Compartmentalized Hydrogel Assay as a tool for assessing tumor-stroma interactions

Elisa D'Arcangelo 1 ; Alison McGuigan 1, 2;

 1.  Institute of Biomaterials and Biomedical Engineering, University of Toronto; 2.  Department of Chemical Engineering and Applied Chemistry, University of Toronto

 In addition to a rapidly proliferating mass of cancer cells, tumor tissue also hosts a variety of stromal cells, which are essential contributors to the growth and metastatic potential of tumors. Stromal cells function as sources of growth factors, metabolic precursors and matrix-remodelling enzymes, are actively recruited to the tumor site and, in the specific case of fibroblasts, transition to an 'activated' state induced by cancer-cell signalling. Upon treatment of the primary cancer cell mass, an activated stroma is left behind, which potentially supports any remaining tumor cells in reaching a proliferative state, leaving patients vulnerable to disease recurrence. In vitro assays to identify mechanisms to revert or reduce the state of tumor-induced stromal activation are required and of particular interest are assays that offer direct insights in to how stromal modulation affects cancer cell invasion.


Activated, cancer-associated fibroblasts (CAFs), are observed in vivo at the invasive edge of the tumor. We have designed a culture device that enables microfluidic molding of 2-compartment hydrogels: each compartment is laden with cancer cells or CAFs, and the two cell populations contact each other along a central compartment edge. We have observed strand- like projections of tumor cells invading into the adjacent, CAF-occupied compartment, suggesting that invasion requires immediate contact with fibroblasts/fibroblast-remodeled hydrogel. We have confirmed that soluble signals produced by CAFs were not sufficient to elicit an invasive change in the cancer cell population, while CAF-mediated hydrogel remodelling resulted in matrix topographies conducive of invasion. CAF-remodelled matrix induces migration of normal epithelial cells, while matrix remodelled by non-activated (quiescent) fibroblasts did not. Additionally, we were able to modulate the extent of cancer cell invasion by treating co-cultures with macrophage-conditioned media. Together, these data suggest that our in vitro platform can model an activated stromal environment, whereby the activation status defines the extent of tumor cell invasion.

The chronic activation of CAFs is known to be epigenetically regulated; we therefore assessed invasiveness after stromal treatment with epigenetic enzyme inhibitors (40 probes provided by the Structural Genomics Consortium) to uncover the epigenetic components necessary for imposing a tumor invasion-permissive stroma at the tumor edge and will present potential targets for in vivo stromal re-programming after treatment of the primary tumor.

Overall, this novel platform that is applicable to the exploration of mechanisms that drive activation-dependent modulation of tumour cell phenotype. It proves to be a valuable functional assay of stromal activation and a tool for cell interaction and pathway discovery.