Final Report Summary - THESEUS (Tumour Heterogeneity and Somatic Evolution of Unstable cancer genomes)
Tumours are dynamic populations of abnormal cells that undergo evolutionary change. The selection pressures that shape the population of cells in a tumour could come from the body, or could come externally, such as in the form of therapy. In time, the evolutionary forces at play would change the underlying genetic makeup of the population, and thereby alter the behaviour of the tumour. Analyses of cancers have shown that cancers can exhibit substantial intra-tumour heterogeneity (ITH) in the genetic makeup of individual cells within the tumour. The underlying genetic instability and diversity of genetic makeup in the population of cancer cells is associated with drug resistance and poor clinical outcome.
Many of the existing systems used to study cancer do not incorporate modelling of the dynamic genetic changes and show limited ITH. In this project, we investigated the genetic abnormalities in cancers that would generate and sustain genetic instability, and generated new ways of studying cancers in mice that integrate the dynamic evolving nature of cancers. We have generated novel mouse models of lung and colon cancers and we have used these strains to examine how evolutionary forces drive genetic changes in the tumour population during cancer progression. We show that these tumours exhibit dynamic changes and are more likely to be resistant to multiple cancer therapies. We observed not only intratumour heterogeneity, but also evidence of early genome doubling events. We observed intratumour heterogeneity in the form of extensive aneuploidy, while only a low number of mutations were identified in the analysed mouse lung tumour samples.
Aneuploidy and ongoing changes in chromosome number, i.e. chromosomal instability (CIN), is commonly observed in cancer. By performing a multi-sample analysis across 25 cancer types we have been able to identify that specific loss of heterozygosity events preceed genome doubling (Watkins et al in revision). To further study the degree of aneuploidy within the mouse lung tumours, we used image cytometry to identify aneuploid cells which were subsequently analyzed by single-cell shallow whole genome sequencing. Our results show that Trp53 mutant tumours have a higher variation in copy number as well as higher overall ploidy. Data from the single cell sequencing also shows that increased ploidy correlates with a higher proportion of the genome being fragmented. We are currently investigating the hypothesis that recurrent genomic copy number gains and losses might contribute to therapy resistance. This would be in line with the analysis of TRACERx clinical samples, where we have shown that lower HLA expression correlates with impaired response to immunotherapy (McGranahan, Furness, Rosenthal et al Science 2016).
Our efforts to identify genes involved in maintaining genome integrity identified SETD2 as playing a role in nucleosome stabilization and DNA repair (Kanu, Grönroos et al Oncogene 2015). In our further studies of clear cell renal cell carcinoma (ccRCC) we were able to identify recurring deterministic evolutionary pathways and show that chromosomal complexity predicts clinical phenotypes (Turajlic, Xu, Litchfield et al. Cell 2018a). We also show that the metastatic potential of ccRCC was correlated with chromosomal complexity (Turajlic, Xu, Litchfield et al. Cell 2018b). Multiple cellular mechanisms are in place to prevent genome instability and we have identified new mechanisms of tolerance to genome instability (Crockford, Zalmas, Gronroos et al., Annals of Oncology 2017, López-García, Cancer Cell 2017). This work will lead to further studies and efforts on how to eliminate or control cancer cell populations that are inherently dynamic and adaptive.
Many of the existing systems used to study cancer do not incorporate modelling of the dynamic genetic changes and show limited ITH. In this project, we investigated the genetic abnormalities in cancers that would generate and sustain genetic instability, and generated new ways of studying cancers in mice that integrate the dynamic evolving nature of cancers. We have generated novel mouse models of lung and colon cancers and we have used these strains to examine how evolutionary forces drive genetic changes in the tumour population during cancer progression. We show that these tumours exhibit dynamic changes and are more likely to be resistant to multiple cancer therapies. We observed not only intratumour heterogeneity, but also evidence of early genome doubling events. We observed intratumour heterogeneity in the form of extensive aneuploidy, while only a low number of mutations were identified in the analysed mouse lung tumour samples.
Aneuploidy and ongoing changes in chromosome number, i.e. chromosomal instability (CIN), is commonly observed in cancer. By performing a multi-sample analysis across 25 cancer types we have been able to identify that specific loss of heterozygosity events preceed genome doubling (Watkins et al in revision). To further study the degree of aneuploidy within the mouse lung tumours, we used image cytometry to identify aneuploid cells which were subsequently analyzed by single-cell shallow whole genome sequencing. Our results show that Trp53 mutant tumours have a higher variation in copy number as well as higher overall ploidy. Data from the single cell sequencing also shows that increased ploidy correlates with a higher proportion of the genome being fragmented. We are currently investigating the hypothesis that recurrent genomic copy number gains and losses might contribute to therapy resistance. This would be in line with the analysis of TRACERx clinical samples, where we have shown that lower HLA expression correlates with impaired response to immunotherapy (McGranahan, Furness, Rosenthal et al Science 2016).
Our efforts to identify genes involved in maintaining genome integrity identified SETD2 as playing a role in nucleosome stabilization and DNA repair (Kanu, Grönroos et al Oncogene 2015). In our further studies of clear cell renal cell carcinoma (ccRCC) we were able to identify recurring deterministic evolutionary pathways and show that chromosomal complexity predicts clinical phenotypes (Turajlic, Xu, Litchfield et al. Cell 2018a). We also show that the metastatic potential of ccRCC was correlated with chromosomal complexity (Turajlic, Xu, Litchfield et al. Cell 2018b). Multiple cellular mechanisms are in place to prevent genome instability and we have identified new mechanisms of tolerance to genome instability (Crockford, Zalmas, Gronroos et al., Annals of Oncology 2017, López-García, Cancer Cell 2017). This work will lead to further studies and efforts on how to eliminate or control cancer cell populations that are inherently dynamic and adaptive.