Document XOrbgON5Y90nB1y4GZVZwaaDd

Eur. kazu, KoJn the aticnts ,200- IES, L. uroline Mature emical t. Rev. feet of lasmin >Taki, ie and mtion. r~ TOXICOLOGY AND APPLIED PHARMACOLOGY 62, 137-151 (1982) i ! The Pharmacokinetics and Macromolecular Interactions of Trichloroethylene in Mice and Rats1 W. T. Stott, J. F. Quast, and P. G. Watanabe Toxicology Research Laboratory, Health A Environmental Sciences, USA, Dow Chemical U.S.A., Midland, Michigan 48640 Received June 10, 1981; accepted September 18, 1981 The Pharmacokinetics and Macromolecular Interactions of Trichloroethylene in Mice and Rats. Stott, W. T., Quast, J. F., and Watanabe. P. G., (1982). Toxicol. Appl. Pharmacol. 62, 137-151. Male B6C3F1 mice metabolized inhaled trichloroethylene (TRI) (600 ppm/6 hr) to a greater extent (262% more) than male Osbome-Mendel rats. Mice metabolized more (332%) inhaled TRI to a hepatic macromolecular binding metabolite in vivo than rats. Oral administration of TRI resulted in treatment-related hepatocellular cytotoxicity in repeated dosing trials in the mouse. Hepatic effects observed in mice treated with a maximum tolerated dose of 2400 mg/kg/day TRI for 3 days were primarily centrilobular hepatocellular swelling with focal hepatocellular necrosis. These effects lead to an enhanced regenerative process as indicated by increased hepatic DNA synthesis activity (220% of control) and incidence of mitotic figures. Treatment of mice (po) with TRI for a 3-week period (5 days/week) resulted in a dose-related increase in hepatocellular swelling with giant and mineralized cells present in the 2400 mg/kg/day dosed animals. Contrasting the mouse data, rats appeared to be less sensitive to a maximum tolerated dose level of TRI, displaying enhanced hepatic DNA syn thesis levels (175% of control) but no histopathology after a similar 3-week treatment with 1 1100 mg/kg/day TRI. RenaUissue injxrth species was not significantly affected bv TRI. An estimate of the extent of TRI interaction with DNA was also determined by measuring the radioactivity associated with purified hepatic DNA, Only a very low level of in vivo TR1DNA interaction was observed in mice given 1200 mg/kg TRI po which is reportedly tumorigenic upon chronic administration (maximum estimate 0.62 0.43 alkylation/10* nucleotides). When coupled with the very weak or negative responses of pure TRI in in vitro mutagenesis assays, the DNA alkylation data indicate a lack of eenotoxic potential. The data in toto suggest an epigenetic mechanism of tumor formation in the B6C3F1 mouse, implying that a tumorieenic response to TRI fxpospre in. these animals would only be evulsni^ upon chronic administration of high, cytotoxic dose levels of TRI. Trichloroethylene (TRI) has been widely used as an industrial solvent and dry clean ing agent. The safety of trichloroethylene has been questioned on the basis of the re sults of a carcinogen bioassay conducted by the National Cancer Institute (NCI) (1976); an increased incidence of hepatocellular car 1 Research sponsored by the Trichloroethylene Pro gram Panel of the Chemical Manufacturers Associa tion. cinoma was observed in male mice given TRI. A confounding factor ih this study (NCI, 1976) was the use of TRI containing approximately 0.2% epichlorohydrin (Henschler et al., 1977) which has previously been shown to cause nasal carcinomas in rats at toxic doses (Laskin et al., 1980). No in creased incidence of tumors was found in a recent 18-month chronic inhalation bioassav of 100 and 5QQ pnm nonepoxide-stabilized Tkl in hamsters, rats, and mice (Henschler 137 0041-008X/82/010137-15$02.00/0 Copyright > 19$2 by Academic Pmt, Inc All rights of reproduction in aay fora reserved. SL 035290 138 STOTT. QUAST, AND WATANABE et al., 1980). In addition, a cohort study of workers exposed to TRI has not revealed any increase in cancer mortality (Axelson et al., 1978). TRI has been reported to be both negative or weakly positive in numerous mutagenicity bioassays. However, many of these studies are of questionable value because they lack enough data for critical evaluation and/or analytical data regarding the presence of mutagenic epoxide-stabilizing compounds in the TRI utilized (Duprat and Gradiski, 1980; Konietzko et al., 1978; Slacik-Erben et al., 1980; Fahrig, 1977; Bartsch et al., 1979; Bronzetti et al., 1978; Greim et al., 1975; Shahin and Von Borstei, 1977; Sim mon et al., 1977; Waskell, 1978). Further, TRI metabolites reportedly bind to protein, RNA, and DNA in vitro and protein in vivo (Laib et al., 1979; Uehleke and PoplawskiTabarelli, 1977; Van Duuren and Bancrjce, 1976; Bolt et al., 1977; Bolt and Filser, 1977; Banerjee and Van Duuren, 1978). The present study was undertaken to pro vide a better understanding of the bioassay results (NCI, 1976) and their relevance in terms of carcinogenic risk to humans ex posed to TRI. This evaluation was ap proached by studying the metabolism, bioac tivation, and cellular interactions of TRI in susceptible (B6C3F1) and nonsusceptible (Osborne-Mendel) mice and rats, respec tively. METHODS Chemicals and radiochemicals. Trichloroethylene (>99.9% purity, amine stabilized) was obtained from the Dow Chemical Company (Freeport, Tex.) [UL14C]Trichloroethylene [UC](TRI) (specific activity *11.3 mCi/mmol) was purchased from Midwest Re search Institute (Kansas City, Mo.). Radiochemical purity was determined to be >99% by radio-gas chromatographic/mass spectroscopic analysts. [6'HJThymidine ([3H]dT) (specific activity * 24.7 mCi/ mmol) was obtained from New England Nuclear (Bos ton, Mass.). All enzymes, nucleosides, and nucleic acid base standards were obtained from Sigma Chemical Company (St. Louis, Mo.). Animals. Male B6C3F1 mice (17 to 29 g) were ob tained from Charles River Breeding Laboratory (Por tage, Mich.) and male Osborne-Mendel rats (190 to 230 g) were obtained from CAMM Research Labora tory (Wayne, N.J.). All animals were fed rat chow (Ralston Purina, St. Louis, Mo.) and given water ad libitum, housed under conditions of a 12-hr light cycle, 21 2*C, 50 13% R.H., and acclimated at least 7 days. Animals were housed in groups except for rats on study which were housed individually. Metabolism study. Sixteen animals were exposed to either 10 or 600 ppm [,4C]TRI for 6 hr in a 30-liter glass inhalation chamber supplied with an airflow either of 3 liters/min (mice) or 10 liters/min (rats). Four an imals per exposure were acclimated 21 day in metabolic cages prior to exposure. Chamber concentrations were monitored continuously by a Wilks Infrared spectrom eter (S. Norwalk, Conn.). The chamber concentrations were maintained within 10% of targeted concentrations for the entire exposure period (9.8 0.3 and 613 6 ppm for mice; 9.4 0.7 and 630 ppm for rats). The specific activity of the chamber [l4C]TRI was analyzed at the beginning and end of exposure by slowly bubbling 0.3 ml of chamber atmosphere through 20 ml of ACS scintillation medium (Amersham Corp., Arlington Heights, 111.) and counting ,4C in a Beckman Model No. LS9000 liquid scintillation counter (Beckman Instru ments, Inc, Anaheim, Calif.). Specific activities equalled 663 fiCi and 174 MCi/mmol [,4CJTRI for mice, and 13 MCi and 4.3 MCi/mmol [>4C]TRI for rats in 10 and 600ppm exposure levels, respectively. At the termination of [,4C]TRI exposure, four animals were placed indi vidually into Rothtype metabolism cages for sequential collection of excreta including expired volatiles. The remaining 12 animals were utilized for studies of raacromolecular binding (see below). Excretion of >4C activity in the urine, feces, and ex pired air was monitored for 30 hr postexposure. Urine was collected at --80*C, carbon dioxide was trapped by bubbling the chamber effluent through a solution of 2-methoxyethanol:ethanolamine (7:3), and expired [l4C}TRI was trapped on activated charcoal columns (6 g). In preliminary experiments, charcoal trapping of [l4C]TRI was observed to be essentially complete, and the trapped >4C was found by GC-MS analysis to be [,4C]TRI; desorption of trapped [,4C]TRI by toluene for 4 hr was >93%. It was also noted in preliminary experiments that >98% of gavage-administered ['"CJTRI was recovered in 48 hr. Therefore in the present study, at approximately 30 hr post [l4C]TRI exposure the an imals were killed by decapitation; the liver, kidneys, and a patch of skin from the back were collected, weighed, and stored at -80*C until analyzed. Carcasses were skinned, weighed, ground, and homogenized with a measured volume of distilled water and stored at --80"C. The micromole equivalents (Mtnol-eq) of [l4C]TRI in SL 035291 TRICHLOROETHYLENE INTERACTIONS IN MICE AND RATS 139 the samples were then determined by direct liquid scin tillation counting (LSC) (urine, COj trap, and toluenedesorbed TRI) or by first oxidizing samples (liver, kid ney, skin, feces homogenate, and carcass homogenate) in a Harvey Biological Material Oxidizer (Hillsdale, N.J.). Results are expressed as total micromole equiv alents of [l4C]TRI present in a sample. Macromolecular binding. At 0, 6, and 24 hr postin halation exposure to 10 or 600 ppm [l4C]TRI for 6 hr (see above), four animals were killed by decapitation. The livers and kidneys were excised, frozen on dry ice, and stored at --80"C. Tissue samples were obtained from mice or rats from the metabolism study at 50 hr postexposure. A 1:1 (tissue:water) homogenate was then prepared, and the irreversible binding of radioactivity to hepatic and renal tissue was determined (Jollow et at., 1973). After exhaustive solvent extraction of the trichloroacetic acid precipilable material, the pellet was solubilized in 1 m potassium hydroxide and neutralized with 1 M hydrochloric acid; the ,4C content was then quantitated by LSC. Protein content of the pelleted materia] was assayed by the method of Lowry et at. (1951). The results are expressed as picomole equiva lents of TRI bound per microgram of protein. . Repeated dosing experiments. For the 3-day study, mice were gavaged with 0 or 2400 mg/kg/day TRI in corn oil (USP grade, 10 ml/kg), and rats received daily doses of TRI (0 and 1100 mg/kg/day TRI) in corn oil (5 ml/kg). All animals received daily sc injections of [3H]thymidine ([3H]dT) 3 to 4 hr after TRI dosing. Mice were administered 1 mCi, 15.2 ug [*H]dT, and 10 ml volume/kg (specific activity* 15.0 Ci/mmol) while rats were administered 0.1 mCi. 1.52 ng [>H]dT, and 1 ml volume/kg (specific activity * 15.9 Ci/mmol): In the first 3-week study, mice were treated as above (5 days/week) except that animals were administered [3HJdT (0.1 mCi, 6.42 ug [JH]dT, and 10 ml volume/ kg, specific activity * 3.77 Ci/mmol) on the last 5 days of TRI dosing. In the larger, second 3-week study, mice received either 0, 250, 500, 1200, or 2400 mg/kg/day (5 days/week) TRI in corn oil (10 ml/kg). Mice re ceived sc injections of ['H]dT as described for the 3day experiments on the 4 days preceding termination of the experiment Rats were dosed with TRI as above and treated with [3H)dT via implantation between the shoulder blades of an osmotic pump (Alza Corp., Palo Alto, Calif.) containing 0.1 mCi and 1 Mg [5H]dT on Day 15 of the experiment Upon termination of an experiment, animals were killed by decapitation, the liver and kidneys (where noted in results) were excised, weighed, placed on dry ice, and stored at --8Q'C. Upon thawing, DNA was extracted from tissues after the procedure outlined by Reitz et al. (1980a) and quantitated by the diphenylamine method of Burton (1956). DNA synthesis indi cated by [JH]dT incorporation was detected by LSC. DNA synthesis activity ([JH]dT incorporation) is ex pressed as disintegrations per minute [3H]dT per mi crogram of DNA. DNA alkylation in vivo: DNA isolation and highpressure liquid chromatography analysis. Mice were dosed by gavage with 1200 mg/kg [,4C]TRI (>500 /iCi/animal, specific activity * 2.4 mCi/mmol) in com oil (10 ml/kg) and killed 5 hr later by decapitation. Livers were immediately excised, frozen on dry ice, and stored at --80C (<2 days). DNA was isolated by a combination of the methods of Marmur (1961) and Kirby (1957). Isolated DNA was free of detectable RNA (orcinol reaction) and glycogen (enthrone reac tion). Quantification of DNA and l4C contents was ac complished as described above on small aliquots of the sample. High-pressure liquid chromatographic (HPLC) anal ysis of the remaining DNA isolate was undertaken by measuring the radioactivity associated with isolated free purine bases or nucleosides. Half the DNA samples were enzymatically digested to nucleosides in 0.05 m Tris0.01 m MgClj buffer (pH 7.8) by 0.05 mg DNAase I, 1 mg snake venom phosphodiesterase (Crotalus atrax), and 1.5 U alkaline phosphatase/mg DNA at 37'C for 16 hr. The remaining half of the samples were digested in 0.1 M HQ for 30 min at 70C, a procedure that cleaves purine bases from the phosphate-sugar back bone of DNA without significant loss of alkylated prod ucts (Lawley, 1976; Swenson and Lawley, 1978). Both types of DNA digests were analyzed by HPLC on a Partisil M-9 10/50 SCX column (Whatman Inc., Clif ton, NJ.) with a Waters Model 6000A pump (Milford, Mass.) equipped with a Perkin-Elmer Model LC 55B Spectrophotometer (Oak Brook, III.). Nucleosides and bases were eluted from the column by an isocratic flow of 2 ml/min of 0.1 m ammonium phosphate buffer (pH 3.5) and detected at 254 nm. Fractions (1.0 ml) of the eluate were collected and following addition of 10 ml Aquasol-2 scintillation medium (New England Nu clear), the ,4C content was determined by LSC. Sig nificant levels of disintegrations per minute sample were determined to be those greater than a possible two sigma counting error above the mean background disintegra tions per minute sample (i.e., mean of background plus possible counting error). This procedure has given ex cellent separation of alkylated nucleosides or bases from normal nucleosides and bases in our laboratory. The pattern of radioactivity elution was then compared to the nucleoside or free base elution pattern allowing for the differentiation of metabolically incorporated (C-l) l4C in normal bases, void volume of 14C associated with trace protein or glycogen contamination (C-l metabolic incorporation or binding) and true DNA alkylated bases. Histopathology. Tissue samples were taken from the central region of the left lobe of the liver and/or right 035292 si* 140 STOTT, QUAST, AND WATANABE kidney prior to freezing of the excised tissue. Samples were fixed in 10% neutral-buffered formalin, sectioned, processed by standard procedures, stained with bematoxylin-eosin, and observed microscopically for evidence of morphologic changes associated with treatment. Statistics. Animals were computer randomized into treatment groups in all repeated dosing experiments and checked for group homogeneity and significant differ ences between groups by Bartlett's test and Dunnett's r test (Steel and Torrie, 1960). All results were com pared by Dunnett's or Student's t test (Steel and Torrie, 1960) with the level of significance set at p < 0.05. RESULTS Metabolism As shown in Table 1, 98 to 99% of the total 14C body burden of male B6C3F1 mice exposed to either 10 or 600 ppm [14C]TRI for 6 hr was metabolized within 50 hr. The primary route of excretion in mice was the urine constituting approximately 75% of the total TRI body burden. Approximately 9% of the [14C]TRI body burden was biotrans formed to "CO* As shown by the general lack of change in the routes of excretion and total metabolism of [14C]TRI, the metabo lism of [l4C]TRI by the B6C3F1 mouse did not appear to be saturated at the 600-ppmTRI exposure level. While small increases in excretion of unmetabolized [C,4]TRI (exhaled) and carcass 14C were observed between the low- and high-exposure levels, these changes were within the potential ex perimental error. In contrast, the metabo lism of [14C]TRI by the male Osborne-Mendel rat appeared to show characteristics of saturation at the 600-ppm exposure level (Table 2). Total metabolism of [UC]TRI (^mol-eq TRI/kg body wt) was decreased in the rats exposed to 600 ppm (79% [14C]TRI body burden) relative to those ex posed to 10 ppm (98% [,4C]TRI body bur den). At the same time exhalation of [,4C]TRI increased 10-fold with increased [ l4C]TRI exposure level. As with mice, the primary route of excretion of [,4C]TRI was via the urine which accounted for 62 and 55% of body burden [,4C]TRI in low- and high-dose rats, respectively. TABLE 1 Recovery of Radioactivity for 50 hr Pqstjnhaiation Exposure of Male B6C3F1 Mice to 10 or 600 ppm 1,1,2[UL-uC]Trichloroethylenefor 6 hr* 10 ppm 600 ppm /imol-eq TRI/ /unol-eq TRI/ kg body wt %* kg body wt %* Expired 1,1,2-TRI CO, Urine Feces Cage wash1 Skin Liver Kidney Carcass Total body burden . Total metabolized 0.617 (0.182) -7.38 (0.383) 58.1 (18.8) 2.92 (0.542) 1.10 (0,446) X33 (1.13) 2.40 (0.401) 0.310(0,310) 1.50 (0.316) 78.5 (17.9) 77.9 (18.1) 0.79 9.4 74.0 3.7 1.4 3.0 3.1 0.39 1.9 99.2 76.2 (27.3) 299 (37.5) 2278 (505) 115 (24.6) 85.5 (65.5) 56.3 (11.9) 7X2 (10.4) 11.1 (3.18) 146 (57.9) 3138 (616) 3062 (592) 2.4 9.5 7X6 3.7 X7 1.8 2.3 0.35 4.6 97.6 * Average of four animals (SD). * Percentage of recovered radioactivity. ' Primarily due to urine. Si 035293 TRICHLOROETHYLENE INTERACTIONS IN MICE AND RATS 141 TABLE 2 Recovery of Radioactivity for 50 hr Postinhalation Exposure of Male Osborne-Mendel Rats to 10 or 600 ppm I,1,2[UL-mC1Trichloroethylenefor 6 hr* 10 ppm 600 ppm /imol-eq TRI/ kg body wt <fc* umol-eq TRI/ kg body wt %* Expired 1,1,2-TRI CO, Urine Feces Cage wash' Skin Liver Kidney Carcass Total body burden Total metabolized 0.778 (0.103) 1.71 (0.295) 22.6 (5.96) 2.52 (0.253) 0.357 (0.204) 3.78 (1.21) 1.01 (0.085) 0.131 (0.009) 2.82 (0.898) 35.8 (5.64) 35.0 (5.64) 2.1 4.8 63.1 7.0 1.0 10.6 2.8 0.37 7.9 97.8 227 (6.40) 31.2 (2.10) 594 (79.8) 38.3 (3.47) 15.1 (10.7) 76.2 (30.3) 16.3 (1.13) 1.72 (0.083) 75.1 (27.7) 1075 (87.2) 847 (81.6) 21.1 2.9 55.3 3.6 1.4 7.1 1.6 0.16 7.0 78.9 * Average of four animals (SD). * Percentage of recovered radioactivity. ' Primarily due to urine. A comparison between the mouse and rat revealed that the mouse metabolized 123% (p < 0.01) more body burden TRI on a per kilogram body weight basis than rats after exposure to 10 ppm and 262% (p<0.01) more [l4C]TRI/kg than rats after exposure to 600 ppm. tivation in rats exposed to 600 ppm while the activation of [14C]TRI in mice remained unsaturated. Maximum binding observed in the liver occurred immediately postexposure (3 hr for the kidneys) and decreased steadily over the next 48 hr. Macromolecular Binding As shown in Table 3, mice activated more [MC]TRI to a reactive, macromolecular binding metabolite than did rats upon ex posure to 10 and 600 ppm [UC)TRI. The species differences in binding, measured as picomole-equivalents as [UC]TRI per mi crogram of protein, were considerably greater (three- to fourfold) in both hepatic and renal tissues following the 600-ppm exposure, and only a modest increase in binding was ob served in hepatic tissue following the 10-ppm exposure. Presumably this finding was due to the saturation of [UC]TRI metabolic ac Repeated Dosing Experiments Repeated gavage administration of dos ages of TRI to mice caused alterations in hepatocellular morphology and in levels of hepatic DNA synthesis but had no observ able effect on renal tissue (Table.'4). Mice dosed for 3 days with 2400 mg/kg/day TRI had increased liver weights (120% of con trol), slightly less DNA per gram of tissue (90% of control), and an increase in hepatic DNA synthesis activity (222% of control). Upon microscopic examination of hepatic tissue, histopathological changes were ob served which included a generalized pattern of centrilobular hepatocellular swelling with I i 03529* SL 142 STOTT, QUAST, AND WATANABE TABLE 3 Hepatic and Renal Macromolecular Binding of 1,1,2[UL-**C]Trichloroethylene Metaboute in Male B6C3F1 Mice and Osborne-Mendel Rats Exposed to 10 or 600 ppm for 6 hr* Exposure (ppm) 10 t600 Tissue Liver Kidney Liver Kidney Test animal Mouse Rat Mouse Rat Mouse Rat Mouse Rat Binding (pmoleq ("ClTRIAig protein) 0.318 (0.004)* 0.268 (0.013) 0.168 (0.014) 0.153 (0.007) 20.4 (2.49)* 4.72 (0.415) 5.06 (0.667)* 1.77 (0.200) Ratio (mouse/rat) 1.2 u 4,3 3.0 ' Average of four animals (SD) at time of maximum binding postexposure (liver m 0 hr, kidney m 3 hr). * p < 0.01. some individual hepatocellular acute necro sis and an increase in mitotic activity. The response to TRI exposure after 3 weeks of repeated gavage dosing over a broad dose range in mouse hepatic tissue is shown in Tables 4 and 5. In an initial 3-week repeated dosing experiment (Table 4), mice dosed with 2400 mg/kg/day had a large in crease in liver weight (170% of control), a substantial decrease (76% of control) in DNA content per gram of tissue (indicative of larger cell size), and an increase in hepatic DNA synthesis activity (181% of control). Histopathological changes observed were a TABLE 4 Organ-to-Body Weight Ratios, DNA Content, DNA Synthesis, and Histopathology Data from Male B6C3F1 Mice Administered 1,1,2-Trichloroethylene by Gavage for 3 Days or 3 Weeks (5 Days/Week)* Exptrimoat (do* lc*d) Pomuifo ort*o waifht/body wi*tlt mg DNA/g tuvuc DNA lyntluu dpn f'HJdT/ as DNA Tnoiod/ control Hutopctbolofy* 3-Dy npesud |i*i|* Control (corn oil) Liver Tnatad (2400 mg/kg) Liver 3-Wart npotud po|t (5 dayi/wart) Cootroi (an oil) Uror Kidney Truicd (2400 mg/kg) Liver Kidney 4.6 (0.19) 5,5 (0,32)* 3.21 (0.15) 1.49(0.14) 1,95 (0.72)* 1.49(0.11) 1.19 (0.10) 2.19 (0.19)* 120.7 (37.6) 261.3 (69 2.99(0.21) 5.63 (0.64) 2.27 (0.13)* 4.91 (0.60) 11.06 025) 10.35 (2.40) 20.02 (6.96)* 1.39(1.11) 2.22 Lit 0.11 - ' Avtnfe of 10-12 animal* (iSD). * No bmopothofofy otecrvcd. +++, hnlopatbolofy obwvoil (m* text for daKriptioo). * p < 0.01. SL 035295 TRICHLOROETHYLENE INTERACTIONS IN MICE AND RATS 143 EIN TABLE 5 Liver-to-Body Weight Ratios, DNA Content, DNA Synthesis, and Histopathology Data from Male B6C3F1 Mice administered 1,1,2-Trichlorethylene by Gavage for 3 Weeks (5 Days/Week)* DNA synthesis Dose level (mg/kg) Percentage liver weight/body weight Mg DNA/g hepatic tissue dpm [3H]dT/Mg DNA Treated/ control Histopathology* Control (corn oil) Treated 250 500 1200 2400 4.85 (0.28) 5.20 (0.53) 5.46 (0.61 )* 6.56 (0.42)** 8.57 (0.46)** 2.72 (0.20) 152 (0.22) 2.39 (0.20)** 2.14 (0.12)** 2.09 (0.36)** 128 (27.5) 108 (34.0) 87.6 (15.4)* 140 (41.0) 156 (52.6) 0.84 0.68 1.09 1.22 - + + ++ +++ * Average of 10-12 animals (SD). r). * -, No histopathology observed, +++ >++>+, severity of observable histopathology. p < 0.05. *p<0.01. in- ), a generalized pattern of centrilobular hepa plasmic reticulum) at 250 and 500 mg/kg/ in tocellular swelling with giant ceil inflam day TRI dose levels to increased centrilob live mation and dystrophic mineralized cells in ular hepatocellular swelling (1200 mg/kg/ itic the centrilobular or midzonal regions of the day TRI group), and to still more severe lobule (Fig. 1). These mineralized (calcified) centrilobular hepatocyte swelling, giant cell cells were indicative of earlier individual cell inflammation, and mineralized cells (2400 necrosis in these animals as seen in the 3- mg/kg/day TRI group) (Fig. 1). While day repeated dosing trial. No renal effect of there was some indication of increased he administered TRI was observed and body patic DNA synthesis at the higher-dose lev OM weights did not differ significantly between els, the biological variability precluded sta EKS treated and control animals. In a second 3- tistical significance contrary to earlier week repeated TRI-dosing experiment (Ta observations. ble 5), mice dosed with 0, 250, 500, 1200, Unlike mice, only slight, nonsignificant or 2400 mg/kg/day showed a clear dose-re changes were observed in hepatic tissue of lated hepatocellular hypertrophic response rats repeatedly dosed for 3 days with a 1100 to administered TRI. A dose-related in mg/kg/day dose of TRI (Table 6). In rats crease in liver weights (up to a maximum repeatedly dosed for 3 weeks (5 days/week) of 177% of control) and a decrease in DNA with 1100 mg/kg/day TRI, an. increase in content per gram of tissue (77% of control) liver weights (118% of control)' with no de were observed in mice repeatedly dosed with crease in DNA content per gram of tissue 2:500 mg/kg/day TRI. Dose-related histo- and no hepatic histopathology was observed. pathological changes in hepatic tissue were An increase in hepatic DNA synthesis ac observed at all dose levels. These changes tivity (175% of control) in these animals was ranged from slight increases in cytoplasmic observed, however, indicating that the in eosinophilic staining of centrilobular hepa- creased liver weight was apparently due to tocytes (often associated with depleted gly an increase in morphologically normal-look cogen or proliferation of smooth endo ing hepatocytes. Renal tissue and body SL 035296 144 STOTT. QUAST. AND WATANABE Fig. I. Photomicrograph* of hepatic tissues from (a) control mouse and (b) a mouse dosed with 2400 mgA8/day for 3 weeks (5 days/week). Note the presence of dystrophic mineralized cells in the centrilobular region shown in the latter micrograph. weight gains were not significantly affected by exposure to TRI. Overall, the OsborneMendel rats response to administered TRI was considerably less than that of the B6C3F1 mouse. In Vivo DNA Alkylation Investigation of the ability of TRI to al kylate hepatic DNA in vivo in mice indi cated that TRI-alkylated DNA to only a 035^97 SL TRICHLOROETHYLENE INTERACTIONS IN MICE AND RATS 145 very small degree (Table 7). In mice dosed with 1200 mg/kg of relatively high-specific* activity [UC]TRI (2.4 mCi/mmol), a max imum estimate of the average DNA alkyl ation level of 0.62 0.42 alkylations/10* nucleotides was observed in three of four treated mice. The possible erroneous contri bution of trace contamination of the DNA isolate with alkylated proteins or of meta bolic incorporation (C-l pool) or TRI-de- SL 035298 146 STOTT, QUAST, AND WATANABE TABLE 6 Organ-to-Body Weight Ratios, DNA Content, DNA Synthesis, and Histopathology Data from Male Osborne-Mendel Rats Administered 1,1,2-Trichloroethylene by Gavage for 3 Days or 3 weeks (5 Days/Week)* Experiment (dote level) 3-Doy repeated f*va|t Control (core oil) Liner Treated (MOfrmg/kf) Liver 3*Wwk repeated lavage (5 days/woek) Control (corn oil) Liver Kidney Treated (1100 mg/kg) Liver Kidney Average of four ***"*> (SD). * No obtarvable histoptthokkfy. *p<kOl. Percent*!* orgin wright/body weight 4.00 (0.19) 4.40 (0.27) 3.77 (0.21) 0,79 (0.04) 4.43 (0.21 r 0.(2 (0.04) Mg DNA/s tissue 2.94 (0.26) 2.65 (0.29) 3.27 (0.40) 3.65 (0.SI) 3.27 <0l26) 3.53 (059) DNA synthesis " dpm pHldT Mg DNA Treated/ control 35.4 (13.5) 42-t (6.7S) 1.21 (0.61) 5.50 (2.34) 2.12(0.96)* 6.31 (1.72) 1.21 1.75 1.15 Histopathology* - - rived radioactivity into normal DNA bases to the alkylation value was eliminated to a large extent by HPLC analysis of isolated DNA digests. This procedure is particularly necessary for compounds such as TRI which undergo considerable catabolism and protein binding in vivo. As shown in Fig. 2, a ma jority of the radioactivity (> counting error) eluted from the column with this particular DNA isolate was present in the void volume or coeluted with normal purines. Void vol ume radioactivity was attributed to trace TABLE 7 In Vivo Alkylation of Hepatic DNA by 1,1,2(UL-uC]Trichloroethylene in Male B6C3F1 Mice Dosed with 1200 mg/kg [,4C]TRI by Gavage Animal Detection limit (alk/10*) Non-Ci dpm' >2 Sigma error Maximum alkylations 10* nucleotides* Maximum CBI' 1 0.12 2 0.52 3 0.15 4 0.18 18.0 ND* 7.1 26.6 0.50 -- 0.27 1.1 0.055 -- 0.030 0.120 ' dpm not associated on HPLC analysis with normal DNA bases from normal metabolic incorporation and not attributable to counting error (i.e., mean background dpm + 2 SD). * "Maximum" alkylations possible due to the relatively large amount of dpm associated with protein binding and metabolic incorporation into normal bases. `Covalent binding index (Lutz, 1979). Comparative hepatic CBI for dimethylnitrosamine, diethylnitrosamine, methylnitrosourea, and aflatoxin B-l are approximately 5500, 125, 640, and 17,000 respectively. CBI " (junol adduct/mol dN)/(mmol/kg dose). 4 None detected. 03^i99 Si* TRICHLOROETHYLENE INTERACTIONS IN MICE AND RATS 147 protein contamination in the isolated DNA and unhydrolyzed nucleotide oligomers since alkylated DNA bases routinely elute consid erably beyond the solvent front and often beyond the normal DNA bases with the SCX column. A relatively large amount of void volume radioactivity also has been re ported by Parchman and Magee (1980) in DNA isolated from [l4C]TRI-exposed rats and mice and by Laib et al. (1979) in in vitro microsome-RNA preparations. Radio activity was associated with purine bases in all samples analyzed either as bases (acid hydrolysis) or as nucleosides (enzyme hy drolysis). This normal purine associated ra dioactivity was identical to that observed in experiments (data not shown) utilizing [uC]formic acid to label the C-l pool in mice and subsequently label purines (and eventually thymine also) within 5 hr of dos ing. The remaining radioactivity (> count ing error) not associated with normal bases or the void volume was assumed to represent true alkylated bases. However, the possibil ity of "spill over" of some [l4cj-labeled pep tides from the void volume, the short reten tion times of these peaks (less than that expected for most alkylated bases), and the fact that no chemical identification of pos sible adducts was undertaken dictate that these alkylation values be viewed with cau tion. Thus, it is important that the base al kylation values obtained from these data should be regarded as a maximum possible alkylation which may have occurred. DISCUSSION The data presented provide a possible ex planation of the observed differences in the tumorigenic response of male B6C3F1 mice and Osborne-Mendel rats to TRI (NCI, 1976) based upon their ability to metabolize more TRI to a reactive compound and sub sequent hepatotoxicity. Mice metabolize more inhaled TRI than rats at both low (10 ppm)- and high (600 ppm)-exposure levels (123 and 260%, respectively). Further, the amount of total in vivo hepatic and renal macromolecular binding of TRI in these an imals was observed to be as much as three- Fig. 2. HPLC strong cation-exchange chromatogram of enzymatic digest of hepatic DNA isolated from a B6C3F1 mouse 6 hr after dosing with 1200 mg/kg ["CJTRI. Normal nucleosides shown ate thymidine (dT), deoxyguanosine (dG), deoxycytidine (dC), and deoxyadenosine (dA). Radioactivity in collected fractions are also plotted (dots). The solid line at 43 dpm/fraction represents the mean back ground dpm/fraction plus counting error (2a). SL 35300 148 STOTT, QUAST, AND WATANABE fold higher in mice than in rats (600-ppm exposure). Thus B6C3F1 mice appear to be capable of metabolizing more inhaled TRI to a toxic, reactive intermediate on a weight basis than Osborne-Mendel rats, especially at relatively high-exposure levels. Humans metabolize TRI (inhaled) simi lar to rodents and excrete trichloroethanol and trichloroacetic acid in the urine (Ertle et al., 1972; Ikeda and Imamura, 1973; Kirnmerle and Eben, 1973; Monster et al., 1976; Muller et al., 1974). Based upon the kinetics of TRI clearance in humans (Monster et al., 1976) (approximately 3.7 liters/hr-kg), hu mans metabolize approximately 20 times less TRI on a weight basis than rats (77 liters/hr - kg) at similar exposure levels of 70 and 65 ppm, respectively (Filser and Bolt, 1979). Thus any risk extrapolation made from rodent data to humans should take into account this fact. As demonstrated by Reitz et al. (1978, 1980b) with chloroform, chem icals which are metabolically activated may be less toxic in man than rodents if man metabolically activates less when expressed on a per kilogram basis. The mechanism by which a compound may cause tumors has important implica tions in terms of the degree of carcinogenic hazard it may pose to humans. This subject has been extensively reviewed by several au thors (Reitz et al., 1980a; Schumann et al., 1980; Stott et al., 1981; Watanabe et al., 1980; Weisburger and Williams, 1980). Briefly, the genetic mechanism of carcino genesis is characterized by a direct inter action (alkylation or intercalation) of the chemical with DNA while epigenetic mech anisms of carcinogenesis do not involve a direct interaction of the chemical with ge netic material. An epigenetic mechanism that reportedly may characterize the tumorigenic action of several chlorinated ethylenes in the B6C3F1 mouse has been that of recurrent cytotoxicity (Reitz et al., 1980a; Schumann et al., 1980). As reviewed by Stott et al. (1981) and Watanabe et al. (1980), this mechanism of tumorigenicity involves the enhancement of the normal cel lular mutation rate during regenerative ac tivity (increased DNA synthesis) in response to a cytotoxic challenge by a chemical. Little alkylation of DNA occurs and the toxico logical impact of the chemical (i.e., cytotox icity) displays characteristics of reversibility and threshold. Thus, a tumorigenic response of animals to chemicals of this type is ob served only after prolonged exposure to toxic dose levels. Consistent with a recurrent cytotoxicity mechanism of tumorigenicity, administra tion to mice of a dose of TRI which was tumorigenic upon chronic administration (2400 mg/kg/day for 3 days) caused a lo calized cell necrosis, an enhanced DNA syn thesis, and a minimal degree of centrilobular hepatocellular swelling. Upon a more pro longed exposure (3 weeks), however, the pri mary response observed in mice was a doserelated centrilobular hepatocellular swelling and the occurrence of mineralized cells, pre sumably a result of hepatocellular necrosis in early stages of TRI dosing. Treatmentrelated hepatic effects were seen in male Osborne-Mendel rats only after prolonged (3 weeks) dosing of a maximum tolerated dose of TRI (1100 mg/kg/day). Increased liver weights and elevated DNA synthesis levels were not coupled with observable bistopathology or decreased DNA content per gram of tissue suggesting a simple increase in the number of normal hepatocytes rather than hepatic degenerative/regenera tive changes as seen in the mouse. Similar differences in the responses of Wistar rats and B6C3F1 mice to repeated administra tion of TRI have been reported by Elcombe et al. (1981). Despite reports of weak mutagenic activ ity in in vitro assays (Bronzetti et al., 1978; Greim et al., 1975), TRI (nonepoxide inhib ited) did not readily bind hepatic DNA of B6C3F1 mice in vivo. The observed maxi mum of only 0.62 0.42 alkylated bases/ -------------------------------SL 035301----------------- TRICHLOROETHYLENE INTERACTIONS IN MICE AND RATS 149 106 nucleotides contrasts sharply with the hundreds of alkylations observed when mice were dosed with powerful genotoxins such as dimethylnitrosamine or methylnitrosourea (Lawley, 1976). While qualitative con siderations are of obvious importance in the assessment of the significance of DNA bind ing, strictly quantitative data relative to dose (convalent binding index) have been useful in the ranking of the carcinogenic potential of many tumorigenic compounds (Lutz, 1979). Thus the genotoxic potential of TRI based upon in vivo DNA binding data and numerous negative bacterial mutagenicity assays appears to be very slight. The observed weak genotoxic potential and hepatotoxicity of TRI in B6C3F1 mice dosed with a high level of TRI are consistent with an epigenetic recurrent cytotoxic mech anism of tumorigenic action. However, the observed hepatic hypertrophic response and reported induction of the hepatic mixedfunction oxygenase enzyme system (Norpoth et al., 1974; Pessayre et al., 1979; Savolainen er al., 1977) of animals treated with TRI suggest another possible epigenetic mechanism of action. As suggested by Tennekes (1979) for several other enzyme-in ducing agents (e.g., phenobarbital), this mechanism may simply involve the enhance ment of preexisting oncogenic factors pres ent in this mouse strain. Indeed, the B6C3F1 mouse reportedly has a very high sponta neous liver tumor rate (5 to 45%) (Elashoff et al., 1979). In either case, the B6C3F1 mouse appears to be predisposed metabolically to produce more reactive TRI metab olite than Osborne-Mendel rats and to re spond to subsequent tissue damage by the production of tumors. REFERENCES Axelson, O., Andersson, K., Hogstedt, C., Holmberg, B., Mouna, G. and de Verdier, A. (1978). A cohort study on trichloroethylene exposure and can cer mortality. J. Occup. Med. 20, 194-196. Banerjee, S,, and Van Duuren, B. L. (1978). Co valent binding of the carcinogen trichloroethylene to hepatic microsomal proteins and to exogenous DNA in vitro. Cancer Res. 38, 776-780. Bartsch, H,, Malaveille, C, Barbin, A., and Planche, G. (1979). Mutagenic and alkylating me tabolites of halo-ethylenes. chlorobutadienes and dichlorobutenes produced by rodent and human liver tissues. Arch. Toxicol. 41, 249-277. Bolt, H. M,, Buchter, A., Wolowskj, L., Gill, D. L,, and Bolt, W. (1977). Incubation of '*C-trichloroethylene vapour with rat liver microsomes: Up take of radioactivity and covalent protein binding of metabolites. Ini. Arch. Occup. Environ. Health 39, 103-111. Bolt, H. M., and Filser, J. G. (1977), Irreversible binding of chlorinated ethylenes to macromolecules. Environ Health Perspect. 21, 107-112. BRONZETTl, G., ZEIGER, E., AND FREZZA, D, (1978). Genetic activity of trichloroethylene in yeast J. En viron Pathol. Toxicol. 1, 411-418. Burton, D. (1956). A study of the conditions and mechanism of the diphenytamine reaction for the es timation of deoxyribonucleic acid. Biochem. J. 62, 315-323. Duprat, P.. and Gradiski, D. (1980). Cytogenetic effect of trichloroethylene in the mouse as evaluated by the micronucleus test IRCS Med. Sci. Ubr. Compend. 8, 182. Elashoff, R. M,, Preston, D. L,, and Fears, J. R. (1979). Comparison and evaluation of some experi mental designs for use in carcinogen screening. J. Nat. Cancer Inst. 62, 1209-1219. Elcombe, C. R., Pratt, I,, and Rose, M. S. (1981). Species differences in the subacute toxicity of tri chloroethylene. Toxicologist 1, 70 (Abstract). Ertle, T., Henschler, D,, Muller, G.. and Sfassovski, M. (1972). Metabolism of trichloroeth ylene in man. I. The significance of trichloroethanol in long-term exposure conditions. Arch. Toxicol. 29, 171-188. Fahrig, R. (1977). The mammalian spot test (fellfiecken-test) with mice. Arch. Toxicol. 38, 87-98. Filser, J. G,, and Bolt, H. M. (1979). Pharmaco kinetics of balogenated ethylenes in rats. Arch. Tox icol. 42, 123-136. Greim, H., Bonse, G,, Radwan, Z., Reichert, D., and Henschler, D. (1975). Mutagenidty in vitro and potential carcinogenicity of chlorinated ethylenes as a function of metabolic oxirane formation. Biochem Pharmacol. 24, 2013-2017. Henschler, D., Eoer, E. Neuoeckbr, T., and Metzler, M. (1977). Carcinogenicity of trichloroethyl ene: Fact or artifact? Arch. Toxicol. 37, 233-236. Henschler, D., Romen, W,, Elsasser, H. M., Reichert, D., Eder, E, and Radwan, Z. (1980). Carcinogenicity study of trichloroethylene by long- SL 035302 150 STOTT, QUASI, AND WATANABE term inhalation in three animal species. Arch, Tox icol 43, 237-248. Ikeda, M., AND Imamura, T. (1973). Biological halflife of trichloroethylene and tetrachloroethylene in human subjects. lot. Arch. Arbeitsmed. 31, 209-224. Jollow, D. L., Mitchell, J. R,, Potter, W. Z,, Davis, D. C,, Gillett, J. R., and Brodie, B. B. (1973). Acetaminophen-induced hepatic necrosis. J. Pharmacol Exp. Ther. 187, 195-202. KlMMERLE, G., and Eben, A. (1973). Metabolism, excretion and toxicology of trichloroethylene after inhalation. 2. Experimental human exposure. Arch. Toxicol 30, 127-138. Kirby, K.`S. (1957). A new method for the isolation of deoxyribonucleic acids; Evidence on the nature of bonds between deoxyribonucleic acid and protein. Biochem. J. 66, 495-504. Konietzko, H. K., Haberlandt, W., Heilbronner, H., Reill, G., and Weichardt, H. (1978). Cytogenetische untersuchungen an trichloSthylen-arbei- tem. Arch. Toxicol 40, 201-206. Laib, R. J., Stockle, G,, Bolt, H. M,, and Kunz, W. (1979). Vinyl chloride and trichloroethylene: Comparison of alkylating effects of matabolites and induction of preneoplastic enzyme difiriencies in rat liver. J. Cancer Res. Clin. Oncol 94, 139-147. Laskin, S., Sellakumar, A. R., Kuschner, M., Nelson, N,, LaMendola, S., Rusch, G. M,, Katz, G. V., Dulak, N. C., and Albert, R. E. (1980). Inhalation carcinogenicity of epichlorohydrin in noninbred Sprague-Dawley rats. J. Nat. Cancer Inst. 65, 751-757. Lawley, P. D. (1976). Methylation of DNA by car cinogens: Some applications of chemical analytical methods. In Screening Tests in Chemical Carcino genesis (M. Montesano, H. Bartsch, and L. Tomatis, eds.), pp 181, 208. 1ARC Sri. Publ. No. 12, Lyon. Lowry, O. H., Rosebrouoh, N. J., Farr, A. L., and Randall, R. J. (1951). Protein measurement with the Folin phenol reagent J. Biol Chem. 193, 265- 275. Lutz, w. K, (1979). In vivo covalent binding of organic chemicals to DNA as a qualitative indicator in the process of chemical carcinogenesis. Mutat. Res. 65, 289-356. Marmur, J. (1961). A procedure for the isolation of deoxyribonucleic acid from micro-organisms. J. Mol. Biol 3, 208-218. Monster, A. C., Boersma, G,, and Duba, W. C. (1976). Pharmacokinetics of trichloroethylene in vol unteers; influence of workload and exposure concen tration. Int. Arch. Occup. Environ. Health 38, 87- 102. Muller, G., Spassovski, M,, and Henschler, D. (1974). Metabolism of trichloroethylene in man. II. Pharmacokinetics of matabolites. Arch. Toxicol 32, 283-295. National Cancer Istitute (1976). Carcinogenesis Bioassay of Trichloroethylene, CAS No, 79-01-6, DHEW Publ. No. (NIH) 76-802. Norpoth, K., Witting, U.. and Springorum. M. (1974). Induction of microsomal enzymes in the rat liver by inhalation of hydrocarbon solvents. Int. Arch. Arbeitsmed. 33, 315-321. Parchman, L. G., AND Magee, P. (1980). Production of l4C0i from trichloroethylene in rats and mice and a possible interaction of a trichloroethylene metabo lite with DNA. In Abstracts of Mt. Soc. Toxicol, Washington, D.C., p. A51. Pessayre, D,, ALLEMand, H., Wandscheer, J. C., Descatoire, V., artigou, J-Y., and Benhamou, J-P. (1979). Inhibition, activation, destruction, and induction of drug-metabolizing enzymes by trichlo roethylene. Toxicol. Appl. Pharmacol 49, 355-363. Reitz, R. H,, Gehring, P. J., and Park, C. N. (1978). Carcinogenic risk estimation for chloroform: An al ternative to EPA's procedures. Food Cosmet. Toxi col 16,511-514. Reitz, R. H., Watanabe, P. G., McKenna, M. J., Quast, J. F,, and Gehring, P. J. (1980a). Effects of vinylidene chloride on DNA synthesis and DNA repair in the rat and mouse: A comparative study with dimenthylnitrosamine. Toxicol. Appl Pharmacol 52, 357-370. Reitz, R. H., Quast, J. F., Stott, W. T,, Watanabe, P. G., and Gehring, P. J. (1980b). Pharmacoki netics and macromolecular effects on chloroform in rats and mice: Implications for carcinogenic risk es timation. In Water Chlorination: Enivronmemal Im pact and Health Effects (R. L. Jolley, W. A. Brungs, R. B. Cumming, and V. A. Jacobs, eds.), VoL 3. pp. 983, Ann Arbor Sri. Publ., Ann Arbor, Mich. Savolainen, H,, Pfaffu, P., Tenoen, M., and Vainio, H. (1977). Trichloroethylene and 1,1,1-trichloroethane: Effects on brain and liver after five days intermittent inhalation. Arch. Toxicol 38, 229-237. Schumann, a. M,, Quast, J. F,, and Watanabe, P. G. (1980). The pharmacokinetics and macromo lecular interactions of perchloroethylene in mice and rats as related to oncogenicity. Toxicol Appl Phar macol 55, 207-219. Shahin, M. M., and Von Borstel, R. C. (1977). Mutagenic and lethal effects of benzene hexachloride, dibutyl phthalate and trichloroethylene in Saccharomyces cerevisiae. Mutat. Res. 48, 173-180. Simmon, V. F,, Kauhanen, K., and Tardiff, R. G. (1977). Mutagenic activity of chemicals identified in drinking water. In Progress in Genetic Toxicology (D. Scott, B. A. Bridges, and F. H. Sobels, eds.), pp. 249. 258. Elsevier/North-Holland, Amsterdam. Slacik-Erben, R,, Roll, R., Franke, G., and Uehleke, H. (1980). Trichloroethylene vapors do not produce dominant lethal mutations in male mice. Arch. Toxicol. 45, 37-44. Steel, R. G. D., and Torrie, J. H. (1960). Principles SL 035303 4 i, M. e rat irch. .ction : and taboicol., ' c., MOU, and chlo-363, ^78). n al~oxi- l. J., Tects )NA with tcol ABE, :oki- *tgs, . pp- AND -triiays 237. ABE, jnoand har~ 77). idc, ha- . G. il in ogy PP- iND not ice. lies r TRICHLOROETHYLENE INTERACTIONS IN MICE AND RATS 151 and Procedures of Statistics. McGraw-Hill, New York. Stott, W. T., Reitz, R. H,, Schumann, A. M., and Watanabe, P. G. (1981). Genetic and nongenetic events in neoplasia. Food Cosmet. Toxicol, 19, in press. Swenson, D. H., and Lawley, P. D. (1978). Alkyl ation of deoxyribonucleic acid by carcinogens di methyl sulfate, ethyl methanesulfonate, N-ethyl-Nnitrosourea, and N-methyl-N-nitrosourea. Biochem. J. 171, 575-587. Tennekes, H. A. (1979). The Relationship Between Microsomal Enzyme Induction and Liver Tumour Formation. Centre Agr. Publ. Docum., Wageningen. Uehleke, H., and Poplawski-Tabarelli, S. (1977). Irreversible binding of ,4C-labeled trichloroethylene to mice liver constituents in vivo and in vitro. Arch. Toxicol. 37, 289-294. Van Duuren, B. L., and Banerjee, S. (1976). Co valent interaction of metabolites of the carcinogen trichloroethylene in rat hepatic microsomes. Cancer Res. 36, 2419-2422. WasKell, L. (1978). A study of the mutagenicity of anesthetics and their metabolites. Mutat. Res. 57, 141-153. Watanabe, P. G., Reitz, R. H., Schumann, A. M., McKenna, M. J,, Quast, J. F., and Gehring, P. J. (1980). Implications of the mechanism of tumorigenicity for risk assessment. In The Scientific Basis of Toxicity Assessment (H. R. Witschi, ed.), Vol. 6, pp. 69, 89. Elsevier/North Holland, New York. Weisburger, J. H., and Williams, G. A. (1980). Chemical Carcinogens. In Toxicology: The Basic Science of Poisons (J. Doull, C. D. Klaassen, and M. O. Andur, eds.), pp. 84, 138. Macmillan, New York. SL 035304