Alterations in genes that function in normal growth and development have been linked to malignant cell transformation. The mononuclear phagocyte colony-stimulating factor (CSF-1 or M-CSF) is a polypeptide growth factor synthesized by mesenchymal cells, which stimulates the survival, proliferation, and differentiation of haematopoietic cells of the monocyte-macrophage series. Multiple forms of soluble CSF-1 are produced by proteolytic cleavage of membrane-bound precursors, some of which are stably expressed at the cell surface. The c-fms proto-oncogene encodes the CSF-1 receptor, which is composed of an extracellular ligand-binding domain linked by a single membrane-spanning segment to a cytoplasmic tyrosine-specific protein kinase domain. Whereas the tyrosine kinase activity of the normal receptor is stimulated by CSF-1, mutations in the c-fms gene can constitutively activate the kinase to provide growth-stimulatory signals in the absence of the ligand. Oncogenic activation of the c-fms gene product appears to involve removal of a negative regulatory tyrosine residue near the carboxyl terminus of the receptor and one or more additional mutations that may simulate a conformational change induced by CSF-1 binding. Expression of the human c-fms gene in mouse NIH-3T3 cells confers a CSF-1 stimulated growth phenotype, indicating that receptor transduction is sufficient for fibroblasts to respond to a haematopoietic growth factor. In contrast, the v-fms oncogene induces factor-independent growth and tumorigenicity in factor-dependent myeloid cell lines, and contributes to the development of proliferative disorders of multiple haematopoietic lineages when introduced into murine bone marrow progenitors. Aberrant expression of an endogenous c-fms gene secondary to pro viral insertion and transcriptional activation has also been implicated in virus-induced myeloblastic leukaemia in mice. The c-fms and CSF-1 genes have been mapped on the long arm of human chromosome 5, a region that frequently undergoes interstitial deletions in certain haematopoietic disorders including acute myelogenous leukaemia. The study of CSF-1 and its receptor should provide information concerning the role of tyrosine kinases in regulating the normal growth and differentiation of haematopoietic cells and in contributing to their malignant transformation.

The macrophage colony-stimulating factor (CSF-1 or M-CSF) is a lineage-specific haematopoietic growth factor required for the proliferation, differentiation, and viability of mononuclear phagocytes (Stanley et al. 1983). CSF-1 also stimulates a variety of specialized macrophage functions including tumoricidal activity (Ralph & Nakoinz, 1987) and the production of other cytokines (Metcalf & Nicola, 1985; Warren & Ralph, 1986). The growth factor is a glycosylated polypeptide homodimer (Stanley & Heard, 1977) whose pleiotropic actions on mononuclear phagocytes are mediated by binding to a single class of high-affinity receptors at the cell surface (Guilbert & Stanley, 1980, 1986). Expression of the receptor for CSF-1 is an early marker of commitment to the monocyte-macrophage series. Low numbers of CSF-1 receptors are present on immature bone marrow progenitors, and a 10-fold increase in receptor number takes place as the cells differentiate (Bartelmez et al. 1985). Circulating monocytes and tissue macrophages maintain high numbers of cell surface CSF-1 receptors (ca. 50000 per cell).

The receptor for CSF-1 is encoded by the c-fms proto-oncogene (Sherr et al. 1985) and is one of a family of oncogene products and growth factor receptors, which exhibit a tyrosine-specific protein kinase activity (Hunter & Cooper, 1985). The enzymatic activity appears to be required for the generation of mitogenic signals by this class of proteins. In this paper, we will briefly review the evidence that the CSF-1 receptor is the c-fms gene product and focus on recent investigations of the interaction between this receptor and its ligand in normal and malignant cells.

In recent years, experimental evidence has been obtained establishing a relationship between malignant transformation and alterations of genes that are involved in normal growth and differentiation. These observations have generally followed from the study of retroviral oncogenes (v-onc genes) that were generated by recombination between replicating viruses and proto-oncogene (c-onc) sequences present in normal cellular DNA. The v-onc genes are responsible for the ability of these viruses to cause tumours in animals and to transform cells in culture. Our interest in the CSF-1 receptor and its ligand evolved from the study of v-fms, a viral oncogene transduced from the feline c-fms proto-oncogene.

The Susan McDonough strain of feline sarcoma virus (SM-FeSV) was isolated from a spontaneously occurring multicentric fibrosarcoma of a domestic cat (McDonoughet al. 1971). SM-FeSV transforms established fibroblast cell lines from several species in culture and induces fibrosarcomas when re-inoculated into kittens (Sarma et al. 1972). Acquired cellular sequences are localized as a contiguous segment in the middle of the SM-FeSV genome, which was designated c-fms (Donner et al. 1982). Nucleotide sequence analysis of the v-fms gene predicted that its product was a transmembrane protein and revealed that the carboxyl-terminal domain exhibited amino acid sequence similarity with the family of tyrosine-specific protein kinases (Hampe et al. 1984).

The product of the v-fms oncogene is an integral transmembrane glycoprotein with an associated tyrosine kinase activity. The glycoprotein contains asparagine (N)linked oligosaccharides, which are processed during transport to the plasma membrane (Anderson et al. 1982, 1984). A single membrane-spanning segment of 26 hydrophobic amino acids near the middle of the protein divides it into two domains (Hampe et al. 1984). The v-fms gene product is oriented in the plasma membrane with its glycosylated amino-terminal domain exposed at the cell surface and its carboxyl-terminal tyrosine kinase domain in the cytoplasm (Rettenmier et al. 1985a). Cell surface expression is required to elicit the transformed phenotype (Roussel et al. 1984; Nichols et al. 1985; Hadwiger et al. 1986). The v-fms-coded glycoproteins on the plasma membrane become associated with clathrin-coated pits and are internalized in endocytic vesicles (Manger et al. 1984). These findings provided a basis for the hypothesis that the viral oncogene was derived from a cellular gene encoding a growth factor receptor. Indeed, precedents for a link between oncogenes and growth factors, or their receptors, were first established by the findings that the v-sis oncogene encodes a chain of the platelet-derived growth factor (PDGF) (Doolittle et al. 1983; Waterfield et al. 1983) whereas the v-erbB oncogene product represents a truncated version of the epidermal growth factor (EGF) receptor (Downward et al. 1984).

Although transcripts of the c-fms gene were first detected in mouse and human placental tissue (Müller et al. 1983a,b) and in human choriocarcinoma (Müller et al. l983a) and murine myeloid leukaemia (Gonda & Metcalf, 1984) cell lines, we were initially unable to precipitate a presumptive c-fms protein from these sources using the then available antibodies to the c-fms gene product. Since the c-fms gene ofSM-FeSV was derived from a feline cellular gene, the viral oncogene product was more likely to exhibit antigenic cross-reactivity with the c-fms product of the cat. Normal adult cat tissues were therefore surveyed for the presence of c-fms transcripts, and the highest levels of the 4·0-kb c-fms mRNA were present in the spleen.

When the v-fms gene products are precipitated with antisera and the immune complexes incubated with [γ-32P]ATP in the presence of manganese ions, the glycoproteins are phosphorylated on tyrosine residues (Barbacid & Lauver, 1981). Using monoclonal antibodies raised against the viral oncogene product, the feline c-fms-coded glycoprotein was similarly identified in detergent homogenates of cat spleen by an immune complex protein kinase assay (Rettenmier et al. 1985b). High levels of the c-fms gene product were also detected in cat peritoneal inflammatory exudates. We reasoned that determination of the phenotype of cells expressing the c-fms-coded glycoprotein might provide an important clue to the identity of its putative ligand. Fluorescence-activated flow cytometry and cell sorting with a monoclonal antibody to a v-fms-coded epitope revealed that the c-fms product was preferentially expressed on mature feline macrophages (Sherr et al. 1985). The restricted distribution of the c-fms gene product suggested that its ligand might be the mononuclear phagocyte colony-stimulating factor CSF-1. A series of collaborative experiments was therefore initiated with Dr E. Richard Stanley who had previously purified and characterized both murine and human CSF-1.

Because antibodies to the feline v-fms protein did not cross-react with the mouse or human c-fms gene product, additional antisera were prepared by immunizing rabbits with a recombinant v-fms-coded polypeptide expressed in bacteria (Furman et al. 1986). The latter antibodies primarily recognize conserved epitopes in the cytoplasmic carboxyl-terminal domain, and specifically precipitated presumptive c-fms products from other species (Sherr et al. 1985; Rettenmier et al. 1986). The mouse c-fms gene product was identified as a glycoprotein of 165K (K = 103 Mr), in close agreement with the molecular weight of the cell surface CSF-1 receptor estimated from chemical cross-linking studies with radiolabelled ligand (Morgan & Stanley, 1984). The murine c-fms gene product also exhibited functional properties of the CSF-1 receptor (Sherr et al. 1985). In membrane preparations, the tyrosine kinase activity of the c-fms-coded glycoprotein was stimulated by CSF-1. Moreover, 125I-ligand-receptor complexes formed at the surface of viable macrophages were quantitatively precipitated by the antisera after detergent lysis. These results established that the c-fms gene product was closely related, and possibly identical, to the CSF-1 receptor.

This conclusion has been confirmed by subsequent investigations. As a consequence of the recombination event that generated SM-FeSV, the v-fms oncogene product retains the complete extracellular domain of its c-fms progenitor (Coussens et al. 1986; Wheeler et al. 1986a). Cells infected by SM-FeSV therefore acquire binding sites for CSF-1, and chemical cross-linking studies demonstrated that specific binding was mediated by the cell surface form of the v-fms-coded glycoprotein (Sacca et al. 1986). Direct genetic evidence that the c-fms gene encodes a functional CSF-1 receptor came from studies using the molecularly cloned c-fms and CSF-1 cDNAs (Roussel et al. 1987). Expression of a human c-fms cDNA in mouse NIH-3T3 cells enables them to form colonies in semi-solid medium when grown in the presence of recombinant human CSF-1. Moreover, co-transfection of NIH-3T3 cells with the human c-fms and CSF-1 genes transforms these cells by an autocrine mechanism. Although the parental cells constitutively produce low amounts of mouse CSF-1 (Sacca et al. 1986), the murine growth factor is not biologically active for human mononuclear phagocytes. Therefore, the requirement for human CSF-1 to stimulate the human c-fms gene product in this system is consistent with the known species specificity of the growth factor. These results also demonstrate that transduction of a CSF-1 receptor, which is normally expressed in mononuclear phagocytes, is sufficient to render fibroblasts responsive to the ligand. The intracellular targets for growth-promoting signals generated by the CSF-1 receptor thus appear to be shared by macrophages and fibroblasts.

A schematic diagram depicting the structure of the v-fms and human c-fms gene products is shown in Fig. 1. The primary translation product of the human c-fms gene is 972 amino acids in length (Coussens et al. 1986). Removal of a 19-residue amino-terminal signal peptide yields a polypeptide of 953 amino acids compared to 922 in the corresponding form of the viral oncogene product (Hampe et al. 1984). The mature cell surface form of the v-fms-coded glycoprotein has an apparent molecular weight of about 140K (Anderson et al. 1984; Manger et al. 1984), although slight variations of size in different cell types are related to differences in the patterns and extent of glycosylation. The mature form of the human c-fms gene product is a glycoprotein of apparent molecular weight 150K (Woolford et al. 1985; Rettenmier et al. 1986). Canonical sequences for A’-1 inked glycosylation (Asn-X-Ser/Thr) are clustered in the extracellular amino-terminal domain, and the carboxyl-terminal domain exhibits sequence similarity with other members of the proteintyrosine kinase gene family. The gpl40V-fms and human gpl50c-fms share extensive amino acid sequence identity in the extracellular (75 %) and cytoplasmic (95 %) domains but differ at their extreme carboxyl termini. As discussed below, this carboxyl-terminal region of divergence includes a single tyrosine residue (Tyr969), which may play a role in modulating the response of the normal receptor to CSF-1.

Fig. 1.

Structure of the v-fms and c-fms gene products. Schematic diagrams of the mature cell surface forms of SM-FeSV gpl40v-fms and the human gp150c-fms are shown. A transmembrane segment of hydrophobic amino acids (black bar) divides the proteins into an extracellular domain for ligand binding and a cytoplasmic tyrosine kinase domain. Potential sites for the addition of asparagine-linked oligosaccharides in the aminoterminal domain are marked by open circles. The portion of the carboxyl-terminal domain that exhibits sequence homology with other protein tyrosine kinases is indicated by asterisks. Within this region of each protein, the position of a critical lysine residue (K) at the ATP-binding site is noted. The segment of 70 fws-specific amino acids that interrupts the core kinase consensus sequence is stippled. The region of carboxyl-terminal sequence divergence between the x-fms and c-fms gene products is indicated by the cross-hatched area in gpl40V-fms. The location of a negative-regulatory tyrosine residue (Y969) near the carboxyl-terminus of gp150c-fms is noted.

Fig. 1.

Structure of the v-fms and c-fms gene products. Schematic diagrams of the mature cell surface forms of SM-FeSV gpl40v-fms and the human gp150c-fms are shown. A transmembrane segment of hydrophobic amino acids (black bar) divides the proteins into an extracellular domain for ligand binding and a cytoplasmic tyrosine kinase domain. Potential sites for the addition of asparagine-linked oligosaccharides in the aminoterminal domain are marked by open circles. The portion of the carboxyl-terminal domain that exhibits sequence homology with other protein tyrosine kinases is indicated by asterisks. Within this region of each protein, the position of a critical lysine residue (K) at the ATP-binding site is noted. The segment of 70 fws-specific amino acids that interrupts the core kinase consensus sequence is stippled. The region of carboxyl-terminal sequence divergence between the x-fms and c-fms gene products is indicated by the cross-hatched area in gpl40V-fms. The location of a negative-regulatory tyrosine residue (Y969) near the carboxyl-terminus of gp150c-fms is noted.

The differences in amino acid sequence of the extracellular domains of the v-fms and human c-fms gene products are largely due to interspecies divergence between cats and man. The amino-terminal domain of the fms-coded glycoprotein lacks the cysteine-rich repeats found in the receptors for EGF and insulin (Coussens et al. 1986). However, the common spacing of individual cysteine residues and short segments of amino acid sequence homology in their ligand-binding domains suggest that the fms product and the receptor for PDGF were derived from a common ancestor (Yarden et al. 1986). Both CSF-1 and PDGF are dimeric growth factors, and the similarity of structure in the extracellular domains of their receptors may be related to conformational requirements for binding dimeric ligands.

The cytoplasmic domain of the fms gene product has sequence similarity with other protein tyrosine kinases. By analogy (Kamps et al. 1984), a lysine residue, located at position 613 of the viral protein and 616 of the normal human receptor, is presumed to be the critical ATP-binding site required for enzymatic activity. A tyrosine residue at positions 807 and 809 of the respective proteins is homologous to the major site of autophosphorylation in many other tyrosine kinases (Hunter & Cooper, 1985). The v-fms gene product is constitutively phosphorylated on tyrosine in transformed cells (Tamura et al. 1986), whereas significant tyrosine phosphorylation of the CSF-1 receptor has not been detected in the absence of ligand. The fms gene products have a segment of about 70 amino acids interposed between the ATP-binding region and the rest of the core kinase consensus sequence. The PDGF receptor (Yarden et al. 1986) and the kit oncogene product (Besmer et al. 1986a) have a similar organization of their respective tyrosine kinase domains. Although there is very little homology among the inserted segments in the kinase domains of these three proteins, the extensive amino acid sequence similarity within their core kinase regions again suggests that they represent a family of closely related receptors. The inserted sequences may affect the activity of these kinases or their interaction with cellular substrates in the signal transduction pathway.

The v-fms and c-fms gene products differ conspicuously at their carboxyl-terminal ends where the 40 carboxyl-terminal amino acids of the normal receptor have been replaced by 11 unrelated residues in the SM-FeSV oncogene product (Hampe et al. 1984; Coussens et al. 1986). The alteration deletes a single tyrosine residue at position 969 (Tyr969) which, by analogy to studies on the products of the v-src and c-src genes (Courtneidge, 1985; Iba et.al. 1985; Cooper et al. 1986), may represent a negative regulatory site of tyrosine phosphorylation. Another independent isolate of feline sarcoma virus contains a v-fms gene encoding a different carboxyl-terminal alteration, which also lacks this terminal tyrosine residue (Besmer et al. 19866). A chimeric N-fms/c-fms construct in which sequences for the normal human c-fms coded carboxyl terminus were substituted in the v-fms gene was more than 10-fold reduced in its transforming efficiency for NIH-3T3 cells (Browning et al. 1986; Roussel et al. 1987). A point mutation replacing Tyr969 with a phenylalanine residue restored the transforming activity of the chimeric construct to the level of y-fms itself (Roussel et al. 1987). NIH-3T3 cells expressing the human c-fms gene with either the wild-type Tyr969 or with the Phe969 mutation were not transformed. However, when plated in semi-solid medium in the presence of recombinant human CSF-1, NIH-3T3 cells expressing the receptor with the Phe969 mutation yielded consistently larger colonies than cells expressing the wild-type c-fms(Tyr969) (Roussel et al. 1987). Moreover, in cells transformed by co-transfection of c-fms alleles with human CSF-1 cDNA, the efficiency of transformation induced by the wild-type c-fms(Tyr969) was about 10-fold less than that of c-fms(Phe969); co-transfection of the latter construct with the CSF-1 gene yielded a transforming efficiency equivalent to that of v-fms.

Taken together, these results suggest that Tyr969 plays a negative regulatory role in the activity of the CSF-1 receptor. Removal of this residue, while not an activating transforming mutation by itself, may up-regulate the receptor kinase activity, thereby enhancing the autocrine transforming efficiency of the receptor in a cotransfection assay with the CSF-1 gene. The fact that v-fms is a transforming gene in the absence of its ligand, whereas c-fms (Phe969) is not, suggests that there are one or more additional mutations in v-fms that contribute to its transforming activity. These latter mutations may mimic a ligand-induced conformational change in the receptor, thus rendering the kinase independent of CSF-1. It should be possible to pinpoint these mutations by comparing the sequence of c-fms with a cDNA of the feline c-fms gene and testing appropriate c-fms mutants for transforming activity.

The CSF-1 receptor exhibits typical properties of a growth factor receptor. Binding of CSF-1 results in autophosphorylation (Sherr et al. 1985; Yeung et al. 1987), rapid internalization, and degradation of both the ligand and the receptor (Guilbert & Stanley, 1986; Wheeler et al. 19866; Downing et al. 1988). The ability to confer a CSF-1-responsive phenotype by expression of the receptor in heterologous NIH-3T3 cells will allow dissection of the components of receptor function through site-directed mutagenesis. For example, appropriate mutants can be constructed to assess the possible role of Tyr809 in activation of the receptor kinase and signal transduction. Reduction of a CSF-1-mediated response by substitution of a phenylalanine residue at this position would suggest that autophosphorylation of Tyr809 in response to ligand binding is a component of receptor activation. By analogy to results in other receptor systems, replacement of lysine616 would be expected to abolish both the kinase activity and signal transduction without affecting ligand binding. Down-modulation of CSF-1 receptors by either CSF-1 or other compounds such as phorbol esters (Wheeler et al. 1986b) could then be assessed directly in the absence of receptor activation.

The intracellular events leading to a CSF-1 response after binding of the ligand are poorly understood. Autophosphorylation at one or more tyrosine residues is probably associated with activation of the receptor. Ligand-receptor complexes are returned to the interior of the cell via the endocytic pathway and rapidly degraded. The signals for receptor down-modulation and the identity of cellular substrates for the receptor kinase are not known. The production of intracellular second messengers in response to receptor activation may be involved. Cells transformed by the v-fms oncogene exhibit an elevated rate of phosphatidylinositol (PI) turnover (Jackowski et al. 1986) with cleavage of phosphatidylinositol 4,5-diphosphate (PIP2) generating inositol-l,4,5-triphosphate to mobilize intracellular calcium and diacylglycerol, which activates protein kinase C. This is associated with increased activities of a membrane-associated PIP2 phospholipase C (Jackowski et al. 1986) and a PI kinase (Kaplan et al. 1987). However, it has been reported that CSF-1 stimulation is not associated with PI turnover in macrophages (Whetton et al. 1986). Thus, a role for PIP2-derived second messengers in signal transduction by the normal receptor has not been established.

The sequence of events leading to changes in nuclear function after CSF-1 stimulation are also unknown. As for a variety of stimuli in many other cell systems, CSF-1 rapidly induces transient high levels of expression of the c-fos proto-oncogene within 15 min in macrophages (Bravo et al. 1987), followed by a sustained increase in c-fos expression at lower levels 4-12h after stimulation (Müller et al. 1985). The addition of CSF-1 also leads to a decrease in the rate of protein turnover within 2h (Tushinski & Stanley, 1983) and stimulation of DNA synthesis within 8—12h (Tushinski & Stanley, 1985). Clearly, considerable effort will be necessary to elucidate the signal transduction mechanism initiated by binding of CSF-1 to its receptor. Working from both ends of the pathway by defining the physiologically relevant substrates for the CSF-1 receptor kinase and identifying CSF-1-inducible genes is a reasonable approach to this difficult and important problem.

CSF-1 is a glycosylated polypeptide homodimer of identical subunits assembled through disulphide bonds (Stanley & Heard, 1977). The biological activity of the growth factor is dependent on its dimeric structure and is abolished by disulphide reduction. The carbohydrate component includes TV-linked oligosaccharides, which do not appear to be required for CSF-1 function (Das & Stanley, 1982). The human growth factor has been purified from urine, and oligonucleotide probes deduced from the amino-terminal amino acid sequence were used to clone genomic sequences coding for CSF-1. Human CSF-1 is encoded by a single gene, which maps to the long arm of chromosome 5 (Pettenati et al. 1987). Transcripts of this gene undergo differential splicing to generate multiple mRNAs (Ladner et al. 1987), which specify at least two biologically active forms of the growth factor (Kawasaki et al. 1985; Wonget al. 1987). A cDNA encoding murine CSF-1 has also been isolated from the mouse L929 fibroblast cell line (DeLamarter et al. 1987).

A schematic diagram of two primary translation products deduced from the nucleotide sequences of various human and murine CSF-1 cDNAs is shown in Fig. 2. Analysis of the sequences suggested that soluble CSF-1 is generated by proteolytic cleavage of a membrane-bound precursor. All of the precursors contain a 32-residue amino-terminal signal peptide for membrane insertion followed by sequences of the secreted growth factor, a presumptive transmembrane segment of hydrophobic amino acids, and a short carboxyl-terminal tail. The cysteine residues and potential sites for TV-linked glycosylation are present on the amino-terminal side of the putative membrane-spanning sequence. These features suggested that the CSF-1 precursor is synthesized as an integral transmembrane glycoprotein cotransla-tionally oriented in the membrane of the endoplasmic reticulum (ER) with its glycosylated amino-terminal domain in the ER cisterna and its carboxyl-terminal tail in the cytoplasm.

Fig. 2.

Primary translation products encoded by biologically active human and mouse CSF-1 cDNAs. Each CSF-1 precursor consists of an amino-terminal signal peptide (black bar) and sequence for the secreted growth factor (stippled), a transmembrane segment (black bar), and a carboxyl-terminal tail. Canonical sequences for TV-linked glycosylation are indicated by open circles. Human CSF-1256 and CSF-1554 are encoded by mRNAs derived by alternative splicing of the primary transcript. A contiguous segment of 298 amino acids present in human CSF-1554 and absent in CSF-1256 is delimited by the dashed lines. Polypeptide sequences from this insert, which are present in the soluble growth factor derived from CSF-1554, are cross-hatched. Murine CSF-1552 is similar in structure to the large human CSF-1 precursor. The extent of amino acid sequence identity between the corresponding regions of the mouse and human CSF-l precursors is indicated.

Fig. 2.

Primary translation products encoded by biologically active human and mouse CSF-1 cDNAs. Each CSF-1 precursor consists of an amino-terminal signal peptide (black bar) and sequence for the secreted growth factor (stippled), a transmembrane segment (black bar), and a carboxyl-terminal tail. Canonical sequences for TV-linked glycosylation are indicated by open circles. Human CSF-1256 and CSF-1554 are encoded by mRNAs derived by alternative splicing of the primary transcript. A contiguous segment of 298 amino acids present in human CSF-1554 and absent in CSF-1256 is delimited by the dashed lines. Polypeptide sequences from this insert, which are present in the soluble growth factor derived from CSF-1554, are cross-hatched. Murine CSF-1552 is similar in structure to the large human CSF-1 precursor. The extent of amino acid sequence identity between the corresponding regions of the mouse and human CSF-l precursors is indicated.

A l·6-kb CSF-1 cDNA cloned from a human pancreatic carcinoma cell line encodes a 256 amino acid CSF-1 precursor (Kawasaki et al. 1985). The 554-residue polypeptide specified by a 4-kb human cDNA from this and other sources includes the complete coding sequence of the smaller clone, with a block of 298 additional amino acids inserted after residue 149 on the amino-terminal side of the transmembrane segment (Wong et al. 1987; Ladner et al. 1987). The two human cDNAs are generated by alternative splicing of the primary transcript (Ladner et al. 1987). The additional coding sequences of the 4-kb cDNA include some tryptic peptides detected in a purified form of human urinary CSF-1 (Wong et al. 1987). Thus, the soluble growth factors specified by these two cDNAs have different carboxyl termini. The 552 amino acid murine CSF-l precursor encoded by a cDNA cloned from mouse L929 cells (DeLamarter et al. 1987) is similar in structure to the product of the 4-kb human cDNA. The sequences of the soluble human and murine growth factors (residues 1-ca. 189) exhibit more than 80% amino acid identity. The remainder of the inserted segment in the larger CSF-l precursor shows less amino acid sequence conservation than the secreted growth factor.

Expression of CSF-l cDNAs in viral vectors has allowed biochemical analysis of the products. The l·6-kb human CSF-l cDNA directs the synthesis of a membranebound 34K glycoprotein, which is rapidly assembled into disulphide-linked dimers and externalized on the plasma membrane (Rettenmier et al. 1987). Proteolysis of this 68K homodimer at the cell surface yields soluble 44K CSF-1 dimers composed of 22K subunits. In cell culture systems, less than 10% of the precursors on the plasma membrane are recovered in the form of extracellular growth factor. The role of the membrane-spanning sequence has been directly tested by insertion of an upstream termination codon after amino acid 158 in the coding sequence of the l·6-kb cDNA (Heard et al. 1987a). The truncated clone produces a biologically active, soluble growth factor, which is not membrane-associated and is efficiently secreted from cells. The similarity in size between the product of the truncated clone and the soluble growth factor released by cleavage of the membrane-bound precursor indicates that the termination codon was placed in the coding sequence of the 1·6-kb cDNA near the site of proteolysis.

Soluble CSF-1 encoded by the 4-kb human cDNA includes additional carboxyl-terminal amino acid residues not present in the secreted product of the smaller clone (Wong et al. 1987). Thus, the sites for proteolytic cleavage are different for the two CSF-1 precursors, and it appears that they are processed by different mechanisms. In contrast to the product of the l·6-kb cDNA, which is detected at the cell surface, the CSF-1 precursor encoded by the 4-kb clone is cleaved within the cell and rapidly secreted. Differences in the mechanism of CSF-1 biosynthesis may reflect diverse physiological roles for various forms of the growth factor. Although the 4-kb mRNA is the major CSF-1 transcript detected in a variety of human cell lines, in at least one inducible system, production of biologically active human CSF-1 is correlated with the appearance of the smaller species (Ralph et al. 1986). In addition, two mRNAs of the corresponding sizes are differently expressed in various mouse tissues (Raja-vashisth et al. 1987). The function of the different CSF-1 precursors in vivo is as yet unclear. One possibility is that product of the 4-kb cDNA is the major source of CSF-1 released into the circulation, whereas the membrane-bound form of CSF-1 encoded by the l·6-kb clone primarily stimulates receptor-bearing cells in circumstances involving direct cell-cell contact.

CSF-1 is produced by mesenchymal cells and interacts with its receptor on mononuclear phagocytes (Tushinski et al. 1982). The growth factor appears in the circulation, and its concentration is determined by monocytes and macrophages, which bind and degrade the hormone after receptor-mediated endocytosis (Bartocci et al. 1987). The pleiotropic effects of CSF-1 include maintaining the viability of mature monocytes and macrophages and stimulating the proliferation and differentiation of committed bone marrow progenitors (Stanley et al. 1983). More immature haematopoietic precursors are also rendered responsive to CSF-1 in the presence of interleukin-3 (Multi-CSF) or haemopoietin-1 (Bartelmez & Stanley, 1985; Stanley et al. 1986). The latter molecule is now known to be the macrophage product, interleukin-1α (Mochizuki et al. 1987), indicating that differentiated macrophages can release mediators that stimulate the production of immature bone marrow progenitors. Positive and negative feed-back mechanisms therefore regulate the production of mononuclear phagocytes. In addition, it has recently been reported that monocytes and macrophages themselves produce CSF-1 in response to phorbolesters and the physiological inducers, γ-interferon and granulocyte-macrophage CSF (Horiguchi et al. 1986, 1987; Rambaldi et al. 1987). This suggests that autocrine or paracrine mechanisms might also be involved in modulating the activity of these cells during inflammatory reactions.

An intriguing area for future investigations is a possible function for CSF-l and its receptor beyond their role in the haematopoietic compartment. Transcripts of the c-fms gene are found in mouse and human placental tissue (Müller et al. 1983a,b), and the c-fms-coded glycoprotein has been identified in two independently derived human choriocarcinoma cell lines (Woolford et al. 1985; Rettenmier et al. 1986). The c-fms product in these malignant trophoblasts is indistinguishable from that detected on normal peripheral blood monocytes, and both cell types display high-affinity binding sites for CSF-l (Rettenmier et al. 1986). Pregnancy results in a 1000fold increase in the concentration of CSF-l in the mouse uterus with the highest levels being detected at term (Bartocci et al. 1986). During the latter half of gestation, placental concentrations of the growth factor are higher than those detected in foetal tissues or in a variety of maternal organs. The simplest interpretation is that CSF-l produced by stromal cells of the uterus stimulates receptors expressed on normal placental trophoblasts during embryogenesis. Thus, although CSF-l was initially identified as a lineage-specific colony-stimulating factor in haematopoiesis, it is likely that this growth factor has additional physiological roles in other circumstances.

Due to its identification as the product of the c-fms gene, the ability of the CSF-l receptor to induce malignant cell transformation has already been tested in a number of contexts. The c-fms gene was isolated from a feline fibrosarcoma, and it transforms a variety of fibroblast cell lines in culture. Because these mesenchymal cells also produce CSF-l, it was possible that transformation was mediated by an autocrine mechanism involving transduction of a competent CSF-l receptor into cells that synthesized the corresponding growth factor. To test this hypothesis, the c-fms gene was expressed in murine haematopoietic cell lines that require colony-stimulating factors for their proliferation and survival in culture and do not produce CSF-l. Introduction of the c-fms gene induced growth factor independence and tumorigen-icity in the CSF-l-dependent macrophage cell line BAC1.2F5 (Wheeler et al. 1986b) and the interleukin-3-dependent myeloid cell line FDC-P1 (Wheeler et al. 1987). The factor-independent cells did not synthesize CSF-l, nor did expression of the v-fms -coded glycoprotein alter the number or affinity of the receptors for the corresponding growth factors required by the parental cells. Thus, transformation of these cells did not involve an autocrine mechanism, and there was no evidence for transmodulation of the murine CSF-l or IL-3 receptors by the c-fms gene product.

The c-fms gene product is expressed at high levels by mature mononuclear phagocytes, and introduction of the c-fms gene into NIH-3T3 cells did not induce transformation (Roussel et al. 1987). These findings indicate that the v-fms gene product is constitutively activated to provide growth-stimulatory signals in the absence of ligand as a result of critical structural alterations. As discussed above, one of the activating mutations involves removal of a carboxyl-terminal tyrosine residue that may modulate the response of the normal receptor to ligand binding. Because the latter mutation alone is insufficient to activate the transforming potential of the c-fms gene, the v-fms-coded glycoprotein must contain one or more additional alterations that contribute to its oncogenic properties.

Expression of the human c-fms gene enables mouse NIH-3T3 fibroblasts to respond mitogenically to exogenously added human CSF-1 (Roussel et al. 1987). Constitutive co-expression of human c-fms and CSF-1 cDNAs transforms these cells by an autocrine mechanism. Rearrangement and expression of the endogenous murine CSF-1 gene has been implicated as a second transforming event leading to tumorigenicity in a CSF-l-dependent macrophage cell line immortalized by a retrovirus containing the c-myc oncogene (Baumbach et al. 1987). These results support the notion that unregulated expression of a growth factor and its receptor in the same cell may contribute to neoplastic progression. However, this circumstance by itself does not appear to be sufficient fully to transform all established cell lines. In the SV40-immortalized BAC1.2F5 mouse macrophage cell line (Morgan et al. 1987), expression of an introduced human CSF-1 gene abrogated the requirement of exogenously added CSF-1 for growth in vitro, but the factor-independent cells were not tumorigenic in nude mice (Roussel et al. 1988). Human CSF-1 is biologically active for stimulation of murine mononuclear phagocytes, and the factor-independent cells produced sufficient amounts of the growth factor to constitutively down-modulate their CSF-1 receptors. The fact that the v-fms gene did render these same cells factor-independent and tumorigenic (Wheeler et al. 1986b) underscores the critical differences between the viral transforming protein and the normal receptor.

The results of these studies in cell culture systems are currently being extended to expression of these genes in vivo. Although SM-FeSV encodes an analogue of the CSF-1 receptor, it was isolated from a mesenchymal tumour and has not been reported to induce haematopoietic neoplasms in its natural host. However, when mouse bone marrow cells containing an introduced v-fms gene were used to repopulate lethally irradiated recipients, diseases of multiple haematological lineages were observed (Heard et al. 1987b). After a latency period, several primary recipients developed myeloproliferative disorders, which lacked evidence of clonality and were not efficiently transplanted to secondary hosts. In contrast, other primary recipients developed in their spleens provirus-positive clones, which were transplantable to secondary, lethally-irradiated recipients and gave rise to clonal erythroleukaemias and B-cell lymphomas. Although cells expressing v-fms did not appear to have an obligate proliferative advantage during serial transfers, the correlation between expression of the v-fms gene and disease indicated that the oncogene product contributed to the development of these disorders. Clearly, similar types of experiments using molecular clones of the c-fms and CSF-1 genes will be of interest.

Evidence has recently been obtained for involvement of the CSF-1 receptor in the development of a naturally occurring leukaemia. The replication-competent Friend murine leukaemia virus induces myeloblastic leukaemia in mice with a latency of 6-12 months. In approximately 20% of these cases the provirus is integrated in a region upstream of the first coding exon of the c-fms gene in host cell DNA (Gisselbrecht et al. 1987). This proviral insertion results in high level expression of c-fms transcripts that encode the full-sized CSF-1 receptor. These myeloblasts might be unusually responsive to CSF-1 due to the high levels of receptor synthesis. Leukaemia might ultimately develop in the clonally expanded population as a result of secondary genetic changes involving either the c-fms gene itself or other cellular genes. The loss of the germ line c-fms allele in some of these tumours is consistent with the possibility that activating mutations may also have occurred in the affected c-fms locus to unmask the latent transforming potential of its product and provide a further proliferative advantage. Whatever the mechanism, the fact that the c-fms gene is a target for proviral insertion resulting in transcriptional activation indicates that this is an initiating event for the development of leukaemia in these animals.

The genes for several CSFs and cell surface receptors have been localized to the long arm of human chromosome 5, including c-fms at 5q33.2-33.3 (Le Beau et al. l986n) and CSF-1 at 5q33.l (Pettenatieí al. 1987). Acquired interstitial deletions of 5q are a frequent cytogenetic abnormality in patients who develop acute myelogenous leukaemia after exposure to toxic chemicals (Rowley et al. 1981; PedersenBjergaard et al. 1984; Le Beau et al. 1986b). Deletions of this region are also characteristic of the ‘5q- syndrome’ (Van den Berghe et al. 1974), a refractory anaemia with thrombocytosis and hypolobulated megakaryocytes (Sokal et al. 1975; Wieniewski & Hirschhorn, 1983). Bone marrow cells from some patients with the 5q- syndrome have been shown to be hemizygous for the c-fms gene (Nienhuis et al. 1985). These results raise the possibility that alterations in the structure or expression of CSF-1 or its receptor may contribute to these haematological disorders in man. A careful dissection of the genes encoding growth factors and their receptors that are clustered in this region of chromosome 5 will be required to resolve these possibilities.

Work in the authors’ laboratory is supported by grants CA38187 and CA20180 from the National Cancer Institute, National Institutes of Health and by the American Lebanese Syrian Associated Charities (ALSAC) of St. Jude Children’s Research Hospital.

Anderson
,
S. J.
,
Furth
,
M.
,
Wolff
,
L.
,
Ruscetti
,
S. K.
&
Sherr
,
C. J.
(
1982
).
Monoclonal antibodies to the transformation-specific glycoprotein encoded by the feline retroviral oncogene v-fms
.
J. Virol
.
44
,
696
702
.
Anderson
,
S. J.
,
Gonda
,
M. A.
,
Rettenmier
,
C. W.
&
Sherr
,
C. J.
(
1984
).
subcellular localization of glycoproteins encoded by the viral oncogene v-fms
.
J. Virol
.
51
,
730
741
.
Barbacid
,
M.
&
Lauver
,
A. V.
(
1981
).
Gene products of McDonough feline sarcoma virus have an in vitro-associated protein kinase that phosphorylates tyrosine residues: Lack of detection of this enzymatic activity in vivo
.
J. Virol
.
40
,
812
821
.
Bartelmez
,
S. H.
,
Sacca
,
R.
&
Stanley
,
E. R.
(
1985
).
Lineage specific receptors used to identify a growth factor for developmentally early hemopoietic cells: Assay of hemopoietin-2
.
J. cell. Physiol
.
122
,
362
369
.
Bartelmez
,
S. H.
&
Stanley
,
E. R.
(
1985
).
Synergism between hemopoietic growth factors (HGFs) detected by their effects on cells bearing receptors for a lineage specific HGF: Assay of hemopoietin-1
J. cell. Physiol
.
122
,
370
378
.
Bartocci
,
A.
,
Mastrogiannis
,
D. S.
,
Migliorati
,
G.
,
Stockert
,
R. J.
,
Wolkoff
,
A. W.
&
Stanley
,
E. R.
(
1987
).
Macrophages specifically regulate the concentration of their own growth factor in the circulation
.
Proc. natn. Acad. Sci. U.S.A
.
84
,
6179
6183
.
Bartocci
,
A.
,
Pollard
,
J. W.
&
Stanley
,
E. R.
(
1986
).
Regulation of colony-stimulating factor 1 during pregnancy
.
J. exp. Med
.
164
,
956
961
.
Baumbach
,
W. R.
,
Stanley
,
E. R.
&
Cole
,
M. D.
(
1987
).
Induction of clonal monocytemacrophage tumours in vivo by a mouse c-myc retrovirus: Rearrangement of the CSF-1 gene as a secondary transforming event
.
Molec. Cell. Biol
.
7
,
664
671
.
Besmer
,
P.
,
Lader
,
E.
,
George
,
P. C.
,
Bergold
,
P. J.
,
Qiu
,
F-H.
,
Zuckerman
,
E. E.
&
Hardy
,
W. D.
(
1986b
).
A new acute transforming feline retrovirus with fms homology specifies a C-terminally truncated version of the c-fms protein that is different from SM-feline sarcoma virus v-fms protein
.
J. Virol
.
60
,
194
203
.
Besmer
,
P.
,
Murphy
,
J. E.
,
George
,
P. C.
,
Qiu
,
F.
,
Bergold
,
P. J.
,
Lederman
,
L.
,
Snyder
,
H. W.
Jr
,
Brodeur
,
D.
,
Zuckerman
,
E. E.
&
Hardy
,
W. D.
(
1986a
).
A new acute transforming feline retrovirus and relationship of its oncogene v-kit with the protein kinase gene family
.
Nature, Lond
.
320
,
415
421
.
Bravo
,
R.
,
Neuberg
,
M.
,
Burckhardt
,
J.
,
Almendral
,
J.
,
Wallich
,
R.
&
Müller
,
R.
(
1987
).
Involvement of common and cell type-specific pathways in c-fos gene control: Stable induction by cAMP in macrophages
.
Cell
48
,
251
260
.
Browning
,
P. J.
,
Bunn
,
H. F.
,
Cline
,
A.
,
Shuman
,
M.
&
Nienhuis
,
A. W.
(
1986
).
“Replacement” of COOH-terminal truncation of v-fms with c-fms sequences markedly reduces transformation potential
.
Proc. natn. Acad. Sci. U.S.A
.
83
,
7800
7804
.
Cooper
,
J. A.
,
Gould
,
K. L.
,
Cartwright
,
C. A.
&
Hunter
,
T.
(
1986
).
Tyr527 is phosphorylated in pp60c-src Implications for regulation
.
Science
231
,
1431
1434
.
Courtneidge
,
S. A.
(
1985
).
Activation of the ρp60c-src kinase by middle T antigen binding or by dephosphorylation
.
EMBO J
.
4
,
1471
1477
.
Coussens
,
L.
,
van Beveren
,
C.
,
Smith
,
D.
,
Chen
,
E.
,
Mitchell
,
R. L.
,
Isacke
,
C. M.
,
Verma
,
I. M.
&
Ullrich
,
A.
(
1986
).
Structural alteration of viral homologue of receptor proto-oncogene fms at carboxyl terminus
.
Nature, Lond
.
320
,
277
280
.
Das
,
S. K.
&
Stanley
,
E. R.
(
1982
).
Structure-function studies of a colony stimulating factor (CSF-l)
.
J. biol. Chem
.
257
,
13 679
13 684
.
Delamarter
,
J. F.
,
Hession
,
C.
,
Semon
,
D.
,
Gough
,
N. M.
,
Rothenbuhler
,
R.
&
Mermod
,
J.-J.
(
1987
).
Nucleotide sequence of a cDNA encoding murine CSF-1 (macrophage-CSF)
.
Nucl. Acids Res
.
15
,
2389
2390
.
Donner
,
L.
,
Fedele
,
L. A.
,
Garon
,
C. F.
,
Anderson
,
S. J.
&
Sherr
,
C. J.
(
1982
).
McDonough feline sarcoma virus: characterization of the molecularly cloned provirus and its feline oncogene (v-fms)
.
J. Virol
.
41
,
489
500
. .
Doolittle
,
R. F.
,
Hunkapiller
,
M. W.
,
Hood
,
L. E.
,
Devare
,
S. G.
,
Robbins
,
K. C.
,
Aaronson
,
S. A.
&
Antoniades
,
H. N.
(
1983
).
Simian sarcoma virus one gene, v-sis, is derived from the gene (or genes) encoding a platelet-derived growth factor
.
Science
221
,
275
277
.
Downing
,
J. R.
,
Rettenmier
,
C. W.
&
Sherr
,
C. J.
(
1988
).
Ligand-induced tyrosine kinase activity of the colony-stimulating factor 1 receptor in a murine macrophage cell line
.
Molec. cell. Biol
.
8
,
1797
1799
.
Downward
,
J.
,
Yarden
,
Y.
,
Mayes
,
E.
,
Scrace
,
G.
,
Totty
,
N.
,
Stockwell
,
P.
,
Ullrich
,
A.
,
Schlessinger
,
J.
&
Waterfield
,
M. D.
(
1984
).
Close similarity of epidermal growth factor receptor and v-erbB oncogene protein sequences
.
Nature, Lond
.
307
,
521
527
.
Furman
,
W. L.
,
Rettenmier
,
C. W.
,
Chen
,
J. H.
,
Roussel
,
M. F.
,
Quinn
,
C. O.
&
Sherr
,
C. J.
(
1986
).
Antibodies to distal carboxyl terminal epitopes in the v-fms-coded glycoprotein do not cross-react with the c-fms gene product
.
Virology
152
,
432
445
.
Gisselbrecht
,
S.
,
Fichelson
,
S.
,
Sola
,
B.
,
Bordereaux
,
D.
,
Hampe
,
A.
,
Andre
,
C.
,
Galibert
,
F.
&
Tambourin
,
P.
(
1987
).
Frequent c-fms activation by proviral insertion in mouse myeloblastic leukaemias
.
Nature, Land
.
329
,
259
261
.
Gonda
,
T. J.
&
Metcalf
,
D.
(
1984
).
Expression of myb, myc and fos proto-oncogenes during the differentiation of a murine myeloid leukaemia
.
Nature, Land
.
310
,
249
251
.
Guilbert
,
L. J.
&
Stanley
,
E. R.
(
1980
).
Specific interaction of murine colony-stimulating factor with mononuclear phagocytic cells
.
J. Cell Biol
.
85
,
153
159
.
Guilbert
,
L. J.
&
Stanley
,
E. R.
(
1986
).
The interaction of 125I-colony-stimulating factor-1 with bone marrow-derived macrophages
.
J. biol. Chem
.
261
,
4024
4032
.
Hadwiger
,
A.
,
Niemann
,
H.
,
KÄbisch
,
A.
,
Bauer
,
H.
&
Tamura
,
T.
(
1986
).
Appropriate glycosylation of thefms gene product is a prerequisite for its transforming potency
.
EMBO J
.
5
,
689
694
.
Hampe
,
A.
,
Gobet
,
M.
,
Sherr
,
C. J.
&
Galibert
,
F.
(
1984
).
Nucleotide sequence of the feline retroviral oncogene v-fms shows unexpected homology with oncogenes encoding tyrosinespecific protein kinases
.
Proc. natn. Acad. Sci. U.S.A
.
81
,
85
89
.
Heard
,
J. M.
,
Roussel
,
M. F.
,
Rettenmier
,
C. W.
&
Sherr
,
C. J.
(
1987a
).
Synthesis, post-translational processing, and autocrine transforming activity of a carboxylterminal truncated form of colony stimulating factor-1
.
Oncogene Res
.
1
,
423
440
.
Heard
,
J. M.
,
Roussel
,
M. F.
,
Rettenmier
,
C. W.
&
Sherr
,
C. J.
(
1987b
).
Multilineage hematopoietic disorders induced by transplantation of bone marrow cells expressing the v-fms oncogene
.
Cell
51
,
663
673
.
Horiguchi
,
J.
,
Warren
,
M. K.
&
Kufe
,
D.
(
1987
).
Expression of the macrophage-specific colony-stimulating factor in human monocytes treated with granulocyte-macrophage colonystimulating factor
.
Blood
69
,
1259
1261
.
Horiguchi
,
J.
,
Warren
,
M. K.
,
Ralph
,
P.
&
Kufe
,
D.
(
1986
).
Expression of the macrophage specific colony-stimulating factor (CSF-l) during human monocytic differentiation
.
Biochem. biophys. Res. Commun
.
141
,
924
930
.
Hunter
,
T.
&
Cooper
,
J. A.
(
1985
).
Protein-tyrosine kinases
.
A. Rev. Biochem
.
54
,
897
930
.
Iba
,
H.
,
Cross
,
F. R.
,
Garber
,
E. A.
&
Hanafusa
,
H.
(
1985
).
Low level of cellular protein phosphorylation by nontransforming overproduced p60c-src
.
Molec. Cell Biol
.
5
,
1058
1066
.
Jackowski
,
S
,
Rettenmier
,
C. W.
,
Sherr
,
C. J.
&
Rock
,
C. O.
(
1986
).
A guanine nucleotidedependent phosphatidylinositol-4,5-disphosphate phospholipase C in cells transformed by the v-fms and v-fes oncogenes
.
J. biol. Chem
.
261
,
4978
4985
.
Kamps
,
M. P.
,
Taylor
,
S. S.
&
Sefton
,
B. M.
(
1984
).
Direct evidence that oncogenic tyrosine kinases and cyclic AMP-dependent protein kinase have homologous ATP-binding sites
.
Nature, Land
.
310
,
589
592
.
Kaplan
,
D. R.
,
Whitman
,
M.
,
Schaffhausen
,
B.
,
Pallas
,
D. C.
,
White
,
M.
,
Cantley
,
L.
&
Roberts
,
T. M.
(
1987
).
Common elements in growth factor stimulation and oncogenic transformation: 85kd phosphoprotein and phosphatidylinositol kinase activity
.
Cell
50
,
1021
1029
.
Kawasaki
,
E. S.
,
Ladner
,
M. B.
,
Wang
,
A. M.
,
van Arsdell
,
J.
,
Warren
,
M. K.
,
Coyne
,
M. Y.
,
Schweickart
,
V. L.
,
Lee
,
M. T.
,
Wilson
,
K. J.
,
Boosman
,
A.
,
Stanley
,
E. R.
,
Ralph
,
P.
&
Mark
,
D. F.
(
1985
).
Molecular cloning of a complementary DNA encoding human macrophage-specific colony stimulating factor (CSF-l)
.
Science
230
,
291
296
.
Ladner
,
M. B.
,
Martin
,
G. A.
,
Noble
,
J. A.
,
Nikoloff
,
D. M.
,
Tal
,
R.
,
Kawasaki
,
E. S.
&
White
,
T. J.
(
1987
).
Human CSF-l: Gene structure and alternative splicing of mRNA precursors
.
EMBO J
.
6
,
2693
2698
.
Le Beau
,
M. M.
,
Albain
,
K. S.
,
Larson
,
R. A.
,
Vardiman
,
J. W.
,
Davis
,
E. M.
,
Blough
,
R. R.
,
Golomb
,
H. M.
&
Rowley
,
J. D.
(
1986b
).
Clinical and cytogenetic correlations in 63 patients with therapy-related myelodysplastic syndromes and acute nonlymphocytic leukemia: Further evidence for characteristic abnormalities of chromosomes No. 5 and 7
.
J. clin. Oncol
.
4
,
325
345
.
Le Beau
,
M. M.
,
Westbrook
,
C. A.
,
Diaz
,
M. O.
,
Larson
,
R. A.
,
Rowley
,
J. D.
,
Gasson
,
J. C.
,
Golde
,
D. W.
&
Sherr
,
C. J.
(
1986a
).
Evidence for the involvement of GM-CSF andfms in the deletion (5q) in myeloid disorders
.
Science
231
,
984
987
.
Manger
,
R.
,
Najita
,
L.
,
Nichols
,
E. J.
, Hakomori, S.-I. &
Rohrschneider
,
L.
(
1984
).
Cell surface expression of the McDonough strain of feline sarcoma virus fins gene product (gpl40fms
.
Cell
39
,
327
337
.
McDonough
,
S. K.
,
Larsen
,
S.
,
Brodey
,
R. S.
,
Stock
,
N. D.
&
Hardy
,
W. D.
Jr
(
1971
).
A transmissible feline fibrosarcoma of viral origin
.
Cancer Res
.
31
,
953
956
.
Metcalf
,
D.
&
Nicola
,
A.
(
1985
).
Synthesis by mouse peritoneal cells of G-CSF, the differentiation inducer for myeloid leukemia cells: Stimulation by endotoxin, M-CSF, and multi-CSF
.
Leukaemia Res
.
9
,
35
50
.
Mochizuki
,
D. Y.
,
Eisenman
,
J. R.
,
Conlon
,
P. J.
,
Larsen
,
A. D.
&
Tushinski
,
R. J.
(
1987
).
Interleukin 1 regulates hematopoietic activity, a role previously ascribed to hemopoietin 1
.
Proc, natn. Acad. Sci. U.S A
.
84
,
5267
5271
.
Morgan
,
C.
,
Pollard
,
J. W.
&
Stanley
,
E. R.
(
1987
).
Isolation and characterization of a cloned growth factor dependent macrophage cell line, BAC1.2F5
.
J. cell Physiol
.
130
,
420
427
.
Morgan
,
C. J.
&
Stanley
,
E. R.
(
1984
).
Chemical crosslinking of the mononuclear phagocyte specific growth factor CSF-1 to its receptor at the cell surface
.
Biochem. biophys. Res. Comm
.
119
,
35
41
.
Müller
,
R.
,
Curran
,
T.
,
Müller
,
D.
&
Guilbert
,
L.
(
1985
).
Induction of c-fos during myelomonocytic differentiation and macrophage proliferation
.
Nature, Land
.
314
,
546
548
.
Müller
,
R.
,
Slamon
,
D. J.
,
Adamson
,
E. D.
,
Tremblay
,
J. M.
,
Müller
,
D.
,
Cline
,
M. J.
&
Verma
,
I. M.
(
1983a
).
Transcription of c-onc genes c-raski and c-fms during mouse development
.
Molec. Cell. Biol
.
3
,
1062
1069
.
Müller
,
R.
,
Tremblay
,
J. M.
,
Adamson
,
E. D.
&
Verma
,
I. M.
(
1983b
).
Tissue and cell type specific expression of two human c-onc genes
.
Nature, Land
.
304
,
454
456
.
Nichols
,
E. J.
,
Manger
,
R.
,
Hakomori
,
S.
,
Herscovics
,
A.
&
Rohrschneider
,
L. R.
(
1985
).
Transformation by the v-fms oncogene product: Role of glycosylational processing and cell surface expression
.
Molec. Cell Biol
.
5
,
3467
3475
.
Nienhuis
,
A. W.
,
Bunn
,
H. F.
,
Turner
,
P. H.
,
Gopal
,
T. V.
,
Nash
,
W. G.
,
O’Brien
,
S. J.
&
Sherr
,
C. J.
(
1985
).
Expression of the human c-fms proto-oncogene in hematopoietic cells and its deletion in the 5q- syndrome
.
Cell
42
,
421
428
.
Pedersen-Bjergaard
,
J.
,
Philip
,
P.
,
Pedersen
,
N. T.
,
Hou-Jensen
,
K.
,
Svejgaard
,
A.
,
Jensen
,
G.
&
Nissen
,
N. I.
(
1984
).
Acute nonlymphocytic leukemia, preleukemia, and acute myeloproliferative syndrome secondary to treatment of other malignant diseases. II. Bone marrow cytology, cytogenetics, results of HLA typing, response to antileukemic chemotherapy, and survival in a total series of 55 patients
.
Cancer
54
,
452
462
.
Pettenati
,
M. J.
,
Le Beau
,
M. M.
,
Lemons
,
R. S.
,
Shima
,
E. A.
,
Kawasaki
,
E. S.
,
Larson
,
R. A.
,
Sherr
,
C. J.
,
Diaz
,
M. O.
&
Rowley
,
J. D.
(
1987
).
Assignment of CSF-1 to 5q33.l: Evidence for clustering of genes regulating hematopoiesis and for their involvement in the deletion of the long arm of chromosome 5 in myeloid disorders
.
Proc. natn. Acad. Sci. U.S.A
.
84
,
2970
2974
.
Rajavashisth
,
T. B.
,
Eng
,
R.
,
Shadduck
,
R. K.
,
Waheed
,
A.
,
Ben-Avram
,
C. M.
,
Shively
,
J. E.
&
Lusis
,
A. J.
(
1987
).
Cloning and tissue-specific expression of mouse macrophage colonystimulating factor mRNA
.
Proc. natn. Acad. Sci. U.S.A
.
84
,
1157
1161
.
Ralph
,
P.
&
Nakoinz
,
I.
(
1987
).
Stimulation of macrophage tumoricidal activity by the growth and differentiation factor CSF-1
.
Cell. Immunol
.
105
,
270
279
.
Ralph
,
P.
,
Warren
,
M. K.
,
Lee
,
M. T.
,
Csejtey
,
J.
,
Weaver
,
J. F.
,
Broxmeyer
,
H. E.
,
Williams
,
D. E.
,
Stanley
,
E. R.
&
Kawasaki
,
E. S.
(
1986
).
Inducible production of human macrophage growth factor, CSF-1
.
Blood
68
,
633
639
.
Rambaldi
,
A.
,
Young
,
D. C.
&
Griffin
,
J. D.
(
1987
).
Expression of the M-CSF (CSF-1) gene by human monocytes
.
Blood
69
,
1409
1413
.
Rettenmier
,
C. W.
,
Chen
,
J. H.
,
Roussel
,
M. F.
&
Sherr
,
C. J.
(
1985b
).
The product of the c-fms proto-oncogene: a glycoprotein with associated tyrosine kinase activity
.
Science
228
,
320
322
.
Rettenmier
,
C. W.
,
Roussel
,
M. F.
,
Ashmun
,
R. A.
,
Ralph
,
P.
,
Price
,
K.
&
Sherr
,
C. J.
(
1987
).
Synthesis of membrane-bound colony-stimulating factor-1 (CSF-1) and downmodulation of CSFT receptors in NIH 3T3 cells transformed by cotransfection of the human CSF-1 and c-fms (CSF-1 receptor) genes
.
Molec. Cell. Biol
.
7
,
2378
2387
.
Rettenmier
,
C. W.
,
Roussel
,
M. F.
,
Quinn
,
C. O.
,
Kitchingman
,
G. R.
,
Look
,
A. T.
&
Sherr
,
C. J.
(
1985a
).
Transmembrane orientation of glycoproteins encoded by the v-fms oncogene
.
Cell
40
,
971
981
.
Rettenmier
,
C. W.
,
Sacca
,
R.
,
Furman
,
W. L.
,
Roussel
,
M. F.
,
Holt
,
J. T.
,
Nienhuis
,
A. W.
,
Stanley
,
E. R.
&
Sherr
,
C. J.
(
1986
).
Expression of the human c-fms proto-oncogene product (colony-stimulating factor-1 receptor) on peripheral blood mononuclear cells and choriocarcinoma cell lines
.
J. clin. Invest
.
77
1740
1746
.
Roussel
,
M. F.
,
Dull
,
T. J.
,
Rettenmier
,
C. W.
,
Ralph
,
P.
,
Ullrich
,
A.
&
Sherr
,
C. J.
(
1987
).
Transforming potential of the c-fms proto-oncogene (CSF-l receptor)
.
Nature, Land
.
325
,
549
552
.
Roussel
,
M. F.
,
Rettenmier
,
C. W.
,
Look
,
A. T.
&
Sherr
,
C. J.
(
1984
).
Cell surface expression of v-fms-coded glycoproteins is required for transformation
.
Molec. Cell Biol
.
4
,
1999
2009
.
Roussel
,
M. F.
,
Rettenmier
,
C. W.
&
Sherr
,
C. J.
(
1988
).
Introduction of a human colony stimulating factor-1 gene into a mouse macrophage cell line induces CSF-l independence but not tumorigenicity
.
Blood
71
,
1218
1225
.
Rowley
,
J. D.
,
Golomb
,
H. M.
&
Vardiman
,
J. W.
(
1981
).
Nonrandom chromosome abnormalities in acute leukemia and dysmyelopoietic syndromes in patients with previously treated malignant disease
.
Blood
58
,
759
767
.
Sacca
,
R.
,
Stanley
,
E. R.
,
Sherr
,
C. J.
&
Rettenmier
,
C. W.
(
1986
).
Specific binding of the mononuclear phagocyte colony-stimulating factor CSF-l to the product of the v-fms oncogene
.
Proc. natn. Acad. Sci. U.S.A
.
83
,
3331
3335
.
Sarma
,
P. S.
,
Sharar
,
A. L.
&
McDonough
,
S.
(
1972
).
The SM strain of feline sarcoma virus. Biologic and antigenic characterization of virus
.
Proc. Soc. exp. Biol. Med
.
140
,
1365
1368
.
Sherr
,
C. J.
,
Rettenmier
,
C. W.
,
Sacca
,
R.
,
Roussel
,
M. F.
,
Look
,
A. T.
&
Stanley
,
E. R.
(
1985
).
The c-fms proto-oncogene product is related to the receptor for the mononuclear phagocyte growth factor, CSF-L Cell
41
,
665
676
.
Sokal
,
G.
,
Michaux
,
J. L.
,
van den Berghe
,
H.
,
Cordier
,
A.
,
Rodhain
,
J.
,
Ferrant
,
A.
,
Moriau
,
M.
,
Debruyere
,
M.
&
Sonnet
,
J.
(
1975
).
A new hematologic syndrome with a distinct karyotype: The 5q- chromosome
.
Blood
46
,
519
533
.
Stanley
,
E. R.
,
Bartocci
,
A.
,
Patinkin
,
D.
,
Rosendaal
,
M.
&
Bradley
,
T. R.
(
1986
).
Regulation of very primitive, multipotent, hematopoietic cells by hemopoietin-1
.
Cell
45
,
667
674
.
Stanley
,
E. R.
,
Guilbert
,
L. J.
,
Tushinski
,
R. J.
&
Bartelmez
,
S. H.
(
1983
).
CSF-l - A mononuclear phagocyte lineage-specific hemopoietic growth factor
.
J. cell. Biochem
.
21
,
151
159
.
Stanley
,
E. R.
&
Heard
,
P. M.
(
1977
).
Factors regulating macrophage production and growth. Purification and some properties of the colony stimulating factor from medium conditioned by mouse L cells
.
J. biol. Chem
.
252
,
4305
4312
.
Tamura
,
T.
,
Simon
,
E.
,
Niemann
,
H.
,
Snoek
,
G. T.
&
Bauer
,
H.
(
1986
).
gpl40v-fms molecules expressed at the surface of cells transformed by the McDonough strain of feline sarcoma virus are phosphorylated in tyrosine and serine
.
Molec. Cell. Biol
.
6
,
4745
4748
.
Tushinski
,
R. J.
,
Oliver
,
I. T.
,
Guilbert
,
L. J.
,
Tynan
,
P. W.
,
Warner
,
J. R.
&
Stanley
,
E. R.
(
1982
).
Survival of mononuclear phagocytes depends on a lineage-specific growth factor that the differentiated cells selectively destroy
.
Cell
28
,
71
81
.
Tushinski
,
R. J.
&
Stanley
,
E. R.
(
1983
).
The regulation of macrophage protein turnover by a colony stimulating factor (CSF-1)
.
J. cell Physiol
.
116
,
67
75
.
Tushinski
,
R. J.
&
Stanley
,
E. R.
(
1985
).
The regulation of mononuclear phagocyte entry into S phase by the colony stimulating factor CSF-L
J. cell Physiol
.
122
,
221
228
.
Van den Berghe
,
H.
,
Cassiman
,
J.-J.
,
David
,
G.
,
Fryns
,
J.-P.
,
Michaux
,
J.-L.
&
Sokal
,
G.
(
1974
).
Distinct haematological disorder with deletion of long arm of No. 5 chromosome
.
Nature, Lond
.
251
,
437
438
.
Warren
,
M. K.
&
Ralph
,
P.
(
1986
).
Macrophage growth factor CSF-l stimulates human monocyte production of interferon, tumor necrosis factor, and colony stimulating activity
.
J. Immun
.
137
,
2281
2285
.
Waterfield
,
M. D.
,
Scrace
,
G. T.
,
Whittle
,
N.
,
Stroobant
,
P.
,
Johnsson
,
A.
,
Wasteson
,
A.
,
Westermark
,
B.
,
Heldin
,
C.-H.
,
Huang
,
J. S.
&
Deuel
,
T. F.
(
1983
).
Platelet-derived growth factor is structurally related to the putative transforming protein p28”‘of simian sarcoma virus
.
Nature, Lond
.
304
,
35
39
.
Wheeler
,
E. F.
,
Askew
,
D.
,
May
,
S.
,
Ilhe
,
J. N.
&
Sherr
,
C. J.
(
1987
).
The v-fms oncogene induces factor-independent growth and transformation of the interleukin-3-dependent myeloid cell line FDC-P1
.
Molec. Cell. Biol
.
7
,
1673
1680
.
Wheeler
,
E. F.
,
Rettenmier
,
C. W.
,
Look
,
A. T.
&
Sherr
,
C. J.
(
1986a
).
The v-fms oncogene induces factor independence and tumorigenicity in CSF-1 dependent macrophage cell line
.
Nature, Lond
.
324
,
377
380
.
Wheeler
,
E. F.
,
Roussel
,
M. F.
,
Hampe
,
A.
,
Walker
,
M. H.
,
Fried
,
V. A.
,
Look
,
A. T.
,
Rettenmier
,
C. W.
&
Sherr
,
C. J.
(
1986b
).
The amino-terminal domain of the v-fms oncogene product includes a functional signal peptide that directs synthesis of a transforming glycoprotein in the absence of feline leukemia virus gag sequences
.
J. Virol
.
59
,
224
233
.
Whetton
,
A. D.
,
Monk
,
P. N.
,
Consalvey
,
S. D.
&
Downes
,
C. P.
(
1986
).
The haemopoietic growth factors interleukin 3 and colony stimulating factor-1 stimulate proliferation but do not induce inositol lipid breakdown in murine bone-marrow-derived macrophages
.
EMBO J
.
5
,
3281
3286
.
Wisniewski
,
L. P.
&
Hirschhorn
,
K.
(
1983
).
Acquired partial deletions of the long arm of chromosome 5 in hematologic disorders
.
Am. J. Hematol
.
15
,
295
310
.
Wong
,
G. G.
,
Temple
,
P. A.
,
Leary
,
A. C.
,
Witek-Giannotti
,
J. S.
,
Yang
,
Y.-C.
,
Ciarletta
,
A. B.
,
Chung
,
M.
,
Murtha
,
P.
,
Kriz
,
R.
,
Kaufman
,
R. J.
,
Ferenz
,
C. R.
,
Sibley
,
B. S.
,
Turner
,
K. J.
,
Hewick
,
R. M.
,
Clark
,
S. C.
,
Yanai
,
N.
,
Yokota
,
H.
,
Yamada
,
M.
,
Saito
,
M.
,
Motoyoshi
,
K.
&
Takaku
,
F.
(
1987
).
Human CSF-1: Molecular cloning and expression of 4 kb cDNA encoding the human urinary protein
.
Science
235
,
1504
1508
.
Woolford
,
J.
,
Rothwell
,
V.
&
Rohrschneider
,
L.
(
1985
).
Characterization of the human c-fms gene product and its expression in cells of the monocyte-macrophage lineage
.
Molec. Cell Biol
.
5
,
3458
3466
.
Yarden
,
Y.
,
Escobedo
,
J. A.
,
Kuang
,
W. J.
,
Yang-Feng
,
T. L.
,
Daniel
,
T. O.
,
Tremble
,
P. M.
,
Chen
,
E. Y.
,
Ando
,
M. E.
,
Harkins
,
R. N.
,
Francke
,
U.
,
Fried
,
V. A.
,
Ullrich
,
A.
&
Williams
,
L. T.
(
1986
).
Structure of the receptor for platelet-derived growth factor helps define a family of closely related growth factor receptors
.
Nature, Lond
.
323
,
226
232
.
Yeung
,
Y. G.
,
Jubinsky
,
P. T.
,
Sengupta
,
A.
,
Yeung
,
D. C. Y.
&
Stanley
,
E. R.
(
1987
).
Purification of the colony-stimulating factor 1 receptor and demonstration of its tyrosine kinase activity
.
Proc. natn. Acad. Sci. U.S.A
.
84
,
1268
1271
.