Prospective Isolation of the Earliest Common Myeloid Progenitor in Adult Hematopoiesis

Abstract 768

In murine hematopoiesis, hematopoietic stem cells (HSCs) and multi-potent progenitors (MPPs) have been identified within the LSK (Lin⁻ Sca-1⁺ c-kit⁺) fraction of bone marrow cells (Morrison SJ and Weissman IL Immunity 1994). We have proposed that the first commitment step at the myeloid versus lymphoid bifurcation occurred outside the LSK fraction, where myeloid or lymphoid lineage-committed progenitors such as common myeloid progenitors (CMPs) and common lymphoid progenitors (CLPs) are prospectively isolated (Akashi et al Nature 2000). However, recent studies revealed that the initial commitment step might occur within LSK fraction. Adolfsson et al showed that a fraction of LSK cells expressing Flt3 at a high level have lost megakaryocyte/erythroid (MegE) potential and were largely primed for granulocyte/monocyte (GM) and lymphoid cells. This population was named lymphoid-primed multipotent progenitors (LMPPs) (Adolfsson et al Cell 2005). In our hands, by utilizing mice harboring a fluorescent reporter for GATA-1 transcription factor, we demonstrated that the initial upregulation of GATA-1 transcription factor was observed within MPP population, which was defined as CD34⁺ LSK fraction, and GATA-1⁺ MPPs were capable of generating only myelo-erythroid cells but lacked lymphoid potential (Arinobu et al Cell Stem Cell 2007). This result strongly suggested that the initial myelo-erythroid commitment occurs at the MPP stage.

To isolate the earliest myelo-erythroid LSK progenitors in normal mice without utilizing the GATA-1 reporter system, we conducted expression profiling of GATA-1⁺ MPP population by cDNA microarray analyses and following validation by FACS analyses. We found that a cell-surface antigen CD41 was specifically expressed at a high level in GATA-1⁺ MPP cells, indicating that CD41^hi MPP cells might be the corresponding population to GATA-1⁺ MPP cells.

CD41^hi MPPs gave rise exclusively to GM and MegE colonies in vitro, but no lymphoid colonies were observed even under lymphoid-inducible conditions such as co-culture with OP9 or OP9-DL1 stromal cell layer. In vivo, CD41^hi MPPs showed differentiation potential into GM cells, erythroid cells and platelets, lacking B or T lymphoid cell-producing potential in congenic transplantation assay. While the GM cell-reconstitution potential of LMPPs and CMPs peaked around day10 and day4 after transplantation, respectively, CD41^hi MPPs had their peak around 3 weeks and relatively long-lasting reconstitution potential, presumably reflecting their immaturity. To further evaluate differentiation potential more directly and quantitatively, CD41^hi MPPs with LMPPs or CMPs were injected together into one individual recipient mouse. In this head-to-head competitive assay system, CD41^hi MPPs generated a plenty of mature GM cells, whose numbers were nearly ten times of those produced from the original CMPs or LMPPs in a reconstitution setting. Thus CD41^hi MPPs possess potent restricted-lineage differentiation potential into GM and MegE lineages.

In addition, the gene expression analysis by single-cell quantitative real-time PCR and cDNA microarray revealed that the CD41^hi MPP population expresses both GM- and MegE-affiliated genes, but not lymphoid genes, reflecting their lineage restriction. More importantly, each single CD41^hi MPP cell expresses either one or both of GATA-1 and PU.1 at a low level, suggesting that the promiscuous expression of these transcription factors might play a critical role in myelo-erythroid lineage commitment (Miyamoto et al Developmental Cell 2002).

Furthermore, to evaluate the functional importance of CD41^hi MPP population, we created the systemic infectious model mice, mimicking the severe peritonitis. CD41^hi MPP pool as well as GMP expands dramatically in response to an increased demand for GM cells, while LMPP does not; thus, CD41^hi MPPs compose physiologically important pathway in myelo-erythropoiesis.

Accordingly, CD41^hi MPP population might represent the earliest branch point for myelo-erythroid development, which resides upstream of the conventional CMP. Based on these data, we propose to redefine the true CMP as CD41^hi MPP population. This CD41^hi earliest myelo-erythroid progenitor should be useful to investigate the molecular mechanisms for hematopoietic lineage fate decision.