Breast cancer progression and outcome are determined not only by characteristics of the tumor epithelial cells, but also by the overall arrangement and activity of the tumor stroma. Our previous work has demonstrated that gene expression signatures in human tumor stroma can predict outcome of breast cancer patients independently of established clinical parameters and molecular subtypes. The goal of this project is to use a systems approach to define the crifical pathways mediating communication between the tumor epithelium and stromal compartments in the poor outcome-associated triple-negative (TN) and HER2-positive (HER2+) human breast cancer molecular subtypes. We will extend our previous work in an attempt to better define stromal subtypes using both mRNA and miRNA expression patterns. This project has three major aims:
Aim 1. To develop coordinate stromal-epithelial expression signatures for TN and HER2+ breast cancer. Stroma and tumor epithelium compartments will be isolated by laser capture microdissection from a set of previously banked tumor samples that are linked to diagnostic, treatment and outcome information. Gene expression profiles will be generated for each, and the resulting data will be analyzed to further characterize specific signatures as well as to compare the results with those obtained from mouse models in projects 1 and 2.
Aim 2. To identify stromal-epithelial gene interaction networks. Using a de novo bioinformatics methodology termed STR-EPI, genes and pathways showing coordinated behavior and potenfially modulating crosstalk between the two compartments will be identified, with a focus on secreted and cell-surface proteins that may directly mediate communication. These results will be validated in available larger sample sets and directly examined experimentally with Projects 2 and 3, and their links to outcome and response to therapy will be investigated.
Aim 3. To identify and integrate stromal-epithelial and serum-derived miRNA signatures. Profiles will be generated from miRNA isolated in parallel from the samples described above, as well as from matched serum samples. These will be analyzed in tandem with mRNA data in order to develop robust and highly accurate clinically accessible biomarkers for stroma/epithelial features and microenvironment-based classification.
The identification of pathways governing interactions between breast cancer cells and the surrounding stroma will permit the development of novel prognostic markers and potential therapeutic targets. miRNA based tests suitable for use in standard clinical material, includingserum, will allow improved patient stratification, staging and therapeutic decision-making, leading to more personalized treatment.
Showing the most recent 10 out of 89 publications
