Pharmacophore modeling strategies for the development of novel nonsteroidal inhibitors of human aromatase (CYP19).

The present study utilizes for the first time the structural information of aromatase, an important pharmacological target in anti-breast cancer therapy, to extract the pharmacophoric features important for interactions between the enzyme and its substrate, androstenedione. A ligand-based pharmacophore model developed from the most comprehensive list of nonsteroidal aromatase inhibitors (AIs) is described and explained, as well. This study demonstrates that the ligand-based pharmacophore model contributes to efficacy while the structure-based model contributes to specificity. It is also shown that a 'merged' model (i.e., a merged structure-based and ligand-based model) can successfully identify known AIs and differentiate between active and inactive inhibitors. Therefore, this model can be effectively used to identify the next generation of highly specific and less toxic aromatase inhibitors for breast cancer treatment.