Dr Philip Dunne - Phenotypic Plasticity in Colorectal Cancer

Introduction

LeQuesne John

The genomic sequencing of colorectal cancer (CRC) has identified many important oncogenic drivers, leading to the establishment of the Vogelstein genetic model for colorectal tumorigenesis. However, substantial phenotypic heterogeneity exists across and within genetically identical tumours, meaning that genetics alone cannot explain the complexity of the tumour ecosystem. To characterise the transcriptional landscape in CRC, numerous studies have developed molecular classification/subtyping systems based on gene-level data that align with genetic alterations underpinning the Vogelstein paradigm.

To move beyond these established gene-level systems, we developed a CRC classification system reflecting phenotypic landscape in CRC, based on pathway-level biological signalling. These pathway-derived subtypes (PDS) are independent of KRAS and other Vogelstein features, and reveal subtle phenotypes related to epithelial differentiation and lineage maturity, reminiscent of those proposed within Waddington’s landscape. This approach reveals how individual cells contribute to the overall phenotypic landscape observe in each human tumour, and how the selective advantage of each individual phenotype evolves during tumour development.

Using these phenotypic landscapes as the basis for biological discovery in both human tumours and genetically engineered mouse models (GEMMs), our team are using a combination of bulk, single cell and spatial transcriptomics to define a more holistic phenotypic map of the tumour landscape and cellular communication networks associated with tumour development and progression in CRC. Furthermore, while data availability has increased, molecular data rarely realises its full potential due to the programming skills required for analysis. Therefore, our team complement our biological analyses with the development of “no-programming-required” publicly available data apps for mechanistic interrogation of these cohorts. This approach facilitates new discoveries and leads to the democratisation of data analysis by removing a major bottleneck in the analysis of complex molecular profiling datasets.


Other funding: 

cruk             MRC logo                   NewAICRIlogo beige