Quantitative responsiveness of murine hemopoietic populations in vitro and in vivo to recombinant multi-CSF (IL-3)

Purified, bacterially synthesized, recombinant Multi-CSF (rMulti-CSF) exhibited in vitro proliferative effects identical to those of native Multi-CSF on a broad range of progenitor cells from normal murine bone marrow and on the continuous cell line FD-CP1. The injection into normal adult mice of 6-200 ng purified rMulti-CSF three times daily for six days induced dose-related rises in peritoneal macrophages, neutrophils, and eosinophils, and in the spleen weight and content of megakaryocytes, mast cells, and progenitor cells, some responses being detectable with the lowest dose injected. No Multi-CSF was detected in the serum, organ extracts, or organ-conditioned medium from normal or endotoxin-injected mice. While the data indicate that even small doses of injected rMulti-CSF can produce detectable hemopoietic changes in normal adult mice, the pattern of these changes and the failure to detect Multi-CSF in vivo raise doubts that Multi-CSF plays a significant role as a regulator of hemopoiesis in normal adult life.     

Nonneoplastic hematopoietic myeloproliferative syndrome induced by dysregulated multi-CSF (IL-3) expression

Post 5-fluorouracil-treated murine marrow cells were infected with a retroviral vector (MPZen) bearing a multi-potential colony stimulating factor (Multi-CSF) cDNA insert and then transplanted into lethally irradiated syngeneic recipients to study the effects of autocrine production of Multi-CSF in normal hematopoietic cells. Extremely high levels (14,000 U/mL) of Multi-CSF were detected in the sera and in media conditioned by various hematopoietic tissues of the transplanted animals. While spleen, peritoneal, and peripheral blood cellularity increased approximately 10-fold, 10-fold, and 50-fold, respectively, bone marrow cellularity decreased twofold. Progenitor numbers were depressed twofold in the bone marrow but elevated more than 100-fold in the spleen and peritoneum. The majority (80%) of transplanted mice died within 5 weeks of transplantation and showed extensive neutrophilic infiltration of the spleen, lung, liver, and muscle, often with mast cell foci; a phenomenon also seen in the skin and intestine. Neither the infected cells from hematopoietic tissues of the primary mice, nor autonomous mast cell-lines that grew from these cells in liquid culture produced any overt disease when transplanted into normal or sublethally irradiated secondary recipients. In contrast, injection into mice of autonomous FDC-P1 cells transformed by the same retroviral construct led to tumor formation in vivo within 4 weeks. Thus, dysregulated Multi- CSF expression by normal hematopoietic cells produces a fatal but nonneoplastic myeloproliferative syndrome.     

Effects of purified bacterially synthesized murine multi-CSF (IL-3) on hematopoiesis in normal adult mice

Normal adult C57BL, BALB/c, and C3H/HeJ mice were injected intraperitoneally three times daily for up to 6 days with 102,000 U (200 ng) per injection of purified, bacterially synthesized, Multipotential colony-stimulating factor (CSF) (Interleukin-3) (rMulti- CSF) and compared with control mice injected with serum/saline with or without added endotoxin (1 ng/mL). Mice injected with rMulti-CSF exhibited tenfold rises in blood eosinophil and twofold to threefold rises in neutrophil and monocyte levels. The spleens from mice injected with rMulti-CSF showed a 50% increase in weight, elevated levels of maturing granulocytes, eosinophils, nucleated erythroid cells and megakaryocytes, and up to 100-fold rises in mast cells. Progenitor cell frequencies in the spleen were elevated sixfold to 18-fold. No significant changes were observed in the marrow. Sixfold to 15-fold rises were observed in peritoneal cell populations of mice injected with rMulti-CSF with evidence of increased peritoneal macrophage phagocytic activity. Livers of C57BL mice, but not of the other strains, exhibited increased numbers of infiltrating hematopoietic cells whereas rises in mast cell numbers were observed in the mesenteric lymph node, skin, and gut in BALB/c and C3H/HeJ mice. Endotoxin was excluded as being responsible for the observed changes except possibly those involving peritoneal macrophage phagocytic activity. The results indicate that the injection of normal mice with rMulti-CSF significantly stimulates the same types of hematopoietic populations as are stimulated in vitro by Multi-CSF and indicate that this and other CSFs should be useful in stimulating hematopoietic repopulation and functional activity in vivo.     

Expression of human interleukin-3 (multi-CSF) is restricted to human lymphocytes and T-cell tumor lines

While the cellular sources for granulocyte-macrophage colony- stimulating factor (GM-CSF) are known to be widely distributed among several cell types, interleukin-3 (IL-3) gene expression has been demonstrated in only certain T-cell clones and in blood mononuclear cells stimulated with phytohemagglutinin (PHA) and phorbol-myristate- acetate (PMA). To determine which blood cells were responsible for this expression, we fractionated PHA/PMA-stimulated mononuclear cells and identified T lymphocytes as the source of IL-3 mRNA. Low-level IL-3 expression was detected as well in several stimulated human T-cell lines. Hematopoietic stromal cells such as fibroblasts and endothelial cells could not be induced to express IL-3 mRNA. The kinetics of IL-3 mRNA induction in mononuclear cells and lymphocytes stimulated with PHA/PMA or anti-CD3 monoclonal antibody (MoAb) and interleukin-1 (IL-1) were similar to those observed for GM-CSF expression.     

Therapeutic efficacy of recombinant interleukin-6 (IL-6) alone and combined with recombinant human IL-3 in a nonhuman primate model of high-dose, sublethal radiation-induced marrow aplasia

Using a nonhuman-primate model of radiation-induced bone marrow aplasia, we examined whether the single, concomitant, or sequential administration of recombinant human interleukin-3 (IL-3) and IL-6 would promote bone marrow regeneration measured by an increase in circulating platelets (PLT) and neutrophils (PMN). Rhesus monkeys were irradiated at 450 cGy and were randomly assigned to one of five treatment protocols, receiving IL-6; IL-3; combined IL-6 and IL-3; sequential IL- 3 and IL-6; or human serum albumin (HSA) as a control. Cytokines or HSA were administered at total dosages of 15 micrograms/kg/day. Complete blood counts and white blood cell differentials were monitored for 60 days postirradiation. Both IL-3 and IL-6 significantly enhanced the regeneration of PLTs and decreased the duration of thrombocytopenia (P = .005) without affecting PMN recovery. The radiation-induced anemia that was observed in the HSA-treated controls was less severe and resolved more quickly in the IL-6 treated animals. Sequential IL-3/IL-6 significantly increased the production of PLTs when compared with the HSA-treated controls (P = .003) and monkeys receiving concomitant IL- 3/IL-6 (P = .041) but did not alter PMN levels significantly (P = .80). Coadministration of IL-6 and IL-3 did not enhance PLT but improved PMN recovery over IL-6 alone. In this primate model of marrow aplasia, IL-6 significantly enhanced the regeneration of PLTs but had no significant effect on PMN production, and did not exacerbate radiation-induced anemia. Furthermore, the use of sequentially administered IL-3 and IL-6 may improve PLT recovery as compared with concurrent IL-3/IL-6 administration, although this protocol is not significantly different in effect than either cytokine alone.     

Influence of cytokines (IL-1α, IL-3, IL-11, GM-CSF) on megakaryocyte-fibroblast interactions in normal human bone marrow

Abstract: The evolution of myelofibrosis accompanying chronic myeloproliferative disorders (CMPDs) is often linked with megakaryopoiesis. However, it is not known whether or to what extent megakaryocytes of normal human bone marrow are capable of stimulating fibroblast growth. For this reason, an in vitro study was performed to elucidate possible cytokine-dependent interactions between megakaryocytes and fibroblasts derived from healthy volunteers. Fibroblast growth was significantly promoted by the presence of megakaryocytes and modulated by additional application of various cytokines. While recombinant human (rh) interleukin (IL)-1α had no obvious effect on fibroblast proliferation, a slight increase was detected on adding granulocyte-macrophage colony stimulating factor (rhGM-CSF). Application of rhIL-3 caused a significant increase in the number of fibroblasts. In contrast, administration of rhIL-11 suppressed the megakaryocyte-dependent growth-promoting effect and co-stimulation with rhIL-3 led to a significant decrease of fibroblast number in comparison to rhIL-3-stimulated co-cultures. Inhibition of cell-cell contact in unstimulated, as well as in rhIL-3-stimulated co-cultures led to a conspicuous impairment of fibroblast growth. A similar effect was observed when neutralizing antibodies directed against platelet-derived growth factor (PDGF) and transforming growth factor (TGF)β₁ were added to rhIL-3-stimulated cultures. Our findings are in keeping with the assumption that interactions between megakaryocytes and fibroblasts involve in cytokine-mediated functional network regulated by factors such as spatial relationship, cytokine stimulation, and low concentrations of mediators, particularly PDGF and TGFβ. In this complex system rhIL-3 seems to play a crucial role in the promotion of these various interrelationships.     

Interleukin-1 (IL-1) inhibits growth of cytomegalovirus in human marrow stromal cells: inhibition is reversed upon removal of IL-1.

A Toledo strain cytomegalovirus (CMV) containing the gene for green fluorescent protein (GFP) under the control of elongation factor-1 promoter was used to study infection of human marrow stromal cells. Two stromal cell lines were used: HS-5, which secretes copious amounts of known cytokines and interleukins; and HS-27a, which does not secrete these activities. CMV growth and spread was monitored by counting green plaques and quantitating GFP intensity. Initial studies indicated that, whereas HS-5 and 27a have similar susceptibilities to infection, as evidenced by the same number of GFP+ cells at day 2, HS-5 appears more resistant to growth and spread of CMV. Furthermore, conditioned media from HS-5 (HS-5 CM) inhibited CMV plaque formation in HS-27a, suggesting that factors secreted by HS-5 are responsible for limiting CMV growth. Neutralizing antibodies against interleukin-1alpha (IL-1alpha) and IL-1beta completely blocked the ability of HS-5 CM to limit viral growth, suggesting that IL-1, which is known to be present in HS-5 CM, is responsible for this effect. When exogenous IL-1beta was added to CMV-infected HS-27a, both the number of plaques and the intensity of GFP was significantly reduced in IL-1-treated HS-27a compared with untreated HS-27a (the number of plaques by day 18 was 20 +/- 3 v 151 +/- 12/well, respectively; GFP intensity was 535 +/- 165 v 6,516 +/- 652/well, respectively, in 4 separate experiments). At day 21, when IL-1beta-treated, CMV-infected cultures were passaged and then cultured in the absence of IL-1beta, CMV growth progressed with the kinetics of the original untreated culture, indicating that the IL-1beta effect is reversible. Because HS-27a expresses the type I IL-1 receptor, we speculate that the antiviral effects are mediated through IL-1-induced changes in cellular gene expression. DNA chip analysis of mRNA from IL-1beta-treated and nontreated HS-27a cells has identified some candidate molecules.     

Human marrow stromal cells: response to interleukin-6 (IL-6) and control of IL-6 expression

Production of interleukin-6 (IL-6) by marrow stromal cells from human long-term marrow cultures and from stromal cells transformed with simian virus 40 was examined. As with other cultured mesenchymal cells, unstimulated stromal cells produced undetectable amounts of IL-6 mRNA when assayed by Northern blots. However, within 30 minutes after exposure of transformed marrow stromal cells to the inflammatory mediators, recombinant human interleukin-1 alpha (IL-1 alpha) or recombinant human tumor necrosis factor alpha (TNF alpha), significant increases in IL-6 expression were observed. The time course of IL-6 mRNA upregulation in transformed marrow stromal cells with IL-1 alpha and TNF alpha differed: The maximal response to TNF alpha was observed at 30 minutes whereas that to IL-1 alpha occurred at 8 hours. Although IL-6 at a concentration of 500 U/mL was inhibitory to adherent transformed marrow stromal cell proliferation, a concentration- dependent stimulation of anchorage-independent colony growth was observed when the cells were plated in semisolid medium with IL-6. The stromal cell colony-stimulating effect of IL-6 was abrogated by a neutralizing antibody to IL-6. Moreover, the heteroserum with anti-IL-6 activity and two anti-IL-6 monoclonal antibodies partially blocked autonomous and IL-1 alpha-induced colony formation, suggesting that colony formation by transformed marrow stromal cells may require IL-6. Clonal-transformed stromal cell lines were derived from the anchorage- independent stromal cell colonies. Both IL-6 mRNA and protein were constitutively produced at high levels. The addition of IL-6 to either long-term marrow culture adherent cells or transformed marrow stromal cells downregulated the expression of collagen I, a major stromal cell matrix protein. Thus, IL-6 affects proliferation of stromal cells and influences their production of extracellular matrix, suggesting that IL- 6 may have indirect as well as direct influences on hematopoietic cell proliferation.     

The effects of interleukin 3 (IL-3) on cells responsive to macrophage colony-stimulating factor (CSF-1) in liquid murine bone marrow culture

We have examined variation between inbred mouse strains in the proliferative response of bone marrow cells in liquid culture to macrophage colony-stimulating factor (CSF-1) and interleukin 3 (IL-3). In all mouse strains, thymidine incorporation was stimulated by CSF-1 and, after an initial lag period, it reached a peak on day 5. In contrast, two mouse strains, A/J and Balb c, had much lower proliferative responses to IL-3 than did the other strains. In A/J there was no increase in thymidine incorporation above the initial baseline, although the fall in incorporation seen in the absence of any added growth factor was prevented. IL-3 also prevented the loss of CSF-1 responsiveness observed when A/J bone marrow cells were incubated in medium alone. The lag phase in the response to CSF-1 was progressively abolished following IL-3 pre-treatment. Thus, the data with A/J mice separate two distinct activities of IL-3 and show that proliferation is not required for the synergistic effect exerted by IL-3 on CSF-1-stimulated macrophage generation from bone marrow. In strains in which IL-3 alone was able to stimulate proliferation, its action was not additive with that of CSF-1, and addition of both factors together did not overcome the lag phase. This suggests that the two factors act on the same cell population, and that the known synergistic effect of IL-3 on macrophage colony formation in soft agar does not result from an increase in the initial rate of proliferation. The possibility that the combination of factors might alter the duration of the growth response in vivo is discussed.     

The use of recombinant cytokines for enhancing immunohematopoietic reconstitution following bone marrow transplantation. I. Effects of in vitro culturing with IL-3 and GM-CSF on human and mouse bone marrow cells purged with mafosfamide (ASTA-Z)

Enhanced colony formation (CFU-GM) in vitro was observed in murine and human bone marrow (BM) cells following pre-incubation for 2-3 days with recombinant murine GM-CSF or natural purified murine IL-3, and recombinant human GM-CSF or IL-3, respectively. Pre-incubation in the presence of both GM-CSF and IL-3 produced additive stimulatory effects. BM cells previously treated in vitro with mafosfamide (ASTA-Z) under conditions identical to those used in the purging of autologous BM grafts, also demonstrated an enhanced cumulative response to combinations of GM-CSF and IL-3, with up to 100-fold increase in CFU-GM as compared with controls (p less than 0.001). In mice, the number of CFU-S was also significantly increased (2-20 times) following incubation of unpurged and purged BM cells in murine IL-3 and/or GM-CSF. Interestingly, the frequency of both CFU-GM and CFU-S in BM cells first purged with ASTA-Z and then cultured with both cytokines was significantly higher (p less than 0.01) than that in fresh, intact BM cells. In addition, mice transplanted with unpurged or purged, cytokine cultured syngeneic BM cells exhibited a significantly (p less than 0.01) earlier reconstitution of peripheral white blood cells and of BM CFU-GM, and a significantly enhanced anti-sheep red blood cell plaque-forming cell response. Overall, the data suggest that it might be possible to enhance immunohematopoietic reconstitution in recipients of unmanipulated, as well as ASTA-Z purged autologous BM following short-term culture of BM cells with recombinant colony stimulating factors prior to bone marrow transplantation.     

Detection of interleukin-6 (IL-6) in human bone marrow myeloma cells by light and electron microscopy

We used an Avidin-Biotin peroxidase complex (ABC) immunoperoxidase technique to evaluate the localization of IL-6 of human bone marrow cells in multiple myeloma (MM). The cellular distribution of IL-6 was determined at light and electron microscopic levels. The author's study indicated that cytoplasmic IL-6 was detected only in the myeloma cells of bone marrow cells. Immunoelectron microscopic (immuno-EM) study showed positive reactivity mainly in the perinuclear space (PNS), well-developed rough endoplasmic reticulum (RER), and Golgi area.     

The effects of interleukin 7 (IL-7) on human bone marrow in vitro.

Interleukin 7 (IL-7) stimulates the proliferation of pre-B cells from long-term murine lymphoid cultures and normal bone marrow. In addition, IL-7 stimulates the proliferation of murine T cells, including fetal and adult thymocytes as well as peripheral T cells. Flow cytometry and cell enumeration analyses were carried out on light-density human bone marrow cells incubated in the presence or absence of IL-7. The data showed no evidence for a proliferative effect of IL-7 on B-lineage cells expressing CD24 or on myeloid cells expressing CD15; however, IL-7 did stimulate the growth of T cells expressing CD3. After 16 days of stimulation the number of CD3+ cells in marrow cultures increased 350% in the presence of IL-7. In contrast, cultures incubated in the absence of IL-7 showed a 50% decrease in the number of T cells, with a preponderance of myeloid lineage cells. Flow cytometry indicated that cells from IL-7-stimulated cultures were mature T cells because they also expressed cell surface antigens for either CD4 or CD8. These studies show that in contrast to the murine system, IL-7 does not appear to stimulate the growth of human pre-B cells from adult human bone marrow. This is consistent with other experiments that suggest that human pro-B cells and not human pre-B cells respond to IL-7. It appears that IL-7 preferentially promotes the growth of T cells from human marrow.     

Human interleukin-5 (IL-5) regulates the production of eosinophils in human bone marrow cultures: comparison and interaction with IL-1, IL-3, IL-6, and GMCSF

Recombinant human interleukin-5 (rhIL-5), in either liquid or semi- solid cultures, selectively induced eosinophil production from normal human bone marrow, with no activity on other cell lineages. The time course of eosinophil production induced by murine IL-5, rhIL-3, and rh granulocyte-macrophage colony stimulating factor (GMCSF) was similar to rhIL-5. The rate of eosinophil maturation in vitro was independent of the stimulating cytokine, mature eosinophils being produced after 4 to 5 weeks in liquid culture with each of these cytokines. The eosinophils produced in response to each cytokine were morphologically indistinguishable, and had the ultrastructural features of maturity except that the electron-dense material in the granules had not formed into crystalline cores. Neither rhIL-1 nor rhIL-6 alone, or in combination with rhIL-5 or rhIL-3, induced eosinophil differentiation or proliferation under the conditions used. rhIL-3 and rhGMCSF induced more eosinophil colonies than rhIL-5, rhIL-5 had an additive, not synergistic, effect on eosinophil colony production when combined with either rhIL-3 or rhGMCSF, suggesting that rhIL-5 stimulates a smaller and possibly different population of eosinophil progenitors. However, rhIL-5 induced the greatest eosinophil production in liquid cultures, suggesting that although it may act on a smaller population of precursors, it is able to stimulate more proliferative steps than either rhIL-3 or rhGMCSF.     

The effect of recombinant human granulocyte-macrophage colony-stimulating factor (rGM-CSF) on 3'-azido-3'-deoxythymidine (AZT)-mediated biochemical and cytotoxic effects on normal human myeloid progenitor cells

Administration of 3'-azido-3'-deoxythymidine (AZT) to patients with acquired immunodeficiency syndrome (AIDS) causes significant anemia and neutropenia. The bone marrow cytotoxicity of AZT has been attributed to deoxyribonucleotide pool perturbations that might result in impaired DNA synthesis in normal bone marrow elements. We examined the effect of human recombinant granulocyte-macrophage colony-stimulating factor (rGM-CSF) on AZT-mediated biochemical perturbations and in vitro growth inhibition of normal bone marrow myeloid progenitor cells. Exposure of nonadherent, bone marrow mononuclear cells (BMMC) to 100 ng/ml of rGM-CSF for 6 h resulted in approximately twofold increments in the mean intracellular deoxycytidine triphosphate (dCTP) and thymidine triphosphate (dTTP) levels. Administration of 10 microM AZT alone to BMMC for 6 h markedly reduced dCTP and dTTP levels and generated significant levels of AZT triphosphate (AZT-TP). Coadministration of rGM-CSF (100 ng/ml) along with AZT (10 microM) partly restored dCTP and dTTP levels and significantly reduced AZT-TP levels. Furthermore, simultaneous exposure of BMMC for 4 h to 100 ng/ml of rGM-CSF reduced the mean DNA incorporation of [3H]AZT (10 microM) from 27.2 to 19.1 pmol/micrograms of DNA. Additionally, the inhibitory effects of AZT (10 microM) on granulocyte-macrophage colony-forming unit (CFU-GM) colony growth were significantly reduced in the presence of 100 ng/ml of rGM-CSF. These in vitro studies suggest that rGM-CSF partly corrects AZT-mediated biochemical perturbations as well as reduces the cytotoxicity of AZT in normal human bone marrow myeloid progenitor cells.     

Implication of Granulocyte-Macrophage Colony Stimulating Factor (GM-CSF) and Interleukin-3 (IL-3) in Children with Acute Myeloid Leukaemia (AML); Malignancy

Granulocyte-macrophage colony stimulating factor (GM-CSF) and Interleukin-3 (IL-3) are increasingly used to stimulate granulopoiesis in neutropenic patients but these are rarely used in the lights of knowledge of the endogenous CSF-levels. In this study we measured serum levels of GM-CSF and IL-3 at diagnosis and after remission in children with acute leukaemia, using an enzyme linked immuno-sorbent assay (ELISA) techniques in 14 patients with acute myeloid leukaemia (AML) and 27 patients with acute lymphoblastic leukaemia (ALL). Twelve healthy age-matched children were used as a reference group. AML patients showed a highly significant increase in serum levels of GM-CSF and IL-3 before induction of therapy (p < 0.0001) compared to the reference control group, with a highly significant decline of both GM-CSF and IL-3 (p < 0.0001) after successful remission. On the other hand, ALL patients showed no significant elevation of GM-CSF and IL-3 at diagnosis (p > 0.5), with no significant difference between preinduction and postinduction serum levels of either (p > 0.5). Since these cytokines are known to be fundamental for the growth of AML cells, we postulate that the pretreatment levels of both GM-CSF and IL-3 could play a role in the pathogenesis of AML.     

Lithium enhancement of megakaryocytopoiesis in culture: mediation via accessory marrow cells

To examine the effect of lithium (Li) on early megakaryocytopoiesis, murine marrow megakaryocytic (CFU-M) and granulocyte-macrophage (CFU-C) progenitors were assayed in vitro with and without addition of lithium chloride (LiCl) to culture. At 2 mM LiCl, the numbers of CFU-M- and CFU- C-derived colonies were increased to 146% +/- 8% and 128% +/- 6% of controls, respectively (p less than 0.005). Enumeration of megakaryocytes per colony showed a 78% increase of colonies (p less than 0.05) containing from 6 to 22 cells, suggesting an increased proliferative capacity of CFU-M in the presence of LiCl. Conditioned media from spleen cells cultured in the presence of both pokeweed mitogen (PWM-SCM) and 2 mM Li increased the numbers of CFU-M and CFU-C to 157% +/- 8% and 183% +/- 8%, respectively (p less than 0.025), compared to control cultures stimulated by PWM-SCM alone. Since the production of active colony-stimulating activities (CSA) from mitogen- stimulated conditioned media requires T lymphocytes, we hypothesized that the enhancement of the growth of early hematopoietic progenitors in marrow cultures was due to a Li-induced CSA production by accessory marrow cells, rather than a direct effect of Li on stem cells. To test this, cyclosporin-A (CyA), a T-lymphocyte function inhibitor known to suppress CSA production in PWM-SCM, was added to marrow cultures in the presence of 2 mM Li. CyA (3 micrograms/ml) abrogated the Li-induced enhancement of CFU-M and CFU-C growth, but had no effect on colony formation when added alone. The data suggest that the Li-induced enhancement of early megakaryocytopoiesis and granulocytopoiesis is due to local production of CSA(s) by an accessory cell population and requires the integrity of T-lymphocyte function.     

Biosynthetic (recombinant) human granulocyte-macrophage colony- stimulating factor: effect on normal bone marrow and leukemia cell lines

To examine the biologic properties of the molecule encoded by the human gene for granulocyte-macrophage colony-stimulating factor (GM-CSF), we expressed the cloned complementary DNA (cDNA) in transfected monkey COS cells and purified the resultant protein. Purified biosynthetic human GM-CSF was added to cultures of normal hematopoietic progenitor cells in semisolid media, and the resulting colonies were characterized cytochemically. Non-adherent light-density bone marrow cells from healthy adult volunteers were maximally stimulated with GM-CSF (approximately 250 pmol/L, and four types of colonies were consistently identified by aspirating the individual colonies and staining with a triple stain for specific and nonspecific esterases and eosinophilic granules. Pure neutrophilic granulocyte (G), mixed granulocyte- macrophage (GM), pure macrophage (M), and pure eosinophil (EO) colonies were observed, the mean incidences on day 8 being 70%, 20%, 5%, and 5%, and on day 14, 7.5%, 16.6%, 50.9%, and 25.0%, respectively. In all types of colonies, complete maturation to segmented forms or typical macrophages was detected. GM-CSF did not enhance the growth of BFU-E from normal peripheral blood buffy coat cells in the simultaneous presence of erythropoietin alone or erythropoietin with purified erythroid-potentiating activity. GM-CSF stimulated HL-60 and KG-1 colony formation twofold and fivefold, respectively; consistent differentiation induction towards monocytic and eosinophilic lineages was observed in HL-60 but not in KG-1. These in vitro findings indicate that GM-CSF is a multilineage stimulator for progenitor cells of G, GM, M, and EO colonies.     

Clinical application of recombinant human colony-stimulating factor (rhG-CSF) in bone marrow transplantation]

Human granulocyte colony-stimulating factor (G-CSF) specifically stimulates granulocyte production and enhances functions of mature granulocytes. It also proliferates myeloid leukemic cells. The molecule was purified and molecularly cloned in 1986. In this study, 51 patients received single daily injections with recombinant human (rh) G-CSF (2-20 micrograms/kg) after bone marrow transplantation were compared with control 36 patients. Blood neurophilic recovery was accelerated remarkably by rhG-CSF administration without any adverse effects including delay in reticulocyte and platelet recoveries. Furthermore, the duration for the management in laminar airflow room and febrile days greater than or equal to 38 degrees C were shortened in rhGS-CSF treated group. In 6 patients with myeloid leukemia in relapse, we also tested the effects of the rhG-CSF-combined conditioning regimen. In all the patients tested, the leukemic cells responded to rhG-CSF. Clinically no relapse was observed and 3 patients are well on day 224-357 for the time being. These findings indicate that rhG-CSF is very useful in bone marrow transplantation.     

Effects of human marrow stromal cells on proliferation by human granulocytic (GM-CFC), erythroid (BFU-E) and mixed (Mix-CFC) colony-forming cells

S ummary . We have investigated the effects of the major components of bone marrow stroma (fibroblasts, fat cells, macrophages and endothelial cells) on colony-forming haemopoietic precursor cells. Selective cultures of the different stromal cell types, grown to confluence, were used as underlayers for agar or methylcellulose cultures containing bone marrow cells. In different experiments, colony-stimulating factor (CSF) was added to stimulated granulocyte-macrophage colony-forming cells (GM-CFC), erythropoietin was used to induce erythroid burst (BFU-E) formation or erythropoietin and medium conditioned by leucocytes in the presence of phytohae-magglutinin (PHA-LCM) was added to induce the formation of colonies of mixed myeloid lineage (Mix-CFC)
Fibroblasts grown from human marrow enhanced granulopoiesis when CSF was present in the cultures but suppressed the formation of BFU-E and mixed colonies. Treatment of the cultures with methylprednisolone induced the formation of fat cells in the fibroblast cultures and prevented both the fibroblast-mediated enhancement of granulopoiesis and the fibroblast-mediated suppression of erythropoiesis. Stromal macrophages reduced granulocyte colony formation but had little effect on the proliferation of BFU-E or mixed colony-forming cells. Endothelial cells stimulated granulopoiesis by releasing CSF into the culture supernatant: supernatant from endothelial cell cultures had no marked effects on either BFU-E or Mix-CFC.
We conclude that different components of marrow stroma have contrasting effects on erythropoiesis and granulopoiesis; thus, marrow stroma may regulate the expression of stem cell differentiation in vivo.     

Culture of human fetal B-cell precursors on bone marrow stroma maintains highly proliferative CD20dim cells

Growth of human B-cell precursors (BCP) was achieved by plating fetal CD10+ surface-mu (s mu)- cells in liquid medium onto bone marrow- derived fibroblastic stromal cell layers deprived of hematopoietic cells. Proliferation of the fetal BCP was strongly potentiated by the addition of interleukin-7 (IL-7) to the cultures. Cultures included both a stroma-adherent and -nonadherent fraction of lymphoid cells, allowing us to expand the number of input BCP to 13-fold. In the presence of exogenous IL-7, proliferation was dose-dependent relative to the number of stromal cells, demonstrating that soluble IL-7 does not act alone to promote optimal growth. We further showed that the lymphoid cells recovered remain CD10+ sIg- BCP and that most cells expressed the maturation-associated CD20 antigen when IL-7 was added to the cultures. Whereas both freshly isolated CD20- and CD20bright BCP proliferated in the presence of stroma, we observed that high- proliferative capacity CD20dim cells were maintained in the cultures. Finally, CD20dim sorted cells were shown to subsequently acquire high levels of CD20 expression in culture, thus demonstrating a partial maturation sequence. The present culture system thus represents a useful model for studying the regulatory signals in early human B lymphopoiesis.