Chem. Res. Toxicol. 1993,6, 711-717
711
Role of Metallothionein in Zinc(11) and Chromium(111) Mediated Tolerance to Carbon Tetrachloride Hepatotoxicity: Evidence against a Trichloromethyl Radical-ScavengingMechanism Phillip M. Hanna,* Maria B. Kadiiska,l Sandra J. Jordan, and Ronald P. Mason Laboratory of Molecular Biophysics, National Institute of Environmental Health Sciences, National Institutes of Health, P.O.Box 12233, Research Triangle Park, North Carolina 27709 Received June I, 199P
The *CC13radical generated during the metabolism of CCl, is readily spin trapped in vivo and in vitro by phenyl N-tert-butylnitrone (PBN) to form the stable PBN/*CCl3 radical adduct, which can then be extracted into organic solvents and detected by ESR spectroscopy. We have used this technique to examine the proposed protective roles of Zn(II), Cr(III),and metallothionein (MT) against carbon tetrachloride toxicity in vivo. Hepatic MT, which is induced by Zn(II), has been proposed to protect against CCLinduced cellular damage by scavenging the free radical metabolites formed. CCLinduced hepatotoxicity was significantly suppressed in male Sprague-Dawley rats pretreated with a single dose of 5 mg/kg Zn(I1) or Cr(II1) according to standard serum assays for liver-specific enzymes, and hepatic M T was elevated after pretreatment with either Zn(I1) or Cr(II1). In vitro, no difference was detected in either the amount of CCLderived free radical metabolites formed or the rate a t which they were formed by microsomes from rats pretreated 24 h in advance with 5 mg/kg Zn(I1) or Cr(II1). Extraction of rat liver with 2:l chloroform/methanol 1h after the administration of a 0.8 mL/kg intraperitoneal or intragastric dose of CCl, also revealed no difference in the amount of trichloromethyl radical spin trapped in vivo following pretreatment with either Zn(I1) or Cr(II1). These results suggest that pretreatment with either Zn(I1) or Cr(II1) does not affect CC4 metabolism nor does the M T significantly scavenge the trichloromethyl free radical metabolite.
Introduction The metabolism of carbon tetrachloride has been studied in some detail as a model for the hepatic metabolism of haloalkanes. It is now well established that carbon tetrachloride is metabolized by the cytochrome P450 mixed-function oxidase system in the smooth endoplasmic reticulum to the trichloromethyl ('CC13) free radical ( I , 2),particularly by the cytochrome P4502~1 isozyme2(3,4). The *CC13radical metabolite may then react with other cellular components directly or may combine with 0 2 to form the trichloromethylperoxyl ('OOCCls) radical (5). These radical metabolites may cause damage by direct covalent binding to subcellular components or by initiation of lipid peroxidation. The initial free radical reactions also trigger a series of pathobiochemicalresponses within the cell which may themselves be responsible for eventual cell death (6-9). Several studies have shown that the pretreatment of rats with Zn(I1) protects against carbon tetrachloride toxicity (10-1 3). Among the various criteria used to assess the extent of CC&-inducedliver damage were elevations in hepatic malonaldehyde, an index of lipid peroxidation, and in serum alanine aminotransferase and aspartate aminotransferase activities. Recently, a single intraperitoneal dose of Cr(III), but not Cr(VI), Cu(II), or Zn(II), was reported to protect both rats and mice against a lethal *Abstract published in Advance ACS Abstracts, September 1,1993. 1 Permanent address: Institute of Physiology, Bulgarian Academy of Sciences,"AcademicianGeorgy Bonchev"Street, Building 23,1113Sofia, Bulgaria. * Also known as cytochrome P460j in rats.
intragastric dose of carbon tetrachloride (14). This latter report was based on results where 4 out of 10of the Cr(II1)treated rats survived 2 weeks after carbon tetrachloride administration, whereas none survived in any of the other metal-treated or control groups. Hepatic MT,3which is significantlyelevated upon Zn(I1) administration, has been proposed to protect against cellular damage by scavenging the CCLderived free radical metabolites (12,13). Even induction of hepatic MT as an inflammatory response to an intramuscular injection of turpentine in rats has been shown to coincide with reduced hepatotoxicity from carbon tetrachloride (15). About 30% (or 20) of the 61 amino acids which make up the MT sequence are cysteine (16, 17). This high sulfhydryl content enables MT to efficiently scavenge oxy-radicals in vitro ( 1 4 1 9 ) . An alternative protective mechanism of MT may be its ability to release Zn(I1) for binding at sites on membrane surfaces, displacing adventitious iron and thereby inhibiting lipid peroxidation (20). In one report, however, it is unclear why Zn(I1) pretreatment had little protective effect against a lethal dose of CCb compared to Cr(II1) pretreatment (14). The transient *CC13radical generated during the metabolism of CCl, is readily spin trapped in vitro and in vivo by PBN to form the stable PBN/*CCl3radical adduct (21-23). Once trapped in vivo, the radical adduct can be extracted into organic solvents and examined by ESR spectroscopy (24). We have used these techniques to 3 Abbreviations used MT,metallothionein; PBN, phenyl N-tertbutylnitrone;ig, intragastric;ALT,alanine aminotxansferaeqLDH,lactate dehydrogenase; SDH,sorbitol dehydrogenase.
This article not subject to U.S.Copyright. Published 1993 by the American Chemical Society
712
Chem. Res. Toxicol., Vol. 6, No. 5, 1993
examine the proposed protective roles of Zn(II), Cr(III), and MT against carbon tetrachloride toxicity in vivo, particularly the ability of MT to scavenge the trichloromethyl metabolite. The results suggest that metallothionein does not play a significant role in scavenging the trichloromethyl radical.
Materials and Methods Materials. Carbon tetrachloride (HPLC grade) and cadmium chloride (99.99 % ) were purchased from Aldrich (Milwaukee,WI). PBN, NADPH (tetrasodium, reduced form, type X), and bovine hemoglobin were purchased from Sigma (St. Louis, MO). [WICarbon tetrachloride (99% minimum 13C) was purchased fromIsotec, Inc. (Miamisburg,OH), and [lWCd]cadmiumchloride (100 mCi) was purchased from Amersham (Arlington Heights, IL). Phosphate buffer, pH 7.4, prepared from a combination of the mono- and disodium salts, was incubatedwithpreequilibrated Chelex 100resin (Bio-Rad,Richmond, CA) to remove trace heavy metal contaminants. All other chemicals were used without further purification. In Vivo Experiments. Male 400 40 g Sprague-Dawley rats (Charles River Breeding Laboratories, Raleigh, NC) were used for all in vivo experiments and were raised on a standard chow mix (NIH Open Formula; Zeigler Brothers, Gardner, PA) with free access to both food and water. Food, but not water, was removed 24 h prior to experiments where CCL was administered as an intragastric (ig) injection. At the specified time prior to sample collection,rats were given an ip injection of either 5 mg of Zn(I1) (as the sulfate dissolved in deionized water)/kg of rat body weight, 5 mg of Cr(II1) (as the nitrate also in deionized water)/kg of rat body weight, or an equivalent volume (1mL/kg) of 0.85% saline. CC4 (0.80 mL/kg of rat body weight‘ ) was administered either ip as a 1:2 CCL/ olive oil solution before anesthesia or ig neat after anesthesia. Controls received either pure olive oil ip or pure water ig. For in vivo spin trapping experiments, an ip dose of 70 mg/kg (as a 28 mg/mL solution) or 7 mg/kg (as a 7 mg/mL solution) of PBN was also administered 1 h prior to sample collection. For experiments where CC4 was administered ip, rats were anesthetized with ether just prior to sample collection at 1, 8, and 24 h post-CCL injection. Blood was withdrawn from the aorta abdominalis for serum analyses, and the liver was removed, rinsed, and frozen immediately on dry ice. For experiments where CC4was administered ig, the rats were first anesthetized with a 50-75 mg/kg ip injection of Nembutal (Abbott Laboratories), which was generallysufficient to maintain anesthesia throughout the experiment, then given an ig injection of either CCL or water. One hour post-CC4 injection, blood was withdrawn for serum analyses and the livers were removed and either used immediately to prepare microsomes for cytochrome P450 analyses and in vitro ESR experiments or frozen on dry ice for further assays and extraction experiments. The rats were killed by exsanguination. Microsomal Experiments. Liver microsomes were prepared as previously described (26). Briefly, livers were weighed, rinsed, and homogenized in ice-cold 0.1 M sodium phosphate, pH 7.4. The homogenate was centrifuged a t loooOg for 20 min, and microsomes were separated from the resulting supernatant by centrifugation at 165000g for 35 min. The microsomal pellet was washed once by resuspension in ice-cold 0.1 M sodium phosphate, pH 7.4, and recentrifuged at 165000g. Following the final suspension in 0.1 M sodium phosphate, pH 7.4, microsomal protein concentrations were determined using the method of Lowry et al. (27). Cytochrome P450 content was measured according to the method of Omura and Sato (28). Extraction of Liver. Liver tissue was extracted similar to the method of Folch et al. (29). Liver tissue (3-5 g) was
*
4 Or 1275 mg/kg.The oral LDm for carbon tetrachloride is 2820 mg/kg for rata and 1500 mg/kg ip (25).
Hanna et al. homogenized directly into 5 volumes of an ice-cold 2:l chloroform/ methanol mixture. The homogenate was then washed with 0.2 mL of 0.85% saline/mL of homogenate and centrifuged for 15 min to separate the aqueous and organic phases. The organic phase was collected, dried with anhydrous sodium sulfate, and concentrated for ESR measurements by blowing a steady stream of Nz over the solution. The final volumes were scaled to 0.25 or 1 mL/g of homogenized liver tissue from rats given 7 or 70 mg/kg PBN, respectively. This solution (0.8-1.0 mL) was transferred to a quartz ESR flat cell (Wilmad, Buena, NJ) and bubbled very gently with Nz for 2 min to remove 02 before the ESR spectrum was recorded. Biochemical Assays. Hepatic MT levels were measured as the Cd-binding capacity of liver homogenate according to the procedure of Onosaka et al. (30) as described by Eaton and Toal (31). Cd binding was measured by the activity of the y-emitter “Cd. Serum activities of lactate dehydrogenase (LDH), alanine aminotransferase (ALT),and sorbitol dehydrogenase (SDH) and the total serum concentration of bile acids were measured according to published procedures (32-35). Analysis of variance (ANOVA) statistical methods were applied to the results of the biochemical assays. Two-way factorial ANOVAs were used in the zinc studies to assess the effects of experimental treatments and duration of treatment, and their interaction. One-way ANOVAs were used in the chromium studies. The natural logarithm was used as a variance stabilizer prior to analysis. Following the application of the ANOVA procedure, pairwise comparisons were made using the RyanEinot-Gabriel-Welsch multiple range test (36). Instrumentation. ESR spectra for the microsomal experimenta were recorded on aBruker ER200D spectrometer operating a t a microwave frequency of 9.77 GHz and a microwave power of 20 mW. ESR spectra for the organic extraction experiments were recorded on a Varian E-109 spectrometer operating a t a microwave frequency of 9.33 GHz and a microwave power of 20 mW. Spectra were recorded using a quartz flat cell (Wilmad) in a TMllo microwave cavity. Other instrumental parameters are given in the figure legends. Spectra were recorded on an IBMcompatible computer interfaced with the spectrometers. A Shimadzu UV 3000 spectrophotometer was used for the cytochrome P450 assay of Omura and Sat0 (28). All other spectrophotometric data were obtained using a Beckman DU-7 spectrophotometer. The lWCdbinding to MT was determined using a Beckman Gamma 5500 radiation counter with a DP 5500 data reduction system.
Results Effect of Zn(I1) and Cr(II1) Treatment on CCld Hepatotoxicity. The hepatotoxicity of CCl, was assessed by the serum activities of LDH, ALT, and SDH. Only 1 h after an ip injection of 0.8 mL/kg CC4, the activities of all three hepatic enzymes were elevated in the serum (Table I). Their serum activities remained highly elevated 24 h after the CCl, injection, indicating significant damage to the liver. Serum activity of LDH was also significantly elevated within the first hour following the ig administration of 0.8 mL/kg CCk, similar to the results following ip administration, but the SDH activity remained unchanged (Table I). Pretreatment by a single ip dose of 5 mg/kg Zn(I1) 24 h prior to the injection of CCl, imparted significant protection against initial CCL-inducedliver damage (Table I). Nevertheless, 24 h after the CC4 injection, serum activities of all three hepatic enzymes were significantly elevated over control levels. Pretreatment 24 h beforehand with an ip injection of 5 mg/kg Cr(II1) instead of Zn(I1) also significantly protected against hepatic damage within
Chem. Res. Toxicol., Vol. 6, No. 5, 1993 713
Role of Metallothionein during CCld Hepatotoxicity
Table I. Effect of a Single Dose of Zn(I1) o r Cr(II1) on CCh Hepatotoxicity (&SEI. ~
treatment controlb
ccv ccv
cc4c
Zn(II)/CC&C Zn(II)/CCLc Zn(II)/CC&c CC4d Cr(III)/CC4d
time (h) 1
8 24 1 8 24 1 1
LDH (IU/L) 178 f 36 794 f 174O 1468 f 430' 960 f 146O 303 f 34 427 f 132f 333 f 27f 1018 f 277' 485 f 44'
ALT (IU/L) 36.8 f 1.5 58.0 f 4.9 176.3 3O.le 211.8 f 96.7* 46.3 f 7.3 64.5 f 15.3 377.3 f 156.3'
total bile acids &mol/L) 26.4 f 8.3 18.4 f 4.9 119.2 f 29.5' 102.3 f 30.1' 16.4 f 1.6 53.6 f 3.6' 166.8 f 24.4' 23.3 f 1.4 22.0 f 1.2
SDH (IU/L) 11.8 f 0.9 35.0 f 3.P 322.5 f 94.3' 239.3 i 95.8O 31.8 f 7.8' 72.0 f 13.8'J 233.3 f 11.7O 17.0 f 6.2 16.3 f 3.2
The Zn(I1) or Cr(II1) was administered ip 24 h prior to CCL administration. Serum was collected for assays at the times indicated following CC4 administration. Rata were anesthetized with ether (N = 4). 0.8 mL/kg CC4 administered ip. Rata were anesthetized with ether (N = 4). d 0.8 mL/kg CC4 administered ig. Rata were anesthetized with Nembutal ( N = 4). e Values are significantly different (p < 0.05) from the control. f Values are significantly different (p < 0.05) from corresponding time poi& of CCL-treated animals. Table 11. Effect of a Single Dose of Zn(1I) on CCld Hepatotoxicity (ME). time ALT SDH total bile treatmentb (h) (IU/L) (IU/L) acids (fimol/L) 13.0 f 1.6 33.6 f 6.5 control 38.0 f 3.7 21.8 f 3.6' 30.9 f 1.7 Zn(I1)C 1 39.4 f 2.6' 50.6 f 3 . g CC4' 1 104.4 f 6 . g 80.6 f 7.6' 33.7 f 4.9 Zn(II)/CC4d 1 65.4 f 11.6' 43.0 f 8.N 4 The Zn(I1) was administered ip simultaneously with CCL, and the serum was collected 1 h later for assays. Rata were anesthetized with ether (N= 5). 5 mg/kg Zn(I1) or 0.8 mL/kg CC4 administered ip. d 5 mg/kg Zn(I1) and 0.8 mL/kg CC4 administered ip simultaneously. e Values are significantly different (p < 0.05) from the CC4treated and Zn(II)/CC4-treated groups. f Values are significantly different (p < 0.05) from all groups.
the first hour of CC14 treatment, similar to the effect of Zn(I1) pretreatment (Table I). In all cases, the total serum concentration of bile acids remained unchanged within the first hour following CC14 administration but were elevated after 24 h. Elevated levels of bile acids in serum is a clinical indication of cholestasis. Zn(I1) pretreatment caused no significant protection against elevated bile acid concentrations in the serum after the first hour (Table I). When Zn(I1) and CC4 were administered ip simultaneously, CCLinduced hepatotoxicity was also diminished relative to treatment with CCl4 alone (Table 11). In this case, the Zn(I1) was injected as a 20 mg/mL solution to minimize the volume of fluid entering the ip cavity. The results in Table I1 suggest that the protective mechanism of Zn(I1) does not involve the induction of protein synthesis. Effect of Zn(II), Cr(III), and CCl4 on Hepatic Metallothionein Concentration. Hepatic concentrations of MT were dramatically elevated following the ip administration of either Zn(II),Cr(III),or CC4. As shown in Figure 1,24 h after the ip injection of 5 mg/kg Zn(II), hepatic MT was induced 1700% over saline-treated controls according to the Wd-binding assay. After 48 h, however, MT concentrations once more approached control levels. Similarly, Cr(II1) administration ip induced MT concentrations as much as 640% after 24 h (Figure 2). CC14 administration ip also induced hepatic MT by 575% after 24 h (Figure 11, which may be due to an inflammatory response (37). In contrast, administration of CC4 ig did not cause a significant elevation of hepatic MT concentrations (data not shown). Effect of Zn(I1) and Cr(II1) Treatment on CC4 Metabolism. Carbon tetrachloride is metabolized to the trichloromethyl radical by the cytochrome P450 enzymatic system (22). A suppression of this CC4-activating enzymatic system may account for the protective effects of
~
W 500-
l\ivZn(
2
I
ac
400-
L)
11)-Treated
U
0 300-
Zn( Il)/CCl,-Treated
0
E
200-
8
I-
100-
2 0
5
10
15
20
25
Time (h) Figure 1. Effect of CC4 administration on hepatic M T concentrations. CC4 (0.8 mL/kg, ip) was administered at 0 h. Zn(I1) (5mg/kg, ip) was administered 24 h before the CC4. The M T concentrations (hSE,N = 4) at the times indicated were determined by the modified "Wd/hemoglobin method described by Eaton and Toal(31). Asterisk indicates significant difference @ < 0.05) from the control at 0 h.
Zn(I1) or Cr(II1) against CCl, hepatotoxicity. It has been shown previously, however, that a single dose of Zn(I1) does not suppress cytochrome P450 in rats (12,38,39).In accord with these reports, no change in the hepatic cytochrome P450 concentration was observed 25 h following treatment with 5 mg/kg Zn(I1) (Table 111). Treatment with 5 mg/kg Cr(II1) also had no significant effect on hepatic cytochrome P450 concentration (Table IV) (14). The lack of a significant effect in vivo on CCl, metabolism by Zn(I1) or Cr(II1)pretreatment was also supported by in vitro spin trapping experiments using NADPHsupplemented microsomes prepared from Zn(I1)- and Cr(II1)-treated rats. The ESR spectrum of a 30-min incubation of NADPH, PBN, CC4, and microsomes from nontreated rats in 0.1 M sodium phosphate, pH 7.4, under a stream of Nz is shown in Figure 3A. The hyperfine coupling constants obtained were aN = 14.02 G and agH = 1.60 G, with an additional coupling of u p c = 9.64 G when 13CC4was used (Figure 3B). Both parts B and C of Figure 3 demonstrate that the predominant ESR signal was derived from a CCL-dependent carbon-centered radical metabolite as reported previously (22). The small underlying, 13C-independent signals in parts A and B of Figure 3 can be simulated using the parameters for lipidderived alkyl and alkoxy1 adducts of PBN previously reported (40). Lipid-derived radicals are presumably formed by CCL-initiated lipid peroxidation. From the result shown in Figure 3D, it is evident that after 30 min there was little difference in CCl, metabolism
714 Chem. Res. Toxicol., Vol. 6,No. 5, 1993
Hanna et al.
i Y
C
I
0
5
10
15
.
25
Table 111. Effect of Zn(II), CCl,, and Zn(II)/CCh Treatments on Hepatic Cytochrome P450 Concentration treatmentn time (h) cytochrome P450b control 100 f 22 Zn(1I)c 1 107 f 5 1
8 24 1 8 24
69 f 11 57 f 6 43 f 7e 56 f 4 52 f 7 44 f 50
CC4d Zn(II)/CC4c Zn(II)/CCu Zn(II)/CC4C Rata anesthetized with ether. b Values are reported as % of controlvalue(0.56nmol/mgmicrosomalprotein,ISE,N = 4). Zn(I1) administered (5 mg/kg ip) 24 h before recorded time or CC4 administration(0.8mL/kg, ip). 0.8 mL/kg administeredip. e Values are significantly different (p < 0.05) from the control. Table IV. Effect of Cr(III), CCl4, and Cr(III)/CCl~ Treatments on the Hepatic Cytochrome P460 Concentration treatmentn time (h) cytochrome P450b control
- CCI,
,
20
Time (h) Figure 2. Induction of hepatic metallothionein (MT)following a single ip dose of 5 mg/kg Cr(II1).The MT concentrations ( S E , N = 4) at the times indicated were determined by the modified "Wd/hemoglobin method described by Eaton and Toal (31). Asterisk indicatessignificant difference (p
