Cancer and Immunogenetics Laboratory - Walter Bodmer

Single cell analysis of differentiation and progression in colorectal cancer using established cell lines and primary cultures

We now have a panel of more than 120 colorectal cancer (CRC) derived cell lines and have worked out efficient procedures for growing out cells from biopsy and resection samples from fresh primary or metastatic CRCs. The cell lines are well characterised with respect to the most common driver mutations and whole genome micro-array mRNA expression data for bulk cultures in exponential growth phase. We also have whole exome DNA sequences for a significant proportion of the cell lines. Over the last 10 years we have worked out techniques for studying the in vitro differentiation of the cell lines, especially in 3D matrigel cultures, as well as for evaluating drug responses, with a major emphasis on monoclonal antibody treatment of the cell lines. Single cell DNA and mRNA analysis would be an enormous help in working out the pathways of differentiation in the cell lines and in evaluating genetic heterogeneity. It will be much more efficient, more informative and less expensive than conventional approaches using bulk cell analysis.

Over the last 4 years my laboratory has been a partner in an EU group (Cellomatic), led by colleagues at the Danish Technical University in Copenhagen, in the development of microfluidic chip trapping of single cells and whole genome DNA amplification with a view to whole genome DNA sequencing of single cells. We are about to get an instrument for doing this more systematically, which has been developed by Unisensor (now part of Philips). There is a good possibility that this instrument could also be used for mRNA analysis. The possible advantages of this instrument are the low cost of the instrument itself and low operating costs with very simply produced chips. A part of our single cell analysis would be to compare the operation of this approach with the Fluidigm C1 and Opera machines.

The main specific interest is to explore the changes in single cell gene expression as single cancer stem cells produce colonies with the three main types of differentiated cells, enterocytes, goblet cells and enteroendocrine cells. Our knowledge of some of the major genes involved in controlling these patterns of differentiation should enable us to sort out what are the controls that lead to these switches in differentiation. Blocking differentiation is probably the most important hallmark of adenocarcinoma progression and this is especially striking for goblet cell differentiation, which seems to be a very early step in the progression of most CRCs. Understanding this process is therefore very important for the possible development of novel treatments that encourage differentiation, rather than simply target cell division. Once the patterns of differentiation have been better understood in the cell lines then the same approaches can be applied to the primary cell cultures. We are also studying the interactions between myofibroblasts, key cells in the colorectal microenvironment, and epithelial cells. There is clear evidence for mutually induced interactions based on complementary growth factors and receptors in the two cell types. The development of these interactions will be studied at the single cell level in co-cultures of myofibroblasts and cell line or primary culture derived epithelial cells, since we have good expressed markers to distinguish the two cell types without the need to separate them physically. Similar experiments could also be done with other cells such as monocytes and dendritic cells, which are also important cells in the tumour microenvironment. Another application of such single cell analysis of mixed cell cultures could be to help to sort out the patterns of immune responses of immune effector cells and their epithelial cell targets in in vitro assays of monoclonal antibodies, including bispecific antibodies, targeting CRC epithelial cells. This has become very important as it is now realised that switching on T-cell inhibitory molecules, such as PD-1 and PDL-1, on the epithelial cells is likely to be a major factor limiting the efficacy of T-cell bispecific monoclonal antibody therapies. In all these types of studies, short term culture such is enabled by the new Fluidigm Opera machine could be very useful.

We have evidence that some CRC lines, and so most probably also some primary tumours, may maintain stably two or more different clones within them. Though this phenomenon, which is different from the obvious genetic heterogeneity of evolving cancers, may mostly be driven by expression differences, chance differences in neutral genetic variants may become a valuable tool in sorting out such heterogeneity.

The most likely source of expression variation in cancers, and indeed in normal adult tissues, is differential methylation of CpG islands in promoter regions. It will surely be quite soon that techniques will be developed that enable reliable single cell DNA methylation sequencing. This will, in my view, add enormously to the power of single cell analysis for studying patterns of differentiation in cancer and in normal tissues.

Hans Clevers and his colleagues have worked out the key requirements for outgrowth of cells from normal colorectal tissue .We therefore envisage being able to confirm the results we obtain on differentiation control using CRC lines and primary cultures, on cultures of normal cell colorectal epithelial cells.

We have had preliminary discussions with Adam Mead and Neil Ashley (who was, as you know, formerly in my laboratory working on aspects of the above research) about the possibility of applying for a join research grant for the sort of single cell analysis of epithelial cell differentiation and function I have described above. We aim to start quite soon doing some initial experiments to demonstrate the value of the single cell approach for the study of epithelial cell differentiation and function.