How viable is cell culture for a model? - (Mar/17/2013 )
I'm sure many of you on here have seen this paper:
If adding serum to culture media causes cancer cells to terminally differentiate into lines that are completely different from their in vivo human counterparts, why do we still add serum then to media to grow cancer cells? This has been known since 2006, so why do we still continue to do it if the accuracy of such models is completely off from what would be expected to occur in vivo?
Are there more studies out there that have looked at whole transcriptomes of other commonly used cell lines that are purchased from vendors that also compared those transcriptomes to cells from a primary tumor? Just curious to see how much cell culture changes the in vivo pheno/genotypes, and if it is a reason why so much science fails when it comes time to translating research into the clinic.
An interesting point to raise.
I think by using serum-free medium containing these growth factors one actually selects for the growth cancer stem cells, thereby enriching them in culture, although it mustn't be forgotten that these kind of cultures are only "enriched" for them, and do contain loads of more differentiated progeny and even some terminally differentiated cells.
I guess a lot of it is due to the great difference in availability, isn't it ? Everyone can get their hands on the established cell lines, while getting primary cells is more difficult.
And taking normal cell lines and just culturing them in EGF+/FGF+/serum- medium won't solve the problem, in my opinion. Take glioblastoma as an example - there are so many papers out nowadays where they take common cell lines like U87 for instance, grow them in said medium and make experiments with the spheroids as "glioblastoma stem cells". I think you really have to use primary cells for that.
Serum is just one problem, what about:-
21% Oxygen growing conditions ???????( tissue oxygenation)
Etc Etc Etc