Cancer Dynamics under a Chemotherapeutic Stress Gradient Using a Microfluidic in vitro Tumor Environment

Abstract

In this dissertation, we use the microfluidic cancer-on-chip system we have developed to explore cancer population dynamics and how cancer acquires drug resistance. The microfluidic cell culture device, the “evolution accelerator” (EA), generates an in vitro landscape of stress heterogeneity across a tumor population. The system allows for high-magnification real-time observations of different cancer cell lines and downstream analysis of cell phenotype as a function of position on the stress landscape. With the EA technology, we investigate the adaptation and evolution dynamics in prostate cancer cell metapopulations under a stress landscape of a chemotherapeutic drug (docetaxel). High-resolution time-lapse scanning provides abundant information about the change in cell morphology, population dynamics, cell motility and cell migration over time on a cellular level.

We further implement this technology to study quantitatively the emergence of polyploid, mesenchymal and stem-like cancer cells in the context of complex heterogeneous yet controllable in vitro environments with a spatially-varying drug concentration. Within our microfluidic stress landscape, we observe: (1) a previously-unobserved surprisingly large number of polyploid giant cancer cells (PGCCs) which emerged in a highly stressful region in response to chemotherapy; (2) the transition of the epithelial to the mesenchymal state; (3) the stem-like characteristics of PGCCs. We argue that the elevated emergence of PGCCs in a high drug environment is due to migration of diploid epithelial cells from regions of low drug concentration, where they proliferate, to regions of high drug concentration, where they rapidly convert to PGCCs. The coexistence of the emerging drug-resistance PGCCs and the altruistic proliferative diploid cells may serve as a survival strategy for the cancer population. This suggests the clinical value of identifying vulnerabilities of PGCCs that might be considered critical targets.

Finally, we present a microfluidic device, the static diffuser, which, unlike earlier work using continuous flows, generates a long-term chemical gradient within tumor microenvironment based on a static diffusion mechanism. Due to the simplicity of the experimental setup, the system allows not only well-controlled continuous microscopic studies of the interaction among various cell types, but also parallel experimentation for up to 18X time-resolved downstream cellular assays. As a proof of concept, we report the co-culture of human bone marrow stromal cell line (HS-5) and bone-metastatic prostate cancer cell line (PC3) using the static diffuser.

Taken together, the experimental platform and cancer studies presented in this dissertation show the power of sophisticated in vitro environments to enable the discovery of new pathways and mechanisms underlying evolution of drug resistance in cancer.