The secret lives of cancer cell lines

Abstract

The extent of genetic and epigenetic diversity between and within patient tumors is being mapped in ever more detail. It is clear that cancer is an evolutionary process in which tumor cell intrinsic and extrinsic forces shape clonal selection. The pre-clinical oncology pipeline uses model systems of human cancer – including mouse models, cell lines, patient-derived organoids and patient-derived xenografts – to study tumor biology and assess the efficacy of putative therapeutic agents. Model systems cannot completely replicate the environment of human tumors and, even within the same cancer model, data are often irreproducible between laboratories. One hypothesis is that ongoing evolutionary processes remain relevant in laboratory models, leading to divergence over time. In a recent edition of Nature, Ben-David and colleagues showed that different stocks of widely used cancer cell lines – a staple of cancer research over many decades – are highly heterogeneous in terms of their genetics, transcriptomics and responses to therapies. The authors find compelling evidence of positive selection based on ongoing mutational processes and chromosomal instability. Thus, the origin, culture conditions and cumulative number of population doublings of cell lines likely influence experimental outcomes. Here, we summarize the key findings of this important study and discuss the practical implications of this work for researchers using cell lines in the laboratory.

DNA Methylation and Bladder Cancer: Where Genotype does not Predict Phenotype

“…The major discrepancies one could think of might be: inter-individual methylation variations, molecular heterogeneity within bladder tissues, treatment variability, biological age heterogeneity, genes susceptible to gain methylation, ethnic background, or the material/cell lines variability. Hynds and colleagues recently, discussed the scale of heterogeneity between strains of cancer cell lines and the accumulation of heavy mutational load over the last decades [10]. Similarly, tumor specific variable methylation levels within neighboring CpGs embedded within long interspersed nucleotide elements (LINE-1) for cancers has also been recently discussed [11].…”

An integrated landscape of protein expression in human cancer

“…Although they usually retain driver mutations, cell lines often contain 'additional' genomic aberrations not present in tumours. They also lack the tumour microenvironment interactions and can undergo divergent evolution during long-term cell culture [1][2][3] . Hence, in order to capture the patient's tumour biology, it is therefore often more desirable to study primary cells or tissue biopsies.…”

Modeling the Diversity of Epithelial Ovarian Cancer through Ten Novel Well Characterized Cell Lines Covering Multiple Subtypes of the Disease

“…In this context, patient tumor-derived cell lines are efficient, inexpensive, and easy to maintain, and thus make for an attractive human pre-clinical cancer model [ 32 , 33 ]. Ovarian cancer being highly heterogeneous, a larger number of models need to be studied to cover the heterogeneity seen in the clinic, and a wide variety of cell lines is necessary to get a representative and accurate picture of the disease.…”