Simulating human tumour heterogeneity using cancer cell line mixtures

Tumours comprise (epi)genetically and phenotypically diverse cellular subpopulations evolving over space and time. This heterogeneity can be observed as differences in morphology or immunohistology of tumour sections, gene expression levels, genomic sequence and structure, proliferative potential, or metastatic ability. In order to unravel how this heterogeneity persists, one may study the clonal structure and evolution of tumours. Understanding intra-tumor heterogeneity, or the differences amongst cells in a single tumour, is of particular importance to facilitate treatment combinations that effectively target all clinically relevant subclones. This clonality-informed approach requires identification and monitoring of clonal cell populations within a tumour. In this study I simulated tumour heterogeneity using cancer cells lines in idealised mixtures. Deep allelic measurements using next generation DNA amplicon sequencing were integrated in a computational modelling program (PyClone), to retrieve cellular prevalence’s and clonal structure within these cell line mixtures. This approach to modeling heterogeneity employed both diploid and aneuploid cell lines with genomic analysis focused on heterozygous alleles and changes in the prevalence of theses alleles when cell lines were mixed in different proportions. As a result I first identified that using NGS and PyClone modeling enables elucidation of population clonal structure as predicted from idealised mixtures of diploid cell lines. However, the aneuploidy cell line mixtures demonstrated a requirement for copy number information to be included as a prior input to clonality modeling in order to avoid misleading interpretations of cellular prevalence and clonal structure. Defining and monitoring clonal heterogeneity in patient tumours is of importance to track functionally relevant subpopulations of tumour cells, enabling oncologists to administer cocktails of therapeutic agents targeting relevant clones irrespective of their numerical abundance. This clonality-informed iii treatment approach is a promising development to tackle the growing challenge of therapy resistant subclones and thus limit cancer recurrence.Medicine, Faculty ofMedical Genetics, Department ofGraduat