More than half of biomedical researchers say that they do not bother to verify the identity of their cell lines, a survey1 suggests — even though scientists have been warned for years that many studies are undermined because the cells they use are contaminated or mislabelled.

Of the 446 survey respondents, 52% said that they did not bother to authenticate their cells (which involves checking their species, tissue-type and sex), although more performed checks for obvious signs of contamination. And of those who did make sure their cells were not mislabelled, slightly under half failed to use the gold-standard DNA-based testing method. The survey was carried out by a task force of researchers concerned about cell authentication, and co-ordinated by the non-profit Global Biological Standards Institute (GBSI) in Washington DC.

Misidentified or contaminated cell lines waste research dollars and hamper the reproducibility of research findings. But many scientists told the GBSI that cost and time constraints deterred them from testing. One-fifth said that they were unaware of the issue, and one-quarter seemed complacent, either seeing no need for testing or believing that they were “careful”, according to the survey results, which are published in the journal BioTechniques1. Most of the respondents — half of whom were senior or mid-career scientists — had great faith in their own abilities, with almost three-quarters rating themselves as “either expert or above average cell-culturists”.

Bigger problems

Little has changed in cell-line authentication and cell-culture practices in the past decade, says Leonard Freedman, the GBSI’s president, and co-author of the study. (A 2004 survey2 suggested that one-third of laboratories ran checks on cell lines.) “While support for change is strengthening, the scientific community has still not embraced cell authentication as an expected part of the research process,” says Freedman.

Freedman says that the problem of cell identity is a “microcosm of the bigger problems contributing to data irreproducibility”. He wants journals to refuse to publish papers unless authors describe how they authenticated the cell lines used in their work. 

In 2013, Nature journals asked authors to give the source of their cell line and whether it had been authenticated, but an analysis earlier this year of a sample of about 60 submitted papers found that only 10% had authenticated their cell lines.

In response, Nature and associated research journals introduced a policy in May requiring authors to check the cell lines used against a database of almost 500 known misidentified cell lines (provided by the International Cell Line Authentication Committee), and to provide details about the source and testing of the cells.

Ultimately, the problem demands that the entire biomedical community embrace the need to systematically authenticate cell lines, says Freedman. “We must commit sufficient time, resources, and expertise to adequately train and educate scientists in best practices for cell-culture experiments,” he says.