Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Cocaine has been found to produce hepatic necrosis in mice (Shuster et al., 1977; Evans and Harbison, 1978; Freeman and Harbison, 1981), rats (Watanabe et al., 1987) and humans (Perino et al., 1987; Wanless et al., 1990; Kanel et al., 1990; Silva et al., 1991; Radin, 1992). Cocaine hepatotoxicity has been studied most extensively in mice, where it appears that cytochrome P450-mediated oxidation of cocaine is required for toxicity to occur (see Kloss et al., 1984; Boelsterli and Goldlin, 1991; and Roberts et al., 1992a for reviews). Through the oxidative pathway, cocaine is N-demethylated to norcocaine (Kloss et al., 1983a), which is further metabolized to N-hydroxynorcocaine (Kloss et al., 1982). N-Hydroxynorcocaine is then oxidized to norcocaine nitroxide (Misra et al., 1979; Kloss et al., 1984). P450 enzymes have been shown to participate in each of these steps of oxidative cocaine metabolism, and there is evidence that flavin-containing monooxygenase may also partially mediate the metabolism of cocaine to norcocaine (Rauckman et al., 1982a; Kloss et al., 1983a). There has been speculation that norcocaine nitroxide, which is a stable radical, undergoes a further, one-electron oxidation to produce a highly reactive nitrosonium species (Charkoudian and Shuster, 1985).

Evidence for the role of oxidative cocaine metabolism in its hepatotoxicity comes principally from experiments that examined the effects of enzyme inducers and inhibitors on cocaine-induced liver injury. In general, pretreatments that increase P450 enzyme activity or decrease the activity of competing esteratic metabolism increase cocaine hepatotoxicity, whereas P450 inhibitors block or diminish liver injury from cocaine (Roberts et al., 1992a). The intermediate metabolites norcocaine and N-hydroxynorcocaine are hepatotoxic when administered directly to mice (Thompson et al., 1979), with potency greater than or equal to that of cocaine. The toxicity of both can be blocked by pretreatment with the P450 inhibitor SKF 525A (Thompson et al., 1979), which suggests that at least one additional oxidative step beyond N-hydroxynorcocaine formation is required to produce the toxic species.

Although evidence clearly points to the nitroxide metabolite as critical in the hepatotoxic effects of cocaine, there has been little study of the effects of this metabolite on the liver. Incubation of norcocaine nitroxide with mouse hepatic microsomal suspensions has been shown to stimulate lipid peroxidation (Rosen et al., 1982; Kloss et al., 1983b), but there has not yet been any direct evaluation of the toxicity of norcocaine nitroxide in vivo. In order to gain a better understanding of the potential role of the nitroxide metabolite of cocaine in its hepatotoxicity, we synthesized norcocaine nitroxide and administered it to mice. To facilitate comparisons with cocaine, a mouse strain (ICR) used previously in studies of cocaine hepatotoxicity in this laboratory was utilized for these experiments. We report here the results of studies examining both temporal aspects and dose-response relationships for toxicity, as well as the importance of both P450 and esteratic metabolism in the hepatic effects of norcocaine nitroxide.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals and reagents. Cocaine hydrochloride, troleandomycin, ketaconazole and chloramphenicol were obtained from Sigma Chemical Co. (St. Louis, MO). Norcocaine nitroxide was synthesized from norcocaine according to method of Rauckman et al. (1982a). Structure was confirmed by ESR, and the purity was found to be 90% by TLC and HPLC. Cimetidine and SKF 525A (2-diethylaminoethyl 2,2-diphenyl valerate) were a generous gift from SmithKline Beecham Pharmaceuticals (Philadelphia, PA). Diazinon (phosphothiotic acid O,O-diethyl O-[6-methyl-2-(1-methylethyl)-4-pyrimidinyl] ester) was purchased from Chem Service, Inc. (West Chester, PA), and sodium phenobarbital, U.S.P., N.F. was obtained from Spectrum Chemical Mfg. Corp. (Gardena, CA).

Animals and treatments. Male ICR mice (Harlan Sprague-Dawley, Indianapolis, IN) weighing 32 to 34 g were housed five per polycarbonate cage on corn cob bedding and given free access to food (Purina 5001, Ralston Mills, St. Louis, MO) and water. The animals were kept in temperature- and humidity-controlled animal quarters (temperature, 22 ± 2°C; humidity, 55 to 65%) with a 12-hr light/dark cycle. Mice were euthanized by CO₂ asphyxiation. Intracardiac blood for measurement of serum ALT activity was drawn immediately after the cessation of respiration. All procedures were approved by the University of Florida Institutional Animal Care and Use Committee.

Serum ALT measurements. Serum ALT activity was measured by the method of Bergmeyer et al. (1978) using an ALT 20 kit (Sigma Diagnostics, Inc., St. Louis, MO).

Histopathology. Liver was harvested, fixed in neutral buffered 10% formalin for at least 24 hr, trimmed, processed, embedded in paraffin, sectioned at 4 to 6 µm and stained with hematoxylin and eosin for light microscopic evaluation. For electron microscopy, tissues were fixed in 2% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.3, washed in 0.1 M cacodylate buffer and post-fixed in buffered 1% osmium tetroxide for 1 hr. Samples were then dehydrated in a graded ethanol series at 10-min intervals followed by a 2% uranyl acetate en bloc stain overnight. Ethanol-dehydrated samples were washed twice in 100% acetone, infiltrated with Embed 812 resin and polymerized at 60°C for 2 days. Ultrathin sections were collected on carbon-coated 0.25% formvar grids and poststained with 2% aqueous uranyl acetate followed by Reynold's lead citrate. Photomicrographs of stained sections were taken on a Hitachi H-7000 electron microscope at 75 kV.

Statistical analysis. Serum ALT activity data were analyzed by a one-way analysis of variance followed by a Student Neuman-Keuls' post-hoc test. Groups were considered significantly different when P  .05. Because of the log-normal distribution of serum ALT activity values, data were log transformed for statistical comparisons.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Time course of norcocaine nitroxide toxicity. The hepatic effects of the administration of a single dose of norcocaine nitroxide (50 mg/kg i.p.) were evaluated over time. Clinical observation of the animals after norcocaine nitroxide treatment found no evidence of CNS stimulation, such as occurs with a comparable dose of cocaine. Serum ALT activities were significantly elevated in response to norcocaine nitroxide, reaching peak levels 12-18 hr after treatment and declining thereafter (fig. 1). Serum ALT activities followed a similar time course in mice treated with cocaine (50 mg/kg i.p.) for comparison. Although the serum ALT activity levels were higher in cocaine-treated mice at each time-point, none of the differences was statistically significant (P > .05).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Time course of serum ALT activities after a single dose of norcocaine nitroxide. Serum ALT activities were measured at varying times after a single dose of norcocaine nitroxide or cocaine, 50 mg/kg i.p. Results are expressed as mean ± S.E.M., N = 5 to 10 animals. Untreated control mice had serum ALT activities of 62 ± 16 I.U./l; mice treated with saline (vehicle for cocaine) had serum ALT activities of xx ± xx I.U./l. Serum ALT activities for treated mice at all time-points (except 6 hr for norcocaine nitroxide) were significantly increased compared with controls, P < .05. There were no significant differences between serum ALT values for norcocaine nitroxide- and cocaine-treated mice.

View larger version (163K): [in this window] [in a new window]   Fig. 2.   Hepatocellular degeneration and necrosis in naive mice after a single dose of norcocaine nitroxide (50 mg/kg i.p.). Top panel) Liver section from a mouse euthanized 12 hr after the norcocaine nitroxide dose. In the centrilobular zone, and extending into the midlobular zone, hepatocytes are swollen, finely granular and pale (cloudy swelling); a few hepatocytes have pyknotic or fragmented nuclei and are necrotic. Hepatocytes adjacent to the central vein are spared. Bottom panel) Liver section from a mouse euthanized 18 hr after the norcocaine nitroxide dose. There is degeneration of nearly all centrilobular and midzonal hepatocytes, with sparing of only a few hepatocytes immediately adjacent to the central vein; many hepatocytes are necrotic, as evidenced by nuclear pyknosis, fragmentation or lysis. C, central; P, portal. Hematoxylin-eosin stain. Bar = 40 µm; original magnification, ×400.

View larger version (163K): [in this window] [in a new window]   Fig. 3.   Subcellular morphology after a single dose of norcocaine nitroxide. Top panel) Representative liver section from vehicle (corn oil)-treated control mouse showing an intact, rounded nucleus and numerous mitochondria. Bottom panel) Representative liver section from a mouse 30 min after administration of norcocaine nitroxide (50 mg/kg i.p.). Dilation of the nuclear membrane, presence of vacuoles caused by dilation of the cytocapillary network and dissociation of ribosomes from the rough endoplasmic reticulum are evident. Original magnification, ×18,000.

Dose-response relationship for norcocaine nitroxide hepatotoxicity. Liver toxicity (based on serum ALT activity) caused by norcocaine nitroxide treatment increased with the dose administered (fig. 4). Mean serum ALT activities 12 hr after administration for mice injected with 10 or 20 mg/kg norcocaine nitroxide were not significantly different from controls, whereas mice injected with 30, 40 or 50 mg/kg norcocaine nitroxide had significantly elevated serum ALT activity levels (P  .05). Serum ALT activities were highest in mice injected with 50 mg/kg i.p., but these results were not significantly different from those in mice injected with 40 mg/kg.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Serum ALT activities after treatment with varying doses of norcocaine nitroxide. Mice were euthanized 12 hr after the norcocaine nitroxide dose. Results are expressed as mean ± S.E.M., N = 8 to 10 animals. *Significantly different from vehicle (corn oil)-treated controls (41 ± 2 I.U./l), P < .05.

Effects of enzyme induction and inhibition on norcocaine nitroxide toxicity. Serum ALT activities and hepatic lesions were evaluated in mice 12 hr after a single dose of norcocaine nitroxide, with or without pretreatment with enzyme inducers or inhibitors (fig. 5). Serum ALT activities were not significantly affected by phenobarbital pretreatment. However, pretreatment with phenobarbital shifted the site of necrosis from the midzonal and centrilobular regions (zones 2 and 3, respectively) to the periportal region (zone 1) of the lobule (fig. 6).

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Effects of hepatic enzyme inducers and inhibitors on norcocaine nitroxide toxicity. Mice were pretreated with inducers and inhibitors of metabolism using regimens described in "Materials and Methods." After pretreatment, mice were administered a single dose of norcocaine nitroxide (50 mg/kg i.p.) and euthanized 12 hr later for measurement of serum ALT activity. Results are expressed as mean ± S.E.M.; panel A, N = 7 to 10 animals; panel B, N = 10 to 20 animals. *Significantly different from mice pretreated with saline, P < .05.

View larger version (160K): [in this window] [in a new window]   Fig. 6.   Effect of pretreatment with phenobarbital or SKF 525A on hepatotoxicity in mice 12 hr after a single dose of norcocaine nitroxide (50 mg/kg i.p.). Top panel) Liver from a mouse pretreated with phenobarbital. Hepatocellular degeneration and necrosis are confined to the periportal region. Bottom panel) Liver from a mouse pretreated with SFK 525A, showing normal morphology. C, central; P, portal. Hematoxylin-eosin stain. Bar = 40 µm. Original magnification, ×400.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Norcocaine nitroxide was found to produce in mice dose-dependent hepatotoxicity remarkably similar to that produced by cocaine. Previous reports have indicated that significant elevations of serum ALT activities result from cocaine doses greater than 30 mg/kg i.p. in this mouse strain (Roberts et al., 1992b). With this as a basis for comparison, the results shown in figure 4 suggest that cocaine and norcocaine nitroxide are similar in potency. This interpretation must be made cautiously, however, given the uncertainty regarding the fate of an administered dose of the nitroxide derivative of cocaine. Previous studies have shown that although norcocaine nitroxide is stable in solution, it may be reduced by sulfhydryl compounds such as glutathione when divalent cations are present (Rauckman et al., 1982a). The rate and extent to which reduction of the nitroxide occurs in vivo are unknown, but such a reaction could cause the loss of much of an i.p. dose of norcocaine nitroxide before it reaches the liver. If this is the case, the toxic potency of norcocaine nitroxide in the liver relative to that of cocaine may be greater than is implied by the comparison presented here.

Norcocaine nitroxide and cocaine were found to produce essentially the same lesion: hepatocellular necrosis predominantly in the midzonal region. Previous studies have shown that phenobarbital induction results in an apparent shift in the cocaine lesion from the midzonal or centrilobular region (depending on the mouse strain) to the periportal region (Roth et al., 1992; Powell et al., 1991). The same shift was observed in this study for hepatic necrosis from norcocaine nitroxide. The basis for the shift in the cocaine lesion has never been explained, but the underlying mechanism appears to be relevant for toxicity from the nitroxide metabolite as well as from cocaine.

Perhaps one of the most striking features of norcocaine nitroxide liver injury is how rapidly it develops. Electron microscopic examination of the liver shows dramatic changes in subcellular structure within 30 min of the norcocaine nitroxide dose (fig. 3). In particular, there are pronounced changes in mitochondria and endoplasmic reticulum. Previous studies in our laboratory and others have found swelling of both smooth and rough endoplasmic reticulum and loss of ribosomes after cocaine administration (Gottfried et al., 1986; Mehanny and Abdel-Rahman, 1991; Roth et al., 1992; Powers et al., 1992). Recent studies have reported mitochondrial inner membrane permeability changes and lipid peroxidation from cocaine in rats (Masini et al., 1996), although reports of alterations in mitochondrial morphology after cocaine treatment have been inconsistent (Gottfried et al., 1986; Powers et al., 1992). Overall, observations regarding the nature and timing of the histopathologic response to norcocaine nitroxide observed in this study appear to be consistent with a role for this metabolite in cocaine hepatotoxicity.

The importance of P450 activity in the bioactivation of cocaine has been well established, and some of the participating P450 forms have been identified. For example, N-demethylation of cocaine to norcocaine has been shown to be mediated by CYP3A in mouse and human liver (LeDuc et al., 1993; Pellinen et al., 1994). The P450(s) that catalyze subsequent oxidative steps in these species have not yet been determined, however. In rats, evidence suggests that CYP3A and CYP2B can N-demethylate cocaine, and CYP2B may be required for subsequent oxidative metabolism of norcocaine (Poet et al., 1996). The results of the present study show that the hepatotoxicity of norcocaine nitroxide is also dependent on P450 activitypretreatment with the P450 inhibitors SKF 525A or cimetidine effectively inhibited norcocaine nitroxide-induced liver injury. Toxicity was also inhibited by pretreatment with P450 inhibitors selective for CYP3A (troleandomycin and ketaconazole) and CYP2B1/2 (chloramphenicol), a result that suggests a role for these specific P450 forms. This latter observation was surprising in that phenobarbital pretreatment, which increases activity of both CYP3A and CYP2B1/2 in rodents (Waxman and Azaroff, 1992; Murayama et al., 1996), did not increase the hepatotoxicity of norcocaine nitroxide (fig. 5). Although these studies clearly demonstrate the importance of P450 activity for norcocaine nitroxide toxicity in the liver, additional studies will be required to identify the participating P450 form(s).

In the case of cocaine, the extent of hepatotoxicity appears to result from a balance between oxidative (P450) and esteratic metabolism. Inhibition of esteratic metabolism increases the hepatotoxicity of cocaine (Thompson et al., 1979), presumably by making more of the drug available for oxidative metabolism leading to bioactivation; in fact, increased levels of cocaine and oxidative metabolites have been measured in mice after cotreatment with the esterase inhibitor diazinon (Benuck et al., 1988). The ability of diazinon to potentiate the hepatotoxicity of norcocaine nitroxide in this study suggests that this metabolite also undergoes significant esteratic metabolism to a non-hepatotoxic product.

At least two mechanisms have been proposed by which terminal oxidative metabolites of cocaine might produce cytotoxicity. In one of these, it is thought that a futile redox cycle develops between N-hydroxynorcocaine and norcocaine nitroxide, oxidation to the nitroxide being mediated by P450 and reduction of the nitroxide to the N-hydroxy form being mediated by flavoproteins (Rauckman et al., 1982b). It is proposed that this futile redox cycle results in the generation of reactive oxygen species that lead to oxidative stress, including lipid peroxidation (Rauckman et al., 1982b; Boelsterli and Goldlin, 1991) (see fig. 7). Some experimental observations argue against this hypothesis. For example, no superoxide formation was detected during the metabolism of N-hydroxynorcocaine to norcocaine nitroxide by rat liver microsomes (Lloyd et al., 1993), and studies using rat hepatocytes have found that preventing the lipid peroxidation from cocaine by cotreatment with an antioxidant (-tocopherol polyethylene glycol 1000 succinate) has no effect on cytotoxicity (Goldlin and Boelsterli, 1991). On the other hand, generation of reactive oxygen species appears to be important in the toxicity of cocaine in hepatocyte cultures (Goldlin and Boelsterli, 1991; Boelsterli et al., 1993), which suggests that oxidative damage is a key feature of cocaine hepatotoxicity but that it occurs through nonperoxidative mechanisms. Another hypothesis holds that norcocaine nitroxide is further oxidized to a reactive metabolite that binds to critical subcellular targets (Evans, 1983) (fig. 7). In regard to this, Charkoudian and Schuster (1985) have proposed that the nitroxide undergoes a one-electron oxidation to a nitrosonium, a species that is characteristically highly reactive.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Postulated relationships between the terminal oxidative metabolism of cocaine and toxicity.

The results of the present study are consistent with both of these proposed mechanisms, including the requirement for P450 enzyme activity. In the case of the reactive metabolite mechanism, the required P450 would presumably be the enzyme that oxidizes the nitroxide to the reactive species. For the futile redox cycle mechanism, norcocaine nitroxide toxicity could be blocked by inhibiting the P450 enzyme responsible for oxidizing N-hydroxynorcocaine to norcocaine nitroxide. Even though the nitroxide was administered directly in these studies, an appropriate P450 inhibitor would break the cycle by preventing oxidation of N-hydroxynorcocaine formed from reduction of the administered nitroxide.

In conclusion, although there is ample evidence that terminal oxidative metabolites of cocaine are responsible for its hepatotoxic effects, there has been little direct study of the toxicological properties of these metabolites. This study confirms the widespread assumption that norcocaine nitroxide is hepatotoxic and indicates that both P450 and esterase activities are important determinants in its toxicity.