Multiple Screening Techniques: A Way to Develop a Chemical-Animal Model

摘要

The primary objective and public health relevance of this investigation was to develop a chemical-animal model with a toxicological and therapeutic approach. The results outlined here are developed from the latest techniques being employed in the chemical and biomedical fields. This research outlines a model building approach that progressed from a preliminary agent screening technique (quantitative structure-activity relationship/structure-activity relationship, QSAR/SAR) and in vivo testing using the Chernoff-Kavlock (CK) assay through to in vitro testing in transgenic adenocarcinoma of the mouse prostate (TRAMP) cell lines. The preliminary investigation involved development of a QSAR/SAR model to predict the teratogenicity of a series of related chemical agents (dopamine mimetics). This QSAR/SAR model was then validated using a complete leave one out cross-validation. The predictivity of a more general QSAR/SAR model of developmental toxicity was then tested experimentally in vivo using the chemical agent retinoic acid. The second model was based on in vivo animal screening using the CK assay. The CK assay involves the dosing of pregnant animals, either mice or rats, during the organogenesis period of fetal development. This assay quantitatively measures effects on fetal viability and growth, and allows for a more qualitative assessment of teratogenicity by recording obvious malformations. The third segment of this study was an in vitro evaluation of the effects of a series of microtubule perturbing agents on cell viability, cell death and gene expression of the TRAMP cell lines. This research could contribute to the development of drug treatments that would be more effective against human prostate cancer.In the first section of my thesis, a mathematical model was generated with experimental data from the literature on a congeneric series of twelve dopamine mimetics. Based on a single physicochemical parameter, the final model is 100% effective at predicting biological activity (teratogenicity) of dopamine mimetics. We also found inconsistencies in the original biological data that might influence the choice of final model.The second section of my thesis involves the experimental validation of a general QSAR/SAR model that predicted retinoic acid would be positive for developmental toxicity. Retinoic acid was therefore tested in a standard mouse CK assay (the same assay used to generate the data used to generate the model) to test the SAR model prediction. Significant increases in the incidence of both fetal death and intrauterine growth retardation were observed in the offspring of the treated mice. Statistical analysis revealed these effects were dose-dependent. These results demonstrated, in a quantitative manner, the developmental toxic effects of retinoic acid in the mouse, as were predicted by the SAR model and as expected from developmental literature.The final segment of my thesis dealt with the preliminary in vitro screening of four promising anticancer agents, Analog II, 4-methoxy Analog II, JR oxime I and TDH 169 on the clonal TRAMP cell lines C1A, C2H and C2N. 4-Methoxy Analog II displayed the most promising antiproliferative effects and apoptosis inducing effects. A microarray analysis of mRNA expression in response to 4-methoxy Analog II was conducted to determine agent-induced expression alterations in the C1A cell line. Upregulation of the apoptosis activating genes Bok and Siva-pending was observed, while the apoptosis inhibiting genes Birc 4, Dad1 and Atf5 were significantly downregulated.