Fatty liver and impaired hepatic metabolism alter the congener-specific distribution of polychlorinated biphenyls (PCBs) in mice with a liver-specific deletion of cytochrome P450 reductase

Polychlorinated biphenyls (PCBs) are persistent organic pollutants that are linked to adverse health outcomes. PCB tissue levels are determinants of PCB toxicity; however, it is unclear how factors, such as an altered metabolism and/or a fatty liver, affect PCB distribution in vivo. We determined the congener specific disposition of PCBs in mice with a liver-specific deletion of cytochrome P450 reductase (KO), a model of fatty liver with impaired hepatic metabolism, and wild-type (WT) mice. Eight-week-old male WT (M-WT, n = 3), male KO (MKO, n = 5), female WT (F-WT, n = 4), and female KO mice (F-KO, n = 4) were exposed orally to Aroclor 1254. PCBs were quantified in adipose, blood, brain, and liver tissues by gas chromatography-mass spectrometry. The Sigma PCB levels followed the rank order adipose similar to liver brain > blood in WT and adipose liver > brain > blood in KO mice. PCB levels were much higher in the liver of KO than WT mice, irrespective of the sex. A comparison across exposure groups revealed minor genotype and sex-dependent differences in the PCB congener profiles (cos Theta > 0.92). Within each exposure group, tissue profiles showed small differences between tissues (cos Theta = 0.85 to 0.98). These differences were due to a decrease in metabolically more labile PCB congeners and an increase in congeners resistant to metabolism. The tissue-to-blood ratio of PCBs decreased for adipose, increased for the brain, and remained constant for the liver with an increase in chlorination. While these ratios did not follow the trends predicted using a composition-based model, the agreement between experimental and calculated partition coefficients was reasonable. Although the distribution of PCBs differs between KO and WT mice, the magnitude of the partitioning of PCBs from the blood into tissues can be approximated using composition-based models. (C) 2020 Elsevier Ltd. All rights reserved.