- Email: [email protected]
Carcinogenic Activity of Dichloroacetic Acid and Trichloroacetic Acid in the Liver of Female B6C3F1 Mice
FUNDAMENTAL AND APPLIED TOXICOLOGY ARTICLE NO.
31, 192–199 (1996)
0091
Carcinogenic Activity of Dichloroacetic Acid and Trichloroacetic Acid in the Liver of Female B6C3F1 Mice MICHAEL A. PEREIRA1 Environmental Health Research and Testing, Inc., 2514 Regency Road, Lexington, Kentucky 40503; and Center for Environmental Medicine, Medical College of Ohio, 3000 Arlington Avenue, Toledo, Ohio 43614 Received July 31, 1995; accepted January 15, 1996
Carcinogenic Activity of Dichloroacetic Acid and Trichloroacetic Acid in the Liver of Female B6C3F1 Mice. PEREIRA, M. A. (1996). Fundam. Appl. Toxicol. 31, 192–199. The concentration–response relationships for the hepatocarcinogenic activity of dichloroacetic acid2 (DCA) and trichloroacetic acid (TCA), two contaminants of finished drinking water, were determined in female B6C3F1 mice. Dichloroacetic acid or trichloroacetic acid at 2.0, 6.67, or 20.0 mmol/liter was administered to the mice in the drinking water starting at 7 to 8 weeks of age and until sacrifice after 360 or 576 days of exposure. The relationships of the yield of foci of altered hepatocytes, hepatocellular adenomas, and hepatocellular carcinomas to the concentration of DCA and TCA in the water were best described by second-order and linear regressions, respectively. The liver-to-body weight ratio increased linearly for both DCA and TCA, as did the vacuolization of the liver induced by DCA. The foci of altered hepatocytes and tumors in the animals treated with DCA were predominantly eosinophilic and contained glutathione S-transferase-p (GST-p, over 80% of the lesions), while the tumors induced by TCA were predominantly basophilic and lacked GST-p, including all 11 hepatocellular carcinomas. Therefore, the carcinogenic activity of DCA and TCA appeared to differ both with respect to their dose–response relationship and to the characteristics of precancerous lesions and tumors. q 1996 Society of Toxicology
Dichloroacetic acid (DCA) and trichloroacetic acid (TCA) have been found in finished drinking water in concentrations ranging from 34 to 160 mg/liter (Uden and Miller, 1983; Krasner et al., 1989). These chloroacetic acids are formed as reaction byproducts during chlorine disinfection of water containing humic acids and other organic substances (Miller and Uden, 1983; Coleman et al., 1984). Herren-Freund et al. (1987) demonstrated that DCA and TCA in the drinking water of female B6C3F1 mice induced hepatocellular adenomas and carcinomas. Others have subsequently reported that 1 Present address: Center for Environmental Medicine, Medical College of Ohio, 3000 Arlington Ave., Toledo, OH 43614. Fax: (419) 381–3089. 2 Abbreviations used: DCA, dichloroacetic acid; GST-p, glutathione Stransferase-p; TCA, trichloroacetic acid.
0272-0590/96 $18.00 Copyright q 1996 by the Society of Toxicology. All rights of reproduction in any form reserved.
AID
FAAT 2144
/
6k0c$$$201
DCA and TCA also induced liver tumors in male and female B6C3F1 mice (DeAngelo et al., 1991; Bull et al., 1990; Anna et al., 1994). The U.S. EPA has used, in part, the results of these carcinogenesis bioassays in mouse liver and the radiomimetic linear multistage (LMS) model to calculate 1006 risk levels of 1.5 and 0.56 mmol/liter for DCA and TCA in drinking water, respectively. The mechanism for the hepatocarcinogenic activity of these two chloroacetic acids appears not to result from genotoxic activity, as in vitro and in vivo tests for genotoxicity have not demonstrated significant activity (Waskel, 1978; Rapson et al., 1980; Herbert et al., 1980; Chang et al., 1992). Also, DCA has been shown to induce liver tumors that did not contain a unique mutation spectrum in either the H- or K-ras oncogene, but rather mutations in the 61st codon of the H-ras oncogene common to spontaneous tumors in B6C3F1 mice, albeit in a smaller percentage of the tumors (Anna et al., 1994). DCA and TCA do induce the proliferation of peroxisomes in the liver of mice (Elcome, 1985; Elcome et al., 1985; Goldworthy and Popp, 1987; Odum et al., 1988), an activity shared with some other hepatocarcinogens. This paper reports the concentration–response relationship for the carcinogenic activity of DCA and TCA that was previously identified by us in female B6C3F1 mice (Herren-Freund et al., 1987). The concentration– response relationship and the characteristics of the lesions induced by DCA and TCA support different pathogeneses for the two chloroacetic acids. MATERIALS AND METHODS Animals and chemicals. VAF (viral antibody-free) female B6C3F1 mice at 5 to 6 weeks of age were purchased from Charles River Breeding Laboratories (Portage, MI). The animals were housed in an AAALACaccredited facility and in accordance with the U.S. Public Health Service ‘‘Guide for the Care and Use of Laboratory Animals.’’ The Institutional Animal Care and Use Committee at Environmental Health Research and Testing approved the experiment. A maximum of five mice/cage were housed using solid-bottom polycarbonate cages with stainless steel wirebar lids. The bedding consisted of SaniChip (hardwood) bedding purchased from P. J. Murphy Forest Products (Montville, NJ). The animals were provided ad libitum with rodent pelletized NIH-07 Open Formula diet (Ziegler Brothers, Gardners, PA) and deionized and filtered (0.2 mm) drinking water. The light cycle consisted of 14 hr of light and 10 hr of dark.
192
05-15-96 14:54:47
ftoxa
AP: Fund Tox
193
CARCINOGENIC ACTIVITY OF DCA AND TCA Experimental design. When the mice were 7–8 weeks of age, they started to receive in their drinking water either DCA or TCA at 2.0, 6.67, or 20.0 mmol/liter and neutralized with sodium hydroxide to pH 6.5–7.5. The control group received 20.0 mmol/liter sodium chloride (control for sodium salt). The DCA was purchased from Eastman Kodak Co. (Rochester, NY) and the TCA and sodium chloride from Fisher Scientific Co. (Pittsburgh, PA). The concentrations of DCA and TCA were chosen so that the high concentration was comparable to those previously used by us to demonstrate carcinogenic activity in female B6C3F1 mice (Herren-Freund et al., 1987). The chloroacetic acid solutions were administered continuously to the mice until they were sacrificed by carbon dioxide asphyxiation after either 360 or 576 days of exposure. An exception was the group of mice exposed to 20.0 mmol/liter DCA in a 72-day cycle consisting of 24 days of exposure followed by 48 days without exposure. The 72-day cycle was repeated until the animals were sacrificed which resulted in the same total dose as that received by the animals continuously exposed to 6.67 mmol/liter, thus allowing comparison of the carcinogenic activity of the same total dose administered by continuous exposure to that of an intermittent exposure. The 24 days of exposure in the intermittent group was chosen to ensure that any enhancement of cell proliferation induced by DCA had ceased (see Results). The number of animals in each treatment group at the start of the study is presented in Table 1. Drinking water consumption and body weights were monitored during the first 4 weeks of the experiment and monthly thereafter. The animals were sacrificed after 360 or 576 days of exposure so that the sacrifice occurred at the end of either five or eight cycles of intermittent exposure to DCA. The terminal sacrifice for the experiment occurred at 576 days when it became apparent that liver tumors as indicated by a distended abdomen were present and that some of the animals receiving the high concentration of TCA were becoming moribund. Thus, it appeared that these animals could not survive another 72 days required to complete a cycle of intermittent exposure to DCA. During the necropsy, the liver was excised, weighed, and examined for lesions. The whole liver was cut into Ç3-mm blocks and along with representative sections of the visible lesions fixed in 10% neutral phosphate-buffered formalin for approximately 18 hr, transferred to 70% alcohol, and within 48 hr embedded in paraffin. The blocks were sectioned at 5 mm and stained with hematoxylin and eosin for histopathological evaluation for foci of altered hepatocytes, hepatocellular adenomas, and hepatocellular carcinomas. The slides were evaluated blind, being identified only by the study number and the random number assigned to the animal. Foci of altered hepatocytes in this study contained six or more cells and hepatocellular adenomas were distinguished from foci by the occurrence of compression at greater than 80% of the border of the lesion. The extent of vacuolation in the liver of DCA-treated mice was scored using a 0 to /3 scale, with 0 indicating the absence of vacuolation, /1 indicating vacuolated hepatocytes in the periportal zone, /2 indicating distribution of vacuolated hepatocytes in the midzone, and /3 indicating maximum vacuolation of the hepatocytes throughout the liver. Cell proliferation and glutathione S-transferase-p. Cell proliferation was determined in treatment groups containing 10 mice each and exposed to either DCA or TCA for 5, 12, or 33 days. Five days prior to sacrifice, mice were implanted with an osmatic minipump (Alzet Corp., Palo Alto, CA) containing 30 mg/ml bromodeoxyuridine (BrDU) and designed to deliver 1 ml/hr. Tissue blocks were sectioned at 5 mm and immunohistochemically stained for BrDU incorporation by the procedure of Sugihara et al. (1986) using monoclonal rat anti-BrDU antibody (Accurate Chemical and Scientific Corp., Westbury, NY). At least 2000 hepatocytes/mouse were evaluated for BrDU-labeled and unlabeled nuclei and the BrDU-labeling index was calculated as the percentage of hepatocytes with labeled nuclei. Glutathione S-transferase-p (GST-p) was detected by the avidin–biotin peroxidase complex method employing rabbit NCL–GST-p from human chronic lymphoblast leukemic spleen (Novocastra Laboratory, Ltd., Newcastle upon Tyne, UK), goat blocking serum, Vector kit PK-4001 and counterstaining with hematoxylin. Serial sections were stained with hema-
AID
FAAT 2144
/
6k0c$$$202
05-15-96 14:54:47
toxylin and eosin. The controls for nonspecific staining included (1) nonimmunized rabbit serum used instead of the rat anti-BrDU antibody or the rabbit NCL–GST-p and (2) the performance of the staining procedures without the primary or the secondary antibody. None of these negative controls demonstrated significant staining.
RESULTS
Drinking Water Consumption and Body and Liver Weights The drinking water consumption during the first 4 weeks of exposure to the chloroacetic acids is presented in Table 1. During the first week of exposure, only the high concentration group of both DCA and TCA exhibited a decrease in drinking water consumption. Thereafter, no effect of exposure to either DCA or TCA including the monthly monitoring of consumption (data not given) was observed except for a sporadic decrease or increase in a single treatment group. The body weights of the animals during the course of the study are presented in Fig. 1. The body weights of the mice that received 20.0 mmol/liter DCA were reduced after 35 weeks of exposure, with neither of the two other DCA treatment groups demonstrating an effect. After 51 weeks of exposure to 20.0 mmol/liter TCA, there was a trend of reduced body weights with some of the times monitored being statistically significant. The two lower concentrations of TCA did not effect the body weight of the animals. Figure 2 presents the concentration relationships of the liver-tobody weight ratio obtained after 360 and 576 days of exposure. The increase in liver-to-body weight ratio was linearly related to the concentration of DCA or TCA in the drinking water; the linear regression coefficients (r) were equaled to 0.9917 and 0.9925 at 360 days and 0.9994 and 0.9861 at 576 days, respectively. DCA exerted a greater effect than TCA upon the ratio. The low number of tumors at 360 days and the fact that the liver-to-body weight ratio did not increase with the increase in tumor response between 360 and 576 days argues against the greater ratio in DCA-treated mice being due to an increased tumor burden. Upon histopathologic evaluation, the liver from the animals treated with DCA contained vacuolated hepatocytes. The extent of vacuolation of the hepatocytes in the mice administered 0, 2.0, 6.67, or 20.0 mmol/liter DCA was scored as 0.0, 0.80 { 0.08, 2.32 { 0.11, or 2.95 { 0.05, respectively, using the scale of 0 to /3. Hence, all three concentrations of DCA resulted in vacuolation of hepatocytes. Foci and Tumor Response Table 2 contains the incidence (number and percentage of animals with a lesion) and Table 3 the yield (lesions/ mouse) of foci of altered hepatocytes, hepatocellular adenomas, and hepatocellular carcinomas in mice administered DCA or TCA. After 360 days of exposure, in mice adminis-
ftoxa
AP: Fund Tox
194
MICHAEL A. PEREIRA
TABLE 1 Drinking Water Consumption in Mice Administered DCA, TCA, or NaCl Group: treatment mmol/liter 1. 2. 3. 4. 5. 6. 7.
Weeks Na
20.0 DCA 6.67 DCA 2.0 DCA 20.0 TCA 6.67 TCA 2.0 TCA 20.0 NaCl
1
40 50 90 38 46 93 134
4.23 5.46 4.82 4.38 4.72 4.58 5.10
{ { { { { { {
0.35b,* 0.20 0.16 0.17* 0.12 0.17 0.20
2 5.56 5.46 4.59 4.78 5.05 4.69 5.00
{ { { { { { {
3 1.12 0.20 0.08 0.16 0.12 0.11 0.18
4.47 5.72 4.36 5.06 4.69 4.49 4.89
{ { { { { { {
4 0.26 0.80 0.12 0.25 0.14 0.13 0.10
4.99 5.55 5.29 5.35 5.65 5.14 4.94
{ { { { { { {
0.54 0.26 0.23 0.24 0.22* 0.19 0.12
a
N indicates the number of animals at the start of the study. Results are means { SE expressed as ml/mouse/day. * Significantly different from Group 7 (NaCl vehicle) by an ANOVA followed by a pair test with Bonferroni’s correction, p value õ0.05. b
tered DCA, the incidence and yield of foci of altered hepatocytes and hepatocellular adenomas were increased only by 20.0 mmol/liter, while after 576 days of exposure 6.67 mmol/liter also increased the yield of foci and adenomas. The higher concentration of DCA at 576 days also increased the incidence and yield of hepatocellular carcinomas. In the group of mice that received intermittent exposure to DCA, i.e., 72-day cycle consisting of 24 weeks of 20.0 mmol/liter DCA followed by 48 days without exposure, the only effect observed was an increase in foci of altered hepatocytes after 576 days. There was no significant difference between the
treatment group that intermittently received 20.0 mmol/liter DCA and the group that received the same total exposure of DCA but administered continuously as 6.67 mmol/liter. With respect to TCA, after 360 days the high concentration group (20.0 mmol/liter) contained an increase in hepatocellular carcinomas, with no other effect observed in this or the other two TCA treatment groups. Treatment for 576 days with 20.0 mmol/liter TCA resulted in an increase in foci of altered hepatocytes, hepatocellular adenomas, and hepatocellular carcinomas, while 6.67 mmol/liter resulted in an increase in foci and carcinomas. No other significant effects
FIG. 1. Body weights for female B6C3F1 mice administered either DCA (A) or TCA (B) in the drinking water. Results are means { SE with * indicating p value õ0.05 by a one-way analysis of variance followed by the Dunnett’s test (Dunnett, 1955).
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
ftoxa
AP: Fund Tox
195
CARCINOGENIC ACTIVITY OF DCA AND TCA
FIG. 2. Liver-to-body weight ratio versus the concentration of DCA and TCA in the drinking water. (A) Sacrifice after 360 days of exposure. (B) Sacrifice after 576 days of exposure. Results are means { SE with linear regression lines obtained using Sigma Plot for Windows (Jandel Scientific, Erkrath, Germany).
were obtained for the foci and tumor response in mice treated with TCA. The yield of total lesions (foci / adenomas / carcinomas) versus the concentration of DCA or TCA in drinking water is presented graphically in Fig. 3. Regression analysis indicated that the response of total lesions was linearly related to the concentration of TCA with a correlation coefficient of
0.9979. The correlation coefficient for the regression analysis of the total lesions produced by DCA was only 0.9627 for first order and 0.9998 for second order. Therefore, Fig. 3 presents the second-order regression line for the total lesion response to DCA. The dose–response relationship for the yield of tumors (adenomas / carcinomas) was also linear for TCA and second order for DCA (data given in Table 3).
TABLE 2 Incidence of Foci and Tumors in Mice Administered DCA or TCAa N Group: treatment mmol/liter 1. 2. 3. 4. 5. 6. 7. 8.
20.0 DCA Intermittent DCAc 6.67 DCA 2.0 DCA 20.0 TCA 6.67 TCA 2.0 TCA 20.0 NaCl
Foci
360
576
360
20 15 20 40 20 19 40 40
19 34 28 50 18 27 53 90
8 (40.0)b,* 0 1 (5.0) 0 0 0 3 (7.5) 0
Adenomas 576 17 14 11 7 11 9 10 10
360
(89.5)* (41.2)* (39.3)* (14.0) (61.1)* (33.3)* (18.9) (11.1)
7 (35)* 0 3 (15) 0 2 (10) 3 (15.8) 3 (7.5) 1 (2.5)
Carcinomas 576
16 3 7 3 7 3 4 2
(84.2)* (8.8) (25.0)* (6.0) (38.9)* (11.1) (7.6) (2.2)
360
576
1 (5) 0 0 0 5 (25)* 0 0 0
5 (26.3)* 1 (2.9) 1 (3.6) 0 5 (27.8)* 5 (18.5)* 0 2 (2.2)
a Mice were administered DCA, TCA, or NACl in their drinking water starting at 7–8 weeks of age. The mice continued to receive the test agents until sacrificed after either 360 or 576 days of exposure. b The results indicate the number of animals with lesions (foci of altered hepatocytes, hepatocellular adenomas or hepatocellular carcinomas) with the number in parentheses being the percentage of animals. c The mice received 20.0 mmol/liter DCA in their drinking water starting at 7–8 weeks of age and for a total of 24 days followed by 48 days without exposure. The 72-day cycle was repeated until the mice were euthanized. * p value õ 0.05 by Fisher’s exact test.
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
ftoxa
AP: Fund Tox
196
MICHAEL A. PEREIRA
TABLE 3 Yield of Liver Lesions/Mouse in DCA- or TCA-Administered Animalsa N Group: treatment mmol/liter 1. 2. 3. 4. 5. 6. 7. 8.
20.0 DCA Intermittent DCA 6.67 DCA 2.0 DCA 20.0 TCA 6.67 TCA 2.0 TCA 20.0 NaCl
Foci/mouse
360
576
360
20 15 20 40 20 19 40 40
19 34 28 50 18 27 53 90
0.60 { 0.22b,** 0 0.05 { 0.05 0 0 0 0.08 { 0.04 0
Adenoma/mouse 576
7.95 0.47 0.39 0.14 1.33 0.41 0.26 0.11
{ { { { { { { {
2.00** 0.11* 0.11* 0.05 0.31** 0.13* 0.08 0.03
360
Carcinoma/mouse
576
0.45 { 0.17** 0 0.20 { 0.12 0 0.15 { 0.11 0.21 { 0.12 0.08 { 0.04 0.03 { 0.03
5.58 0.09 0.32 0.06 0.61 0.11 0.08 0.02
{ { { { { { { {
1.14** 0.05 0.13* 0.03 0.22* 0.06 0.04 0.02
360
576
0.10 { 0.10 0 0 0 0.50 { 0.18* 0 0 0
0.37 { 0.17* 0.03 { 0.03 0.04 { 0.04 0 0.39 { 0.16* 0.22 { 0.10* 0 0.02 { 0.02
a Mice were administered DCA, TCA, or NaCl in their drinking water starting at 7–8 weeks of age. The mice continued to receive the test agents until sacrificed after either 360 or 576 days of exposure. b The results indicate the number of animals with lesions (foci of altered hepatocytes, hepatocellular adenomas, or hepatocellular carcinomas)/mouse / SE. Results are significantly different from Group 8 by the Kruskal–Wallis test: *p value õ0.05; **p value õ0.01.
The foci of altered hepatocytes and the tumors obtained from this study were basophilic, eosinophilic, or mixed containing both characteristics (Table 4). DCA induced a predominance of eosinophilic foci and tumors, with over 80% of the foci and 90% of the tumors in the 6.67 and 20.0 mmol/liter concentration groups being eosinophilic. Only approximately half of the lesions were characterized as eo-
sinophilic with the rest being basophilic, in both the treatment group intermittently exposed to DCA and the group administered 2.0 mmol/liter DCA. The percentage of the lesions characterized as eosinophilic in the group intermittently exposed to DCA was significantly less than in the groups that were administered either 6.67 or 20.0 mmol/liter DCA (p value õ0.05). The eosinophilic foci and tumors consistently stained immunohistochemically for the presence of GST-p, while basophilic lesions did not stain for GSTp, except for a few scattered cells or small areas comprising less than 10% of the lesion. The foci of altered hepatocytes in the TCA treatment groups were approximately equally distributed between basophilic and eosinophilic in tincture. However, the tumors were predominantly basophilic lacking GST-p (21 of 28, 75%) including all 11 hepatocellular carcinomas. Since a higher percentage of tumors compared to foci were basophilic, it would suggest that in TCA-treated animals basophilic foci compared to eosinophilic foci have a greater tendency to progress tumors. The limited numbers of lesions, i.e., 14, in the sodium chloride (vehicle control) group were characterized as 64.3, 28.6, and 7.1% basophilic, eosinophilic, and mixed, respectively. Cell Proliferation
FIG. 3. Yield of total lesions (foci of altered hepatocytes / hepatocellular adenomas / carcinomas) versus the concentration of DCA and TCA after 576 days of exposure. A second-order regression line is drawn for DCA and a linear regression line for TCA. Results are means { SE and Sigma Plot for Windows (Jandel Scientific, Erkrath, Germany) was used to obtain the regression lines.
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
The BrDU-labeling index in DCA- and TCA-treated mice is presented in Figs. 4A and 4B, respectively. Five days of exposure to DCA resulted in a concentration-dependent increase in the BrDU-labeling index. However, after 12 and 33 days of exposure to DCA, the BrDU-labeling index was no longer altered, although at 12 days there was a nonsignificant trend for an increase in the BrDU-labeling index. TCA also increased the BrDU-labeling index only after 5 days of exposure with all three concentrations producing approximately the same increase.
ftoxa
AP: Fund Tox
197
CARCINOGENIC ACTIVITY OF DCA AND TCA
TABLE 4 Characterization of Foci and Tumors in Mice Administered DCA or TCA Tumorsa
Foci Group: treatment mmol/liter 1. 2. 3. 4. 5. 6. 7. 8.
20.0 DCA Intermittent DCA 6.67 DCA 2.0 DCA 20.0 TCA 6.67 TCA 2.0 TCA 20.0 NaCl a b
N
Basophilic
Eosinophilic
Mixed
N
Basophilic
Eosinophilic
Mixed
150 16 11 7 22 11 13 10
3.3b 56.3 18.2 42.8 36.4 45.5 38.5 70.0
96.7 43.7 81.8 57.2 54.6 54.5 61.5 30.0
0 0 0 0 9.1 0 0 0
105 4 10 3 18 6 4 4
2.9 75.0 10.0 0 61.1 100 100 50.0
96.1 25.0 90.0 100 22.2 0 0 25.0
1.0 0 0 0 16.7 0 0 25.5
Tumors include both hepatocellular adenomas and hepatocellular carcinomas. Results are the percentages possessing the characteristics for the number of foci or turmos indicated under N, respectively.
DISCUSSION
Dichloroacetic acid and trichloroacetic acid are found as byproducts of chlorination in finished drinking water at concentrations of 34 to 160 mg/liter (Uden and Miller, 1983; Krasner et al., 1989). The chloroacetic acids are also metabolites of trichloroethylene and tetrachloroethylene (Elcome, 1985; Goldworthy and Popp, 1987; Odum et al., 1988; Dekant et al., 1984), two widely used industrial solvents that are common contaminants of surface water, groundwater, and drinking water, as well as soil and lysates of landfill hazardous waste disposal sites (Coleman et al., 1976; Con-
glio et al., 1980; Westrick et al., 1984). Exposure of B6C3F1 mice to DCA or TCA in the drinking water has been shown to produce liver tumors (Herren-Freund et al., 1987; DeAngelo et al., 1991). The U.S. EPA using a LMS model has applied these previous results to calculate 1006 risk levels of 1.5 and 0.56 mmol/liter for DCA and TCA in drinking water, respectively. Determination of the dose–response relationship for the carcinogenic activity of DCA and TCA would improve the estimation of their carcinogenic hazard to humans. The tumorigenic activity of TCA exhibited a linear relationship to its concentration in drinking water, in contrast
FIG. 4. BrDU-labeling index in 2000 hepatocytes from mice exposed to DCA (A) or TCA (B). The results are means { SE for groups of 10 mice each with * indicating p õ 0.05 (an ANOVA followed by a paired test with Bonferroni’s correction).
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
ftoxa
AP: Fund Tox
198
MICHAEL A. PEREIRA
to the second-order relationship exhibited by DCA. In male B6C3F1 mice, DeAngelo et al. (1991) also appeared to have observed an exponential dose–response relationship for DCA, in which 26.9 but not 3.84 mmol/liter increased the yield of liver tumors. Their reported tumor response in male mice of 4.0 tumors/animals after 420 days of exposure to 26.9 mmol/liter DCA (DeAngelo et al., 1991) was similar to the 5.95 tumors/animals observed by us in female mice exposed to 20.0 mmol/liter. Thus, female and male mice would appear to be of similar sensitivity to DCA. The exponential increase in the tumorigenic potency of DCA did not correlate with the linear relationships for DCA (1) induced increase in the liver-to-body weight ratio; (2) enhancement of the BrDU-labeling index after 5 days of exposure; and (3) induced vacuolation of the liver that stained with PAS, indicating an accumulation of glycogen. This would suggest that the dose of DCA to the liver does not increase exponentially with its concentration in the drinking water. DCA administered intermittently as a 72-day cycle consisting of 24 days of exposure to 20.0 mmol/liter followed by 48 days without exposure resulted in a very low yield of foci and tumors. This group was included in the study in order to determine whether the carcinogenic activity of intermittent exposure to DCA is equal to the activity of the same total dose given continuously. However, both the intermittent exposure to 20.0 mmol/liter DCA and the continuous exposure to 6.67 mmol/liter DCA induced only a very low yield of foci and tumors. The total exposure to DCA in these two treatment groups was one-third that of the group continuously exposed to 20.0 mmol/liter, so that it was expected that the yield of foci and tumors would be one-third that of the 20.0 mmol/liter treatment group. Instead, the yield of these lesions was less than 1/20th of the group exposed continuously to 20.0 mmol/liter. One possible explanation for the low tumorigenic potency of intermittent exposure to 20.0 mmol/liter DCA is that the lesions induced by DCA were not stable but regress during the 48 days in which the animals were not exposed to DCA. We have unpublished evidence that foci of altered hepatocytes and hepatocellular adenomas initiated by methylnitrosourea and promoted by DCA regress upon termination of exposure to DCA. Thus, at the beginning of the next exposure to DCA, the pathogenesis of the lesions would commence with less advanced lesions than those present when the previous exposure ceased. The characteristics of the foci and tumors induced by DCA and TCA further suggest different mechanisms of action. The foci and tumors induced by DCA were greater than 80% eosinophilic and contained GST-p; however, the tumors induced by TCA were predominantly basophilic lacking GST-p. The characteristics of foci and tumors of mouse liver have been previously reported to be carcinogen specific (Ward et al., 1983; Dragani et al., 1985; Weghorst and Klaunig, 1989; Hatayama et al., 1993). Liver tumors in
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
mice not administered a carcinogen or those induced by diethylnitrosamine have been reported to be predominantly basophilic and lacking GST-p (Ward et al., 1983; Dragani et al., 1985; Weghorst and Klaunig, 1989; Hatayama et al., 1993). When diethylnitrosamine-initiated mice were promoted with phenobarbital or 1,4-bis[2-(3,5-dichloropyridyloxy)]benzene (TCPOBOP), the resulting foci and tumors were then predominantly eosinophilic (Ward et al., 1983; Dragani et al., 1985; Weghorst and Klaunig, 1989; Hatayama et al., 1993; Diwan et al., 1992) and in female but not male mice promoted with phenobarbital contained GST-p (Hatayama et al., 1993). Chronic administration to mice of phenobarbital (Evans et al., 1992) or nitrofen (Hoover et al., 1980) also resulted in eosinophilic tumors. In contrast, promotion of diethylnitrosamine-initiated mice with di(2-ethylhexyl)phthalate, a peroxisome proliferator, resulted in foci of altered hepatocytes and tumors that were basophilic (Ward et al., 1983; Standeven and Goldworthy, 1993). Marsman and Popp (1994) have shown in rats that peroxisome proliferators also induce basophilic foci and tumors. Both DCA and TCA are peroxisome proliferators in mice and rats (Elcome, 1985; Elcome et al., 1985; Goldworthy and Popp, 1987; Odum et al., 1988). In summary, the basophilic staining of the tumors induced by TCA is consistent with other peroxisome proliferators, while the eosinophilic staining of those induced by DCA is not consistent. Furthermore, the differences in the characteristics of the tumors induced by DCA and TCA as well as the different shape of their concentration–response relationships would suggest different pathogeneses and mechanisms. ACKNOWLEDGMENTS This study was supported in part by a grant from the American Water Works Association Research Foundation. I thank Dr. Ronald A. Lubet for his helpful suggestions during the preparation of the manuscript.
REFERENCES Anna, C. H., Maronpot, R. R., Pereira, M. A., Foley, J. F., Malarkey, D. E., and Anderson, M. W. (1994). Ras proto-oncogene activation in dichloroacetic acid-, trichloroethylene- and tetrachloroethylene-induced liver tumors in B6C3F1 mice. Carcinogenesis 15, 2255–2261. Bull, R. J., Sanchez, I. M., Nelson, M. A., Larson, J. L., and Lansing, A. J. (1990). Liver tumor induction in B6C3F1 mice by dichloroacetate and trichloroacetate. Toxicology 63, 341–359. Chang, L. W., Daniel, F. B., and DeAngelo, A. B. (1992). Analysis of DNA strand breaks induced in rodent liver in vivo, hepatocytes in primary culture, and a human cell line by chlorinated acetic acids and chlorinated acetaldehydes. Environ. Mol. Mutagen. 20, 277–288. Coleman, W. E., Lingg, R. D., Melton, R. G., and Kopfler, F. C. (1976). The occurrence of volatile organics in five drinking water supplies using gas chromatography/mass spectroscopy. In Identification and Analysis of Organic Polutants of Water (L. H. Keith, Ed.), pp. 305–327. Ann Arbor Sciences Publ., Ann Arbor, MI. Coleman, E. W., Munch, J. W., Kaylor, W. H., Streicher, R. P., Ringhand,
ftoxa
AP: Fund Tox
CARCINOGENIC ACTIVITY OF DCA AND TCA H. P., and Meier, J. R. (1984). Gas chromatography/mass spectroscopy analysis of mutagenic extracts of aqueous chlorinated humic acid. A comparison of the byproducts to drinking water contaminants. Environ. Sci. Technol. 18, 674–678. Conglio, W. A., Miller, K., and Mackeever, D. (1980). The Occurrence of Volatile Organics in Drinking Water. Criteria and Standards Division, Science and Technology Branch, Environmental Protection Agency, Washington, DC. DeAngelo, A. B., Daniel, F. B., Stober, J. A., and Olson, G. R. (1991). The carcinogenicity of dichloroacetic acid in the male B6C3F1 mouse. Fundam. Appl. Toxicol. 16, 337–347. Dekant, W., Metzler, M., and Henschler, D. (1984). Novel metabolites of trichloroethylene through dechlorination reactions in mice and humans. Biochem. Pharmacol. 33, 2021–2027. Diwan, B. A., Lubet, R. A., Ward, J. M., Hrabie, J. A., and Rice, J. M. (1992). Tumor-promoting and hepatocarcinogenic effects of 1,4-bis[2(3,5-dichloropyridyloxy)]benzene (TCPOBOP) in DBA/2NCr and C57BL/6NCr mice and an apparent promoting effect on nasal cavity tumors but not on hepatocellular tumors in F344/NCr rats initiated with N-nitronitrosodiethylamine. Carcinogenesis 13, 1893–1901. Dragani, T. A., Manenti, G., Galliani, G., and Della Porta, G. (1985). Promoting effects of 1,4-bis[2-(3,5-dichloropyridyloxy)]benzene in mouse hepatocarcinogenesis. Carcinogenesis 6, 225–228. Dunnett, C. W. (1955). A multiple comparison procedure for comparing several treatments with a control. J. Am. Stat. Assoc. 50, 1096–1121. Elcome, C. R. (1985). Species differences in carcinogenicity and peroxisome proliferation due to trichloroethylene: A biochemical human hazard assessment. Arch. Toxicol. Suppl. 8, 6–17. Elcome, C. R., Rose, M. S., and Pratt, I. S. (1985). Biochemical, histological, and ultrastructural changes in rat and mouse liver following the administration of trichloroethylene: Possible relevance to species differences in hepatocarcinogenicity. Toxicol. Appl. Pharmacol. 79, 365–376. Evans, J. G., Collins, M. A., Lake, B. G., and Butler, W. H. (1992). The histology and development of hepatic nodules and carcinoma in C3H/ He and C57BL/6 mice following chronic phenobarbitone administration. Toxicol. Pathol. 20, 585–594. Goldworthy, T. L., and Popp, J. A. (1987). Chlorinated hydrocarbon-induced peroxisomal enzyme activity in relation to species and organ carcinogenicity. Toxicol. Appl. Pharmacol. 88, 225–233. Hatayama, I., Nishimura, S., Narita, T., and Sato, K. (1993). Sex-dependent expression of class pi glutathione S-transferase during chemical hepatocarcinogenesis in B6C3F1 mice. Carcinogenesis 14, 537–538. Herbert, V., Gardner, A., and Coleman, N. (1980). Mutagenicity of dichlor-
AID
FAAT 2144
/
6k0c$$$204
05-15-96 14:54:47
199
oacetate, an ingredient of some formulations of pangamic acid (tradename ‘‘vitamin B15’’). Am. J. Clin. Nutr. 33, 1179–1182. Herren-Freund, S. L., Pereira, M. A., Khoury, M. D., and Olson, G. (1987). The carcinogenicity of trichloroethylene and its metabolites, trichloroacetic and dichloroacetic acid, in mouse liver. Toxicol. Appl. Pharmacol. 90, 183–189. Hoover, K. L., Ward, J. M., and Stinson, S. F. (1980). Histopathologic differences between liver tumors in untreated (C57BL/6 1 C3H)F1 (B6C3F1 ) mice and nitrofen-fed mice. J. Natl. Cancer Inst. 65, 937– 948. Krasner, S. W., McGuire, M. J., Jacangelo, J. G., Patania, N. L., Reagen, K. M., and Ajeta, E. M. (1989). The occurrence of disinfection byproducts in U.S. drinking water. J. Am. Water Works Assoc. 81, 41–53. Marsman, D. S., and Popp, J. A. (1994). Biological potential of basophilic hepatocellular foci and hepatic adenoma induced by the peroxisome proliferator, Wy-14,643. Carcinogenesis 15, 111–117. Miller, J. W., and Uden, P. C. (1983). Characterization of nonvolatile aqueous chlorination products of humic substances. Environ. Sci. Technol. 17, 150–157. Odum, J., Green, T., Foster, J. R., and Hext, P. M. (1988). The role of trichloroacetic acid and peroxisome proliferation in the differences in carcinogenicity of perchloroethylene in the mouse and rat. Toxicol. Appl. Pharmacol. 92, 103–112. Rapson, W. H., Nazar, M. A., and Butsky, V. V. (1980). Mutagenicity produced by aqueous chlorination of organic compounds. Bull. Environ. Contam. Toxicol. 24, 590–597. Standeven, A. M., and Goldworthy, T. L. (1993). Promotion of preneoplastic lesions and induction of CYP2B by unleaded gasoline vapor in female B6C3F1 mouse liver. Carcinogenesis 14, 2137–2141. Sugihara, H., Hattori, T., and Fukuda, M. (1986). Immunohistochemical detection of bromodeoxyuridine in formalin-fixed tissues. Histochemistry 85, 193–195. Uden, P. C., and Miller, J. W. (1983). Chlorinated acids and chloral in drinking water. J. Am. Water Works Assoc. 75, 524–527. Ward, J. M., Rice, J. M., Creasia, D., Lynch, P., and Riggs, C. (1983). Dissimilar patterns of promotion by di(2-ethylhexyl)phthalate and phenobarbital of hepatocellular neoplasia initiated by diethylnitrosamine in B6C3F1 mice. Carcinogenesis 4, 1021–1029. Waskel, L. (1978). Study of the mutagenicity of anesthetics and their metabolites. Mutat. Res. 57, 141–153. Weghorst, C. M., and Klaunig, J. E. (1989). Phenobarbital promotion in diethylnitrosamine-initiated infant B6C3F1 mice: Influence of gender. Carcinogenesis 10, 609–612. Westrick, J. J., Mello, J. W., and Thomas, R. F. (1984). The groundwater supply survey. J. Am. Water Works Assoc. 76, 52.
ftoxa
AP: Fund Tox
31, 192–199 (1996)
0091
Carcinogenic Activity of Dichloroacetic Acid and Trichloroacetic Acid in the Liver of Female B6C3F1 Mice MICHAEL A. PEREIRA1 Environmental Health Research and Testing, Inc., 2514 Regency Road, Lexington, Kentucky 40503; and Center for Environmental Medicine, Medical College of Ohio, 3000 Arlington Avenue, Toledo, Ohio 43614 Received July 31, 1995; accepted January 15, 1996
Carcinogenic Activity of Dichloroacetic Acid and Trichloroacetic Acid in the Liver of Female B6C3F1 Mice. PEREIRA, M. A. (1996). Fundam. Appl. Toxicol. 31, 192–199. The concentration–response relationships for the hepatocarcinogenic activity of dichloroacetic acid2 (DCA) and trichloroacetic acid (TCA), two contaminants of finished drinking water, were determined in female B6C3F1 mice. Dichloroacetic acid or trichloroacetic acid at 2.0, 6.67, or 20.0 mmol/liter was administered to the mice in the drinking water starting at 7 to 8 weeks of age and until sacrifice after 360 or 576 days of exposure. The relationships of the yield of foci of altered hepatocytes, hepatocellular adenomas, and hepatocellular carcinomas to the concentration of DCA and TCA in the water were best described by second-order and linear regressions, respectively. The liver-to-body weight ratio increased linearly for both DCA and TCA, as did the vacuolization of the liver induced by DCA. The foci of altered hepatocytes and tumors in the animals treated with DCA were predominantly eosinophilic and contained glutathione S-transferase-p (GST-p, over 80% of the lesions), while the tumors induced by TCA were predominantly basophilic and lacked GST-p, including all 11 hepatocellular carcinomas. Therefore, the carcinogenic activity of DCA and TCA appeared to differ both with respect to their dose–response relationship and to the characteristics of precancerous lesions and tumors. q 1996 Society of Toxicology
Dichloroacetic acid (DCA) and trichloroacetic acid (TCA) have been found in finished drinking water in concentrations ranging from 34 to 160 mg/liter (Uden and Miller, 1983; Krasner et al., 1989). These chloroacetic acids are formed as reaction byproducts during chlorine disinfection of water containing humic acids and other organic substances (Miller and Uden, 1983; Coleman et al., 1984). Herren-Freund et al. (1987) demonstrated that DCA and TCA in the drinking water of female B6C3F1 mice induced hepatocellular adenomas and carcinomas. Others have subsequently reported that 1 Present address: Center for Environmental Medicine, Medical College of Ohio, 3000 Arlington Ave., Toledo, OH 43614. Fax: (419) 381–3089. 2 Abbreviations used: DCA, dichloroacetic acid; GST-p, glutathione Stransferase-p; TCA, trichloroacetic acid.
0272-0590/96 $18.00 Copyright q 1996 by the Society of Toxicology. All rights of reproduction in any form reserved.
AID
FAAT 2144
/
6k0c$$$201
DCA and TCA also induced liver tumors in male and female B6C3F1 mice (DeAngelo et al., 1991; Bull et al., 1990; Anna et al., 1994). The U.S. EPA has used, in part, the results of these carcinogenesis bioassays in mouse liver and the radiomimetic linear multistage (LMS) model to calculate 1006 risk levels of 1.5 and 0.56 mmol/liter for DCA and TCA in drinking water, respectively. The mechanism for the hepatocarcinogenic activity of these two chloroacetic acids appears not to result from genotoxic activity, as in vitro and in vivo tests for genotoxicity have not demonstrated significant activity (Waskel, 1978; Rapson et al., 1980; Herbert et al., 1980; Chang et al., 1992). Also, DCA has been shown to induce liver tumors that did not contain a unique mutation spectrum in either the H- or K-ras oncogene, but rather mutations in the 61st codon of the H-ras oncogene common to spontaneous tumors in B6C3F1 mice, albeit in a smaller percentage of the tumors (Anna et al., 1994). DCA and TCA do induce the proliferation of peroxisomes in the liver of mice (Elcome, 1985; Elcome et al., 1985; Goldworthy and Popp, 1987; Odum et al., 1988), an activity shared with some other hepatocarcinogens. This paper reports the concentration–response relationship for the carcinogenic activity of DCA and TCA that was previously identified by us in female B6C3F1 mice (Herren-Freund et al., 1987). The concentration– response relationship and the characteristics of the lesions induced by DCA and TCA support different pathogeneses for the two chloroacetic acids. MATERIALS AND METHODS Animals and chemicals. VAF (viral antibody-free) female B6C3F1 mice at 5 to 6 weeks of age were purchased from Charles River Breeding Laboratories (Portage, MI). The animals were housed in an AAALACaccredited facility and in accordance with the U.S. Public Health Service ‘‘Guide for the Care and Use of Laboratory Animals.’’ The Institutional Animal Care and Use Committee at Environmental Health Research and Testing approved the experiment. A maximum of five mice/cage were housed using solid-bottom polycarbonate cages with stainless steel wirebar lids. The bedding consisted of SaniChip (hardwood) bedding purchased from P. J. Murphy Forest Products (Montville, NJ). The animals were provided ad libitum with rodent pelletized NIH-07 Open Formula diet (Ziegler Brothers, Gardners, PA) and deionized and filtered (0.2 mm) drinking water. The light cycle consisted of 14 hr of light and 10 hr of dark.
192
05-15-96 14:54:47
ftoxa
AP: Fund Tox
193
CARCINOGENIC ACTIVITY OF DCA AND TCA Experimental design. When the mice were 7–8 weeks of age, they started to receive in their drinking water either DCA or TCA at 2.0, 6.67, or 20.0 mmol/liter and neutralized with sodium hydroxide to pH 6.5–7.5. The control group received 20.0 mmol/liter sodium chloride (control for sodium salt). The DCA was purchased from Eastman Kodak Co. (Rochester, NY) and the TCA and sodium chloride from Fisher Scientific Co. (Pittsburgh, PA). The concentrations of DCA and TCA were chosen so that the high concentration was comparable to those previously used by us to demonstrate carcinogenic activity in female B6C3F1 mice (Herren-Freund et al., 1987). The chloroacetic acid solutions were administered continuously to the mice until they were sacrificed by carbon dioxide asphyxiation after either 360 or 576 days of exposure. An exception was the group of mice exposed to 20.0 mmol/liter DCA in a 72-day cycle consisting of 24 days of exposure followed by 48 days without exposure. The 72-day cycle was repeated until the animals were sacrificed which resulted in the same total dose as that received by the animals continuously exposed to 6.67 mmol/liter, thus allowing comparison of the carcinogenic activity of the same total dose administered by continuous exposure to that of an intermittent exposure. The 24 days of exposure in the intermittent group was chosen to ensure that any enhancement of cell proliferation induced by DCA had ceased (see Results). The number of animals in each treatment group at the start of the study is presented in Table 1. Drinking water consumption and body weights were monitored during the first 4 weeks of the experiment and monthly thereafter. The animals were sacrificed after 360 or 576 days of exposure so that the sacrifice occurred at the end of either five or eight cycles of intermittent exposure to DCA. The terminal sacrifice for the experiment occurred at 576 days when it became apparent that liver tumors as indicated by a distended abdomen were present and that some of the animals receiving the high concentration of TCA were becoming moribund. Thus, it appeared that these animals could not survive another 72 days required to complete a cycle of intermittent exposure to DCA. During the necropsy, the liver was excised, weighed, and examined for lesions. The whole liver was cut into Ç3-mm blocks and along with representative sections of the visible lesions fixed in 10% neutral phosphate-buffered formalin for approximately 18 hr, transferred to 70% alcohol, and within 48 hr embedded in paraffin. The blocks were sectioned at 5 mm and stained with hematoxylin and eosin for histopathological evaluation for foci of altered hepatocytes, hepatocellular adenomas, and hepatocellular carcinomas. The slides were evaluated blind, being identified only by the study number and the random number assigned to the animal. Foci of altered hepatocytes in this study contained six or more cells and hepatocellular adenomas were distinguished from foci by the occurrence of compression at greater than 80% of the border of the lesion. The extent of vacuolation in the liver of DCA-treated mice was scored using a 0 to /3 scale, with 0 indicating the absence of vacuolation, /1 indicating vacuolated hepatocytes in the periportal zone, /2 indicating distribution of vacuolated hepatocytes in the midzone, and /3 indicating maximum vacuolation of the hepatocytes throughout the liver. Cell proliferation and glutathione S-transferase-p. Cell proliferation was determined in treatment groups containing 10 mice each and exposed to either DCA or TCA for 5, 12, or 33 days. Five days prior to sacrifice, mice were implanted with an osmatic minipump (Alzet Corp., Palo Alto, CA) containing 30 mg/ml bromodeoxyuridine (BrDU) and designed to deliver 1 ml/hr. Tissue blocks were sectioned at 5 mm and immunohistochemically stained for BrDU incorporation by the procedure of Sugihara et al. (1986) using monoclonal rat anti-BrDU antibody (Accurate Chemical and Scientific Corp., Westbury, NY). At least 2000 hepatocytes/mouse were evaluated for BrDU-labeled and unlabeled nuclei and the BrDU-labeling index was calculated as the percentage of hepatocytes with labeled nuclei. Glutathione S-transferase-p (GST-p) was detected by the avidin–biotin peroxidase complex method employing rabbit NCL–GST-p from human chronic lymphoblast leukemic spleen (Novocastra Laboratory, Ltd., Newcastle upon Tyne, UK), goat blocking serum, Vector kit PK-4001 and counterstaining with hematoxylin. Serial sections were stained with hema-
AID
FAAT 2144
/
6k0c$$$202
05-15-96 14:54:47
toxylin and eosin. The controls for nonspecific staining included (1) nonimmunized rabbit serum used instead of the rat anti-BrDU antibody or the rabbit NCL–GST-p and (2) the performance of the staining procedures without the primary or the secondary antibody. None of these negative controls demonstrated significant staining.
RESULTS
Drinking Water Consumption and Body and Liver Weights The drinking water consumption during the first 4 weeks of exposure to the chloroacetic acids is presented in Table 1. During the first week of exposure, only the high concentration group of both DCA and TCA exhibited a decrease in drinking water consumption. Thereafter, no effect of exposure to either DCA or TCA including the monthly monitoring of consumption (data not given) was observed except for a sporadic decrease or increase in a single treatment group. The body weights of the animals during the course of the study are presented in Fig. 1. The body weights of the mice that received 20.0 mmol/liter DCA were reduced after 35 weeks of exposure, with neither of the two other DCA treatment groups demonstrating an effect. After 51 weeks of exposure to 20.0 mmol/liter TCA, there was a trend of reduced body weights with some of the times monitored being statistically significant. The two lower concentrations of TCA did not effect the body weight of the animals. Figure 2 presents the concentration relationships of the liver-tobody weight ratio obtained after 360 and 576 days of exposure. The increase in liver-to-body weight ratio was linearly related to the concentration of DCA or TCA in the drinking water; the linear regression coefficients (r) were equaled to 0.9917 and 0.9925 at 360 days and 0.9994 and 0.9861 at 576 days, respectively. DCA exerted a greater effect than TCA upon the ratio. The low number of tumors at 360 days and the fact that the liver-to-body weight ratio did not increase with the increase in tumor response between 360 and 576 days argues against the greater ratio in DCA-treated mice being due to an increased tumor burden. Upon histopathologic evaluation, the liver from the animals treated with DCA contained vacuolated hepatocytes. The extent of vacuolation of the hepatocytes in the mice administered 0, 2.0, 6.67, or 20.0 mmol/liter DCA was scored as 0.0, 0.80 { 0.08, 2.32 { 0.11, or 2.95 { 0.05, respectively, using the scale of 0 to /3. Hence, all three concentrations of DCA resulted in vacuolation of hepatocytes. Foci and Tumor Response Table 2 contains the incidence (number and percentage of animals with a lesion) and Table 3 the yield (lesions/ mouse) of foci of altered hepatocytes, hepatocellular adenomas, and hepatocellular carcinomas in mice administered DCA or TCA. After 360 days of exposure, in mice adminis-
ftoxa
AP: Fund Tox
194
MICHAEL A. PEREIRA
TABLE 1 Drinking Water Consumption in Mice Administered DCA, TCA, or NaCl Group: treatment mmol/liter 1. 2. 3. 4. 5. 6. 7.
Weeks Na
20.0 DCA 6.67 DCA 2.0 DCA 20.0 TCA 6.67 TCA 2.0 TCA 20.0 NaCl
1
40 50 90 38 46 93 134
4.23 5.46 4.82 4.38 4.72 4.58 5.10
{ { { { { { {
0.35b,* 0.20 0.16 0.17* 0.12 0.17 0.20
2 5.56 5.46 4.59 4.78 5.05 4.69 5.00
{ { { { { { {
3 1.12 0.20 0.08 0.16 0.12 0.11 0.18
4.47 5.72 4.36 5.06 4.69 4.49 4.89
{ { { { { { {
4 0.26 0.80 0.12 0.25 0.14 0.13 0.10
4.99 5.55 5.29 5.35 5.65 5.14 4.94
{ { { { { { {
0.54 0.26 0.23 0.24 0.22* 0.19 0.12
a
N indicates the number of animals at the start of the study. Results are means { SE expressed as ml/mouse/day. * Significantly different from Group 7 (NaCl vehicle) by an ANOVA followed by a pair test with Bonferroni’s correction, p value õ0.05. b
tered DCA, the incidence and yield of foci of altered hepatocytes and hepatocellular adenomas were increased only by 20.0 mmol/liter, while after 576 days of exposure 6.67 mmol/liter also increased the yield of foci and adenomas. The higher concentration of DCA at 576 days also increased the incidence and yield of hepatocellular carcinomas. In the group of mice that received intermittent exposure to DCA, i.e., 72-day cycle consisting of 24 weeks of 20.0 mmol/liter DCA followed by 48 days without exposure, the only effect observed was an increase in foci of altered hepatocytes after 576 days. There was no significant difference between the
treatment group that intermittently received 20.0 mmol/liter DCA and the group that received the same total exposure of DCA but administered continuously as 6.67 mmol/liter. With respect to TCA, after 360 days the high concentration group (20.0 mmol/liter) contained an increase in hepatocellular carcinomas, with no other effect observed in this or the other two TCA treatment groups. Treatment for 576 days with 20.0 mmol/liter TCA resulted in an increase in foci of altered hepatocytes, hepatocellular adenomas, and hepatocellular carcinomas, while 6.67 mmol/liter resulted in an increase in foci and carcinomas. No other significant effects
FIG. 1. Body weights for female B6C3F1 mice administered either DCA (A) or TCA (B) in the drinking water. Results are means { SE with * indicating p value õ0.05 by a one-way analysis of variance followed by the Dunnett’s test (Dunnett, 1955).
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
ftoxa
AP: Fund Tox
195
CARCINOGENIC ACTIVITY OF DCA AND TCA
FIG. 2. Liver-to-body weight ratio versus the concentration of DCA and TCA in the drinking water. (A) Sacrifice after 360 days of exposure. (B) Sacrifice after 576 days of exposure. Results are means { SE with linear regression lines obtained using Sigma Plot for Windows (Jandel Scientific, Erkrath, Germany).
were obtained for the foci and tumor response in mice treated with TCA. The yield of total lesions (foci / adenomas / carcinomas) versus the concentration of DCA or TCA in drinking water is presented graphically in Fig. 3. Regression analysis indicated that the response of total lesions was linearly related to the concentration of TCA with a correlation coefficient of
0.9979. The correlation coefficient for the regression analysis of the total lesions produced by DCA was only 0.9627 for first order and 0.9998 for second order. Therefore, Fig. 3 presents the second-order regression line for the total lesion response to DCA. The dose–response relationship for the yield of tumors (adenomas / carcinomas) was also linear for TCA and second order for DCA (data given in Table 3).
TABLE 2 Incidence of Foci and Tumors in Mice Administered DCA or TCAa N Group: treatment mmol/liter 1. 2. 3. 4. 5. 6. 7. 8.
20.0 DCA Intermittent DCAc 6.67 DCA 2.0 DCA 20.0 TCA 6.67 TCA 2.0 TCA 20.0 NaCl
Foci
360
576
360
20 15 20 40 20 19 40 40
19 34 28 50 18 27 53 90
8 (40.0)b,* 0 1 (5.0) 0 0 0 3 (7.5) 0
Adenomas 576 17 14 11 7 11 9 10 10
360
(89.5)* (41.2)* (39.3)* (14.0) (61.1)* (33.3)* (18.9) (11.1)
7 (35)* 0 3 (15) 0 2 (10) 3 (15.8) 3 (7.5) 1 (2.5)
Carcinomas 576
16 3 7 3 7 3 4 2
(84.2)* (8.8) (25.0)* (6.0) (38.9)* (11.1) (7.6) (2.2)
360
576
1 (5) 0 0 0 5 (25)* 0 0 0
5 (26.3)* 1 (2.9) 1 (3.6) 0 5 (27.8)* 5 (18.5)* 0 2 (2.2)
a Mice were administered DCA, TCA, or NACl in their drinking water starting at 7–8 weeks of age. The mice continued to receive the test agents until sacrificed after either 360 or 576 days of exposure. b The results indicate the number of animals with lesions (foci of altered hepatocytes, hepatocellular adenomas or hepatocellular carcinomas) with the number in parentheses being the percentage of animals. c The mice received 20.0 mmol/liter DCA in their drinking water starting at 7–8 weeks of age and for a total of 24 days followed by 48 days without exposure. The 72-day cycle was repeated until the mice were euthanized. * p value õ 0.05 by Fisher’s exact test.
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
ftoxa
AP: Fund Tox
196
MICHAEL A. PEREIRA
TABLE 3 Yield of Liver Lesions/Mouse in DCA- or TCA-Administered Animalsa N Group: treatment mmol/liter 1. 2. 3. 4. 5. 6. 7. 8.
20.0 DCA Intermittent DCA 6.67 DCA 2.0 DCA 20.0 TCA 6.67 TCA 2.0 TCA 20.0 NaCl
Foci/mouse
360
576
360
20 15 20 40 20 19 40 40
19 34 28 50 18 27 53 90
0.60 { 0.22b,** 0 0.05 { 0.05 0 0 0 0.08 { 0.04 0
Adenoma/mouse 576
7.95 0.47 0.39 0.14 1.33 0.41 0.26 0.11
{ { { { { { { {
2.00** 0.11* 0.11* 0.05 0.31** 0.13* 0.08 0.03
360
Carcinoma/mouse
576
0.45 { 0.17** 0 0.20 { 0.12 0 0.15 { 0.11 0.21 { 0.12 0.08 { 0.04 0.03 { 0.03
5.58 0.09 0.32 0.06 0.61 0.11 0.08 0.02
{ { { { { { { {
1.14** 0.05 0.13* 0.03 0.22* 0.06 0.04 0.02
360
576
0.10 { 0.10 0 0 0 0.50 { 0.18* 0 0 0
0.37 { 0.17* 0.03 { 0.03 0.04 { 0.04 0 0.39 { 0.16* 0.22 { 0.10* 0 0.02 { 0.02
a Mice were administered DCA, TCA, or NaCl in their drinking water starting at 7–8 weeks of age. The mice continued to receive the test agents until sacrificed after either 360 or 576 days of exposure. b The results indicate the number of animals with lesions (foci of altered hepatocytes, hepatocellular adenomas, or hepatocellular carcinomas)/mouse / SE. Results are significantly different from Group 8 by the Kruskal–Wallis test: *p value õ0.05; **p value õ0.01.
The foci of altered hepatocytes and the tumors obtained from this study were basophilic, eosinophilic, or mixed containing both characteristics (Table 4). DCA induced a predominance of eosinophilic foci and tumors, with over 80% of the foci and 90% of the tumors in the 6.67 and 20.0 mmol/liter concentration groups being eosinophilic. Only approximately half of the lesions were characterized as eo-
sinophilic with the rest being basophilic, in both the treatment group intermittently exposed to DCA and the group administered 2.0 mmol/liter DCA. The percentage of the lesions characterized as eosinophilic in the group intermittently exposed to DCA was significantly less than in the groups that were administered either 6.67 or 20.0 mmol/liter DCA (p value õ0.05). The eosinophilic foci and tumors consistently stained immunohistochemically for the presence of GST-p, while basophilic lesions did not stain for GSTp, except for a few scattered cells or small areas comprising less than 10% of the lesion. The foci of altered hepatocytes in the TCA treatment groups were approximately equally distributed between basophilic and eosinophilic in tincture. However, the tumors were predominantly basophilic lacking GST-p (21 of 28, 75%) including all 11 hepatocellular carcinomas. Since a higher percentage of tumors compared to foci were basophilic, it would suggest that in TCA-treated animals basophilic foci compared to eosinophilic foci have a greater tendency to progress tumors. The limited numbers of lesions, i.e., 14, in the sodium chloride (vehicle control) group were characterized as 64.3, 28.6, and 7.1% basophilic, eosinophilic, and mixed, respectively. Cell Proliferation
FIG. 3. Yield of total lesions (foci of altered hepatocytes / hepatocellular adenomas / carcinomas) versus the concentration of DCA and TCA after 576 days of exposure. A second-order regression line is drawn for DCA and a linear regression line for TCA. Results are means { SE and Sigma Plot for Windows (Jandel Scientific, Erkrath, Germany) was used to obtain the regression lines.
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
The BrDU-labeling index in DCA- and TCA-treated mice is presented in Figs. 4A and 4B, respectively. Five days of exposure to DCA resulted in a concentration-dependent increase in the BrDU-labeling index. However, after 12 and 33 days of exposure to DCA, the BrDU-labeling index was no longer altered, although at 12 days there was a nonsignificant trend for an increase in the BrDU-labeling index. TCA also increased the BrDU-labeling index only after 5 days of exposure with all three concentrations producing approximately the same increase.
ftoxa
AP: Fund Tox
197
CARCINOGENIC ACTIVITY OF DCA AND TCA
TABLE 4 Characterization of Foci and Tumors in Mice Administered DCA or TCA Tumorsa
Foci Group: treatment mmol/liter 1. 2. 3. 4. 5. 6. 7. 8.
20.0 DCA Intermittent DCA 6.67 DCA 2.0 DCA 20.0 TCA 6.67 TCA 2.0 TCA 20.0 NaCl a b
N
Basophilic
Eosinophilic
Mixed
N
Basophilic
Eosinophilic
Mixed
150 16 11 7 22 11 13 10
3.3b 56.3 18.2 42.8 36.4 45.5 38.5 70.0
96.7 43.7 81.8 57.2 54.6 54.5 61.5 30.0
0 0 0 0 9.1 0 0 0
105 4 10 3 18 6 4 4
2.9 75.0 10.0 0 61.1 100 100 50.0
96.1 25.0 90.0 100 22.2 0 0 25.0
1.0 0 0 0 16.7 0 0 25.5
Tumors include both hepatocellular adenomas and hepatocellular carcinomas. Results are the percentages possessing the characteristics for the number of foci or turmos indicated under N, respectively.
DISCUSSION
Dichloroacetic acid and trichloroacetic acid are found as byproducts of chlorination in finished drinking water at concentrations of 34 to 160 mg/liter (Uden and Miller, 1983; Krasner et al., 1989). The chloroacetic acids are also metabolites of trichloroethylene and tetrachloroethylene (Elcome, 1985; Goldworthy and Popp, 1987; Odum et al., 1988; Dekant et al., 1984), two widely used industrial solvents that are common contaminants of surface water, groundwater, and drinking water, as well as soil and lysates of landfill hazardous waste disposal sites (Coleman et al., 1976; Con-
glio et al., 1980; Westrick et al., 1984). Exposure of B6C3F1 mice to DCA or TCA in the drinking water has been shown to produce liver tumors (Herren-Freund et al., 1987; DeAngelo et al., 1991). The U.S. EPA using a LMS model has applied these previous results to calculate 1006 risk levels of 1.5 and 0.56 mmol/liter for DCA and TCA in drinking water, respectively. Determination of the dose–response relationship for the carcinogenic activity of DCA and TCA would improve the estimation of their carcinogenic hazard to humans. The tumorigenic activity of TCA exhibited a linear relationship to its concentration in drinking water, in contrast
FIG. 4. BrDU-labeling index in 2000 hepatocytes from mice exposed to DCA (A) or TCA (B). The results are means { SE for groups of 10 mice each with * indicating p õ 0.05 (an ANOVA followed by a paired test with Bonferroni’s correction).
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
ftoxa
AP: Fund Tox
198
MICHAEL A. PEREIRA
to the second-order relationship exhibited by DCA. In male B6C3F1 mice, DeAngelo et al. (1991) also appeared to have observed an exponential dose–response relationship for DCA, in which 26.9 but not 3.84 mmol/liter increased the yield of liver tumors. Their reported tumor response in male mice of 4.0 tumors/animals after 420 days of exposure to 26.9 mmol/liter DCA (DeAngelo et al., 1991) was similar to the 5.95 tumors/animals observed by us in female mice exposed to 20.0 mmol/liter. Thus, female and male mice would appear to be of similar sensitivity to DCA. The exponential increase in the tumorigenic potency of DCA did not correlate with the linear relationships for DCA (1) induced increase in the liver-to-body weight ratio; (2) enhancement of the BrDU-labeling index after 5 days of exposure; and (3) induced vacuolation of the liver that stained with PAS, indicating an accumulation of glycogen. This would suggest that the dose of DCA to the liver does not increase exponentially with its concentration in the drinking water. DCA administered intermittently as a 72-day cycle consisting of 24 days of exposure to 20.0 mmol/liter followed by 48 days without exposure resulted in a very low yield of foci and tumors. This group was included in the study in order to determine whether the carcinogenic activity of intermittent exposure to DCA is equal to the activity of the same total dose given continuously. However, both the intermittent exposure to 20.0 mmol/liter DCA and the continuous exposure to 6.67 mmol/liter DCA induced only a very low yield of foci and tumors. The total exposure to DCA in these two treatment groups was one-third that of the group continuously exposed to 20.0 mmol/liter, so that it was expected that the yield of foci and tumors would be one-third that of the 20.0 mmol/liter treatment group. Instead, the yield of these lesions was less than 1/20th of the group exposed continuously to 20.0 mmol/liter. One possible explanation for the low tumorigenic potency of intermittent exposure to 20.0 mmol/liter DCA is that the lesions induced by DCA were not stable but regress during the 48 days in which the animals were not exposed to DCA. We have unpublished evidence that foci of altered hepatocytes and hepatocellular adenomas initiated by methylnitrosourea and promoted by DCA regress upon termination of exposure to DCA. Thus, at the beginning of the next exposure to DCA, the pathogenesis of the lesions would commence with less advanced lesions than those present when the previous exposure ceased. The characteristics of the foci and tumors induced by DCA and TCA further suggest different mechanisms of action. The foci and tumors induced by DCA were greater than 80% eosinophilic and contained GST-p; however, the tumors induced by TCA were predominantly basophilic lacking GST-p. The characteristics of foci and tumors of mouse liver have been previously reported to be carcinogen specific (Ward et al., 1983; Dragani et al., 1985; Weghorst and Klaunig, 1989; Hatayama et al., 1993). Liver tumors in
AID
FAAT 2144
/
6k0c$$$203
05-15-96 14:54:47
mice not administered a carcinogen or those induced by diethylnitrosamine have been reported to be predominantly basophilic and lacking GST-p (Ward et al., 1983; Dragani et al., 1985; Weghorst and Klaunig, 1989; Hatayama et al., 1993). When diethylnitrosamine-initiated mice were promoted with phenobarbital or 1,4-bis[2-(3,5-dichloropyridyloxy)]benzene (TCPOBOP), the resulting foci and tumors were then predominantly eosinophilic (Ward et al., 1983; Dragani et al., 1985; Weghorst and Klaunig, 1989; Hatayama et al., 1993; Diwan et al., 1992) and in female but not male mice promoted with phenobarbital contained GST-p (Hatayama et al., 1993). Chronic administration to mice of phenobarbital (Evans et al., 1992) or nitrofen (Hoover et al., 1980) also resulted in eosinophilic tumors. In contrast, promotion of diethylnitrosamine-initiated mice with di(2-ethylhexyl)phthalate, a peroxisome proliferator, resulted in foci of altered hepatocytes and tumors that were basophilic (Ward et al., 1983; Standeven and Goldworthy, 1993). Marsman and Popp (1994) have shown in rats that peroxisome proliferators also induce basophilic foci and tumors. Both DCA and TCA are peroxisome proliferators in mice and rats (Elcome, 1985; Elcome et al., 1985; Goldworthy and Popp, 1987; Odum et al., 1988). In summary, the basophilic staining of the tumors induced by TCA is consistent with other peroxisome proliferators, while the eosinophilic staining of those induced by DCA is not consistent. Furthermore, the differences in the characteristics of the tumors induced by DCA and TCA as well as the different shape of their concentration–response relationships would suggest different pathogeneses and mechanisms. ACKNOWLEDGMENTS This study was supported in part by a grant from the American Water Works Association Research Foundation. I thank Dr. Ronald A. Lubet for his helpful suggestions during the preparation of the manuscript.
REFERENCES Anna, C. H., Maronpot, R. R., Pereira, M. A., Foley, J. F., Malarkey, D. E., and Anderson, M. W. (1994). Ras proto-oncogene activation in dichloroacetic acid-, trichloroethylene- and tetrachloroethylene-induced liver tumors in B6C3F1 mice. Carcinogenesis 15, 2255–2261. Bull, R. J., Sanchez, I. M., Nelson, M. A., Larson, J. L., and Lansing, A. J. (1990). Liver tumor induction in B6C3F1 mice by dichloroacetate and trichloroacetate. Toxicology 63, 341–359. Chang, L. W., Daniel, F. B., and DeAngelo, A. B. (1992). Analysis of DNA strand breaks induced in rodent liver in vivo, hepatocytes in primary culture, and a human cell line by chlorinated acetic acids and chlorinated acetaldehydes. Environ. Mol. Mutagen. 20, 277–288. Coleman, W. E., Lingg, R. D., Melton, R. G., and Kopfler, F. C. (1976). The occurrence of volatile organics in five drinking water supplies using gas chromatography/mass spectroscopy. In Identification and Analysis of Organic Polutants of Water (L. H. Keith, Ed.), pp. 305–327. Ann Arbor Sciences Publ., Ann Arbor, MI. Coleman, E. W., Munch, J. W., Kaylor, W. H., Streicher, R. P., Ringhand,
ftoxa
AP: Fund Tox
CARCINOGENIC ACTIVITY OF DCA AND TCA H. P., and Meier, J. R. (1984). Gas chromatography/mass spectroscopy analysis of mutagenic extracts of aqueous chlorinated humic acid. A comparison of the byproducts to drinking water contaminants. Environ. Sci. Technol. 18, 674–678. Conglio, W. A., Miller, K., and Mackeever, D. (1980). The Occurrence of Volatile Organics in Drinking Water. Criteria and Standards Division, Science and Technology Branch, Environmental Protection Agency, Washington, DC. DeAngelo, A. B., Daniel, F. B., Stober, J. A., and Olson, G. R. (1991). The carcinogenicity of dichloroacetic acid in the male B6C3F1 mouse. Fundam. Appl. Toxicol. 16, 337–347. Dekant, W., Metzler, M., and Henschler, D. (1984). Novel metabolites of trichloroethylene through dechlorination reactions in mice and humans. Biochem. Pharmacol. 33, 2021–2027. Diwan, B. A., Lubet, R. A., Ward, J. M., Hrabie, J. A., and Rice, J. M. (1992). Tumor-promoting and hepatocarcinogenic effects of 1,4-bis[2(3,5-dichloropyridyloxy)]benzene (TCPOBOP) in DBA/2NCr and C57BL/6NCr mice and an apparent promoting effect on nasal cavity tumors but not on hepatocellular tumors in F344/NCr rats initiated with N-nitronitrosodiethylamine. Carcinogenesis 13, 1893–1901. Dragani, T. A., Manenti, G., Galliani, G., and Della Porta, G. (1985). Promoting effects of 1,4-bis[2-(3,5-dichloropyridyloxy)]benzene in mouse hepatocarcinogenesis. Carcinogenesis 6, 225–228. Dunnett, C. W. (1955). A multiple comparison procedure for comparing several treatments with a control. J. Am. Stat. Assoc. 50, 1096–1121. Elcome, C. R. (1985). Species differences in carcinogenicity and peroxisome proliferation due to trichloroethylene: A biochemical human hazard assessment. Arch. Toxicol. Suppl. 8, 6–17. Elcome, C. R., Rose, M. S., and Pratt, I. S. (1985). Biochemical, histological, and ultrastructural changes in rat and mouse liver following the administration of trichloroethylene: Possible relevance to species differences in hepatocarcinogenicity. Toxicol. Appl. Pharmacol. 79, 365–376. Evans, J. G., Collins, M. A., Lake, B. G., and Butler, W. H. (1992). The histology and development of hepatic nodules and carcinoma in C3H/ He and C57BL/6 mice following chronic phenobarbitone administration. Toxicol. Pathol. 20, 585–594. Goldworthy, T. L., and Popp, J. A. (1987). Chlorinated hydrocarbon-induced peroxisomal enzyme activity in relation to species and organ carcinogenicity. Toxicol. Appl. Pharmacol. 88, 225–233. Hatayama, I., Nishimura, S., Narita, T., and Sato, K. (1993). Sex-dependent expression of class pi glutathione S-transferase during chemical hepatocarcinogenesis in B6C3F1 mice. Carcinogenesis 14, 537–538. Herbert, V., Gardner, A., and Coleman, N. (1980). Mutagenicity of dichlor-
AID
FAAT 2144
/
6k0c$$$204
05-15-96 14:54:47
199
oacetate, an ingredient of some formulations of pangamic acid (tradename ‘‘vitamin B15’’). Am. J. Clin. Nutr. 33, 1179–1182. Herren-Freund, S. L., Pereira, M. A., Khoury, M. D., and Olson, G. (1987). The carcinogenicity of trichloroethylene and its metabolites, trichloroacetic and dichloroacetic acid, in mouse liver. Toxicol. Appl. Pharmacol. 90, 183–189. Hoover, K. L., Ward, J. M., and Stinson, S. F. (1980). Histopathologic differences between liver tumors in untreated (C57BL/6 1 C3H)F1 (B6C3F1 ) mice and nitrofen-fed mice. J. Natl. Cancer Inst. 65, 937– 948. Krasner, S. W., McGuire, M. J., Jacangelo, J. G., Patania, N. L., Reagen, K. M., and Ajeta, E. M. (1989). The occurrence of disinfection byproducts in U.S. drinking water. J. Am. Water Works Assoc. 81, 41–53. Marsman, D. S., and Popp, J. A. (1994). Biological potential of basophilic hepatocellular foci and hepatic adenoma induced by the peroxisome proliferator, Wy-14,643. Carcinogenesis 15, 111–117. Miller, J. W., and Uden, P. C. (1983). Characterization of nonvolatile aqueous chlorination products of humic substances. Environ. Sci. Technol. 17, 150–157. Odum, J., Green, T., Foster, J. R., and Hext, P. M. (1988). The role of trichloroacetic acid and peroxisome proliferation in the differences in carcinogenicity of perchloroethylene in the mouse and rat. Toxicol. Appl. Pharmacol. 92, 103–112. Rapson, W. H., Nazar, M. A., and Butsky, V. V. (1980). Mutagenicity produced by aqueous chlorination of organic compounds. Bull. Environ. Contam. Toxicol. 24, 590–597. Standeven, A. M., and Goldworthy, T. L. (1993). Promotion of preneoplastic lesions and induction of CYP2B by unleaded gasoline vapor in female B6C3F1 mouse liver. Carcinogenesis 14, 2137–2141. Sugihara, H., Hattori, T., and Fukuda, M. (1986). Immunohistochemical detection of bromodeoxyuridine in formalin-fixed tissues. Histochemistry 85, 193–195. Uden, P. C., and Miller, J. W. (1983). Chlorinated acids and chloral in drinking water. J. Am. Water Works Assoc. 75, 524–527. Ward, J. M., Rice, J. M., Creasia, D., Lynch, P., and Riggs, C. (1983). Dissimilar patterns of promotion by di(2-ethylhexyl)phthalate and phenobarbital of hepatocellular neoplasia initiated by diethylnitrosamine in B6C3F1 mice. Carcinogenesis 4, 1021–1029. Waskel, L. (1978). Study of the mutagenicity of anesthetics and their metabolites. Mutat. Res. 57, 141–153. Weghorst, C. M., and Klaunig, J. E. (1989). Phenobarbital promotion in diethylnitrosamine-initiated infant B6C3F1 mice: Influence of gender. Carcinogenesis 10, 609–612. Westrick, J. J., Mello, J. W., and Thomas, R. F. (1984). The groundwater supply survey. J. Am. Water Works Assoc. 76, 52.
ftoxa
AP: Fund Tox