Cut your finger, and hordes of platelets rush to the site of the wound and plug the leak. But outside of the body, blood banks and haematologists struggle to obtain sufficient numbers of platelets to treat diseases such as anaemia and leukaemia, because the specialized blood-clotting cells only have a shelf life of five days and cannot be stored frozen. Embryonic stem (ES) cells offer a potential platelet factory especially because the nucleus-free platelets can be irradiated before transfusion to eliminate any undifferentiated ES cells. But researchers' attempts to produce platelets in vitro have been held up by an inability to culture stem cell–derived platelets with sufficient quantity or quality in the lab. Now, a new technique for obtaining high-quality, ES cell–derived platelets may unclot some of these problems.

Hiromitsu Nakauchi, of the University of Tokyo, and his colleagues isolated a group of ES cells from mouse embryoid bodies that expressed two key markers at day six of development: c-Kit, a surface protein found on blood stem cells, and the alpha IIb integrin subunit, a platelet-specific marker. By day 12, these c-Kit+/alpha IIb+ cells had tenfold higher expression of the platelet receptor glycoprotein Ib alpha (GPIbA) gene than any other cell type, and they displayed all the transcription factors and markers indicative of platelet formation1. Over time, however, these cells started to shed GPIbA and two other glycoproteins, which impaired normal blood-clot formation. Nakauchi's team figured out that the glycoproteins were lost because of enzymes called metalloproteinases. The researchers found that adding inhibitors to block these enzymes increased GPIbA expression, improved blood-clot formation in vitro and increased the number of ES cell–derived platelets in vivo following transfusion into irradiated mice.

“It's an improvement to previously published procedures,” says Harald Langer, of the National Institutes of Health's National Cancer Institute in Bethesda, Maryland. Labs around the world, including Naukauchi's2 and that of Langer's former supervisor Meinrad Gawaz, of University Hospital Tübingen in Germany3, had obtained platelets from CD34+ progenitor blood cells before, but “it seems that with this very elegant embryonic stem cell system, it's much more efficient,” he says.

There are still a few kinks to work out, however. Although the study implicates metalloproteinases in degrading the ES cell–derived platelets, the authors did not pinpoint which exact enzymes were active, Langer notes. Further work will also be needed before the technique can be scaled up to the level necessary to be clinically useful, he adds.

Nakauchi says he's now working to extend his new technique to human ES cells, but he hasn't yet found the markers that signify human platelet progenitors. So, he plans to introduce exogenous genes, first to induce a haematopoietic cell fate and then to stimulate platelet-specific differentiation. In this way, he hopes to increase the volume of platelets he can produce. “By combining this two-step system, we should be able to increase the efficiency 100-fold,” he says. Nakauchi also has unpublished data showing successful platelet formation from both human and mouse induced pluripotent stem cells. If these attempts succeed, blood banks could one day receive a steady flow of blood-clotting cells.