The immortalized hCMEC/D3 was purchased from the American Type Culture Collection (ATCC). The cells were cultured (37˚C; 5% CO₂) in high-glucose Dulbecco's modified Eagle's medium (DMEM; HyClone; Cytiva) with 10% fetal bovine serum (FBS; HyClone; Cytiva). The cells in the logarithmic growth stage were selected for use in the experiments. The cells subjected to OGD were slightly washed with sugar-free Earle's balanced salt solution (EBSS; MilliporeSigma) medium three times. The cells were then cultured with a sugar-free EBSS and under the conditions of oxygen deprivation (93% N₂, 5% CO₂) for 6, 12, 24 and 48 h, respectively. Normal control cells were cultured in a conventional CO₂ incubator. The transient transfection of pcDNA 3.1 vectors [pcDNA-empty vector (NC)/pcDNA-Meg3; final concentration, 100 nM], miRNA mimics (miR-122-5p mimic/NC mimic; final concentration, 50 nM), or short interfering (si)RNAs [Meg3-siRNA/NDRG family member 3 (NDRG3) NDRG3-siRNA/NC-siRNA; final concentration, 50 nM] was performed using Lipofectamine 2000^® (Invitrogen; Thermo Fisher Scientific, Inc.) when the cell-density reached 70%. Cell transfection was performed for a duration of 48 h at 37˚C. The cells were examined 48 h post-transfection. miR-122-5p mimic (cat. no. miR10000421-1-5) and NC mimic (cat. no. miR1N0000001-1-5) were purchased from Guangzhou RiboBio Co., Ltd. pcDNA-NC (NR_002766.2), pcDNA-Meg3 (NR_002766.2), Meg3-siRNA, NDRG3-siR NA and NC-siRNA were purchased from Shanghai GenePharma Co., Ltd. The sequence of miR-122-5p mimics was 5'-UGGAGUGUGACAAUGGUGUUUG-3'. The sequence of NC mimics was 5'-UACUGAGAGACAUAAGUUGGUC-3'.

The cell density was 0.2x10⁶/cm² for RNA extractions. TRIzol^® (Thermo Fisher Scientific, Inc.) was used to extract total RNA from the hCMEC/D3 cells. Extracted RNA was reverse transcribed into cDNA for miRNA and mRNA detection using the SuperScrip™ IV First-Strand Synthesis system (Invitrogen; Thermo Fisher Scientific, Inc.) and Mir-X miRNA First-Strand Synthesis kit (Takara Bio, Inc.) according to the manufacturer's protocols, respectively. The expression of miR-122-5p, Meg3 and NDRG3 was determined by RT-qPCR using a Mir-X miRNA qRT-PCR TB Green kit or TB Green Fast qPCR mix (Takara Bio, Inc.) according to the manufacturer's protocols, respectively. The sequences of primers were as follows: Meg3 forward, 5'-TCATCCGTCCAC CTCCTTGTCTTC-3' and reverse, 5'-GTCCTCTTCATCCTTTGCCATCCTG-3', miR-122-5p forward, 5'-TGGAGTGTGACAATGGTGT-3' and universal antisense from Mir-X miRNA qRT-PCR TB Green kit (no. 638314; Clontech; Takara Bio, Inc.), NDRG3 forward, 5'-CTGGTGGAAGGTCTTGT-3' and reverse, 5'-GCGCCCATTATAGGAACCCA-3', β-actin forward, 5'-CTCCATCGTCCACCGCAAATGCTTCT-3' and reverse, 5'-GCTCCAACCGACTGCTGTCACCTTC-3', U6, Mir-X miRNA from qRT-PCR TB Green kit (cat. no. 638314; Takara Bio, Inc.). RT-qPCR was performed with a program of 5 min at 95˚C and then 45 cycles at 95˚C (30 sec), 95˚C (5 sec), 55˚C (30 sec), and 72˚C (30 sec). The 2^-ΔΔCq method was used to calculate the relative expression (15). RT-qPCR reactions were performed in triplicates.

Total proteins were collected from hCMEC/D3 cell lysates using RIPA buffer (Cell Signaling Technology, Inc.). The protein concentration was determined using a BCA kit (Beyotime Institute of Biotechnology). Total protein (30 µg/sample) was separated via 10% SDS-PAGE and transferred to nitrocellulose membranes. The membranes were blocked with 5% skimmed milk powder overnight at 4˚C. The corresponding protein antibodies used were as follows: NDRG3 (cat. no. ab131266, 1:1,000), Bax (cat. no. ab32503, 1:1,000), Bcl-2 (cat. no. ab32124, 1:1,000), VEGFA (cat. no. ab46154, 1:1,000), VEGF receptor 2 (VEGFR2, cat. no. ab134191, 1:1,000) and β-actin (cat. no. ab8227, 1:2,000) (all from Abcam). The incubation conditions for the primary antibodies were 37˚C for 1 h. Subsequently, they were incubated with secondary horseradish peroxidase-conjugated goat anti-rabbit IgG antibody (cat. no. 1706515, Bio-Rad Laboratories, 1:2,000) for 30 min at 37˚C. The bands were visualized and quantitated using the ECL system (Thermo Fisher Scientific, Inc.). β-actin is used as an internal control. For densitometry analysis of Western blot data ImageJ 1.48 software (National Institutes of Health) was used.

StarBase (https://starbase.sysu.edu.cn/starbase2/index.php) was applied to predict the binding sites between the Meg3 3'-untranslated region (3'-UTR) and miR-122-5p. TargetScan (https://www.targetscan.org/vert_71/) was applied to identify the potential target genes of miR-122-5p. The analyzed data were downloaded from StarBase and TargetScan. A fragment of Meg3 or the NDRG3 3'-UTR, containing the predicted miR-122-5p binding site, was generated by PCR using specific primers. The resulting PCR amplicons were cloned into the pmiR-RB-REPORT TM luciferase reporter plasmid (Shanghai GeneChem Co., Ltd.). 293T cells were purchased from Procell Life Science & Technology Co., Ltd. (cat. no. CL-005). The 293T cells were seeded in 96-well plates at a density of 50-70% cells/well, and transfected with the Meg3-wt, Meg3-mut, NDRG3 3'UTR-wt/mut, together with miR-122-5p/NC mimic using the Lipofectamine 2000^® (Invitrogen; Thermo Fisher Scientific, Inc.). At 24 h following transfection, the luciferase activities were recorded with a plate-style luminometer using the Promega Dual-Luciferase system (Promega Corporation), with Renilla luciferase as normalization.

The cultures were supplemented with CCK-8 solution (Dojindo Laboratories, Inc.) and incubated for 1 h at 37˚C. Cell proliferation was measured at a wavelength of 450 nm using a microplate reader (Varioskan Flash; Thermo Fisher Scientific, Inc.).

A total of 150 µl dissolved Matrigel matrix adhesive was added to each pre-cooled 24-well plate and placed at 37˚C for 30 min and allowed to solidify. hCMEC/D3 cells (4x10⁵) were added to each well and cultured for 12 h in an incubator at 37˚C. Lumen formation was observed under a Leica DMI6000B light microscope (Leica Microsystems, Inc.), and five fields were randomly selected for each group to obtain images.

The migration of hCMEC/D3 cells was assessed using a wound healing assay. The cells were seeded into a 6-well plate with serum-free medium. Once the cells reached 90% confluency, a scratch was made across the surface of the well using a sterile 200 µl micropipette tip. The cells were cultured at 37˚C for 24 h, and the distance between the two edges was observed.

The apoptosis of the hCMEC/D3 cells was analyzed using Annexin V, according to the manufacturer's protocol. Briefly, the cells were washed with PBS (Invitrogen; Thermo Fisher Scientific, Inc.) and the cell concentration was adjusted to 1.0x10⁶ cells/ml. The cells were subsequently suspended in 150 µl buffering solution. Subsequently, the cells were stained with 10 µg/ml Annexin V-FITC (Roche Diagnostics GmbH) and 5 µl PI (Roche Diagnostics GmbH) at 4˚C for 20 min in the dark. Apoptotic cells were then analyzed using a BD FACSCelesta flow cytometer with FACSDiva software v8.0.1.1 (Becton-Dickinson and Company). Apoptotic rate (%)=percentage of early + percentage of late apoptotic cells.

The data are presented as the mean ± standard deviation. Statistical analysis was performed using SPSS software (version 19.0; IBM Corp.). One-way analysis of variance with Tukey's post hoc test and the unpaired Student's t-test were used for comparisons between groups. P<0.05 was considered to indicate a statistically significant difference.

First, the present study investigated the expression of lncRNA Meg3 in the OGD model using hCMEC/D3 cells. It was found that the expression of lncRNA Meg3 was increased in a time-dependent manner in response to hypoxia and OGD when compared with normal conditions (Fig. 1A). To verify the transfection efficiency, the expression of Meg3 in Meg3 siRNA (si-Meg3) or pcDNA-Meg3-transfected-hCMEC/D3 cells was detected. lncRNA Meg3 expression was successfully silenced or upregulated in OGD-exposed hCMEC/D3 cells, respectively (Fig. 1B). In addition, Meg3 siRNA decreased Meg3 expression, while pcDNA-Meg3 transfection increased Meg3 expression in OGD-exposed hCMEC/D3 cells (Fig. 1C). Moreover, the data demonstrated that hCMEC/D3 cell proliferation, migration and angiogenesis were markedly enhanced in the si-Meg3 + OGD-treated group compared with the group exposed to OGD only (Fig. 1D-H). Meg3 silencing inhibited the apoptosis of OGD-exposed hCMEC/D3 cells (Fig. 1I and J). By contrast, pcDNA-Meg3 transfection exerted the opposite effects (Fig. 1D-J). In addition, western blot analysis revealed an increase in Bax, and a decrease in Bcl-2, VEGFA and VEGFR2 expression in the OGD-exposed hCMEC/D3 cells compared with the control cells (Fig. 1K-L). However, these expression levels were significantly reversed after Meg3 silencing (Fig. 1K-L). Thus, the aforementioned data indicated that Meg3 knockdown promoted hCMEC/D3 cell proliferation, migration and angiogenesis, as well as inhibiting cell apoptosis in response to OGD-induced injury.

Subsequently, the present study aimed to identify the possible target miRNAs of Meg3. It was found that miR-122-5p expression was decreased in a time-dependent manner in the OGD model of hCMEC/D3 cells (Fig. 2A). Therefore, it was hypothesized that lncRNA Meg3 could interact with miR-122-5p. A Meg3-wt luciferase reporter vector and a Meg3-mut luciferase reporter vector with mutations on a predicted miR-122-5p binding site in Meg3 were constructed (Fig. 2B). A luciferase reporter assay was then conducted. miR-122-5p mimic markedly decreased the luciferase activity of 293T cells transfected with Meg3-wt; however, miR-122-5p mimic did not alter the luciferase activity of 293T cells transfected with Meg3-mut compared with the NC mimic group (Fig. 2C). Moreover, in miR-122-5p mimic-transfected hCMEC/D3 cells, the RNA level of Meg3 was significantly decreased (Fig. 2D). Furthermore, as shown in Fig. 2E, Meg3 knockdown promoted the miR-122-5p level in OGD-exposed hCMEC/D3 cells, and the overexpression of Meg3 exerted the opposite effect.

To further explore whether miR-122-5p mediates the regulatory effects of Meg3 on OGD-induced cell damage, the hCMEC/D3 cells were co-transfected with miR-122-5p mimic/anti-miR-122-5p and pcDNA-Meg3 (Meg3) prior to OGD. It was found that miR-122-5p overexpression restored hCMEC/D3 cell proliferation, migration and angiogenesis inhibited by pcDNA-Meg3, while miR-122-5p knockdown exerted the opposite effects (Fig. 3A-D). In addition, compared with the control group, the OGD-exposed hCMEC/D3 cells exhibited a marked induction in cell apoptosis (Fig. 3F-G). However, this induction in apoptosis was abolished after miR-122-5p silencing (Fig. 3F-G). Western blot analysis of the Bax, Bcl-2, VEGFA and VEGFR2 expression levels revealed consistent changes with the expression levels in hCMEC/D3 cells following the OGD challenge and transfection with pcDNA-Meg3 + miR-122-5p mimic or pcDNA-Meg3 + anti- miR-122-5p (Fig. 3H and I)

Subsequently, online database searches were performed to predict the downstream target genes of miR-122-5p. Notably, NDRG3 was found to be a potential target gene of miR-122-5p (Fig. 4A). To investigate whether miR-122-5p directly targets the 3'UTR of NDRG3, luciferase reporter assays were performed. The data indicated that miR-122-5p mimic markedly reduced the luciferase activity of 293T cells transfected with NDRG3 3'UTR-wt and miR-122-5p mimic (Fig. 4B). Moreover, miR-122-5p mimic markedly suppressed NDRG3 expression (Fig. 4C and D). Furthermore, Meg3 knockdown inhibited NDRG3 expression in OGD-exposed hCMEC/D3 cells, and Meg3 overexpression exerted the opposite effects (Fig. 4E and F). Moreover, the increase in NDRG3 expression at both the mRNA and protein level induced by Meg3 overexpression was significantly attenuated by miR-122-5p mimic and enhanced by a miR-122-5p inhibitor (Fig. 4G and H). Overall, these results suggested that NDRG3 was a target gene of miR-122-5p, and that Meg3 promoted NDRG3 expression by sponging miR-122-5p.

To confirm that NDRG3 is a functional target gene of lncRNA Meg3, efficient siRNA was designed to knock down NDRG3 expression at both the RNA and protein level (Fig. 5A and B). Subsequently, knockdown experiments were performed by silencing NDRG3 expression in hCMEC/D3 cells transfected with pcDNA-Meg3 (Meg3). The exacerbating effect of Meg3 overexpression on OGD-induced injury was abolished by NDRG3 knockdown, including an increase in the proliferation, migration and angiogenesis, as well as a decrease in the apoptosis of hCMEC/D3 cells subjected to OGD (Fig. 5C-K).