Document 6Rbp7jwprDQw87zVkzx8X7GE

The Science of the Total Environment, 47 (1985) 385--413 Elsevier Science Publishers B.V., Amsterdam -- Printed in The Netherlands 385 CARCINOGENIC AND MUTAGENIC PROPERTIES OF CHEMICALS IN DRINKING WATER R.J. BULL College of Pharmacy, Washington State University, Pullman, Washington 99164 ABSTRACT Isolated cases of careless handling of industrial and domestic waste has lead to a wide variety of dangerous chemicals being inadvertently introduced into drinking water. However, chemicals with established carcinogenic and mutagenic properties that occur with a high frequency and in multiple locations are limited in number. To date, the chief offenders have been chemicals of relatively low carcinogenic potency. Some of the more comnon chemicals are formed as by-products of disinfection. The latter process is generally regarded as essential to the production of a "microbiologically safe" drinking water. Consequently, any reductions in what may be a relatively small carcinogenic risk must be balanced against a potential for a higher frequency of waterborne infectious disease. The results of recent toxicological investigations will be reviewed to place the potential carcinogenic and mutagenic hazards frequently associated with drinking water into perspective. First, evidence for the carcinogenicity of certain volatile organic compounds such as trichloroethylene, tetrachloroethylene and carbon tetrachloride is considered. Second, the carcinogenic activity that can be ascribed to various by-products of chlorination is reviewed in some detail. Finally, recent evidence that other chemicals derived from the treatment and distribution of drinking water is highlighted as an area requiring move systematic attention. INTRODUCTION A wide variety of chemicals have been found in drinking waters around the world. Modern analytical techniques have identified such a diversity of chemicals that it renders comprehensive review of toxic and potentially carcinogenic properties of each substance virtually impossible. However the sources of toxic chemicals in finished drinking water are derived from three general sources: 1. Contamination of source water. 2. By-products of treatment processes. 3. Leaching from the distribution system. Examples of contaminants from each source will be discussed in turn. CONTAMINANTS OF SOURCE A great variety of chemicals can contaminate sources of drinking water. By far the most frequent contaminants in this category are trichloroethylene 0048-9697/85/803.30 <B 1985 Elsevier Science Publishers B.V. SL 034495 and tetrachloroethylene because of their large production volumes and widespread use in coinnerce. The carcinogenicity of trichloroethylene, tetrachloroethylene and to a lesser degree carbon tetrachloride are somewhat more controversial. Consequently, this section will provide an updated review of the data bearing on the carcinogenic properties of these chemicals. Consideration of other contaminants will be for comparative purposes only. A summary of pertinent carcinogenicity and mutagenicity data associated with contaminants of source waters is provided in Tables 1 and 2. Carbon Tetrachloride Studies of carbon tetrachloride (Cameron and Karunaratne, 1936; Reuber and Glover 1967a; Reuber and Glover 1967b) have documented the ability of J carbon tetrachloride to induce preneoplastic changes in the liver of rats (e.g., hyperplastic nodules and cholangiofibrosis). Subsequent studies (Reuber and Glover, 1970) demonstrated that subcutaneous administration of carbon tetrachloride in corn oil at a dose of 3080 mg/kg twice weekly gave l rise to hepatocellular carcinomas in male Japanese rats, Osborne-Mendel rats, and Uistar rats. Black rats and Sprague-Dawley rats failed to survive long enough to develop carcinomas following the same treatment. The authors' called attention to the fact that carcinoma development appeared to be inversely related to the degree of liver cirrhosis across the strains tested. However, this relationship may be fortuitous since those strains developing severe cirrhosis failed to survive. The single level of treatment utilized in this study precludes a clear resolution of this issue. A more recent study (NCI, 1976) of the carcinogenicity of carbon tetrachloride in Osborne-Mendel rats failed to confirm its carcinogenic effects at doses of up to 160 mg CCl^/kg in corn oil administered 5 times weekly for 78 weeks (47 and 94 mg/kg for males, 80 and 159 mg/kg for females). However, it should be noted that the doses utilized were much lower than used in previous studies. Eschenbrenner and Miller (1943) demonstrated that carbon tetrachloride was able to induce hepatomas in Strain A/J mice. Tumor yield was generally dose-related, but the effect was enhanced as the spacing between doses was Increased from 1 to 5 day Intervals to the same total dose of carbon tetrachloride. In subsequent experiments Eschenbrenner & Miller (1946) examined the relationship between liver necrosis and tumor development and found that the liver became resistant to necrotizing doses of carbon tetrachloride with chronic treatment. This observation was most probably related to the inactivation of cytochrome P-450 by carbon tetrachloride observed in modern metabolic studies (Glende et al_., 1975). Overall, there a rra n t 11 *** SL 03A496 t 387 appeared to be a relationship between obvious necrosis and the development of hepatomas. However, hepatomas could be induced by doses of carbon tetrachloride that did not induce obvious necrosis in the liver. The authors pointed out that they could not rule out the possibility of a slightly greater rate of liver cell replacement at these lower doses. They did find evidence of cirrhosis at doses not producing obvious necrosis, indicating some degree of liver injury. Kiplinger and Kensler (1963) also observed the induction of hepatomas following oral administration of carbon tetrachloride in C3H mice relative to historical controls, however, this study did not include concurrent controls. The National Cancer Institute (NCI, 1976) documented a high incidence of hepatocellular carcinomas in the B6C3F1 mice administered at doses of 1250 and 2500 mg/kg by gavage in a corn oil vehicle. Despite the fact of its recognized carcinogenic effects in rodents there is little convincing evidence that carbon tetrachloride possesses genotoxic properties (IARC, 1979b; USEPA, 1984). On the other hand, the ability of hepatonecrotic doses of carbon tetrachloride to promote the yield of liver tumors in mice is well documented (Pound and McGuire, 1975). Therefore, the bulk of the evidence indicates that the cancer producing activity of carbon tetrachloride are secondary to its hepatoxic effects and carbon tetrachloride might be considered as a carcinogen that does not act by a "genotoxic" mechanism. Trichloroethylene Trichloroethylene (TCE) was first shown to be carcinogenic in a study sponsored by the National Cancer Institute (NCI, 1976). In this study, male B6C3F1 mice were exposed to time-weighted average doses of 1169 and 2339 mg/kg body weight, whereas, females received 869 and 1739 mg/kg of TCE dissolved in corn oil by gavage for a period of 78 weeks. In the same study both male and female Qsborne-Mendel rats were administered doses of 549 and 1097 mg/kg TCE by the same means and for the same duration of exposure. Oose-related Increases in the incidences of hepatocellular carcinomas were observed in both male and female mice, although the response was much more marked in the males. There were no statistically significant.increases in tumor yield in either male or female Osborne-Mendel rats. Inhalation studies conducted by Henschler et al_. (1980) at concentrations of TCE in air of 100 and 500 ppm for 6 h/day, 5 days/week for a period of 18 months demonstrated an Increased Incidence of malignant lymphoma in female NMRI mice but was without apparent effect in male mice, Wistar rats or Syrian 034497 SL .^*^'52?5<SSSSP43>:)-.. SL 034498- TABLE 1. Evidence for Carcinogenicity of Carbon Tetrachloride, Trichloroethylene and Tetrachoroethylene Species A Strain Chemical (Sex) Carbon Tetrachloride C3H Hfc? (HtFJ Strain L Highest Dose 0.04 ml 2 to 3 times weekly for 8-16 weeks 0.04 ml 46x over a 4 month period (no Route Oral II Vehicle Olive Oil M Result Hepatoma Hepatoma Reference Edwards, et al., 1941 Edwards, et al., 1941 H Strain A Nice (M&F) 2.5g/kg to a total Oral Olive Oil Hepatoma Eschenbrenner of 30 doses & Killer, 1943 41 B4C3F1 Mice (M&F) 2500/mg/kg 5 X weekly Oral 2-5* Hepatocellular NCI. 1976 for 78 wks solution carcinoma In corn oil H C3H Mice 1.6 g/kg 3 X weekly Oral Corn Oil Hepatoma Kiplinger & for 10 wks Kensler, 1963 W C3H Mice 0.04 ml 2 times Rectal Olive Oil Liver Tumors Confer & weekly for 20-26 Stenger, weeks 1965 44 Osborne-Mendel 160 mg/kg 5 X weekly Oral Corn Oil Negative NCI, 1976 Rat (M&F) 78 weeks M Syrian golden 12.5 ul/hamster weekly Oral Corn Oil Liver cell Della Porta hamsters no controls carcinomas et.al., 1976 H Japanese rat (M) 2g/kg bw 2 X weekly s *c Corn Oil Hepatocellular Reuber & carcinoma Glover, 1970 0 Osborne-Mendel (M) H * N H s.c* MH Hepatocellular Reuber & carcinoma Glover, 1970 H Uistar - (M) 4 U M 41 s.c. MM Hepatocellular Reuber & Carcinoma Glover, 1970 Species & Strain Chemical (Sex) Carbon Tetrachloride Black rat (H) TABLE 1. (Continued) Hiqhest Dose 2g/kg ow 2 X weekly Route Vehicle s.c. Corn Oil 1* Sprague-Dawley (M) M H M H U *1 H Trichloroethylene M It N N B6C3F1 Mice (M&F) Osborne-Mendel (H&F) ICR Nice (F) Sprague-Dawley rats (F) Mice (H&F) 2339 mg/kg TWA for 7B weeks 1097 mg/kg TWA for 78 weeks 0, 150 & 450 ppm in air 7 h 0, 150 & 450 ppm in air 7 h per day, 5 days a week 500 ppm for 18 months Ora) Oral Inh. Inh. Inh, H It Tetrachloroethylene M H Rats (H&F) 500 ppm for 18 months Hamsters (H&F) 500 ppm for 18 months B6C3F1 Hice (H&F) 1072 mg/kg TWA for 78 weeks Osborne-Hendel (H&F) Sprague-Dawley (H&F) 600 ppm in air for 12 months Inh. Inh, Oral Oral Inh. Corn Oil Corn Oil Air Air Air Air Air Corn Oil Corn Oil - Result (Died early) Reference Rebuer & Glover, 1970 Reuber & Glover, 1970 Hepatocellular NCI, (1976) carcinoma Negative NCI, (1976) Pulmonary Fukuda, et adenocarcinoma al., 1983 Negative H It Halignant lymphoma Negative Negative Henschler et al., 1980 W II uu Hepatocellular NCI, (1977) carcinoma Negative NCI, (1977) Negative ftampy et al. 1978 TABLE 2. Evidence for Genotoxicity of Trichloroethylene and Tetrochloroethylene Chemical Trichloroethylene Tleesst Sbyy:stem ST Typfrimurium Special Conditions TA100 +/- mouse and rat S-9 fraction in desiccators Hiqhest Dose Tested 1.5 ml in desiccator 2.5 ml cytotoxic Result Pos. req. S-9 Reference Simmon et al., 1977 S. Typhimurium TA1950 TA!951 TA19S2 TA1535 TA1538 TA100 TA98 Spot test no S-9 used in vitro 0.05 ml/plate Heg. neg. neg. pos. pos. neg. neg. Cerna & Kypenova 1977 TA1950 Host mediated lUcQ i 1/2 LOcQ pos* Cerna & yi TA1951 pos. Kypenova r* TA1952 pos. 1977 ,o s * E.Coli K12 gal * +/- mouse microsomes arg + phenobarbital pre- nad + treated HTR (forward mutation) 3.3 mM in media arg req. metabol1c activation Greim et al., 1975 Covalent binding in rat hepatic microsomes In Vitro pos. Van Duuren & Banerjee, 1976 Covalent binding to tissue proteins in Wistar rats. In Vitro 1000 ppm for 5 hours. (approx, the same level of binding seen with vinyl chloride). pos. Bolt and Filser 1977 yit< -I'vV'Xri A ' '' ' 391 SL 034501 Chemical Trichloroethylene M TABLE 2. (Continued) Test System_______ Covalent binding to protein in rat liver microsomes Special Conditions, In vitro S. Typhimurium TA100 +/- Aroclor 1254 induced rat liver $-9 fraction Highest Dose Tested Unlike vinyl chloride binding is not specific to SH containing protein. Result pos. 100 ug/ml in top agar neg. with and Mi thout S-9 Reference Bolt and Filser 1977 Henschler et al., 1977 Saccharomvces cerev(siae~X?18S-14C .+/- liver s'9 20 u1/,nl fraction Pos. with S-9 only Shahin 1 Von Borstel 1977 Fischer Rat F1706 embryo cell system Covalent binding to hepatic proteins in the S.O. Rat Saccharomyces cerev(siae 07 1.1 x 104 M In vivo and In vitro Suspension test, 40 mti +/- mouse liver 10,000 x g supernatant Pos. Price et al., 1977 Pos. Allemand, et al., 1978 Pos. point mut. and give conversion required S-10 to be active Bronzetti, et al., 1978 392 SL 034502 Chemical Trichloroethylene M M II Tetrachloroethylene Test System Sa c c ha rootyces cerevlslae b? " D4 * D4 Schizosaccharomyces podtbe Schizosaccharomyces pombe DMA binding In In vitro & in vTvo E. Coli K12 TABLE 2. (Continued) Special Conditions Host mediated Host mediated Hiqhest Dose Tested 400 mg/kg 400 mg/kg 3700 mg/kg (22 repeated doses of 150 mg/kg + 400 mg/kg day of sacrifice). Result pos. ilv reversion + trp conversion Pos. at ade and trp loc. pos. trp and ade loci Reference Bronzetti, et al., 1978 MH MH PI +/- mouse and rat 22 mM liver $-9 fraction pretreated phenobarbital or B-naphthoflavone neg. at ade loci 1.3.4. 5&9 Rossi, et al. 1983 PI Host mediated 2 9/^9 neg. Rossi, et al. 1983 Calf thymus, in vitro mouse, in vivo gal * arg * rad + HTR +/- House microsomes, 0.9 mM in media phenobarbital pre treated pos. in vitro neg. in vivo neg. Bergman, 1983 Greim et al. 1975 1 Test System S. typnimurium S. typhimurium Hacromolecular binding in rat and liver tested in vivo Saccharomyces cerevisiae TABLE 2. (Continued) ___ Special Conditions ______ Highest Dose Tested TA1950 Spot test-no S-9 used TA1951 in vitro TA1952 TA1535 TA1538 TA100 TA98 Result neg. neg. neg. neg. neg. pos. neg. TA1950 Host mediated TA1951 Female ICR mice TA1952 LDS0 and 1/2 IDM Pos. no dose response evident Reference Cerna & Kypenova, 1977 Cerna A Kypenova, 1977 500 mg/kg (oral dose) Binding to Schumann et macro al. 1980 molecules but no binding detected in purified DNA D7 Suspension assay +/- 85 rti Neg. mouse liver S-9 9/kg (single Neg. Host mediated assay dose) 2g/kg x 12 + Neg. A g/kg Bronzetti et al. 1983 393 SL 034503 394 hamsters. More recently, Fukuda et ak (1983) were able to show a small increase in the incidence of pulmonary adenocarcinomas in female ICR Swiss mice exposed to ISO and 450 ppm TCE (16 and 15%, respectively, versus a control Incidence of 2%) in air 7 h per day, 5 days a week for 104 weeks. No evidence of carcinogenic activity was observed in female Sprague-Oawley rats subjected to the same treatment. Consequently, evidence that TCE is carcinogenic is limited to mice, although three different tumors have been identified in each of the mouse strains that have been studied. As a result of the initial observation that TCE was carcinogenic in mice and because it comes from the same chemical class as vinyl chloride (chloroethylene), an established human carcinogen, there have been many studies of the compound's genotoxic properties. Simmon et ak (1977) reported that TCE increased the reversion rate of 5. typhimurium strain TA100 only in the presence of a 9000 X g supernatant of homogenates made of mouse or rat liver. The experiments in this study were conducted in a desiccator and avoided the volatilization of TCE that is encountered in the standard plate assay of the Ames test when such precautions are not taken. Cerna and Kypenova (1977) were unable to demonstrate mutagenic activity of TCE in the spot test utilizing a variety of strains of S. typhimurium in the absence of a metabolic activation system, but were able to demonstrate positive activity in host-mediated assays in female ICR mice using strains TA1950, TA1951 and TA1952. Greim et ak (1975) demonstrated that TCE was capable of more than doubling the spontaneous mutation rate at the arg+ locus of E. coli Strain K12 after incubation with a liver microsomal preparation taken from mice that had been pretreated with phenobarbital. No activity was attributable to TCE in the absence of metabolic activation. Henschler et ak (1977) failed to confirm the mutagenic activity of TCE in S, typhimurium strain TA100 both in the presence and absence of a metabolic activation system derived from Aroclor 1254 pretreated rats. In addition to studies in bacterial systems, TCE has been shown to induce mutations in one yeast test system by two different research groups (Shahin S von Borstel, 1977; Bronzetti et ak, 1978) either in the presence of a metabolic activation system or in host-mediated assays, but negative under similar circumstances in another (Rossi et ak, 1983). Price et ak (1978) were able to demonstrate that TCE was capable of transforming Fischer rat embryo cells. Transformed cells were capable of growth in semisolid agar and produced undifferentiated fibrosacromas when Inoculated into newborn Fischer rats. Henschler et ak, (1977) suggested that the mutagenic activity and the carcinogenic activity of TCE might be accounted for by the presence of epichlorohydrin and 1,2-epoxibutane. However, these chemicals produced a mutagenic effect in the absence of a metabolic activation system, an observation that was not observed in prior positive tests of TCE by other authors. It has been dearly shown that TCE is metabolized to a form capable of covalently interacting with tissue proteins at a level comparable to that observed with vinyl chloride (Van Duuren & Banerjee, 1976; Bolt 4 Filser, 1977; Allemand et al_. 1978). Only very low levels of interaction of TCE with ONA in vivo have been demonstrated (Stott et a1_., 1982; Bergman, 1983). Under in vitro conditions and in the presence of a metabolic activation system, a higher degree of binding of a TCE metabolite to calf thymus DNA has been observed (Bergman, 1983). The evidence that TCE is carcinogenic was considered to be limited by IARC (1979d). At that time the evidence that TCE has genotoxic properties was obscured by reports of impurities present in the technical grade solvent (Henschler et al_., 1977). However, the mutagenic activity associated with impurities have always shown activity in the absence of metabolic activation, whereas samples of TCE have been routinely negative in the absence of metabolic activation. As indicated above, positive results have been specifically observed under circumstances where metabolic activation was possible in E. coli, S. typhimurlum, and Saccharomyces cerevisiae. Moreover demonstration of transformation in Fischer rat embryo cells supports the notion that TCE is able to act as a carcinogen. Negative results using bacterial and yeast systems in vitro appear to involve failure to consider the volatility of TCE or employed lower concentrations than utilized in positive studies. The reasons for the discrepancy in results using two yeast strains in host-mediated assays is not as clear. The lack of strong evidence of TCE interaction with DNA in vivo (Stott et aK, 1982; Bergman, 1983) presents a problem in invoking a genotoxic event as being responsible for the carcinogenic response. The difficulty in demonstrating such an interaction is in sharp contrast to that observed with vinyl chloride, a closely related chemical (Green and Hathway, 1978). On the other hand, it is difficult to completely rule out the possibility that the minor interaction observed (Stott et al_., 1982) could be responsible for some of the apparent genotoxic effects of TCE. The observations of pulmonary adenocarcinomas (Fukuda et aj_., 1983) and malignant lymphoma (Henschler et al_., 1980) in addition to hepatocellular carcinomas previously reported (NCI, 1976), indicate that TCE has some carcinogenic activity in mice. It should be pointed out, however, that these 3*505 396 tumors have an appreciable background incidence in the strains of mice utilized. In the liver of mice, doses that produced liver tumors produced definite histological evidence of hepatocellular damage and a dose*related hypertrophic response to trichloroethylene (Stott et aU, 1982). These data suggest that liver tumors in long-term studies may have resulted from repeated tissue damage and hypertrophy. Tetrachloroethylene To date, the evidence that tetrachloroethylene (perchloroethylere, PCE) is carcinogenic is limited to a study conducted by the National Cancer Institute (1977). The results of this study indicated that time weighted average doses of 536 and 1072 mg/kg to male and 386 and 772 mg/kg given to female B6C3F1 mice by gavage using a corn oil vehicle for 78 weeks resulted in the development of hepatocellular carcinomas. No evidence of increased tumor incidence was observed in Osborne-Mendel rats given time weighted average doses of 471 and 941 mg/kg in males and 474 and 949 mg/kg in females administered under the similar conditions in the same study. Rampy et a|. (1978) also failed to observe increased tumor incidences in Sprague-Dawley rats exposed to 600 ppm TCE in air for a period of 12 months. Evidence that PCE has genotoxic properties is also limited. PCE failed to increase the mutation frequency of E. coll K12 strain (Greim et al_. 1975) in the presence or absence of mouse liver microsomes. It was reported (in abstract form) to produce an increase in the reversion rate of S. typhimurium strain TA100 in the absence of metabolic activation in the spot test, but was inactive in all other strains employed (Cerna & Kypenova, 1977). On the other hand, these same authors indicated that PCE was positive in a host-mediated assay utilizing S. typhimurium strains TA1950, TA1951 and TA1952 in female ICR mice at the LD50 and 1/2 of the LD50 in the same abstract. There was no evidence of dose-response. Bronzetti et at. (1983) were unable to demonstrate an effect of PCE using Saccharomyces cerevisiae strain 07 in a suspension assay in the presence or absence of mouse liver S-9 fraction or in a host-mediated assay. It 1$ notable that these same authors had previously demonstrated positive results with TCE under both of these conditions (Bronzetti et al_. 1978). The IARC (1979e) evaluated the evidence that PCE was carcinogenic and found only limited evidence that it was active in mice. There is no substantive basis on which to modify that opinion at the present time. SL 034506 397 BY-PRODUCTS OF TREATMENT PROCESSES The triha1omethanes were the first chemicals that were recognized to be by-products of the chlorination of drinking water (Bellar et ak, 1974; Rook, 1974). Trichloromethane (chloroform), dichlorobromomethane, dibromochloromethane, and tribromomethane (bromoform) are the most comnonly encountered tribalomethanes found in drinking water. Iodinated derivatives are less commonly observed. A recent, comprehensive review of its carcinogenic, mutagenic, and teratogenic effects has been published by Davidson et ak, (1982). The present section will primarily concern itself with data that has become available since this earlier review. More recently, it has become quite apparent that the tribalomethanes (THM) are only one class of by-products produced in the chlorination of drinking water. Many of the non-THM by-products have been shown to be mutagenic in bacterial' systems, but of these only the haloacetonitriles and chlorinated phenols have been shown carcinogenic in experimental animals. Thus, only these two additional classes will be covered in the present review. However, chlorine is known to increase mutagenic activity in drinking water (e.g. Zoeteman et ak 1982; Kool et ak 1982; Meier and Bull, 1985) and in reactions with humic acid, (Meier et ak 1983; Coleman et ak, 1984; Meier et ak, 1985). On the other hand, it should be remembered that chlorination is the only practice extensively studied. Many by-products remain to be identified with chlorine as welT as other alternative modes of disinfection and even those which have been identified have been poorly characterized toxicologically. In addition to by-products of drinking water disinfection, other treatment processes have the potential for the introduction of carcinogens into drinking water. In this light the carcinogenicity of acrylamide is discussed for illustrative purposes. Chloroform A study sponsored by the National Cancer Institute (NCI, 1976) demonstrated that B6C3F1 mice of both sexes developed hepatocellular carcinomas while male Osborne-Mendel rats had increased yields of renal tumors following chronic exposure to chloroform (Table 3). Chloroform was adnlnistered to both mice and rats in a com oil vehicle by stomach tube. Because of the primary source of chloroform In drinking water Is side reactions of chlorination and the necessity for producing microbiologically safe drinking water a more thorough examination of the dose response relationships Involved in chloroform-induced cancer was undertaken (Jorgenson et ak 1985). The doses used in this study was extended to a lower range and 03^507 SL 034508 Species and Strain A Mice TABLE 3. Evidence for Carcinogenicity of Chloroform in Experimental Animals Vehicle Olive Oil Highest Dose Tested 2350 mg/kg Route oral Tumor Liver C57BL Nice Toothpaste GO mg/kg/d oral none CBA Mice Ci/l 1CI Mice Sprague-Dawley Rat Beagle Dog M Osborne-Mendel Rat 1 N N B6C3F1 Mice M Osborne-Mendel Rat B6C3F1 Mice Toothpaste loothpaste Toothpaste Toothpaste Toothpaste 41 Corn Oil H U It Corn Oil U Drinking Water Drinking Water GO mg/kg/d 60 mg/kg/d 60 mg/kg/d 60 mg/kg/d 20 mg/kg/d II 180 mg/kg/d H U tt 477 mg/kg/d M 130 mg/kg/d 400 mg/kg/d oral oral oral oral oral M oral N H M oral M oral oral none none kidney none none Total Neoplasms Kidney Thyroid Cholangiofibromas Cho1angiocarcinomas Liver Lymphoma Kidney None References Eschenbrenner & Miller, 1945 Roe et al., 1979 ti U n H it Reuber, 1979 NCI, 1976 Reuber, 1979 It M M It NCI, 1976 Reuber, 1979 EPA-NC1 Study. EPA-NCI Study. V0 CD * r *, c' ; ......... ffv 399 the group sizes were substantially expanded. Chloroform was administered in the drinking water. This design was to avoid problems that may have been associated with gavage dosing and the corn oil vehicle. The expanded group size at low doses would provide a better basis for extrapolation of the results to estimate human risks. Serum and tissue biochemical measurements were made in separate groups of animals on parallel treatments. Additional control groups were utilized for both species in which were restricted in their water consumption to the amount consumed by the high dose groups for each species. This provided a check on the effects that might be secondary to decreased water consumption. Only female B6C3F1 mice and male Osborne-Mendel rats were included in the experiment as these were the sexes which gave the greatest response in the NCI (1976) study. In the male Osborne-Mendel rat very similar results were observed in the Jorgenson et al_. study (1985) as had been obtained in the prior NCI study (1976). In particular, there was a dose-related Increase in renal tubular adenomas and adenocarcinomas. The results obtained with the female B6C3F1 mice were, however, substantially different than observed in the original NCI study. There was a complete absence of any dose-related increases in the incidence of hepatocellular carcinomas in mice. This result was particularly significant, since the daily doses of chloroform in the two high exposure groups bracket the low dose which gave rise to an 80X incidence of hepatocellular carcinomas in the earlier NCI study. These results are interesting in the context of evidence of chloroform carcinogenicity that was obtained prior to the NCI study (Table 3). Eschenbrenner and Miller (1945) noted that hepatocarclnogenic effects of chloroform were confined to doses which produced frank liver necrosis. Studies quoted in Table 3 that failed to produce increases in tumor incidence involved dose levels lower than those used in either the NCI (1976) or Jorgenson et aK (1985) studies. Roe et al_. (1979), however, did find increased- incidence of renal tumors in ICI mice when chloroform was administered in a toothpaste base at a dose of 60 mg/kg per day. Renal tumors were not produced in the C57BL, CBA, and CF/1 strains at the same dose in the same study. No liver tumors were not observed in any of these mouse strains. It has been argued that the carcinogenicity was secondary to tissue necrosis followed by regenerative hyperplasia largely because of the early observations of Eschenbrenner and Miller (1945). Chloroform at doses found capable of Inducing tumors produces substantial Increases of trltiated thymidine incorporation into DNA of both the liver and kidney of the 86C3F1 mouse (Reitz et al., 1982). This result provides evidence of a regenerative 03459 400 response and is consistent with this hypothesis. Along the same line, Moore et al_. (1982) provided evidence that doses of 60 mg/kg of chloroform in corn oil resulted in damage to the kidney and evidence of tubular regeneration as measured by tritiated thymidine uptake. These same doses administered in a toothpaste base were without effect. At 240 mg/kg similar effects were seen with both vehicles. This type of response was not strongly apparent in the liver or kidney of Osborne-Mendel rats. In animals receiving parallel treatments to those used in the Jorgenson et aK study (1985) there was a dose-related increase in the liver fat content apparent in chloroform-treated mice (Jorgenson et al_., 1982). This effect was observed only at much higher doses in the rat. Before the changes in liver fat occurred in the rat they were preceded by substantial a depression of serum triglycerides. Data on serum triglycerides of the mice under these conditions are not presently available. Based on the rat data one can speculate that the use of corn oil may have contributed to the development of hepatocellular carcinomas in the first NCI study through some type of interaction with the hepatoxic effects of chloroform. The difficulty that has been encountered trying to demonstrate that chloroform possesses genotoxic properties (Table 4) indirectly supports the thesis that chloroform induced tumors are the result of overt tissue necrosis. A number of independent workers have failed to show a direct interaction of chloroform or a metabolite with DNA in the mouse or rat liver or kidney (Diaz-Gomez and Castro, 1980; Reitz, et al_. 1982; Pereira et al_. 1982). On the other hand, chloroform has been shown capable of inducing chromosome breakage and increased sister chromatid exhange frequencies in human lymphocytes in vitro and increased sister chromatid exchange in the bone marrow of mice treated with chloroform _Vn vivo (Morimoto and Koizumi, 1983). It has also been found to enhance the transformation of Syrian Hamster Embryo (SHE) cells by S7 adenovirus (Hatch et ah, 1983). It is notable that the concentrations of chloroform tested in these positive experiments tend to be higher than those utilized in similar experiments that had negative results (Kirkland et at., 1981). It must be said, however, that the exact mechanism by which chloroform induces cancer remains elusive. Attempts to directly demonstrate an alternative mechanism for chloroform-induced cancer have produced equivocal data. For example, Pereira et aK (1982) failed to clearly demonstrate either tumor Initiating or tumor promoting activity In the rat liver. In suamary, the extent to which chloroform In drinking water represents a carcinogenic hazard to man is still uncertain. There is no doubt that chloroform Is capable of producing renal tumors in at least one strain of mice and rats. Chloroform can also induce liver tumors in mice, but this effect seems dependent on the vehicle used. In view of the apparent dependence of the carcinogenic effects of chloroform in the mouse liver on the vehicle in which it was administered, it is suggested that chloroform's carcinogenic activity in the rat may be more appropriate for estimating carcinogenic risk in humans. Haloacetonitrlles The dihalogenated chlorine and bromine derivatives of acetonitrile have been the principal species identified in chlorinated drinking water. (Trehy & Bieber, 1981; Oliver, 1983). Investigations into the carcinogenicity of these chemicals have been confined to the three members of the chlorinated series, bromochloroacetonitrile and dibromoacetonitrile. Simmon et al_. (1977) first documented that dichloroacetonitrile (DCAN) was mutagenic in Salmonella typhimurium. Bull et al_. (1985) confirmed the mutagenic properties of DCAN in Salmonella and demonstrated that bromochloroacetonitrile (BCAN) is mutagenic under similar circumstances. Chloroacetonitrile (CAN), trichloroacetonitrile (TCAN) and dibromoacetonitrile (DBAN) were found to be Inactive in the presence or absence of a 9000 X g supernatant fraction (S-9) of a liver homogenate taken from rats previously treated with Arochlor 1254. All five haloacetonitriles induced sister chromatid exchange (SCE) in Chinese hamster ovary cells, jn vitro, in the absence of rat liver S-9 fraction. None of these same five haloacetonitriles, however, were found capable of inducing mutagenic activity in the mouse micronucleus assay in vivo (5 male and 5 female CO-1 mice in each group) at doses of 0, 12.5, 25 and 50 mg/kg bw for 5 consecutive days in this same study. Direct evidence of the carcinogenicity of the haloacetonitriles is limited to initiation/promotion experiments in the mouse skin. Experiments employing the oral route of administration (a total dose of 150 and 300 mg/kg bw split between 6 individual doses) followed by a 20 week promotion schedule with 12-0-tetradecanoyl-phorbol-13-acetate (TPA) failed to demonstate a significantly elevated incidence of either benign or malignant skin tumors within a 1 year observation period (Bull et aK, 1984). On the other hand, higher doses applied topically (1200, 2400 and 4800 mg/kg, also split between 6 applications), yielded significantly higher tumor incidences with CAN, BCAN and DBAN. DCAN and TCAN produced small increases in cumulative tumor yields, but the changes were not significantly different from controls. SL 034511 402 SL 034512 TABLE 4. Evidence for Genotoxic Activity of Chloroform Test System Special Conditions Highest Dose Tested Result S. TvDlmurlum TA1535 TA1537 M TA153B M TA98 TA100 Syrian Hamster Embryo CellsEnhanced viral transformation V79 cells - 8-azaguanine */- House A Rat Liver and kidney S-9 M M 1* M Applied In sealed chamber Applied In flow through system, then sealed 10 mg/plate II N II M 0.5 ml/chamber 31 In air neg. ii n ii pos. neg. Human Lymphocyte - Sister Chromatid Exchange Induction Human Lymphocyte Chromosome Damage E. Coll MP2p and WP2uvr A*p Human Lymphocyte - Sister Chromatid Exchange. In vitro House Bone Marrow - Sister Chromatid Exchange. jj vivo Rat Liver, DIM - binding In vivo 4/ - Rat Liver S-9 +/ - Rat Liver S-9 4/ - Rat Liver S-9 400 ug/plate 400 ug/plate 10 mg/plate 6 mg/ml 4 x 200 mg/kg/1 neg. neg. neg. pos. pos. neg. Reference Van Abbe et a). 1982 II U II Hatch et al., 1983 Sturrock, 1977 Kirkland et al., 1981 II II Morimoto & Koisuml 1983 M Diaz-Gomez & Castro, 19B0 1 * s . r' \ .' Test System Sprague Dawley Rat, DNA binding In vivo B6C3F1 mice, DKA binding, B6C3F1 mice, DNA binding In vivo tn t-1 TABLE 4. (Continued) Special Conditions Highest Dose Tested Liver N.A. Kidney Liver Liver N.A. N.A. N.A. Result neg. neg. neg. neg. Reference Pereira et al., 1982 It H Reiti et al., 1982 034513 obJ 1 404 A preliminary investigation of the haloacetonitriles ability to increase the incidence of lung tumors in strain A/J mice was reported by Bull and Robinson (1985). Ooses of 10 mg/kg administered three times weekly for 8 weeks significantly increased lung tumor yields at 9 months of age with CAN, TCAN, and SCAN (Bull and Robinson, 1985). However, the tumor incidence was increased to only 32, 28 and 31S respectively. Control incidence of this spontaneous tumor was 105. Although statistically significant, this small increase in lung tumors is difficult to interpret because of the variable background rate of this tumor in A/J mice. These data indicate that certain members of the haloacetonitrile class do possess weak carcinogenic properties. This evidence has been confined to the topical route of administration and coupled to the subsequent application of a potent tumor promoting agent, TPA. There is no evidence available concerning the ability of these compounds to induce cancer by a systemic route of administration. Because of their relatively common occurrence as a result of the chlorination of drinking water it is critical that studies examining this issue be initiated. Chlorinated Phenols Chlorination of drinking water often gives rise to small quantities of chlorinated phenol derivatives. The principle products are 2-chlorophenol, 2,4-dichlorophenol and 2,4,6-trichlorophenol. There is substantive evidence of carcinogenic properties for only 2,4,6-trichlorophenol. The National Cancer Institute (NCI, 1979) conducted a feeding study of the effects of 5,000 and 10,000 ppm of 2,4,6-trichlorophenol in the diet of F344 rats and 5,214 and 10428 ppm in the diet of B6C3F1 mice. The incidence of lymphomas or leukemias was increased in male rats in a dose-related manner. Hepatocellular carcinomas were increased in both male and female mice and the incidences were dose-dependent. Chlorinated Aldehyde and Ketone Derivatives A number of chlorinated aldehydes and ketones have been identified in drinking water or related matrices following chlorination. A number of these by-products have been shown to be mutagenic (Bull and Robinson, 1985). Preliminary results from our studies, indicate that at least two of these chemicals, 2-chloropropenal and 1,3-dichloroacetone are capable of Initiating tumors in mouse skin (unpublished results). SL 034514 405 Other Treatment Chemicals A wide variety of chemicals are necessary in the production of a safe and wholesome drinking water. For the most part these chemicals are generally regarded as safe and pose no substantive hazard to the health of the consuming public. However, if the synthesis of some products is improperly controlled they can contain contaminants that are of potential concern. An example of such products is the group of polymeric chemicals that are conmonly used clarifying water, the coagulant aids. The coagulant aids are high molecular weight polymers that are essentially non-toxic and are removed from the finished water by precipitation and filtration. However, the monomeric components used in the synthesis of these polymers are often reactive chemicals that are highly toxic and if sufficiently stable in water could be expected to be in the finished drinking water. One example of such a monomer is acrylamide. Acrylamide has been recently shown capable of initiating papillomas and carcinomas in the skin of mice whether administered by the oral, Intraperitoneal or topical routes of administration (Bull et aU, 1984a). It was also shown to be capable of increasing the yield of lung adenomas in Strain A/J mice in the same study. More recently, very similar results have been reported with ICR-Swiss mice (Bull 1984b). Although acrylamide is inactive in the standard plate assay of the Ames' test (Bull et,aK, 1984a) it does increase the frequency of sister chromatid exchange in Chinese hamster ovary cells, jm vitro (Bull et al_., 1983). Shlraishi and Yamamoto (1978) and Shiraishi (1978) reported that acrylamide produces aneuploid and polyploid cells in both bone marrow cells and spermatogonia of mice treated with acrylamide by oral or intraperitoneal administration. Bull et aK (1983) observed dose-related Increases in spermhead abnormalities in mice with acrylamide. More recently, Vanhorick and Moens (1983) found that acrylamide was a weak inducer of SV40 ONA amplification in SV40-tran$formed Chinese hamster cells as measured by H) situ hybridization techniques. However, acrylamide synergistically enhanced the Induction of SV40 DNA synthesis by a number of other carcinogens. DISTRIBUTION SYSTEM Contribution of toxic chemicals from the distribution system to finished drinking water can be considerable under certain circumstances. It Is beyond the scope of this paper to deal comprehensively with this subject. However, It should be pointed out that certain products used in drinking water distribution are made of monomeric components that are recognized carcinogens. The most obvious example is polyvinyl chloride pipe which is a polymer of the SL 034515 406 well established human carcinogen vinyl chloride (IARC, 1979). Modern manufacturing techniques have reduced the monomer content of these products to negligible concentrations, but users of such products should carefully avoid substandard materials. Another product that was widely used in lining of certain water mains, storage tanks and pipes are coal tar and asphaltic-based paints. The high polyaromatic hydrocarbon content of coal tar paints is partially, but not completely responsible for the high carcinogenic activity of these products in the mouse skin (Robinson et aK, 1984). Although still containing carcinogenic components, the asphaltic based paints used in contact with potable water are approximately 3 orders of magnitude less potent than their coal tar analogs. Because of the extremely low water solubility of the polyaromatic hydrocarbons that are partially responsible for the carcinogenic activity of coal tar these chemicals would not be expected to be rapidly leached from coatings. However, the extent to which these coatings provide a relatively constant reservoir for the leaching of carcinogen from the surface of the distribution system has not been well studied. However, some information exists to suggest that as these coatings age they can begin to shed fine particulates containing high concentrations of polyaromatic hydrocarbons into the finished drinking water. DISCUSSION It is clear from the previous discussion that chemical carcinogens find their way into drinking water through a variety of means. Obviously, industrial contamination is only one and in most cases a minor source of chemicals in drinking water. Evidence that the chlorination by-products 2,4,6-trichlorophenol and the haloacetonitriles have carcinogenic properties indicate that we must look beyond the trihalomethanes as the only potential carcinogenic hazards associated with disinfection. However, whatever the carcinogenic risk these products might represent they must be balanced against the obvious reduction of disease from waterborne infectious disease that results from chlorination. Finally there 1$ a clear need to carefully scrutinize products that are utilized in the treatment and distribution of drinking water since these products can be a major source of harmful chemicals presented at the tap for human consumption. Despite the need to accept the general notion that carcinogenicity In experimental animals signals cause for concern about human exposures to the same chemical, it must be recognized that considerable uncertainty underlies the extrapolation of that information to man. These uncertainties are both SL 034516 407 qualitative and quantitative and have been discussed in detail elsewhere (IARC, 1982c). For example, if a chemical induces malignant tumors in more than one species, particularly if it produces tumors that are rare in the particular strain or species its carcinogenic properties are widely accepted (IARC, 1982c). On the other hand, a chemical that increases the incidence of a tumor that has a high spontaneous incidence in the test animal employed are felt to provide only limited evidence of carcinogenicity. Certain of the low molecular weight chlorinated hydrocarbons fall into this category. Other chemicals from the same class, such as vinyl chloride, cause little controversy because they produce malignant tumors in multiple sites and different species. It should be recognized that this view does not suggest that increased incidence of spontaneous tumors is of less concern than induction of rare tumors. Rather It is a statement related to the wide variety of factors that can modify the yield of a spontaneous tumor, whereas induction of a rare tumor probably requires the properties of a complete carcinogen. It is ironic that most of the chemicals that occur frequently in drinking water fall into a class of carcinogens whose carcinogenic effects are controversial. In general, the controversy involves those chemicals which induce liver tumors, in particular those whose activity is limited to the induction of liver tumors in mice. Part of this concern is illustrated by the fact that the yield of hepatocellular carcinomas in B6C3F1 mice can be increased substantially by simple performance of a partial hepatectomy (Newberne et al.., 1982). The chemicals most directly involved In this controversy are trichloroethylene and tetrachloroethylene, but also involve considerations of the carcinogenicity of carbon tetrachloride and chloroform. Carbon tetrachloride should in many ways be considered the prototype for this group, although its carcinogenicity seems somewhat more generally accepted than the other three compounds. As pointed out above, carbon tetrachloride is capable of producing liver tumors in a variety of species. Despite many attempts (many of which have not been published), it has been virtually impossible to demonstrate any genotoxic activity with carbon tetrachloride (IARC, 1979b; USEPA, 1984). However, the studies that have demonstrated carbon tetrachloride Induced liver cancer has involved very high doses, from 1000 to more than 2000 mg/kg administered In an vegetable oil vehicle 2 to 5 times weekly for extended periods of time. The experimental results with trichloroethylene in many ways contrast sharply with that obtained with carbon tetrachloride. Trichloroethylene has been shown to only induce hepatocellular carcinomas in mice (NCI, 1976). SL 034517 Although of a rather low potency, the genotoxic properties of trichloroethylene have been demonstrated by a number of investigators. In this respect, evidence of its carcinogenic activity would have to be considered stronger than that of carbon tetrachloride. However, the fact remains that trichloroethylene has only been shown capable of increasing the yield of neoplasms only in strains of mice which have high background incidence of these same neoplasms (Henschler et ak, 1980; Fukuda et ak 1983; NCI, 1976b). Examination of the livers of mice receiving chloroform in drinking water demonstrated a dose-related increase in liver fat content that extend to levels much lower than those utilized for the original NCI coreinogenes is study (Jorgenson et ak, 198Z). Similar changes were not observed in the livers of Osborne-Mendel rats until much higher levels of chloroform were administered in drinking water. These results suggest that corn oil was in some way involved in the liver pathology that develops in mice treated with chloroform. If so, this would imply that tumors Induced by other hepatoxic agents may depend heavily on the corn oil vehicle coimonly used in testing of non-polar chemicals. Such an interpretation would suggest that results that have been obtained with chemicals such as trichloroethylene and tetrachloroethylene may not be extrapolated to other species without appropriate qualification. In fact, these data also call into similar question the carcinogenicity of carbon tetrachloride since It has only been shown capable of producing liver tumors when administered in a vegetable oil vehicle. Tetrachloroethylene increases hepatic ONA synthesis in B6C3F1 mice but not in rats following doses of 1000 mg/kg per day for 11 days (Schumann, et ak, 1980). Similarly, Reitz et ak (1982) found substantial increases In ONA synthesis of the liver and kidney of B6C3F1 mice treated with single doses of 60 and 240 mg/kg of chloroform and much smaller increases in the liver and kidney of male Osborne-Mendel rats. These data indicate that the toxic effects of these chemicals in the mouse liver results in substantial tissue regeneration of dose levels utilized in the carcinogenesis bioassays of these chemicals. Consequently, the possibility exists that these chemicals are capable of Increasing liver tumor development from spontaneously initiated cells in the same way that carbon tetrachloride has been shown to Increase hepatomas in the liver of dlethylnitrosamines initiated mice (Pound and McGuire, 1978). These questions become extremely important when attempting to estimate the risk to human beings consuming drinking water. Presumably the damage that 3*5l8 f i r` 409 is produced by initiators of cancer (i.e. those that act through damage to DNA) produce a certain degree of irreversible or non-repairable damage that is directly related to the dose of carcinogen. Theoretically, this would lend itself to a linear relationship to dose at low response rates and risks at low concentrations can be reliably estimated from experiments conducted at high doses. Increased tumor incidence due to non-genotoxic effects of chemicals or interactions with nutritional state presumably involve reversible processes. Consequently, the assumption of a linear relationship with dose at low response rates may not be justified. This may substantially affect any estimate of the health hazard associated with the low concentrations of these chemicals that occur in drinking water that is based upon the high doses used in animal experiments. The data available at present is not suited for projecting the risk to humans from the use of materials in the treatment and distribution of drinking water. Identification of carcinogenic properties of by-products of disinfection other than the trihalomethanes, acryamide and coal tar paints, however, does indicate that the use of such material in contact with potable water requires closer scrutiny. The industry should examine critically time-honored practices in the treatment and distribution of drinking water to gradually replace these materials with safer alternatives. In summary, evidence that a chemical Induces cancer in experimental animals does suggest that there is a potential risk to man regardless of the mechanism involved. Although it is certainly possible that a chemical that produces cancer in experimental animals would not do so in man, it 1$ difficult to develop a prudent course of action on this assumption because the evidence which presently exists would suggest that the reverse is the general case. If tumorlgenesls is dependent upon tissue damage, it should be remembered that these agents are also capable of inducing liver damage in humans. Consequently, it should be recognized that the argument presented regarding relationship of tissue damage to tumorigenicity is more of a quantitative consideration in the judgement of risks in man and less of a question of across-species extrapolation than discussions of the issue would ordinarily imply. Therefore, the appearance of all the above chemicals in drinking water is cause for concern and to be avoided wherever possible and practical. SL 034519 410 REFERENCES Allemand, H.t Pessayre, 0., Descatoire, V., Degott, C., Feldmann, G., and Benhamou, J.P., (1978) Metabolic activation of trichloroethylene into a chemically reactive metabolite toxic to the liver. J. Pharmcol. Exp. Therapeut.. 204, 714-723. Bellar, T.A., Lichtenberg, J.J. and Kroner, R.C., (1974) The occurrence of organohalides in chlorinated drinking waters. J. Am. Mater Works Assn., 66, 703-706 Bergman, K,, (1983) Interactions of trichloroethylene with DNA in vitro and with RNA and DNA of various mouse tissues in vivo. Arch. Toxicol.. 64, 181-193. Bieber, T.I. and Trehy, M.L., (1983) Dihaloacetonitriles in chlorinated natural waters. In: Jolley, R.L., et al. eds., Water Chlorination: Environmental Impact and Health Effects, 4, Ann Arbor MI: Ann Arbor Science Publi., Inc., pp 85-96. Bronzetti, G,, Zeiger, E., and Frezza, 0., (1978) Genetic activity of trichloroethylene in yeast. J. Toxicol. Environ. Health, 1, 411-418. Bronzetti, G., Bauer, C., Corsi, C., Del Carratore, R.. Gal 11, A., Nieri, R., and Paolini, M., (1983) Genetic and biochemical studies on perchloroethylene in vitro and in vivo. Mutation Research 116, 323-331. Bolt, H.M. and Filser, J7G., (1977) ""Irreversible binding of Chlorinated ethylenes to macromolecules. Environ. Health Persp. 21, 107-112. Bull, R.J., Meier, J.R. and Robinson, M. (1983) Evidence of genotoxic effects with acrylamide. Pharmacologist 25, 640. Bull, R.J., Robinson, M., Laurie, R.b., Stover, G.D., Greisiger, E., Meier, J.R., and Stober, J., (1984a) Carcinogenic effects of acrylamide in sencar and A/J mice. Cancer Res. 44, 107-111. Bull, R.J., Robinson, M. and Stober, J.A7 (1984b) Carcinogenic activity of acrylamide in the skin and lung of Swiss-ICR mice. Cancer Letters 24, 209-212. Bull, R.J., Meier, J.R., Robinson, M., Ringhand, P., Laurie, 0. and Stober, J., (1985) Evaluation of mutagenic and carcinogenic properties of brominated and chlorinated acetonitriles: By-products of chlorination. Fundam. Appl. Toxicol. In press. Bull, R.J. and Robinson, M~ (1985) Carcinogenic activity of haloacetonitriles and haloacetone derivatives in the mouse skin and lung. In: Jolly R.L., et ak eds., Water Chlorination: Environmental Impact and Health Effects, 5, In Press. Cameron. and Karunaratne, W.A.E., (1936) Carbon tetrachloride cirrhosis in relation to liver regeneration. J. Pathol. Bacteriol. 42, 1-21. Cerna, M. and Kypenova, H., (1977) Mutagenic activity of chloroethylenes analyzed by screening system tests. Mutat. Res. 46, 214-215. Davidson, I.W.F., Sumner, D.D. and Parker, J.C., (1982*J A Review of its metabolism, teratogenic, mutagenic and carcinogenic potential. Drug and Chemical Toxicol. 5, 1-87. Oiaz-Gomez, M.I. and Castro, J.A., (1980) Covalent binding of chloroform metabolites to nuclear proteins - no evidence for binding to nuclei acids. Cancer Lett. 9, 213-218. Eschenbrenner, A.B., and Miller, E., (1943) Studies on hepatomas. I. Size and spacing of multiple doses in the induction of carbon tetrachloride hepatomas. J, Natl. Cancer Inst. 4, 385-388. Eschenbrenner, A.B., and"Miller, E., (1?45) Induction of hepatomas in mice by repeated oral administration of chloroform with observations on sex differences. J. Natl. Cancer Inst. 2, 251-255. Eschenbrenner, A.B., and Miller, E., ((1>4E) Liver neurosis and the induction of carbon tetrachloride hepatomas in strain A mice. J. Natl. Cancer Inst. 6, 325-341. sL 034520 412 Moore, D.H., Chasseaud, L.F., Majeed, S.K., Prentice, O.E., Roe, F.J.C., and Van Abbe, N.J. (1982) The effect of dose and vehicle on early tissue damage and regenerative activity after chloroform administration to mice. Fd. Chem. Toxic.. 20, 951-954. Morimoto, K. and Koizimu, A., (1983) Trihalomethanes induce sister chromatid exchanges in human lymphocytes in vitro and mouse bone marrow cells in vivo. Environ. Research., 32, 72-79. NCI (1976a) National Cancer Institute Carcinoqenesis Bioassay of Chloroform. NTIS No.'T>B2CT0T8/A5: NCI (1976b) National Cancer Institute Carcinogenesis Bioassay of Trichloro- ethylene'T~Wn-C5-TR-2.~NTB-7fi-802. NCI (1977) National Cancer tnstitute~ffloassav of Tetrachloethylene for Possible Carcinogenicity. (NIH) 77-813. NCI (1979) Bioassay of 2.4.6-trichlorophenol for possible carcinoqenicity. PHEW Publ. No. (NIH) 73-1711............................... Newberne, P.M., de Camrargo, J.L'.V., and Clark, A. (1982) Chloine deficiency partial hepatectomy and liver tumors in rats and mice. Toxicol. Pathol., 10, 95-109. Newberne, P.M., Weigert, J., and Kula, N., (1979) Effects of dietary fat on hepatic mixed-function oxidases and hepatocellular carcinoma induced by aflatoxin B, in rats. Cancer Res. 39, 3986-3991. Oliver, B.G., (1983) Oihaloacetonitriles in drinking water: Algae and fulvic acid as precursors. Env. Sci. Technl., 17, No. 2, 80-83. Pereira, M.A., Lin, L.-H.C., Lippitt, H.M. an3""Herren, S.L. (1982) Trihalomethanes as initiators and promoters of carcinogenesis. Environ. Hlth. Persp , 46, 151-156. Pound7 A.W. and McGuire, L.J., (1978) Influence of repeated liver regeneration on hepatic carcinogenesis by diethylnitrosamine in mice. Br. J. Cancer, 37, 595-602. Price, P.J., Hassett, C.m, Mansfield, J.I. (1978) Transfonning activity of trichloroethylene and proposed industrial alternatives. In Vitro, 14, 290-293. Rampy, L.U., Quast, J.F., Balmer, M.F., Leong, B.K.J. and Gehring, P.J. (1978) Results of a long-term inhalation toxicity study in rats of a perchloroethylene (tetrachloroethy1ene) formulation. Toxicology Research Lab,. Dow Chemical Co., Midland, MI 48640. Reitz, R.H., Fox, T.R, and quast, J.F., (1982) Mechanistic considerations for carcinogenic risk estimation: chloroform. Environ. Hlth. Persp. 46, 163-168. Reuber, M.0. and Glover, E.L. (1970) Cirrhosis and carcinoma of the liver male rats given subcutaneous carbon tetrachloride. J. Natl. Cancer Inst.. 44, 419-423. Reuber, M.D. and Glover, E.L., (1976a) Hyperplastic and early neoplastic lesions of the liver in buffalo strain rats of various ages given subcutaneous carbon tetrachloride. J. Natl. Cancer Insti.. 38, 891-895. Reuber, M.D. and Glover, E.L. (1976b) Cholangiofibrosis in the liver of Buffalo strain rats injected with carbon tetrachloride. Exp. Pathol. 48, 319-322. Reuber, M.D. (1979) Carcinogenicity of chloroform. Environ. Hlth. Persp. 31, 171-182. Robinson, M., Bull, R.J., Munch, J. and Meier, 0. (1984) Comparative carcinogenic and mutagenic activity of coal tarr and petroleum asphalt paints used in potable water supply systems. J. Appl. Toxicol.. 4, 49-56. " Roe, F.J.C., Palmer, A.K., Worden, A.N., Van Abbe, N.J., (1979) Safety evaluation of toothpaste containing chloroform. I. Long-term studies in mice. J. Environ. Path. 5 Toxicol., 2, 799-819. SL 034521 413 Rook, J.J. (1974) Formation of haloforms during chlorination of natural water. Water Treat. Sxamin., 23, 234-243. Rossi, A.M., Migliore, L., Ba'r'aTe, R7 and Loprieno, N., (1983) In vivo and in vitro mutagenicity studies of a possible carcinogen, trichloroethylene, and its two stabilizers, epichlorohydrin and 1,2-epoxybutane. Terataqenesis Carcinogenesis S Mutagenesis, 3, 75-87. Shahln, m.m! and von Borstel, R.C. (1977) Mutagenic and lethal effects of -benzene hexachloride, dibutyl phthalate and trichloroethylene in Saccharonyces cerevisiae. Mutat. Res.. 48, 173-180. Shiraishi, Y. (19781 Chromosome aberations induced by monomeric acrylamide in bone marrow and germ cells of mice. Mutat. Res., 57, 313-324. Shiraishi, Y. and Yamamoto, K., (1978) Chromosome aberrations induced by ' monomeric acrylamide in germ cells of mice. Proc. Japan Acad., 54, Ser. B., 272-276. Slnnon, V.F., Kauhanen, K., Tardiff, R.G. (1977) Mutagenic activity of chemicals identified in drinking water. Progress in Genetic Toxicology, pp 249. Stott, W.T., quest, J.F and Watanabe, P.G. (1982) The pharmacokinetics and macromolecular intention of trichloroethylene in mice and rats. Toxicol. Appl. Pharmacol. 62, 137-151. Trehy, M.L. and Bieber, T.I., (1981) Detection, identification and Quantitative analysis of dihaloacetonitriles in chlorinated natural waters. Thl Keith. L.H", ed. Vol. 2, Advances in the Identification and Analysis of Orqanic Pollutants in Water.. (Ann Arbor. MI: Ann Arbor 5dei.ce PubI. .Inc.), pp, 941-975.-------- r USEPA (1984) Research and Health Assessment Document for Carbon Tetrachloride, EPA 600/8-82-001. Van Duuren, B.L. and Banerjee, S., (1976) Covalent Interaction of metabolites of the carcinogen trichloroethylene in rat hepatic microsomes. Cancer Research. 3, 2419-2422. Vanhovick, M, and Moens, W. (1983) Carcinogen-mediated induction of SV40 DNA amplification is enhanced by acrylamide in Chinese hamster C060 cells. Carcinogenesis, 4, 1459-1463. Welsburger, J.H. and Williams, G.M. (1983) The distinct health risk analysis required for qenotoxic carcinoqens and promotinq aqents. Environ. Hlth. Persp., 50, 233-245. ' Withey, J.R.,Toll ins, B.T. and Collins, P.G., (1983) Effect of vehicle i on the pharmacokinetic and uptake of four halogenated hydrocarbons from ^ the gastrointestinal tract of the rat. J. AodI. Toxiconl.' "3. 249-253. % ft,' , $ ~~Xot LiV. R^ Environ. Scl. Technol. 1987, 21, 366-370 _7) Chiou, C- T.; Schmedding, D. W. Environ. Sci. Technol. 1982.16, 4-10. (38) Watarai, H.; Tanaka, M.; Suzuki, N. Anal. Chem. 1982,54, 702-705. (39) Konemann, H.; van Leeuwen, K. Chemosphere 1980, 9, 3-19. (40) Chiou, C. T. Environ. Sci. Technol. 1985,19, 57-62. (41) Davies, R. P.; Dobbs, A. J. Water Res. 1984,18,1253-1262. (42) Garten, C. T., Jr.; Trabalka, J. R. Environ. Sci. Technol. 1983.17, 590-595. (43) Rapaport, R. A.; Elsenreich, S. J. Environ. Sci. Technol. 1984.18, 163-170. (44) Sugiura, K,; Ito, N.; Matsumoto, N.; Mihara, Y.; Murata, K; Tsukakoshi, Y.; Goto, M. Chemosphere 1978, 7,731-736. (45) Yalkowski, S. H.; Valvani, S. C.; Mackay, D. Residue Rev. 1983, 85, 43-55. (46) Mackay, D. Environ. Sci. Technol. 1982,16, 274-278. (47) Arbuckle, W. B. Environ. Sci. Technol. 1983,17,537-542. (48) Wasik, S. P.; Miller, M. M.; Tewari, Y. B.; May, W. E.; Sonnefeld, W. J.; DeVoe, H.; Zoller, W. H. Residue Rev. 1983, 85, 29-42. (49) Miller, M. M.; Ghodbane, S.; Wasik, S. P.; Tewari, Y, B.; Martire, D. E. J. Chem. Eng. Data 1984, 29, 184-190. (50) Yalkowsky, S. H.; Valvani, S. C. J, Pharm. Sci. 1980, 69, 912-922. (51) Bharath, A.; Mallard, C.; Orr, D.; Ozbum, G.; Smith, A. Bull. Environ. Contam. Toxicol. 1984, 33, 133-137. (52) Pearlman, R. S.; Yalkowsky, S. H.; Banerjee, S. J. Phys. Chem. Ref. Data 1984, 13, 555-562. (53) Muir, D. C. G.; Marshall, W. K.; Webster, G. R. B. Che mosphere 1985,14, 829-833. (54) Miller, M. M.; Wasik, S. P.; Huang, G.-L.; Shiu, W.-Y.; Mackay, D. Environ. Sci. Technol. 1985,19, 522-529. (55) Owens, J. W.; Wasik, S. P.; De Voe, H. J. Chem. Eng. Data 1986, 31, 47-51. Received for review June 9,1986. Accepted October 29,1986. Kinetics of the Depletion of Trichloroethene Gladys Barrio-Lage, ` Frances Z. Parsons, and Raja S. Nassar Drinking Water Research Center, Florida International University, Tamiami Campus, Miami, Florida 33199 The depletion of trichloroethene (TCE) was studied in microcosms containing water and three types of natural sediment ranging in composition from highly organic to a calcareous sedimentary rock. The depletion rates varied slightly in the different sediments. The first-order rate constant fcj for the depletion of TCE ranged from 8.7 x HT4 and 4.9 X HT* h"J in soils contaminated with TCE prior to microcosm preparation to 3.4 X 10~* and 4.6 X KT4 h_l for Boils with a large organic content to 3.2 X HT* h'1 for crushed rock microcosms. Depletion was found to follow nonlinear forms of the Michaelis-Menten kinetics in the organic sediments; however, microcosms containing crushed rock and water followed a linear form of the equation. Km values were found to be dependent on the percent of total organic carbon in the sediment. Introduction Several papers (1-6) have appeared in the literature recently reporting the biotransformation of trichloroethene (TCE) in different environments. All refer to the products of transformation but none to the kinetics of depletion. Kinetic activity is important in predicting the persistence of pollutants and their biotransformation products in the environment. Field studies using reclaimed waste water injected into an aquifer indicated that TCE exhibited half-life trans formation in 300 days (7). Vogel and McCarty (5) obtained 100% transformation of tetrachloroethene (PCE) and trichloroethene (TCE) to vinyl chloride (VC) after 10 days in a 110-cm anaerobic column operated under methanogenic conditions. This was an extremely fast transfor mation rate compared to that reported by others (2,4, 6, 8), which were done under conditions more closely simu lating the environment. The purpose of this work was to study the kinetics of the depletion of trichloroethene under different kinds of environmental conditions found in aquifers. The aquifers in southeastern Florida are composed of materials ranging from highly organic muck to calcareous sedimentary rock and offered the opportunity to study TCE transformation Table I. Biomass and Organic Content of Sediment0 organic content, % organic Cb biomass concn, nmol of lipid phosphate0 muck TV* Tie VB sand Tt T, rock 25 2 <1 1.1 X 10* 1.5 x 10* 2.1 x 10* 2.6 x 10* <1 "Reference 9. `Percent organics calculated by weight loss on ignition at 550 C (to avoid loss of C02 from carbonate sediments). "Biomass expressed as concentration of lipid phosphate; 50 jimol of lipid phosphate equals lg of microbial dry weight (ref 15). Biomass data from D. C. White and D. Hedrick, Florida State University, Tallahassee, FL. Each value represents the average of six bottles. * X0 through TK refer to incubation times; T0 = 24 h, Tt * 6 months, and Tle 2 years. in environmental materials having different compositions. One aquifer material, a silty sand, had previously been exposed to TCE through a major spill from a storage tank. This provided the opportunity to examine the adaptability of indigenous microorganisms to TCE transformation. Experimental Procedures Chemicals. Trichloroethylene (TCE), 98%, was from Aldrich Chemical Co., Milwaukee, WI. Preparation of Microcosms. Three types of sediment were obtained for the preparation of microcosms (see Table I). Organic muck was collected from the Everglades, a graminoid wetland that is the recharge basin for the Biscayne Aquifer in southern Florida. The second type of natural sediment, VB sand, is highly siliceous and was obtained from a site that was previously contaminated by a TCE spill in Vero Beach, FL. The rock used for the preparation of the third type of microcosm was obtained from the Everglades near the muck collection site because muck lies over the oolitic bed rock, which forms the un confined aquifer. Samples A and B in Table II are from two collection sites in the same area. Sample A was taken 366 Environ. Sci. Technol., Vol. 21, No. 4, 1987 0013-936X/87/0921-0366$01.50/0 1987 American Chemical Society SL 034523 from the bottom of a shallow canal and sample B from near ground-.surface of the canal bank. The rock was crushed, and all sediments were passed through a 6.34-mm sieve. Microcosms and controls were then prepared as previously described (8). Each microcosm was spiked to contain 5 mg/L of TCE with a solution of TCE in nitrogen-purged distilled water. The microcosms were spiked 2 weeks after construction to allow equilibration and oxygen depletion to occur inside the test and control bottles and thus simulate original conditions of the sample site. All microcosms and controls were allowed to incubate in the dark at 25 C for measured time periods of up to 6 months. Repeated sampling of a single microcosm in preliminary studies caused contami nation and a change in the volume of the contents and introduced a gas phase (head space). For this reason, replicate microcosms were constructed as described above, for each schedule test period, so that each microcosm was used only once per analysis. Although the microcosms were prepared homogeneously, variability of activity oc curred and was accounted for by calculating the mean of replicate runs. Viability of active sediments and sterility of the sterile controls and spiking solutions were determined by streaking these materials on plates of R^A medium (2,10) and observing development of microbial growth. Instrumentation. A Tracor Model 222 gas chromato graph with a 244 cm X 2.5 mm i.d. stainless steel column, packed with 60/80-mesh Tenax GC, and a Hall electrolytic conductivity detector Model 700, operating in the halogen mode, were employed for analysis. Nitrogen carrier gas at 40 mL/min and hydrogen reaction gas at 50 mL/min were supplied. The column oven was programmed to hold isothermal for 6 min at 40 C while 1-5 mL of microcosm contents (or dilution thereof) or standards were purged with N2 directly on the head of the column (11). The column temperature was then increased from 40 to 220 C at 8 deg/min. The temperature of the detector was kept at 850 20 C. The detection limit of this method is 0.1 jtg/L. Replicate samples were rerun in several cases to maintain a reproducibility better than 5%. Selected samples were verified by gas chromatogra phy/mass spectroscopy (GC/MS) on a Finnigan 4500 GC/MS system interfaced to a Tekman LSC-2 purgeand-trap system. The gas chromatographic column was a 6 ft X 2 mm i.d. 0.2% Carbowax 1500 on 80/100 Carbopack B column. Standard Solutions. Stock aqueous solutions of TCE were prepared at 500 mg/L (ppm), by volumetric dilution, similar to the spiking solutions. Aliquots of the stock solution were diluted with water to achieve solutions of the desired concentration. All bottles were wrapped in aluminum foil to avoid photodecomposition and kept at 4 C. The system was checked daily against standard solutions and was recalibrated when the deviation was greater than 3%. Results and Discussion Substrate depletion curves of TCE in several types of sediment and water microcosms were analyzed by linear and nonlinear forms of the Michaelis-Menten equation. Microcosms prepared with muck and water and those with VB sand and water showed a nonlinear depletion of TCE, as shown in Figures 1 and 2. These two sets of data points were analyzed by a method previously described by Duggleby and Morrison (12) for the analysis of progress curves by nonlinear regression. However, the microcosms pre pared with crushed rock and water showed a more linear shape of the curve, and it was analyzed by a linear form Figure 1. Depletion of TCE in muck and water microcosms. The symbols represent the experimental data; the solid line is the theoretical curve calculated from the kinetic parameters given in Table II. The initial concentration is Indicated on the ordinate. Figure 2. Depletion of TCE In VB sediment and water microcosms. The symbols represent the experimental data; the solid line is the theoretical curve calculated from the kinetic parameters given in Table II. of the Michaelis-Menten equation. Monod kinetics were considered; however, bacterial growth, if it occurs, is less than that which gives sigmoidal substrate depletion be havior. Michaelis-Menten enzyme kinetics is a "hyperbolic rate model" (13), which may apply to surface-catalyzed Environ. Sci. Technol., Vol. 21, No. 4. 1987 367 SL 03*524 Table 11. Kinetic Parameters Describing the Depletion of TCE in Sediment and Water Microcosms constants site A muck site B V, pM/h Km, aM Kp, /iM kit h'1* 1.1 x nr2 32.0 1.4 X nr2 3.4 X 1C4 1.9 x 10-* 60.0 4.1 x 101 4.5 X KT* `First-order rate constant obtained by V/Km. Vero Beach sand site A site B 1.3 x 10`2 14.0 7.8 8.7 x 10-4 5.0 X nr3 10.0 1.1 X 10`2 4.9 X nr4 reactions as well as enzyme kinetics {13), and has been found to be a good mathematical model to describe the depletion and formation of biodegradable organic com pounds (13,14). The progress curves shown in Figures 1 and 2 were an alyzed by the method described by Duggleby and Morrison (12). It is a procedure based on the Gauss-Newton method for nonlinear regression developed to analyze progress curve data. The rate equation that fits data to a one substrate/one product model, exhibiting competitive product inhibition, is rock, site B 2.4 X 10`3 7.4 3.2 X KT* where St and Pt represent the concentration of substrate and product, respectively, at time t, z is the expected amount of product formed during reaction (z = S0 - St), S0 is the initial substrate concentration, V is the maximum rate of substrate depletion, and Km and Kv are the half saturation constant and the product inhibition constant, respectively. B> use of Duggleby and Morrison's method (12), initial estimates of the parameters V, Ka, and Kp are obtained by casting the integrated rate equation into a linear form. The transformation used is one in which the integrated equation is divided by t and then z/t is. treated as the dependent variable as follows; (z/t) -V + KJ(l/t) In (1 - z/S0)] + (Km/Kp)(z/t + ((S0 + P0)/t) In (1 - x/S0)] (2) where P0 is the initial product concentration. The data are then fitted to this equation by multiple linear re gression treating the terms in square brackets as the in dependent variables, and V, Km, and Km/Kp as the pa rameters to be estimated. From this analysis, initial es timates of V, Km, and Kp are readily obtained by an it erative method that resulted in a reduction of the sum of squares due to errors in S (or z). When the initial esti mates of the parameters are close to their true value, it usually converges after three to five iterations. Figures 1 and 2 demonstrate the fitting of this method to the data obtained from the muck and the VB sand sediments, respectively. The symbols represent the ex perimental data, each being the average of replicate runs, while the solid lines are theoretical curves calculated from the kinetic parameters given in Table II. As shown in Table II, rock samples did not show a Kp value, because the data obtained from the rock and water microcosms were Unear when the substrate concentration was plotted vs. incubation time. Very good fitting was obtained (Figure 3) when the foUowing equation was used no calculate V and Km of rock microcosms: t/[ln (So/S)] = (1/V)[(S0 - S)/(ln (S0/S))] + Ka/V (3) Figure 3 shows the Michaelis-Menten fit of the data for 368 Environ. Scl. Technol.. Vol. 21. No. 4, 1987 iinim like I - .U7M?0DI Figure 3. Michaelis-Menten (eq 3) fit of the data from the depletion of TCE in crushed rock and water microcosms. the depletion of TCE in crushed rock microcosms. Km and V values reported in Table II for rock samples were ob tained from a linear regression analysis of the MichaelisMenten fit of the data and subsequently divided by the microcosm volume to obtain the reported values. It is believed that because crushed rock had a much simpler composition than the other sediments, and produced smaller concentration of cis-l,2-dichloroethene (CIS) (9), competitive product inhibition was not observed in the rock, as in the other sediments. As previously reported (9), Vero Beach sand, which had been exposed to a TCE spill prior to microcosm preparation, showed complete transformation of TCE to CIS, while muck and crushed rock sediments transformed less than 5% of the TCE to CIS (or less CIS was accumulated). The first order rate constants (kj) in Table II were ob tained from V/Km. Dividing eq 2 by Km, the first term on the right-hand side of the equation is V/Km. This term is the first-order rate constant (kj, which is significant at the beginning of the reaction when z is small and product inhibition effects are negligible. In Table II, k1 values for muck and VB sand are very similar, with VB sand having a slightly higher value than muck, and rock shows the smallest kh This supports expected results; that is, because Vero Beach sediment had been exposed to a TCE spill prior to this study, the microorganisms there were expected to be adapted and cause TCE transformation more rapidly than organisms in the unexposed rock and muck. Prea daptation of microorganisms to TCE as substrate increased SL 034525 Figure 4. Dependence of the Mchaefe-Menten constant, Km, on the percent organic content of sediments. the transformation rate slightly but not dramatically. Crushed rock, which contained the least microbial biomass and organic carbon (see Table I), showed the slowest transformation rate. Physical and chemical effects, including mass-transfer effects influenced by sorption/desorption of TCE on soils, were accounted for by comparison of values from active microcosms with those from sterile control microcosms. The concentration of TCE in sterile controls, throughout the incubation period, was found to remain almost con stant at 2450 440 pg/L, and thus changes that occurred in nonsterile microcosms were credited to microbial ac tivity. Very good agreement was obtained between V, Km, and ki shown in Table II for muck and the kinetic parameters previously reported (8) for the depletion of CIS, trans1,2-dichloroethene (TRANS), and 1,1-dichloroethene (1,1-DCE) in the same type of sediment and water mi crocosms. This indicates that ethenes with two or three chlorine atoms are depleted at similar rates in separate situations. Suflita et al. (14) indicated that the first-order decay rate for 3,5-dichlorobenzoate is slower than the rate obtained with 3-chlorobenzoate. We did not find a rate increase in the dichlorinated ethenes compared with the trichlorinated ethene; however, Sulflita's (14) results for both chlorinated species were obtained in sequence, in the same incubation vessel, and our results for both chlorinated species were obtained in separate incubation containers. In previous studies (8), Km values ranged from 29.6 to 65.3 pM for the depletion of CIS, TRANS, and 1,1-DCE in water and muck microcosms, Km values shown in Table II range from 32.5 to 49.8 pM for depletion of TCE in the same type of microcosms, which demonstrates very good agreement between the earlier (8) and present studies. As seen in Table II, Km values increased with increased or ganic content of sediments, and proportionality is apparent between percent organic carbon and Ka as illustrated in Figure 4. The chemical reaction that results in the de chlorination of TCE is dependent on the availability of electrons, hydrogen, and several mediators (16). The or ganic content of the sediments contributes to this milieu as shown by the positive relationship between organic content and Km. Conclusions The depletion of TCE in different sediments in micro cosms simulating groundwater environments can be de scribed by Michaelis-Menten kinetics. In microcosms where the sediments contained significant amounts of organic material, a nonlinear form of the equation gave the best fitting of data; however, crushed rock microcosms with low organic content were best described by a linear form of the equation. Muck and VB sand, which were very different in organic matter and microbial biomass, showed similar depletion rates, with VB sand depleting TCE at a slightly faster rate than the muck sediment. This indicates that the depletion rate of TCE, in general terms, depended not only on or ganic content but also on the microbial biomass. Crushed rock that contained little organic matter and microbial biomass also biotransformed TCE. As expected, it showed the smallest value for of the sediments studied. The depletion of TCE in the crushed rock microcosms was not a physical or chemical effect because sterilized control microcosms containing crushed rock did not show deple tion of TCE in the same incubation time. The kinetic parameters obtained for TCE in muck microcosms were similar to the kinetic parameters for cis-l,2-dichloroethene, trcms-l,2-dichloroethene, and 1,1dichloroethene previously reported (8) with the same kinds of microcosms. This indicates that chlorinated ethenes with two or three chlorines in separate situations deplete at similar rates. The Michaelis-Menten constant, Km, was observed to be dependent on the organic content of the sediment. Acknowledgments We thank J. F. Morrison and R. G. Duggleby from the Australian National University, Canberra, A.C.T., for making the procura progress curve analysis program available to us. We gratefully acknowledge the technical assistance of Jesus Escobar. Registry No. TCE, 79-01-6. Literature Cited (1) Parsons, F. Z.; Wood, P, R.; DeMarco, J. J.~Am. Water Works Assoc. 1984, 76, 56-59. (2) Parsons, F. Z.; Lage, G. B. J.--Am. Water Works Assoc. 1985, 77, 52-59. (3) Bouwer, E. J.; McCarty, P. L. Appl. Environ. Microbiol. 1983, 45, 1286-1294. (4) Kleopfer, R. D.; Easley, D. M.; Haas, B. B., Jr.; Deihl, T. G.; Jackson, D. C.; Wurrey, C. J. Environ. Sci. Technol. 1985, 19, 277-280. (5) Vogel, T. M.; McCarty, P. L. Appl. Environ. Microbiol. 1985, 49, 1080-1083. (6) Parsons, F.; Barrio-Lage, G.; Rice, R. J. Environ. Toxicol. Chem. 1985, 4, 739-742. (7) Roberts, P. V.; Schreinger, J. E.; Hopkins, G. C. Water Res. 1982, 16, 1025-1035. (8) Barrio-Lage, G.; Parsons, F. Z.; Nassar, R. S.; Lorenzo, P. A. Environ. Sci. Technol. 1985, 20, 96-99. (9) Parsons, F. Z.; Lage, G. B.; Lorenzo, P.; Nassar, R, Sixth Annual Meeting of the Society of Environmental Toxi cology and Chemistry, Nov 1985, St. Louis, MO. (10) Reasoner, D. J.; Geldreich, E. E. Appl. Environ. Microbiol. 1985, 49, 1-7. SL 034526 Environ. Sci. Technol., Vol. 21, No. 4, 1987 369 Environ. Sd. Technol. 1987. 21, 370-373 (11) Mehran, M. J. Chromatogr. Sei. 198S, 24, 546-548. (12) Duggleby, R. G-; Morrison, J. F. Biochim. Biophys. Acta 1977, 481, 297-312. (13) Goring, A. I.; Hamaker, J. W. Organic Chemicals in the Soil Environment; Dekker: New York, 1972; Chapter 4. (14) Sulflita, J. M.; Robinson, J. A.; Tiedje, J. M. Appl. Environ. Microbiol. 1983, 45,1466-1473. (15) White, D. C.; Davis, W. M.; Nickels, J. S.; King, J. D.; Bobbie, R. J. Oecologia 1979, 40, 51. (16) Kobayashi, H.; Rittmann, B. E. Environ. Sci. Technol. 1982, 16, 170A-183A. Received for review December 6, 1985. Revised manuscript received July 21,1986. Accepted November 3,1986. This work was supported by the U.S. Environmental Protection Agency, Ground Water Research Branch, under Contract CR809994-02 to the Florida State University and Subcontract 281308-500 to the Florida International University. Structure-Activity Relationships for Sorption of Linear Alkylbenzenesulfonates Vincent C. Hand* and Glenwood K. Williams Human and Environmental Safety Division, Procter and Gamble Company, Cincinnati, Ohio 45217 Sorption of ten radiolabeled linear alkylbenzenesulfonate (LAS) homologues and isomers onto four river sediments increased with increasing alkyl chain length and as the phenyl position approached the end of the chain. Initial solution concentrations of LAS were varied between 10 and 1000 ppb. The value of the sorption partition coefficient (JCd) increased by factor of 2.8 for each addi tional methylene group in the homologous series C10 LAS to Cj4 LAS. The value of Kd varied by 4 orders of mag nitude, from 3 to 26000 L/kg, as sediment type, chain length, and phenyl position were varied. Sorption and desorption were rapid (<8 h) and nearly reversible in a single-replacement desorption experiment. Introduction Sorption plays a significant role in the environmental fate and effects of materials released to the aqueous en vironment, largely determining the distribution of material between environmental compartments. Most systematic studies of sorption have focused on relatively hydrophobic organics (see ref 1 and references cited therein), which adsorb by partitioning into the organic phase of the sed iment. Recently, the sorption of polychlorinated phenols has been shown to be affected by the fraction of ionized species in solution, and thus by pH (2,3). We have studied the sorption of linear alkylbenzenesulfonates (LAS), e.g., 1, which are ionized under all environmentally relevant conditions. 1 Linear alkylbenzenesulfonate is used primarily as a surfactant in detergents and cleaning products. The commercial material consists of a complex mixture of homologues and isomers. Individual LAS homologues or isomers are identified by alkyl chain length and by the position of attachment of the benzenesulfonate on the chain. The mixture most frequently used in detergent formulations has an average chain length of 12-13 and a phenyl position between 2 and 6. Most LAS is disposed to sewage and removed during primary and secondary sewage treatment (4). Low, but measurable, concentrations of LAS have been reported in river and estuarine sediment (5, 6). Two previous laboratory studies (7,8) determined iso therms for parts-per-million mixtures of LAS sorbing to sediments. Both reports described a correlation between the sorption of LAS and the fraction of organic carbon on the sediment, suggesting a hydrophobic sorption mecha nism. We describe the sorption of a series of radiolabeled LAS homologues and isomers to four different sediments. This study differed from past studies on LAS sorption in the concentration range studied and in the systematic comparison of the sorption of a series of LAS homologues and isomers. With radiolabeled LAS, sorption measure ments were extended to environmentally relevant con centrations, 10-1000 ppb. The simplicity of radiochemical analyses facilitated the determination of a mass balance for each experiment. Systematic measurement of sorption for a series of homologues or isomers provided another probe of sorption mechanisms, which differed from the more conventional variation of sediment properties. Comparing physical and chemical properties among the members of a homologous or isomeric series is not new to the study of surfactancy (9) but has been rare in envi ronmental studies. The observed changes in sorption with changes in LAS structure were consistent with a hydrophobic sorption mechanism. Experimental Section Materials. Linear sodium alkyl[U-uC]benzenesulfonates were synthesized by New England Nuclear. Radiochemical purity was >97% as determined by thinlayer chromatography (TLC) and/or gas-liquid chroma tography (GLC). Isomer distributions and specific activ ities are described in Table I. Stock solutions were pre pared in alcohol (methanol/ethanol/2-propanol, 5:90:5) and stored at 4 C. Sediment characteristics are reported in Table II. Dry sediments EPA B1 and EPA 5, along with the corre sponding characterization data, were kindly provided by Dr. Samuel Karickhoff of the U.S. Environmental Pro tection Agency. Sediments RC 3 and RC 4 were collected from Rapid Creek, SD, 0.8 and 7 km downstream from the Rapid City sewage treatment plant, dried, and charac terized by standard methods at CTL Engineering, Co lumbus, OH. These sites correspond to sampling sites 1 and 2 in reference 5. Proximity to the sewage outfall causes the RC 3 sediment to appear as a black ooze con taining relatively high organic carbon. Natural waters (Table III) were centrifuged (lOOOOg, 40 min) before use to remove particles with diameters above 0.1pm, SL 034527 370 Environ. Sci. Technol., Vol. 21, No. 4, 1987 0013-936X/87/0921-0370S01.50/0 1987 American Chemical Society