Correction: Myeloproliferative leukemia protein activation directly induces fibrocyte differentiation to cause myelofibrosis.

Owing to the insufficient specificity of the anti-myeloproliferative leukemia protein (MPL) antibody in the original version of this Article, Figure 6 and parts of Figures 2a, 4e, and 5a do not represent the correct information. The corrected version of Figure 6 is in this correction and those of Figures 2a, 4e, and 5a are shown in the supplemental information.

Myeloproliferative leukemia protein activation directly induces fibrocyte differentiation to cause myelofibrosis.

Myelofibrosis (MF) may be caused by various pathogenic mechanisms such as elevation in circulating cytokine levels, cellular interactions and genetic mutations. However, the underlying mechanism of MF still remains unknown. Recent studies have revealed that fibrocytes, the spindle-shaped fibroblast-like hematopoietic cells, and the thrombopoietin (TPO)/myeloproliferative leukemia protein (MPL; TPO receptor) signaling pathway play a certain role in the development of MF. In the present study, we aimed to investigate the relationship between fibrocytes and MPL activation. We showed that TPO or a TPO receptor agonist directly induces fibrocyte differentiation using murine fibrocyte cell lines and a murine MF model. Conversely, elimination of macrophages expressing MPL by clodronate liposomes reversed the MF phenotype of the murine model, suggesting that fibrocyte differentiation induced by MPL activation contributes to the progression of MF. Furthermore, we revealed that SLAMF7high MPLhigh monocytes in human peripheral blood mononuclear cells were possible fibrocyte precursors and that these cells increased in number in MF patients not treated with ruxolitinib. Our findings confirmed a link between fibrocytes and the TPO/MPL signaling pathway, which could result in a greater understanding of the pathogenesis of MF and lead to the development of novel therapeutic interventions.

Macropinocytosis and TAK1 mediate anti-inflammatory to pro-inflammatory macrophage differentiation by HIV-1 Nef.

Macrophages (MΦ) are functionally classified into two types, anti-inflammatory M2 and pro-inflammatory M1. Importantly, we recently revealed that soluble HIV-1 proteins, particularly the pathogenetic protein Nef, preferentially activate M2-MΦ and drive them towards an M1-like MΦ, which might explain the sustained immune activation seen in HIV-1-infected patients. Here, we show that the preferential effect of Nef on M2-MΦ is mediated by TAK1 (TGF-ß-activated kinase 1) and macropinocytosis. As with MAP kinases and NF-κB pathway, Nef markedly activated TAK1 in M-CSF-derived M2-MΦ but not in GM-CSF-derived M1-MΦ. Two Nef mutants, which were unable to activate MAP kinases and NF-κB pathway, failed to activate TAK1. Indeed, the TAK1 inhibitor 5Z-7-oxozeaenol as well as the ectopic expression of a dominant-negative mutant of TAK1 or TRAF2, an upstream molecule of TAK1, inhibited Nef-induced signaling activation and M1-like phenotypic differentiation of M2-MΦ. Meanwhile, the preferential effect of Nef on M2-MΦ correlated with the fact the Nef entered M2-MΦ more efficiently than M1-MΦ. Importantly, the macropinosome formation inhibitor EIPA completely blocked the internalization of Nef into M2-MΦ. Because the macropinocytosis activity of M2-MΦ was higher than that of M1-MΦ, our findings indicate that Nef enters M2-MΦ efficiently by exploiting their higher macropinocytosis activity and drives them towards M1-like MΦ by activating TAK1.

IL-34 and M-CSF share the receptor Fms but are not identical in biological activity and signal activation.

Macrophage colony-stimulating factor (M-CSF) regulates the production, survival and function of macrophages through Fms, the receptor tyrosine kinase. Recently, interleukin-34 (IL-34), which shares no sequence homology with M-CSF, was identified as an alternative Fms ligand. Here, we provide the first evidence that these ligands indeed resemble but are not necessarily identical in biological activity and signal activation. In culture systems tested, IL-34 and M-CSF showed an equivalent ability to support cell growth or survival. However, they were different in the ability to induce the production of chemokines such as MCP-1 and eotaxin-2 in primary macrophages, the morphological change in TF-1-fms cells and the migration of J774A.1 cells. Importantly, IL-34 induced a stronger but transient tyrosine phosphorylation of Fms and downstream molecules, and rapidly downregulated Fms. Even in the comparison of active domains, these ligands showed no sequence homology including the position of cysteines. Interestingly, an anti-Fms monoclonal antibody (Mab) blocked both IL-34-Fms and M-CSF-Fms binding, but another MAb blocked only M-CSF-Fms binding. These results suggested that IL-34 and M-CSF differed in their structure and Fms domains that they bound, which caused different bioactivities and signal activation kinetics/strength. Our findings indicate that macrophage phenotype and function are differentially regulated even at the level of the single receptor, Fms.

Detection of murine adult bone marrow stroma-initiating cells in Lin(-)c-fms(+)c-kit(low)VCAM-1(+) cells.

We attempted to characterize the phenotype of cells which initiate fibroblastic stromal cell formation (stroma-initiating cells: SICs), precursor cells for fibroblastic stromal cells, based on the expression of cell surface antigens. First, we stained adult murine bone marrow cells with several monoclonal antibodies and separated them by magnetic cell sorting. SICs were abundant in the c-kit(+), Sca-1(+), CD34(+), VCAM-1(+), c-fms(+), and Mac-1(-) populations. SICs were recovered in the lineage-negative (Lin(-)) cells but not the Lin(+) cells. When macrophage colony-stimulating factor (M-CSF) was absent from the culture medium, no stromal colony appeared among the populations enriched in SICs. Based on these findings, the cells negative for lineage markers and positive for c-fms (M-CSF receptor) were further divided on the basis of the expression of c-kit, VCAM-1, Sca-1 or CD34 with a fluorescence-activated cell sorter. SICs were found to be enriched in the Lin(-)c-fms(+)c-kit(low) cells and Lin(-)c-fms(+)VCAM-1(+) cells but not in Lin(-)c-fms(+)Sca-1(+) cells and Lin(-)c-fms(+)CD34(low) cells. As a result, the SICs were found to be present at highest frequency in Lin(-)c-fms(+)c-kit(low)VCAM-1(+) cells: a mean of 64% of the SICs in the Lin(-) cells were recovered in the population. In morphology and several characteristics, the stromal cells derived from Lin(-)c-fms(+)c-kit(low)VCAM-1(+) cells resembled fibroblastic cells. The number of Lin(-)c-fms(+)c-kit(low)VCAM-1(+) cells in bone marrow of mice injected with M-CSF was higher than that in control mice. In this study, we identified SICs as Lin(-)c-fms(+)c-kit(low)VCAM-1(+) cells and demonstrated that M-CSF had the ability to increase the cell population in vivo.

Expression of Jagged1 gene in macrophages and its regulation by hematopoietic growth factors.

OBJECTIVE: Serrate/Jagged and Delta are cell surface ligands for Notch receptors that may influence hematopoietic cell fate decisions and are known to be expressed in bone marrow stromal cells. In a series of screenings of cDNAs constructed by a cDNA library subtraction technique, we identified Jagged1, one of the Notch ligands, as a gene up-regulated by macrophage colony-stimulating factor (M-CSF) in bone marrow macrophages. Therefore, we compared stromal cells and macrophages for expression of Notch ligands including Jagged1 and analyzed the regulation of their expression by cytokines. MATERIALS AND METHODS: Murine bone marrow macrophages were prepared by culturing femoral bone marrow cells with M-CSF. Primary bone marrow fibroblastic stromal cells were prepared by a culture system that we recently developed. The expression of Notch ligands was analyzed by either Northern blot analysis or reverse transcriptase polymerase chain reaction. RESULTS: The bone marrow macrophages expressed Jagged1 but not Jagged2 and Delta1 at a level that was detectable by Northern blot analysis. Expression of the Jagged1 gene was markedly up-regulated by growth factors for the cells, i.e., M-CSF, granulocyte-macrophage colony-stimulating factor, and interleukin-3. Expression of Jagged2 and Delta1 seldom was affected by the stimuli. The primary bone marrow fibroblastic stromal cells, and murine stromal cell lines, such as PA6 and ST2, also expressed Jagged1 transcript, at levels comparable to the steady-state level in macrophages. However, expression of the Jagged1 gene was little affected when these cells were stimulated with fibroblastic growth factor and platelet-derived growth factor. CONCLUSIONS: We demonstrated that bone marrow macrophages as well as stromal cells constitutively produced Jagged1 and that the expression was markedly up-regulated by hematopoietic growth factors, M-CSF, granulocyte-macrophage colony-stimulating factor, and interleukin-3. The results highlight the involvement of macrophages and these growth factors in hematopoietic cell fate decisions via the production of Jagged1.

Erythropoietin synthesis by tumour tissues in a patient with uterine myoma and erythrocytosis.

We report a patient with uterine myoma (leiomyoma) and erythrocytosis in whom erythropoietin (Epo) production in the leiomyoma tissue was identified by reverse transcription polymerase chain reaction (RT-PCR) and enzyme-linked immunosorbent assay (ELISA). A 48-year-old Japanese woman with uterine myoma showed marked erythrocytosis (haemoglobin: 20.2 g/dl, haematocrit: 61.1%, red blood cells: 6.51 x 10(12)/1). After hysterectomy, erythrocytosis rapidly disappeared. In the leiomyoma tissue collected from the patient, Epo mRNA expression was confirmed using RT-PCR. Furthermore, ELISA showed that the Epo protein level was significantly increased compared with those in control tissues. It is suggested that the pathogenesis of erythrocytosis in patients with uterine myoma involves ectopic Epo production by leiomyoma tissues.

Antitumor immunity induced by irradiated tumor cells producing macrophage colony-stimulating factor.

We previously reported that administration into mice of mouse lymphoid leukemia L1210 cells engineered to secrete macrophage colony-stimulating factor (M-CSF) could lead to tumor rejection. Here, we demonstrate that inoculation with irradiated M-CSF-producing cells protects mice against a subsequent challenge with unmodified parental tumor cells. We used 2 experimental protocols: the inoculation with irradiated M-CSF-producing L1210 cells (EM5) before the challenge with parental cells and after the challenge with parental cells. Both protocols effectively improved the survival rate of mice compared with protocols in which irradiated non-M-CSF-producing L1210 cells (EM-mock) were inoculated. Inoculation with 1 x 10(2) irradiated EM5 cells was sufficient to prolong the survival time of mice subsequently challenged with 1 x 10(4) parental cells. In vivo depletion experiments with administration of antibodies suggested the involvement of CD4+ T cells, CD8+ T cells, and natural killer (NK) cells in the antitumor effect. Consistent with these findings, the cytotoxic T lymphocyte activity of splenocytes from EM5-inoculated mice was higher than that from EM-mock-inoculated mice, and L1210 tumors were heavily infiltrated by CD4+ T cells and NK cells as well as macrophages in EM5-inoculated mice.

p56(dok-2) as a cytokine-inducible inhibitor of cell proliferation and signal transduction.

p56(dok-2) acts as a multiple docking protein downstream of receptor or non-receptor tyrosine kinases. However, the role of p56(dok-2) in biological functions of cells is not clear. We found that transcription of the p56(dok-2) gene in macrophages was increased markedly in response to cytokines such as macrophage colony-stimulating factor (M-CSF), granulocyte/macrophage-CSF and interleukin-3 (IL-3). Forced expression of p56(dok-2) inhibited M-CSF-, granulocyte-CSF-, IL-3- and stem cell factor-induced proliferation of myeloid leukemia cells, M-NFS-60. The p56(dok-2)-overexpressing cells showed an impaired induction of c-myc but not of c-jun, junB or c-fos when stimulated with M-CSF. Consistent with these results, the peritoneal cavity of the hairless (hr/hr) strain of mutant mice, whose cells expressed less p56(dok-2) than wild-type mice, contained more macrophages than that of +/hr mice. Moreover, the inhibition of endogenous p56(dok-2) expression in macrophage-like tumor cells, J774A.1, by stable expression of antisense p56(dok-2) mRNA accelerated cell proliferation. The study identifies a novel role for p56(dok-2) as a molecule that negatively regulates signal transduction and cell proliferation mediated by cytokines in a feedback loop.

Effect of cytokines on the proliferation/differentiation of stroma-initiating cells.

A culture system that identifies the precursor of murine bone marrow fibroblastic stromal cells (stroma-initiating cells, SIC) has been developed. In this system, mature fibroblasts are depleted by adherence to plastic dishes and the nonadherent cells are seeded at a low density, which results in the formation of colonies composed of fibroblastic cells. Macrophage colony-stimulating factor (M-CSF) has been shown to accelerate the colony formation in the system. In this study, we examined the stroma-inducing activity of a number of cytokines. Neither granulocyte-CSF, stem cell factor, interleukin (IL)-1, IL-6, transforming growth factor, epidermal growth factor, insulin-like growth factor, platelet-derived growth factor, nor fibroblast growth factor showed the activity. Similarly, tumor necrosis factor (TNF) did not show any stroma-inducing activity, but the factor inhibited the stromal colony formation induced by M-CSF. In this study, we found that granulocyte/macrophage-CSF (GM-CSF) and IL-3, as well as M-CSF had the stroma-inducing activity. Neither an additive nor synergistic effect was observed when the three factors were assayed in various combinations. The stroma-inducing activity of M-CSF, GM-CSF and IL-3 was observed even if lineage-negative bone marrow cells were used as target cells, suggesting that mature hematopoietic cells such as macrophages and granulocytes were not involved in the induction of stromal colony formation by these factors. Our results raise the possibility that GM-CSF and IL-3 as well as M-CSF stimulate the proliferation or differentiation of the precursor of bone marrow fibroblastic stromal cells.

Association of Cbl with Fms and p85 in response to macrophage colony-stimulating factor.

Tyrosine phosphorylation of Cbl and its association with signal-transducing molecules in response to macrophage colony-stimulating factor (M-CSF) were analyzed by using cell lines which express the wild-type and a mutant M-CSF receptor, Fms. We found that in a clone, F723 TF-1 cells expressing mutant Fms in which tyrosine 723 had been substituted with phenylalanine, the M-CSF stimulation-dependent association between Cbl and Fms was markedly impaired. However, phosphorylation of Cbl and its association with the p85 subunit of phosphatidylinositol 3-kinase were induced in these mutant cells as seen in the wild-type fms transfectant. These results suggest that phosphorylation of tyrosine 723 is particularly important for the recruitment of Cbl to the M-CSF receptor, but is not required for the phosphorylation and binding of Cbl to signal-transducing molecules such as p85.

In vivo stimulatory effect of macrophage colony-stimulating factor on the number of stroma-initiating cells.

The in vivo effect of human macrophage colony-stimulating factor (M-CSF) on the number of cells that formed stromal colonies in an in vitro culture system (stroma-initiating cells; SICs) was investigated. We found that the number of SICs in the femurs of C57BL/6 mice was significantly increased by the treatment with M-CSF. We also found that the SICs were resistant to at least three different chemotherapeutic reagents, 5-fluorouracil (5-FU), cytarabine, and cyclophosphamide, because the femoral cells of mice treated with these reagents contained higher numbers of SICs than those of untreated mice. M-CSF treatment also increased the number of SICs of the reagent-pretreated mice. The SICs detected in our culture system were present only in Mac-1(-)CD45(-) cells, and the M-CSF treatment of 5-FU-pretreated mice actually increased the number of Mac-1(-)CD45(-) SICs. The Mac-1(-)CD45(-) SICs collected from mice that were pretreated with 5-FU and then treated with M-CSF formed stromal colonies under in vitro culture conditions that did not contain M-CSF but did contain a high concentration of fetal calf serum. This result suggested that SICs collected following the treatment procedure did not necessarily require the presence of M-CSF for their in vitro proliferation. Our study indicated that M-CSF has the ability to increase the number of progenitor or precursor cells for bone marrow stromal cells in vivo system.

Association of CrkL with STAT5 in hematopoietic cells stimulated by granulocyte-macrophage colony-stimulating factor or erythropoietin.

CrkL is an adapter protein comprising Src homology (SH) 2 and SH3 domains. We investigated the molecule(s) associated with CrkL in factor-dependent cell lines. In the granulocyte-macrophage colony-stimulating factor (GM-CSF)-dependent cell lines TF-1 and UT-7, an approximately 95-kDa tyrosine-phosphorylated protein was precipitated along with CrkL after GM-CSF stimulation. The same protein was also observed when we used the erythropoietin (EPO)-dependent cell line UT-7/EPO, in an EPO stimulation-dependent manner. We identified it as STAT5 (signal transducer and activator of transcription 5, 96 kDa) by STAT5-specific antibodies. The direct binding of the SH2 domain of CrkL to STAT5 was demonstrated in far Western blotting and pull-down experiments using the glutathione S-transferase (GST) fusion construct CrkL-SH2. The addition of the oligopeptide containing phosphotyrosine 694 in STAT5A impaired the association between GST-CrkL-SH2 and STAT5. Furthermore, in a gel shift assay using prolactin-inducible element (PIE) as the probe, the DNA binding activity of STAT5 was inhibited by the interaction with GST-CrkL-SH2 in vitro. Finally, we found that STAT5 associated with CrkL did not bind to PIE sequence. These results suggest that CrkL participates in the Janus kinase (JAK)-STAT pathway by direct association with STAT5.

Identification of alternatively spliced transcripts encoding murine macrophage colony-stimulating factor.

We have isolated a novel cDNA encoding macrophage colony-stimulating factor (M-CSF) from a murine stromal cell line, ST2. The cDNA included an entire coding sequence of the M-CSF gene but contained an additional sequence of 140 base pairs (bp). Northern blot analysis demonstrated that other murine cell lines such as a fibroblastic cell line (L) and a stromal cell line (PA6) also expressed the transcripts corresponding to the clone. The nucleotide sequence analyses of the cDNA and the cloned M-CSF genome revealed that the 140-bp insertion sequence was part of intron 1 which separated exon 1 and exon 2: the former contained part of the amino acid residues of the signal sequence and the latter the rest of the signal sequence and the first 22 amino acid residues of the mature protein. The insertion of the 140-bp intron sequence not only changed the amino acid sequence of the signal peptide but also generated an in-frame termination codon. However, instead of the dysfunction of the original initiation codon, the 140-bp insertion sequence contained a putative ATG initiation codon that preserved the original open reading frame. Finally, we found that the cDNA directed the expression of a secreted and biologically active M-CSF protein when it was introduced into COS7 cells and M-CSF activity in the culture supernatants was measured using an M-CSF-dependent cell line. These results indicate the presence of an alternatively spliced M-CSF transcript which utilizes an alternate initiation codon in order to specify active M-CSF protein.

Properties of primary murine stroma induced by macrophage colony-stimulating factor.

The ability of purified human macrophage colony-stimulating factor (M-CSF) to accelerate the formation of stromal cells from murine bone marrow cells was investigated. The liquid culture of the marrow cells with M-CSF resulted in the formation of monolayers of macrophages on day 7. When the M-CSF was removed on that day and the residual adherent cells were cultured in the absence of M-CSF for an additional 7 days, many colonies appeared with cells that were morphologically distinguishable from M-CSF-derived macrophages. The appearance of the colonies was dependent on the concentration of M-CSF used at the beginning of the culture. Each colony was isolated as a single clone and analyzed. All clones were negative for esterase staining. These cells did not express M-CSF receptor mRNA and did not show a mitogenic response to M-CSF. On the contrary, these cells could be stimulated to proliferate by fibroblast growth factor and platelet-derived growth factor. The polymerase chain reaction analysis of these cells demonstrated constitutive expression of mRNA for M-CSF, stem cell factor, and interleukin (IL)-1, but not IL-3. Some clones expressed mRNA for granulocyte/M-CSF and IL-6. We also examined the ability of the cells to maintain murine bone marrow high proliferative potential colony-forming cells (HPP-CFC) in a coculture system. Most of the clones showed a significant increase in total HPP-CFC numbers after 2 weeks of coculture, although the extent of stimulation differed among clones. These results suggested that the colonies established by M-CSF were composed of functional stromal cells that were phenotypically different from macrophages.

Biologic activity of proteoglycan macrophage colony-stimulating factor.

We compared the biologic activities of 85-kDa macrophage-CSF (85-kDa M-CSF), which is fully active, and a proteoglycan M-CSF (PG-M-CSF). Both originate from the same precursor, but the latter retains the carboxyl-terminal portion, which must be proteolytically removed from the precursor to generate 85-kDa M-CSF and which is uniquely modified by a chondroitin sulfate glycosaminoglycan chain. PG-M-CSF supported the formation of murine macrophage colonies such as 85-kDa M-CSF. Furthermore, PG-M-CSF stimulated the proliferation of murine bone marrow macrophages, an M-CSF-dependent murine cell line, and an M-CSF-responsive human cell line established by transfer of the human M-CSF receptor gene. PG-M-CSF and 85-kDa M-CSF had equivalent specific biologic activities on a molar basis in all bioassays. The activity of PG-M-CSF was not affected by enzymatically removing the glycosaminoglycan chain when assayed by the formation of macrophage colonies and proliferation of the bone marrow macrophages. We analyzed the phosphorylation on tyrosine residue(s) of the M-CSF receptor in response to these M-CSFs that trigger mitogenic responses. PG-M-CSF rapidly (within 10 min) induced receptor phosphorylation in human cells with the same potency as 85-kDa M-CSF. These results indicate that PG-M-CSF is not a latent form or precursor of 85-kDa M-CSF but a fully biologically active cytokine.

Identification of binding domains for basic fibroblast growth factor in proteoglycan macrophage colony-stimulating factor.

We recently demonstrated that proteoglycan macrophage colony-stimulating factor (PG-M-CSF) binds basic fibroblast growth factor (bFGF) and neutralizes the biological activity of bFGF. In this study, we identified the binding sites of PG-M-CSF for bFGF. We examined the binding of bFGF to overlapping 12-mer peptides with the sequence of the putative binding region. High affinity binding was detected at two peaks; one consisted of the three adjacent peptides, 212-223, 213-224 and 214-225 and the other, of the three adjacent peptides, 246-257, 247-258 and 248-259. The synthetic peptide (212VDPGSAKQRPPRST225) did not inhibit bFGF binding to another peptide (246PQPRPSVGAFNPGM259), and vice versa. However, both peptides inhibited the bFGF-induced but not platelet-derived growth factor-induced stimulation of DNA synthesis in murine Balb/c 3T3 cells.

Immunohistochemical identification of proteoglycan form of macrophage colony-stimulating factor on bone surface.

Several studies using op/op mice have shown that macrophage colony-stimulating factor (M-CSF) was necessary for osteoclast formation in vivo. Previously we reported that osteoblastic cells produced two molecular forms of M-CSF; one is an 85-kDa M-CSF, and the other is a proteoglycan form of M-CSF (PG-M-CSF) which has a binding affinity for bone-derived collagens and is extractable from human bone. In this study, we performed immunostaining of human bone using a newly established anti-PG-M-CSF antibody, and showed positive staining PG-M-CSF, probably produced by bone lining cells, on the bone surface. This observation suggests that the bone surface is suitable for osteoclast formation because of the presence of PG-M-CSF.

Augmentation of cancer chemotherapy by preinjection of human macrophage colony-stimulating factor in L1210 leukemic cell-inoculated mice.

Human macrophage colony-stimulating factor (hM-CSF) is a potent stimulator of the effector functions of monocytes/macrophages. We investigated the antitumor effects of this factor in CDF1 male mice inoculated with L1210 cells, a mouse B-cell leukemia line. Mice preinoculated with various numbers of L1210 cells on day 0 were given intravenous injections of vehicle (human serum albumin; HSA) (100 micrograms/kg/day) or hM-CSF (20 micrograms/kg/day) for 3 days from day 1. In mice preinoculated with 10(2) L1210 cells but not with 10(3) or more L1210 cells, a marked increment in survival rate was observed with hM-CSF treatment. We next examined the effect of hM-CSF treatment combined with chemotherapy on the survival of mice that had been preinoculated with 10(5) L1210 cells. In our system, the administration of 4.9 mg/kg adriamycin (ADM) alone slightly prolonged survival of the tumor-bearing mice, but all of the mice died within 20 days. When hM-CSF was injected for 3 days before this ADM treatment, the invasion and proliferation of tumor cells in the liver and spleen were markedly inhibited and 50% of the mice were still alive at day 50. We detected inhibitory activity toward L1210 growth in serum of mice administered with hM-CSF, and the degree of the inhibitory activity was correlated with the level of nitrite (NO2-) in the serum. When L1210 cells were co-cultured with peritoneal macrophages from mice intraperitoneally injected with hM-CSF, the uptake of [3H]thymidine in L1210 cells was inhibited. The inhibition was abolished by the addition of NG-monomethyl-L-arginine, an inhibitor of NO2- synthesis, suggesting that the reactive nitrogen oxide intermediate is involved in hM-CSF-induced inhibition of L1210 growth.

Proteoglycan form of macrophage colony-stimulating factor binds low density lipoprotein.

We recently isolated a proteoglycan form of macrophage colony-stimulating factor (PG-M-CSF) that carries a chondroitin sulfate glycosaminoglycan chain. Here, we examined the interaction of PG-M-CSF with low density lipoprotein (LDL). When LDL preincubated with PG-M-CSF was fractionated by molecular size sieving chromatography, it was eluted earlier than untreated LDL. When LDL was preincubated with chondroitin sulfate-free 85-kD M-CSF instead of PG-M-CSF, the elution profile of LDL remained unchanged, indicating specific interaction between PG-M-CSF and LDL. The level of PG-M-CSF binding in the wells of a plastic microtitration plate precoated with LDL was significant, this binding being completely abolished by pretreatment of PG-M-CSF with chondroitinase AC, which degrades chondroitin sulfate. The addition of exogenous chondroitin sulfate or apolipoprotein B inhibited the binding of PG-M-CSF to LDL in a dose-dependent manner, indicating that the interaction between PG-M-CSF and LDL was mediated by the binding of the chondroitin sulfate chain of PG-M-CSF to LDL apolipoprotein B. PG-M-CSF was also demonstrated in the arterial wall, and there were increased amounts of PG-M-CSF in atherosclerotic lesions. The in vitro interaction between PG-M-CSF and LDL thus appears to have physiological significance.