Development of a Microphysiological System to Model Human Cancer Metastasis From the Colon to the Liver

We describe a novel device to mimic the metastasis of cancer cells from the colon into the liver in a human model. The colon mimic is connected to the liver model by a gravity-driven recirculating unidirectional flow of a blood surrogate and can mimic the five steps of the metastatic cascade: invasion in the colon, intravasation into the bloodstream, systemic transportation, extravasation into the liver, and colonization in the liver. The colon mimic uses established normal colon epithelial organoid cells (NL) and human umbilical vein endothelial cells (HUVEC) plated on opposite sides of a membrane. To better mimic the colon structure the NL side of the membrane is exposed to air to establish an air-liquid interface. The liver mimic consists of human liver sinusoidal endothelial cells (HHSEC) and epithelial hepatic cells (HepG2 C3A) plated in Matrigel on opposite sides of a membrane. Labeled colorectal cancer cells/clusters (CA) from organoids are introduced into an established normal colon epithelial cell (NL) layer from the same patient before assembly of the system or alternatively NL organoids and fluorescently labeled CA organoids from the same patient were prepared as a ratio of 10:1 NL:CA and established together before assembly of the system. Cell viability is greater than 85% in this system. We demonstrate that over 5 days of operation that the five steps of the metastatic cascade are replicated. This novel device allows an in vitro estimate of metastatic capability (as measured by using percentages of the labeled areas per device through ImageJ) in response to selected variables. In this study, the metastatic capability depends on the source of cancer cells (e.g., the patient), the clumping of cancer cells, glucose concentration, and oxygen levels (hypoxia). For the first time, this new in vitro system mimics all five steps of the metastatic cascade in a single device and provides a new device to probe and observe the process of metastasis in a human-based model in only 5 days. The rapid observation is due to the use of a high concentration of cancer cells in the colon (e.g. 10%) and the absence of the immune system. Our device makes it possible to probe aspects of each step of metastasis and interactions between steps.