The aim of the present study was to determine if the human erythroid (E) and megakaryocytic (MK) lineages were closely linked to the existence of a bipotent burst-forming unit (BFU) E/MK progenitor. In methylcellulose cultures, BFU-E/MK colonies were observed at day 12 and closely resembled mature BFU-E with the exception that the erythroid component was surrounded by MK. These colonies were quite different from the colony forming unit (CFU)-GEMM-derived colonies, which were composed of a larger number of erythroblasts and which developed later in culture. The existence of these bilineage colonies composed of 100 to 1,000 erythroblasts intermingled with a few MK and without granulocytic cells was confirmed by the plasma clot technique and immunoalkaline phosphatase labeling of the MK. To investigate if this bipotent progenitor belonged to the compartment of primitive progenitors, CD34+ marrow cells were subfractionated according to expression of the CD38 antigen. The bipotent BFU-E/MK progenitor as well as a large fraction of MK progenitors were found in the CD34+ CD38+/- or in the CD34+ CD38- cell fractions. Growth of this bipotent BFU-E/MK progenitor required the combination of stem cell factor (SCF), Interleukin-3 (IL-3), and Epo in serum free conditions. Addition of IL- 6 had only a marginal effect, whereas megakaryocyte growth and development factor (MGDF) was not an absolute requirement, but slightly increased the plating efficiency of CFU-MK and of BFU-E/MK progenitors when combined with SCF, IL-3, and Epo. In contrast, when these cultures were performed in the presence of 30% fetal calf serum, no BFU-E/MK colonies were observed irrespective of the combination of growth factors used, including the presence of MGDF; however, inclusion of the MS-5 cell line restored the growth of this bipotent progenitor. In contrast, in cultures performed in the presence of human normal or aplastic plasma, MS-5 had only a slight effect on the cloning efficiency but improved MK cytoplasmic maturation and MK size, suggesting that the main effect of MS-5 is to diminish the inhibitory effect of the fetal calf serum on the MK differentiation. The clonal origin of bipotent BFU-E/MK colonies was demonstrated in liquid culture of single CD34+ CD38low cells by immunophenotyping individual clones. At day 12, 30% of the clones contained erythroblasts (glycophorin A+) and some MK (CD41+) without granulocytes (G) or macrophages (M) (CD14+ and CD15+). At day 20, clones containing erythroblasts and MK were rare (5%). In contrast multilineage clones could be frequently detected at this time without passage from BFU-E/MK clones at day 12 to GEMM at day 20. These results suggest that a bipotent BFU-E/MK progenitor may be a nonrandom step in the hierarchical development of stem cells.

In an attempt to understand better the regulation of stem cell function in chimeric immunodeficient mice transplanted with human cells, and the filiation between progenitor cells identified in vitro and in vivo, we assessed the different compartments of hematopoietic progenitors found in the marrow of CB17-severe combined immunodeficiency (SCID) mice (34 mice, 9 experiments) after intravenous injection of 2 to 3 x 10(7) cord blood mononuclear cells. On average 6.3 +/-4 x 10(5) human cells were detected per four long bones 4 to 6 weeks after the transplant predominantly represented by granulomonocytic (CD11b+) and B lymphoid (CD19+) cells. Twenty five percent of these human cells expressed the CD34 antigen, of which 90% coexpressed the CD38 antigen and 50% the CD19 antigen. Functional assessment of progenitor cells (both clonogenic and long-term culture-initiating cells [LTC-IC]) was performed after human CD34+ cells and CD34+/CD38- cells have been sorted from chimeric CB17-SCID marrow 3 to 10 weeks after intravenous (IV) injection of human cells. The frequency of both colony-forming cells and LTC-IC was low (4% and 0.4%, respectively in the CD34+ fraction) when compared with the frequencies of cells with similar function in CD34+ cells from the starting cord blood mononuclear cells (26% +/- 7% and 7.2% +/- 5%, respectively). More surprisingly, the frequency of LTC-IC was also low in the human CD34+ CD38- fraction sorted from chimeric mice. This observation might be partly accounted for by the expansion of the CD34+ CD19+ B-cell precursor compartment. Despite their decreased frequency and absolute numbers, the differentiation capability of these LTC-IC, assessed by their clonogenic progeny output after 5 weeks in coculture with murine stromal cells was intact when compared with that of input LTC-IC. Furthermore the ratio between clonogenic progenitor cells and LTC-IC was similar in severe combined immunodeficiency (SCID) mice studied 4 weeks after transplant and in adult marrow or cord blood suspensions. Results generated in experiments where nonobese diabetic (NOD)-SCID mice were used as recipients indicate a higher level of engraftment but no change in the distribution of clonogenic cells or LTC-IC. These results suggest that the hierarchy of hematopoietic differentiation classically defined in human hematopoietic tissues can be reconstituted in immunodeficient SCID or NOD-SCID mice.

Previously, it was believed that megakaryocytopoiesis was regulated by two types of humoral factors: megakaryocyte colony-stimulating factor (MK-CSF), which acts on progenitors inducing their proliferation, and thrombopoietin (TPO), a megakaryocyte(s) (MK) maturational factor that induces platelet formation. The recently cloned Mpl-ligand (Mpl-L) seems to have both properties in vivo and in vitro and has also been called TPO. However, it cannot be excluded that a part of these activities is due to a synergistic effect with growth factors present in the serum or synthesized by accessory cells. To delineate the precise TPO (Mpl-L) biologic activities, we performed serum-free cultures at limiting cell dilution. Target cells were adult human marrow CD34+CD41+ cells, which represent a highly selected population of late MK progenitor or transitional cells. Cells were purified using a flow cytometer equipped with an automatic cloning design unit. We determined that the recombinant molecule had a biologic activity that reached a plateau at 10 ng/mL. At this concentration, a linear relationship between the average MK number per well and the number of cells seeded (between 1 to 50 cells per well) was observed. At one cell per well, 60% of the wells contained a single MK at day 5 of culture. Half of these wells contained only one large MK, whereas the other half contained several MK (up to 25), demonstrating that TPO has direct proliferative biologic activity. In contrast, at limiting dilution, none of the other cytokines tested (stem cell factor [SCF], interleukin- 6 [IL-6], and erythropoietin [Epo]) were effective, whereas IL-3 showed a mild effect. However, a combination of SCF plus IL-6 plus IL-3 produced similar results as TPO alone. Addition of the other cytokines to TPO did not enhance the cloning efficiency of the CD34+CD41+ cells but increased twofold the average number of MKs per clone. MKs reached a ploidy of 32N and 64N in the presence of TPO. The mean ploidy value was approximately 6 and was not modified by addition of the other cytokines. At the ultrastructural level, a majority of the MKs showed maturational defects related to an imbalance between the synthesis of alpha-granules and demarcation membranes. However, a fraction (about 30%) had a cytoplasmic maturation that exactly mimicked that of marrow MKs. In addition, proplatelet-shedding MKs were observed in the cultures, even at limiting dilution. Such a result was not observed with any other individual cytokines, including the combination of three cytokines.(ABSTRACT TRUNCATED AT 400 WORDS)