The study protocol was approved by the Institutional Review Board of the First Affiliated Hospital of China Medical University. A total of 90 cases of archived primary tumor specimens embedded in paraffin were obtained from the Department of Pathology. For 15 of the 90 patients, paraffin-embedded adjacent non-cancerous normal brain tissues were also obtained. These specimens were from 90 patients diagnosed with astrocytoma who underwent resection at the First Affiliated Hospital of China Medical University between 2012 and 2014. Informed consent was obtained from all these patients. Histological diagnosis was performed and the astrocytoma tissues were classified as grades I-IV according to the WHO classification guidelines (2007).

Surgically excised astrocytoma specimens were fixed in 10% neutral buffered formalin, embedded in paraffin, and 4-µm-thick sections were prepared. Astrocytoma tissue sections were used for immunostaining with the Maixin Elivision plus kit (Fuzhou Maixin Biotechnology Co., Ltd., Fuzhou, China). The sections were deparaffinized and antigen retrieval was performed in citrate buffer for 2 min using an autoclave. Endogenous peroxidase was then blocked using hydrogen peroxide. Eya2 rabbit polyclonal antibody (1:300 dilution; 11314-1-AP; Proteintech Inc., Chicago, IL, USA) was used to incubate the tissue sections at 4°C overnight. Incubation with polymer secondary antibody was performed after washing with phosphate-buffered saline (PBS). The DAB plus kit (Fuzhou Maixin Biotechnology Co., Ltd.) was used to develop the staining. Counterstaining was carried out using hematoxylin, and the sections were dehydrated in alcohol before mounting.

Strong nuclear staining was regarded as positive. Samples with <15% strong nuclear staining were regarded as having a negative/low expression. Tumor samples with ≥15% strong nuclear staining were considered to have a Eya2 high expression.

The U251, U87, A172 cell lines were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA). The cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (HyClone, Logan, UT, USA). The astrocytoma cells were cultured in sterilized bottles. the Eya2 plasmid was obtained from OrigGene Technologies, Inc. (Rockville, MD, USA) and transfection was performed using Lipofectamine 3000 reagent. siRNA against Eya2 (Eya2 siRNA), Six1 siRNA and control siRNA were all purchased from Dharmacon (GE Healthcare Dharmacon Inc., Lafayette, CO, USA). DharmaFECT reagent (GE Healthcare Dharmacon Inc.) was used to transfect the siRNA. ERK inhibitor (PD98059) was purchased from Sigma-Aldrich (St. Louis, MO, USA). The cells were treated with PD98059 at the concentration of 20 µM for 6 h.

RNA was isolated from the cells and tissues using TRIzol reagent (Life Technology, St. Louis, MO, USA) according to the manufacturer's instructions. cDNA was obtained using the PrimerScript RT Master Mix kit (Takara, Dalian, China). 20 µl of reverse-transcription reaction solution consisted of 4 µl 5X RT Master Mix and 800 ng RNA. The reaction procedure was 85°C, 2 min and 37°C, 30 min. Quantitative PCR was performed using SYBR-Green PCR Master mix (Takara Bio, Inc., Shiga, Japan) in a total volume of 20 µl on 7900 Real-Time PCR system (Applied Biosystems, Foster City, CA, USA) as follows: 95°C for 30 sec, 40 cycles of 95°C for 5 sec, 60°C for 30 sec. β-actin was used as the reference gene. The relative levels of gene expression were calculated using the 2^−ΔΔCq method. The sequences of the primers are listed as follows: Eya2 forward, 5′-CACTCCCTGAAGGCACTAAACCTCATC-3′ and reverse, 5′-CTGCATTATCCTCTCGAAGCAGCTCTC-3′; cyclin D1 forward, 5′-TGGAGGTCTGCGAGGAACA-3′ and reverse, 5′-TTCATCTTAGAGGCCACGAACAT-3′; cyclin E forward, 5′-AGCCAGCCTTGGGACAATAAT-3′ and reverse, 5′-GAGCCTCTGGATGGTGCAAT-3′; MMP9 forward, 5′-CCTCTGGAGGTTCGACGTGA-3′ and reverse, 5′-TAGGCTTTCTCTCGGTACTGGAA-3′; actin forward, 5′-ATAGCACAGCCTGGATAGCAACGTAC-3′ and reverse, 5′-CACCTTCTACAATGAGCTGCGTGTG-3′.

Cell invasion assay was performed using a Transwell chamber. The inserts were coated with 18 µl Matrigel from BD Biosciences (San Jose, CA, USA) with a dilution rate of 1:4. At 48 h after siRNA transfection, the cells were trypsinized and 100 µl of serum-free DMEM were used to dilute cells, which was added to the upper chamber followed by incubation for a further 20 h. DMEM with 15% FBS was added to the lower chamber. The cells that passed through the filter were fixed, stained with hematoxylin and counted under a microscope (BX53; Olympus, Tokyo, Japan).

For colony formation assay, following transfection, the cells were plated into 6-cm dishes and cultured for 2 weeks. The cells were then stained with Giemsa (Sigma-Aldrich). to observe the colonies. The colonies with >50 cells were counted using a microscope (CX23; Olympus).

For CCK-8 assay, the cells were seeded in 96-well plates and 20 µl CCK-8 solution were added to each well. CCK-8 solution was incubated with the cells for 4 h, and the medium was then removed, and the remaining formazan was dissolved using DMSO and measured at 450 nm using a microplate reader (5119100, Multiskan™ FC; Thermo Fisher Scientific, Inc., Carlsbad, CA, USA).

Total proteins were extracted using Pierce cell lysis buffer (Pierce, Thermo Fisher Scientific, Inc., Rockford, IL, USA). The proteins were quantified by the Bradford method. A total of 50 µg of protein was transferred onto PVDF membranes following separation by SDS-PAGE. The membranes were incubated at 4°C overnight with primary antibodies against Eya2 (1:1,000; 11314-1-AP; Proteintech), Six1 (1:1,000; SAB2104425; Sigma-Aldrich), cyclin D1 (2978), cyclin E (20808), p-ERK (9101), ERK (9102), MMP9 (3852) and GAPDH (2118) (1:1,000; all from Cell Signaling Technology, Inc., Boston, MA, USA). Following incubation with HRP-coupled anti-mouse (sc-2789)/rabbit (sc-2357) IgG (1:1,000; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) at 37°C for 2 h, the bound proteins were visualized using an ECL kit (Pierce) and detected using a DNR Bio-Imaging System (DNR Bio-Imaging Systems, Ltd., Jerusalem, Israel). The densitometry of western bands was measured using ImageJ software.

A sufficient amount of antibody was added to 500 µg protein and gently rotated at 4°C overnight. The immunocomplex was captured by adding 25 µl protein A/G agarose beads (Cell Signaling Technology, Inc.) and rotated at 4°C for 2 to 3 h. The mixture was centrifuged at 1,500 × g for 5 min and the precipitate was washed 3 times with cold PBS, resuspended in sample buffer and boiled for 5 min to dissociate the immunocomplex from the beads. The supernatant was then collected by centrifugation (1,500 × g for 5 min at 4°C) and subjected to western blot analysis.

SPSS version 16.0 software (SPSS Inc., Chicago, IL, USA) for Windows was used for all statistical analyses. The Chi-square test was used to investigate the clinical data. A t-test was used to compare other data. A value of p<0.05 was considered to indicate a statistically significant difference. The experiments were repeated 3 times.

We examined Eya2 protein expression in 90 cases of astrocytoma tissues and 15 adjacent non-cancerous tissues using immunohistochemistry. We found a negative/low expression of Eya2 in glial cells in normal tissue (Fig. 1A). Positive nuclear Eya2 staining was observed in 36.7% (33/90) of the astrocytoma tissues (Fig. 1B–F). We also analyzed the correlation between Eya2 expression and clinical characteristics. A high Eya2 expression was found to positively correlate with an advanced tumor grade. As shown in Table I, the positive rates of Eya2 protein in grade I (pilocytic astrocytoma), grade II (diffuse astrocytoma), grade III (anaplastic astrocytoma) and grade IV (glioblastoma) astrocytoma were 0% (0/8), 26.7% (8/30), 45.5% (10/22) and 50% (15/30), respectively (p=0.03). No correlation was found between Eya2 and age (p=0.62) and gender (p=0.82).

In order to explore the biological function of Eya2 in astrocytoma cells, we analyzed its expression level in 3 astrocytoma cell lines (U251, U87 and A172) (Fig. 2A). Transfection with the Eya2 expression plasmid was performed in the A172 cell line, which was fouond to have a low endogenous expression of Eya2. siRNA knockdown was carried out in the U251 cell line, which was found to have a high expression of Eya2 (Fig. 1A). As shown in Fig. 2B, the protein and mRNA levels of Eya2 were significantly increased following transfection with the expression plasmid, and were decreased following transfection with siRNA.

The role of Eya2 in cell growth was examined by CCK-8 cell viability assay (Fig. 3A). The results revealed that transfection with the Eya2 expression plasmid increased the cell growth rate and the silencing of Eya2 inhibited cell proliferation. Colony formation assay revealed that transfection with the Eya2 expression plasmid significantly increased the colony number in the A172 cells. On the contrary, the knockdown of Eya2 in the U251 cells significantly decreased the colony number (Fig. 3B). Tran swell invasion assay was also performed to examine the role of Eya2 in the invasion of both the A172 and U251 cell lines. The results revealed that the knockdown of Eya2 decreased the number of invading U251 cells, and that Eya2 overexpression in the A172 cells facilitated cell invasion (Fig. 3C).

The above-mentioned results indicated that Eya2 promoted astrocytoma cell proliferation and invasion. Thus, we further examined the effect of Eya2 on cell cycle progression. As shown in Fig. 4A, Eya2 overexpression in the A172 cells facilitated the G1/S transition, while the knockdown of Eya2 in the U251 cells decreased the percentage of cells in the S phase. We further examined the levels of cell cycle-related proteins and found that Eya2 overexpression positively regulated cyclin D1 and cyclin E mRNA and protein expression (Fig. 4B). However, the knockdown of Eya2 in the U251 cell line downregulated the mRNA and protein expression of cyclin D1 and cyclin E. In addition, Eya2 overexpression upregulated the mRNA and protein levels of MMP9, a protein that is closely associated with cancer invasiveness (Fig. 4B).

To explore the potential mechanisms of the Eya2 regulation of cell invasion, we examined the changes in several signaling pathways previously reported to be involved in cancer invasion (17–21). We found that Eya2 overexpression upregulated ERK phosphorylation, while Eya2 knockdown downregulated ERK phosphorylation (Fig. 4B), suggesting that the ERK pathway may be responsible for the invasion-promoting effects of Eya2. To confirm the role of ERK as the mediator of invasiveness, the ERK inhibitor, PD98059 (20 µM, 6 h), was used to treat the A172 cells transfected with the Eya2 expression plasmid. As shown in Fig. 5A, treatment with the ERK inhibitor blocked the promoting effect of Eya2 on cell invasion. In addition, western blot analysis and PCR revealed that treatment with the ERK inhibitor abolished the effect of Eya2 on MMP9 upregulation in the A712 cell line at both the mRNA and protein level (Fig. 5B).

Eya2 has been previously regarded as a Six1 partner, which has been reported to be an important cancer-related protein (22,29). To confirm this, immunoprecipitation was performed. We immunoprecipitated Eya2 from U251 cell lysate and analyzed it for Six1 binding. As shown in Fig. 5C, Eya2 co-immunoprecipitated with Six1 in the U251 cells. We also examined the protein expression of Six1 in 3 astrocytoma cell lines. We found that Six1 expression was high in the U251 cells and low in the U87 and A172 cells, which closely correlated with the Eya2 levels (Fig. 2A). To validate the involvement of Six1 in the biological role of Eya2, we adopted Six1 siRNA to deplete its endogenous expression in the A172 cells (Fig. 5D). Western blot analysis and real-time PCR analysis demonstrated that the depletion of Six1 down-regulated MMP9 expression in the A172 cells. In addition, Eya2 overexpression failed to upregulated MMP9 mRNA and protein in the Six1-depleted cells, which indicated Six1 that is required for the biological effects of Eya2.