Hepatic effects of a highly purified 2,2',3,4,4',5,5'-heptachlorbiphenyl (PCB 180) in male and female rats.

PCB 180 (2,2',3,4,4',5,5'-heptachlorobiphenyl) is a persistent and accumulating polychlorinated biphenyl abundantly present in food and the environment. In this study, we used highly purified PCB 180 (dioxinlike impurities: 2.7 ng TEQ(WHO)/g PCB 180) in a 28-day toxicity study in young adult Sprague-Dawley rats. Male and female rats were given total doses of 3, 10, 30, 100, 300, 1000 or 1700 mg/kg b.w. PCB 180 by gavage. Increased liver weights were observed at ≥ 300 mg/kg b.w. in males and females. No increases in serum ALT or ALP activities were found. A significant increase in liver pentoxyresorufin O-dealkylase (PROD) activity was found in males at ≥ 10 mg/kg b.w. and in females at ≥ 30 mg/kg b.w. In both genders, a significant induction of hepatic 7-ethoxyresorufin O-deethylase (EROD) activity was also observed in males at ≥ 10 mg/kg b.w. and in females at ≥ 300 mg/kg b.w. Western blotting showed that mainly cytochromes P450 (CYPs) 2B1/2 and 3A1 were induced while slight effects were seen on CYP1A1, CYP1A2 and CYP1B1. However, no induction of CYP1A1, 1A2 and 1B1 was found on the mRNA level, except for a slight effect in females at 1000 mg/kg b.w. Furthermore, hepatic UDP-glucuronosyltransferases (UGTs) 1A1 and 1A6 were markedly induced in males and slightly induced in females. The hepatic concentrations of apolar retinoids were decreased in males at ≥ 30 mg/kg b.w. and in females at ≥ 300 mg/kg b.w. Taken together our findings show that pure PCB 180 leads to hepatic changes in a dose range which did not cause CYP1A1 induction but causes centrilobular liver hypertrophy, affects drug-metabolizing enzymes involved in the metabolism of exogenous and endogenous substrates and leads to changes in liver retinoid levels. A benchmark dose (BMD) approach is presented in order to model lowest effective dose levels for these effects. Comparison of PCB 180 liver level related to BMDL₅ for hepatic hypertrophy in rats with human data on 'total' hepatic PCB levels in individuals without history of specific exposure suggests a relatively small margin of tissue burden in the range of 37-fold. Our results show that the highly pure non dioxin-like PCB 180 exerted strong effects different to dioxin-like compounds and that the low TEQ contamination allowed a characterization of the PCB as non-dioxinlike.