Isolation of normal human megakaryocytes

S ummary . This paper reports a simple procedure for obtaining human megakaryocytes with a high purification and high recovery yield. Bone marrow cells, obtained from surgically removed ribs, were separated by a two-step procedure. Initially, a single cell suspension was enriched in megakaryocytes by equilibrium density centrifugation, the low density cell fraction was subsequently layered over a shallow continuous albumin gradient in a glass sedimentation chamber. Megakaryocytes averaged 0.03 ± 0.02% of all nucleated cells in the starting marrow cell suspension, after this procedure an average 80 ± 15% of the initial megakaryocyte population was recovered with a purity of 94 ± 4%. Previous methods, based upon the use of a two-step procedure, are reviewed. The theory of velocity sedimentation is discussed with regard to the differences in the methodology used, which account for the different results I obtained.     

Isolation of human megakaryocytes by immunomagnetic beads

A simple method was developed to purify human megakaryocytes to homogeneity from normal bone marrow aspirates. An initial separation of marrow between 1.020 and 1.050 g/ml. Percoll density cut was used to enrich megakaryocytes. After washing, the cells were suspended with immunomagnetic beads which were coated with sheep anti-mouse IgG antibody and treated with anti-human glycoprotein (GP) IIb/IIIa monoclonal antibody, or the cells were treated with human platelet GP IIb/IIIa monoclonal antibody and suspended with the immunomagnetic beads which were coated with sheep anti-mouse IgG antibody. Megakaryocytes were selectively separated using a magnet. All of the isolated cells were morphologically recognizable megakaryocytes. 1.5-3.1 x 10(4) megakaryocytes were obtained from 1.7-4.5 x 10(8) bone marrow nuleated cells. These cells were all positive in immunoenzymatic staining for GP IIb/IIIa. Megakaryocytes obtained by this method responded to recombinant human GM-CSF (rhGM-CSF) showing an increased 3H-thymidine (3H-dT) incorporation. These data show that this method is useful for obtaining pure megakaryocyte populations which can be submitted to comprehensive biological studies.     

Isolation of large numbers of enriched human megakaryocytes from liquid cultures of normal peripheral blood progenitor cells

Investigations linking human megakaryocyte development and cell biology have been hindered by an inability to obtain large, relatively pure megakaryocyte cell preparations from in vitro stem cell cultures. We report here that such preparations can be generated from liquid cultures of normal human peripheral blood mononuclear cells stimulated by a serum source of megakaryocyte colony stimulating activity (Meg- CSA, the 0% to 60% ammonium sulfate protein fraction of aplastic canine serum). Adherent-depleted peripheral blood mononuclear cells are suspended at 5 x 10(5) to 10(6) cells/mL in supplemented liquid culture medium, platelet-poor human plasma 20% (vol/vol) and 1 to 2 mg/mL serum Meg-CSA protein. After 12 to 14 days of incubation, megakaryocytes constitute 3.0 +/- 2.9% (mean +/- SD, n = 8) of the unseparated cultured cell population. Megakaryocytes can be enriched by counterflow centrifugal elutriation to a purity of 58 +/- 14% (+/- SD) with a recovery of 13 +/- 7% and a viability of 67 +/- 19%. This algorithm results in the average isolation of approximately 3 x 10(5) enriched megakaryocytes from a 100-mL starting volume of peripheral blood. Cultured megakaryocytes exhibit normal light and ultrastructural morphology by Wright-Giemsa staining and electron microscopic analysis. After a 12-day culture interval, enriched megakaryocyte preparations exhibit morphologic stage distributions that are similar to normal human marrow. Stage distributions move rightward with culture duration indicating partial synchrony of megakaryocyte maturation. On cytospin preparations, megakaryocyte diameter averages 30.2 +/- 1.5 microns and increases with maturation stage. Flow cytometric analyses demonstrate the expression of platelet glycoproteins (GP) Ib and IIb/IIIa by the cultured megakaryocytes. The modal ploidy of the enriched cells at day 12 of culture is 16N and most remaining megakaryocytes are 8N or 32N. Liquid culture of serum Meg-CSA-stimulated human peripheral blood mononuclear cells represents a valuable investigative tool that should permit studies of human megakaryocyte biology that have not been possible in the past.     

Flow cytometric analysis of normal human megakaryocytes

Megakaryocytes from normal routine human bone marrow aspirates were analyzed by flow cytometry for size, fine cell structure and granularity, membrane expression of glycoprotein (GP) IIb/IIIa and ploidy. Marrow cells were initially enriched for megakaryocytes by a Percoll density gradient and megakaryocytes were labeled with a fluoresceinated monoclonal antibody directed to the GPIIb/IIIa complex. The cells were fixed with paraformaldehyde and stained with propidium iodide (PI) for DNA quantitation. Using two-color flow cytometry, megakaryocytes were identified by their high membrane immunofluorescence and their ploidy was determined according to the relative fluorescence intensity of the PI. Forward light scatter (FSC), correlating with cell size, 90 degrees side light scatter (SSC), reflecting primarily cell internal fine structure and granularity, and total cell membrane fluorescence were examined. To evaluate independently the relationship between size and cell membrane fluorescence obtained by flow cytometry, megakaryocytes were sorted directly on slides and analyzed by a laser-based anchored cell analyzer (ACAS). There was a strong correlation among size, SSC, and the level of membrane fluorescence. The mean diameter of megakaryocytes was 28.1 +/- 12.3 micron. The modal ploidy distribution was 16N with approximately one-fifth of the cells less than or equal to 4N. The mean FSC and SSC levels increased with increasing ploidy. However, the marked overlap observed between the ranges of these parameters in adjacent ploidy classes suggested that size and SSC increase continuously rather than by discrete steps as is characteristic for ploidy. The total surface membrane fluorescence was correlated with cell size (r = 0.98) as measured by FSC or directly by the ACAS (r = 0.85), and with cell ploidy (r = 0.99) indicating an augmentation in total membrane GPIIb/IIIa expression with an increase in cell size and ploidy. However, estimated GPIIb/IIIa fluorescence density was inversely correlated with FSC suggesting that the GPIIb/IIIa surface epitope density is decreased with increasing cell maturity. We conclude that flow cytometry is a useful technique for the rapid analysis of human megakaryocytes obtained by marrow aspiration, and should be applicable to studies of pathologic states.     

Purification of human megakaryocytes by fluorescence-activated cell sorting

For direct studies of growth control, a method was developed to purify viable human megakaryocytes to homogeneity from routine normal bone marrow aspirates. An initial separation of marrow over a 1.050 g/mL Percoll density cut was used to enrich megakaryocytes. After washing, the cells were specifically labeled with a fluoresceinated monoclonal antibody or F(ab')2 fragment to the platelet glycoprotein (GP) IIb/IIIa complex. Megakaryocytes were selectively sorted by using Becton Dickinson FACStar flow cytometer on the basis of a fluorescence intensity greater than 50-fold that of control cells. To increase resolution and purity the sorting rate was adjusted to one cell in 13 formed drops, and negative events that coincided with positive ones were aborted. Two thirds of the isolated cells were large, morphologically recognizable megakaryocytes with a forward light scatter fourfold that of the main cell population. Microscopic examination showed these cells to be greater than or equal to 98% megakaryocytes with a diameter of 20 to 46 microns and a ploidy range of 2N to 64N with a mode of 16N. The small highly fluorescent cells were 10 to 21 microns in diameter, and their ploidy range from 2N to 32N with main ploidy classes of 2N and 4N. The majority of these small cells also positively reacted with monoclonal antibody to platelet GPIb. The isolated cells were cultured in either Iscove's or leucine, lysine-deficient RPMI 1640 medium with 10% human plasma. The cells were maintained in culture more than three days and were capable of synthesis of both DNA and protein as assessed by radiolabeled thymidine and amino acid incorporation. Moreover, the isolated megakaryocytes were capable of responding to recombinant granulocyte-macrophage colony-stimulating factor. The data show that human megakaryocytes can be purified from routine marrow aspirates on the basis of a lineage marker and that they are capable of growth in vitro.     

Binding and regulation of thrombopoietin to human megakaryocytes

Thrombopoietin (TPO, c-Mpl ligand) is considered to play an important role in the regulation of megakaryocytopoiesis and platelet production by activating the cytokine receptor c-Mpl. We have examined the binding of ¹²⁵I-TPO to the human megakaryocytic cell line, CMK, and to primary human megakaryocytes. Scatchard analysis of TPO binding to its cognate receptor in megakaryocytic cells suggested the existence of a single class of c-Mpl receptors. CMK cells exhibited 1223 receptors per cell with a dissociation constant ( K _d) of K _d = 223 p M , whereas primary human megakaryocytes exhibited 12 140 receptors per cell and a dissociation constant of K _d = 749 p M . The pretreatment of CMK cells and primary bone marrow megakaryocytes with TPO resulted in a decreased binding of TPO to the c-Mpl receptors. This down-regulation was observed within 3 h and was not inhibited by cycloheximide. Phorbol ester, an activator of protein kinase C, also inhibited TPO binding to the c-Mpl receptors by reducing the number of these receptors. The pretreatment of CMK cells with IL-3, IL-6 and DMSO, all of which induced the differentiation of CMK cells, did not affect the binding of TPO to the c-Mpl receptors. These results suggest an additional mechanism, where protein kinase C may help to regulate the binding of TPO to these cells.     

Purification, characterization, and in vitro differentiation of cytotrophoblasts from human term placentae

Highly purified functional cytotrophoblasts have been prepared from human term placentae by adding a Percoll gradient centrifugation step to a standard trypsin-DNase dispersion method. The isolated mononuclear trophoblasts averaged 10 microns in diameter, with occasional cells measuring up to 20-30 microns. Viability was greater than 90%. Transmission electron microscopy revealed that the cells had fine structural features typical of trophoblasts. In contrast to syncytial trophoblasts of intact term placentae, these cells did not stain for hCG, human placental lactogen, pregnancy-specific beta 1-glycoprotein or low mol wt cytokeratins by immunoperoxidase methods. Endothelial cells, fibroblasts, or macrophages did not contaminate the purified cytotrophoblasts, as evidenced by the lack of immunoperoxidase staining with antibodies against vimentin or alpha 1-antichymotrypsin. The cells produced progesterone (1 ng/10(6) cells . 4 h), and progesterone synthesis was stimulated up to 8-fold in the presence of 25-hydroxycholesterol (20 micrograms/ml). They also produced estrogens (1360 pg/10(6) cells . 4 h) when supplied with androstenedione (1 ng/ml) as a precursor. When placed in culture, the cytotrophoblasts consistently formed aggregates, which subsequently transformed into syncytia within 24-48 h after plating. Time lapse cinematography revealed that this process occurred by cell fusion. The presumptive syncytial groups were proven to be true syncytia by microinjection of fluorescently labeled alpha-actinin, which diffused completely throughout the syncytial cytoplasm within 30 min. Immunoperoxidase staining of cultured trophoblasts between 3.5 and 72 h after plating revealed a progressive increase in cytoplasmic pregnancy-specific beta 1-glycoprotein, hCG, and human placental lactogen concomitant with increasing numbers of aggregates and syncytia. At all time points examined, occasional single cells positive for these markers were identified. RIA of the spent culture media for hCG revealed a significant increase in secreted hCG, paralleling the increase in hCG-positive cells and syncytia identified by immunoperoxidase methods. We conclude that human cytotrophoblasts differentiate in culture and fuse to form functional syncytiotrophoblasts.     

Expression of adhesion antigens of human bone marrow megakaryocytes, circulating megakaryocytes and blood platelets.

There is evidence that mature megakaryocytes migrate into sinusoids, enter the blood and fragment in the vascular bed. We wondered whether differences in expression of adhesion antigens could be associated with the egress of megakaryocytes from bone marrow into the peripheral blood or the fragmentation into platelets. Megakaryocytes from human marrow were purified by counterflow centrifugal elutriation followed by a glycoprotein Ib-dependent agglutination procedure. Megakaryocytes from central venous blood and pulmonary arteries were purified by counterflow centrifugal elutriation alone. Adhesion antigens were labelled in an immunohistochemical assay. Both bone marrow megakaryocytes and platelets from healthy volunteers stained > 75% positive for CD36, CD41, CD42, Cdw49b (alpha subunit VLA2), Cdw49e (alpha subunit VLA5), Cdw49f (alpha subunit VLA6) and CD62. Circulating megakaryocytes, although > 75% positive for CD41, had, unlike platelets and bone marrow megakaryocytes, a reduced and remarkable heterogeneous (5-100% positive) labelling with antibodies against Cdw49b, Cdw49e, Cdw49f. These results could be confirmed by comparing the bone marrow megakaryocytes, circulating megakaryocytes and platelets from 7 patients that were recovered and processed at the same time. Morphologically mature, circulating megakaryocytes have, unlike bone marrow megakaryocytes, a heterogeneous expression of adhesion antigens, especially of Cdw49b, Cdw49e, and Cdw49f.     

Expression of CD4 by human megakaryocytes.

The CD4 antigen, which serves as the receptor for human immunodeficiency virus type 1 (HIV-1) on T cells, has been detected on human megakaryocytes. Recent evidence of impaired thrombopoiesis in HIV-1-related thrombocytopenia suggested that these cells could be directly infected by the virus and prompted a search for a receptor on megakaryocytes of normal subjects that could permit entry of HIV-1. Bone marrow specimens from uninfected normal control subjects were centrifuged over Ficoll-Hypaque (1.077 g/ml) and analyzed by three-color analysis with a flow cytometer utilizing monoclonal antibodies against CD4 and a glycoprotein present on the surface of megakaryocytes and platelets (GPIIb/IIIa; CD41), as well as 7-aminoactinomycin D, a stain for DNA. Cells presumed to be megakaryocytes were identified by having a DNA content greater than tetraploid and staining brightly with anti-CD41. Approximately 0.4% of the nucleated cells of the marrow met these criteria. Twenty-five percent of these megakaryocytes stained as brightly as CD4+ T cells. Several clones of antibody recognizing different epitopes of the CD4 molecule gave similar results. Platelets were CD4-. Staining of megakaryocytes with anti-CD4 was confirmed by direct microscopic examination of Percoll-gradient-enriched megakaryocytes employing two-color (CD4-phycoerythrin and CD41-fluorescein) immunofluorescence analysis and phase-contrast microscopy. The proportion of double-labeled cells among 112 phase-contrast-identifiable megakaryocytes from five bone marrow specimens varied between 20% and 26% with a mean and SD of 22% +/- 2.5%. Thus some human megakaryocytes express CD4 on their surface that should be capable of binding the HIV-1 gp120 envelope protein. This could serve as a portal of entry for HIV-1.     

Terminal cytoplasmic maturation of human megakaryocytes in vitro

Several studies suggest that serum factors (thrombopoietins) regulate thrombopoiesis by altering the number, size, ploidy, and maturation rate of megakaryocytes (MK). Various in vivo systems have been used to quantitate these events. In this study, an in vitro system was developed to monitor terminal cytoplasmic maturation of isolated human MK. MK enriched by elutriation, which eliminated the MK progenitors, were suspended in culture with serum from either normal donors (NABS) or patients with aplastic anemia (AAS). In cultures composed of small platelet glycoprotein-positive mononuclear cells and morphologically immature MK, development was characterized by sequential shifts in MK through morphologically recognizable maturation stages I, II, III, and IV over eight days of incubation (I and II only; then I, II, III; II, III, IV; III and IV; then IV only). Platelet formation coincided with the appearances of stage IV cells. Cultures composed of a mixture of all stages followed a similar maturation sequence, only at an accelerated rate. AAS resulted in the more rapid appearances of the mature cells in either system. This study indicates that human MK can undergo terminal cytoplasmic maturation in vitro, and that altering culture conditions (AAS for NABS) can accelerate the rate of maturation. Three major events occur during megakaryocytopoiesis: proliferation of the progenitor cells, polyploidization, and cytoplasmic maturation. Now it is possible to study the terminal steps of differentiation independent of proliferative events.     

Detection of vitamin D binding protein on the surface of cytotrophoblasts isolated from human placentae

Vitamin D binding protein (DBP), a Mr 56,000-58,000 alpha 2-glycoprotein, is the major serum protein involved in the transport of vitamin D sterols. Recently it has been suggested that DBP may also be involved in immunoglobulin G binding to cells. Because the trophoblast is involved in the transport of molecules such as vitamin D and immunoglobulin G to the fetus, we asked whether DBP could be detected on the surface of human placental trophoblast cells. Cytotrophoblasts purified from human term placentae were fixed and made permeant with Triton X-100 and examined by indirect immunofluorescence after incubation with a monoclonal antibody to DBP. Greater than 90% of these cells stained positively, whereas no staining was observed with nonimmune antiserum. The presence of DBP on/in the surface of cytotrophoblasts could also be demonstrated by fluorescent cytometry. When cell surface-associated proteins of cytotrophoblasts were radioiodinated, a Mr 57,000 radiolabeled protein could be immunoisolated from the cell lysate with a purified monospecific polyclonal antibody to DBP. Immunoisolation of this radiolabeled protein was prevented by the addition of excess unlabeled human DBP to the cell lysate before incubation with antibody. This Mr 57,000 radiolabeled protein could also be isolated by affinity chromatography selecting for proteins that bind to globular actin. When cytotrophoblasts were incubated with [35S]methionine for 3 or 18 h, active synthesis of DBP could not be demonstrated by immunoisolation techniques. These studies demonstrate the presence of DBP on the surface of well washed, human cytotrophoblasts. This DBP may be maternally derived, since active synthesis of DBP could not be demonstrated.     

Immunological study of in vitro maturation of human megakaryocytes

Human megakarocyte colonies were grown from the bone marrow in plasma clot or methyl cellulose cultures. Maturation of the megakaryocytic cells was sequentially studied from day 5 to day 16 of culture by fluorescent labelling with a panel of monoclonal and polyclonal antibodies against different platelet glycoproteins (Gp), P1 A1 antigen, factor VIII RAg platelet factor 4 (PF 4), fibrinogen and platelet-derived growth factor (PDGF). Expression of Gp Ib was also studied by immunogold technique at electron microscopy. The first cells identifiable by these antibodies were found at day 5 of culture. They had the size of a lymphocyte. These small megakaryocyte precursors already expressed all the platelet antigens, HLA-DR and transferrin receptors and were devoid of erythroid or myeloid markers. Among the platelet antigens, Gp IIIa was the most sensitive marker for the identification of these precursors. However, double-fluorescent labelling demonstrated that the different platelet markers were coexpressed in a large majority of cells. Interestingly, cytoplasmic markers demonstrated that these small megakaryocyte precursors were themselves heterogenous by morphological criteria. During maturation, expression of Gps, particularly of Gp Ib, increased while the labelling pattern of anti factor VIII RAg and anti PF 4 antibodies switched from diffuse to granular staining. PDGF could also be detected in the megakaryocytes grown in culture.