Novel three-dimensional organotypic liver bioreactor to directly visualize early events in metastatic progression

Metastatic seeding leads to most of the morbidity from carcinomas. However, little is known of this key event as current methods to study the cellular behaviors utilize nonrepresentative in vitro models or follow indirect subsequent developments in vivo. Therefore, we developed a system to visualize over a multiday to multiweek period the interactions between tumor cells and target organ parenchyma. We employ an ex vivo microscale perfusion culture system that provides a tissue-relevant environment to assess metastatic seeding behavior. The bioreactor recreates many features of the fluid flow, scale, and biological functionality of a hepatic parenchyma, a common site of metastatic spread for a wide range of carcinomas. As a test of this model, prostate and breast carcinoma cells were introduced. Tumor cell invasion and expansion could be observed by two-photon microscopy of red fluorescent protein (RFP)- and CellTracker-labeled carcinoma cells against a green fluorescent protein (GFP)-labeled hepatic tissue bed over a 14-day period. Tumors visible to the naked eye could be formed by day 25, without evident necrosis in the >0.3-mm tumor mass. These tumor cells failed to grow in the absence of the supporting three-dimensional (3D) hepatic microtissue, suggesting paracrine or stromal support function for the liver structure in tumor progression. Initial ultrastructural studies suggest that early during the tumor-parenchyma interactions, there are extensive interactions between and accommodations of the cancer and host cells, suggesting that the tumor-related epithelial-mesenchymal transition (EMT) reverts, at least transiently, to promote metastatic seeding. In sum, our 3D ex vivo organotypic liver tissue system presents a critical vehicle to examine tumor-host interactions during cancer metastasis and/or invasion. It also circumvents current limitations in assays to assess early events in metastasis, and provides new approaches to study molecular events during tumor progression.