Besides illuminating basic biology, a better handle on how the blood supply is replenished could reveal ways to treat blood cancers and other diseases. Hedgehog signalling, vital in normal development, was thought to play an important role in haematopoiesis, but two papers published recently in Cell Stem Cell question whether this is really the case.

Iannis Aifantis and his team at New York University investigated a conditional knockout mouse in which the hedgehog (Hh) effector, a protein called Smoothened (Smo), is deleted in adult mice1. In this model, deletion was triggered by chemically inducing expression of interferon-alpha. To their surprise, they detected no difference between the knockouts and controls when they measured haematopoiesis, peripheral blood count and colony formation in blood samples in vitro. The cells could also compete with wild-type cells when transplanted into nude mice, and overexpressing the protein didn't expand the haematopoietic stem cell (HSC) numbers or abilities.

"This was a big surprise", says Aifantis, revealing that he had the whole set of experiments repeated just to be sure of the results. Gary Gilliland's team at Harvard University in Boston reported similar findings in the same mouse knockout model, confirming that Hh does indeed appear to be dispensable for haematopoiesis.2

Interestingly, these two studies contradict a lot of the published literature, which suggests a role for Hh signalling in HSC regulation. "Not only did we not prove this, we actually disproved it," Aifantis continues. "It took a long time, but we knew we had something interesting."

"This is surprising because previous work using both gain- and loss-of-function approaches has suggested a critical role for Hh in the development of haematopoietic cells," says Tannishtha Reya of Duke University in Durham, North Carolina. "It's scientifically important to examine every aspect; we need to look at all the experiments in an integrated way." Earlier this year, Reya's lab showed, in a different mouse model in which Smo was deleted at an earlier stage, that blocking Hh signalling reduced the regenerative capacity of HSC. So why the difference?

In Reya's opinion, it comes down to three things: timing, context, and the method of deletion. Both Reya and Aifantis believe that because the vav-Cre model that Reya's lab used is active from embryonic stages, the timing of the deletion could be the key. Early deletion may prevent other pathways from compensating and thus may reveal defects more readily. In terms of context, the mouse model used by Gilliland and Aifantis requires interferon expression, which has been reported to activate HSCs, explains Reya. (Aifantis believes the effect of interferon is negligible.) Other differences could be due to whether the gene is deleted in germline, giving rise to animals that lack the gene entirely, or if the gene is deleted conditionally, using techniques that apply only in certain tissues or under certain circumstances.

Particularly interesting was that Aifantis's paper showed that the Smo knockout had no effect on T cell development, yet an earlier paper from his lab using a different mouse system had the opposite result. In addition, a recent germline deletion of Smo studied by yet another lab also caused a defect in T cell development3. "This was the biggest surprise", said Aifantis, because it further questions the importance of the backgrounds of the mice.

There are certainly clinical implications for these studies, agree Aifantis and Reya. Most studies, including these, demonstrate that loss of Hh signaling allows haematopoietic cells to retain at least some functional capacity. This suggests that the use of small molecule inhibitors of Smo may be valuable in the treatment of leukaemias, in which Hh plays a role, because they will be able to kill cancer cells without severely affecting normal stem cells.

These papers certainly leave a lot of questions unanswered. It will be important to look at early haematopoiesis and use alternative mouse models to gain a fuller picture of what is really going on, says Reya. Aifantis agrees, adding that "we may be looking for answers before we know if they even exist".

Related articles

Self-renewing blood and leukaemia cells need hedgehog