Differentiation and drug response evaluation in colorectal cancer

Supervisor:  Prof Sir Walter Bodmer

Derangement of cellular differentiation is a key step in the progression of all carcinomas. It is generally the least differentiated cancers that tend to be the most aggressive. Understanding the molecular basis of differentiation and its derangement in cancers is likely to provide novel targets for the development of new treatments, as well as new approaches to diagnosis and early detection.

In a cancer it is the cancer stem cells (CSCs) that drive the growth of the cancer and that still differentiate to varying extents. In the case of colorectal cancer (CRC), the differentiation is into the three cell types (columnar, goblet, enteroendocrine) normally found in the colorectal crypt. Depending on the degree of differentiation retained by a tumour, the proportion of cancer stem cells in the tumour may vary considerably.

We use a large extensively characterised panel of more than 100 colorectal cancer (CRC) derived cell lines, together with newly derived organoid cultures from primary CRCs, for in vitro studies of the control of differentiation into absorptive enterocyte and mucus producing goblet cell types. We also use patient derived myofibroblast cell lines in co-culture with the CRC derived cell line cells to mimic the epithelial microenvironment. With these resources, we have developed a system (using colony morphology, combined with surface and other biochemical markers) for detecting and quantifying the degree of differentiation within our large panel of CRC derived cell lines. This has enabled the development of simple assays for differentiation at the in vitro cellular level. These assays can be used to test the effects on cellular differentiation and CSC growth of knocking in and knocking out any chosen gene function. Novel candidate genes for testing involvement in controlling stemness or regulating differentiation have been identified by whole-genome mRNA expression comparisons between CRC lines that retain the capacity to differentiate versus those that do not. We can also study the expression profiles of selected subsets of cells from a given cell line and do single cell genomic analysis to follow differentiation patterns in the cell lines under different culture conditions that mimic the in vivo situation. Genes that show promise in an initial siRNA knock down screen are further validated by forced expression. Following this, the best candidates can be further investigated using fluorescent reporter constructs for key genes, such as CDX1, NOTCH1, and BMI1, already known to be involved in differentiation control. For this approach to be successful it is essential to work with a sufficiently large panel of cell lines to be representative of the wide range of variation in patterns of differentiation and stem cell growth in CRCs.

It is already clear from our studies that there is a much greater variety of patterns of differentiation in CRCs than is usually assumed or detected. Our goal is to define the molecula pathways of differentiation control much more clearly than has been possible so far, and so enable a much finer classification of CRCs through identification of the particular relevant gene functions that are altered in any given cancer. This should also suggest novel targets for drug development. The classification can readily be tested on primary tumour material, enabling investigation of the associations of different patterns of differentiation with prognosis and treatment outcomes. Using our cell line panel, with this extra feature of classification, also provides the basis for a finer preclinical analysis of drug responses. For this, our focus is on immune mediated responses to monoclonal antibody derived mono- and combination therapies.

For further information please contact:  Prof Sir Walter Bodmer