In this issue of Blood, Lux and colleagues and Ghiaur and colleagues shed new light on the controversy of whether hematopoietic stem cells (HSCs) arise in the yolk sac (YS) or the aorta-gonad-mesonephros (AGM) region.
Early evidence indicated that HSCs, like gonadal stem cells and melanoblast precursors, arise in the YS and then migrate into the developing embryo.1 Contrary later evidence, initially from chicken-quail chimeras, appeared to show that YS migrants in the fetal liver (FL) are transient and that persisting definitive HSCs are of AGM origin.2 The importance of the AGM region was supported by evidence that it is the special site of induction of regulatory networks of factors necessary for the development of HSCs, including those coded by genes such as Notch1, SCL, GATA-2, Runx1, and Fli1.3 As strong as this evidence is, doubts have persisted regarding the YS-AGM relationship.
Two studies using mutant mice have reopened this question. Lux and colleagues studied Ncx1−/− mice that lacked a beating heart and circulation. Before death at embryonic day 11, YS hematopoiesis was normal but the body (AGM/FL) lacked hematopoietic cells. They concluded that all definitive HSCs are derived from the YS. A similar conclusion was reached by Ghiaur and colleagues based on studies of Rac1−/− mice, whose cells exhibit defective migration to hematopoietic sites. Again, hematopoiesis was normal in the YS but none was present in the embryo.
It could be argued in both cases that HSCs of any origin would be unlikely to develop or seed in a dying embryo. If this is not an issue, neither study actually documents the migration of YS cells to the FL. The studies merely show that YS-cell migration is necessary for FL hematopoiesis. An explanation that would be compatible with all of the conflicting information is to propose that HSCs do arise in the YS, but to become definitive, they must first migrate to the AGM region, where local inductive influences convert the cells to definitive HSCs. In essence, the AGM region is not a maternity ward in which HSCs are born de novo, but rather is a highly specialized finishing school transforming promising YS cells into genuine HSCs. The developing HSCs in the AGM region would then be YS migrants and not cells derived from local HSC precursors.
Two views of the origin of HSCs. The left panel shows the more common view that HSCs are born in the AGM region. To incorporate the present findings of Lux and colleagues and Ghiaur and colleagues, the right panel proposes a YS origin for HSCs, but with mandatory reprogramming of these cells in the AGM region.
Two views of the origin of HSCs. The left panel shows the more common view that HSCs are born in the AGM region. To incorporate the present findings of Lux and colleagues and Ghiaur and colleagues, the right panel proposes a YS origin for HSCs, but with mandatory reprogramming of these cells in the AGM region.
Why might these developmental questions matter? If the YS is the sole origin of HSCs, then the initiation of hematopoiesis is a finite event that, after the disappearance of the YS, can never be repeated in adult life. If, on the other hand, the AGM region is the true source of HSCs, hematopoiesis has been initiated within the body and, in principle at least, might one day be able to be recapitulated in adult life. If the finishing school proposal is correct, we need to accept that the initiation of hematopoiesis, like the formation of the gonads, is a finite developmental event.
Conflict-of-interest disclosure: The author declares no competing financial interests. ■
