Cellular Tracking in a Three-Dimensional Microengineered Tumor Model

Background: Cancer cell metastasis is a highly complex multi-step process involving a myriad of biophysical and biochemical parameters. The mechanistic understanding of cancer cell metastasis is largely limited due to a severe lack of physiologically relevant, three-dimensional (3D) tumor models. 3D macroscale hydrogels encapsulated with cancer cells offer immense advantages over standard two-dimensional cell culture assays as they recapitulate gradients of diffusion, biomimetic 3D morphologies, and physiological cell responses in vitro.However, the majority of previous hydrogel-based tumor models lack specific organization and cannot quantify cellular invasion through live cell tracking. In this study, we utilize gelatin methacrylate (GelMA) hydrogel and a novel, two-step photolithography technique to microengineer a tumor model with highly organized circular tumor constructs surrounded by a 3D matrix. Methods: Invasive breast cancer MDA-MB-231 cells, and non-invasive MCF10A mammary cells were encapsulated separately in GelMA hydrogel and micropatterned to form circular constructs (500 μm diameter, 100 μm height) using photolithography techniques. Subsequently, pure GelMA was patterned in between the constructs to form an area representative of the surrounding matrix. Results: In this platform, cancer cells were able to invade the surrounding matrix from the circular tumor constructs, giving it the unique ability to quantify the cells’ invasive phenotype. MDA-MB-231 cells proliferated profusely upon encapsulation and 13.89 ± 0.94% of total cells invaded the surrounding matrix by day 5 of culture. MCF10A cells, on the other hand, quickly clumped to form 3D cellular clusters, with only 1.08 ± 0.24% of the cells disseminating from the tumor construct by day 5. Our microengineering technique allowed us to independently define separate biophysical properties within the tumor constructs compared to that of the surrounding matrix. Owing to this differential, we performed live cell imaging and observed that MDA-MB-231 cells inside the tumor region had low speed and high angular directionality compared to the higher speed and low directionality of cells migrating through the surrounding matrix. Conclusion: The microengineered tumor presented herein compartmentalizes two aspects of the tumor microenvironment: a distinguished tumor and its surrounding stroma. Its ability to quantify numerous parameters of cellular invasion makes it a robust platform for fundamental biological studies, high-throughput drug testing, and personalized diagnostics.