Towards a better understanding of early drug-induced regulatory mechanisms of liver tumorigenesis

Non-genotoxic carcinogens (NGC) form a group of molecules that do not directly bind DNA (1) but that produce perturbations
in the gene expression and epigenetic state of cells which facilitate tumor formation, typically through the promotion of
preexisting neoplastic cells into neoplasms. The molecular events underlying the NGC-induced transformation of normal hepatocytes
to altered hepatocellular foci are still unclear and no acute early molecular markers for NGC are available for drugs under development.
According to regulatory expectations, drug safety is tested in both short term in vitro and long-term
in vivo studies in several experimental animals (rodent and non-rodent species) prior testing on human.
As the safety assessment in experimental animals has been very successful in predicting toxicity of
biologically active chemicals in humans, differences in species biochemistry or pathophysiology be-
tween human and rodents have raised doubts regarding the appropriateness of extrapolating some rodent
tumor findings to humans. A better understanding of NGC mode of action on cellular mechanism is
believed to help addressing the relevance of rodent assays to human risk assessment and help in
early prediction of NGC in drug development.
In this dissertation we have adapted innovative bioinformatic approaches to toxicogenomic data from
comprehensive in vivo experiments in order 1) to identify key early regulatory interactions underlying liver
drug-induced non-genotoxic carcinogenesis and 2) to examine potential species-specificity (human-mice)
in receptor-dependent mechanisms underlying liver tissue molecular responses to NGC. The outcome of
this research provides with novel mechanism-based candidate biomarkers for NGC, and allows for a better
understanding of early mechanisms and pathways underlying drug-induced toxicity in rodents and their
relevance to human.