The regulation of haemopoiesis in long-term bone marrow cultures: I. Role of L-cell CSF

The effects of L-cell conditioned medium which contains granulocyte/macrophage colony stimulating factor (CSF); of highly purified L-cell CSF; and the antiserum directed against L-cell CSF, have been investigated in long-term murine bone marrow cultures. Treatment of cultures with CSF containing conditioned medium led to a rapid decline in haemopoiesis. However, this inhibition of in vitro haemopoiesis is probably caused by materials other than CSF, since the addition of highly purified L-cell CSF had no appreciable effect upon long-term haemopoietic cell proliferation or differentiation. Furthermore, the inhibitory activity of L-cell conditioned medium was not abrogated following neutralization of the CSF activity by CSF antiserum. The direct addition of CSF antiserum did not inhibit granulocyte or macrophage formation. These results suggest that long-term cultures of murine marrow cells may show extensive interactions with stromal cells which are not influenced by exogenous stimulatory or inhibitory factors.     

The nature of 12-O-tetradecanoylphorbol-13-acetate (TPA)-stimulated hemopoiesis, colony stimulating factor (CSF) requirement for colony formation, and the effect of TPA on [125I]CSF-1 binding to macrophages

The tumor-promoting phorbol diester, 12-O-tetradecanoylphorbol-13-acetate (TPA) was found to act both independently of and synergistically with the mononuclear phagocyte specific colony stimulating factor (CSF-1) to stimulate the formation of macrophage colonies in cultures of mouse bone marrow cells. In contrast, TPA did not synergize with other CSF subclasses that stimulate the formation of eosinophil, eosinophil-neutrophil, neutrophil, neutrophil-macrophage, and macrophage colonies, nor with either of the two factors required for megakaryocyte colony formation, megakaryocyte CSF, and megakaryocyte colony potentiator. In serum-free mouse bone marrow cell cultures TPA retained the ability to independently stimulate macrophage colony formation. However, TPA-stimulated colony formation was suboptimal and delayed in serum-free cultures that could support optimal colony formation in the presence of CSF-1. In addition, TPA did not directly compete with [125I]CSF-1 at 4 degrees C for its specific, high-affinity receptor on mouse peritoneal exudate macrophages. However, a 2-hour preincubation of the cells with TPA at 37 degrees caused almost complete loss of the receptor. Thus, TPA is able to mimic CSF-1 in its effects on CSF-1 responsive cells in some aspects (the spectrum of target cells, the morphology of resulting colonies, and the ability to down-regulate the CSF-1 receptor) but it is not able to mimic CSF-1 in other ways (TPA alone cannot stimulate the full CSF-1 response, TPA does not stimulate the most primitive CSF-1 responsive cells, and TPA does not bind to the CSF-1 receptor).     

Tumor-promoting phorbol esters inhibit the binding of colony-stimulating factor (CSF-1) to murine peritoneal exudate macrophages

L-cell colony-stimulating factor (CSF-1) is a sialoglycoprotein of molecular weight 70,000 daltons that specifically stimulates macrophage colony formation by single committed cells from normal mouse bone marrow and by various classes of more differentiated tissue-derived mononuclear phagocyte colony-forming cells (Stanley et al., 1978). CSF-1 interacts with target cells by direct and specific binding to membrane receptors (CSF-1 receptors) that are present only on cells of the mononuclear phagocyte series and their precursors. We studied the effect of tumor-promoting phorbol esters on the binding of 125I-labeled CSF-1 (125I-CSF-1) to murine peritoneal exudate macrophages (PEM). Biologically active TPA (12-O-tetradecanoyl phorbol-13-acetate) inhibits the binding of 125I-CSF-1 to its receptor on PEM. This inhibition exhibits temperature, time, and concentration dependence. At 37 degrees C, maximum inhibition occurred at about 10(-7) M; inhibition was 50% at 5 X 10(-9) M. At 0 degrees C, the inhibitory activity of TPA is diminished. The action of TPA on PEM is transient. Treated cells recover their 125I-CSF-1-binding activity whether TPA is later removed or not. The process of recovering CSF-1-binding activity is completely blocked by the addition of cycloheximide. When several phorbol derivatives were tested for their inhibitory activities, only biologically active phorbol esters were found to possess such activities. Furthermore, the inhibitory activities of various phorbol esters are proportional to their tumor-promoting activities. Inhibition appears to be due to a reduction in the total number of available CSF-1 receptors rather than a decrease in receptor affinity.     

Effect of lithium on granulopoiesis in culture.

Lithium carbonate therapy is associated with polymorphonuclear leukocytosis. In vitro studies have shown that lithium ions stimulate formation of granulocytic colonies. In a study undertaken to determine how lithium acts, colony-forming cells uncontaminated by monocytes (which elaborate colony-stimulating factor [CSF] in vitro) were obtained by means of a two-step cell separation procedure. The effects of lithium on colony formation were then studied in (a) cultures stimulated by humoral CSF, (b) cultures in which monocytes were relied upon to synthesize CSF de novo and (c) unstimulated cultures. Lithium enhanced the action of CSF but did not stimulate colony formation in the absence of CSF. In monocyte-stimulated cultures, colony formation increased with lithium concentrations up to 1 mmol/L but this increase paralleled that in CSF-stimulated cultures and therefore was not due to increased CSF production by monocytes. At higher concentrations of lithium, colony formation decreased in the monocyte-stimulated cultures but increased in the CSF-stimulated cultures. A lithium concentration of 4 mmol/L gave the greatest enhancing effect on colony formation in CSF-stimulated cultures and a concentration greater than 1 mmol/L inhibited de novo synthesis of CSF by monocytes.     

Characterization of the factor in L-cell conditioned medium capable of stimulating colony formation by mouse marrow cells in culture

Medium harvested from cultures of mouse L-cells contains “conditioning factor activity” (CFA) that may be detected by its ability to stimulate colony formation by mouse marrow cells in culture. The active material has been characterized and appears to be a glycoprotein with properties similar to those reported for the colony-stimulating activity in human urine. Characterization of trypsin-digested material indicated that only part of the molecule is essential for biological activity. The CFA has been partially purified from serum-free L-cell conditioned medium, and evidence has been obtained that radioactivity derived from labelled valine or glucosamine may be incorporated into the factor. L-cell conditioned medium appears to be a convenient source of partially-purified, highly active material, suitable for use in studies on its mechanism of action.     

Antibody-mediated suppression of bone marrow colony formation in vitro.

Antisera to mouse brain reacts with hematopoietic stem cells in the mouse bone marrow. We have examined the effect of anti-mouse brain serum (AMBS) on the development of in vitro colonies from mouse bone marrow cells. The addition of 5% AMBS to the cultures markedly decreased the numbers of colonies formed to an average of 10% of the number obtained with normal rabbit serum. AMBS suppressed formation induced by colony stimulating factors (CSF) derived from three different sources; serum from endotoxin treated mice, mouse L-cell conditioned media, and human peripheral mononuclear cell conditioned media. The suppressive activity was quantitatively recovered in the IgG fraction of AMBS. Divalent F(ab')2 fragments were as effective as the intact IgG in decreasing colony formation. Fab fragments were not suppressive. These results suggest that colony formation is induced via a dynamic interaction between CSF and the progenitor cell membrane, and that antibody directed at cell membrane antigen(s) interferes with the generation of the induction signal.     

Interleukin 2 inhibits in vitro granulocyte-macrophage colony formation

We have previously shown that murine bone marrow cells cultured with interleukin 2 (IL-2) produce interferon-alpha/beta (MuIFN-alpha/beta) and that IFN-alpha/beta can suppress in vitro granulocyte-macrophage colony-forming cell formation (GM-CFC). In this study, IL-2 was directly assessed for its ability to inhibit in vitro granulocyte and/or macrophage colony-forming cell formation (GM-CFC/M-CFC). C57BL/6 bone marrow cells were cultured with different colony-stimulating factors (CSF), i.e., partially purified macrophage-CSF (M-CSF) or recombinant granulocyte and macrophage CSF (GM-CSF) in the presence or absence of different IL-2 preparations. Partially purified mouse IL-2 or recombinant human or mouse IL-2 (rHuIL-2 and rMuIL-2) totally inhibit GM-CFC and M-CFC formation at 7 days of culture. The level of inhibition mediated by IL-2 was concentration-dependent, with as little as 1 U/ml giving total inhibition of colony formation. The ability of IL-2 to inhibit colony formation was completely abolished by treatment with antisera to IL-2. MuIFN-alpha/beta and MuIFN-gamma appeared to play no role in IL-2-induced myelo-suppression in that addition of antisera to these IFN failed to block IL-2-induced suppression. Myelo-suppression mediated by IL-2 was independent of the concentration of CSF used in the bone marrow cultures. Suppression was also not dependent upon the initial presence of T cells or natural killer (NK) cells. Bone marrow cells depleted of Thy-1+, Lyt-1+, Lyt-2+, NK-1.1+, Asialo GM1+, or Qa-5+ cells were as susceptible to IL-2 induced suppression as untreated or complement-treated bone marrow cells. These results suggest that IL-2 may play an important role in regulating different aspects of hematopoiesis.     

Effects of a murine mammary tumor on in vivo and in vitro hemopoiesis

The effects of an autologous transplanted mammary tumor (RIII-T3) on hemopoiesis in RIII mice are described. Tumor-bearing animals died 30 to 40 days after inoculation and displayed splenomegaly, extreme neutrophilia, and moderately increased monocyte levels in the spleen, peripheral blood, and bone marrow. The precursors of neutrophils and monocytes, granulocyte/macrophage colony-forming cells (GM-CFC) were elevated in the spleen, bone marrow, and peripheral blood. RIII-T3-conditioned medium stimulated bone marrow GM-CFC and caused the myelomonocytic cell line, WEHI-3B, to differentiate in vitro. The conditioned medium did not stimulate erythroid, megakaryocyte, or eosinophil colony formation. When conditioned medium was fractionated, two peaks of activity corresponding to GM-CSF and G-CSF were observed, suggesting that the extreme neutrophilia observed in tumor-bearing animals may result from chronic exposure of the hemopoietic system to these hemopoietic hormones.     

Inhibition of murine CFU-C by vindesine: restoration of colony growth by colony stimulating factor

Vindesine (VDS) is a new vinca-alkaloid related to vinblastine and vincristine that blocks production of the microtubules in the mitotic phase of the cell cycle. Studies were undertaken to investigate the inhibitory effect of VDS on normal murine bone marrow cell proliferation and the possible interactions between this compound and L-cell derived colony stimulating factor (CSF). One X 10(7) murine bone marrow cells were exposed to various concentrations of VDS, ranging from 0.1 to 1.5 micrograms/ml for 1 h at 37 degrees C. Following this period, the cells were plated in agar in the presence of 100 units of CSF. A dose-dependent inhibition of colony formation was noted with increasing doses of the drugs. To determine whether an increased dose of CSF could overcome the inhibitory effect of VDS, further studies compared colony growth in response to 100 and 200 units of CSF. Virtually no inhibition of colony growth was detected in VDS-treated cells exposed to this higher dose of CSF while a dose-dependent reduction in CFU-C was noted with 100 units of CSF. Preincubation of cells with VDS and CSF prevented the inhibition that occurred with VDS alone. The addition of anti-CSF serum during the preincubation phase abolished the protective effect of CSF. The studies show that short-term exposure of marrow cells to VDS causes a dose-dependent inhibition of in vitro colony formation; this inhibition is prevented by increasing doses of CSF in agar culture or by simultaneous preincubation with CSF. The CSF action appears specific as its protective effect is neutralized by antibody to CSF, suggesting a potential role for CSF in preventing the antimitotic activity of VDS.     

Effect of 12-O-tetradecanoylphorbol-13-acetate (TPA) on the growth inhibitory and increased phosphatidylinositol (PI) responses induced by epidermal growth factor (EGF) in A431 cells

Epidermal growth factor (EGF) inhibited the growth of A431 human epidermoid carcinoma cells. The tumor promoting, phorbol ester, 12-O-tetradecanoylphorbol-13-acetate (TPA) also retarded A431 cell growth. Addition of both TPA and EGF inhibited cell growth in an additive or synergistic manner depending upon the initial plating density of the cultures. EGF increased the production of diacylglycerol (60-70%) and stimulated the synthesis of phosphatidylinositol (PI) from 3H-inositol (three- to fourfold increase). Both of these responses were attenuated in the presence of TPA. TPA alone stimulated the production of diacylglycerol (DG) but had little effect on PI synthesis. The biological effect of TPA appeared to be mediated by the presence of a high-affinity receptor for phorbol esters on A431 cells. Moreover, the binding of 125I-EGF to A431 cells was unaffected by TPA, suggesting that the antagonistic effects of TPA were occurring distal to the EGF receptor. These findings also indicated that although TPA and EGF both inhibited A431 cell growth, this effect could be dissociated from changes in PI synthesis but may be dependent upon transient changes in DG production.     

The effect of DLE fractions on GM-progenitors of haematopoietic stem cells in vitro

Dialysable leucocyte extract (DLE) prepared from buffy coats of human blood, potentiates the effect of Colony-stimulating factor (CSF) on the growth of granulocyte-macrophage colony forming cell (GM-CFC) colonies in vitro. This relative increase of the number of colonies is apparent when diluted CSF (present in lung conditioning medium) as a control, and DLE, in a wide range of concentrations are added to the culture of mouse bone marrow cells. Fractionation of DLE on Amicon membranes revealed that the activity resides in molecules of 0-5kD. Molecules 5-10kD have no potentiating effect. DLE and its fractions (0-5kD, 0-1kD), except fractions 0-500 D and 5-10kD, when added undiluted i.e. at the initial concentration, exerted a suppressive effect: colonies are not formed despite the presence of CSF. In a pilot experiment, it was shown that DLE is able to stimulate colony-forming activity of earlier progenitors of erythroid cells (BFUe), under the influence of erythropoietin.     

Characterization of myogenesis from adult satellite cells cultured in vitro

We describe several characteristics of in vitro myogenesis from adult skeletal muscle satellite cells from the rat and several amphibian species. The timing of cell proliferation and fusion into myotubes was determined, and in urodeles, myogenesis from satellite cells was clearly demonstrated for the first time. Growth factors are known to stimulate satellite cell proliferation. Acidic FGF mRNA was present in rat satellite cells during proliferation but it was not detected in myotubes. Fibronectin was synthesized in satellite cells during proliferation and expelled into the extracellular medium when the myotubes differentiated. We suggest that fibronectin plays a part in the formation of myotubes, as this process was inhibited by anti-fibronectin IgG. Adult satellite cells might differ from fetal myoblasts since they were observed to exhibit the opposite response to a phorbol ester (TPA) to that of the myoblasts. We therefore examined the possibility that the different levels of protein kinase C activity and different phorbol ester binding characteristics in the two cell types account for these opposite responses. Our results suggest that the difference is not connected with the phorbol ester receptor but might be caused by events subsequent to protein kinase C activation. Localized extracellular proteolytic activity might have a role in cell mobilization and/or fusion when satellite cells are activated. We showed that the content of plasminogen activators, chiefly urokinase, was larger in tissues from slow twitch muscles which regenerate more rapidly than fast muscles. The urokinase level rose sharply in cultures when cells fused into myotubes, and was twice as high in slow muscle cells as in fast ones. We also found that, in vitro, slow muscle satellite cells displayed greater myogenicity, but that phorbol ester inhibited their mitosis and myogenicity. We conclude that satellite cells acquire characteristics which differentiate them from myoblasts and correspond to the fast and slow muscles from which they originate.     

Effects of human leukocyte conditioned medium on mouse haemopoietic progenitor cells.

Medium conditioned by human peripheral blood leukocytes (HLCM) was studied for its in vitro effects on haemopoietic progenitor cells (CFU-s and CFU-c) present in mouse bone marrow. HLCM has poor colony stimulating activity in semi-solid cultures of mouse bone marrow cells, but invariably increases the number of colonies obtained in the presence of plateau levels of semi-purified colony stimulating factor (CSF). In liquid cultures, HLCM appears to contain a potent initiator of DNA synthesis in CFU-s, an activity which coincides with an increased CFU-s maintenance and causes a three- to four-fold increase in CFU-c number. It is apparent from this study that HLCM, in addition to stimulating colony formation in cultures of human bone marrow cells, has a profound in vitro effect on primitive haemopoietic progenitor cells of the mouse, which cannot be attributed to CSF.     

Purification of a human urinary colony-stimulating factor

Colony-stimulating factor (CSF), a protein required for the in vitro formation of colonies composed of granulocytes and/or macrophages, was isolated from the urine of anemic patients by using a seven-step procedure. The purified, homogeneous CSF had a specific activity of 1.9 X 10(8) U/absorbance unit at 280 nm (AU). This represents an overall purification of 25,330-fold and a total recovery of 3.8%. Upon iodination of the protein, the radioactivity migrated on sodium dodecyl sulfate (SDS) gel electrophoresis as a single peak with an apparent molecular weight of 46,000; reduction with mercaptoethanol caused dissociation to a single component of molecular weight 23,000. Only the dimer is active in stimulating colony formation. Urinary CSF stimulates formation of colonies comprising only macrophages in the mouse bone marrow cell culture assay. A neutralizing antibody raised against mouse L-cell CSF did not neutralize the activity of the urinary CSF but did bind it. This may indicate that the relative positions of antibody binding sites and the active sites are different in these two glycoproteins.     

Effects of okadaic acid on mouse hemopoietic cells

Effects of okadaic acid, a potent non-12-O-tetradecanoyl-phorbol-13-acetate(TPA)-type tumor promoter, on mouse hemopoietic cells were investigated. Okadaic acid stimulated mouse bone marrow cells to form granulocyte-macrophage colony-forming unit (CFU-GM) colonies without added colony stimulating factors(CSFs). At the concentration of 1.82 x 10(-8) M, colony formation of 77 +/- 14 colonies/1 x 10(5) bone marrow cells was observed. Observations on the effects of other cells on the CSF induction suggested that okadaic acid primarily stimulated the functions of macrophages, and the CSF production from macrophages might be attributed to the CFU-GM colony formation. On the other hand, the erythroid colony-forming unit(CFU-E) colony formation stimulated by     

Inhibition of mouse bone marrow granulocyte-macrophage colony formation by factors derived from natural killer cells: an in vitro model for hybrid resistance

Investigations were performed to study whether soluble factors produced by NK-cells could mediate "hybrid resistance" in vitro. NK-cells enriched from spleens of B6D2F1 hybrid mice were incubated with parental B6 bone marrow, and the effect of the derived supernatants on the development of granulocyte-macrophage colony forming cells (GM-CFC) was assessed. Cell free supernatants obtained from low density cells (LDC) of B6D2F1 hybrids stimulated with bone marrow cells (BMC) from B6 mice inhibited GM-CFC formation. The inhibition was similar using B6, D2 or B6D2F1 bone marrow cells as the targets for GM-CFC growth. Our findings suggest that NK cells from F1 hybrid mice when stimulated with BMC from B6 mice release inhibitory factors, different from IFN-gamma and that this production may represent a mechanism of natural resistance to parental H-2b bone marrow grafts.     

Identification of a common signal associated with cellular proliferation stimulated by four haemopoietic growth factors in a highly enriched population of granulocyte/macrophage colony-forming cells.

We have prepared a population of bone marrow cells that is highly enriched in neutrophil/macrophage progenitor cells (GM-CFC). Four distinct haemopoietic growth factors can stimulate the formation of mature cells from this population, although the proportions of neutrophils and/or macrophages produced varied depending on the growth factor employed: interleukin 3 (IL-3) and granulocyte/macrophage colony-stimulating factor (GM-CSF) stimulated the formation of colonies containing both neutrophils and macrophages; macrophage colony-stimulating factor (M-CSF) produced predominantly macrophage colonies; and granulocyte colony-stimulating factor (G-CSF) promoted neutrophil colony formation. Combinations of these four growth factors did not lead to any additive or synergistic effect on the number of colonies produced in clonal soft agar assays, indicating the presence of a common set of cells responsive to all four haemopoietic growth factors. These enriched progenitor cells therefore represent an ideal population to study myeloid growth-factor-stimulated survival, proliferation and development. Using this population we have examined the molecular signalling mechanisms associated with progenitor cell proliferation. We have shown that modulation of cyclic AMP levels has no apparent role in GM-CFC proliferation, whereas phorbol esters and/or Ca2+ ionophore can stimulate DNA synthesis, indicating a possible role for protein kinase C activation and increased cytosolic Ca2+ levels in the proliferation of these cells. The lack of ability of all four myeloid growth factors to mobilize intracellular Ca2+ infers that these effects are not achieved via inositol lipid hydrolysis.(ABSTRACT TRUNCATED AT 250 WORDS)     

Protein kinase C-mediated negative-feedback inhibition of unstimulated and bombesin-stimulated polyphosphoinositide hydrolysis in Swiss-mouse 3T3 cells.

Bombesin-related peptides stimulate a rapid increase in polyphosphoinositide hydrolysis in Swiss-mouse 3T3 cells. These peptides generate an increase in the efflux of 45Ca2+ from pre-labelled cells, a response consistent with an inositol trisphosphate-mediated mobilization of intracellular Ca2+. The bombesin-stimulated release of cellular 45Ca2+ is inhibited by tumour-promoting phorbol esters (e.g. 12-O-tetradecanoylphorbol 13-acetate, TPA). Although there are several possible sites of action at which this effect might occur, our results indicate that TPA induces an uncoupling of bombesin-stimulated hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2) without decreasing cellular binding of bombesin. In cultured cells, protein kinase C can be down-modulated by a prolonged incubation of the cells with phorbol esters. Such pretreatment greatly decreased the inhibitory effect of TPA on bombesin-stimulated PIP2 hydrolysis, suggesting that this action of the phorbol ester is mediated via protein kinase C. Since diacylglycerol is an endogenous activator of protein kinase C and a direct product of PIP2 hydrolysis, these results suggest that protein kinase C inhibition of polyphosphoinositide hydrolysis may function as a negative-feedback pathway. Cells in which protein kinase C has been down-modulated show elevated basal and bombesin-stimulated production of inositol phosphates, providing evidence that such a feedback loop limits polyphosphoinositide turnover in both unstimulated and mitogen-stimulated cells.     

Action of granulopoiesis-stimulating cytokines rhG-CSF, rhGM-CSF, and rmGM-CSF on murine haematopoietic progenitor cells for granulocytes and macrophages (GM-CFC)

The aim of this study was to provide new data to the knowledge of mechanisms by which recombinant human granulocyte colony-stimulating factor (rhG-CSF), recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) and recombinant murine granulocyte-macrophage colony-stimulating factor (rmGM-CSF) enhance the numbers of colonies growing from hematopoietic progenitor cells for granulocytes and macrophages (GM-CFC) in the murine bone marrow. The in vitro technique for cultivating GM-CFC from normal bone marrow cells was used. For evaluation of stimulatory actions of the drugs studied, the factors themselves or sera of mice given these factors were added to the cultures. The factors or the sera were present in the cultures either as the only potentially stimulatory agents or acted jointly with a suboptimum concentration of recombinant murine interleukin-3 (rmIL-3). It was found that both rhG-CSF and rmGM-CSF stimulate the proliferation of GM-CFC by a combination of direct mechanisms (direct actions on the target cells) and indirect effects (effects mediated through the induction of other cytokines and/or growth factors in the murine organism). The rhGM-CSF exhibited somewhat weaker in vitro effects in comparison with the other two factors and only indirect effects were noted. Additional in vivo experiments documented that, in spite of differences in mechanisms of action of the individual drugs studied on murine bone marrow cells in vitro, equal in vivo doses of the factors induce quantitatively similar effects on the production of GM-CFC in vivo.     

The role of colony-stimulating factor in granulopoiesis

The proliferation and maturation of granulocytic-monocytic stem cells appears to be controlled by a series of closely related glycoproteins termed “colony-stimulating factors” (CSFs). Recently, we devised a 6-step scheme for the purification of murine fibroblast (L-cell)-derived CSF. Ten liter pools of conditioned media were concentrated by ultrafiltration, precipitated by ethanol, and separated on DEAE cellulose, Con-A Sepharose, and Sephadex G 150. The CSF was separated from trace contaminants, including endotoxin, by density gradient centrifugation. The purified material was radioiodinated and used to define the serum half-life and in vivo distribution. Following IV injection there was a biphasic serum clearance with a t½ of 24–40 min and 2–2½ hours in the first and second phases. Approximately 25% of the tracer was excreted in the urine at 6 h; however, urinary radioactivity was due to low molecular weight peptides. Simultaneous studies by radioimmunoassay showed a similar rapid serum clearance of unlabeled CSF but virtually no urinary CSF activity. Thus, assays for urinary CSF may not provide useful measures of in vivo CSF activity. Further in vitro studies have defined the interaction of CSF with responsive cells in the marrow. Varying doses of CSF were incubated with 10⁷ marrow cells for intervals of 24–48 h. The major increment in cell-associated radioactivity occurred between 6 and 16 h. The reaction was saturable with 1–2 ng/ml CSF. Binding was prevented by cold CSF, but not by other proteins. Irradiation yielded only a minimal reduction in CSF binding. The interaction of CSF with marrow cells appeared to require new protein synthesis, as binding was completely inhibited by cycloheximide and puromycin. Irradiated mice injected with antibodies to CSF showed an inhibition of granulopoiesis by marrow cells in peritoneal diffusion chambers; however, granulopoiesis in the intact bone marrow was unaffected. Granulpoiesis in long-term marrow cultures was also unaffected by anti-CSF. These different responses may be due to accelerated clearance of injected CSF in nonirradiated mice or to extensive stromal interactions that modulate and perhaps control granulocytic differentiation in the intact bone marrow microenvironment.