Abstract
Metastatic cancer, characterized by the growth and invasion of abnormal cells in several organs beyond a tumor’s point of origin, is the leading cause of cancer-related death worldwide. The migration mechanisms in cancer cells involve a complex cascade of events that are triggered primarily by interactions with various biophysical and biochemical cues within the microenvironmental niche. In fact, the secondary site of cancer colonization has long been understood as “non-random choice” as it is dependent on the cells’ interactions with each of these cues. For decades, experimental research on tumor progression has been conducted using live-cell imaging analysis with 2D and 3D in vitro tumor mimicking models. Still, it has remained difficult to systematically analyze the individual roles of each of these signaling factors due to the technological challenges behind the acquisition of physiologically relevant data in vivo and in vitro. However, modern fabrication techniques, recent advances in biomaterials, and the new generation of data analysis based on machine learning could together lead to next-generation tumor-on-a-chip platforms. This next-generation tumor-on-a-chip system could allow the significance of the tumor microenvironment and its impacts on cell migration to be studied. This thesis starts in chapter one with an introduction to past and current research on cell migration and metastasis, will focus on in vitro models and the mechanical and chemical features of the cellular microenvironment known to influence both processes. Towards developing the next generation of tumor-on-a-chip platforms, chapter two presents the design of a biomaterial scaffold that mimics the mechanics of the matrix surrounding cancerous masses. Crosslinking density of collagen fibrils around solid tumors is often enhanced and is known as a prognostic indicator of tumor malignancy via palpation of external surfaces. The xi varied mechanical properties of tumor niches are often recapitulated in cell-laden biomaterials. However, one of the major limitations of the current systems is the inability to reproduce the stiffening mechanism of natural tissues—due to increasing crosslinking densities—instead, relying on unwanted alterations of other properties, such as porosity or density to tune the stiffness of the matrix. To overcome those limitations, an artificial extracellular matrix (ECM) analog made of gelatin-based hydrogels was synthesized by precisely adjusting its degree of methacrylation. Controlling the functionalization of the matrix resulted in independent control over its elasticity, thus recapitulating the stiffening mechanism in tumors found in vivo. The new biomimetic model was used to produce a spatiotemporal cell migratory analysis that unraveled the reciprocal relationship between tumor invasion and ECM mechanics in breast cancer. Following the successful reproduction of mechanical characteristics of the ECM, the addition of chemical maps to the system is presented in chapter three, achieved by housing the mechanically tunable biomimetic ECM within an environment capable of generating gradients of chemical factors. A wide array of growth factors and chemokines diffuse concurrently through the tumor matrix, as cancer cells communicate with surrounding stromal cells using chemical signals. This chemical crosstalk is considered crucial during the invasion process. Understanding this communication requires models that accurately capture and mimic its complexity, enabling the deciphering of the chemical sources triggering cancer metastasis. Here, additive manufacturing was utilized to fabricate a tumor-on-chip platform to delineate the effects of single and competing chemokines diffusing passively in a metastatic tumor stroma microenvironment. This platform was used to capture the “decision making” process in single cells, leading to directional bias in cellular migration.