Identification and characterization of a low-affinity granulocyte- macrophage colony-stimulating factor receptor on primary and cultured human melanoma cells

Hematopoietic growth factor receptors are present on cells of normal nonhematopoietic tissues such as endothelium and placenta. We previously demonstrated functional human granulocyte-macrophage colony- stimulating factor (GM-CSF) receptors on small cell carcinoma of the lung cell lines, and others have reported that certain solid tumor cell lines respond to GM-CSF in clonogenic assays. In the current study, we examine human melanoma cell lines and fresh specimens of melanoma to determine whether they have functional GM-CSF receptors. Scatchard analyses of 125I-GM-CSF equilibrium binding to melanoma cell lines showed a mean of 542 +/- 67 sites per cell with a kd of 0.72 +/- 0.14 nmol/L. Cross-linking studies in the melanoma cell line, M14, showed a major GM-CSF receptor species of 84,000 daltons. Under the conditions tested, the M14 cells did not have a proliferative response to GM-CSF in vitro, nor was any induction of primary response genes detected by Northern analysis in response to GM-CSF. Studies to determine internal translocation of the receptor-ligand complex indicated less than 10% of the 125I-GM-CSF internalized was specifically bound to receptors. Primary melanoma cells from five surgical specimens had GM-CSF receptors; Scatchard analysis was performed on one sample, showing 555 sites/cell with a kd of 0.23 nmol/L. These results indicate that human tumor cells may express a low-affinity GM-CSF receptor protein that localizes to the cell surface and binds ligand, but lacks functional components or accessory factors needed to transduce a signal.     

In vitro characterization of the human recombinant soluble granulocyte- macrophage colony-stimulating factor receptor

We have cloned, expressed, and partially purified a naturally occurring, truncated, soluble form of the human granulocyte-macrophage colony-stimulating factor (GM-CSF) receptor alpha subunit to investigate its biochemical and biologic properties. The soluble receptor species lacks the transmembrane and cytoplasmic domains that are presumably removed from the intact receptor cDNA by a mechanism of alternative splicing. The resulting soluble 55- to 60-kD glycosylated receptor species binds GM-CSF with a dissociation constant (kd) of 3.8 nmol/L. The soluble GM-CSF receptor successfully competes for GM-CSF binding not only with the transmembrane-anchored GM-CSF receptor alpha subunit but also with the native oligomeric high-affinity receptor complex. In addition, in human bone marrow colony-forming assays, the soluble GM-CSF receptor species can antagonize the activity of GM-CSF. Our data suggest that the soluble GM-CSF receptor may be capable of acting in vivo as a modulator of the biologic activity of GM-CSF.     

Biologic properties in vitro of a recombinant human granulocyte- macrophage colony-stimulating factor

Recombinant human granulocyte-macrophage colony-stimulating factor (rH GM-CSF) was purified to homogeneity from medium conditioned by COS cells transfected with a cloned human GM-CSF cDNA and shown to be an effective proliferative stimulus in human marrow cultures for GM and eosinophil colony formation. The specific activity of purified rH GM- CSF in human marrow cultures was calculated to be at least 4 X 10(7) U/mg protein. Clone transfer experiments showed that this proliferation was due to direct stimulation of responding clonogenic cells. Acting alone, rH GM-CSF did not stimulate erythroid colony formation, but in combination with erythropoietin, increased erythroid and multipotential colony formation in cultures of peripheral blood cells. rH GM-CSF had no proliferative effects on adult or fetal murine hematopoietic cells, did not induce differentiation in murine myelomonocytic WEHI-3B cells, and was unable to stimulate the survival or proliferation of murine hematopoietic cell lines dependent on murine multi-CSF (IL 3). rH GM- CSF stimulated antibody-dependent cytolysis of tumor cells by both mature human neutrophils and eosinophils and increased eosinophil autofluorescence and phagocytosis by neutrophils. From a comparison of these effects with those of semipurified preparations of human CSF alpha and -beta, it was concluded that rH GM-CSF exhibited all the biologic activities previously noted for CSF alpha.     

Identification of a growth factor for primary murine stroma as macrophage colony-stimulating factor

Knowledge of the stromal microenvironment is crucial for understanding the hematopoietic system. We took advantage of an assay that permits analysis of primary stroma-initiating cells (SICs) on the clonal level, and further characterized SICs and the factors that regulate SICs. Stroma formation in this assay is dependent on a high-molecular-weight factor secreted by the stromal cell line AC3.U. Here we show that this factor is identical to macrophage colony-stimulating factor (M-CSF), and that purified M-CSF is sufficient for induction of stroma formation. M-CSF, isolated from the line AC3.U, as well as from L929 cells and COS cells transfected with an expression vector encoding M- CSF, migrated in two peaks as 160- and 650-kD species after gel filtration. These molecular-weight species encompassed all stroma- inducing activity, and both stimulated macrophage colony formation. Affinity chromatography and blocking studies with antibodies specific for M-CSF and c-fms confirmed M-CSF as the sole factor in the supernatant of the stromal cell line AC3.U that promotes stroma formation. Culture of marrow, for as little as 1 week, depleted M-CSF- dependent SIC while increasing the incidence of replatable, factor- independent SIC. This suggests that culture changes the properties of SICs, perhaps by inducing differentiation into mature stromal cells. Thus, our results show a novel function of M-CSF as an important modulator of stroma formation.     

Purified murine granulocyte/macrophage progenitor cells express a high-affinity receptor for recombinant murine granulocyte/macrophage colony-stimulating factor

Purified recombinant murine granulocyte/macrophage colony-stimulating factor (GM-CSF) was labeled with 125I and used to examine the GM-CSF receptor on unfractionated normal murine bone marrow cells, casein-induced peritoneal exudate cells, and highly purified murine granulocyte/macrophage progenitor cells (CFU-GM). CFU-GM were isolated from cyclophosphamide-treated mice by Ficoll-Hypaque density centrifugation followed by counterflow centrifugal elutriation. The resulting population had a cloning efficiency of 62-99% in cultures containing conditioned medium from pokeweed mitogen-stimulated spleen cells and 55-86% in the presence of a plateau concentration of purified recombinant murine GM-CSF. Equilibrium binding studies with 125I-labeled GM-CSF showed that normal bone marrow cells, casein-induced peritoneal exudate cells, and purified CFU-GM had a single class of high-affinity receptor with an approximate Ka of 10(8)-10(9) M-1. CFU-GM expressed an average of 3783 +/- 4 receptors per cell; normal bone marrow cells, 1518 +/- 242 receptors per cell; and peritoneal exudate cells, 2025 +/- 216 receptors per cell. Affinity crosslinking studies demonstrated that 125I-labeled GM-CSF bound specifically to two species of Mr 180,000 and 70,000 on CFU-GM, normal bone marrow cells, and peritoneal exudate cells. The Mr 70,000 species is thought to be a proteolytic fragment of the intact Mr 180,000 receptor. The present studies indicate that the GM-CSF receptor expressed on CFU-GM and mature myeloid cells are structurally similar. In addition, the number of GM-CSF receptors on CFU-GM is twice the average number of receptors on casein-induced mature myeloid cells, suggesting that receptor number may decrease as CFU-GM mature.     

Identification through chemical cross-linking of distinct granulocyte- macrophage colony-stimulating factor and interleukin-3 receptors on myeloid leukemic cells, KG-1

Granulocyte/macrophage colony-stimulating factor (GM-CSF) and interleukin-3 (IL-3) each bind specifically to a small number of high- affinity receptors present on the surface of the cells of the acute myelogenous leukemia line, KG-1. Through chemical cross-linking of IL-3 and GM-CSF to KG-1 cells, we identified distinct binding proteins for each of these cytokines with approximate molecular masses of 69 and 93 Kd, respectively. Although these two binding proteins are distinct, GM- CSF and IL-3 compete with each other for binding to KG-1 cells. Other cell lines, which express receptors for either factor but not for both do not display this cross-competition for binding with IL-3 and GM-CSF. These findings imply that distinct IL-3 and GM-CSF binding proteins are expressed on the cell surface and that an association exists between these proteins on KG-1 cells.     

Differentiation-linked changes in tyrosine phosphorylation, functional activity, and gene expression downstream from the granulocyte- macrophage colony-stimulating factor receptor

The HL-60 model of myeloid maturation was used to test whether changes in signaling from the granulocyte-macrophage colony-stimulating factor (GM-CSF) receptor accompany maturation-related changes in cellular responses to GM-CSF. Receptor expression, tyrosine phosphorylation, functional activity, and c-fos gene expression were measured. Functional GM-CSF receptors were present throughout differentiation as both uninduced and dimethyl sulfoxide (DMSO)-induced HL-60 cells responded to GM-CSF, albeit in different ways. Uninduced promyelocytes proliferated in response to GM-CSF, whereas DMSO-induced cells lost the capacity to proliferate but did respond with increased expression of beta 2-integrins, enhanced respiratory burst activity, and metabolism of arachidonic acid. GM-CSF-stimulated upregulation of c-fos mRNA expression was not detected in immature cells but developed after 2 to 4 days with DMSO in line with a marked increase in responsiveness to stimulation with phorbol ester, showing that increased expression of c- fos is predominantly a feature of mature phagocytes. GM-CSF stimulated the tyrosine phosphorylation of a broadly similar range of proteins in both uninduced and DMSO-treated HL-60 cells, but protein bands were more heavily phosphorylated in DMSO-induced cells. Phosphorylation was rapid in onset and very transient in immature cells. Phosphorylation of several proteins, in particular a 130-kD band, was more sustained in DMSO-induced cells. These differences in signaling were not because of numerical differences in receptors, because reduction of GM-CSF concentration to trigger equivalent numbers of high-affinity receptors delayed the onset of phosphorylation in DMSO-induced cells. We conclude that there are maturation-related changes in signaling downstream from the GM-CSF receptor.     

A phase I/II study of sequential interleukin-3 and granulocyte- macrophage colony-stimulating factor in myelodysplastic syndromes

In this phase I/II study, 9 patients with myelodysplastic syndromes (MDS) were treated with interleukin-3 (IL-3) followed by granulocyte- macrophage colony-stimulating factor (GM-CSF). Each treatment cycle was 28 days long and administered as follows: 1 microgram/kg/d IL-3 on days 1 through 7 and 3 micrograms/kg/d GM-CSF for days 8 through 21, followed by a 7-day rest period. IL-3 dose escalations were planned, but the dose of GM-CSF was fixed. Three patients had refractory anemia, 4 had refractory anemia with ringed sideroblasts, and 2 had refractory anemia with excess blasts. Six patients were dependent on red blood cell transfusions, 1 on platelet transfusions, and 2 on both. The absolute neutrophil count improved in 7 (77%) patients and the platelet count improved in 3 (33%) patients during therapy. Hemoglobin levels were unchanged. A clinically relevant response was seen in only 1 patient with thrombocytopenia, and he received five cycles of therapy. The neutrophil count decreased in 2 patients and the platelet count decreased in 4 patients during treatment. The toxicity of the treatment was significant. In the first cohort of 3 patients, 1 patient developed supraventricular tachycardia and congestive heart failure. In the second group, 1 patient developed progressive granulocytopenia and died of gram-negative septicemia. Because of the disparate toxicity, 3 more patients were treated at the same dose level. One of these experienced a high fever and bone pain requiring hospitalization. Because of these adverse effects, the IL-3 dose was not escalated and all patients received 1 microgram/kg/d for 7 days. We believe that sequential therapy with IL-3 and GM-CSF at these dose levels causes unacceptable toxicity in patients with MDS. The major toxic events occurred during weeks 4 and 5 after starting treatment and may have been primarily caused by GM-CSF therapy. Although neutrophil counts improve in most patients, the effect on red blood cells and platelets is minimal. At present, this form of therapy remains problematic and appears to have a limited potential in the management of MDS.     

Cloning of the low-affinity murine granulocyte-macrophage colony-stimulating factor receptor and reconstitution of a high-affinity receptor complex.

A cDNA clone (clone 71) that encodes a low-affinity receptor for murine granulocyte-macrophage colony-stimulating factor (GM-CSF) has been isolated by direct expression. This molecule is the homologue of the human GM-CSF receptor alpha subunit, although homology between these molecules is surprisingly low (less than 35% amino acid identity). The cDNA encodes a polypeptide of 387 amino acids, which contains the conserved features of the hematopoietin receptor superfamily. When expressed in COS-7 cells, this clone encodes a protein that binds radiolabeled murine GM-CSF with low affinity. Coexpression of clone 71 with a cDNA corresponding to a low-affinity interleukin 3 (IL-3) receptor (AIC2A) did not alter the affinity of binding of either GM-CSF or IL-3. However, coexpression of clone 71 with the IL-3 receptor-related cDNA AIC2B generated high-affinity binding sites for murine GM-CSF but not murine IL-3. These studies show that clone 71 and AIC2B are capable of forming an alpha beta complex capable of binding murine GM-CSF with high affinity, while AIC2A appears not to be a component of the murine GM-CSF receptor.     

Identification of a viability domain in the granulocyte/macrophage colony-stimulating factor receptor beta-chain involving tyrosine-750.

The granulocyte/macrophage colony-stimulating factor (GM-CSF) receptor (GMR) is a heterodimeric receptor expressed by myeloid lineage cells. In this study we have investigated domains of the GMR beta-chain (GMR beta) involved in maintaining cellular viability. Using a series of nested GMR beta deletion mutants, we demonstrate that there are at least two domains of GMR beta that contribute to viability signals. Deletion of amino acid residues 626-763 causes a viability defect that can be rescued with fetal calf serum (FCS). Deletion of residues 518-626, in contrast, causes a further decrement in viability that can be only partially compensated by the addition of FCS. GMR beta truncated proximal to amino acid 517 will not support long-term growth under any conditions. Site-directed mutagenesis of tyrosine-750 (Y750), which is contained within the distal viability domain, to phenylalanine eliminates all demonstrable tyrosine phosphorylation of GMR beta. Cell lines transfected with mutant GMR beta (Y750-->F) have a viability disadvantage when compared to cell lines containing wild-type GMR that is partially rescued by the addition of FCS. We studied signal transduction in mutant cell lines in an effort to identify pathways that might participate in the viability signal. Although tyrosine phosphorylation of JAK2, SHPTP2, and Vav is intact in Y750-->F mutant cell lines, Shc tyrosine phosphorylation is reduced. This suggests a potential role for Y750 and potentially Shc in a GM-CSF-induced signaling pathway that helps maintain cellular viability.     

Neutrophil migration is defective during recombinant human granulocyte- macrophage colony-stimulating factor infusion after autologous bone marrow transplantation in humans

We have previously reported that continuous intravenous (IV) administration of recombinant granulocyte-macrophage colony-stimulating factor (rHuGM-CSF) to humans following high-dose alkylating agent chemotherapy and autologous bone marrow support (ABMS) results in myeloid bone marrow maturation, accelerated granulocyte recovery, and reduced treatment-related toxicity. However, we found that leukocyte counts declined rapidly after discontinuation of rHuGM-CSF therapy, which suggests possible growth factor effects on leukocyte margination and migration. For these reasons we studied granulocyte margination by using 111In-labeled autologous granulocytes and found similar granulocyte margination before (21.5% +/- 13.4%) and during continuous IV rHuGM-CSF infusion (23.3% +/- 9.6%). Phagocytosis of Cryptococcus neoformans and granulocyte hydrogen peroxide production was similar before and during rHuGM-CSF infusion and similar to patients treated with the same high-dose chemotherapy and ABMS but not receiving growth factor. However, migration of granulocytes to a sterile inflammatory site was markedly reduced during continuous rHuGM-CSF infusion (1.2 +/- 0.9 WBCs/cm2, 24 hr) as compared with baseline (39.6 +/- 17.7 WBCs/cm2/24 hr; P less than .0008). These findings may be of relevance when extravascular granulocytes are required for host defense.     

Granulocyte/macrophage colony-stimulating factor stimulates monocyte and tissue macrophage proliferation and enhances their responsiveness to macrophage colony-stimulating factor

Granulocyte-macrophage colony-stimulating factor (GM-CSF) is a specific humoral growth factor that stimulates both neutrophilic granulocyte and macrophage production by bone marrow hematopoietic progenitor cells. GM- CSF also stimulates the proliferation and clonal growth of both tissue macrophages and blood monocytes. Although at low concentrations GM-CSF was unable to support the long-term growth of tissue macrophages, it greatly enhanced their responsiveness to macrophage CSF (M-CSF, or CSF- 1). This effect was also observed by treating macrophages with GM-CSF for a short time. GM-CSF did not compete with M-CSF for binding to M- CSF receptors nor was it inactivated by treatment with anti-M-CSF antiserum. Treatment of tissue macrophages with GM-CSF led to a rapid but transient downregulation of M-CSF receptors; prolonged incubation at 37 degrees C, however, resulted in a restoration and upregulation of M-CSF receptors. Identical effects were observed with both native or recombinant GM-CSF. This study suggests that GM-CSF regulates tissue macrophage production by two modes of action: (a) direct stimulation of macrophage proliferation, and (b) enhancement of their responsiveness to M-CSF.     

Differential activation of leukotriene biosynthesis by granulocyte- macrophage colony-stimulating factor and interleukin-5 in an eosinophilic substrain of HL-60 cells

Cytokines can stimulate eosinophils to produce cysteinyl leukotrienes (LTs) in the lung that provoke tissue destruction associated with asthma. Priming of an eosinophilic substrain of HL-60 cells (HL-60#7) with recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) before ionophore challenge was found to produce an apparent 45% increase in total LT production in a dose-dependent manner (ED50 = 150 pmol/L) that could be accounted for by a decrease in the time required for maximal formation of LTs. GM-CSF had no effect on the kinetic parameters of LTC4 synthase and therefore probably acts upstream of this catalytic event. Incubation with interleukin-5 (IL-5), however, had no effect on LT biosynthesis. This differential priming ability was not a consequence of different receptor populations or differences in the affinity or stability of the ligand-receptor complexes of GM-CSF and IL-5. GM-CSF and IL-5 each displayed similar populations of high-affinity binding sites and neither GM-CSF nor IL-5 were able to cross-compete for the other's receptor binding sites. Analysis of phosphotyrosine patterns suggest that IL-5 is incapable of transducing a signal in eosinophilic HL-60#7 cells even though IL-5 and GM-CSF receptors mediate signal transduction via a common beta-chain component that is also necessary for high-affinity binding. Overall, this unique system may permit the dissection of distinct events responsible for specific intracellular signals transduced separately by GM-CSF or IL-5.     

Identification and molecular cloning of a soluble human granulocyte-macrophage colony-stimulating factor receptor.

Granulocyte-macrophage colony-stimulating factor (GM-CSF) plays an important role in hematopoiesis and host defense via interaction with specific cell-surface receptors in target tissues. We identified a truncated, soluble form of the low-affinity GM-CSF receptor (GMR) in chorio-carcinoma cells. Low-affinity GMR cDNAs encoding both the membrane-bound and soluble receptors were obtained by PCR using primers corresponding to the published sequence. Clones encoding the soluble receptor were identical in sequence to the membrane-bound form but contained a 97-nucleotide internal deletion. The amino acid sequence of this deleted cDNA predicts a protein that lacks the 84 C-terminal amino acids of the membrane-bound receptor, including the transmembrane and cytoplasmic domains, and contains 16 different amino acids at its C terminus. Expression of the soluble GMR cDNA in murine psi-AM cells as well as GM-CSF-dependent myeloid 32Dc13 cells produced a secreted protein that retained its capacity to bind GM-CSF in solution. RNase protection analysis indicates that this variant cDNA is derived from a naturally occurring mRNA. Soluble receptors have been identified for several other hematopoietin receptors and may be a general feature of this class. The striking similarity between the soluble form of the GMR and other hematopoietin receptors suggests that soluble binding proteins may play an important role in regulating the broad spectrum of biological responses mediated by these cytokines.     

Kinetic resolution of two mechanisms for high-affinity granulocyte-macrophage colony-stimulating factor binding to its receptor

Granulocyte-macrophage colony-stimulating factor (GM-CSF) is an important hematopoietic cytokine that exerts its effects by interaction with the GM-CSF receptor (GMR) on the surface of responsive cells. The GM-CSF receptor consists of two subunits: GMRalpha, which binds GM-CSF with low affinity, and GMRbeta, which lacks intrinsic ligand-binding capability but complexes with GMRalpha to form a high-affinity receptor (GMRalpha/beta). We conducted dynamic kinetic analyses of GM-CSF receptors to define the role of GMRbeta in the interaction of ligand and receptor. Our data show that GMRalpha/beta exhibits a higher k(on) than GMRalpha, indicating that GMRbeta facilitates ligand acquisition to the binding pocket. Heterogeneity with regard to GM-CSF dissociation from GMRalpha/beta points to the presence of loose and tight ligand-receptor complexes in high-affinity binding. Although the loose complex has a k(off) similar to GMRalpha, the lower k(off) indicates that GMRbeta inhibits GM-CSF release from the tight receptor complex. The two rates of ligand dissociation may provide for discrete mechanisms of interaction between GM-CSF and its high-affinity receptor. These results show that the beta subunit functions to stabilize ligand binding as well as to facilitate ligand acquisition.     

Cooperative effects of interleukin-3 (IL-3), IL-5, and granulocyte- macrophage colony-stimulating factor: a new myeloid cell line inducible to eosinophils

The cytokines interleukin-3 (IL-3); IL-5, and granulocyte-macrophage colony-stimulating factor (GM-CSF) are known to contribute to the proliferation and differentiation of eosinophil progenitors. Recently, it was determined that the cellular receptors for these three cytokines share a common beta-chain while having unique alpha-chains. Thus, there is considerable interest in how these cytokines and their receptors interact in promoting production of eosinophils. We have established a cell line (AML14) from a patient with acute myelogenous leukemia that will consistently exhibit eosinophilic differentiation in suspension in response to IL-3, IL-5, and GM-CSF. Proliferation with only modest differentiative effects was observed in response to a single cytokine. Combinations of two cytokines gave variable results, with GM-CSF + IL-3 and IL-3 + IL-5 causing more proliferation than a single cytokine but little more differentiation. The combination of GM-CSF + IL-5 caused marked enhancement of eosinophilic differentiation with only modest augmentation of proliferation. The combination of all three cytokines was most effective in stimulating both proliferation and eosinophilic differentiation (up to 70% of cells) of AML14 cells. Specific binding of GM-CSF and IL-5 to AML14 cells can be conveniently studied by flow cytometric methods, and cross-competition of these two cytokines for their respective receptors was demonstrated. IL-3 was shown to partially compete for IL-5 binding on AML14 cells. Although specific IL- 3 binding could not be demonstrated by flow cytometry, mRNA for the alpha-chains of the IL-3, IL-5, and GM-CSF receptors and the beta-chain common to all three receptors was detected in AML14 cells. The AML14 cell line may be a useful model for the study of cooperative interactions of IL-3, IL-5, GM-CSF, and their respective receptors in the promotion of eosinophil progenitor growth and differentiation.