Background
Glioblastoma multiforme (GBM), classified as a grade IV astrocytoma, has an extremely poor prognosis [
1]. Long-term survival of patients with malignant gliomas has not improved substantially despite the development of multimodality treatments, including cytoreductive surgery, adjuvant radiation therapy, and cytotoxic chemotherapy. In order to develop additional therapeutic strategies, further understanding of the molecular genetics, biology and immunology of gliomas is desired.
GBMs are distinguished pathologically from lower-grade anaplastic astrocytomas by the presence of necrosis and microvascular hyperplasia, a florid form of angiogenesis [
2]. Above all, a striking feature of GBMs is the presence of increasing neovascularization [
3]. Many studies have demonstrated that glioma growth is dependent on the generation of tumor-associated blood vessels [
4,
5], therefore, use of antiangiogenic strategies is considered as a promising approach for the treatment of malignant gliomas.
There has been important progress in the elucidation of the molecular pathogenesis of malignant gliomas. Two common and highly specific genetic events associated with the GBM histology are epidermal growth factor receptor (EGFR) amplification and loss of the phosphatase and tensin homologue on chromosome 10 (PTEN) [
6,
7]. Many studies have revealed that EGFR is functionally dysregulated in various tumors. Dysregulation of signal transduction processes affects a variety of downstream biological processes associated with gene transcription and protein translation, cell proliferation, migration, adhesion, invasion, and angiogenesis [
8]. Abnormalities of EGFR signaling have also been reported to be observed frequently in GBMs [
9]. EGFR gene amplification or overexpression is detected in approximately 40% of patients with these tumors [
10,
11].
The EGFR variant type III (EGFRvIII), the most common mutation of EGFR in GBMs, is reported to be present in 25% to 33% of all cases of GBMs, but only in those showing EGFR amplification and overexpression [
12]. EGFRvIII overexpression has been shown to induce tumor growth of GBMs [
13] and reported to be correlated with a poor prognosis in clinical settings [
14,
15]. This EGFR variant is the result of deletion of exons 2 to 7 including the extracellular ligand-binding domain, and its receptor tyrosine kinase is constitutively active [
9]. Because it is not present in normal tissues, it is considered as a potential target for tumor-specific therapy. Currently, considerable effort is being made for the development of anti-EGFRvIII agents, such as vaccines and specific antibodies [
7,
16,
17].
EGFR signaling promotes not only cell growth, but also angiogenesis by induction of proangiogenic factors such as the vascular endothelial growth factor (VEGF) and interleukin-8 (IL-8) [
18]. Although the NF-kB/IL-8 pathway contributes to tumor angiogenesis in EGFRvIII-overexpressing glioblastomas [
19], the EGFRvIII signaling pathways involved in the promotion of angiogenesis have not yet been clearly elucidated. In this study, we show the involvement of EGFRvIII in tumor angiogenesis in LN229, a GBM cell line, and that the induction of angiopoietin-like 4 (Angptl4) expression by c-Myc is involved in EGFRvIII-induced angiogenesis.
Discussion
Although EGFRvIII has been shown to promote tumor growth of gliomas through various signaling pathways [
9], the key signal molecules involved in the alteration of the tumor microenvironment have not yet been fully elucidated. In this study, we investigated whether EGFRvIII contributes to tumor angiogenesis, and showed dramatic increases in the microvessel density and vascular permeability in tumor xenografts of LN229-vIII as compared to LN229-WT in mice, consistent with the results of a previous study [
19]. Considering that hypervascularity is a distinctive pathological characteristic of malignant gliomas [
31,
32], the EGFRvIII expression status may have a great impact on the clinical picture. Although EGFR is known to promote angiogenesis by induction of proangiogenic factors, such as VEGF-A and interleukin 8 [
18,
33], no dramatic induction of angiogenesis by wtEGFR was observed in our experiments. This difference leads to the speculation that constitutive activation of EGFR may trigger striking induction of various transcripts, including pro-angiogenic factors. In order to examine the molecular mechanisms underlying the induction of angiogenesis by EGFRvIII, the expressions of 60 angiogenic factors in LN229 cells were examined by real-time PCR analysis. Although VEGF-A is a representative angiogenic factor and a possible therapeutic target for glioblastoma [
34], VEGF-A induction by EGFRvIII was observed only to a certain extent
in vivo (data not shown), and not at all
in vitro (Additional file
3: Table S1). Among the 60 angiogenic factors, we first found that Angptl4 expression was significantly induced by EGFRvIII overexpression, and that Angptl4 acts as a pro-angiogenic factor in tumor xenografts. Recently, Bonavia, et al. showed that the NF-kB/IL-8 pathway plays important roles in EGFRvIII-induced angiogenesis and growth in gliomas [
19], however, no significant change of the IL-8 expression was observed in our in vitro experiment (Additional file
3: Table S1). It is likely that the differences between our results and those of the previous report are related to differences in the cell lines.
The molecular mechanisms of Angptl4-induced angiogenesis in malignant gliomas still remain largely unknown. Angptl4 is expressed in the liver, adipose tissue and placenta, as also in ischemic tissues [
35]. It is a member of the angiopoietin family and is a target of members of the peroxisome proliferator-activated receptor (PPAR) family, which are known as metabolic-response transcription factors [
36]. It has been reported that expression of Angptl4 is upregulated under various conditions including hypoxia and caloric restriction, and transcription factors such as PPARγ and Smad have been shown to regulate its expression [
35,
37,
38]. Increased Angptl4 expression has been shown in a variety of tumor tissues, such as oral Kaposi’s sarcoma, esophageal squamous cell carcinoma, gastric cancer, and colorectal cancer [
39‐
42]. Since a number of reports have indicated the effects of Angptl4 on angiogenesis, including endothelial cell proliferation, migration, differentiation, endothelial cell adhesion, and vascular permeability [
43‐
46], it seems likely that Angptl4 contributes to the increased angiogenesis and vascular permeability in gliomas formed by EGFRvIII cells. Moreover, it has been demonstrated that Angptl4 disrupts vascular endothelial cell-cell junctions and promotes lung metastasis of breast cancer cells expressing transforming growth factor-β [
35], while preventing metastasis of melanoma cells [
47] and also inhibiting angiogenesis [
48]. These diverse and often conflicting results suggest that Angptl4 exhibit tissue-specific activity and act in accordance with the prevailing cellular environment.
Our results suggest that Angptl4 transcription is regulated, at least partially, by EGFRvIII/ERK/c-Myc-mediated signaling. EGFR activation induces Ras/MEK/ERK phosphorylation, and phosphorylated ERK activates various transcription factors. It has been shown that MAPK signaling contributes to Angptl4 expression [
25]. Myc is known as an ERK-activated transcription factor [
30]. Wild-type EGFR expression, as compared to mock, increased tumor growth and Angptl4 expression
in vivo, and also activated ERK phosphorylation in the LN229 cells; however, the degree of activation was not significantly different from that induced by EGFRvIII expression (data not shown). These data suggest that, although the MAPK pathway plays an important role in c-Myc activation, other factors are also involved in the marked activation of c-Myc and induction of Angptl4 expression in the LN229-vIII cells. The promoter region of Angptl4 contains the consensus sequence of c-Myc, ‘CACGTG’. The results of the ChIP assay revealed enhanced binding between c-Myc and the promoter region of Angptl4 in LN229-vIII cells, suggesting that the transcriptional regulation of Angptl4 by c-Myc might contribute to the induction of angiogenesis in gliomas. An MEK inhibitor was also found to markedly inhibit Angptl4 expression in EGFRvIII-overexpressing LN229 cells. In a previously reported study, combined use of an MEK inhibitor with a PI3K inhibitor effectively suppressed the growth of gliomas [
49]. MEK inhibitors have been examined in clinical trials for various cancers, and their potential usefulness in the treatment of gliomas has been suggested.
Conclusions
In conclusion, we demonstrated in this study that EGFRvIII induces Angptl4 expression through the ERK/c-Myc pathway, and that Angptl4 is a possible inducer of tumor angiogenesis in gliomas expressing EGFRvIII. Since EGFRvIII strongly induces neovascularization in the tumors, expression of EGFRvIII or Angptl4 may be a possible biomarker for predicting the effectiveness of antiangiogenic therapy, as well as serve as a therapeutic target, although further studies are needed.
Methods
Cell culture
The human glioblastoma cell lines LN229 (American Tissue Culture Collection) were maintained in Dulbecco’s minimal essential medium (DMEM, Sigma) supplemented with streptomycin (100 μg/ml), penicillin (100 units/mL), and 10% heat-inactivated fetal bovine serum (FBS) at 37°C under 5% CO
2 in a humidified chamber. The cDNA for wild-type EGFR or EGFRvIII was transfected into LN229 cells by a retrovirus vector, as described previously [
17], and the transfected cells were selected by GFP expression from the viral expression vector using a cell sorter (BD Biosciences).
Cell proliferation assay
LN229 cells (1,000 cells/well) were seeded into a 96-well microtiter plate. After incubation for 24-96 h at 37ºC, the cell viability was measured with a Cell Counting Kit-8 (Dojindo, Tokyo, Japan) in accordance with the manufacturer’s instructions.
RNA isolation, reverse-transcription PCR, and real-time PCR
Total RNA was isolated using Isogen (Nippon Gene, Tokyo, Japan) and the resulting RNA was reverse-transcribed with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Tokyo, Japan). Real-time PCR assay was performed on a StepOnePlus (Applied Biosystems) using the TaqMan Gene Expression Assays or a TaqMan Array Gene Signature 96-Well Plate (Angiogenesis, human, Applied Biosystems). The relative real-time PCR quantification was based on a comparative quantitation method.
Western blotting
Western blotting was performed as described previously [
50], with some modifications. The cells were washed with ice-cold PBS and lysed with M-PER (PIERCE, Tokyo, Japan) containing protease and phosphatase inhibitors. The protein concentration was determined using a BCA protein assay kit (PIERCE). The protein samples were mixed with SDS-PAGE sample buffer (2% SDS, 10% glycerol, 6% 2-mercaptoethanol, 50 mM Tris-HCl; pH 6.8), and an equal amount of proteins in each sample was subjected to SDS-PAGE. The separated proteins were transferred to a PVDF membrane (Millipore, Bedford, MA) and blocked with 5% skim-milk in TBST (0.9% NaCl, 0.1% Tween20, 20 mM Tris-HCl; pH7.4). The primary antibodies used were anti-EGFR antibody (BD Pharmingen, NJ) and anti-actin antibody (Sigma). Horseradish peroxidase (HRP)-conjugated antibodies (Cell Signaling Technology, Tokyo, Japan) were used as the secondary antibodies. The PVDF membrane was developed with the ECL reagent (GE Healthcare, Buckinghamshire, UK).
Tumor xenograft model
LN229 cells were subcutaneously implanted (3.0 × 10
6 cells/mouse) into the posterior flanks of 4-week old female BALB/c nu/nu mice. The tumor sizes were monitored as described previously [
51]. Animal studies were carried out according to the Guideline for Animal Experiments, drawn up by the Committee for Ethics in Animal Experimentation of the National Cancer Center, which meet the ethical standards required by law and the guidelines about experimental animals in Japan.
Microvessel density analysis
After tumor implantation, the mice were sacrificed under diethyl ether anesthesia, and the tumors were dissected and weighed. Immunostaining was performed as described previously [
52]. The tumor tissues were embedded and frozen with dry ice/ethanol. Tumor frozen sections (7 μm) were prepared and air-dried for at least 1 h. The sections were fixed with cold acetone, blocked in goat serum for 10 min at room temperature, and then incubated with anti-mouse CD31 rat monoclonal antibody (BD Pharmingen) for 18 h at 4°C. The sections were then stained with ABC Elute kit, or anti-rat IgG-Alexa fluor 555 conjugates (Molecular Probes, Inc.) for immunohistochemistry and immunofluorescent staining, respectively. After mounting the sections, the images were examined and scanned with Biozero (Keyence, Tokyo, Japan) at 20 × magnification. For quantitative analysis, the vascular area/mm
2 in the tumors was quantified by counting the CD31-positive area in independent hotspots of at least four different microscopic fields in each of five mice/group, using the ImageJ software. The four fields were averaged in each tumor and the averages for each animal used to express the final count ± SEM.
Vascular permeability
The in vivo vascular permeability assay was performed as described previously with some modifications [
53]. The tumor-implanted mice were intravenously injected with TexasRed conjugated dextran (50 mg/kg/mouse, Mw 70,000, Molecular probes, Inc). At 6 h after the injection, Alexa647-conjugated Isolectin IB4 (Molecular probes, Inc) was injected for fluorescent staining of the blood vessels. After 10 minutes, perfusion fixation was performed under ether anesthesia and the tumors were extracted from the mice. The extracted tumors were frozen and sectioned as described above. The sections were fixed with 4% paraformaldehyde, mounted, and observed by fluorescent microscopy as described above.
Enzyme-linked immunosorbent assay (ELISA)
LN229 cells were seeded (3.0 × 105 cells) in a 35-mm dish and incubated overnight. The medium was refreshed and the culture dish was incubated for a further 48 h at 37°C. The culture medium was collected and centrifuged at 1,000 g for 10 min. The supernatant was recovered and ELISA for Angptl4 was performed using the Human Angiopoietin-like 4 DuoSet ELISA kit (R&D Systems, Minneapolis, MN) with a sensitivity of 1.25 ng⁄mL, an intra-assay coefficient of variation of 0.6–7.6%, and an inter-assay coefficient of variation of 8.5–11.2%. The assay was performed in accordance with the manufacturer’s instructions. The remaining cells on the dishes were lysed and the amount of protein was measured by a BCA protein assay. Tumor tissues extracted from the mice were homogenized in PBS (–) and centrifuged at 10,000 × g for 10 min at 4°C. The supernatant was collected and ELISA was performed as described above. Duplicate measurements were performed in a single experiment.
Electrophoretic mobility shift assay (EMSA)
Nuclear fractions were extracted from the LN229 cells using a Nuclear Extraction kit (Panomics, Redwood City, CA). The EMSA binding assay was carried out using a Panomics EMSA “gel shift” kit in accordance with the manufacturer’s instructions. Assays were conducted using a biotin-labeled double-stranded oligonucleotide having a consensus recognition sequence for Myc/Max purchased from Panomics. Protein-DNA complexes were separated using nondenaturing PAGE. The oligonucleotides were secondarily probed with HRP-conjugated streptavidin and developed with the component solution by LAS4000.
RNAi experiment
The RNAi experiment was performed with the Lipofectamine RNAiMAX reagent (Invitrogen, Tokyo, Japan) in accordance with the manufacturer’s instructions. The sequences of siRNA for c-Myc were 5′-AGA CCU UCA UCA AAA ACA UTT-3′ (sense) and 5′-AUG UUU UUG AUG AAG GUC UCG-3′ (antisense), which were designed by Ambion, and the non-silencing control siRNA was purchased from Invitrogen. After incubation with the siRNA for 48 h at 37°C, the mRNA expressions of c-Myc and Angptl4 were quantitatively determined by real-time PCR. Short hairpin RNA targeting the Angptl4-including entry vector was designed and prepared by Invitrogen. The shRNA was subcloned to a retrovirus vector and used in the experiments as described in a previous study [
17].
ChIP assay
The ChIP assay was performed using the ChIP IT Express kit (Active Motif, Carlsbad, CA), in accordance with the manufacturer’s instructions. LN229 cells were fixed with 1% formaldehyde for 10 min. The cells were then washed, lysed, and sonicated to reduce DNA lengths to the range of 200 to 1500 bp. The chromatin/DNA complexes were incubated with antibodies to c-Myc (Santa Cruz Biotechnology, CA) or IgG (Cell Signaling Technology) overnight at 4°C. The immune complexes were precipitated, eluted, reverse-crosslinked, and treated with proteinase K. After extraction of the DNA fragments, real-time PCR analysis was performed using Power SYBR green PCR master mixes (Applied Biosystems). The primer for the promoter of Angptl4 was purchased from BioScience (Fredrick, MD), and was as follows: forward, 5′-TAC TAG CGG TTT TAC GGG CG-3′; reverse, 5′-TCG AAC AGG AGG AGC AGA GAG CGA-3′. The predicted PCR product included a c-Myc binding sequence. Relative enrichment was comparatively calculated using IgG negative control as described in eBioScience instructions.
Statistical analysis
Significant differences were analyzed by an unpaired Student’s t-test or analysis of variance (ANOVA) with Tukey’s post-hoc test using the GraphPad Prism software (Version 5.0). p < 0.05 was considered to indicate statistically significant difference.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
YK and FK conceived the idea, designed the experiments, and drafted the manuscript. YK, YK and YK performed the experiments. FK, TM and TT edited the manuscript. All authors read and approved the manuscript.