Background
Glioblastoma multiforme (GBM) is the most aggressive malignancy among primary brain tumors with very low median survival. Due to the feature of high heterogeneity and infiltration, GBM’s etiology and pathophysiologic mechanisms remain unclear, resulting in limited treatment approaches. However, mounting evidence demonstrates that immune response is involved in development and progression of GBM [
1], and therefore, thorough investigation of the underlying molecular and immunologic mechanism will be beneficial for novel interventions.
GBMs are complex solid tumors containing neoplastic and non-neoplastic cells. The majority of the non-neoplastic cells are tumor-associated macrophages (TAMs), either of peripheral origin or brain resident microglia, which account for 30% of total tumor mass and are reported to play several roles in GBM progression including proliferation, motility, survival, and immunosuppression [
1,
2]. TAMs are recruited to GBM environment and release a wide array of chemokines and cytokines in response to those factors produced by GBMs [
1]. As a result, TAMs facilitate tumor proliferation, survival, and migration, acting as an anti-immunological and pro-tumorigenic role in the tumor microenvironment.
Microglia/macrophages exist in a spectrum of phenotypes. The classically activated microglia/macrophages, M1 phenotype, stimulate anti-tumor immune response through secretion of pro-inflammatory cytokines, such as IFN-gamma, IL1β, iNOS, etc., whereas the alternatively activated or M2 phenotypes promote tumor survival via producing anti-inflammatory cytokines like IL4, TGFβ, IL10, etc. [
3]. Within the tumor microenvironment, TAMs are forced forwards M2 phenotypes by GBM cells via secreting a wide variety of factors such as IL10, IL4, IL6, M-CSF, macrophage inhibitory factor (MIF), TGFβ, and prostaglandin E2 (PGE2), and subsequently support tumor growth and invasion [
4]. However, the pathogenesis of elevation of these anti-immune factors and reduction of pro-immune factors in GBM remains elucidated.
Enhancer of zeste homolog 2 (EZH2) is the core catalytic subunit of Polycomb repressive complexes 2 (PRC2) and can silence a bundle of tumor suppressor genes through methylation of lysine 27 of histone 3 (H3K27) of target genes [
5]. It has been reported to aberrantly express in multiple solid tumors. In particular, accumulated data including our study verify that EZH2 is highly abundant in GBM samples and that expression levels of EZH2 are positively correlated with GBM grades and unfavorable survival [
6,
7], which is therefore suggested as a biomarker for diagnosis and prognosis. In our previous study, we demonstrate EZH2 promotes GBM cell proliferation, cell cycle progression, inhibition of apoptosis, and invasion, serving as an oncogene [
8]. Recently, EZH2 is reported to be involved in adaptive immune response, whereas little information about EZH2 in innate immune response is available, especially in GBM.
In the present study, we aimed to investigate the influence of EZH2 on immune functions of microglia or macrophages in a co-culturing model with GBM cells. We demonstrated that EZH2 inhibition in GBM induced elevation of M1 makers (iNOS and TNFα) and reduction of a pool of M2 markers in murine microglia and human PBMC-derived macrophages. Furthermore, we found that EZH2 inhibition in GBM cells enhanced phagocytic capacities of co-culturing microglia through activation of iNOS.
Methods
Isolation of human monocyte-derived macrophages
Human peripheral blood mononuclear cells (PBMCs) were isolated using sequential Ficoll and Percoll density gradient centrifugations as described previously [
9]. Blood samples were collected from healthy volunteers, and the procedure was approved by the ethics committee of Sun Yat-sen Memorial Hospital, Sun Yat-sen University. Monocytes were purified using CD14 microbeads according to the manufacturer’s directions (Miltenyi Biotec) and then were treated with 20 ng/ml macrophage colony stimulating factor (M-CSF, switching monocytes to M2 phenotype macrophages) and cultured at 37 °C and 5% CO
2 in ultra-low attachment flasks (Corning) for 5 day in the RPMI 1640 medium with 10% fetal bovine serum (FBS, Gibco).
Isolation and culture of microglia
Primary microglia were prepared and isolated using 1-day-old C57BL/6 mice as described previously [
10]. Briefly, cells were isolated by trypsinization, mechanically dissociated and plated into 75-cm
2 flasks (BD Falcon) pre-coated with 0.4 μg/ml poly-L-lysine (Sigma-Aldrich) in DMEM-F12 medium containing 10% FBS, 100 U/mL penicillin and 0.1 mg/mL streptomycin (Invitrogen). Cells were maintained at 37 °C in a humidified atmosphere with 5% CO
2. The culture medium was changed after 1 day and then twice a week. Ten to fourteen days later, floating and loosely attached microglia were manually shaken off, were centrifuged (200 g, 6 min), and were suspended in the culture medium and plated at a final density of 2–3 × 10
5 cells/ml. Adherent cells (96% positive for microglia marker Iba1) were incubated for 48 h prior to experiments.
Cell culture and co-culture of GBM with macrophage/microglia
Human U87 and U251 GBM cell lines were obtained from China Academia Sinica cell repository (Shanghai, China). Cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM, Hyclone) supplemented with 10% FBS and incubated at 37 °C with 5% CO2. Murine GL261 GBM cell line was obtained from the Clinical Research Center of Nan-Fang Hospital of Southern Medical University (Guangzhou, China) and were cultured in DMEM-F12 medium containing 10% FBS.
Co-cultivation of PBMC-derived macrophages or mouse primary microglia with GBM cells (U87, U251, or GL261) were performed in 12-well Boyden chambers (Corning). GBM cells were seeded on the 0.4-μM inserts, which are permeable to supernatants but not to cellular components, with the cell density of 1 × 105 cells/insert. Macrophages or microglia were seeded in the lower chambers (5 × 104 cells/cm2) and grown for indicated periods of time.
GBM conditioned medium (GCM) was collected after 24 h culture and applied to microglia in a volume of 2 ml (35-mm dish) or 100 μl (96-well plate). The fresh medium was used as controls.
EZH2 siRNA transfection and EZH2 function inhibition
EZH2 siRNA (siEZH2) and negative control (NC) oligonucleotides were purchased from GenePharma (Shanghai, China, see Additional file
1). Oligonucleotides were transected into U87, U251, and GL261 cells using the transfection reagent Lipofectamine2000 (Invitrogen) following the manufacturer’s instructions with oligonucleotides at a working concentration of 100 nM. The transfected cells were incubated for 24 or 48 h before subsequent experiments.
The EZH2 functional inhibitor, DZNep, was purchased from Sigma-Aldrich. DZNep was added to U87 and U251 cell medium at a concentration of 5 μM for indicated time duration.
Gene expression microarray and bioinformatics analysis
GBM sample preparation: U87 cells were treated with siEZH2 and NC for 48 h and then total RNA was extracted using a TRIzol reagent (Life Technologies, USA) according to the manufacturer’s instructions. The same samples were repeated twice and then these three copy samples were mixed for next processes.
Microarray: The quantification and quality were assessed by spectrophotometry and an Agilent 2100 Bioanalyzer (Agilent Technologies, USA). The RNA samples at 2 μg without degradation was amplified, labeled, and hybridized. Gene expression profile was analyzed using the Agilent Whole Human Genome Oligo Microarray kit, 4 × 44 K (Agilent Technologies, USA), consisting of approximately 41,000 genes and transcripts. Ratio was calculated by comparison between siEZH2 and NC group. The experimental data were analyzed using the SBC Analysis System, including gene ontology (GO) and pathway enrichment analysis. Biological processes terms were performed by a functional enrichment analysis tool, FunRich_V3 (
http://www.funrich.org/).
RNA extraction and quantitative real-time PCR
Total RNA was extracted using Trizol reagent (Invitrogen) according to its protocol and was reversely transcribed using PrimeScript RT Reagent Kit (Takara, Otsu, Shiga, Japan). Quantitative real-time PCR (qPCR) was performed using SYBR Green PCR master mix (Takara) on an Applied Biosystems StepOnePlus™ System (Applied Biosystems, Foster City, CA, USA). All qPCR reactions were performed in triplicate, and relative quantifications (RQs) were calculated using the 2
-△△Ct method. The sequences of PCR primers were listed in Additional file
1.
Protein isolation and western blotting
Cells were washed twice with ice-cold PBS, and lysed in RIPA buffer (Pierce, Waltham, MA, USA). Protein lysates were separated by 8–10% SDS-PAGE and then were electrophoretically transferred to PVDF membrane (Millipore, Lake Placid, NY, USA). After blocked in 5% non-fat milk, the membrane was incubated with either rabbit anti-human EZH2 (1:1000, Cell Signaling Tech., Beverly, MA, USA), rabbit anti-human GAPDH antibody (1:1000, CST), followed by HRP (horseradish peroxidase)-labeled goat-anti-mouse or goat-anti-rabbit IgG (1:5000, Santa Cruz). Protein levels were detected using ECL detection solution (Pierce) and visualized on Bio-Rad ChemiDocXRS (Bio-Rad Laboratories, Hercules, CA, USA).
Phagocytosis assay
Microglia were plated onto 35-mm dishes at a density of 2 × 105/ml and were co-incubated with different conditions of GBM cells for 24 h. Then the fluorescence-labeled latex beads (Sigma) were added at the concentration of 4 μl/ml for 90 min at 37 °C. Later, cells were washed three times with PBS to remove non-phagocytized beads and were fixed with 4% paraformaldehyde. Next, bead phagocytosis of microglia was observed under confocal microscope (Olympus fv10i). In order to quantify phagocytosis, cell numbers with low (< 2 beads/cell), medium (2–10 beads/cell), or high (≥ 10 beads/cell) phagocytic activity were counted using a fluorescent microscope (× 20 magnification). In addition, murine GL261 cells were pre-treated with siEZH2 and DZNep for 24 h, and then, GL261 were removed and 1400 W (a specific iNOS inhibitor, Sigma) was added to microglia for another 24 h with a working concentration of 500 μM. Next, fluorescent latex beads were added to microglia for 90 min. Amounts of latex beads phagocytized by microglia were detected by flow cytometry. All these experiments were repeated thrice.
Flow cytometry
PBMC-derived macrophages co-cultured with U87 cells for 24 h were washed and resuspended in PBS. For intracellular markers, cells were fixed and permeabilized with permeabilization reagents (CALTAGTM Laboratories) and then incubated with anti-human CD68-FITC (macrophage marker, BD Bioscience, CAS: 562117) or anti-human CD206-APC (M2 microglia marker, BD Bioscience, CAS: 561763). After the final washing step, cells were analyzed by flow cytometry (BD). Flow cytometry was also used to detect microglia amounts containing fluorescent beads, as mentioned above.
Nitric oxide production
Murine primary microglia were co-cultured with GBM cells GL261 with different treatments for 24 h. Aliquots of cell culture supernatant were taken at 24 h for total NO production by using the Griess assay according to the protocol of the manufacturer (Beyotime Institute Biotechnology, Jiangmen, China), as absorbance at 540 nm from the ELISA plate reader [
11]. The amount of nitrite was normalized to the amount produced by untreated microglia.
Cell proliferation assay
Primary microglia were seeded in a 96-well culture plates at 2 × 104 cells/well in 100 μl for 48 h to achieve resting state. The medium was then replaced with 100 μl conditioned medium from GL261 cells, which were pre-treated with siEZH2 and DZNep with working concentrations of 100 nM and 5 μM, respectively. 24 h later, cell proliferation was determined by the CCK-8 assay Kit. Additionally, cytokines TGFβ1(Peprotech, CAS100-21), TGFβ2 (Peprotech, 100-35B), and 1400 W (an iNOS inhibitor, sigma) were added to primary microglia for 24 h at working concentrations of 10 μg/ml, 10 μg/ml, and 500 uM followed by CCK-8 assay. Besides, PBS were used as control. The absorbance of the solution was measured on a SpectraMax M5 multimode microplate reader at 450 nm. Cell viability of each group was expressed as a percentage relative to control.
Statistical analysis
All quantified data were repeated thrice at least and expressed as mean ± SD. Comparison between groups was analyzed by one-way analysis of variance (ANOVA) followed by Bonferroni post hoc test. P < 0.05 was set as statistical significance.
Discussion
In the present study, we demonstrate that EZH2 plays a role in GBM innate immune response for the first time. We find that EZH2 inhibition in GBM cells enhances phagocytic capacities and viabilities of co-culturing microglia with iNOS and TGFβ2 dependent. Mechanic researches show that knockdown of EZH2 inhibits expression of anti-inflammatory factors while promotes expression of pro-inflammatory factors in GBM cells. More importantly, EZH2 suppression in GBM contributes to polarization shift of microglia and PMMC-derived macrophage, reflected by increase of M1 markers and reduction of M2 markers.
The oncogene EZH2 is of considerable interest as a potential therapeutic target in GBMs. It has been confirmed that EZH2 is highly expressed in GBM samples and cell lines [
7,
12], and abundance of EZH2 is associated with high GBM grades and poor survival [
6,
7]. Numerous studies report that knockdown of EZH2 by siEZH2 or function suppression by DZNep induce GBM growth inhibition [
13‐
15] and EZH2 is well recognized to take part in multiple tumor processes such as cell cycle [
14], proliferation [
16], apoptosis [
16], invasion and mobility [
16], GBM stem cell differentiation and maintenance [
17,
18], tumor angiogenesis [
19], etc. Our previous study demonstrates that miR-138 effectively inhibits GBM cell proliferation in vitro and tumorigenicity in vivo through directly blocking an EZH2-mediated signal loop [
8]. Recently, the relation between EZH2 and immune response is paid more attention. One study reports that EZH2 mediates the humoral immune response and drives lymphomagenesis through formation of bivalent chromatin domains at critical germinal center (GC) B cell promoters [
20]. EZH2 is also believed to be involved in T cell function [
21]. Another study reports that downregulation of EZH2 results in significant increases in CIITA and HLA-DRA expression as well as increased cell surface expression of MHC II in breast cancer [
22]. These studies strongly suggest that EZH2 plays vital roles in immune response and inflammation. However, up to date, no reports were found to focus on EZH2-related immune response in GBM, as we have shown in the present study.
Accumulated evidence demonstrates that there is a strong interaction between GBM and microglia in the tumor microenvironment. It has been confirmed that GBM cells establish an immune-deficit microenvironment to promote TAMs acquiring the M2 phenotype [
23]. When microglia polarize to M2 phenotypes, they in turn contribute to local immunosuppression and support GBM growth and migration [
23]. To find out exact genes that are changed in TAMs, Szulzewsky et al. used microarray analysis to compare expression profiles of TAMs and naive control cells and demonstrated that expression of Gpnmb and Spp1 are highly upregulated in both murine and human TAMs [
24]. However, there are no other studies about these two genes on GBMs.
Our findings indicate that EZH2 inhibition in GBM decreases expression of M2 markers and increases expression of M1 markers in co-culturing microglia. Two of them, iNOS and TGFβ, are altered much more significantly than others. The NOS has three isoforms, including endogenous NOS (eNOS), neuronal NOS (nNOS) and inducible NOS (iNOS), responsible for production of NO from the amino acid L-arginine [
25‐
27]. iNOS are inducible in several types of cells, including epithelial, mesenchymal, and myeloid cells [
28], and are reported to be aberrant in various human tumors such as breast and stomach cancer [
29‐
32]. Actually, iNOS has been emphasized in GBM pathogenesis by the fact that inhibition of iNOS by genetic and pharmacological approaches impedes glial cell proliferation, invasiveness, and tumor growth in vivo [
33], and that repression of GBM iNOS in vivo leads to a reduction in both microglia recruitment and tumor expansion [
34]. iNOS usually exerts a function of promoting microglia/macrophage polarization toward M1 phenotypes. In our study, we show that relative to microglia co-incubated with GBM without intervention, iNOS increases more than 2 times in microglia co-cultured with EZH2-suppressed GBM cells. Together with reductions of other cytokines shown in Fig.
2, these data suggest that iNOS is involved in EZH2-mediated microglia polarization shift.
TGFβ is a multifunctional cytokine that is involved in tissue homeostasis and embryonic development [
35,
36]. There are three different isoforms, including TGFβ1, TGFβ2, and TGFβ3. Among them, TGFβ1 and TGFβ2 are involved in brain tumor development and progression, particularly in high-grade GBM [
37‐
41]. Besides, the high activity of the TGF-β signaling pathway in human GBM tissues has been associated with poor prognosis [
42]. Our findings demonstrate that EZH2 inhibition in GBM leads to reductions of TGFβ1 and TGFβ2 while TGFβ2 promotes microglia viabilities. This suggests that TGFβ2 is involved in EZH2-mediated GBM progression and growth, which needs more concentration.
Also, we show that EZH2 inhibition in GBM accelerates phagocytosis ability of co-culturing microglia. Ellert-Miklaszewska et al. report primary microglia from rats polarize into alternatively activated cells, as exposed to GBM conditioned medium (GCM) for 6 h [
43]. Furthermore, GCM strongly augments motility of microglia in a scratch assay and phagocytosis of fluorescent latex beads [
43]. Interestingly, another study shows that microglia are activated after 1 h of co-culture with GBM, but longer in contact with tumor cells appears to silence phagocytic properties of microglia [
44]. In this study, we attempt to mimic the real tumor microenvironment and 24-h co-culture is set. Consistent with the above long incubation study, our results show that co-culture of GBM with microglia for 24 h does not change phagocytic ability of microglia. Moreover, we observe that EZH2 inhibition in GBM promotes upregulation of iNOS and production of NO from microglia and improves microglia phagocytosis, which can be reversed by the specific iNOS inhibitor. One recent study has identified a critical role for NO in the microglia phagocytic capacity under inflammatory conditions [
45]. They show that both addition of exogenous NO and induction of NO generation improve microglia phagocytic capacity independent of cGMP signaling [
45], which is supportive of our results. Therefore, our data provides mechanistic evidence of how EZH2 in GBM influences phagocytic activity of microglia.