Cancer Is Genetically Diverse, But Cancer Cell Lines Are Not
By Anette Breindl
BioWorld Today Science Editor
Cancer is famous (and difficult to treat) because of the dizzying variation of genetic alterations that can cause it – tumors can acquire tens of thousands of mutations over the course of an individual patient's disease.
Unfortunately, the cancer cell lines that are still one of the main research tools to study those genes do not illustrate the same pattern of gene expression as the cancers from which they are derived. That's the conclusion of a study published this week by scientists from the National Cancer Institute.
The experiments, which were published in the Oct. 31, 2011, issue of the Proceedings of the National Academy of Sciences, began as studies of cancer cell drug resistance, which is the main focus of senior author Michael Gottesman's work.
Gottesman said it is not clear whether the effect is specific to drug resistance genes or not, but the most likely answer is "sort of."
"I'm not saying this is true for all genes," Gottesman told BioWorld Today. The conditions of cell culture probably have the strongest effect on genes that affect the ability to grow and divide, and to do so under a broad range of environmental conditions, since conditions in a culture dish are quite different from those inside a tumor. Still, "it's probably a more general problem, but we don't have the data to be sure."
The standard tools to study such drug resistance have long been cultured cell lines and microarrays, Gottesman said. "We, and many others, have published literally hundreds of genes whose expression changes in culture in response to drug treatment."
For the work now published in PNAS, Gottesman and his team took advantage of a recently developed method, Taqman-based quantitative RT-PCR, which can give a quantitative estimate of how much gene expression changes under different circumstances.
Microarrays are fairly imprecise overall – able to tell whether there is an increase or decrease in the amount of mRNA, and to some extent, whether that increase is large or small. But the new method is "a much more sensitive and accurate way of measuring RNA."
"We began to look at human cancers to see whether the genes we had identified in culture were important in primary samples," Gottesman said. And the results were sobering.
Gottesman and his team compared the changes in expression levels of nearly 400 drug resistance genes in more than 70 different cancer cell lines and primary samples from five different cancers. In all cases, "the cell lines look much more like each other than they look like the tumors from which they were derived," to a degree that "it would be hard to extrapolate" what's going on in tumors from looking at cultured cell lines.
Gottesman said he expects two reactions to his data. "Some people will be in denial," he said – or, if not in outright denial, in something of a rearguard action, saying that the problem exists only if the goal is to directly apply findings from cell culture to clinical practice.
But a more productive reaction, in Gottesman's view, is to use the information to come up with a better set of tools to study drug resistance.
As such, the findings are "good in the sense that they will allow us to retool, to build better systems. . . . It's an open-ended search."
Possibilities include xenografts of primary cells, that is, tumor cells from biopsy samples, not from cultured cell lines, or more realistic culture conditions. There is also an ongoing effort to develop genetically engineered mouse models, or GEMMs, that are more realistic models of how cancers develop and evolve over time.
No matter what the solution is, it is clear that whatever replaces cultured cell lines will be less convenient. And growing primary cells, whether in better cultures or xenografts, may have some of the same problems associated with it that the current culture systems do.
One extreme possibility is that the mere fact that a cell is able to grow in culture shows that "you may have already selected for the mutations you don't want."
On the other hand, at some point such considerations come up against the realities of research. "We do need an in vitro system," Gottesman said. Studying drug resistance mechanisms is "just too difficult to do clinically."
Until a better system is developed, "we have to be very careful extrapolating and, when we study cultured cells, realize that they are a very artificial system."
Outside of the US
In the U.S. and Canada: +1-800-336-4474
Outside the U.S.: +44-203-684-1796
Hours: Monday - Friday, 8:00am - 6:00 pm EST