Background
Giant cell tumors (GCTs) of bone are rare primary skeletal neoplasms that occur in young adults [
1]. The histological phenotype of GCTs is characterized by a large number of osteoclast-like giant multi-nuclear cells (MNCs), which is why this tumor is called an osteoclastoma or giant cell tumor. Apart from the MNCs, GCTs contain two types of mononuclear cells. One cell type has a round morphology and resembles monocytes (monocyte-like cells), while the other is a spindle-shaped, fibroblast-like stromal cell (stromal cells) [
2]. Primary cell cultures of GCTs revealed that the stromal cells are likely the proliferating cell type in GCTs because the monocyte-like cells and MNCs are lost after several culture passages [
3]. Based on these observations, the current hypothesis for the cellular origin of GCTs is that the stromal cells in GCTs are tumor cells, the monocyte-like cells are reactive macrophages and/or osteoclast precursor cells (OPCs), and the MNCs are reactive osteoclasts (OCs) [
4].
Recently, it was reported that these stromal cells secrete several cytokines and differentiation factors, including TGF-β [
5], MCP-1[
6], RANKL [
7] and M-CSF [
8]. These soluble factors could function as monocyte chemoattractants and stimulate osteoclast differentiation, suggesting that the stromal cells stimulate blood monocytes to migrate into the tumor tissue and enhance
in situ osteoclastogenesis, leading to extended osteolysis by OCs.
We previously reported that the vascular endothelial growth factor (VEGF)-Flt-1 (type-1 VEGF receptor)-focal adhesion kinase (FAK) pathway may be involved in the chemotaxis and cell proliferation of OPCs and contribute to arthritic joint destruction [
9]. VEGF overexpression has also been associated with the biological aggressiveness of GCTs [
10]. Therefore, we hypothesized that the stromal cells in GCTs produce VEGF that recruits OPCs to the neoplastic lesions. In this study, we examined clinical GCT samples in order to determine the possible role of the VEGF-Flt-1-FAK pathway in the pathogenesis of bone destruction in GCTs.
Methods
Patients and tissue specimens
The Institutional Review Board of Kyushu University School of Medicine, Fukuoka, Japan approved the protocol to obtain and examine surgical GCT specimens. Twenty-one GCT patients were surgically treated in the Department of Orthopaedic Surgery, Kyushu University. All tumor specimens were formalin-fixed and paraffin-embedded, and 5-mm sections were cut from one representative block for molecular analyses.
Agents
Sprague-Dawley rats were purchased from KBT Oriental (Saga, Japan). Recombinant human VEGF was obtained from Genzyme/Techne (Minneapolis, MN). Anti-VEGF, -Flt-1 and -Flk-1 Abs were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-FAK Ab was obtained from Upstate Biotechnology (Lake Placid, NY). Antibodies specific for the phosphotyrosine residue at position 397 in FAK (pY-FAK Ab) and anti-tyrosine phosphorylated Flt-1 (pY-Flt-1) were purchased from Invitrogen (Carlsbad, CA) and Oncogene (San Diego, CA), respectively. The VEGF receptor tyrosine kinase (RTK) inhibitor (ZD4190) was purchased from Calbiochem (San Diego, CA).
Cell culture
Rat osteoclast precursor cells (rOPCs) were harvested using by the modified method as previously described [
11](Takeshita S et al. 2000). Briefly, the femurs and tibias of 1-day-old Sprague-Dawley rats were aseptically resected. The bone ends were cut and the marrow cavity was flushed with α-MEM. The marrow cells were collected, washed and cultured in α-MEM containing 10% FCS and rhM-CSF (100 ng/mL) supplemented with 100 U/mL penicillin and 100 mg/mL streptomycin. After three days of culture, the cells were vigorously washed to remove the nonadherent cells, detached by pipetting and subcultured. After culturing for an additional three days, the cells were harvested and used as rOPCs.
Immunohistochemistry and immunofluorescence
Immunohistochemistry was performed as previously described [
12]. Surgical specimens were initially decalcified for two weeks in an EDTA-containing buffer and embedded in paraffin. The endogenous peroxidase activity was quenched by incubating the sections for an additional 30 min in absolute methanol and 3% hydrogen peroxide. The slides were then incubated with the appropriate primary Abs, followed by biotinylated secondary Abs and peroxidase-conjugated streptavidin. The signals were detected using 3-amino-9-ethylcarbazole in N,N-dimethylformamide. To examine the pY-Flt-1 and pY-FAK levels in GCT samples, the staining intensity of each specimen was scored as follows: 1 (weak staining; less than 10% of cells were positive), 2 (intermediate staining; 10-50% positive) and 3 (strong staining; >50% positive). All molecular variables were scored by one investigator, who was blinded to the clinical stages of the patients.
For immunofluorescence, the samples were incubated with the primary Abs overnight at 4°C. The samples were washed in PBS and then incubated with FITC or TRITC-conjugated secondary Abs. Then, the sections were mounted and examined by confocal laser scanning microscopy.
Tissue culture of giant cell tumors of bone
Primary cultures of GCTs were obtained from surgical samples of lytic bone lesions. As previously described [
6], fresh tumor tissues were minced in DMEM containing 10% FBS supplemented with 100 U/mL penicillin and 100 μg/mL streptomycin. The cell suspension containing small tissue pieces was plated in a 10 cm-culture dish and incubated at 37°C in a humidified atmosphere with 5% CO
2 and 95% air. Half of the culture medium was replaced every three days with fresh DMEM containing 10% FBS. When the cells reached confluency, the primary cultures were scraped and subcultured. After several passages, the multinucleated giant cells and monocyte/macrophage-like round cells progressively disappeared from the cultures and only the proliferating spindle-shaped cells remained. At passage eight, the cells were cultured with serum-free DMEM for 24 h and the conditioned medium was collected, filtered through 2.5 μm filters, and used as GCT-conditioned medium (GCT-CM).
Immunoblotting
When the cells reached approximately 70% confluency, they were harvested and solubilized in lysis buffer [20 mM Tris (pH 7.4), 250 mM NaCl, 1.0% NP40, 1 mM EDTA, 50 mg/mL leupeptin, and 1 mM phenylmethylsulfonyl fluoride]. The protein quantity was determined with a Bradford protein assay (Bio-Rad, Hercules, CA). The samples were separated on 4-12% gradient pre-cast MOPS-polyacrylamide gels (Novex, San Diego, CA) and blotted onto nitrocellulose filters. After transfer, the filters were pre-treated with TBS containing 5% dry milk and 0.05% Triton X for 2 h at room temperature and then incubated with the indicated primary antibodies for 2 h at room temperature. After several washes, the membranes were probed with the appropriate horseradish peroxidase-conjugated secondary Abs at room temperature for 1 h. After the final wash, the immunoreactivity of the blots was detected using an enhanced chemiluminescence system (Amersham, Arlington Heights, IL).
Enzyme-linked immunosorbent assay (ELISA) for VEGF
The VEGF levels in GCT-CM were determined using an ELISA kit from R&D Systems (Minneapolis, MN).
Cell proliferation assay
rOPCs cells seeded in culture plates were incubated in serum-free media with various reagents (GCT-CM, VEGF and ZD4190) for 24 h. The cell growth rate was determined using a Celltiter-Glo Luminescent Cell Viability Assay Kit (Promega, Madison, WI) according to the manufacturer's protocol.
Chemotaxis assay
The chemotaxis assay was performed using transwell chambers (Costar, Cambridge, MA) as previously described [
13‐
15]. Briefly, rOPCs were suspended in serum-free α-MEM containing 1% bovine serum albumin and seeded in the upper chamber. The lower chamber was filled with serum-free α-MEM supplemented with or without various cytokines. Polyvinylpyrrolidone-free polycarbonate filters with 8.0-μm pores were coated with type IV collagen and inserted between the two chambers. Then, the cells were allowed to migrate for 6 h at 37°C. After this incubation period, the cells that had migrated to the lower side of the filter were fixed, stained and counted using five fields/filter under a microscope.
Statistical analysis
The results obtained from the chemotaxis and cell proliferation assays are expressed as the means ± SD and were statistically analyzed by the Student's t-test. The association between the expression levels of various molecular factors (pY-FAK and pY-Flt-1) and the clinical stages were analyzed using the Mann-Whitney U test.
Discussion
The association between VEGF expression and angiogenesis has been detected in many solid tumors. In addition, VEGF-induced vascularization during bone development is critical for the formation of OCs [
18,
19]. Therefore, VEGF may be involved in both angiogenesis and osteoclastogenesis. It has been reported that the level of VEGF gene expression in GCTs correlates with the clinical stage at presentation defined by Enneking's surgical staging system [
10], suggesting that the production of VEGF by tumor cells and the induction of angiogenesis may partially contribute to tumor progression. In this study, VEGF was clearly expressed in stromal cells, monocyte-like cells and MNCs in GCTs. CD68, an intracellular glycoprotein, was expressed in monocyte lineage cells, including OPCs and OCs [
20]. Therefore, it is possible that the infiltrating MNCs and monocyte-like cells in GCTs mature into OCs and OPCs, respectively. In contrast, the stromal cells did not express CD68, suggesting that they did not originate from the monocyte-macrophage lineage.
In endothelial cells, the VEGF signals were mainly mediated via Flk-1, the type-II VEGF receptor [
21]. However, in monocytic lineage cells, most VEGF signals were transmitted via Flt-1, as was previously shown [
9]. In regard to the effect of VEGF on monocytes migration, VEGF stimulated the chemotaxis of human monocytes corresponding to the previous report [
22]. (Control; 5 ± 1 cells, 10 nM VEGF: 50 ± 5 cells, 10 nM MCP-1: 99 ± 3 cells) We also found that VEGF treatment induced the tyrosine phosphorylation of FAK (pY-FAK). In this study, we first demonstrated that CD68 and Flt-1 co-localized in MNCs and monocyte-like cells, which are thought to be OPCs in GCTs. However, these cells did not express Flk-1. We also indicated that these cells expressed an activated and tyrosine-phosphorylated Flt-1 (pY-Flt-1) as shown in Figure
5. In addition, pY-FAK was expressed in monocyte-like cells in GCT surgical specimens. These results support the hypothesis that VEGF is released from stromal cells, pOCs and OCs. Then, through paracrine and autocrine mechanisms, the secreted VEGF activates the VEGF-Flt-1-FAK pathway. The activation of this signaling pathway might be involved in the migration of these cells into the lesion at the site of bone destruction in GCTs.
We recently showed that VEGF stimulates the chemotaxis and cell proliferation of RAW cells, a model of mouse OPCs. Thus, we investigated the biological effects of VEGF in GCTs using GCT-CM and rOPCs. Consistent with the immunohistochemistry results, GCT-CM contained VEGF and treating rOPCs with GCT-CM resulted in the tyrosine phosphorylation of FAK within cells. GCT-CM also stimulated the chemotaxis and proliferation of rOPCs. All of these GCT-CM-induced effects were inhibited by adding ZD4190, a VEGF RTK inhibitor, to the GCT-CM. It was recently reported that VEGF treatment induces the formation of osteoclasts in osteopetrotic (op/op) mice that lack functional macrophage colony-stimulating factor [
23]. Thus, it is possible that the VEGF produced by GCTs directly stimulates the formation of MNCs within the tumor. These results suggest that the effects of GCT-CM, including the stimulation of chemotaxis and proliferation of rOPCs but not osteoclastogenesis, were partially dependent on VEGF-Flt-1-FAK signaling and that this signaling plays important roles in recruiting OPCs into the GCT tissue. On the other hand, ZD4190 did not completely block the basal level of chemotaxis and proliferation of rOPCs. Therefore, we assumed that many other cytokines, including TGF-β1[
5], MCP-1[
6] and M-CSF, in GCTs influence the chemotaxis and growth of rOPCs. Meanwhile, a recent study showed that GCTs enhanced osteoclastogenesis via paracrine VEGF secretion under local hypoxic conditions and indicated that this might be a critical mechanism for the pathogenesis of GCTs [
24]. However, when we harvested GCT-CM under normoxia, GCT-CM did not enhance the osteoclastogenesis of OPCs. Therefore, the role of VEGF in osteoclastogenesis in GCTs
in vivo should be further investigated.
To assess the pathological significance of the VEGF-Flt-1-FAK pathway, we also examined the correlation between the pY-Flt-1 and pY-FAK expression levels and the clinical stage of GCTs at presentation. The biological aggressiveness of the tumors was classified as previously described [
25]. In the present study, we demonstrated that the pY-Flt-1 and pY-FAK expression levels correlated with clinical stage of the tumor. A relatively high level of pY-Flt-1 and pY-FAK expression was observed in stage III GCTs compared with stages I-II GCTs. Although a larger number of tumors are needed to confirm these clinical correlations, our results suggest that activation of the VEGF-Flt-1-FAK pathway may contribute to the clinical progression of GCTs.
Conclusions
In conclusion, our results suggest that the VEGF-Flt-1-FAK pathway is potentially involved in recruiting OPCs in GCTs. This pathway, in concert with other factors such as TGF-β and MCP-1, may stimulate the recruitment and cell proliferation of OPCs into GCTs, resulting in tumor progression. In this study, ZD4190, a p.o.-active VEGF RTK inhibitor, disrupted VEGF signaling mediated by Flt-1 as well as Flk-1, indicating that ZD4190 administration may simultaneously inhibit VEGF-induced angiogenesis and the recruitment and proliferation of OPCs in GCTs. Therefore, it is conceivable that VEGF RTK inhibitors may be a useful clinical therapeutic for GCTs.
Acknowledgements
This work was supported by a Grant-in Aid for Scientific Research (19390397 and 19791036) from the Japan Society for the Promotion of Science, Grants-in-aid for Clinical Research Evidenced Based Medicine, and for Cancer Research from the Ministry of Health, Labour and Welfare of Japan. This work was also supported by a grant from the Japan Orthopaedics and Traumatology Foundation, Inc. No. 177.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
YM conceived of the study, carried out the experimental studies, and drafted the manuscript. YO, JF, SK, TF, KI and MK carried out experimental studies. SM, KH and AS participated in the design of the study and performed the data analysis. YI participated in its design and helped to draft the manuscript. All authors read and approved the final manuscript.