Engineered ECM models: Opportunities to advance understanding of tumor heterogeneity

Abstract

Intratumoral heterogeneity is a negative prognostic factor for cancer and commonly attributed to microenvironment-driven genetic mutations and/or the emergence of cancer stem-like cells. How aberrant extracellular matrix (ECM) remodeling regulates the phenotypic diversity of tumor cells, however, remains poorly understood due in part to a lack of model systems that allow isolating the physicochemical heterogeneity of malignancy-associated ECM for mechanistic studies. Here, we review the compositional, microarchitectural, and mechanical hallmarks of cancer-associated ECM and highlight biomaterials and engineering approaches to recapitulate these properties for in vitro and in vivo studies. Subsequently, we describe how such engineered platforms may be explored to define the spatiotemporal dynamics through which cancer-associated ECM remodeling regulates intratumoral heterogeneity and the cancer stem-like cell phenotype. Finally, we highlight future opportunities and technological advances to further elucidate the relationship between tumor-associated ECM dynamics and intratumoral heterogeneity.

Introduction

Intratumoral heterogeneity is a hallmark of cancer and is characterized by the presence of different cancer cell subpopulations that severely limit patient outcomes because of their varied proliferative, invasive, and therapy resistance capabilities [1,2]. Historically, tumor cell heterogeneity has been attributed to oncogenic mutations that increase cell fitness in response to environmental pressures or chemotherapy [3]. However, phenotypic differences caused by epigenetic reprogramming as well as transient changes in gene expression, phosphoproteomics, and metabolic signaling are equally important [2,4,5]. Moreover, the self-renewal and therapy resistance of cancer stem-like cells (CSCs) contribute to clonal diversity within tumors [6]. Indeed, an increase in CSCs because of transformation of tumor cells or environmental selection pressures promotes tumor development, metastasis, and treatment response [6]. Which role the tumor microenvironment (TME) plays in the emergence of CSCs and which effect this has on tumor heterogeneity are not well understood.

Within the TME, cancer cell phenotypes are regulated through crosstalk with tissue-resident stromal cells including cancer-associated fibroblasts (CAFs), adipocytes, endothelial cells, and infiltrating immune cells [7]. Much emphasis has been placed on how secretory functions of these cells control tumor heterogeneity and progression. Yet, their impact on the physical properties of the TME may be similarly critical [8]. In particular, CAFs are widely studied for their role in changing the quantity, biochemical composition, and mechanical properties of extracellular matrix (ECM) in tumors [8], and these alterations regulate the genotype and phenotype of tumor cells as well as CSC quantity and functions [9]. Nevertheless, CAF-dependent ECM changes are not homogeneous, but are subject to spatial and temporal variations (Figure 1). How ECM heterogeneity is functionally linked to tumor heterogeneity remains unclear due in part to the lack of relevant model systems.

Both in vivo and in vitro studies have advanced our understanding of how tumor cell interactions with the ECM affect tumor progression and therapy response. However, the high cost and species-dependent differences between humans and mouse models, as well as shortcomings associated with 2-D cell culture, make it challenging to isolate mechanistic links between ECM remodeling and tumor cell state. Engineered model systems can recapitulate and isolate TME-associated ECM changes to probe their effect on tumor heterogeneity as a function of CSC enrichment. Indeed, simply switching tumor cell culture from conventional 2-D to 3-D culture impacts several hallmarks of malignancy, including cellular metabolism [10], invasion [11], and therapy resistance [12]. Furthermore, 3D culturing of cancer cells enriches for CSCs in part through activation of the epithelial-to-mesenchymal transition and altering cytokine secretion [13]. Here, we will summarize current knowledge of ECM changes in the TME, highlight model systems to mimic these changes for mechanistic studies, and outline strategies to further improve the impact engineered ECM models have on our understanding of tumor heterogeneity and the role of CSCs in this process (Figure 1).