View All Posters

P1011
Application of Cellular Systems Biology to Cytotoxicity Profiling
Presenter Lawrence Vernetti, Cellumen, USA
Additional Authors: L.A. Vernetti, K. Johnston, A. Gough, W. Irwin, K. Guiliano, D.L. Taylor
Toxic agents modulate the activity of biological systems by disrupting the integration and interaction of networks of genes, proteins and metabolic processes that give rise to systemic (dys)function. Thus, toxicity profiling is ideally performed with a systems biology approach. Data will be presented to highlight the value of a Cellular  Systems Biology approach to correlate cytotoxicity with toxic liabilities in vivo. Specifically, cytotoxicity models have been developed for HepG2 cells and primary rat hepatocytes which combined with computational classifiers provide insight into the toxic liabilities of multiple chemical series. These models utilize panels of  biomarkers representing a broad array of toxic mechanisms which can be simultaneously captured by image processing and include markers of nuclear features, cytoskeletal stability, oxidative stress, stress kinase pathway activity, mitochondrial function, DNA damage response and mitotic arrest. These biomarkers are extended in the rat hepatocyte panel by inclusion of measures for cytochrome C, peroxisomal proliferation, and phospholipidosis. A set of 137 compounds with drug safety reference data was screened through the two panels in a 384 well HCS format. A sub-set of the in vitro cytotoxicity compound profiles and corresponding in vivo toxicity scorings were used to create an objective automated classifier that generates (a) similarity to toxins with known mechanisms of action; (b) the rank order for toxicity within the compound set; (c), identification of rodent liver toxins, and (d) an overall safety index. The classifier was tested against the remainder of the compound set and shown to correctly identify moderate and significant in vivo toxins. An assessment of the combined 64 cell feature set from the HepG2 and rat hepatocyte panels will be presented. This Cellular Systems Biology screening approach can be used to prioritize large sets of compounds according to in vivo toxic liability and provide early information on potential toxic mechanisms of action.