PO-491 Single-cell phenotypic profiling of breast cancerpatient-derived tumour xenografts using mass cytometry

Abstract

IntroductionDespite the advances in patient stratification and the understanding of basic molecular mechanisms underlining breast cancer (BC), 30% of patients with early stages of the disease later relapse. To offer more effective treatments to these patients, we need to adapt pre-clinical drug development to account for the main feature of human cancer: heterogeneity. Our group has established a biobank of 160 BC patient-derived tumour xenografts (PDTXs) and has robustly shown that PDTXs and their matching patient-derived tumour cells (PDTCs) preserve most of the originating sample’s molecular features, including intra-tumour clonal diversity (Bruna et al. Cell 2016). This integrated platform permits high-throughput (HT) assessment of drug responses ex vivo in PDTCs and their validation in vivo in PDTXs. Thus, this PDTX/PDTC biobank represents a powerful resource for pre-clinical drug development in BC.Material and methodsHere, we aim to add a crucial layer of molecular characterisation at single-cell resolution by studying the phenotypic tumour heterogeneity in BC PDTXs and its role in drug response mechanisms. To achieve this goal, we employed the state-of-the-art HT single-cell method called mass cytometry (CYTOF), which combines the principles of flow cytometry and mass spectrometry allowing the simultaneous analysis of about 40 molecular events using metal-conjugated antibodies.Results and discussionsWe designed and validated a BC-specific panel of 35 antibodies to investigate: a)the epithelial tumour compartment; b)the stroma tumour compartment; c)oncogenic signalling; d)cell cycle and apoptosis. We used this panel in CyTOF experiments to perform an in-depth phenotypic characterisation of a representative collection of BC cell lines and molecularly distinct PDTXs. Using data-driven unsupervised methods to analyse the CyTOF data focusing on the epithelial compartment, we show that we capture different degrees of both inter- and intra-tumour functional heterogeneity. Importantly, we show that CyTOF can be used to characterise the heterogeneity in signalling dynamics within the different tumour compartments under the perturbation of therapy.ConclusionCyTOF is a novel approach to characterise heterogeneity and drug responses in BC PDTXs. The integration of high-level features, such as single-cell phenotypic profiles, and functional features, such as drug response data, obtained with CyTOF will facilitate the pre-clinical development of novel therapeutic strategies and will help to uncover mechanisms of drug resistance.