Patient-derived Organotypic Model of Colorectal Cancer Cells for Tumor-Stromal Interactions

Introduction: Colorectal cancer is ranked as the third most common cancer. Recent studies have shown that interaction between tumor cells and the tumor microenvironment (TME) can promote cancer progression and therapeutic resistance. Cancer associated fibroblasts (CAFs) are the most abundant stromal cells in the TME. Physiological tumor models are critical to expedite the understanding of disease mechanisms and identifying effective drugs in preclinical studies. The three-dimensional (3D) organotypic model studies the interaction between cancer cells and CAFs. We aimed to establish organotypic model of primary colorectal tumor cells and CAFs and studied potential of targeting tumor-stromal interactions to reduce cancer stemness, drug resistance and growth.

Materials and

Methods: CN0375 and 817829, colorectal cancer cells derived from patients were obtained through the NIH/NCIPDMR. Human colorectal CAFs were obtained from Neuromics. Single cells of colorectal tumor cells and CAFs were seeded in 1:2 ratio in 25 µl of growth factor reduced, phenol red free Matrigel in each well of a 96-well plate. The Matrigel was polymerized for 30 minutes at 37 °C. Then, 200 µl F12 medium containing a rock inhibitor (Y-27632) was added. The culture medium was replaced every two days and the morphology and metabolic activity of organoids with/without CAFs were evaluated daily. To characterize organotypic model, immunofluorescence was done using vimentin as a fibroblast marker, actin as marker for apical side and Hoechst for nucleus. Images were captured with a confocal microscope and composite images were made in ImageJ. To identify specific mechanisms of interactions between CAFs and CRC organoids, we used a phosho-RTK array to detect activities of 49 different receptor tyrosine kinases (RTK) in CRC cells stimulated with conditioned medium of CAFs. Drug studies with oncogenic pathway inhibitors were performed and analyzed. qPCR and western blots were performed to understand mechanisms of interactions of cancer cells and CAFs in the organotypic model.

Results, Conclusions, and Discussions: In the absence of CAF, the organoid was smaller than organoids in presence of CAFs on day 3 (Fig 1a). The number of organoids based on their size and results showed that there were a greater number of organoids i.e., 50 organoids with greater size i.e., 250 µm diameters in presence of CAFs to without CAF (Fig 1b). These results suggested growth and proliferation of organoids in the presence of CAFs. From immunofluorescence of staining and 3D construction of the Z-stack of confocal images of organoids, we confirmed that CAFs surrounded the organoids closely resembling spatial distribution of these cells in solid tumors, stromal cells surround masses of cancer cells (Fig 1c). Using a phosho-RTK array we found that while EGFR, MSP (Marcophage stimulating protein) and Ephrin (Eph) were highly active in CRC cells at a basal level, CAFs predominantly activated MET on CRC cells by 1.6-fold (Fig 1d). An immunoassay to determine prominent soluble factors of CAFs showed only hepatocyte growth factor (HGF), which is the MET ligand, was abundantly present at a concentration of 6.7 ng/ml (Fig 1e) and thus, identified specific mechanism of interaction of CAFs and cancer cells. We studied effects of CAFs on the response of CRC organoids to MEK inhibition and found that CAFs shifted up the dose response curve (Fig 1f). Interactions of CAFs and cancer cells promoted resistance to MEK inhibition suggesting that targeting of CAFs-cancer cells signaling is a potential therapeutic approach. Using a combination of MEK and MET suppressed organoid growth and cancer stemness. This study offers humanized tumor models for mechanistic studies of drug resistance and personalized cancer medicine.

Acknowledgements (Optional):