Cocaine–protein targets in mouse liver

Abstract

Cocaine has been shown to be hepatotoxic in mice, rats and humans. N-Oxidative metabolism of cocaine is required for this effect, and it has been proposed that binding of cocaine reactive metabolites formed via this pathway might be responsible for cytotoxicity. To explore this hypothesis, cocaine–protein adducts in liver following cocaine treatment in naive ICR mice were examined by Western blot analysis and compared with those formed in mice pretreated with phenobarbital or β-naphthoflavone. Phenobarbital and β-naphthoflavone pretreatments have been shown previously to shift the hepatic necrosis in ICR mice from the midzonal region to periportal and perivenular regions, respectively. Similar patterns of cocaine–protein adduction were detected in naive, phenobarbital-pretreated and β-naphthoflavone-pretreated mice, however, suggesting a consistent set of target proteins regardless where within the lobule toxicity occurs. To confirm that Western blot analysis using anti-cocaine antibody was capable of detecting all of the major cocaine–protein adducts, a separate experiment was conducted in which mice were treated with 14 C -labeled cocaine and cocaine–protein adducts were detected fluorographically. This technique detected essentially the same protein adducts as the Western blots. Two of the protein adducts were isolated, subjected to N-terminal sequence analysis, and found to have homology with hsp 60 and transferrin. Western blot analysis using anti-hsp 60 and anti-transferrin antibodies following two-dimension PAGE separation was used to confirm the identity of these protein targets. Impairment of function of either protein could plausibly contribute to cocaine hepatotoxicity, although this remains to be demonstrated.