PHENOTYPIC AND CHEMOTHERAPY RESPONSE PROFILING OF P53 WILD-TYPE AND MUTANT HUMAN BREAST CANCER CELL LINES

Abstract

Anthracycline-based chemotherapy is the mainstay neoadjuvant therapy for breast cancer. However, it is efficacious in only 60% of patients while carrying substantial toxicity. The application of a predictive marker of response may spare predicted ‘poor responders’ from the toxicity. Previously, we demonstrated a gene expression signature that predicts chemotherapy resistance which is linked to TP53 integrity. Further investigation showed that p53 signatures predict response in only ER+ tumors. We hypothesized that the loss of p53 confers an elevated chemotherapy sensitivity in ER+ breast tumors. We engineered isogenic p53 mutant ER+ breast cancer cell lines and assayed their cell cycle kinetics and chemotherapy sensitivity. Our results demonstrated that the loss of p53 is necessary to abrogate p53-mediated cell cycle arrest and produce an increase in apoptosis. Therefore, p53 signatures may be utilized as a predictive marker of response for patients with ER+ breast tumor and spare ‘poor responders’ from toxicity. Since ER+ p53 wild-type breast tumors are associated with anthracycline resistance, new anticancer drugs against that subgroup of tumors are needed. Phenotypic drug screening approach, which do not focus on isolated targets but instead classify compounds by their impact on cell physiology, is highly suitable for this purpose. Current cell-based phenotypic assays require fixation and staining for phenotypic markers, which reduce screen throughput and introduce potential variations and artifacts. Here we describe a high-content live-cell phenotypic assay, which streamlines the process of cytological profiling and provides a consistent platform for empirically evaluating drug action. Importantly, when combined with chemical similarity clustering, the phenotypic assay provided an inference of structure-activity relationships. Finally, a small-scale phenotypic screen of natural products enabled classification of unknown compounds against the cytological profiles of commercial compounds. Hence, the phenotypic screen provides a new and robust opportunity for accelerating the evaluation of compound activity during high-throughput drug screens.