The schwannoma RT4-D6P2T cells (ATCC, Manassas, VA, USA) were cultured in RMPI-1640 (HyClone, Hudson, NH, USA) medium with 10% fetal bovine serum (Gibco, Carlsbad, CA, USA) at 37°C, 95% air and 5% CO₂ incubator (Thermo Fisher Scientific, Waltham, MA, USA), or at 37°C, 94% N₂, 5% CO₂ and 1% O₂ tri-gas incubator (Thermo Fisher Scientific) for hypoxic culture, and were subcultured every 2–3 days. HIF-1α-shRNA-pRNAT-U6.1/Neo interference plasmids (Auragene, Changsha, China), or miR-210 complementary Locked Nucleic Acid (anti-miR-210) (GeneCopopia, Rockville, MD, USA) and lipo2000 (Invitrogen Life Technologies, Carlsbad, CA, USA) mixed liquor was added into a petri dish with 30–50% cells (Nest, Wuxi, China) for 4–6 h, and then the culture medium was removed and replaced by fresh culture medium. After the replacement, the resulting solution was put in a 37°C CO₂ incubator for 24–72 h culture and then the cells were collected for further study.

Total RNAs were extracted from the cells using TRIzol (Invitrogen Life Technologies) reagent, and prepared with a RevertAid First Strand cDNA Synthesis kit (Fermentas Inc., Ontario, Canada). SYBR green qPCR assay was carried out to find the relative expression levels of miR-210, compared with the internal control gene U6, and calculated using the 2^−∆∆Ct method. The rat-miR-210 primers (RmiRQP3171), U6 primers (RmiRQP9003) and qPCR Mix were purchased from GeneCopoeia.

Cells (10⁷) were collected and put into 200 µl RIPA (Invitrogen Life Technologies) lysate for extracting total proteins, of which 50 µg were used for western blot analysis. The EFNA3 and β-actin antibodies were from Sigma (St. Louis, MO, USA), HIF-1α, P62 and elf4E antibodies were from Abcam (Cambridge, MA, USA), VEGF antibodies were from Immunoway (Suzhou, Jiangsu, China).

Cell cycle: the cells were treated for 24 h, and then Trypsins (Auragene) were used to dissociate the cells, and PBS was used to wash and resuspend the cells. Ethyl alcohol (70%) was added into the resuspension, then the resuspension was cultured at 4°C for 18 h. Cells were again washed with PBS and the washed cells were resuspended in Staining Solution (50 µg/ml of PI, 1 mg/ml of RNase A, 0.1% Triton X-100 in PBS). The stained cells (1×10⁶) were then analyzed with a flow cytometer (Beckman Coulter, Brea, CA, USA). Cell apoptosis: cells (5×10⁵) were resuspended in Annexin V-FITC and propidium iodide (PI) labeling solution, and then Annexin V Apoptosis Detection kit FITC (eBioscience, Affymetrix, San Diego, CA, USA) was used for immediate testing. Cell migration: Cellular invasion ability was tested by Transwell assay. The cells were cultured under normoxic/hypoxic for 24 h, and then were dissociated and inoculated in serum-free medium in the upper chamber with well placed Matrigel for another 24 h in normoxic/hypoxic conditions. The remaining cells in the upper chamber were carefully removed, and the cells that entered into the Matrigel were stained by the staining solution with 75% ethyl alcohol and crystal violet (Beyotime Biotechnology, Jiangsu, China) for 25 min. Dried at room temperature, and then were photographed with an inverted microscope (Motic, Xiamen, China). Ultrastructural observation: 24 h after the treatment, the cells were placed, respectively, in normoxic and hypoxic conditions for another 24 h culture, dissociated with Trypsin-EDTA (Auragene), washed twice with PBS, immobilized with 2.5% glutaraldehyde precooled at 4°C overnight, post-immobilized with 2% osmic acid, dehydrated with ethanol-acetone, embedded with epoxy resin, sliced, and then observed with a transmission electron microscope (Hitachi, Tokyo, Japan) under 80 kV.

The schwannoma tissue microarray (SO808) was purchased from Alenabio (Shangxi, China), and hematoxylin and eosin (H&E) staining procedures were as previously described (28). Immunology and Histology Chemistry (IHC) test: after being dewaxed, hydrated, antigen retrieved and H₂O₂ confined, the tissue chip was exposed to rabbit anti-VEGF polyclonal antibodies (1:100, Immunoway), incubated overnight at 4°C, then exposed to goat anti-rabbit secondary antibodies (1:800, Jackson Immuno Research Inc., West Grove, PA, USA), incubated at room temperature for 30 min, and the VEGF protein expression was observed with an optical microscope (Upototech, Changchun, China). In Situ Hybridization (ISH) test: Rat-miR-210 probe sequence: 5′-ACAGATCAGCCGCTGTCACACGCAC-3′, synthesized at BGI Tech (Shenzhen, China). Hybridization probe mixed solution (1:500) was instilled into the chips, using the IsHyb In Situ Hybridization kit (Biochain, Newark, CA, USA) according to the instructions, then observed and photographed the results with an optical microscope (Upototech). Immunofluorescent assay (IF) test: the cells were inoculated in a 6-well dish (the density <80%), after being immobilized, the anti-LC3B II/I antibodies (ab48394) by Abcam were used for IF test, with the primary antibodies diluted at a ratio of 1:1000, and the green fluorescent protein expression was observed with an inverted microscope (Motic).

BSP primers were designed according to the predicted CpG island in Rat-miR-210 promoter region: forward 5′-GGAAGGATATGTTTTGGATTGTATTAA-3′, reverse 5′-TAAAACCCACCCTACAAAACTACAAC-3′, the PCR products were cloned into pGEM-T Easy vector, 15–20 clones were randomly selected for sequencing, and 10 correct clones were selected for methylation analysis.

The data are shown as the mean ± SE. Statistical analysis were performed using SPSS 11.0 software, Student's t-test was used to analyze the differences between two groups, with 95% confidence interval. Pearson's Chi-square test was used to analyze the relationships of microvascular density (MVD) vs. VEGF and MVD vs. miR-210, with 99% confidence interval. P<0.05 was considered to indicate a statistically significant difference; and P<0.01, highly significant.

As hypoxia is one of the key factors stimulating tumor vasculogenesis, we stained central neurilemmoma tissue microarray (with 30 cases of schwannoma tissues from different sources, 2 spots for each case) by H&E, scored microvascular density (MVD), and then tested the expression of VEGF and miR-210 using the same chips by IHC and ISH, respectively (Fig. 1A). The bar chart of Pearson's Chi-squared test for MVD/VEGF and MVD/miR-210 relevance is shown in Fig. 1B, and the scores and P-values as shown in Table I. VEGF positive stain became darker and miR-210 expression level increased with the tumor tissue MVD elevation, the MVD vs VEGF and MVD vs. miR-210 relevant levels had statistical significance at the level of confidence for interval of 99% (P=0.005 <0.01, P=0.025 <0.05).

In order to explore the functions of miR-210 in schwannoma cells, we used RT4-D6P2T cells as a model for related research. First, to find whether hypoxia could affect miR-210/EFNA3 expression, the expression levels of miR-210 and EFNA3 were detected under hypoxia for 0, 12, 24 and 48 h in RT4-D6P2T cells. As shown in Fig. 2, miR-210 expression increased slightly after 12 h, increased by 4–6 times after 24 h, and further increased after 48 h, but not as much as 24 h. While the expression trend of EFNA3 was opposite to that of miR-210. The expression of EFNA3 significantly decreased after 12 h hypoxic culture, and could hardly be detected after 24 h or 48 h. This indicated that hypoxia induced the expression of miR-210 in RT4-D6P2T. EFNA3, one of the target genes of miR-210, is still negatively regulated by miR-210 during hypoxic culture and shows a reverse expression trend to miR-210. This result was consistent with previous results (21). Thus, miR-210/EFNA 3 may be involved in schwannoma cell hypoxia. As expression of both miR-210 and EFNA3 changed significantly after 24 h in hypoxia, so we chose the hypoxic of 5% O₂ for 24 h as the following experimental conditions.

As hypoxia changes miR-210/EFNA3 expression in schwannoma RT4-D6P2T cell line, we wanted to explore the effects of hypoxia on the functions of such cells, to find which function exerted hypoxia-induced miR-210/EFNA3 expression. We tested cell cycle, apoptosis and autophagy of the cells after 24 h hypoxic culture. Apoptosis increased greatly from ~3% under normoxic conditions to ~20% (Fig. 3A). The cell cycle arrested in G1 phase under hypoxic conditions (Fig. 3B).

Considering hypoxia could induce cell autophagy, we measured the cell autophagy under hypoxic culture by IF and western blotting. According to the results shown in Fig. 3C, during hypoxic culture, cellular morphology changed greatly, cells became larger and round in shape, which was associated with the cellular edema caused by too much water in cells due to hypoxia-induced sodium potassium pump function decrease. Besides, the IF staining results showed that the expression level of LC3BII increased, and the fluorescent signal around the cell nuclei significantly increased (Fig. 3C). According to Transwell experiment, although the quantity of the cells entering into Matrigel under the normoxic conditions was not significantly different with that in hypoxia, the cell morphology changed significantly, and under the hypoxic conditions, intercellular interval was wider, cells became narrower and fusiform shape (Fig. 3D). This may contribute to hypoxia-induced epithelial to mesenchymal transition phenomenon. According to the cell ultrastructure image by a transmission electron microscope (Fig. 3E), we found that during hypoxia, double membrane vacuole structures occurred in cytoplasm, and apoptotic bodies in cytoplasm significantly increased. Similarly, the results of western blotting on the autophagy related protein P62 and elf4E also showed that, after hypoxia P62 protein expression significantly decreased, while the expression level of the autophagy effector elf4E increased (Fig. 3F). Therefore, we concluded that hypoxia enhanced apoptosis and autophagy. In addition, according to Fig. 3F, HIF-1α and VEGF expression increased after 24 h hypoxia; and considering miR-210 expression increased and EFNA3 expression decreased during hypoxia together (Fig. 2A), we speculated that during hypoxia, the miR-210 enhancement of schwannoma cells was induced by HIF-1α, and miR-210 regulated vascularization by targeting EFNA3.

To investigate whether miR-210 upregulation during hypoxia is induced by HIF-1α, we carried out HIF-1α silencing by HIF-1α-shRNA, and then placed the HIF-1α loss-of-funciton cells under normoxic/hypoxic conditions, respectively for 24 h. Hypoxia can induce miR-210 upregulation, but after HIF-1α interference, miR-210 expression level decreased significantly (Fig. 4A). This indicated that miR-210 expression did relate to HIF-1α expression, and HIF-1α may positively regulate miR-210 expression.

We further studied the mechanism of how hypoxia induced miR-210 upregulation. We predicted that there were abundant CpG islands in the promoter region of rat miR-210 (Fig. 4B) by the online software Lilab (http://www.urogene.org/cgi-bin/methprimer/methprimer.cgi), CpG Island Searcher (http://cpgislands.usc.edu/), and EMBOSS explorer (http://emboss.bioinformatics.nl/cgi-bin/emboss/cpgplot). Moreover, the early methylation-specific PCR (MSP) preliminary experiment results showed that under hypoxic conditions, the miR-210 promoter region methylation level in cells decreased significantly compared with that under normoxic conditions (data not shown). We speculated that during hypoxia, demethylation occurred in the miR-210 promoter region, which enabled more hypoxia-induced HIF-1α to combine with HRE in the miR-210 promoter region and activated miR-210 translation. So, we first tested the methylation islands of the miR-210 promoter region by Bisulfite sequencing PCR (BSP) under hypoxia, and then assessed whether HRE combined with HIF-1α was included in these demethylated CpG islands. The results showed that there were 17 CpG locuses in the upstream −252 to −554 bp region of the miR-210 coding region, and under the normoxic conditions, the methylation level of this CpG islands segment was ~72.35%, while after 24 h hypoxic culture, the methylation level dcreased significantly to ~44.70% (Fig. 4C).

In order to clarify the function of miR-210 on schwannoma cells during hypoxia, we built miR-210 knockdown cell model and then placed the model cells into normoxic/hypoxic conditions for 24 h, respectively. As shown in Fig. 5A, compared with the negative control group (anti-scramble), the proportion of the cells in the S+G2 phase significantly reduced in anti-miR-210 group, but the quantity difference of the S+G2 phase between the normoxia and the hypoxia group does not significant. The cell apoptosis rates increased significantly in anti-miR-210 group in both the normoxia and the hypoxia group (Fig. 5B). LC3B II fluorescence intensities of anti-miR-210 group are lower than those of the anti-scramble group, and higher (under hypoxic conditions) than those of the cells cultured under normoxic conditions (Fig. 5C).

According to the transmission electron microscope scanning results (Fig. 5E), under normoxic conditions, cellular matrix density was uniform and with few autolysosomes; after hypoxic culture, double membrane vacuole structures occurred in cellular matrix, mitochondria breakage was more serious, some damaged mitochondria were covered by dual-membranes, and autophagy-induced cavitation in cells of the miR-210 group were much less than those of the anti-scramble group. This indicated that miR-210 could mediate cell autophagy. Western blot results on the autophagy related factors (P62 and elf4E) were consistent with the aforementioned results (Fig. 5F). The expression level of HIF-1α and VEGF protein also decreased when the miR-210 was knocked down, while EFNA3 expression increased (Fig. 5F). This indicated that miR-210/EFNA3 was related with the HIF-1α/VEGF-mediated neovascularization paths.