Interactions between epithelial and mesenchymal cells are at the basis of human development and tissue renewal and repair. Similarly, tumour-stroma interactions are increasingly recognised as a key element affecting cancer onset, progression, and resistance to treatment. As an example, in the case of colorectal cancer (CRC), gene expression signatures defining a poor prognosis subgroup reflect a peculiar stromal gene expression, with abundant cancer-associated fibroblasts. Improving our knowledge about paracrine interactions between cancer cells and the tumour microenvironment is therefore likely to provide new means to develop effective therapeutic approaches.

To this aim, RNA sequencing on a large series of CRC patient-derived and cell line-derived mouse xenografts has been recently performed. Considering that in xenografts the stromal cells are contributed by the mouse host, the data have been processed to generate species-specific transcriptional profiles, obtaining for each xenograft a stromal (mouse) and a cancer cell (human) transcriptome. Notably, different CRCs give rise to xenografts with widely different stromal profiles.

These results put researcher in the optimal position to analyse the expression of ligands and receptors of different origin (stromal = mouse, neoplastic = human), to pinpoint candidate paracrine interactions and explore their occurrence and biological associations. In particular, through preliminary analyses, two novel candidate paracrine interactions associated with the accrual by cancer cells of a particularly abundant cancer-associated fibroblasts-populated stroma have been identified.
In this scenario, ARTUSI (Computational analysis and functional validation of epithelial-mesenchymal interactions supporting tumour onset and progression using patient-derived models of colorectal cancer) aims at three major objectives:

• to validate in vitro and in vivo the biological and, most interestingly, therapeutic relevance of the already identified candidate paracrine interactions;
• to perform a thorough computational analysis of the xenograft RNA-seq dataset, focused on ligand/receptor pairs and the “secretome”, to generate an interaction map depicting consistency, heterogeneity and significance of candidate paracrine interactions in CRC;
• to explore with a similar strategy the possible value of CRC xenografts in zebrafish for modelling tumour-stroma interaction, with the advantages of reduced time and effort, increased scalability and facilitated intravital imaging compared to mouse xenografts.

The project will potentially impact on the understanding of epithelial-mesenchymal interactions and CRC biology, improving diagnosis and clinical management. The quantification of tumor stromal composition at genomic level will provide rationales to design biomarkers for the stratification of disease outcome and response to treatment. Secondly, ARTUSI’s unique approach to systematic characterization of the interplay between tumor and stromal cells will provide unprecedented resolution in identifying new actionable targets.