Characterizing the Landscape of Chromothripsis in Human Cancer Cell Lines

Chromothripsis is a catastrophic event that generates genomic rearrangements within one or more chromosomes, resulting in oscillating copy number states. Recent genomic analyses of patients’ tumors reveal that this phenomenon is highly prevalent (~40%) across human cancers. To date, it is unknown whether human cancer cell lines can be used for properly modeling this genomic phenomenon.

Improving on an existing algorithm, we developed an accurate pipeline for chromothripsis detection in cell lines. We then applied this algorithm to ~1,700 human cancer cell lines and mapped their chromothripsis landscapes. Among our findings, we show that ~43% of the human cancer cell lines are chromothriptic, that in over half (51% ) of the chromothriptic cell lines >1 chromosome is affected, and that in certain tumor types (e.g., esophageal and breast cancer) the vast majority of cell lines are chromothrptic (94.5% and 82.09% respectively).

We compared the genomic properties of chromothripsis between human cancer cell lines and tumors, revealing that some features are highly conserved between tumors and cell lines, whereas others are not. For example, in both tumor samples and cell lines of prostate cancer ~50% of the samples are chromothriptic, whereas in breast cancer the prevalence of chromothripsis is drastically higher in cell lines (~82%) vs. tumors (~45%).

Harnessing the rich genomic and functional data available for human cell lines, we currently work on identifying genomic features, cellular pathways and genetic and chemical dependencies of chromothriptic cell lines. This study provides an important novel resource for cell line-based chromothripsis research.