Cancer gets physical: understanding how the tumor microenvironment contributes to tumor progression

Abstract

An abnormal multicellular architecture and a stiffened extracellular matrix (ECM) are defining characteristics of breast cancer, and yet, most in vitro tumor models fail to recapitulate the aberrant tumor microenvironment or accurately predict in vivo cellular responses to therapeutics. This dissertation aims to fill this gap in knowledge by developing and applying a suite of novel in vitro tools to investigate how the physical properties of the tumor microenvironment drive cancer progression. Our approach to develop and apply in vitro tools rests on three independent, but synergistic pillars. First, we established a 3D in vitro tumor model that mimics critical cell-cell and cell-ECM interactions by embedding multicellular spheroids within 3D collagen matrices. We assessed the in vivo relevance of our 3D collagen embedded spheroid model by quantifying the presence of highly malignant cancer stem cells (CSCs) before and after chemotherapeutic treatment with either paclitaxel or cisplatin. By characterizing the CSC response within two other commonly used in vitro models—a 2D monolayer and a 3D collagen model in which single cells are diffusely embedded—we found the CSC response to be model-dependent. Our results therefore highlight the need to screen potential CSC-specific chemotherapy drugs within in vitro models that recapitulate the in vivo 3D multicellular tumor architecture. Second, through integrating computational and experimental approaches, we developed a mathematical model of the transcriptional regulators—Yes-associated protein (YAP) and transcriptional co-activator with PDZ-binding motif (TAZ)—which are often overactive in late-stage cancers despite mutations within their upstream signaling pathway being rare. Here, dysregulated cytoskeletal tension and disrupted apical-basal polarity, two defining characteristics of breast cancer, have been suggested to promote overactive YAP/TAZ signaling. Therefore, we developed a computational model to study how overactive cytoskeletal tension, due to increased ECM stiffness, leads to aberrant YAP/TAZ signaling in cancer. The model revealed that simultaneous alterations in cell mechanics and cell-cell adhesion signaling synergistically converge on YAP/TAZ activity and lead to its overactivation, a process poorly understood in cancer progression. Finally, in an effort to decouple the effects of collagen fiber density and network mechanics on cancer cell behavior, we developed a highly tunable in vitro 3D interpenetrating network (IPN) platform consisting of a primary collagen network reinforced by a secondary visible-light-mediated thiol-ene PEG network. The IPN platform is cytocompatible, inherently bioactive, and mechanically tunable, which makes it a useful tool for studying mechanotransductive signaling pathways. Moreover, while this thesis work focused on in vitro applications, our approach raises the interesting possibility of altering the physical properties of the tumor microenvironment as a potential therapeutic. In summary, this work addresses the question of how the physical properties of the tumor microenvironment affect cancer progression by deploying three distinct, but complementary approaches, and suggests that addressing the physical aspects of cancer progression may improve clinical outcomes.