The human tumour micro-environment modelled in in vitro biomatrices and applied to cancer drug discovery

Lead Research Organisation: University of Nottingham
Department Name: School of Life Sciences


New anti-cancer drugs are being developed to target the interactions between malignant cells and normal human cells, such as fibroblasts, that are recruited to support the tumour. Human tumour cells can be grown in rodents but are supported by rodent fibroblasts. The cross-talk between human and rodent cells may not be optimal to evaluate new drugs, limiting the relevance of using xenograft models for such drug screening. We plan to model this environment in vitro, initially using cells derived from patient specimens grown within a matrix to act as a support. The tumour type we will focus on is colorectal cancer in which cells spread to the liver. Growth factors produced by liver fibroblasts are known to support the growth of these invading colorectal tumour cells and we will investigate whether this pathway is functional in our systems as a proof of concept. The cell interactions observed in the biomatrix system will be compared to those present in freshly isolated liver metastases from patients with colorectal cancer. We will also investigate the possibility of using more readily available human fibroblast-like cells to broaden the applications of the technology to different types of cancer and apply reporter systems which are currently being developed by our group to allow real-time monitoring of cell growth and environmental signals without having to harvest the cells.
Once validated the findings/protocols will be published and, through the applicants? extensive collaborative links with the pharmaceutical industry, will be disseminated to spread good practice and maximise their uptake and use, overall resulting in reduced inappropriate use of animal models.

Technical Summary

The paracrine signalling pathways within a tumour micro-environment play an important role in epithelial: mesenchymal transition (EMT) and metastasis. Cancer-associated fibroblasts (CAF), in particular, have important roles in supporting these processes. Stromal:epithelial interactions have become a focus of cancer drug discovery but are not optimally modelled by xenograft systems. For example, transplanted human epithelial cells are infiltrated with murine stroma but within the hepatocyte growth factor (HGF)/C-met axis, an important paracrine signalling pathway, murine HGF from the stroma does not bind the human C-met receptor on human epithelial cells within xenografts.
Therefore to replace the inappropriate use of xenografts to examine these paracrine interactions in drug discovery applications, we will model the tumour microenvironment in vitro, thereby aligning with the objectives of the NC3Rs.
To test the utility of this approach we will model a colon liver metastasis niche using early-passage patient-derived epithelial cells and matched CAFs in a biomatrix that has been shown to support multiple cell types in co-culture, using a functional HGF/C-met pathway as a measure of success. We will apply the refinement of real-time fluorescent/bioluminescent reporters to assess expansion of the different cell-types within the culture and monitor environmental signals such as hypoxia, EMT and apoptotic potential concurrently, without the need to harvest the cells. These will be validated against fresh human colorectal liver metastasis samples. Furthermore replacement of CAFs with commercially-available human mesenchymal stem cells (MSCs) will also be assessed as use of these should enable broader application of this in vitro system to cells derived from different tumour types. The final assessment of drug response will be performed in a 96-well format to examine the utility of this in vitro approach to a higher through-put format maximising its uptake by the pharmaceutical industry in drug discovery applications.
This refined in vitro system therefore has the potential to replace the need for animals in investigating the biology of the tumour micro-environment for both new target identification and lead optimisation.


10 25 50