The marrow of Gata1^low mice contains progenitor cells with the myeloid, but not with the mastocytic, antigenic profile. (A) Diagrammatic scheme of the antigenic profiling used to define different progenitor cell types in this study. All myeloid and mast cell generating activity is present among the cKit⁺ cells of the Lin⁻Sca1⁻Ly6c⁻FcϵRIα^[−] cells of normal marrow.^18,21  The myeloid generating cells are divided by CD34 and CD16/CD32 staining into GMPs (CD34⁺CD16/CD32⁺, not shown), MEPs (CD34^lowCD16/CD32^low), and CMPs (CD34⁺CD16/CD32^low).¹⁸  GMPs correspond to progenitor cells functionally defined as CFU-Gs, CFU-Ms, and CFU-GMs, and are not investigated in detail in this study. MEPs are cells restricted toward megakaryocyte and erythroid differentiation and include those cells once functionally defined as CFU-MK_day3 and CFU-MK_day7, CFU-E, and BFU-E. CMPs give rise to GMPs, MEPs, and MCPs, and correspond to cells functionally defined as capable to generate in vitro CFU-mix and in vivo CFU-S_{day 12}. The c-Kit⁺ bone marrow cells can be also divided by T1/ST2 staining into T1/ST2⁻ cells that contain all of the myeloid cells described (myeloid progenitors, MPs) and T1/ST2⁺ cells, defined as MCPs, that generate c-Kit⁺/T1/ST2⁺ cells and mast cells within 7 days both in vitro (in semisolid and liquid cultures) and in vivo (in extramedullary sites).²¹  (B) Prospective identification and isolation of MEPs, CMPs, and MCPs from the marrow of Gata1^low and wild-type littermates. Lin⁻c-Kit⁺ cells from normal marrow (the c-Kit staining of the CD34-CD16/CD32 analysis is not shown, for clarity) were further divided by CD34-CD16/CD32 (on the left) and T1/ST2 (on the right) staining according to prospective gates (indicated by rectangles) corresponding to MEPs or CMPs and MPs and MCPs, respectively. Wild-type MPs are defined as cells with the c-Kit⁺/T1/ST2⁻ phenotype and contain, by definition, a mixture of CMPs, MEPs, and GMPs. In the case of wild-type cells, all of the sorted fractions were 80% to 95% pure, as shown by the presented reanalysis. The Lin⁻c-Kit⁺ cells from Gata1^low bone marrow were also divided by CD34-CD16/CD32 staining according to the prospective gates corresponding to MEPs or CMPs (∼90% pure, by reanalysis). In contrast, T1/ST2⁺ cells were not detectable among Gata1^low Lin⁻c-Kit⁺ cells. The absence of MCPs among the marrow cells from Gata1^low mice was confirmed by arbitrarily sorting T1/ST2⁻ cells into the one third-lower (MP^low) and one third-higher (MP^high) portion of the staining, as indicated, and by showing that the profile of the 2 sorted populations was largely overlapping (< 60% pure). Results from 2 representative experiments are shown. Similar results were observed in at least 3 independent experiments per group. The mean (± SD) frequency of each population observed in multiple experiments is summarized in Table 1. (C) Gata1 expression in different classes of progenitor cells prospectively isolated from the marrow of wild-type and Gata1^low mice, as indicated. Results are presented as mean (± SD) of at least 3 independent determinations for progenitor cell type. Values of Gata1 expression statistically different (P < .05) from those expressed by the corresponding CMPs are indicated by a single asterisk, whereas differences statistically significant between values observed in populations with the same profile purified from Gata1^low and wild-type littermates are indicated by double asterisks.