Search

Granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin (IL) -3 induced tyrosine phosphorylation of 92-kDa protein in normal human monocytes. We identified this 92-kDa protein as STAT5, but not as STATs1, 3, and 6 nor c-fes and vav protooncogene products, and demonstrated its translocation to the nucleus, enhancement of specific DNA binding capacity, and potentiation of trancriptional activity by GM-CSF. N-formyl-methionyl-leucyl-phenylalanine (FMLP) and phorbol myristate acetate (PMA) induced tyrosine phosphorylation of 42- and 44-kDa proteins, which were identified as extracellular signal-regulated kinase (ERK), in human monocytes. In marked contrast to neutrophils and MO7e cells, GM-CSF did not induce tyrosine phosphorylation and activation of ERK in monocytes. Among upstream signaling molecules of ERK, Shc was constitutively associated with Grb2 and was not tyrosine-phosphorylated by GM-CSF and FMLP, and Sos1 and c-Raf-1 were not phosphorylated by GM-CSF, IL-3, TNF, and FMLP in monocytes, whereas all these signaling molecules were affected and/or utilized by GM-CSF in MO7e cells. In contrast to neutrophils, p38 was constitutively phosphorylated and agonist-dependent phosphorylation and activation was not detected in human monocytes. Superoxide release stimulated by FMLP was inhibited partially by PD98059 or SB203580, a specific inhibitor of ERK or p38 pathway, and was almost completely inhibited by the combination of both inhibitors, whereas PMA-induced superoxide release was resistant to these two inhibitors in monocytes. PD98059 inhibited GM-CSF-dependent proliferation of MO7e cells. Present results indicate trancriptional roles of STAT5 and functional roles of ERK and/or p38 in normal human monocytes stimulated by physiological receptor-mediated agonists GM-CSF and FMLP. Possible roles of ERK in proliferation of transformed cells were also suggested.

The neutrophil superoxide (O2-)-producing capacity in 57 patients with chronic myeloproliferative disorders (MPDs) and eight patients with chronic myelomonocytic leukemia (CMML) was investigated. O2- release in neutrophils stimulated by chemotactic peptide was markedly increased in all types of chronic MPD, including chronic myelogenous leukemia in both chronic phase and blastic crisis, polycythemia vera, and essential thrombocythemia, but was normal in CMML, which is thought to be a myelodysplastic disorder rather than MPD. Increase in O2(-)-producing capacity in MPD was also observed when other receptor-mediated agonists such as interleukin-8 and concanavalin A were used, but not when phorbol ester, a direct activator of protein kinase C, was used as the triggering agonist of O2- release. Priming effects of granulocyte colony-stimulating factor (G-CSF), granulocyte-macrophage CSF (GM-CSF), and tumor necrosis factor (TNF) on chemotactic peptide-induced O2- release was observed in all patients with MPD and CMML, though fold enhancement of priming effects was much less in MPD compared with normal subjects. In addition, the priming effects of TNF were less than those of GM-CSF in 10 cases, whereas the priming effects of TNF were consistently and markedly greater than those of GM-CSF in normal subjects. Tyrosine phosphorylation of 42-kDa protein stimulated by G-CSF, GM-CSF, and TNF was observed in CML neutrophils to be identical to that in normal neutrophils. Present results indicate specific potentiation of the receptor-mediated route of signaling that is linked to the respiratory burst and downregulated responsiveness to cytokines in neutrophils in patients with all types of chronic MPD, suggesting in vivo priming of patient neutrophils via certain mechanism by cytokines or related stimuli in these hematological disorders.

Tumor necrosis factor (TNF), like granulocyte-macrophage colony-stimul ating factor (GM-CSF), rapidly primed human monocytes for enhanced release of superoxide (O-2) stimulated by receptor-mediated agonists, N-formyl-methionyl-leucyl-phenylalanine (FMLP) and concanavalin A (Con A), but not by phorbol myristate acetate (PMA), which bypasses the receptors to stimulate the cells. The optimal priming was obtained by pretreatment of suspended monocytes with 10 U/mL TNF for 10 minutes at 37 degrees C. The potency of the maximal priming effect was TNF> GM-CSF, and the combined effect of TNF and GM-CSF was greater than that of each cytokine alone. GM-CSF induced an increase in cytoplasmic pH but TNF did not. These findings suggest that TNF and GM-CSF activate monocytes through different mechanisms. TNF and GM-CSF by themselves never triggered O-2 release in suspended monocytes or monocytes adherent to endothelial cells, although both cytokines triggered massive release of O-2 in human neutrophils. In additions, TNF and GM-CSF induced tyrosine phosphorylation of a 42-kD protein in neutrophils but not in monocytes. These findings suggest that the TNF-receptor- or GM-CSF-receptor-mediated signaling pathways for triggering O-(2) release is active in neutrophils but inactive or defective in monocytes. TNF also enhanced phagocytosis of sialidase-treated autologous erythrocytes by monocytes, and this effect was further potentiated in the presence of autologous fresh serum. The significant enhancement of erythrophagocytosis was obtained at 1 U/mL TNF. At this concentration of TNF, the expression of C3bi-receptor (CD11b/CD18) was upregulated. These findings show that TNF rapidly primes human monocytes for enhanced release of O-(2) and erythrophagocytosis and suggest that TNF activates monocytes through autocrine or paracrine mechanisms at the inflammatory sites inasmuch as TNF is primarily produced by activated monocytes/macrophages.

Both monocyte chemotactic and activating factor (MCAF) and N-formyl-methionyl-leucyl-phenylalanine (FMLP) stimulated an increase in cytoplasmic free Ca2+ ([Ca2+]i) and changes in intracellular pH (pHi) in human monocytes in parallel at lower concentrations and stimulated superoxide (O2-) release and changes in transmembrane potential in parallel at higher concentrations. The changes in pHi were characterized by initial rapid acidification followed by sustained alkalinization, and the changes in transmembrane potential were characterized by initial depolarization followed by partial repolarization. The time courses of all responses stimulated by MCAF and FMLP were similar to each other, although the magnitude of all responses was less in MCAF-stimulated cells. MCAF by itself was a very weak stimulus for inducing O2- release. However, MCAF primed monocytes and enhanced O2- release stimulated by FMLP. The priming effect of MCAF was maximal within 5 minutes of preincubation, and the dose-response curves for priming were identical to those for triggering of an increase in [Ca2+]i. Treatment of monocytes with the intracellular Ca2+ chelator, 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA), abolished not only the increase in [Ca2+]i but also the changes in pHi (both acidification and alkalinization) induced by MCAF or FMLP. MCAF further potentiated FMLP-induced O2- release in tumor necrosis factor (TNF)-, granulocyte-macrophage colony-stimulating factor (GM-CSF)-, or IL-3-primed monocytes. These findings suggest that MCAF, alone or in concert with other cytokines, primes monocytes for enhanced release of 02-, and that MCAF- or FMLP-induced intracellular acidification and alkalinization are closely associated with an increase in [Ca2+]i, but not O2- release.

Together, interleukin-1 alpha (IL-1 alpha) and IL-1 beta primed human neutrophils for enhanced release of superoxide (O2-) stimulated by chemotactic peptide, chemokine, and plant lectin, and alone, each triggered O2- release in a dose-dependent manner. The maximal priming and triggering effect was obtained by high concentrations (50 to 500 ng/mL) of IL-1 alpha or IL-1 beta, though IL-1 beta was more effective than IL-1 alpha at suboptimal concentrations. Priming effect of IL-1 was very rapid and maximal within 10 minutes, whereas O2- release triggered by IL-1 was gradual and continued for 90 to 120 minutes. Combined stimulation of human neutrophils with IL-1 plus granulocyte colony-stimulating factor (G-CSF) or granulocyte-macrophage CSF (GM-CSF) resulted in additive priming effect, and combined stimulation of neutrophils with IL-1 plus G-CSF, GM-CSF, or tumor necrosis factor (TNF) resulted in additive triggering effect, even when the maximal concentration of each cytokine was used. These priming and triggering effects of IL-1 alpha and IL-1 beta on the respiratory burst in human neutrophils were completely inhibited by IL-1 receptor antagonist (IL-1ra). Furthermore, only the net effect of IL-1 was inhibited by IL-1ra, even when human neutrophil was stimulated with IL-1 plus other cytokines to release O2-. Present results indicate that IL-1 does stimulate the respiratory burst activity in human neutrophils via receptor-mediated mechanism and suggest that the post-IL-1-receptor signaling pathways linked to the activation system of the respiratory burst are independent from those utilized by other cytokines in human neutrophils.

The superoxide (O2-)-releasing capacity in response to N-formyl-methionyl-leucyl-phenylalanine (FMLP) and the priming effects of recombinant human granulocyte colony-stimulating factor (rhG-CSF) and granulocyte-macrophage colony-stimulating factor (rhGM-CSF) on FMLP-induced O2-release were investigated in neutrophils from 13 patients with aplastic anemia (AA). The O2(-)-releasing capacity of AA neutrophils (0.85 +/- 0.36 nmol/5 min/1 x 10(5) cells, n = 13) was significantly (p < 0.01) increased as compared with that of normal neutrophils (0.24 +/- 0.12 nmol/5 min/1 x 10(5) cells, n = 17). There was no close relationship between the O2(-)-releasing capacity and the peripheral blood neutrophil count or the plasma concentration of C-reactive protein. The plasma concentrations of G-CSF and GM-CSF were not elevated to the detectable levels (< 0.1 ng/ml and < 0.2 ng/ml, respectively) in all patients tested. FMLP-induced O2(-)-release was further enhanced by pretreatment of cells with rhG-CSF or rhGM-CSF for 10 min at 37 degrees C, except that no significant priming by rhG-CSF was observed in five patients. The priming effect of rhGM-CSF was consistently greater than that of rhG-CSF in all patients. The i.v. administration of rhGM-CSF (6 micrograms/kg body weight/day) to one patient resulted in an increase in neutrophil O2(-)-release stimulated by FMLP. These findings indicate that neutrophils from AA patients are already primed in vivo for enhanced release of O2- and that these neutrophil functions are further potentiated by rhG-CSF or rhGM-CSF.

Blood cells from patients with aplastic anemia (AA) were evaluated for the ability to produce interleukin 3 (IL-3) and granulocyte-macrophage colony-stimulating factor (GM-CSF) on stimulation with phytohemagglutinin (PHA) or antilymphocyte globulin (ALG) by the use of an IL-3-dependent cell-line, TF-1, and the GM-CSF-IRMA kit. The IL-3 levels in patients with active AA were significantly lower, both in PHA-stimulated conditioned medium (CM) and in ALG-CM, than those of normal healthy donors (HD; p < 0.01). The degree of reduced IL-3 production in AA patients correlated well with the severity of neutropenia; the level of IL-3 returned to normal after successful treatment with ALG plus methylprednisolone (ALG therapy). On the other hand, GM-CSF production in AA patients varied widely and was only significant in remission patients in PHA-CM; in this case production was higher than that in active AA patients (p < 0.05) or in HD (p < 0.01). Sensitivity to PHA or ALG stimulation was evaluated by the ratio of IL-3 concentrations in ALG-CM versus PHA-CM (ALG/PHA index). The index varied widely from < 0.1 to > 10 in AA patients, contrasting to the clustered values in HD. Seven of the eight patients who had an ALG/PHA index of > 1.0 showed a good clinical response to ALG therapy. However, 12 of the 14 patients who had a lower index (< 1.0) failed to respond. The ALG/PHA index might have an ability to predict the response to ALG therapy.

We investigated granulocyte colony-stimulating factor (G-CSF) receptors on neutrophils from three patients with chronic myelogenous leukemia (CML) in the chronic phase, in comparison with four normal volunteers. Because we experienced some difficulties in radioiodinating intact recombinant human G-CSF, we developed a new derivative of human G-CSF termed YPY-G-CSF. It was easy to iodinate this protein using the lactoperoxidase method because of two additional tyrosine residues, and its radioactivity was higher than that previously reported. The biological activity of YPY-G-CSF as G-CSF was fully retained. Scatchard analysis demonstrated that CML neutrophils had a single class of binding sites (1400 +/- 685/cell) with a dissociation constant (Kd) of 245 +/- 66 pM. The number of sites and Kd value of CML neutrophils were not significantly different from those of normal neutrophils (p greater than 0.9). Cross-linking studies revealed two specifically labeled bands of [125I]YPY-G-CSF-receptor complexes with apparent molecular masses of 160 and 110 kd on both normal and CML neutrophils. This is the first report describing two receptor proteins on neutrophils. According to the analyses of the proteolytic process of these cross-linked complexes and proteolytic mapping, we assume that alternative splicing or processing from a single gene may generate two distinct receptor proteins that bind specifically to G-CSF but have different fates in intracellular metabolism.

Granulocyte colony formation by human bone marrow cells in serum-free cultures was studied using purified recombinant granulocyte colony-stimulating factor (rG-CSF). The cloning efficiency of the serum-free cultures was about 80% that of the serum-containing cultures. In addition to purified G-CSF, four ingredients were found to be essential to granulocyte colony formation: bovine serum albumin (BSA), iron-saturated human transferrin, cholesterol, and L-alpha-phosphatidylcholine. Their optimal concentrations were also investigated. Insulin was not indispensable for granulocyte colony formation, but its addition did increase the number of granulocyte colonies. Hydrocortisone was found to be inhibitory to granulocyte colony formation at high concentrations.

To obtain permanent preparations for morphologic examination of hemopoietic cells cultured in methylcellulose medium, a new method designated as "the membrane filtration technique" has been devised. Permanent preparations of erythroid colonies, bursts and leukemic colonies, which have been grown mainly in methylcellulose medium, are illustrated in this paper. The stages of differentiation of cells within these colonies can be easily recognized by means of Wright-Giemsa staining.

[3H]thymidine uptake by NFS-60 cells in microcultures was found to increase in a linear fashion with the increasing doses of purified recombinant human granulocyte colony-stimulating factor (rhG-CSF). Such increases were found neither with rhG-CSF samples pretreated with rabbit anti-rhG-CSF serum nor with other human colony-stimulating factors such as granulocyte-macrophage colony-stimulating factor (hGM-CSF) or macrophage colony-stimulating factor (hM-CSF). Based on these findings, sera from normal persons and patients with severe infections or various hematological disorders were tested after dialysis using this system in order to determine whether G-CSF levels in sera can be estimated or not. In ten normal persons, five patients with acute myelogenous leukemia (AML M1, M2, and M3), five with myelodysplastic syndrome, and four with chronic myelogenous leukemia, no increases in [3H]thymidine uptake were found within the dose range of 0.4 microliters to 50 microliters. In contrast, linear dose responses parallel to a G-CSF standard curve were observed in one patient with a severe bacterial infection, four with aplastic anemia, two with acute myelomonocytic leukemia (AMMoL) (M4), and two with idiopathic neutropenia tested. From the standard curve, the probable levels of G-CSF were calculated as follows: approximately 200 pg/ml with infection, 130-220 pg/ml with aplastic anemia, 150 and 200 pg/ml with AMMoL, and 1120 and 1200 pg/ml with idiopathic neutropenia. The activities of sera were reduced by the anti-rhG-CSF serum pretreatment in the same way as documented in the case of rhG-CSF. Furthermore, the level in a patient with a severe infection became undetectable soon after elimination of the infection and blood neutrophil counts had returned to normal. These findings indicate that the microbioassay system will be useful for measuring circulating G-CSF levels which would fluctuate in accord with requirements for stimulating neutrophil production or with abnormal production of hG-CSF.

The binding of recombinant erythropoietin (EPO) to fetal mouse liver cells (FMLC) was investigated using a radioiodinated derivative which retained full biological activity. FMLC were fractionated using a preformed Percoll density gradient. Using the fractionated FMLC, the ability to form CFU-E colonies in a semisolid culture was examined, and the binding of [125I]EPO was measured. The highest specific binding of [125I]EPO was observed in a fraction with a density between 1.062 and 1.076 g/ml. The same fraction showed the highest ability to form CFU-E-derived colonies. After suspension culture of FMLC with EPO for 2 days, differentiated erythroid cells with higher density markedly increased. The specific binding of [125I]EPO to these cells almost disappeared with differentiation. Scatchard analysis with cells of the CFU-E-enriched fraction showed a nonlinear curve, suggesting the existence of two classes of binding sites. One binding site was high-affinity (Kd1 = 0.41 nM), and the other low-affinity (Kd2 = 3.13 nM). These results suggest that the expression of EPO receptors on the erythroid cells is highest in CFU-E.

A practically endotoxin-free colony stimulating factor from human urine (CSFHU) was prepared and its granulopoietic effect on normal and cyclophosphamide treated mice was examined. When normal C57BL/6N mice were injected intraperitoneally with 2.5 X 10(6) units/kg of the CSFHU daily for a 5-day period, the numbers of progenitor cells (CFUC) in the femur and spleen were significantly increased. The CFUC in the femur and spleen reached a maximum at day 3 (270%) and day 5 (250%) after the initial injection, respectively. The increase in number of CFUC in the femur exhibited a dose-dependency with respect to the CSFHU and a significant increase was observed even at 4 X 10(5) U/kg (P less than 0.05). However, neither granulocytosis nor monocytosis occurred in normal C57BL/6N mice injected with the CSFHU. In cyclophosphamide induced leukopenic C3H/HeN mice, daily injections of the CSFHU at 2.5 X 10(6) U/kg for 5 days stimulated the restorative granulocyte production (P less than 0.05) as well as the CFUC recovery in both the femur and spleen. These findings suggested that the CSFHU might be involved in granulocyte production in vivo.

A phase I/II study of recombinant human granulocyte colony-stimulating factor (rhG-CSF) in 24 leukemia patients was conducted at our institute. Recombinant human G-CSF (50-200 micrograms/m2/day) was administered i.v. In seven allogeneic bone marrow transplantation (BMT) recipients, treatment with rhG-CSF was started 5 days after BMT. Neutrophils began to increase within 3 days after the start of rhG-CSF administration in five of seven patients. The mean duration necessary for recovery of neutrophils to greater than 500/microliters was 11.3 days after BMT with rhG-CSF; 26.8 days is the figure for recovery without rhG-CSF from Japanese historical data. In seven out of eight patients who received rhG-CSF administration after the first remission-induction chemotherapy, the neutrophil counts increased from less than 300/microliters to greater than 4000/microliters within 10 days. Blasts did not increase in all patients including four acute nonlymphocytic leukemia (ANLL) patients. Severe infections such as septicemia and pneumonia, which were unable to be controlled by antibiotics only, were successfully treated with rhG-CSF and antibiotics. rhG-CSF either stimulated or inhibited myeloid leukemic cells in some refractory cases. Mild bone pain occurred in one patient while receiving rhG-CSF i.v. rhG-CSF seems to have the ability to shorten the period of neutropenia, prevent infections after allogeneic BMT and remission-induction chemotherapy for acute leukemia, and support therapy for infections.

We have investigated direct and monocyte-macrophage (Mono/M phi)-mediated indirect effects of recombinant immune interferon (IFN-gamma) on the growth of established leukemic cell lines (K562, KG1, ML1, HL60, U937, and THP1). The direct antiproliferative effects of IFN-gamma on these leukemic cells were mild or negligible, when estimated by 3H-thymidine incorporation. Indirect effects were assessed by the growth pattern of leukemic cells cocultured with Mono/M phi that were pretreated with INF-gamma. While the leukemic cell growth was slightly suppressed by untreated Mono/M phi, this suppression was significantly augmented by the treatment of Mono/M phi with IFN-gamma (10-10,000 U/ml). In addition, the indirect effects of IFN-gamma on leukemic cell growth were examined at different stages of maturation of Mono/M phi. The augmentation of cytotoxicity was detected only when mature Mono/M phi were treated with INF-gamma. This suggests that IFN-gamma acts on tissue macrophages and augments their cytotoxicity against leukemic cells.

Recombinant human insulin-like growth factor I (IGF-I) increased human and murine erythropoietic colony formation in serum-free culture. In order to investigate the effects of purified factors such as IGF-I on hemopoietic progenitor cells, we have established a serum-free culture system which supports the clonal growth of CFU-E- and BFU-E-derived colonies. Exogenously supplied ingredients were bovine serum albumin (BSA), transferrin, lipid suspensions, 2-mercaptoethanol, and recombinant human erythropoietin (epo). Among these, BSA and cholesterol were found to be essential ingredients. The optimum concentration of BSA sufficient to grow BFU-E was 3%. Erythroid colony and burst formation of human and murine marrow cells was enhanced twofold (p less than 0.05) by a physiological concentration of recombinant human IGF-I. Potentiation was observed in a dose-dependent manner between 10(-9) and 10(-7) M. A few murine CFU-E colonies were formed in the absence of epo. These results suggest that IGF-I has a supportive effect on the proliferation and differentiation of erythroid precursor cells stimulated by epo and that its action is synergistic with that of epo.

When pieces of a human pulmonary carcinoma which was producing colony-stimulating factor (CSF) were transplanted into nude mice a marked granulocytosis of 10(6)/mm3 was observed in the nude mice 4 weeks after the transplantation. In order to elucidate the control mechanism of hemopoiesis, we studied the kinetics of the hemopoietic stem cells of these mice. The number of GM-CFC, CFUE, Meg-CFC and CFUS in the whole body increased to 6 times, 8 times, 6 times and 36 times as much respectively as those of the control mice. The increase in the number of these progenitors occurred mainly in the spleen. The sera of these mice contained only CSF. The cycling fraction of GM-CFC and CFUS determined by thymidine suicide technique was increased. From these findings, it could be considered that CSF produced in these mice acted directly or indirectly on CFUS inducing their differentiation into committed stem cells.

We conducted a randomized crossover study comparing the hemopoietic effect of partially purified human urinary colony-stimulating factor (CSF-HU, an active drug) and human serum albumin (HSA, a control drug) in 24 patients with malignant lymphoma, solid tumors, or multiple myeloma who were receiving two consecutive courses of the same chemotherapeutic regimen. Patients received daily 2-4 X 10(6) units of CSF-HU or an equal amount of protein HSA for five days after the end of the courses of chemotherapy. Assignment to CSF-HU or HSA was determined by the envelope method. The average number of blood granulocytes of 24 cases on day 7 after chemotherapy was 2116 +/- 1649 in CSF-HU-infused courses, which was significantly higher than in HSA-infused courses (1520 +/- 1022) (p less than 0.05). The average time that patients had fewer than 2000 granulocytes/mm3 was 7.6 +/- 4.4 days in CSF-HU-infused courses and 10.3 +/- 5.0 days in HSA-infused courses (p less than 0.02). Fever greater than 38 degrees C was the most frequent side effect, occurring in 32% of the patients receiving CSF-HU infusions. A reduction in the neutropenic interval in CSF-HU-infused courses was observed in patients with fever, as well as in those without fever. Infusions of CSF-HU did not change the number of other hematological parameters, such as erythrocytes, platelets, monocytes, and lymphocytes. These results suggest that CSF-HU infusions may partially protect the patients from granulocytopenia after anticancer chemotherapy.

Specific high-affinity receptor(s) for insulin-like growth factor I have been identified in fetal mouse liver cells (FMLC) rich in late erythroid progenitors (CFU-E). Competition for [125I]IGF-I binding by IGFs and insulin demonstrated the presence of Type-I IGF receptors. Scatchard analysis of the binding data revealed a single class of receptors (Kd, 1.2 nM; R0, 600 sites per cell). Erythroid colony formation and DNA synthesis by these cells were enhanced by IGF-I alone or in combination with erythropoietin (Epo). Subfractionations of FMLC using Percoll density gradients showed that a significant part of [125I]IGF-I binding was observed in the CFU-E-enriched fraction and that the erythroid colony formation was mostly enhanced by IGF-I in the same fraction. IGF-I stimulated the phosphorylation of the beta-subunit of the Type-I receptors. These results indicate that IGF-I modulates the Epo-stimulated proliferation and differentiation of erythroid progenitors via its specific receptors.

A modified culture system has been developed to grow and quantitate megakaryocyte colonies from mouse bone marrow more efficiently than described in other reports. Using this method, it was shown that 30% of CFU-M in normal marrow cells and 70 to 90% of CFU-M in regenerating marrow cells were killed by high specific activity 3H-TdR in vitro. These results indicate that CFU-M are proliferating even in normal adult hemopoietic tissue, but that the proportion of cells that are proliferating is greater in regenerating marrow than in normal intact mice.

In order to clarify the regulation of granulopoiesis by hydrocortisone in humans, we investigated the effect of glucocorticoids on the formation of granulocyte and/ or macrophage colonies. By means of the dual esterase staining techniques applied to whole mount preparations of agar culture dishes, we examined the granulocyte-macrophage colony type. It was revealed that hydrocortisone stimulated the formation of neutrophil-containing colonies and inhibited macrophage colony formation. There was a significant increase in neutrophil colonies when the cells were preincubated with hydrocortisone for only 24 h. Delayed addition of hydrocortisone to the cultures was less effective in increasing the proportion of neutrophil colonies than addition at the beginning of culture. Moreover, addition of hydrocortisone to T-lymphocyte- and phagocyte-depleted bone marrow cells also increased the number of neutrophil colonies and decreased macrophage colonies in the presence of CSF. These results suggest that hydrocortisone may affect granulocyte-macrophage precursors (CFUGM) in an early period of their differentiation.

Sera obtained from seven normal volunteers and 11 patients with aplastic anemia were assayed for two types of colony-stimulating activity (CSA) using monolayer agar cultures containing human unfractionated and monocyte-depleted bone marrow cells. Sera obtained from normal volunteers had low CSA for unfractionated bone marrow cells and no CSA for monocyte-depleted bone marrow cells. On the other hand, sera obtained from patients with aplastic anemia (patients' sera) had moderate CSA for both unfractionated and monocyte-depleted bone marrow cells. The patients' serum CSA titer for unfractionated bone marrow cells rose markedly with the addition of a small amount of diluted control rabbit serum (control serum) into the CSA assay system, while it did not rise with addition of the same amount of diluted rabbit antiserum against partially purified human urinary colony-stimulating factor (CSF) (antiserum). The patients' serum CSA titer for monocyte-depleted human bone marrow cells rose with the addition of control serum as well as antiserum. These findings indicate that the addition of rabbit serum enhances colony formation by human bone marrow cells in the presence of aplastic anemia sera as a source of CSA, and that sera of patients with aplastic anemia contain two antigenically different types of CSA: one that is active on human unfractionated bone marrow cells and partially neutralized by the addition of antiserum, and another that is active on human monocyte-depleted bone marrow cells and is not neutralized by the addition of antiserum.

In order to study the changes in erythroid precursor cells in primary acquired and secondary sideroblastic anemia, bone marrow cells from 4 patients with primary acquired sideroblastic anemia (PASA) and 3 patients with refractory anemia with excess of myeloblasts (RAEM) or erythroleukemia associated with an excess of ringed sideroblasts were cultured for erythroid colony-forming units (CFU-E). The number of CFU-E was markedly decreased in all 7 cases, and erythroid colonies formed consisted exclusively of normal-appearing erythroblasts, while ringed sideroblasts were observed in scattered single erythroblasts or in small aggregates of erythroblasts in primary as well as in secondary sideroblastic anemia. These findings may indicate the presence of 2 populations of erythroid progenitor cells in the bone marrow of patients with primary acquired and secondary sideroblastic anemia. A slight to moderate decrease in granulocyte-macrophage colony-forming units (GM-CFU) was observed in 3 cases of PASA. The decrease in GM-CFU, however, was marked in sideroblastic anemia associated with RAEM or erythroleukemia.

The effects of 12-O-tetradecanoylphorbol-13-acetate (TPA) on human T lymphocyte lymphocyte colony formation in vitro were investigated. The number of T lymphocyte colonies was increased 4-5 times over that of controls by the addition of TPA (10(-7) - 10(-9) M) to phytohemagglutinin (PHA)-containing cultures. Few colonies were observed when stimulated with TPA in the absence of PHA. In the cultures containing a sufficient amount of exogenous T cell growth factor (TCGF), the enhancement of T lymphocyte colony formation by TPA was not observed. TPA enhanced TCGF production by peripheral lymphocytes stimulated with PHA. The optimal concentrations of TPA for T lymphocyte colony formation were similar to those for TCGF production. These findings suggest that TPA enhanced T lymphocyte colony formation by stimulating endogenous TCGF production. Interestingly, T lymphocyte colony formation was not inhibited even at high concentrations of TPA that usually inhibit myeloid and erythroid colony formation. This difference may be due to different sensitivities to TPA between T lymphocyte colony-forming cells and myeloid and erythroid colony-forming cells.

In this paper, a new technique for the cytochemical examination of human hemopoietic cells grown in agar gel is described. All colonies in the agar gel in a plastic dish can be observed as a whole plate preparation after being stained cytochemically. Granulocytes and monocytes/macrophages in a cluster or a colony were easily distinguished from each other by esterase activity in response to naphthol AS-D chloroacetate of alpha-naphthyl acetate. To illustrate this technique's possibilities a time course study on the formation of granulocyte and monocyte/macrophage clusters or colonies from normal marrow cells was performed.

We have previously shown that a factor termed NAP-IF has the capacity to induce neutrophil alkaline phosphatase (NAP) in postmitotic granulocytes (PMGs). Recently, this factor found in cystic fluid of a human squamous cell carcinoma was shown to be identical to granulocyte colony-stimulating factor (G-CSF). In this study we examined whether NAP activity inducible with G-CSF could be modulated by other factors that are present in vivo or those that are known to induce differentiation of hemopoietic cells. Purified natural and recombinant G-CSF (nG-CSF and rG-CSF) induced NAP in PMGs from both normal individuals and patients with chronic myelogenous leukemia. Interferons (IFNs) suppressed expression of NAP by G-CSF. IFN-gamma was a potent inhibitor of G-CSF stimulation: IFN-gamma at 100 U/ml inhibited by greater than 90% the induction of NAP by G-CSF. In contrast, retinol (10(-6) M, a nearly physiological concentration) or all-trans-retinoic acid (10(-6) M) significantly enhanced NAP activity in vitro. Furthermore, the simultaneous addition of 10(-6) M retinol partially reversed the inhibitory action of IFN-gamma on the NAP induction by G-CSF. Our results suggest that NAP activity, which is often abnormal in a variety of diseases, may reflect G-CSF levels in vivo perhaps in concert with a number of other factors including IFNs and retinoids.

Severe hematopoietic injury in mice was induced by using either 5-fluorouracil, adriamycin, mitomycin C, or vinblastine. Daily subcutaneous administration of purified human recombinant granulocyte colony-stimulating factor (rG-CSF; 0.3-10.0 micrograms/day) markedly accelerated recovery from the drug-induced granulocytopenia in a dose-dependent manner, as reported previously. On the other hand, daily intraperitoneal administration of dolichyl phosphate (Dol-P) also enhanced granulopoiesis to accelerate recovery from granulocytopenia, although the effect of Dol-P was relatively moderate as compared with that of rG-CSF. A synergistic recovery of granulopoiesis was observed when Dol-P was administered together with rG-CSF to the mice treated with anti-cancer drugs. Joint use of Dol-P (1 mg/day) and rG-CSF (0.3 micrograms/day) was as effective as a higher dose of rG-CSF (3 micrograms/day). Joint use of Dol-P (1 mg/day) and rG-CSF (3 micrograms/day) was sometimes more effective.

We have reported that the exogenous addition of dolichyl phosphate (Dol-P) enhances the colony-forming capacities of early erythroid progenitors (BFU-E), late erythroid progenitors (CFU-E), and granulocyte-macrophage progenitors (CFU-GM) in adult mouse bone marrow, and that dolichol (Dol) enhances that of only CFU-E (Int. J. Cell Cloning 3:313, 1985). Compactin (2.5-10 microM), a specific inhibitor of mevalonate biosynthesis that causes a decrease of endogenous Dol biosynthesis, inhibited colony formation of CFU-GM. Exogenous addition of Dol-P partially prevented this inhibition, but Dol and the other mevalonate metabolites, such as cholesterol, coenzyme Q10, and isopentenyladenine, could not. In addition, we have found that the colony-forming capacity of CFU-E in fetal mouse liver was not enhanced by exogenous Dol or Dol-P. But the decrease of colony formation or DNA synthesis of fetal CFU-E in the presence of compactin was prevented by the exogenous addition of Dol or Dol-P.

Marked loss of colony stimulating activity was noted when the highly purified human urinary colony stimulating factor (CSFHU-5) dissolved in 0.1 M Tris-HCl buffer (pH 7) was passed through ethylene oxide sterilized Millipore filters. The addition of 0.5 mg/ml polyethylene glycol (PEG) or 10 mg/ml bovine serum albumin (BSA) to the CSFHU solutions resulted in almost complete protection against the CSF loss. A less pure fraction (CSFHU-2) did not lose activity upon filtration. More than 90% of the 125I-CSFHU dissolved in 0.1 M Tris-HCl buffer (pH 7) bound to the ethylene oxide sterilized Millipore filters. The addition of PEG or BSA to the 125I-CSFHU solutions resulted in almost complete prevention of 125I-CSFHU binding to the membranes. These findings indicate that Millipore membranes can bind significant quantities of the highly purified CSFHU, and the addition of PEG to all buffer systems is a useful way to prevent the CSF loss throughout the purification.

Human urinary colony-stimulating factor (CSF-HU) has been highly purified using procedures containing DEAE cellulose, phenyl Sepharose CL-4B, Sephadex G-200, hydroxylapatite, and high performance liquid chromatography. The final preparation had a specific activity of 3.3 X 10(7) U/mg protein. Although the purified CSF-HU was not active on human monocyte-depleted bone marrow cells, it stimulated human peripheral blood monocytes obtained from five healthy volunteers to produce human active granulocytic colony-stimulating factor (G-CSF), which stimulated human monocyte-depleted bone marrow cells to form granulocytic colonies. The human G-CSF-producing activity of CSF-HU was not neutralized by polymyxin B, which is known to inhibit the effect of endotoxin. Newly produced G-CSF had an approximate molecular weight of 24,000 daltons as judged by chromatography on Sephadex G-150. These results indicate that CSF-HU stimulates human monocytes to produce human G-CSF in vitro.

In order to study the pathogenesis of aplastic anemia in man, hemopoietic stem cells were investigated in 'aplastic mice' the aplasia being induced by the immunological method. C3H/He (H-2k, Mlsc) received 600 rad whole body x-irradiation followed by the transplantation of 10(7) lymph node cells prepared from B10.BR mic e (H-2k, Mlsb). The C3H/He mice developed pancytopenia and marrow hypoplasia 21 days after these treatments. The total number of nucleated cells, CFU-S and CFU-C in the marrow and the wet weight and CFU-C of the spleen were markedly reduced. These findings are consistent with those of aplastic anemia in man and the model may provide a useful tool for the investigation of the pathogenesis of this anemia. Control mice that received irradiation only recovered from the damage 21 days later, while control mice that receive lymph node cells only showed no hematological changes.

Specific binding sites for erythropoietin (Epo) were shown in normal and anemic rat bone marrow cells using [125I]labeled human recombinant Epo. When rats were treated once or several times with phenylhydrazine or malotilate, or by phlebotomy, the serum Epo level determined by RIA began to increase rapidly. Thereafter, both the number of erythroid colony-forming unit (CFU-E)-derived colonies and the Epo binding capacity of bone marrow cells increased almost simultaneously in response to induced anemic states, suggesting that the amount of Epo binding in bone marrow cells may reflect in vivo erythropoiesis. Scatchard analysis of the binding data from normal rats revealed the presence of a single class of binding sites (Kd = 0.18 +/- 0.04 nM, 38 +/- 5 sites/cell). In anemic states, the apparent average receptor number per cell increased (52-62 sites/cell) without changing in binding affinity toward Epo. Furthermore, [125I]Epo was cross-linked to the cell surface molecule of approximately 165 kd in nonreducing conditions and 75 kd in reducing conditions. Autoradiographic analysis indicated that Epo receptors were distributed on immature erythroid cells. Proerythroblasts were the most heavily labeled, whereas orthochromatic erythroblasts and cells of myeloid and lymphoid lineages were not labeled. Calculations based on Scatchard and autoradiographic analysis showed that proerythroblasts have 390 receptor sites per cell, twice as many as basophilic or polychromatophilic erythroblasts have. These results are consistent with the stage-specific action of Epo in physiological differentiation of erythroid cells.

The effect of recombinant human erythropoietin (Ep) and granulocyte colony-stimulating factor (G-CSF) on colony formation by human hemopoietic progenitors was examined in a methylcellulose culture system. In the serum-containing culture system, granulocyte-macrophage (GM) colonies and erythroid bursts were formed by non-phagocytic mononuclear cells only in the presence of Ep. To exclude the effect of the serum, which may have hemopoietic factors, we replaced the serum with bovine serum albumin, transferrin, and lipids. In serum-free culture, recombinant Ep supported erythroid colony formation, but not erythroid burst formation. While G-CSF could support the proliferation of macrophages (30%) as well as neutrophils in the presence of fetal calf serum (FCS), it supported mainly neutrophils (97%) in serum-free culture. In this culture system, G-CSF could not induce burst formation in the presence of Ep. By using a serum-free culture system, we found that human G-CSF is a lineage-specific hemopoietic factor which acts on granulocyte-committed progenitor cells and not on early erythroid progenitor cells.

To study the pathogenesis of aplastic anemia, phagocytic cells from the bone marrow of patients with aplastic anemia were investigated. From 10 of 10 patients with this disease the concentration of CFUc in the bone marrow cells significantly increased after the removal of phagocytic cells. On the other hand, in bone marrow from only 4 of 27 controls, 22 non-aplastic anemia patients and 5 normal volunteers, the concentration of CFUc increased after the same treatment. From these findings, it may be inferred that phagocytic cells from the bone marrow of patients with aplastic anemia suppress proliferation and differentiation of granulocytic progenitor cells, and that this phenomenon is related to the development of this disease.