J. Cairns, H. Li, C. Ung, L. Wang; Mayo Clinic, Rochester, MN

BACKGROUND: Taxane and anthracycline are two most effective chemotherapeutics in the treatment of triple negative breast cancer (TNBC). The aim of this study was to investigate the mechanisms for drug response phenotype by using a network biology platform.
METHODS: We developed Phenotype Mapping (P-Map), an extended version of a reverse engineering network tool, CellNet, by taking into account the properties of gene expression variations. We hypothesized that networks conferring same or convergent cellular phenotypes have less gene expression variations. Candidates that fulfilled the desired criteria were inferred from 276 lymphoblastoid cell lines transcriptome data with tabulated drug sensitivity for each cell line. Reverse engineered and protein interaction networks enriched with these genes were dissected for further validations in TNBC lines using specific siRNA knockdown, followed by MTS assay. Immunoblotting, qRT-PCR, flow cytometry and immunostaining were used to study the drug response mechanisms.
RESULTS: P-Map identified 27 genes for taxane response, and 24 for anthracycline response. 15 of the 27 genes significantly altered both docetaxel and paclitaxel response, and 13 of the 24 genes altered epirubicin and doxorubicin response (p<0.01). CDK6, ABCB1 and ABCG2 were regulated by candidates, as well as the percentage of CD44+/CD24- cells (23.14% vs 10.3%; p<0.01) and the activity of cancer stem like cell marker, ALDH (p<0.01). Western blot results revealed that some of these candidates affected the activities of AKT, mTOR, EGFR, NOTCH, ErbB, and Chemokine signaling pathways.
CONCLUSION: P-Map provides a platform for identifying novel drug response mechanism by identifying genes that contribute to variations in response to breast cancer chemotherapy.