Lethal injury by diquat redox cycling in an isolated hepatocyte model

Hepatocyte isolated by collagenase perfusion of livers of male Fischer-344 rats, and treated with 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) (50 microM for 30 min at 37 degrees C) to inhibit glutathione reductase, were significantly more vulnerable to cytotoxicity of the bipyridyl herbicide diquat than similarly treated cells of Sprague-Dawley rats. Without compromise of cell defenses by BCNU, diquat was not cytotoxic to hepatocytes from either strain. Microsomal enzyme induction with phenobarbital (80 mg/kg ip for 3 days before hepatocyte isolation) did not potentiate killing of Fischer hepatocytes by diquat. Specific activities of NADPH-cytochrome P-450 reductase in isolated Fischer and Sprague-Dawley rat liver microsomes utilizing 1 mM diquat as acceptor were 0.085 +/- 0.017 and 0.076 +/- 0.028 mumol/mg.min (mean +/- SEM, N = 5), respectively, indicating the capacity for very active redox cycling of diquat by this route in both strains. The serine protease inhibitor, phenylmethylsulfonyl fluoride (100 microM), had no effect on diquat cytotoxicity, but both leupeptin (100 micrograms/ml) and antipain (50 or 100 microM) were able to delay, through not completely prevent, diquat-induced cell death. The phospholipase inhibitors, chlorpromazine (50 or 100 microM) and dibucaine (50 or 100 microM), similarly delayed but did not prevent cell death. Diquat increased the rate of hepatocyte phospholipid hydrolysis, measured as release into the suspending medium of [14C]arachidonic acid previously incorporated into hepatocyte lipids, but although chlorpromazine decreased phospholipid hydrolysis to the control rate, only partial protection against diquat cytotoxicity was seen. These data suggest that activation of phospholipase A2 and proteases by elevation of cytosolic free Ca2+ cannot account entirely for the loss of cell viability observed in the presence of cytotoxic concentrations of diquat.