Different Copy Number Alteration Results from Fresh Frozen Tissue Sample and matched Cell Line

Question

[berry]

Hi everyone,

I have whole genome sequencing data and I am trying to analyse copy number alterations. My samples are fresh frozen tissue samples and I have some matched cell lines from the same tumor tissue. However, sometimes, the results from the cell cultures don't match with the results of fresh frozen.

One of the results is super extreme, I have a male patient and in cell line I have 3 copies of X chromosomes.

Other than this, sometimes in cell cultures I have additional deletions or amplifications. Do you have any idea what can cause this results? What would you assume? Should I elevate the abnormal cell line results?

Cheers!

Answer 1

[i.sudbery]

There are two things going on here:

The prosaic thing is that no tools is absolutely accurate and two samples from the same person will always produce slightly different results.

The deeper truth, however, is that cells from human bodies, even tumour cells, were never meant to survive outside the human body. Thus in order to be immortalised and live in dishes, they undergo changes to survive. Some of the changes you are seeing will be driving this immortalisation process, however, in order do make these changes, cells often first have to mutate all sorts of systems that are supposed to prevent the sort of genomic instability that this process requires. Once disabled all sorts of other changes are free to take places.

Even if these are not immortalised cell lines, cancer often disables the checks against genome instability, allowing changes to occur willy-nilly.

Finally, as cancer is charactorised by genomic instability, the original sample is likely to be heterogeneous - that is each cell will have a different panel of CNVs. Your frozen sample is like reading an average of all these. However, when a small sample is taken and propogated as a cell line, a different combination of these is likely to come to the fore, some might be selected for (see above), but for others, its just a process of drift, like a population going through a bottleneck.

Thus, TL;DR - you should not expect a cell line and a frozen sample from the sample patient to have the same CNVs.

[berry]

Thank you for your time and comment!