OS06.3 Characterisation of human brain tumour initiating cells leading invasion

Abstract

Abstract Introduction: Glioblastoma (GBM) belongs to the most aggressive cancers. Main pathophysiological feature of GBM is fast and widespread invasion of the brain parenchyma by a subpopulation of progenitor tumour cells, rendering localized therapies ineffective. In addition, brain tumour initiating cells (BTIC) from GBM may repopulate the tumour site after initially effective treatments. Thus, the invasive progenitor tumour cell fraction is a strategic prime target for early and sustained anti-invasion therapies. To date, the dependency on and interaction with the microenvironment of progenitor cells and their adaptive development from the tumour cell fraction has not been thoroughly investigated. Methods: With organotypic brain slice cultures (OBSC), single cell RNA expression analysis and bioinformatics workup, we established a novel integrated approach to investigate the expression profile of invasion leading cells. A micromanipulator adapter was developed to isolate different three subpopulations within human GBM progenitor cells: leader, follower, and stationary, from the OBSCs. In comparison, we use long-term intravital imaging to analyse the behaviour and interaction of the three cellular fractions with each other and the microenvironment in vivo. Imaging through a cranial window in the brain of a living mouse enables us to visualize single tumour cells. Using a photoswitchable vector expressed in BTICs we specifically label and isolate leader, follower, and stationary cells separately to compare leader cell behaviour in vivo, in situ, and in the patient situation. Findings: The in situ OBSC model is a valuable tool to analyse migration and invasion in conditions simulating normal brain tissue including some microenvironmental influences. This allowed us to monitor human BTIC cell invasion in real time. Invasive cells emerge from the initial cell population, driving invasion into the surrounding tissue. Those leader cells differ in morphology and behaviour from stationary cells, which remain at the implantation site. A subsequent microarray analysis of single cells revealed that leader cells have a markedly distinct expression pattern in comparison to follower and stationary cells. This leader cell-specific expression signature is currently verified in fresh patient tumour biopsies, other BTICs analysed in the OBSC as model, and in the cranial window model. Updated results will be presented at the WFNOS meeting. Outlook: We showed that invasive leader cells have a specific functional and molecular profile. A thorough characterization of these cells in respect to their interaction with and response to the microenvironment that may be involved in the process of transforming cells into invasive leader cells is ongoing. We hypothesize that a deep characterization of these events will shed light on the chronology of one of the gliomagenesis milestones, namely invasion.