Novel Evidence That a Quiescent Murine Population of Bone Marrow (BM)-Residing, Developmentally Early, Very Small Sca-1⁺Lin^–CD45^– Cells Is Highly Responsive to Prolonged Bleeding by in Vivo Proliferation and Differentiation Into CD45⁺ Hematopoietic Stem/Progenitor Cells (HSPCs)

Several phenotypes have been proposed for long-term repopulating hematopoietic stem cells (LT-HSCs) in murine bone marrow (BM). However, evidence from our and other laboratories has accumulated that adult murine tissues contain a population of developmentally early, so-called very small embryonic-like stem cells (VSELs), which we have proposed as playing an important role as precursors of LT-HSCs (Exp. Hematology 2011;39:225, Leukemia 2012;25,1278). As we reported, these cells are kept quiescent in the BM microenvironment by erasure of the somatic imprint in differentially methylated regions (DMRs) of some developmentally crucial, paternally imprinted genes (Igf2-H19, RasGRF1, and p57^Kip2), which proper expression is required for proliferation and expansion of pluripotent stem cells (e.g., embryonic stem cells) (Leukemia 2009;23:2042). However, we also demonstrated that these cells may be specified into the hematopoietic lineage in vitro in co-cultures over OP9 stromal cells.

We hypothesized that these very small cells, which can be specified into the hematopoietic lineage ex vivo in an “artificial” OP9 microenvironment, should also be able to become specified into HSPCs in vivo in the normal BM microenvironment in situations of hematopoietic stress that promote the formation of new HSPCs.

Normal C57Bl6 mice were bled (twice a week, 200 μl/bleeding) for 4 weeks, and by the end of each week were injected with bromodeoxyuridine (BrdU) to label cells that are in the cell cycle. These mice were subsequently sacrificed and BM cells, flushed from BM cavities as well as from crushed/collagenase-treated bones to recover cells associated with endosteal niches, were obtained from both control and bled mice. In these cell suspensions, we measured i) the total number of Sca-1⁺Lin^–CD45⁺ HSPCs and small Sca-1⁺Lin^–CD45^– VSELs by FACS and ii) the number of cycling BrdU⁺ HSPCs and VSELs. Moreover, by employing RQ-PCR, we measured the expression of genes regulating the early stages of hematopoiesis and imprinted genes that keep VSELs quiescent in the cell cycle. We also tested the ability of VSELs from control and bled mice to differentiate into CD45⁺ HSPCs in OP9 co-cultures and their ability to reconstitute hematopoiesis in lethally irradiated mice.

We observed that the number of cycling BrdU⁺ VSELs increased from ∼1 ± 0.03% (control) to ∼26 ± 4% and ∼32 ± 6% among BM cells derived from flushed and crushed bones, respectively. Furthermore, in comparison with control animals, BM VSELs isolated from mice after chronic bleeding expressed lower levels of pluripotency markers such as Oct-4 and Nanog, upregulated expression of pro-proliferative mRNA whose expression is regulated by paternal imprinting (Igf2, IGF-1R, and RasGRF1), and downregulated expression of mRNA for paternally imprinted, proliferation-inhibiting H19 and p57^Kip2genes. At the same time, the number of BM HSPCs increased from ∼17 ± 3% to 35 ± 7% and 1 ± 0.02% to 40 ± 5% in flushed and crushed bone-derived cells, respectively. Most importantly, we observed that VSELs isolated from bled mice highly upregulated the expression of genes involved in early stages of hematopoiesis, including Ikaros, Lmo2, GATA-2, HoxB4, PU.1, Scl and c-myb, and this correlated with their accelerated ability to become specified into CD45⁺ HSPCs in co-cultures over OP9 stroma. Finally, VSEL-derived CD45⁺ HSPCs, when isolated from OP9 cultures, grew methylocelulose colonies from all major hematopoietic lineages and were able to reconstitute hematopoiesis in lethally irradiated recipients.

Our data, obtained in an in vivo murine model of hematopoietic stress from chronic bleeding, strongly support the notion that developmentally early murine Sca-1⁺Lin^–CD45^– VSELs represent a population of quiescent stem cells in BM that become specified into the hematopoietic lineage in vivo. We propose that, in order to establish the relationship of these cells to other LT-HSC phenotypes described in BM as well as to construct a complete developmental hierarchy, their hematopoietic potential should be compared side-by-side with other BM-derived stem cells isolated using different phenotypic criteria.

Ratajczak:Neostem Inc: Member of SAB Other.