Haematopoietic stem cells make blood cells by first generating multipotent progenitors. Though they can yield all sorts of mature blood cells, the progenitors cannot sustain themselves, and so each one can produce only a finite number of cells. A looming question in stem cell biology is why haematopoietic stem cells can self-renew but their progenitors cannot. Reporting in Nature, Michael Clarke, Bolaji Akala and colleagues at Stanford University, in California, have now uncovered some of the molecular mechanisms that limit progenitors' ability to expand indefinitely1.

The researchers created triple-mutant mice lacking proteins that are repressed by Bmi1, a protein necessary to sustain adult haematopoietic stem (HS) cells, and found that the frequency of cells that were able to regenerate blood systems increased at least tenfold. The ability to experimentally create long-lived progenitors also hints at how cells could naturally become long-lived malignant cells in blood cancers.

The study focused on three proteins repressed by Bmi1: Trp53, p16Ink4a and p19Arf, of which the latter two are produced by alternative reading frames of a single gene called Cdkn2a. Individually deleting the loci encoding these proteins could not completely rescue HS cell function in mice that lacked Bmi1; however, when all three loci were deleted in genetically engineered mice, bone marrow from the triple-mutant mice was particularly efficient in reconstituting blood systems in normal mice whose natural blood-making capacity had been destroyed. The mice receiving the transplants produced every mature blood cell type, all of which appeared functional.

However, when the researchers examined the bone marrow of triple-mutant mice, they did not see many more HS cells; the triple mutant progenitors had acquired the ability to reconstitute blood. Nonetheless, several in vivo and in vitro tests showed that these progenitors were distinct from HS cells.

To understand why these progenitor cells were so potent, the researchers compared triple-mutant and wild type progenitor cells in ex vivo cultures. The triple-mutant progenitors died less often, with rates of apoptosis twofold lower than those of wild type progenitors. They were also better able to form proliferating colonies. However, more specific proliferating progenitors (myeloid progenitors and granulocyte-macrophage progenitors) were not able to reconstitute blood systems. This finding suggests that self-renewal constraints increase with differentiation and that several mechanisms regulate expansion capability in vivo. Understanding how such regulation can be dismantled at the multipotent progenitor stage can help explain both how cancers get started and how blood systems sustain themselves healthily.