The study of megakaryocytopoiesis has been based largely on in vitro assays. We characterize an in vivo model of megakaryocyte and platelet development in which human peripheral blood stem cells (PBSCs) differentiate along megakaryocytic as well as myeloid/lymphoid lineages in sublethally irradiated nonobese diabetic/severe combined immunodeficient (NOD-SCID) mice. Human hematopoiesis preferentially occurs in the bone marrow of the murine recipients, and engraftment is independent of exogenous cytokines. Human colony-forming units-megakaryocyte (CFU-MK) develop predominantly in the bone marrow, and their presence correlates with the overall degree of human cell engraftment. Using a sensitive and specific flow cytometric assay, human platelets are detected in the peripheral blood from weeks 1 to 8 after transplantation. The number of circulating human platelets peaks at week 3 with a mean of 20 x 10(9)/L. These human platelets are functional as assessed by CD62P expression in response to thrombin stimulation in vitro. Exogenous cytokines have a detrimental effect on CFU-MK production after 2 weeks, and animals treated with these cytokines have no circulating platelets 8 weeks after transplantation. Although cytokine stimulation of human PBSCs ex vivo led to a significant increase in CFU-MK, CD34+/41+, and CD41+ cells, these ex vivo expanded cells provided only delayed and transient platelet production in vivo, and no CFU-MK developed in vivo after transplantation. In conclusion, xenogeneic transplantation of human PBSCs into NOD/SCID mice provides an excellent in vivo model to study human megakaryocytopoiesis and platelet production.

In human polymorphonuclear leukocytes (PMN), changes in the actin architecture are critical for the shape changes required for chemotaxis and phagocytosis. Barbed-end capping proteins are likely to regulate actin assembly in PMN. The previously identified barbed-end blocking proteins in PMN, gelsolin and CapG, require Ca(2+) to initiate capping of actin filaments. Because chemoattractants can stimulate PMN actin assembly by a calcium-independent signal transduction pathway, we sought to purify a calcium-independent barbed-end capping activity from PMN cytoplasmic extracts. A Ca(2+) -insensitive actin polymerization inhibitory activity was partially purified from human PMN [Southwick & Stossel (1981) J. Biol. Chem 256, 3030]. Using five column chromatography steps, we purified the protein to homogeneity as assessed by silver staining. Purification was associated with an increase in specific activity of greater than 40 X. Western blot analysis identified the protein as the nonmuscle isoform of the heterodimeric capping protein capZ. Human PMN capZ has an apparent disassociation constant of 3 nM for capping in the presence or absence of micromolar Ca(2+), as assessed by both pyrenylactin elongation and depolymerization assays. Similar to the activity reported for the actin polymerization inhibitor, activity of PMN capZ was inhibited by increasing the KC1 concentration from 0.1 M to 0.6 M. The capping function was also inhibited by phosphatidylinositol 4,5-bisphosphate (PIP(2)) micelles, with half-maximal inhibition occurring at 5.5 micrograms mL(-1). PMN capZ did not nucleate actin assembly, sequester actin monomers, or sever actin filaments. Quantitative Western blot analysis revealed that capZ levels corresponded to 0.7-1.0% of the total human PMN cytoplasmic protein. Given its abundance and high affinity for barbed filament ends, capZ is likely to play an important role in the calcium-independent regulation of actin filament assembly associated with PMN chemotaxis.