Abstract PO-086: Local quantification of protein expression on frozen tissue sections to evaluate tumor heterogeneity

Abstract

Abstract The discernment of intra-tumor heterogeneity is critical to understand therapy failure and differences of prognosis between patients with similar disease states. Part of the heterogeneity can be appreciated at the protein level, where quantitation can aid the elucidation of complex biological processes. We propose a workflow to quantify protein expression on a “local” level on frozen breast tumor tissue sections, providing a measure and easy visualization of the phenotypical heterogeneity present in the tissue. In this workflow, we extract 8-10 regions of interest (ROIs) of ~100 µm in diameter from frozen tissue sections of 5 patients using a microfluidic probe (MFP). The MFP is a microfluidic scanning probe device that confines a biochemical on a selected region of a tissue, with the capacity of working in liquid environments, and which allows the extraction of individual regions (footprints) in less than 3 min. The rapid extraction together with the liquid environment in which the tissue is present are key for minimizing protein degradation. The extracted proteins are analyzed through an antibody microarray with a panel of 13 proteins that aids a proteomic-based classification of molecular subtypes of breast cancer. Protein expression is quantified based on the greyscale intensity of the corresponding spot in the microarray and normalized by the area of the extracted region. Further analysis determines the presence and extension of phenotypical clusters. Using local protein expression, we generated an expression map of the 13 analyzed proteins. We observe large variations of protein concentrations in the case of beta-actin, estrogen receptor (ER) and EGFR. Cytokeratins (CK)5, CK8, CK17, progesterone receptor and GAPDH presented lower variations in concentration between the analyzed regions in this set of patients. We then evaluated whether the observed differences were statistically significant using Cohen size D effect, showing higher heterogeneity among some proteins (beta-actin, ER, or androgen receptor) and some patients (Patient 1). Finally, we performed a hierarchical clustering analysis across the patients to understand the presence and distribution of phenotypical variants. Patients 3 and 5 presented one single phenotypical variant, while the other patients had up to four. Quantitative methods to describe heterogeneity on the proteomic level are essential to improve our comprehension of tumor biology and thus develop more precise cancer diagnostic approaches. Techniques such as the one presented here can advance our understanding of the spatial location of phenotypical variants within the tumor mass, hinting towards a more accurate choice of treatment modalities. Citation Format: Anna Fomitcheva Khartchenko, Peter Schraml, Govind V. Kaigala. Local quantification of protein expression on frozen tissue sections to evaluate tumor heterogeneity [abstract]. In: Proceedings of the AACR Virtual Special Conference on Tumor Heterogeneity: From Single Cells to Clinical Impact; 2020 Sep 17-18. Philadelphia (PA): AACR; Cancer Res 2020;80(21 Suppl):Abstract nr PO-086.