Document DvQkqdq42MY5y2XaQLOZgM2gB

K OI.OGV AND APPUIID PHARMACOLOGY 103, 77-89 (1 ^SJO) Perchloroethylene-lnduced Rat Kidney Tumors: An Investigation of the Mechanisms Involved and Their Relevance to Humans Trevor Green, Jenny Odum, John A. Nash, and John R. Foster Imperial Chemical Industries pic. Central Toxicology Laboratory. Alderley Park. Macclesfield. Cheshire, SKI04TJ. United Kingdom ReceivedJuly 17, 1989; accepted October 23. 1989 Perchloroethylene (tetrachloroethylene) is a colorless, volatile liquid of low acute toxicity which is used extensively as a dry cleaning agent and industrial degreaser. Two lifetime animal carcinogenicity studies have been re ported in which perchloroethylene was ad ministered to rats and mice by oral gavage (NCI, 1977) and by inhalation (NTP, 1985). Both studies reported a significant increase in hepatocellular carcinoma in male and female BfiCjF! mice and in the inhalational study a low incidence of renal adenoma and adeno carcinoma in male F344 rats. Reduced sur vival ofOsbome-Mendel rats during the oral study prevented any conclusions being drawn from that experiment. An increased inci dence of mononuclear cell leukemia was also seen in F344 rats in the inhalation study. A mechanism has been proposed for the development ofliver tumors in mice exposed to perchloroethylene which is consistent with the absence of liver tumors in rats (Odum et ai, 1988). Mononuclear cell leukemias occur at very high rates in control F344 rats, roak- 0041-008X/90 S3.00 77 Copyright 1990 by Academic not, Inc- All rights ofreproduction in any form reserved. 78 GREEN ET AL. ing interpretation of this finding difficult. The Although these studies suggest several pos incidence of adenomas and adenocarcino sible mechanisms for the renal cancer ob mas in the kidneys of male F344 rats was con served after inhalational exposure to perchlo sidered not to be statistically significant. roethylene, the conclusions are drawn from a However, kidney adenomas are rarely seen in limited number of experiments conducted in F344 rats and adenocarcinomas have not male rats given high oral doses of perchloro been seen in historical controls at the testing ethylene. It is not known whether these laboratory (NTP, 1985). Thus, the low inci- mechanisms are consistent with the observed denceofludncy tumors in rats may beoftoxi species differences nor has the relevance of cological ifnot statistical significance. these mechanisms to humans been deter Several studies have sought an explanation mined. In addition the quantitative extrapo for the kidney tumors seen, in male rats ex lation of results from gavage studies is diffi posed to perchloroethylene. >4ouse liver tu cult when the most appropriate route of ex rnon are believed to be a itsvit of percMoro* posure for humans, and the one used in the ethylene-induced hqntic pqxaisoTOeiH-oliP animal bioassay, is by inhalation. erattan, but earlier studies have shown only In the present study the various mecha margroal mcreases in peroxisomes in the rat nisms of perchloroethylene-induced renal Jfidney (Goldsworthy and Popp, 1987; Odum cancer have been investigated in rats and et ai, 1988). Male rats given faigh doaes of mice of both sexes. Comparisons have been perchkrroethyiene by onri gavage bave been made between animals dosed by oral gavage found to accumulatethe protein ^t-gMtu Mnd exposed by inhalation, and finally, in vi hn in renal proximal tubulareeftx and to ex tro techniques have been used to compare the hibit all the features of protein droplet ne metabolic pathways of perchloroethylene in phropathy (Green et ai, 1986, Goldsworthy rats and mice with those in humans. et al,, 1988). The strong correlation between this epigenetic effect and male rat-specific kidney cancer has led to suggestions that this METHODS is the mechanism responsible (Goldsworthy etal., 1988). Materials Perchloroethylene is metabolized primar ily by oxidation to trichloroacetic acid (Yllner, 1961; Daniel, 1963; Dekant et at., 1985). Evidence of a second metabolic pathway for perchloroethylene, which may be related to the development of kidney tumors, has also been reported (Dekant et al., 1986; Odum and Green, 1987), The presence of a second 1,1,2,2-Tetrachioroethylene (Analar grade, 99.9%) was obtained from BDH Chemicals Ltd. (Poole, Dorset, U.K.), Biochemicals were obtained from Sigma Chemi cal Co. (Poole, Dorset, U.K.). 1,1,2,2,-Tetrachloro{ 1,2->4C]ethylene, specific activity 55 mCi/mmol was obtained from Imperial Chemical In dustries pic. Physics and Radioisotope Services (Billingham, Cleveland, U.K.). The material had a chemical and radiochemical purity of greater than 98% as determined pathway, which is believed to be analogous to that known for hexachlorobutadiene (Nash et al., 1984), was suggested following the identi fication of N-acetyl-S-( 1,2,2-trichlorovinyl)L-cysteine in urine from rats dosed with per by radio-gas chromatography. .S-(l,2,2-Trichlorovinyl)L-cysteine (TCVC), A/-acetyUS-(l,2,2-trichlorovitiyl)-Lcysteine (NAc-TCVC), and S-(-1,2,2-trichlorovinyl)glutathione (TCVG) were synthesized as described by Moore and Green (1988). chloroethylene, The cysteine conjugate of perchloroethylene is known to be activated Animals and Dosing by the renal enzyme 0-lyase to a nephrotoxic intermediate, which has also been shown to be mutagenic in a modified Ames bacterial mutation assay (Green and Odum, 1985). Male and female Fischer 344 rats (160-190 g) and male and female B4C3F, mice (22-25 g) were supplied by Charles River (Margate, Kent, U.K.). The animals were housed in temperature-controlled rooms fitted with a 12- est several posfe|l cancer obBre to perchloi drawn from a s conducted in s of perchlorowhether these h the observed e relevance of s been deterative extrapo;tudies is diffi.e route of exle used in the in. irious mechaiduced renal t in rats and ins have been >y oral gavage finally, in vi) compare the roethylene in ans. Y, grade, 99.9%) (Poole, Dorset, i Sigma Chemi- specific activity al Chemical In dices (Billinga chemical and i as determined frichlorovinyl)chlorovinyi)-Lhlorovinyl)gIu- described by >0-190 g) and sre supplied by : animals were ittcdwitha 12- PERCHLOROETHYLENE-INDUCED KIDNEY TUMORS 79 hr lighting cycle. Feed (PCD diet. Special Diets Services Ltd ._Witham. Essex. U.K..) and water were provided ad libitum. Gavage studies. A group of ten male rats each received single doses of pcrchloroethylenc (1500 mg/kg) in com oil (10 ml/kg) by gavage daily for 42 days. A control group often male rats received corn oil (10 ml/kg) alone. Inhalation studies. Rats and mice of both sexes, five animals per group, were exposed to 400 ppm perchloroethylene, 6 hr/day, for 28 consecutive days. Exposures were whole body in stainless-steel chambers (Doe and Tinston, 1981) having an internal volume approximately 3.4 m\ The chambers were air conditioned to have a nominal temperature of 22'C and a relative humidity of 40-60%. The air flow through the chambers was 300 liters/min. Atmospheres were generated by passing vapor ized perchloroethylene into the input air ofthe chamber. Controls were exposed to air alone. In a further study male rats, three per group, were ex posed to 10 or 100 ppm perchloroethylene for 6 hr, or to 1000 ppm 6 hr/day for 1, 5, or 10 days. Toxicity Studies All ofthe animals were killed by overexposure to halothane. In the gavage study the rats were killed 24 hr after the final dose. During this period urine was collected. Blood samples were collected by cardiac puncture at ter mination and kidneys removed for histopathological ex amination. Rats exposed to 400 ppm perchloroethylene by inhalation were killed 18 hr after the end ofthe final 6hr exposure and the kidneys removed for histopathology. Body weights were recorded during the experiments and organ weights at termination. Blood and urine samples from the gavage study were analyzed for biochemical markers of both liver and kid ney damage. Urine samples were combined to give a sufficiently large sample for the biochemical assays and the subsequent metabolism studies. Plasma urea, alka line phosphatase (ALP), alanine aminotransferase (ALT), and urinary glucose, protein, ALP, and iV-acetyl-0-D-glucosaminidase (NAG) were determined as described by Nash et al. (1984). Urine samples were analyzed for a2u-globulin by the method ofStonard eta! (1987). Slices ofkidney were fixed in 10% neutral buffered formol saline, dehydrated through an ascending ethanol se nes, and embedded in paraffin wax. Sections (5 pm) were cut and stained with hematoxylin and eosin for light mi croscopic examination. Analysis of kidney sections for hyaline droplet formation was performed on the hematoxylin/eosin-stained sections under UV light on a Reichert Polyvar photomicroscope. Tissues were also fixed in 3% glutaraldehyde in 0.1 m sodium phosphate buffer, dehydrated, and embedded in epoxy resin. Sec tions (1 (im) were cut and stained with I % toluidine blue in 1% borax for light microscopy. Suitable areas were then selected from the proximal tubules of the kidney for electron microscopy. Ultrathin sections ofthese areas were stained with uranyl acetate and lead citrate and viewed and photographed in a JEOL JEM 100CX elec tron microscope. Metabolism and Pharmacokinetics Gavage studies. Radiolabeled perchloroethylene (10-- 40 uCi per rat) was incorporated into the doses given on Days 1,17, and 42 of the gavage study. After dosing the animals were transferred to metabolism cages for the sep arate collection ofurine and feces for 24 hr. On Day 42 of the experiment two rats were fitted with biliary cannulas immediately before the final dose. Bile was collected for 24 hr. Inhalation studies. Urine was collected from rats and mice on Days f, 7, and 14 of exposure to 400 ppm petchioroethyiene. The animals were placed in metabolism cages for the 18-hr period between exposures. Rats ex posed to 10 or 100 ppm perchloroethylene for 6 hr and to 1000 ppm 6 hr/day for 1,5, or 10 days were placed in metabolism cages at the end of the last exposure period in each case. Urine was collected for 18 hr. Analysis ofmetabolites in urine and bile. Bile from rats dosed with perchloroethylene by gavage was initially an alyzed for the presence of S-(1.2,2-trichlorovinyl)glutathione by thin-layer chromatography alongside an au thentic standard. The samples were analyzed on silica gel GF plates with a developing solvent of ethyl apetate: methanoEacetic acidrwater (60:15:15:10, v/v). - Bile (6.5. ml) was also incubated with `Y-glutamyltranspeptidase (GGT, 50 units) at 37`C for 3 hr. Protein was precipitated with ice-cold ethanol (13 ml) and removed by centrifugation. The supernatant was concentrated un der reduced pressure and chromatographed on prepara tive thin-layer plates using the same solvent system. Au thentic TCVG, which had been incubated with GGT in the same way as the bile sample, was run as a standard on each plate. An area ofeach plate, 3 cm on either side of the RF of the hydrolyzed TCVG (0.7) was removed and eluted with methanol (100 ml). The eiuate was evap orated to low volume (1 ml) under reduced pressure and an excess ofan ethereal solution ofdiazomethane added. When the reaction was completed the excess diazometh ane was removed, and 10 ml ofdichforomethane added followed by 2 ml of trifluoroacetic anhydride. The sam ple was allowed to stand at room temperature for 15 min, the dichloromethane evaporated off, and the residue dis solved in methanol (0.1 ml). The GGT-hydrolyzed TCVG standard was also eluted from the plate and derivatized in the same way. The samples were analyzed by gas chromatography-mass spectrometry using a VG 7070E instrument fitted with a methyl silicone column (15 m X 0.53 mm) operated at I00*C for 1 min followed by l0*C/min up to 200*C. The flow rate was 5 ml/min he lium. 80 GREEN ET AL. For comparison with the Day 42 urine samples used for the biochemical assays the urine samples taken at the other time points ofthe gavage study were similarly com* bined, as were the samples at each time point of the 400* ppm inhalation study. Samples (2 ml) were acidified with concentrated hydrochloric acid (0.1 ml) and extracted with ethyl acetate (8 ml). The solvent phase was sepa rated. concentrated to 0.1 ml. and methylated with ethe real diazomethane solution as described previously. Ex cess diazomethane was removed and the residue dis solved in methanol (0.2 ml). Authentic NAc-TCVC was derivatized in the same way. The samples were analyzed by gas chromatography-mass spectrometry using the equipment described above. The column was operated from 130 to 2I5*C at a rate of 15`C/min after a 1-min delay. Under these conditions the methyl ester of NAcTCVC had a retention time of 7.5 min. Individual urine samples from rats exposed to 10, 100, and 1000 ppm perchloroethylene were analyzed for tri chloroacetic acid and NAc-TCVC by gas chromatogra phy and gas chromatography-mass spectrometry, re spectively. Both metabolites were extracte4 into ethyl acetate at pH I and methylated with diazomethane as described above. Trichloroacetic acid was determined as described by Odum el at (1988) and NAc-TCVC as de scribed above. In Vitro Metabolism Studies Renal metabolism. Activity of the renal enzyme ^-ly ase with TCVC as substrate was measured in cytosolic fractions from rat, mouse, and human kidney. Human kidney samples were obtained immediately'following surgery from kidneys removed as a result of malignancy or renal failure. The tissues were in each case from parts of the kidney which were morphologically normal. Ani mal kidney samples were taken from animals killed by overexposure to halothane. Cytosol fractions were prepared by homogenizing kid ney samples in 4 vol of ice-cold 0.25 m sucrose/10 mM Tris-HCl buffer, pH 7.5, with a Teflon/glass homogenizer. The homogenates were centrifuged at 105,000# for 1 hr and the supernatant (cytosol) was removed and stored at -70`C until use. Protein was determined by the method of Lowry el al.( 1951). /3-Lyase activity was measured using a method based on that ofStevens and Jakoby (1983). The reaction mix ture contained, in a final volume of 1 ml, 50 mM Na/ K phosphate buffer, pH 8.0, lactate dehydrogenase (0.1 unit), NADH (100 mm), approximately I mg of protein, and TCVC (0.2-7.0 mM). The reaction (37*C for 5 min) was started by the addition ofTCVC and was monitored at 340 nm. Michaelis-Menten constants K,, and V^ were obtained from Lineweaver-Burk plots ofs/v versus s. Hepatic metabolism. Conjugation of perchloroethyl ene with glutathione was assayed in microsomal and cy tosolic fractions of livers from rats, mice, and humans. Human liver samples were obtained from brain dead re nal transplant donors after compliance with ethical and legal requirements. Animals were killed by overexposure to halothane and the livers removed immediately. Livers from rats and mice were homogenized with a Teflon/glass homogenizer in sucrose (250 mM):EDTA (5.4 mM):Tris-HCl buffer, pH 7.5 at 4'C to give 25% (w/ v) homogenates. These were centrifuged at 9000g for 20 min, and the supernatants were transferred to fresh tubes and centrifuged at 105,000# for a further 70 min. The supernatants (cytosol) were stored at -70*C and the pel lets (microsomal fraction) were resuspended in the same buffer and centrifuged at 105,000# for I hr. The washed microsomal pellets were suspended in buffer and stored at -70*C. Human liver samples were homogenized in a Waring blender and cytosol and microsomal fractions prepared as for rats and mice. Incubations contained, in a final volume of 3 ml, 0.1 m phosphate buffer, pH 7.4, glutathione (5 mM), gluta thione reductase (5 units), and either 25 mg cytosolic or 15 mg microsomal protein. The reaction was started by the addition of [14C]perchloroethylene (10 mM; 5 pCi) and the samples were incubated for up to 2 hr at 37*C. The [l4C]perchloroethylene used in these studies was pu rified immediately before use by preparative gas chroma tography. The reaction was stopped by the addition of 50% trichloroacetic solution (0.1 ml) and the precipitated protein removed by centrifugation. Excess perchloroeth ylene was removed from the supernatant by extraction with diethyl ether (3x5 ml). The aqueous-residue was then derivatized using J -fluoro-2,4-dinitrobenzene as de-, scribed by Reed et at. (1980) prior to analysis by highpressure liquid chromatography. Aliquots of the derivatized supernatant (50 *d) were analyzed using a 4.9-mm x 25-cm, l0-*im Lichrosorb NH; column eluted with a solvent gradient containing 3 m sodium acetate, pH 4.5, over 40 min. Fractions of 1 ml were collected and assayed for radioactivity by liquid scintillation counting. TCVG and TCVC were also de rivatized and analyzed under the same conditions. RESULTS Male rats given a daily ays showed an increase in kidney/body weight ratios (1.3fold) and evidence of kidney damage. Sig nificant increases in urinary volume (+35%), glucose (+45%), ALP (+90%), and NAG (+57%) levels compared to controls indicated mild proximal tubular damage. Examination of the kidneys from these animals at the light microscope 1 in SL 035485] PERCHLOROETHYLENE-INDUCED KIDNEY TUMORS 81 "nits higher than the first, corresponding to ;The droplets the addition of a further methyl group, pre were increased not only in number but also in sumably at the nitrogen of cysteine. The de size and shape, larger angular droplets being rivatized extract ofGGT-hydrolyzed bile was dualized (Fig. 1). In addition to the droplets, analyzed by GC/MS in the selective ion de tabular casts were seen at the corticomedul* tection mode, focusing on ions m/e 360 and lary junction, with distortion and enlarge 362 from the first peak and m/e 374 and 376 ment of the tubules containing them. Focal from the second peak. The bile extract was areas of basophilic proximal tubular regener found to contain all four ions in the correct ation were also apparent. ratios with retention times identical to those Rats and mice of both sexes were also ex of the standard (9.5 and 10.5 min) (Fig. 2). posed to perchloroethykne by inhalation at Urine.from both the gavage and inhalation dose levels of up to 400 ppm for 28 days.do studies was analyzed for the presence of N- this experiment kidney/body weight ratios . acetyl-5-(1,2,2-trichlorovinyl) - l - cysteine. were unchanged, nor were there any com'-' Gas chromatograms of methylated urine ex- pound-related changes in the kidneys of ei tracts were shown to contain a peak (7.8 min) ther species at the light or electron micro with a mass spectrum identical to that of scope levels. Increases in hyaline 4roplets authentic (V-acetyl-SH 1,2,2-trichlorovinyl)- were not seen in male rat kidneys, nor could L-cysteine methyl ester (Fig. 3). Quantitation any increase in a-2u-globulin be detected bio Of the levels Of chemically in the kidneys of these animals n> . ______________ ________ . (data not shown). The kidneys taken from rats exposed by inhalation 4o tOOO ppmper- chloroethylene for to days did show a dear iationalf increase in hyaline droplets in the proximal tubules. This was not accompanied by the urrrte'QTiOCHTft samFconditionsoTexposure cast formation or cellular regeneration that (Table 1). There was some evidence that the were seen after gavage dosing. levels in mouse urine were increasing with in Bile collected from male rats for the 24 hr creasing duration of exposure. following the final dose of perchloroethylene The activity of cysteine conjugate /3-lyase on Day 42 of the gavage study was analyzed with the cysteine conjugate of perchloroeth- by thin-layer chromatography and found to ylene as substrate was determined in kidney contain radioactive material identical with cytosol fractions from mouse, rat, and hu authentic S-( 1,2,2-trichlorovinyl)gIutathione mans. Kinetic constants obtained from plots (Rf 0.3). The identity of this metabolite was ofs/v against s are shown in Table 2. The re confirmed following hydrolysis of the bile action was linear with protein concentration with -y-glutamyltranspeptidase in compari over the range used and had a pH optimum son with authentic S-( 1,2,2-trichlorovinyl)- of 8.0. Addition of aminooxyacetic acid (0.1 glutathione. Hydrolysis of the standard with mM) to those experiments reduced activity by GGT resulted in the formation of SK 1,2,2- 50%. The Km was lower and the higher trichlorovinyl)-L-cysteine rather than the ex for the rat compared to either mouse or hu pected S-{ 1,2,2-trichlorovinyl)cystemylgly- man kidney 0-lyase. A comparison of V--/ cine, presumably as a result ofcontamination Km for the three species gave similar values of the GGT with a peptidase which cleaved for mouse and human, whereas the value for the cysteinylglycine bond. The mass spec the rat was up to 30-fold greater, with the trum of the derivatized product is shown in value for the male rat being approximately Fig. 2. The derivatization procedure resulted twice that for the female (Table 2). The varia in two peaks, the second having a mass 14 tion in 0-lyase activity in human kidney sam- SL 035686 Fig. t. Light micrographs taken from the kidney ofa mate rat given perchloroethylene, 1500 mg/kg body wt/day for 42 days, shows (a) the cortex where proximal tubules in the S3 region exhibit an accumulation of hyaline droplets in the cytoplasm ofthe cells, and (b) the corticomedullary area ofthe same kidney. There is an accumulation of proteinaceous debris in the tubules in this region, which appear thin walled due to a compressed layer ofepithelial cells and dilated due to the debris. H&E, X400. ttt PERCHLOROETHYLENE-INDUCED KIDNEY TUMORS 83 ! 'kg body lation of y There due to a 100i 8060- 212 376 20 69 248 1 268 344 100 150 200 250 300 350 400 450 m/e Fig. 2. Chemical ionization (isobutane) mass spectra obtained for the AMrifluoroacetyl methyl ester derivatives ofS-( 1,2,2-trichlorovinyl)-L-cysteine. Inset: Chromatogram of a GGT-hydrolyzed bile extract from male rats dosed orally with perchloroethylene (1500 mg/kg) showing the presence ofthe N-trifluoroacetyl methyl ester derivative ofS-( 1,2,2-trichlorovinyl)cy$teine by selective ion detection. SL 035488 84 GREEN ET AL. 100! 43 80 60 55 88 105 Miami40H ca C<C*> 20 69 77 0 .50 100 Fig. 3. Electron impact mass spectrum ofthe methyl ester ofA'-acetyl-S^ 1,2,2-trichlorovinyl)-L-cysteine extracted from the urine of male rats dosed orally with perchloroethylene (1500 mg/kg). pies was remarkably small. In 11 individuals the rates determined for a single substrate concentration ranged from 0.1 to 0.56 nmol/ min/mg protein. K,, and V__values ob tained for 7 of those individuals were also re markably close (Table 2). Hepatic conjugation of perchloroethylene with glutathione in rat liver fractions was found to occur primarily in the cytosolic frac tion, the rate (18.2 pmol/min/mg protein) be- TABLE 1 (V-ACETYL-S-( 1,2,2-TRICHLOROVINYL)-L-CYSTEINE LEVELS IN RAT AND MOUSE URINE after Exposure to Perchloroethylene (V-Acetyl-S-/ 1,2,2-trichlorovinyl)-L-cysteine (jig/ml) Dose Duration (days) Male Rat female Mouse Male female Inhalation exposure. 400 ppm, 6 hr/day 1 0.75 1.30 0 0 7 2.04 0.90 0.07 0 Oral, 1500 mg/kg/day 14 0.55 1.04 0.20 0.15 1 23.00 ___ a -- WWW. 17 41.11 -- -- ,_ 42 32.73 -- ---- Note. Measurements were carried out on pooled urine samples from five animals (inhalation study) or three animals (oral study). " --, Not determined. SL 035489 Ear perch I 5 ' 350 m/e L-cysteine microsomal otein). The dly between jeing five to *use liver cy. In these ex ceectled in huft"mal or cy^ctilon ofthe Id be at least ; le (*>g/ml) louse Female 0 0.15 ___________ t | 1 three animals PERCHLOROETHYLENE-INDUCED KIDNEY TUMORS TABLE 2 Metabolism of S-(L2,2-trichlorovinylR-cysteine by Kidney Cytosolic /3-Lyase Species" Sex A',,(mM) En*. (nmol/min/mg) \ll M 0.68 0.06 t;al F 1.26 0.21 Mouse M 5.69 2.22 Mouse F 4.43 1.42 Human (n - 3) M 2.53 0.09 Human (n = 4) F 2.67 2. II 4.00 0.11 3.64 0.41 1.15 0.31 1.66 0.27 0.49 0.07 0.64 0.54 Note. Values (means SD) were calculated from plots of.s/v against s. " n = 4 unless otherwise stated. 85 5.88 2.88 0.20 0.37 0.21 0.24 m order of magnitude less than that in the rat. The extremely low rates of glutathione conjugation found, even in the rat, prevented the accurate determination of Km and V__ for this pathway, the data given beiri'g for a substrate concentration of 10 mM. Assay of the same cytosol fractions with l-chloro-2,4dinitrobenzene as a substrate for the glutathione-S-transferase enzymes gave considerably higher rates, rat cytosol (430 nmol/min/mg protein) being similar in activity to human cytosol.(442 nmol/min/mg protein). A comparison of the metabolism of perchloroethylene by the main cytochrome P450 pathway, based on urinary trichloro acetic acid levels, with that by the minor glu tathione pathway, based on urinary A-acetylS-( 1,2,2-trichlorovinyl)-L-cysteine levels, is shown in Fig. 4. The data shown are the uri nary levels of each metabolite following sin gle exposures to dose levels of 10, 100, and 1000 ppm. Exposure to 1000 ppm perchloroethylene for 10 consecutive days had no sig nificant effect on the levels of either metabo lite compared to a single exposure (data not shown). It is clear from Fig. 4 that the gluta thione pathway is minor at low dose levels but begins to increase significantly following saturation of the cytochrome P450 pathway. DISCUSSION Early investigations of the mechanism of perchloroethylene-induced cancer centered around the formation and reactivity of the epoxide believed to be formed during the oxi dative metabolism of this chemical (Bonse et al., 1975; Greim et al., 1975; Henschler, 1977; Bolt et al., 1982; Bartsch et al., 1979). Against such a mechanism was a marked lack ofdetectable genotoxicity in short-term tests, suggesting that other nongenotoxic mecha nisms may be important. This was found to be the case for the liver tumors seen in extion, mediated ! Fig. 4. Dose-dependent metabolism of perchlorocthylene in the male rat. Rats were exposed to 10, 100, or 1000 ppm perchloroethylene for 6 hr. Urine was col lected for the following 18 hr and analyzed for trichloro acetic acid (TCA) and A-acetyl-5-( 1,2,2-trichlorovinyl>L-cysteine (TCVC-NAc). Each point represents the mean SD of three animals. The data point at 400 ppm (TCVC-NAc) is taken from Table I. 86 GREEN ET AL. ini^j<itl<iiffljti m liii liliini ii i tii acid, prondn^KKIRPiaqation both for the liver tumcttaMgsihialce and the species differences * \ | IT*8**4 '***j~ (Odum et al., 1988). From the present study it would appear that the low incidence of male rat kidney tumors is not related to epoxide formation or peroxi some proliferation (Odum et al,, 1988) but to a combination of insults including chronic toxicity, hyaline droplet formation, and the formation of mutagenic cysteine conjugates." Prior to the NTP lifetime inhalation study (NTP, 1985) previous lifetime carcinogenic ity studies in rats have been compromised by poor survival .within the dosed groups (NCI, 1977; Rampy et ai, 1978). In these studies survival was reduced, in least in part, as a re sult of chemical-induced nephropathy which could be distinguished from the age-related nephropathy normally seen in F344 rats. Al though mortality was slight in the latest NTP study there was evidence again of kidney damage characterized by tubular enlarge ment and hyperplasia. In the present shortterm gavage studies kidney damage was ap parent both histologically and from changes in plasma and urine markers of toxicity. Sus tained cell damage or cytotoxicity has long been believed to either result in or promote the development of cancer, such a mecha nism having been proposed for a number of chemicals (Reitz, 1987). Consequently, the possibility that the chronic nephropathy seen in the perchloroethylene cancer studies is causally related to the very low incidence of renal tumors in male rats cannot be ignored. At the same time the significance of this type of effect is very difficult to quantify. Mature male iate|$uea high-oral doses of droplets) in. tabular cells. This effect, whidn^as been reported previously with per chloroethylene (Goldsworthy et ai, 1988), is seen in male rats but not in female rats nor in mice of either sex. Hyaline or protein droplet formation was accompanied by cost forma tion and proximal ttibnlar cell regeneration (tubular basophilia), wjjuch believed to be / part of a cycle of necrosis ifaf Regeneration that results in cancer (Bruner, 1984; Charbonneau et ai, 1988; Trump et ai, 1984; f Strasser et ai, 1988; Kanerva et ai, 1987). a significant number of apparently nongenotoxic male rat-specific renal carcinogens have been linked with this effect. Hyaline droplet nephropathy, which is consistent with the male rat specificity ob served in the perchloroethylene cancer bioas says, was first seen in the present studies fol lowing gavage dosing of perchloroethylene. The same response was not seen in male rats exposed to perchloroethylene by inhalation at the dose levels used in the NTP (1985) study, even after exposure to 400 ppm daily for up to 28 days. Thus, although there is a well-established link between hyaline droplet nephropathy and .renal cancer in male rats, the lack of effect at the top dose level used in the NTP study questions the significance of this phenomenon in the development of tu mors in that study. Hyaline droplet formar tion is frequently seen after a single dose of many of the agents that are known to cause this effect (Trump et al,, 1984). The presentstudies measuring hyaline droplet formation in sexually mature male rats exposed to 400 ppm perchloroethylene were continued for up to 28 days without effect. However, al though droplets were not seen even over this larger time scale, the possibility that hyaline droplet formation is a contributory factor to the development of renal tumors in male rats over the lifetime of the animals cannot en tirely be ruled out. The difference in response between male rats dosed with perchloroethylene by gavage and those dosed by inhalational exposure ap pears to be due to differences in the magni tude of the dose. Although difficult to quan tify exactly, 400 ppm by inhalation is a sig nificantly lower dose than 1500 mg/kg by gavage. The^lfef that WJWffi' the thresh0^1^TO^jWBiyau?^yiIc iwnnnawas ooflfirftiecr seen lter*x*K>6u^ chloroet would a; dose sim later con non, cai m. wei The p are belie tion beti olites an to exist genotoxi roethyle by a sea ation v gate is n pathway is excre Odum a metabol conjugal shown ti d-lyase a lerial it 1985). / pathway be a gei ment of lack of r the Aim failure c number and actr In the gate of p from or teine a orally ai chloroei suremei conjuga activity jugate, 1 betweer of these being tf by this i one coi lieved to be i regeneration . 1984; Char et al.. 1984; tat., 1987). A| itly nongcnocinogens have: hy, which is specificity ob: cancer bioasnt studies foliloroethylene. n in male rats by inhalation NTP (1985)' >00 ppm daily ugh there is a yaline droplet in male rats, ; level used in ignificance of jpment of turoplet formasingle dose of to cause ^^he present >let formation ;posed to 400 :ontinued for However, al;ven over this v that hyaline tory factor to rs in male rats Is cannot en- ^etween male me by gavage i exposure apm the magniicult to quanation is a sig00 mg/kg by m was below ; this response droplets were 300 ppm per- PERCHLOROETHYLENE-INDUCED KIDNEY TUMORS 87 . This dose level appear to approximate the threshold dose since hyaline droplets were seen but the later consequences of hyaline droplet forma::on, cast formation and cellular regenera tion, were not developed at this dose level. The possible mechanisms discussed so far are believed not to involve any direct interac tion between perchloroethylene or its metab olites and DNA. A third mechanism appears to exist which does involve the formation of genotoxic metabolites in the kidney. Pcrchloroethylene has been found to be metabolized >y a second minor pathway involving conju gation With glutathione (Fig. 5). The conju gate is metabolized by the mercapturic acid pathway and the ^-acetylcysteine conjugate is excreted in urine (Dekant et al.". 1986; Odum and Green, 1987). In addition to being metabolized to the mercapturate, the cysteine conjugate of perchloroethylene has also been shown to be a substrate for the renal enzyme /3-lyase and to be mutagenkm tft&Aifies bac terial mutation assay (Green and Odum, 1985). As a result of the discovery of this pathway it must.be assumed that there may be a genotoxic component to the develop ment of tumors in the male rat kidney. The lack of response of perchloroethylene itselfin the Ames assay is almost certainly due to a failure of the standard assay to replicate the number of steps involved in the formation and activation ofthis conjugate in vivo (Fig. 5). In the present study the glutathione conju gate of perchloroethylene was detected in bile from orally dosed rats, and the iV-acetylcysteine conjugate in urine from rats dosed orally and from rats and mice exposed to per chloroethylene by inhalation. In vitro mea surement of the rates of hepatic glutathione conjugation of perchloroethylene, and of the activity of renal /3-lyase with the cysteine con jugate, has enabled comparisons to be made between rats, mice, and humans. Tfce-vesQtts of these studies were consisteat'^sf^jhejat being the species sr^^^hfe^^^^^pcer one conjugation ana renal p-lyase cleavage wereboth higher in i rftdfeetb- sues. In vivo the levels ofthe /v-aowyicysteine conjugate in rat urine ^wert significantly higher than those in mouse urine (Table 2). The higher levels of A-acetylcysteine conju gate in urine from orally dosed rats compared to the levels in urine from rats exposed by in halation are consistent with the differences in dose and behavior ofthe two pathways shown in Fig. 4. Glutathione conjugation increases markedly at higher dose levels following satu ration of the cytochrome P450 pathway. Three mechanisms have fafea identified which may cauge opment exposed <4o chronic toxic corfpPKSh^PlfW'^ailYway.lt seems feasi ble that at least two if not all three of these mechanisms contribuite. 'nMMtaxicdutathione pathways chioniccytotom^t^ The glutathione//S-lyase mechanism alone may not be sufficient to in duce tumors in the rat kidney since the levels of iV-acetylcysteine conjugate in female rat urine and the activity of /3-lyase in female rat kidney are not too dissimilar to those found in the male. Male rat-specific protein droplet nephropathy or increased sensitivity of the male rat to the cytotoxic effects of perchloro ethylene may be essential factors in tumorde velopment. to l'tvtve tittle relevance to human risk assessment. Chemically induced nephropathy or cytoxicity and hyaline droplet nephropathy have both been shown to be threshold phe nomena and as such have no relevance at oc cupational exposure levels. Furthermore hya line droplet formation appears to be re stricted to the male rat and is generally considered not to be relevant to humans. Glutathione conjugation of perchloroethyl ene also seems to be a high-dose phenome non in the rat, increasing only after the major 88 GREEN ET AL. Cl Cl GSH Cl Cl \ glutathione- \/ C=C --------------- C=C / \ S-tronsferase / \ Cl Cl Cl SG - glu - gly N--acetyl transferase. Cl Cl OC NHL \ \ * Cl S-CHj-CH COOH B-lyose NHCOCH I3 Cl s-o^-ch I Urine COOH Cl Cl Cl \ I + C=C <""> H-- C -- C-5 s\ I Cl SH Cl interaction with proteins, DNA Fig. 5. Proposed metabolism of perchloroethylene by the glutathione conjugation pathway. cytochrome P450 pathway has become satu rated. Comparisons of the activity of the re nal Myase did detect a rate in humans which was comparable to that in the mouse, but was an order of magnitude lower than that in the rat. However, the in vitro, studies uainghti man liver tissues foiled to detect a rate for the conjugation of perchtoroethylcne with gluta thione. Consequently the presence ofrenal 0lyasein humankidney iaofao toxicological apuficanoe fbr pctchiotoethyieDe since cur rent evidence suggrats that the hver is the source ofthe substrate for this enzyme (Nash et ai, 1984). In conclusion, it would appear that the low incidence of cancer seen in the kidneys of male rats exposed to perchloroeth ylene arises from a combination of events which are largely unique to the male rat and appear to have little relevance to the determi nation of human risk from exposure to this chemical. REFERENCES Bartsch, H., Malaveille, C., Barbin, A., and Planche, G. (1979). Mutagenic alkylating metabo lites ofhalo-ethylenes, chlorobutadienes and dichlorobutenes produced by rodent or human liver tissues. Evidence foroxirane formation by P450-!inked micro somal mono-oxygenases. Arch Toxicol41,249-277, Bolt, H. M., Laib, R. J., and FlLSER, J. G. (1982). Re active metabolites and carcinogenicity of halogenated ethylenes. Biochem. Pharmacol. 31,1-4. \Bonse, G., Urban, T,, Reichert, D., and Hensch- ler, D. (1975). Chemical reactivity, metabolic oxirane formation and biological reactivity ofchlorinated eth ylenes in the isolated perfused rat liver preparation. Biochem. Pharmacol. 24, 1829-1834, Bruner, R. H. (1984). Pathological findings in labora tory animals exposed to hydrocarbon fuels of military interest. In Renal Effects ofPetroleum Hydrocarbons (M. A. Mchlman, G. P. Hemstreet, J. J. Thorpe, and N. K.. Weaver, Eds.), pp. 133-140. Princeton Scien tific, Princeton, NJ. Charbonneau, M., Strasser, J,, Lock, E. A., Turner, M. J., AND Swenberg, J. A. (1988). 1,4Dichlorobenzene induced nephrotoxicity: Similarity with unleaded gasoline induced effects. In Nephrotox icity: Extrapolationfrom in Vitro to in Vivo andfrom Animal to Man (P. A. Bach and E. A. Lock, Eds.). Ple num Press, New York. Daniel, J. W. (1963). The metabolism of 34Cl-labelled trichloroethylene and tetrachloroethylene in the rat Biochem, Pharmacol. 12,705-802. Dekant, W., Haug, R., and Henschler, D. (1985). Absorption, elimination and metabolism of tetrachlo roethylene. Arch. Pharmacol. 329, Suppl. R24. Dekant, W., Metzler, M., and Henschler, D. (1986). Identification of S-1,2,2-trichlorovinyl-N-acetylcysteine as a urinary metabolite of tetrachloroethy lene: Bioactivation through glutathione conjugation as XL 0354931 ay. O-linked micro41, 249-277. G. (1982). Reofhalogenated 4. AND HENSCH.'tabolicoxirane orinatedethfpreparation. lings in laborajels of military Hydrocarbons J. Thorpe, and inceton Scien- -OCK, E. A., V. (1988). 1,4ity: Similarity In Nephrotoxt'ivo andfrom jck, Eds.). Ple- >f "Cl-labelled :ne in the rat. -:r, D. (1985). n oftetrachloI. R24, NSCHLER, D. ovinyl-W-acerachloroethy:onjugation as PERCHLOROETHYLENE-INDUCED KIDNEY TUMORS 89 possible explanation of its ncphrocarcinogcnicity. J. Biochem. Toxicol. 1,57-72. Doe, J. E., and Tinston, D. J. (1981). Novel chambers for long term inhalation studies. In Proceedings ofthe Inhalation and Technology Symposium, Kalamazoo. Michigan, 1980(B. J. K. Leong, Ed.). Ann Arbor Sci ence, Ann Arbor, MI. Goldworthy, T. L., Lyght, O., Burnett, V. L., and Popp, J. A. (1988). Potential role of <i-2u-globu!in. protein droplet accumulation and cell replication in the renal carcinogenicity of rats exposed to trichloro ethylene, petchloroethylene and pentachloroethane. Toxicol. Appl. Pharmacol. 96,367-379. Goldworthy, T. L., and Popp, J. A. (1987). Chlori nated hydrocarbon-induced peroxisomal enzyme ac tivity in relation to species and organ carcinogenicity. Toxicol. Appl. Pharmacol., 225-233. Green, T., and Odum, J. (1985). Structure/activity studies of the nephrotoxic and mutagenic action of cysteine conjugates of chloro and fluoroalkenes. Chem.-Biol. Interact. 54, 15-31. , Green, T., Odum, J., Foster, J. R., and Hext, P. M. (1986). Perchloroethylene induced hepatic peroxi some proliferation in mice and renal hyaline droplet formation in rats. Toxicologist 6(1), 1262. Greim, H., Bonse, G., Radwan, Z., Reichert, D., and Henschler, D. (1975). Mutagenicity in vitro and potential carcinogenicity of chlorinated ethylenes as a function of metabolic oxirane formation. Bio chem. Pharmacol. 24,2013-2017. Henschler, D. (1977). Metabolism and mutagenicity of halogenated olefins. A comparison of structure and activity. Environ. Health Perspect. 21,61-64. Kanerva, R. L., Ridder, G. M., Lefever, F. R., and ALDEN, C. L. (1987). Comparison of short term renal effects due to oral administration ofdecalin or d-limo- nene in young adult Fischer-344 rats. Food Chem. Toxicol. 25,345-353. Lowry, O. H., Rosebrough, N. J., Farr, a. L., and Randall, R. J. (1951). Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193,265-275. Moore, R. B., and Green, T. (1988). The synthesis of nephrotoxic conjugates of glutathione and cysteine. Toxicol. Environ. Chem. 17, 153-162. Nash, J. A., King, L. J., Lock, E. A., and Green, T. (1984). The metabolism and disposition of hexa- chloro- 1,3-butadiene in the rat and its relevance to nephrotoxicity. Toxicol. Appl. Pharmacol. 73, 124137. National Cancer Institute (NCI). (1977). Bioassay ofTet- rachloroethylenefor Possible Carcinogenicity. Techni cal Report Series No. 13. U.S. Department of Health, Education and Welfare. Publication No. (NIH) 77-813 PB 272940. National Toxicology Program (NTP). (1985). Technical Report on the Toxicology and Carcinogenesis Studies ofTctrachloroethvlene (Perchloroethylene) in F344/N Rats and B6C3F, Mice (Inhalation Studies). DHSSNIH Report 85-2657. Odum, J., and Green, T. (1987). Perchloroethylene metabolism by the glutathione conjugation pathway. Toxicologist7, 1077. Odum, J., Green. T., Foster, J. R., and Hext, P. M. (1988). The role of trichloroacetic acid and peroxi some proliferation in the differences in carcinogenicity of perchloroethylene in the mouse and rat. Toxicol. Appl. Pharmacol. 92, 103-112. Rampy, L. W,, Quast, J. F,, Leong, B. K. J., and Gehring, P. J. (1978). Results oflong-term inhalation studies on rats of I, I, l-trichloroethane and perchloro ethylene formulations. In Proceedings ofthe First In ternational Congress on Toxicology (G. L. Plaa and W. A. M. Duncan, Eds.), p. 562. Academic Press, New York. Reed, D. J., Batson, J. R., Beatty, P. W,, Brodie, A. E., Ellis, W. w,, and Potter, D, W. (1980). High performance liquid chromatography analysis of nano mole levels of glutathione, glutathione disulfide and related thiols and disulfides. Anal. Biochem. 106, 5562. Reitz, R. H. (l987),Role ofcytotoxicity in the carcino genic process. In Nongenotoxic Mechanism in Carci nogenesis (B. E. Butterworth and T. J. Slaga, Eds.), Banbury Report 25. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Stevens, J,, and Jakoby, W. B. (1983). Cysteine conju gate 0-iyase. Mol. Pharmacol. 23,761 -765. Stonard, M. D., Gore, C. W., Oliver, G. J. A., and Smith, 1. K. (1987). Urinary enzymes and protein pat terns as indicators of injury to different regions of the kidney. Fundam. Appl. Toxicol. 9,339. Strasser, J. R,, Charbonneau, M., Borghoff, S. J., Turner, M. J., and Swenberg, J. A. (1988). Renal protein droplet formation in male Fischer 344 rats af ter isophorone treatment. Toxicologist 8,136. Trump, B. F., Upsky, M. M., Jones, T. W,, Heatfield, B. M., Higginson, J., Endicott, K., and Hess, H. B. (1984). An evaluation of the significance of experimental hydrocarbon toxicity to man. In Re nal Effects ofPetroleum Hydrocarbons (M. A, Mehlman, G. P. Hemstreet, J. J. Thorpe, and N, K. Weaver, Eds.), pp. 273-288. Princeton Scientific, Princeton, NJ. YLLNER, S. (1961). Urinary metabolites of l4C-tetrachloroethylene in mice. Nature (London) 191,820.