Scientists Have Conducted Decades of Research on Mislabeled Cell Lines

When I was a student working in a cancer research lab, I ate, breathed and slept cell culture. This is the art of growing tissue (in my case, human skin cells) on the bottoms of plastic petri dishes, where they are kept alive in a specially-concocted brew of nutrients and growth factors.

I became obsessed with the technique because I had to: many types of human tissue cells are fickle to grow and require near-constant attention. They can also easily become contaminated. Because tissue culture cells grow at human body temperature, invasion by bacteria was a constant annoyance. To avoid this, I constantly bathed my hands and forearms in disinfecting isopropyl alcohol. All the work was done inside a ventilation device called a fumehood. In short: you have to keep things very clean.

Despite my fastidious efforts, many experiments were ruined, either because the tissue culture plate became contaminated with bacteria or mould, or because a rogue cell from another experiment slipped into the confines of the petri dish. Because cells grow exponentially in culture, any cells of dubious provenance can easily take over the whole dish. Suddenly, you're dealing with a whole new type of culture, without even knowing it.

Such mistakes may seem like schoolboy errors, but it turns out that having the wrong kind of cells in your culture dish is distressingly common. In August, the journal Science Translational Medicine released a paper showing that a line of brain cancer cells that's been a stalwart of biological research for more than 50 years does not consist of the type of brain tissue they thought it was made from—it's a completely different kind of brain cell.

This is a huge problem, because different types of cells can react differently to the various drugs and chemicals they're used to test. Experts worried it could invalidate results.

And this isn't the first time a problem like this has been identified. "[A] substantial proportion of cell lines is mislabeled or replaced by cell lines derived from a different individual, tissue or species," notes a 2010 report, which states that this glaring problem has been known to the research community since 1950. Science is already being wracked by a reproducibility crisis, in which published and peer-reviewed results just can't be replicated. A recent survey by more than 1,500 researchers in Nature, for example, revealed that more than 70 per cent tried and failed to reproduce another scientist's experiments.

Read More: A Huge Study Found Less than Half of Psychology Findings Were Reproducible

To understand the scope of this, you need to step back and look at how these cell lines are made. They're usually taken from a tumour, then replicated over and over, sometimes for decades. They are workhorses of biological research. Many of them are part of a library of cell cultures called the American Type Culture Collection (ATCC). They are used as test beds for everything from cancer drugs to new cosmetics.

"[Misidentification of cell lines] is a somewhat common thing," Vuk Stambolic, a professor in the Department of Medical Biophysics at the University of Toronto, who uses cell lines extensively in his research, told me in an interview. "There is a famous 'triple-negative breast cancer' cell line MDA-MB-435 that has been in the literature forever, that turns out to be a melanoma [skin cancer] line." Problems with MDA-MB-435 have cropped up since 2000.

Unwilling to give up on this very necessary form of research, scientists are now hunting for new methods to maintain the integrity of a cell line. Recently, the standard is SNP (single nucleotide polymorphism) genotyping: a genetic fingerprinting technique that looks for unique, single-letter differences in the genetic code as they exist between cell types to provide a reliable identification system for their origins.

That technique is also serving to underline the scope of the problem. "Only with SNP genotyping are we beginning to see the gravity of these errors," said Stambolic.

A consortium uses SNP genotyping to monitor the quality of cell lines: the International Cell Line Authentication Committee, or ICLAC. It hosts a database that lists more than 400 misidentified lines. But this is a voluntary organization and can't hope to genetically fingerprint each cell line in existence.

The solution may lie in self-regulation. Since 2013, some Nature journals require verification of a cell line's identity through DNA fingerprinting before research results based on work done on them can be published.

"People don't know," said Stephane Angers, a professor in the Faculty of Pharmacy at the University of Toronto. Angers uses cell lines to study how cells communicate with one another. "What people call a particular cell line in their lab is often very different from another lab. Now there's more and more guidelines about running a few basic tests to validate what people are working with. We just got asked that for one of our papers. They asked that we provide identification for each cell line we used."

Although cells grown in dishes continue to be used extensively in biological research, their days are likely numbered. The advent of personalized medicine means that, increasingly, doctors and scientists will be performing research on a patient's own cells, instead of anonymous (and possibly mislabeled) cell lines that have been the standard for decades.

"While we are by no means abandoning cell lines," said Stambolic, "most labs, including mine, are growing tissue directly out of patient biopsies and surgical specimens in the form of organoids [miniature, lab-grown organs] and find them more representative of the cancer problem, albeit much harder to manipulate than cell lines. This is an evolution of the cell line idea, and deeply rooted in the tremendous work done in cell lines so far."

Want more Motherboard in your life? Then sign up for our daily newsletter.