We hypothesize that epigenetic control of gene expression in addition to genomic amplifications, deletions and mutations between basal-like and claudin-low breast cancer is a major contributor for enrichment of the claudin-low characteristics in tumors that develop resistance to therapy. We will use two models to test our hypothesis. In vivo the C3Tag GEMM phenotypically resembles triple-negative human breast cancers by gene expression profiles, and shares genetic similarities including the acquisition of KRAS gene amplification and overexpression during tumor development. C3Tag tumors treated with inhibitors of MEK, downstream of KRAS signaling, ultimately progress through therapy. Progression was accompanied by reprogramming of tumors to signal through alternative pathways, with a transformation from “basal-like” to “claudin-low”. The in vitro model is SUM229 human breast cancer cells that maintain equilibrium between two populations: one basal-like with enhanced expression of epithelial markers (EpCAM+), and one with depleted epithelial markers (EpCAM-) that profiles as claudin-low. Exome sequencing, RNA-seq and epigenetic profiling in these two systems will define mechanisms for selection of the claudin-low phenotype during therapy. Biosensors for ERK, Ras, RAC, Cdc42 and Rho will be used for live imaging to define the dynamic changes in signaling networks differentially regulated in basal-like versus claudin-low cells. Bioinformatics and computer vision algorithms will be used to analyze the multiple datasets and develop computational models to understand the genetic and epigenetic mechanisms that characterize the basal-like versus claudin-low phenotype. These models will be used to predict novel combination therapies involving targeted kinase inhibitors and drugs that inhibit epigenetic reprogramming.
PI: Timothy Elston, PhD
Category: Team Science