The goal of this project is to identify tissue-specific molecular targets associated with chemical-induced carcinogenesis and pharmacophores for new anticancer drugs that “interact” with them.  The success of this project will result in the identification of new pharmacophores associated with a high degree of cell type-specificity and the cellular components they affect.  This project uses structure-activity relationship (SAR) modeling of tissue-specific carcinogens followed by comparative genomics and proteomics analyses to identify the molecular targets they interact with or influence.

The Specific Aims for this project are:

1. SAR Models: Update “control” SAR models for carcinogenicity and “experimental” SAR models for tissue-specific carcinogenesis.  The control models will be based on the traditional SAR method of contrasting carcinogens to non-carcinogens; the experimental models will be based on a new method of contrasting carcinogens from one site to those of all other sites.  Control models will handle the problem of “what makes a chemical a carcinogen” and the experimental models will tackle the problem of “what makes a carcinogen target a specific site.”
2. Toxicophore Identification: From SA1, define sets of compounds (i.e., with similar structure and biological activity) associated with tissue-specific carcinogenesis and refine these models with ligand-based 3D SAR to develop small libraries of suitable compounds for in vitro testing.
3. Target Identification: Using the compounds from SA2 as chemical probes, assay site-matched cell lines in order to identify what tissue-specific molecular targets they interact with through comparative genomics and proteomics analyses.

Our rationale is that although carcinogens show a degree of tissue specificity, it is not complete, which makes the study of specific mechanisms difficult and, to overcome this, we will divide and tackle the problem.  The first step involves SAR modeling that contrasts carcinogens from one tissue to carcinogens from all others.  Since carcinogens are on both sides of the SAR equation, carcinogenicity per se drops out, and we hypothesize that by controlling our analyses as such, features associated with targeting the carcinogen to the specific tissue will become discernable.  The second involves using small sets of structurally-defined compounds with tissue-specific activity for comparative genomics and proteomics analyses to identify new molecular target(s) associated with their tissue specificity.  New drugs based on this approach will therefore predominately interact with, and be toxic only to, (cancer) cells possessing the particular molecular target.