Precision-cut liver slice culture: An in vitro tool for assessing hepatotoxic interactions of organohalides

Abstract

A serious impediment to the application of fundamental toxicology to the protection of human and environmental health is that most organisms are exposed to mixtures of chemical agents while the majority of toxicology research elucidates the toxic actions of individual agents. Our current state of knowledge is insufficient for predicting the effects of a combination of agents based on the dose-response characteristics of the agents administered singly. Even when dose-response data from mixtures and their individual components are available, no clear consensus exists as to which means are appropriate for determining if an interaction such as synergy or antagonism is indicated by those data. Sound mathematical analysis of toxic interaction is an essential ingredient in this pursuit. Experimental designs and means of data analysis permitting precarious conclusions remain in common use, and impede the characterization and elucidation of xenobiotic interactions. This thesis critiques some of the approaches used to address xenobiotic interaction, and offers specific and novel techniques and guidelines for improved approaches. Increasingly large numbers of toxicants exceed our current ability to assess toxicity. The development of in vitro methods offers an increased ability to examine larger numbers of toxicants and their combinations than conventional in vivo approaches given the finite resources available. This thesis presents evidence supporting the validation of precision-cut liver slice culture as an in vitro model for investigating hepatotoxic interactions of defined binary mixtures. Toxic interactions observed in vivo were demonstrated in the in vitro liver slice culture in two strains of rat. No intrinsic bias was detected by challenging this approach with a sham interaction (one compound combined with itself). Structure-activity based predictions of toxic interaction were demonstrated in liver slice culture. Two separate means of data analysis arrived at the same interpretations of the data for all of the experimental results described above. No toxic interactions were found in a limited but rigorous test of a bacterial toxicity assay, suggesting that interactive toxic responses are sensitive to the choice of biological model. Preliminary experiments were conducted for assessing the effect of mechanistic probes (metabolic manipulations) on established toxic interactions.