 85-91. Kodama S., Yamamoto A. and Matsunaga A. (1997). S-Oxygenation of thiobencarb in tap water processed by chlorination, J. Agric. Food Chem., 45(3), 990-994. Ministry of Labour Japan, Mutagenicity assay for the occupational safety and health act. Test guideline and GLP (1991), Occupational Health Service Center, Tokyo, Japan (in Japanese). Onodera S., Ishikura S., Kagawa Y. and Tanaka K. (1976). Chemical changes of some organic compounds in chlorinated Water. I. Formation of the P=O analogs from the P=S type of organophosphorous pesticide, The Journal of Hygienic Chemistry, 22(4), 196-205 (in Japanese). Onodera S., Maeda S. and Saitoh T. (1995). Mutagenicity and mass spectrometric characterization of oxygen analogs (P=O) derivatized from organothiophosphorus pesticides, Journal of Environmental Chemistry, 5(3), 617-624 (in Japanese). Takanashi H., Urano K., Hirata M., Hano T. and Ohgaki S. (2001). Method for measuring mutagen formation potential (MFP) on chlorination as a new water quality index, Water Research, 35(7), 1627-1634. Takanashi H., Kishida M., Hidaka Y., Kofune H., Ohyama K., Nakajima T. and Ohki A. (2007). Mutagenicity of chlorinated agricultural chemicals, Proceeding of The 2 nd IWA-ASPIRE Conference & Exhibition, P023. Takanashi H. and Urano K. (1998). Statistical procedures for estimating the detection limit and determination limit of the Ames Salmonella mutagenicity assay, Science of the Total Environment, 221(1), 31-42. Tanabe A., Mitobe H., Kawata K., Yasuhara A. and Shibamoto T. (2001). Seasonal and spatial studies on pesticide residue in surface waters of the Shinano River in Japan, J. Agri. Food Chem., 49(8), 3847-3852. United States National Library of Medicine (2009). TOXNET － Toxicology Data Network; Chemical Carcinogenesis Research Information System, http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen? CCRIS (accessed 25 February 2010) . butamifos, its structural isomer, and their related compounds, Chemosphere, 78( 4), 482 - 487 . Kishida M., Kumabe T., Takanashi H., Nakajima T., Ohki A., Miyake. surface waters of the Shinano River in Japan, J. Agri. Food Chem., 49 (8) , 384 7- 385 2. United States National Library of Medicine (2009). TOXNET － Toxicology