Revision of the Human Hematopoietic Tree: Granulocyte Subtypes Derive from Distinct Hematopoietic Lineages

The classical model of hematopoiesis predicts a dichotomous lineage restriction of multipotent hematopoietic progenitors (MPPs) into common lymphoid progenitors (CLPs) and common myeloid progenitors (CMPs). However, this idea has been challenged by the identification of lymphoid progenitors retaining partial myeloid potential (e.g., LMPPs), implying that granulocytes can arise within both the classical lymphoid and the myeloid branches. Here, we resolve this issue by using cell-surface CD133 expression to discriminate functional progenitor populations. We show that eosinophilic and basophilic granulocytes as well as erythrocytes and megakaryocytes derive from a common erythro-myeloid progenitor (EMP), whereas neutrophilic granulocytes arise independently within a lympho-myeloid branch with long-term progenitor function. These findings challenge the concept of a CMP and restore dichotomy to the classical hematopoietic model. [Display omitted] •Basophils and eosinophils, but not neutrophils, derive from the erythro-myeloid lineage•Hematopoietic progenitors with long-term potential reside in the CD133+/CD34+ fraction•Most progenitors with erythroid potential reside in the CD133low/CD34+ fraction•Cultured CD133+/CD34+ cells lose erythro-myeloid but retain SCID-repopulating potential The classical hematopoietic model predicts binary segregation of lymphoid and myeloid potentials. The discovery of progenitors with partial lymphoid and myeloid potentials has challenged this model. Here, Giebel and colleagues analyzed the development of human hematopoietic stem and progenitor populations in various in vitro and in vivo assays. Upon demonstrating that granulocyte subtypes arise in distinct lineages, they provide a revised model of hematopoiesis in which eosinophils and basophils develop from erythro-myeloid progenitors (EMPs) and neutrophils from lymphoid-primed multipotent progenitors (LMPPs).