Validation and assessment of protein and mRNA biomarkers in breast cancer: Focus on MET and ESR1

Breast cancer is the second leading cause of cancer death among women in the United States with approximately 39,500 deaths per year and over 450,000 deaths worldwide. Currently, expression of the breast cancer biomarkers estrogen receptor (ER), progesterone receptor (PgR) and Her2 determined by immunohistochemistry (IHC) are used to predict patient prognosis and to provide the rationale behind patient treatment. Unfortunately, these biomarkers do not accurately predict which patients will be resistant to a given therapy (either de novo or acquired resistance). In this dissertation, I discuss the major first step of antibody validation required for developing a new protein biomarker and sought to utilize these methods to validate the Met receptor as a prognostic marker in breast cancer and generated GFP tagged Met constructs to develop an in vitro model for the subcellular localization of the receptor. Here I document that commercial Met receptor antibodies used in this study were not completely specific when tested by Western blotting whereas non-commercial monoclonal Met antibodies Met4 and Met19s were highly specific. However Met4 and Met19s antibodies demonstrated a lack of reproducibility on the Yale breast cancer cohort YTMA 49 and led to inconclusive results regarding the prognostic value of Met protein expression. Initial studies suggest that nuclear Met could also serve as a prognostic marker although stable expression of full-length Met receptor with a C-terminal GFP tag or Met-GFP truncation constructs in cancer cell lines revealed no evidence of Met localized to the nucleus in cell fractionation studies. I also conclude that the previously identified 60 kDa fragment thought to be a cleavage product of Met is most likely an artifact of non-specific commercial Met antibodies.;Additional biomarker studies have demonstrated the ability of mRNA gene signatures such as Oncotype DX to predict the likelihood of recurrence for patients with ER-positive, tamoxifen treated, lymph node-negative breast cancer and as such I sought to test the hypothesis that in situ measurement of ER mRNA (ESR1) can provide additional prognostic or predictive value to assessment of ER protein. Finally, the RNAscope method for in situ detection of ER mRNA was combined with AQUA to quantify ESR1 levels on two Yale breast cancer cohorts, demonstrating a non-linear relationship between ER mRNA and protein and that ESR1 is predictive of response to endocrine therapy.