Identification of Human Monocyte Lineage Progenitors Lacking Granulocyte Potential: Revision of Phenotypic Definition of Human GMP.

Abstract 1614

To understand development and homeostasis of the immune system, it is crucial to understand origin and pathways for myeloid cells. In human hematopoiesis, all myeloid cells including granulocytes, monocytes, macrophages and dendritic cells are derived from granulocyte/macrophage progenitors (GMPs: CD34⁺CD38⁺IL-3R⁺CD45RA⁺) (PNAS, 2002, vol.99: 11872–11877), but putative progenitor populations downstream of GMPs that are committed to either monocyte or granulocyte lineages have not been isolated. Here we newly identified the human progenitor of monocyte lineage, which is capable of differentiation into monocytes, macrophages and dendritic cells but not into granulocytes. To search for the branch point toward the monocyte/macrophage lineage, we first tried to identify molecules that are differentially expressed in monocytes and granulocytes by utilizing genome-wide cDNA microarray analysis. We found that CD86, a functional ligand for naïve T-lymphocytes, and human T-cell immunoglobulin mucin-3 (TIM-3), a surface molecule playing a role in phagocytosis, were preferentially expressed in monocytes. The expression patterns of these molecules in hematopoiesis were then tested by a multi-color flowcytometric analysis. The bone marrow hematopoietic stem cells, and progenitor populations such as common myeloid progenitors (CMPs), and megakaryocyte/erythrocyte progenitors (MEPs) did not express either of these molecules. However, GMPs were subdivided into 3 subpopulations including CD86^-TIM-3^-, CD86⁺TIM-3^-, and CD86⁺TIM-3⁺ populations. In clonogenic methylcellulose colony-forming unit (CFU) assays, CD86^-TIM-3^- cells gave rise to all types of myeloid colonies including CFU-granulocyte macrophage (CFU-GM), CFU-granulocyte (CFU-G), and CFU-macrophage (CFU-M). CD86⁺TIM-3^- cells formed CFU-M and CFU-GM, but lacked CFU-G activity. Notably, CD86⁺TIM-3⁺ cells gave rise exclusively to CFU-M. These data suggest that the CD86^-TIM-3^- population is the true GMP, which is progressively expressing CD86 and then TIM-3 to become CD86⁺TIM-3⁺ monocyte progenitor population. The liquid culture assays of these GMP subfractions further strengthen this hypothesis: CD86^-TIM-3^- cells gave rise to both CD86⁺TIM-3^- cells and CD86⁺TIM-3⁺ cells, CD86⁺TIM-3^- cells differentiated into only CD86⁺TIM-3⁺ cells, and CD86⁺TIM-3⁺ cells into neither CD86^-TIM-3^- cells nor CD86⁺TIM-3^- cells. Collectively, CD86^-TIM-3^- population is the GMP, and these cells progressively lose granulocyte potential to become CD86⁺TIM-3⁺ monocyte lineage progenitors. Furthermore, in the liquid cultures in the presence of TNF-alpha, CD86⁺TIM-3⁺ GMP gave rise to myeloid dendritic cells. Collectively, the CD86⁺TIM-3⁺ fraction of human GMP populations is the long-sought monocyte lineage progenitors that are common to monocytes and their descendent macrophages and dendritic cells. Our data also indicate that the human GMP populations could more strictly be purified by adding the definition of negative expression of CD86 and TIM-3. The development of methods for prospective purification of these new progenitor populations revises the model of human myeloid cell differentiation, which should be useful to further understand human immune systems.