This project has focussed on delineating how pancreatic tumour cells interact with non-mutated host cells and how these interactions ultimately regulate tumour development and tumour cell function. At the outset of this proposal I hypothesised that interactions between tumour cells and cells in the microenvironment may promote tumour development.
Throughout this project we have 1) developed, benchmarked and validated new models to study interactions between tumour and host cells; 2) mapped the composition of the tumour microenvironment in both primary and metastatic tumours and identification and characterisation of a novel bona-fide tumour suppressive ‘host’ cell population; and 3) Mapped novel signalling interactions between tumour cells and cells in the host microenvironment.
1) Technological advances have enabled isolation and expansion of epithelial pancreatic cells in vitro (outside the body), known as organoids. This, in turn has provided ability for isolation and expansion of human patient samples for in vitro testing of novel therapies and for development of novel models through genetical engineering. This this project we have made significant advances in both areas.:
Firstly, we have used genetic engineering to develop a series of models with specific genetic alterations which are being used to understand how individual genes in isolation and in combination are important for the establishment of a tumour favourable microenvironment. This model has allowed us to work systematically in defining key signals and receptors that control early stages of pancreatic cancer development.
Secondly, we have developed a novel, fully controllable, in vitro model of pancreatic cancer and host cells. While the microenvironment is undeniably an important player in controlling tumour development and therapeutic response in patients, current models poorly reflect this complexity. Together with collaborators at MIT and University of Manchester, we have interrogated how tumour cells interact with and depend on signals from the tissue scaffold (extracellular matrix) and used this to develop a model that now allows us to grow host cells with tumour cells to delineate how these interactions develop to, hopefully, reveal new targets for therapies (Below et al, Chastney et al, Gough et al, Humphries et al).
2) Improved understanding of the host microenvironment depends on firstly mapping out and cataloguing components, after which functional interrogation becomes essential for future therapeutic development. Using high-dimensional analysis by mass cytometry we mapped the cellular constituency in primary tumours of pancreatic cancer and subsequently compared these interactions across primary and metastatic sites (Hutton et al, Blanco-Gomez in prep). A major novel observation from these studies was the identification of a novel, tumour suppressive fibroblast ‘host’ population. These cells are present in all normal, primary and metastatic tumours analysed thus far. Functional analysis suggests that the tumour suppressive fibroblast population may be a lineage and have fixed characteristics. Future analyses will be critical to unravel how these host cells may be used to improve therapies.
3) Emerging tumours induce a local microenvironment pliable to tumour progression, which over time also enable rapid adaptation to therapies and metastasis. Defining the identify and effects of the tumour cell signals that drive this reaction and in turn revealing which ‘host’ cell populations mediate these effects are important for future intervention strategies. Using in vitro models we have studied how tumour cell heterogeneity underpins diversity of heterocellular interactions and how these interactions engage distinct but clone selective tumour cell signals (McCarthy et al under review). Moreover, some of these cellular interactions form metabolic support network opening up a novel approach of disrupting tumours dependencies (Halbrook et al). We have also described a novel signalling circuitry which is driven by tumour cells but depends on subsequent interactions between fibroblasts and immune suppressive myeloid cells, to drive tumour growth and metastasis (Lee, Hogg et al). These studies continue to be important in defining specific heterocellular interaction networks that control tumour development, metastasis and response.
