Protection against carbon tetrachloride-induced hepatotoxicity by pretreating rats with the hemisuccinate esters of tocopherol and cholesterol.

Previous studies have demonstrated that alpha-tocopheryl hemisuccinate (TS) protects hepatocyte suspensions from chemical-induced toxicity. It has been suggested that TS cytoprotection is related to unique properties of the TS molecule or is dependent on the cellular release and activity of unesterified alpha-tocopherol (T). To test the unique cytoprotective nature of TS in vivo, the protective ability of T and tocopherol esters against carbon tetrachloride (CCl4)-induced hepatotoxicity in rats was examined. Hepatoprotection [determined by serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels and histopathology] was not observed after T (or tocopheryl acetate and tocopheryl nicotinate) administration, even though this treatment resulted in a fivefold elevation in hepatic T content. Only pretreatment with TS (100 mg/kg, intraperitoneally) resulted in partial hepatoprotection against CCl4 (2.9 g/kg, orally) toxicity. These findings suggest that hepatoprotection results not from the cellular accumulation of T but rather from the intact TS molecule. To test this hypothesis, the hepatoprotective capacity of cholesteryl hemisuccinate (CS), unesterified cholesterol, and cholesteryl acetate (CA) was examined against CCl4 toxicity. As observed with the tocopherol derivatives, pretreatment with unesterified cholesterol or CA demonstrated no protective ability. However, when rats were pretreated with CS (100 mg/kg), the hepatotoxic effects of CCl4 (elevated serum AST and ALT levels and centrilobular necrosis) were completely prevented. The prevention of CCl4-induced hepatotoxicity by CS and TS do not appear to result from an alteration in hepatic drug metabolism. These data clearly demonstrate that CS and TS are unique and powerful cytoprotective agents against CCl4 hepatotoxicity in vivo.(ABSTRACT TRUNCATED AT 250 WORDS)

. From these studies, we have con- Attempting to exclude this possibility, we investigated the protective effect of in vivo TS administration (and other tocopherol analogs) on carbon tetrachloride-induced toxicity in rats. Furthermore, to assess the importance of the tocopherol molecule in cytoprotection, we also investigated the protective properties of another lipophilic succinate derivative, cholesteryl hemisuccinate (CS).
The method of Janoff et al. (13) was modified to prepare the tris salt of TS. Trizma base (7.5 mM or 913 mg) was dissolved in 25 ml of hot methanol. In a separate flask, TS (7.5 mM or 4.0 g) was dissolved in 25 ml of diethyl ether. Each solution was cooled to room temperature, mixed, and placed on a rotoevaporator under negative pressure. Once solvents were removed, we washed the resulting gelatinous white solid in diethyl ether (4 ml), stirred it, and allowed it to sit overnight. The ether was removed with a rotoevaporator and the resulting white solid dried in a desiccator containing calcium chloride. The melting point of the final product (4.5 g, 91% yield) was 88-910C, and the purity was greater than 98+% as determined by acid/base titration of the amount of trizma base and TS, free acid present in the final product.
A * e Iee -My. ea.
For the experiments described in Tables 1-3, dosing solutions of T, TA, and TS free acid were prepared by dissolving the compounds in ethanol (2 parts) followed by peanut oil (3 parts) at a concentration of 100 mg/ml. In the experiments described in Tables 4 and 5, dosing solutions of cholesterol, CA, TS, and CS, tris  salt were prepared at 25 mg/ml by dissolving the compounds in ethanol (1 part) followed by 3 parts olive oil (for cholesterol and CA) or polyethylene glycol 400 (for TS and CS). In the experiments described in Tables 6 and 7, dosing solutions (25   mg/kg) of TN, TS free acid and cholesterol were dissolved in olive oil (25 mg/ml), whereas the tris salts of TS and CS were fine suspensions (sonicated) of 25 mg/ml saline. The dosing solution of CC14 was dissolved in peanut oil (1 g/ml), and metyrapone was dissolved in corn oil (50 mg/ml). We prepared all dosing solutions immediately before administration.
Male Sprague-Dawley rats (175-275 g), given food and water ad libitum, were used for these experiments. Experimental animals were given an intraperitoneal (ip) injection or oral gavage of tocopherol analog (100 mg/kg body weight), cholesterol analog (100 mg/kg), or vehicle and fasted (as noted, several experiments used only fed rats). Twenty four hours later, each animal received CC14 (1.0 to 2.9 g/kg body weight) or peanut oil (vehicle) by oral gavage. One hour after CC14 administration, each animal received food and water ad libitum. We assessed hepatotoxicity 24, 48, or 72 hr after CC14 administration and monitored survival for 7 days in the acute toxicity experiments (Table 1).
Hepatotoxicity was determined by measuring serum AST and ALT levels and liver histopathology. Using ether anesthesia, we collected serum from the inferior aorta (at sacrifice) or from the tail vein at 24 or 48 hr after CC14 administration. We determined serum AST and ALT levels with a Roche Cobas Bio Clinical Analyzer using the spectrophotometric method adopted from Bergmeyer et al. (14).
For histopathology, we sacrificed experimental animals 48 or 72 hr after CC14 administration with ether anesthesia and exsanguination. The liver (or other tissues) was immediately removed, weighed, sliced into 2-to 3-mm sections, and fixed in 10% neutral buffered formalin. After 48 hr of fixation, tissues were paraffin infiltrated on a Miles Tissue-Tek VIP tissue processor. Four micron histologic sections were cut on a Reichert Histostat Rotary microtome and mounted on glass slides.
All sections were stained with Harris hematoxylin and eosin (H&E) on a Shandon Varistain 24-3. In several experiments (described in Figure 2), liver tissue was cut 8Tocopherol analogs and vehicle were administered and rats fasted for 24 hr before CCI4 treatment. Survival is noted as + and results represent three to nine rats tested per treatment. LD5 values were estimated by the up-and-down method of Dixon (15) and Bruce (16). ND, not determined. into 1-mm cubes and fixed in 2% phosphate-buffered glutaraldehyde for 3 hr. After a phosphate buffer rinse, the tissues were postfixed 1 hr in 1% osmium tetroxide, dehydrated with ethanol, and embedded in Poly/Bed epoxy resin (Polysciences Inc., Warrington, Pennsylvania). Sections were cut 1 ,um thick and stained with 1% toluidine blue. Histologic sections were examined (as described below) by a pathologist without knowledge of experimental treatment. To quantify CC14-induced hepatocellular necrosis, each histologic liver section was rated for necrosis on a scale of 0-4 or was examined for the percentage of liver necrosis as measured with a Zeiss Videoplan system. We examined multiple liver histologic cross-sections from each experimental animal and scored them as to the extent of liver necrosis using a scale of 0-4. The criteria for each score are as follows: 0 = no necrosis; 1 = necrosis limited to proximal perivenular (centrilobular) layers of hepatocytes without extending into the midzonal (mid-third between portal triad and terminal venule/central vein) region; 2 = necrosis of entire perivenular zone extending into the midzonal region, up to onehalf the span of the hepatic lobule; 3 = necrosis of more than one-half of the hepatic lobule extending to within a few cell layers of the portal triad; and 4 = complete necrosis To determine the percentage of liver necrosis (% of liver), a histologic liver section was displayed on the screen of a Zeiss (Oberkochen, Germany) Videoplan monitor using a Hitachi CCTV camera mounted on a Zeiss light microscope with a 1x objective. Using the Videoplan digitizer tablet, stylus, and the data acquisition program, the Videoplan system calculated the total area of the liver section after drawing the outline of the liver. Next, the Videoplan system calculated the percentage of liver necrosis (% of total liver showing necrosis) after drawing the outline of the necrotic areas in the liver section. For each experimental liver, we examined two liver sections; each section was measured twice, and a mean value was calculated from these four determinations. A tissue control and a constructed square control were examined before analyzing experimental samples. Using a square with sections of 20, 40, 60, 80, and 100% drawn on graph paper as a template, the accuracy of the Videoplan in determining the area of each section (expressed as a % of the total area) with a 2% error was confirmed daily. Using a liver section control (from a CC14-treated rat), the accuracy and reproducibility of the Videoplan and its operator were confirmed daily. We analyzed this control liver section for percent liver necrosis on five separate occasions with a mean and SE of 52.7 + 2.2%. Before analyzing experimental samples, we determined the percent necrosis in the liver control section within 2 SEs of the calculated mean value (48-57%).
To determine the effect of tocopherol analog administration on the acute toxicity of CCI4, we estimated the LD 0 by a modified up-and-down method o Dixon (15) and Bruce (16). In brief, the dose of CC14 administered was adjusted by a factor of 1.3 until successive lethal and nonlethal doses were found. The true LD value for each treatment is assumed to f5?l between these successive CC14 doses with opposite effects. The initial dose was 2.9 g CCl4/kg, and survival was monitored for at least 7 days after CC14 treatment to ensure recovery.
Tissue concentrations of T, TA, and TS were determined by the HPLC method of Fariss et al. (1). Liver, brain, kidney, heart, and lung were prepared for analysis by a freeze-clamp technique (immediately frozen with liquid N2, pulverized to a powder with a mortar and pestle, and stored at -800C). For analysis, 25-50 mg of powdered tissue was weighed (or 100 ,ul of plasma or blood added) in a microcentrifuge tube and treated as described for Volume 101, Number 6, November 1993 viable hepatocytes in the previous method (1). We determined the tissue concentration of T from the initial tissue extraction, and the T ester concentration (TS or TA) was determined by measuring the amount of T released from base hydrolysis of the first extract (1). A standard curve for each tocopherol analog was analyzed with each set of experimental samples. The limit of detection for T, TA, and TS by this method is 10 nmol/g tissue and 5 nmol/ml plasma or blood.
We investigated the effect of TS and CS on microsomal drug metabolism using the pentobarbitone (PB) sleeping-time test (17. Male Sprague Dawley rats (160-200 g) were given an intraperitoneal injection of vehicle (olive oil); TS in olive oil (100mg/kg); cholesterol in olive oil (64 mg/kg) plus tris succinate in saline (59 mg/kg); or CS, tris salt in saline (100 mg/kg) plus olive oil (same dose as C treatment) and fasted for 24 hr. Next, PB (50 mg/kg, intraperitoneally) was administered to each rat, and the sleeping time was recorded. The absence of the animal's righting reflex was used to indicate sleep. At the start of sleeping time, we placed the animal in the supine position, and the sleeping time ended once the animal assumed a prone position. As a control, an inhibitor of cytochrome P-450, metyrapone (100 mg/kg) or its vehicle (corn oil) was administered intraperitoneally to rats 30 min before PB (50 mg/kg), and the sleeping time was determined. The pretreatment regimen of each animal was unknown to the investigator during the sleeping time determination.
Data were analyzed for significance (p<0.05) using single-factor analysis of variance (ANOVA) and Scheffe's multiple comparison test (StatView II for Macintosh, version 1.04, Abacus Concepts Inc., Berkeley, California). The Student's t test (unpaired) was used to determine the significant difference (p<0.05) between the two treatments in each PB sleeping time experiment.

Results
The estimated LD50 determined for CC14 was 2.5 g/kg in rats receiving vehicle pretreatment (Table 1). Rats pretreated with a single dose of T or TA (100 mg/kg) before an oral gavage of CC14 were not afforded protection against the lethal effects of this chlorohydrocarbon. In fact, the administration of unesterified T potentiated CC14 acute toxicity (estimated LD50 of 1.5 g/kg). In contrast, rats pretreated with TS (100 mg/kg) were protected against CCI4-mediated toxicity as demonstrated by a 76% increase in the estimated LD50 to 4.4 g/kg (Table 1). Interestingly, when TS was administered orally, no pro-  Values are the means ± SEM (n = 3-4). ND, not detected; nd, not determined. *p < 0.001, as compared to other tocopherol analog treatments. tection was observed. Histological examination of rat tissues (liver, lung, kidney, and heart), obtained 48 hr after a lethal dose of CC14, demonstrated extensive periportal and midzonal liver necrosis, but histopathology was unremarkable for the other tissues examined. Liver histology was normal in TS-pretreated rats that survived 7 days after CC14 treatment.
The protective effect of TS administration against CC14-induced hepatotoxicity is shown in Table 2 and Figure 1. Each rat received an approximate LD50 dose of CC14 (2.9 g CCl4/kg for all treatments except 1.7 g/kg for T-pretreated rats). Rats pretreated with vehicle demonstrated a massive increase in liver-associated serum enzymes AST (15,600 U/1) and ALT (2,820 U/1), 48 hr after CC14 administration. In addition, approximately 58% of the liver was found to be necrotic in each animal from this treatment group (Table  2, Fig. IB). Pretreatment of rats with T or TA did not significantly alter the hepatotoxic effect of CC14 treatment as indicated by serum enzyme levels, percent liver necrosis, and histopathology (  IC). These findings are in agreement with the acute toxicity data in that comparable lethal doses of CC14 given to TA-and Tpretreated rats resulted in similar hepatotoxic damage. In contrast, rats pretreated with TS were significantly protected from CC14 hepatotoxicity, with serum AST and Environmental Health Perspectives ,~~~~~--*g.sll 9X,E *e. ALT levels of 3430 and 630 U/1, respectively (77% reduction as compared to CC14 only treatment). Administration of the cytoprotective agent TS also significantly reduced CCl4-mediated hepatic necrosis by 50% from 58% to 29% (Table  2) and appeared to limit the extent of damage (ballooning degeneration) to periportal and proximal midzonal hepatocytes ( Fig.   1 D). Hepatic T levels in TS-treated rats and fat T levels in T-treated rats were significantly elevated as compared to the other treatment groups 48 hr after CC14 treatment. Tocopherol esters were not detected in the liver 48 hr after CC14 treatment.
The tissue (liver, brain, kidney, heart, lung, whole blood, and plasma) concentration of T and tocopherol esters (TA and TS) were measured 24 hr after tocopherol an'alog administration (no CC14), and the results are shown in Table 3. This time course (24 hr after) was used to determine the tissue (liver) concentration of T, TA, and TS at the time of CC14 exposure. Rats pretreated with T for 24 hr demonstrated a fivefold increase in the amount of unesterified T in the liver (128 nmol/g) as compared to vehide-treated rats (28 nmol/kg). However, this T pretreatment did not result in protection. A twofold increase in liver T concentration was observed with TA and TS pretreatments, but only the administration of TS resulted in protection against CC14 toxicity and hepatic TS accumulation (Table 3). Thus, TS-mediated protection appears to depend on the hepatocellular accumulation of TS. This conclusion was confirmed by our finding that the oral administration of TS eliminates both protection and liver and tissue accumulation of TS (Table 3). Furthermore, the tissue distribution results in Table 3 demonstrate that 24 hr after a single dose of TS (100 mg/kg), a significant (p<0.05 versus vehicle control) accumulation of TS was found in all tissues examined, except the brain. However, 48 hr after TS administration, tissue TS levels were not detectable or were not significantly different from vehicle control.
To investigate the importance of the tocopherol molecule in cytoprotection, we also examined the protective properties of another lipophilic hemisuccinate derivative, CS, as well as unesterified cholesterol and CA. Data from these studies ( Table 4) clearly demonstrate that CS tris salt is a powerful protective agent against the hepatotoxic effects of a sublethal dose of CC14 (1 g/kg). In fact, CS pretreatment completely protected rats from hepatotoxicity at 24, 48, and 72 hr after CC14 exposure as measured by serum ALT levels and hepatic necrosis. However, cholesterol or CA pretreatment did not significantly protect rats from the hepatotoxic effects of CC14 administration (Table 4). A histological examination of the hepatic terminal venules (Fig. 2) revealed that treatment with 1 g CCI4/kg results in perivenular necrosis and collapse (necrosis score 1) with large globule fatty degeneration of midzonal hepatocytes (fat globules turn black with osmium postfixation; Fig. 2B). In contrast, CS pretreatment prevented CCl4-induced perivenular necrosis (necrosis score 0) and dramatically reduced the fatty degeneration of hepatocytes (Fig. 2C).
To determine the influence of fasting on the cytoprotective capacity of TS and CS, rats were pretreated with TS, CS, tris salt, or cholesterol plus sodium succinate plus trizma base and fed for 24 hr before CC14 administration (2 g/kg). Data from these studies (Table 5) demonstrate that CS tris salt pretreatment completely protected fed rats from hepatotoxicity at 24, 48, and 72 hr after CC14 exposure as measured by serum ALT levels and hepatic necrosis. However, TS and cholesterol plus succinate plus trizma pretreatments did not afford rats significant hepatoprotection. Interestingly, fed rats given 2 g CCl4/kg demonstrated less hepatic damage with greater variability (ALT levels) within treatment groups than fasted rats receiving 1 g CCl4/kg (Table 4). In both fed and fasted rats, liver damage (as measured by ALT levels) was observed predominantly at 24 and 48 hr after CC14 exposure, but peak enzyme leakage varied within a treatment group between these two time points. Though serum ALT levels returned to near control levels by 72 hr after CC14 treatment, hepatic necrosis was still detectable by histopathology at this time point (Tables 4 and 5).
Because of the striking protective capacity of CS tris salt, the protective abilities of the free acid and tris salt of TS were compared with CS tris salt and TN pretreatment against a sublethal hepatotoxic dose of CC14 (1 g/kg). The results of this study, shown in Table 6, indicate that the administration of TN   0.05 ± 0.02* 0.03 ± 0.01* 0.03 ± 0.0* 0 ± 0 'Tocopherol and cholesterol analogs and vehicle were administered (100 mg/kg, ip) 24 hr before a single oral dose of CC14 (2.0 g/kg). Rats were fed before CCI4 administration. Serum ALT levels were determined at 24, 48, and 72 hr after CCI4 treatment. The liver necrosis score was determined for specimens obtained 72 hr after CCI4 treatment. Values are the means ± SEM (n = 4-5). *p < 0.01 as compared to vehicle (+CCI4).  8The pretreatment compQunds were administered and rats fasted for 24 hr before a single dose of PB (50 mg/kg, ip). Metyrapone or vehicle (corn oil) were given 30 min before PB. Values are the means ± SEM (n=3-5). not afford rats significant protection against the hepatotoxic effects of CC14. The percent protection [as determined by ALT values with negative (no CC14) and positive (+CCl4) controls assumed to be 100% and 0% protection] for TN and TS free acid were 0% and 46%, respectively. In contrast, the administration of the tris salt form of TS provided highly significant hepatoprotection (75% protection). How-ever, the tris salt of CS proved to be the most effective protective agent tested, providing virtually complete protection against CCl4-mediated hepatotoxicity.
To assess the effect of TS and CS pretreatment on hepatic drug metabolism, the influence of these cytoprotective agents on pentobarbitone-mediated sleeping time in the rat was determined ( Table 7). The pretreatment of rats with metyrapone, a known inhibitor of cytochrome P-450mediated drug metabolism, resulted in a 650% increase in PB sleeping time. However, a significant increase in PB sleeping time was not observed in rats pretreated with TS free acid (19% increase) or CS tris salt (26% increase) as compared to vehicle controls and nonprotective conditions (cholesterol and tris succinate).

Discussion
Because exposure to CC14 results in extensive cellular lipid peroxidation, numerous investigators have examined the protective role of cellular a-tocopherol, the predominant membrane-bound antioxidant and free radical scavenger, in CCl4-induced hepatotoxicity. Attempts to supplement endogenous stores of cellular T, however, have resulted in conflicting reports on the cytoprotective abilities of vitamin E administration against CC14 toxicity. The published reports range from protection (18)(19)(20) and no effect (20)(21)(22) to potentiation (20,23) of CC14 toxicity after vitamin E administration. The present study demonstrates distinct differences in the cytoprotective nature of various tocopherol analogs. As measured by survival, liverassociated serum enzyme levels, and histopathology, protection against CC14 hepatotoxicity was not observed after pretreating rats (24 hr) with a single dose (100 mg/kg) of unesterified T or the acetate and nicotinate esters of T. Thus, the supplementation of hepatic T levels (fivefold increase) in a presumably physiologically active configuration (in vivo administration and disposition) failed to protect rats from the toxic effects of CC14. In fact, we found that T administration enhanced the toxic effect of CC14, while TA and TN administration did not alter CCl4-mediated toxicity. The variable findings, reported here and in numerous other reports, on the protective abilities of vitamin E administration (18)(19)(20)(21)(22)(23), illustrate the need for additional studies on the influence of vitamin E dosing regimens on T subcellular disposition and chemical-induced toxicity. These studies are necessary to understand and have confidence in the use of T or other tocopherol analogs as a therapeutic strategy for CC14 or other chemical poisonings.
In contrast to T, TA, and TN, the administration of the hemisuccinate ester of T protected rats from the lethal and hepatotoxic effects of 2.9 g CCl4/kg body weight and resulted in a significant accumulation of hepatocellular TS. These data support the results and conclusions from our previous in vitro studies, demonstrating that TS is a unique and effective cytoprotective agent. Our finding that TS administered orally results in no hepatic accumulation of TS (presumably due to Volume 101, Number 6, November 1993 TS hydrolysis by gut esterases, as tissue T levels are elevated in the absence of tissue TS; Table 3) and no protection against CC14 toxicity also supports our hypothesis that the presence of cellular TS is essential for cytoprotection against CC14 toxicity. Interestingly, the protection observed for TS free acid in rats treated with a lethal dose of CC14 (2.9 g/kg) was not observed when the concentration of CC14 administered was reduced to a sublethal dose in fasted and fed rats. Though TS free acid pretreatment did reduce CCl4-induced hepatotoxicity by 46% (in fasted rats) when sublethal concentrations of CC14 were administered (Table 6), the degree of protection was not significant (p<0.05) as determined by ANOVA using Scheffe's multiple comparison test. Wolfgang et al. (21) have also reported the modest protective abilities of TS free acid administration against a sublethal dose of CC14. However, the administration of the tris salt of TS did result in significant hepatoprotection (p<0.01). An explanation for the apparent difference in the cytoprotective abilities of the free acid and the tris salt of TS is unknown. We speculate that differences in hydrophilicity between these two forms of TS might influence the bioavailability of TS after administration and hence explain the variability in cytoprotective ability. This hypothesis is in agreement with Carini and co-workers' (8) explanation for enhanced protection against CCl4-induced lipid peroxidation with the water-soluble form of TS, TS polyethylene glycol ester, as compared to TS free acid.
One possible explanation for the unique cytoprotective abilities of TS is that the subcellular distribution of TS (as compared to T administration) is distinctive, and thus T released from TS at this novel subcellular site is essential for protection. In support of this hypothesis, we found that TS pretreatment resulted in hepatic T levels 48 hr after CC14 treatment that were significantly higher than levels observed for nonprotective vitamin E pretreatments (even though these pretreatments increased adipose T levels by 10-fold; Table 2). These findings indicate a unique ability of TS to maintain hepatocellular T levels during a toxic insult or they may merely be the result, rather than the cause, of cytoprotection (cellular T is lost after cell death). We support the later explanation, but this question can only be answered in vitro, by separating viable from nonviable hepatocytes before analysis during a toxic CC14 insult. To argue against a novel cellular disposition for TS, our tissue distribution data, measured 24 hr after a single dose of T ester, does not show a significant difference in the disposition of TS and TA. Table 3, similar concentrations of TS and TA were found in each tissue examined. Furthermore, the release of T as an explanation for TS cytoprotection does not appear to be the critical event because the substitution of the tocopherol portion of the TS molecule with cholesterol results in a compound (CS) with remarkable protective properties against CCl4-induced hepatotoxicity. Thus, our studies clearly demonstrate that the release of tocopherol is not essential for the cytoprotection observed. Because protection was not observed after cholesterol and CA administration, and cholesterol has no reported antioxidant and free radical scavenging abilities, we again conclude that the protection observed results from the cellular accumulation and the unique properties of CS and TS. The protection against CC14 hepatotoxicity observed with CS administration was striking, with complete protection against CC14-induced necrosis and a dramatic reduction in fatty degeneration as demonstrated by light microscopy. These results indicate that CS pretreatment not only maintained cell viability during a toxic insult with CC14 but also maintained hepatic function (processing of lipid) during this insult. Like TS, the tris salt of CS is a more effective cytoprotective agent than the free acid form (personal observation). Again, we speculate that the insoluble nature of CS free acid appears to preclude its protective abilities.

As shown in
As we are the first to describe CS and TS cytoprotection in vivo, little is known about the mechanism of TS and CS protection against CCl4-mediated hepatotoxicity. One possible explanation is that TS and CS pretreatments simply delay the onset of the toxic response rather than provide true protection. This possibility is unlikely because we have shown that TS protection against the lethal effects of CC14 continues for at least 7 days with no liver histopathology observed at that time. In addition, using both histopathology and serum ALT levels, we demonstrated that CS completely protected fed and fasted rats from the hepatotoxic effects of CC14 for 72 hr (a period of time in which serum AST levels, for CCl4-treated rats without protection, have nearly returned to normal endogenous levels). It is important to point out the advantages of using both liver histopathology and serum ALT levels to assess hepatotoxicity. We have found, in agreement with previous reports, that the time course of hepatic AST leakage after CC14 treatment can vary significantly from animal to animal in a treatment group. Whereas the leakage of cellular enzymes is a transient event, liver necrosis is evident for 3-5 days. Thus, by determining the extent of liver necrosis (score) 48 or 72 hr after a CC14 insult, in combination with serum enzyme levels, we gain considerable confidence in reporting the degree of hepatotoxicity or hepatoprotection.
Because the hepatic metabolism of CC14 is essential for the expression of CC14 toxicity, one obvious explanation for this protection would be the inhibition of drug metabolism by CS and TS administration. Previous reports have demonstrated that numerous drugs that potentiate or inhibit CC14 toxicity similarly affect the metabolism of pentobarbitone, resulting in a decrease or increase in sleeping time, respectively (24)(25)(26)(27). As our studies demonstrated no significant alteration in pentobarbitone sleeping time after CS or TS pretreatment, we conclude that inhibition of CC14 metabolism probably is not responsible for the cytoprotection observed with these compounds. However, additional studies examining the effect of these protective agents on hepatic P-450 IIEI activity and spectral P-450 levels are required.
Another possible explanation for the unique protective properties observed after TS and CS pretreatment is the release of succinate from hepatocellular TS or CS. Previous studies indicate that the administration of succinate can protect mitochondria, cells, and organisms from a variety of toxic insults (28)(29)(30) and that succinate is a preferred mitochondrial substrate with respect to energy production (oxidative phosphorylation) and the reduction of pyridine nucleotides (31,32). Because the cellular uptake and accumulation of succinate is severely limited by its hydrophilicity (33), the administration of large concentrations of succinate (g/kg body weight, in vivo) are required to alter the cellular energy status or provide cytoprotection. In contrast, the administration of TS or CS provides a lipophilic carrier for succinate, thus promoting both cellular uptake and accumulation. Interestingly, Simon et al. (34) reported that the majority of intravenously administered TS was found as intact TS in the hepatic microsomal and mitochondrial fractions. In the present study, we found that approximately 120 nmol succinate/g liver is released from hepatocellular TS during a 24-hr period (the hepatic TS concentration at 24 and 48 hr after TS administration were 120 nmol/g and 0 nmol/g ,respectively; Table  3). Assuming that the normal endogenous succinate concentration in fasted rat liver is 170 to 270 nmol/g (35,36), the amount of succinate released from hepatocellular TS would result in a 45-70% increase in succinate levels over a 24-hr period. This substantial increase in succinate, released and used in the hepatocyte, might enable the maintenance of cellular metabolic and .e e -* * repair processes required for survival during a toxic CC14 insult. Additional experimental evidence will be required to confirm the role of succinate in TS and CS protection against CCl4-induced hepatotoxicity. Because an analytical method to measure tissue CS concentrations has not yet been developed, we can only assume (due to common physicochemical properties) that the hepatic accumulation and dispositon of CS is similar to that found for TS. A final explanation for TS-and CSmediated hepatoprotection is that the amphipathic TS and CS molecules are the protective agents. Data from the present study suggest that the ionic nature of TS or CS is important for protection. If the acidic nature of the tocopherol or cholesterol molecule is eliminated, as seen with T, TA, TN, C or CA, the protective properties of these compounds are eliminated. Similar findings were also reported in our previous in vitro studies on TS-mediated cytoprotection (1,5,6). Once in the cell, the TS and CS molecules might alter membrane stability or function by interacting with membrane-bound proteins and/or with the unsaturated fatty acid portion of phospholipid (tocopherol or cholesterol portion) as well as with the polar region of the phospholipid (ionic succinate moiety) (37). In fact, a unique membraneassociated property of TS and CS (not found with other nonionic cholesterol or tocopherol analogs) is their ability to spontaneously form multilamellar liposomes (bilayers) in aqueous solution (13). However, this bilayer organization of TS can be disrupted by increasing the calcium concentration to 2.5 mM or greater (38). This calcium concentration (>2.5mM) also eliminated TS-induced protection against toxic injury in isolated hepatocytes (1).
Numerous reports indicate that TS and CS administration do indeed stabilize membranes (13,37,(39)(40)(41) and alter membrane enzymatic and receptor activity (40)(41)(42)(43)(44)(45). For example, Brase and Westfall (42) reported that the addition of TS stimulates rat liver phenylalanine hydroxylase activity, while Chelliah and Fariss (44) have recently discovered that both TS and CS are potent inhibitors of purified electric eel-derived acetylcholinesterase activity but not butyrylcholinesterase activity. Numerous reports also attest to the effect of CS on the physical state of cellular membranes (increased viscosity), which can influence a variety of membrane functions (39)(40)(41). For example, Levy et al. (39) demonstrated a marked reduction in cellular ion leakage after human erythrocyte membranes were enriched with CS. Kolena and Kasal (40) reported that the incubation of rat testicular membranes with CS and other dicar-boxylic acid esters of cholesterol resulted in an increase in both membrane lipid microviscosity and in human chorionic gonadotropin (hCG)-receptor binding. The elimination of the free carboxylic acid group of CS and the other cholesterol esters abolished their stimulatory effect on both the leutenizing hormone/hCG receptor and membrane microviscosity. Using synaptosomal brain membranes, Lazar and Medzihradsky (41) demonstrated that the addition of CS restored the microviscosity and opioid receptor binding capacity of membranes fluidized by fatty acids. It is well known that numerous physiological and pathological conditions, including CC14 toxicity (46)(47)(48), can modulate the lipid composition and fluidity of biological membranes resulting in diminished structural integrity and altered membranebound enzyme activity (e.g., calcium ATPase activity). We speculate that the activity of CS and TS on cellular membranes, as described above, might prevent or reverse the membrane solvent effect (49) and lipid peroxidative (46)(47)(48) effect associated with CC14 hepatotoxicity.
In summary, we have found that supplementation of endogenous unesterified T levels does not protect the liver from the toxic effects of CC14 In contrast, the administration of the succinate esters of tocopherol and cholesterol afford rats protection against CCl4-induced toxicity. In fact, rats pretreated with CS were completely protected from CCl4-mediated liver damage. The observed protection appears to depend on the hepatocellular accumulation of TS and CS and does not appear to be related to an alteration in hepatic drug metabolism. Future investigations on the mechanism of CS and TS protection against CC14 toxicity should provide valuable information about the critical cellular events responsible for CC14 toxicity and potential strategies to protect ourselves from this toxic chemical and other chlorohydrocarbons.