Sir

Your Editorial 'Stem-cell clarity' (Nature 459, 615–616; 2009) calls for reason in deliberations by the US National Institutes of Health (NIH) on public comments about proposed NIH guidelines for stem-cell research. We agree that rules barring the use of the 21 previously approved human embryonic stem-cell lines at the US National Stem Cell Bank (NSCB) are bad policy, and stand in the way of scientific progress. We trust that the NIH will permit use of approved and other lines derived by the US National Academies' standards.

The draft guidelines reignited a contentious issue that dominated the Bush years — the need for a diversity of lines for research and therapeutic use. In 2001, the NIH announced 78 registered lines. In the end, only 21 were available. Our research using data on cell-line shipments indicates that fewer federally approved lines are available than is generally believed, and only a handful of those are commonly used (J. McC., J.O.-S. and C.T.S. Cell Stem Cell 4, 107–110; 2009).

No more than 18 cell lines have ever been available from NSCB. The 1,217 unique requests for materials between 1998 and 2008 are quite clustered. Only three lines — H1, H7 and H9 — routinely made it to the bench. These three have quickly become experimental standards. Preliminary results from a systematic search of 534 peer-reviewed publications using human embryonic stem-cell lines during the same period show that H9 was used in 83.3% of these, H1 in 60.9% and H7 in 24.2%, and that 68.2% used one or more in combination with other NSCB lines. Also, derivatives of H9 will be transplanted in the world's first clinical trial for spinal injury.

Over time, researchers routinely discard ill-behaved or less useful lines. But our data raise serious questions about how best to calibrate the number and variety of a controversial research tool that is in short supply. Now that the NIH has made a tentative step towards a coherent national policy, the agency should take a pragmatic approach to the dual problems of diversity and fragmented ethical oversight that are the Bush administration's legacy. The field needs a repository containing an adequate number of high-quality, ethically derived, diverse lines. We must work to ensure availability and widespread use of diverse materials.

Every effort should be made to investigate the provenance of existing lines carefully and expeditiously. Local oversight committees must individually grapple with this time-consuming and exacting task; currently, they often repeat each other's efforts. To avoid redundancy and to speed the accession of new lines, we suggest the creation of a national clearing house specifically charged with certifying provenance. If there are indeed 700 lines worldwide (as the NIH hopes), then this is a necessary step to build a rich and diverse resource that can be used widely and efficiently.