Under former US president George W. Bush, fewer than two dozen human embryonic stem cell lines could be studied with federal funding. That number could soon extend into the hundreds, pending ethical review by the US National Institutes of Health. However, research led by Christopher Scott of Stanford University in California shows that of the 20-odd lines available for funding, researchers have so far depended primarily on just 2 of the oldest human embryonic stem cell lines1.

Scott and colleagues collected data from two major repositories of human embryonic stem cells (hESCs). Of 1,217 requests made to the National Stem Cell Bank in Madison, Wisconsin, between March 1999 and December 2008, 1,052 were for just 3 of the approximately 17 lines available and eligible for funding; of those requests, 941, or 77%, were for just 2 lines (H1 and H9). The research didn't examine informal lab-to-lab transfers, which might show up as acknowledgements or coauthorship, says study coauthor Jason Owen-Smith of the University of Michigan. "The work we have done, though, suggests that if a lot of folks are avoiding the banks, they are still using the lines that are most requested from the banks. The pattern of concentration we report for publication is even more striking than for cell-line requests." An analysis of over 500 peer-reviewed articles on research using hESC lines and published between 1999 and 2008 found that roughly two-thirds used just the three most popular lines from the National Stem Cell Bank.

The focus on just a few lines may have been enhanced by US policy, but is not unexpected, says Peter Andrews of the University of Sheffield in the United Kingdom, who also chairs the International Stem Cell Initiative, which has funded and coordinated studies comparing dozens of hESC lines. "Many years ago when people were working on mouse embryonal carcinoma cells and teratocarcinomas, almost all the cell lines commonly used were derived from one mouse tumour," he says. There is a danger that researchers will inadvertently focus work on an outlier; however, he says, particularly in the early days of a field, it makes sense for practitioners to study just one or two members of a class in depth — looking at more is not only difficult, but variation could cause noisy results that obscure important lessons.

Nonetheless, Andrews notes that differences between lines are becoming apparent, and it's unclear if they result from genetic differences, derivation conditions or other factors. Understanding the source and scope of these differences will become more important, so Andrews expects researchers to move out to other lines as the field expands. Indeed, this may be starting to happen: the Harvard Stem Cell Institute in Cambridge, Massachusetts, supplies Harvard University–derived lines that are not yet eligible for federal funding. It received 957 requests for its 28 lines between its establishment in April 2004 and December 2008. The top two requested lines (HUES1 and HUES9) made up less than one-quarter of all requests, many of which were from researchers outside the United States.

Scott says more research is necessary to understand how government policy has influenced convergence on so few lines. Nonetheless, he says, "if researchers had hundreds of lines as a starting point rather than 18, then we might be looking at dozens of well-characterized reference cultures across varied disease and genetic backgrounds rather than just a few." The next step is to quickly establish a registry of ethically procured lines, says Scott, who also heads The Stem Cell Advisors a nonprofit consultancy. "Then science can winnow and choose and settle on the best lines for the best purposes."

Related articles

Characterising stem cells requires consortia

Stem cell research: more support brings more papers