Serum samples were obtained from 15 patients with cerebral ischemia who were also diagnosed with ischemic stroke by MRI, as well as 15 healthy participants at the Workers' Hospital of Tangshan City (Hebei, China). All experimental protocols were approved by the Ethics Committee of the Workers' Hospital of Tangshan City. Written informed consent was obtained from all patients. All samples were flash-frozen in liquid nitrogen, and stored at −80°C until further molecular analysis. The demographics and clinical characteristics of the 15 cerebral ischemic patients and 15 healthy controls are provided in Table I.

Total RNA was isolated from the sera of patients with cerebral ischemia by a miRNAeasy mini kit (Qiagen, Inc., Valencia, CA, USA). The purity and quantity of total RNA were evaluated by NanoDrop ND-1000 Spectrophotometry (Thermo Fisher Scientific, Inc., Waltham, MA, USA) and Agilent's 2100 Bioanalyzer (Agilent Technologies, Inc., Santa Clara, CA, USA). Total RNA (200 ng) was labeled and hybridized with the miRCURY™ LNA Array (version 16.0; Exiqon; Qiagen, Inc.). Following washing, Axon GenePix 4000B microarray scanner (Axon Instruments; Molecular Devices, LLC, Sunnyvale, CA, USA) was used to scan the fluorescence intensity of the microarray. Scanned images were then imported into the GenePix Pro6.0 program (Axon Instruments; Molecular Devices, LLC) for grid alignment and data extraction. Finally, the heat map of the 57 miRNAs with the most evident differences was created using a method of hierarchical clustering with GeneSpring GX, version 7.3 (Agilent Technologies, Inc.).

HUVECs were obtained from the cell bank of the Chinese Academy of Sciences (Shanghai, China), and maintained in M199 medium supplemented with 20 mg/ml endothelial cell growth supplement (Upstate Biotechnology, Inc., Lake Placid, NY, USA) and 10% fetal bovine serum (FBS; HyClone; GE Healthcare Life Sciences, Logan, UT, USA) at 37°C in a humidified incubator with an atmosphere of 95% air and 5% CO₂ (normoxic conditions). For hypoxia, HUVECs were cultured in a hypoxia incubator (Sanyo Electric Co., Ltd., Osaka, Japan) under hypoxic conditions (5% CO₂, 94% N₂ and 1% O₂,) for 6 h at 37°C. Cells cultured under normoxic conditions were used as controls.

Total RNA was isolated from serum samples and HUVECs cells using TRIzol reagents (Invitrogen; Thermo Fisher Scientific, Inc.) following the manufacturer's instructions and reverse transcription was performed using PrimeScript™ RT reagent kit (Takara Biotechnology Co., Ltd., Dalian, China) at 42°C for 30 min and 85°C for 5 sec. For the detection of miRNA, RT-qPCR assays were performed using the TaqMan miRNA Assay (Thermo Fisher Scientific, Inc.) following the manufacturer's instructions. For detection of the mRNA levels, qPCR was performed on an ABI PRISM 7300 sequence detection system in an SYBR Green I Real-Time PCR kit (Applied Biosystems; Thermo Fisher Scientific, Inc.). The RT-qPCR reaction system (30 µl) contained 5 µl cDNA, 15 µl 2X qPCR mix, 1 µl upstream primer, 1 µl downstream primer and 8 µl double distilled H₂O. The PCR protocol was: 95°C for 15 min, followed by 40 cycles of 94°C for 15 sec, 55°C for 30 sec and 70°C for 30 sec and a final extension step at 72°C for 5 min. U6 and GAPDH functioned as the normalization controls in the expression analysis of miR-451 and MIF, respectively. The relative expression of RNAs was calculated using the 2^−ΔΔCq method (17). Each reaction was conducted in triplicate. The primers utilized for RT-qPCR analysis were as follows: miR-451 forward, 5′-AAAGTCGACAAGCTCTCTGCTCAGCCTGTC-3′ and reverse, 5′-AAAATATCTCGAGCCCCCACCCCTGCCTTGT-3′; U6 forward, 5′-TGCGGGTGCTCGCTTCGCAGC-3′ and reverse, 5′-CCAGTGCAGGGTCCGAGGT-3′; MIF forward, 5′-GGCCTCACTTACCTGCACC-3′ and reverse, 5′-AACCATTTATTTCTCCCGACC-3′; GAPDH forward, 5′-GCAACTCCCACTCTTCCACC-3′ and reverse, 5′-GTCATACCAGGAAATGAGCTTGACA-3′. The RT-qPCR assays were performed in triplicate and the change in expression level was calculated using the 2^−ΔΔCq method.

HUVECs were seeded at a density of 6×10⁴/cm² in 6-well plates and allowed to settle for 24 h to ensure that 40–50% confluence was achieved prior to transfection. miR-451 mimic (5′-AAACCGUUACCAUUACUGAGUU-3′) and its negative control (NC; 5′-UCGCUUGGUGCAGGUCGGGAA-3′), miR-451 inhibitor (5′-AACUCAGUAAUGGUAACGGUUU-3′) and negative control (NC; 5′-CAGUACUUUUGUGUAGUACAA-3′) were synthesized by Shanghai GenePharma Co., Ltd. (Shanghai, China). The MIF small interfering RNA (siRNA, (5′-ACACCAACGUGCCCCGCGCdTdT-3′) and NC siRNA (5′-GCGCGGGGCACGUUGGUGUdTdT-3′) were purchased from Santa Cruz Biotechnology, Inc. (Dallas, CA, USA). In addition, to enhance the expression of MIF, the coding domain sequences of MIF mRNA were amplified by PCR, and inserted into pcDNA 3.0 vector (Invitrogen; Thermo Fisher Scientific, Inc.), and named pcDNA-MIF. Cells were cultured to 80% confluence, followed by transfection with miR-451 mimics (50 nM), control mimics, miR-451 inhibitor (50 nM) or control inhibitor using Lipofectamine 2000^® (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. MIF expressing vector (2 µg) was transfected into HUVECs with Lipofectamine^® 2000 following manufacturer's protocol. After 24 h, the transfected cells were subjected to hypoxic conditions for 6 h. The transfection efficiency was determined by RT-qPCR.

Confluent HUVEC monolayers were 0.25% trypsinized and plated onto 24-well plates that were coated with Matrigel (Becton, Dickinson and Company, Franklin, Lakes, NJ, USA) and incubated in M199 medium for 24 h at 37°C. Then, the transfected cells were subjected to hypoxic conditions for 6 h as aforementioned and tube-like structures were observed under a routine light microscope and images captured from five randomly selected microscopic fields. Tube formations were calculated by counting the number of branches, which was conducted using ImageJ software (Version 1.46, National Institutes of Health, Bethesda, MD, USA).

Invasion activities of HUVECs were analyzed using Boyden chambers with 8-mm pore membranes coated with Matrigel (Becton, Dickinson and Company) following manufacturer's protocol. Briefly, following 24 h transfection, HUVECs (1×10⁴ per well) were seeded into the upper chambers in 200 µl serum-free M199 medium, and the lower chambers were filled with 500 µl of M199 medium containing 10% FBS. Following exposure to hypoxic or normoxic conditions for 6 h, cells on the surface of the filter were fixed with 4% formaldehyde, stained with 0.5% crystal violet at 37°C for 30 min, and counted under a routine light microscope (Phenix Optical Instrument Group Company, Jiangxi, China). Cells were counted in five randomly chosen microscopic fields.

HUVECs (1×10⁶ per well) were plated onto 6-well plates and incubated in M199 for 24 h at 37°C. Following 24 h transfection, the cells were scrapped with a 10 µl pipette tip, fresh serum-free medium was added and then cells were exposed to hypoxic or normoxic conditions for 6 h. Initial images were acquired as a reference and, following 6 h secondary images were taken corresponding to the photographed region capture initially. Wound healing was evaluated by measuring the distance of the wounded region with an absence of cells using ImageJ software (Version 1.46, National Institutes of Health, Bethesda, MD, USA).

TargetScan (version 7.0; www.targetscan.org/) and PicTar (version 2006; http://pictar.mdc-berlin.de) target gene prediction software were used to select MIF as a target gene of miR-451.

The 3′-untranslated region (UTR) of MIF and the mutated sequence were inserted into the pGL3 control vector (Promega Corporation, Madison, WI, USA) to construct the wild-type (wt) MIF-3′-UTR vector and mutant MIF-3′-UTR vector, respectively. For the luciferase reporter assay, HUVECs were transfected with the corresponding vectors using Lipofectamine 2000^® (Invitrogen; Thermo Fisher Scientific, Inc.); a 48 h post-transfection, the dual-luciferase reporter assay system (Promega Corporation) was used to measure luciferase activity. To correct for differences in transfection and harvesting efficiencies, Renilla luciferase activity was used to normalize the firefly luciferase activity. All experiments were performed in triplicate.

Total protein was extracted from HUVECs cells using radioimmunoprecipitation lysis buffer (Beyotime Institute of Biotechnology, Shanghai, China). Concentrations of total cellular protein were determined using a BCA assay kit (Pierce; Thermo Fisher Scientific, Inc.). Total protein samples (40 µg) were analyzed by 8% SDS-PAGE and transferred to polyvinylidene difluoride membranes (GE Healthcare, Chicago, IL, USA) by electroblotting. Membranes were blocked with 5% nonfat milk at room temperature for 1 h, followed by incubation overnight at 4°C with primary antibodies. Primary antibodies against MIF (cat. no. sc-130329; Santa Cruz Biotechnology, Inc., Danvers, MA, USA; 1:1,000 dilution), phospho (p)-VEGF Receptor 2 (cat. no. 2478; VEGFR2; Tyr1175; Cell Signaling Technology, Inc.; 1:1,000 dilution), VEGF (cat. no. 2463; Cell Signaling Technology, Inc.; 1:1,000 dilution) and total VEGFR2 (cat. no. 9698; Cell Signaling Technology, Inc.; 1:1,000 dilution) and β-actin (cat. no. sc-58673; Santa Cruz Biotechnology, Inc.; 1:2,000 dilution) were incubated with the membrane at 4°C overnight. Following incubation with anti-rabbit IgG (H+L; DyLight™ 680 Conjugate; cat. no. 5366; Cell Signaling Technology, Inc.; 1:10,000 dilution), bands were detected using an enhanced chemiluminescence kit (GE Healthcare). The intensity of the bands of interest was analyzed by ImageJ software (version 1.46; National Institutes of Health).

Statistical analysis was performed using SPSS software (version 18.0; IBM Corp., Chicago, IL, USA). Data were presented as the mean ± standard deviation. Student's t-test or one-way analysis of variance followed by Tukey post hoc test was used to analyze the difference among/between sample groups. P<0.05 was considered to indicate a statistically significant difference.

To explore the potential involvement of miRNAs in CII, the present study performed miRNA microarray profiling in serum samples from patients with cerebral ischemia. The miRNA microarray identified 35 miRNAs that were upregulated and 22 miRNAs that were downregulated in the ischemia group when compared with the control group (Fig. 1A). Among them, miR-451 was the most significantly downregulated, and previous studies have revealed that miR-451 functions as a suppressor of angiogenesis in hepatocellular carcinoma (18) and human osteosarcoma (19), but little is known regarding its role in the regulation of angiogenesis in CII. Therefore, the present study decided to focus on miR-451 in CII for further study.

To confirm the microarray findings, the miR-451 expression levels were determined by RT-qPCR in all samples. miR-451 was significantly downregulated in the serum samples from patients with cerebral ischemic compared with the normal participants (Fig. 1B), which is consistent with the results observed following microarray analysis. All of these results suggest that the alterations in miR-451 expression may serve important roles in CII.

To assess the biological role of miR-451 in angiogenesis, the present study applied HUVECs under hypoxic conditions to mimic ischemia in vitro (20,21) and the expression of miR-451 was determined by RT-qPCR. As shown in Fig. 2A, hypoxia treatment induced a significant decrease in the expression of miR-451 in HUVECs, which reached a peak at 6 h and then raised to near basal levels by 12 h. Therefore, the present study selected the 6 h time point as the subsequent experimental condition to study the influence of miR-451 on angiogenesis. As angiogenesis is defined as new microvessel formation via branching off from pre-existing vessels, which involves multi-step biological processes, including proliferation, migration and the formation of tube-like vascular structures (22–24), capillary-like tube formation and migration assays are normally used to evaluate angiogenesis in vitro (25). To further investigate the role of miR-451 in tube formation and migration, HUVECs were transfected with miR-451 mimics or mimics NC, followed by normoxic or hypoxia treatment. It was observed that the expression of miR-451 was enhanced following miR-451 mimics transfection under normoxia or hypoxia conditions (Fig. 2B). Furthermore, the tube formation assays demonstrated that overexpression of miR-451 significantly reduced the ability of HUVECs to form tubular structures during hypoxia (Fig. 2C). The Transwell assay revealed that the number of invaded cells was significantly reduced in miR-451 mimics group when compared with mimics NC group (Fig. 2D). In addition, the results also revealed that miR-451 mimics decreased the cell migration distance when compared with the mimics NC group (Fig. 2E).

Given the importance of the VEGF/VEGFR axis in angiogenesis following CII (26–29), the present study sought to determine whether miR-451 affects this axis in HUVECs. The results of western blotting revealed that the expressions of VEGF and p-VEGFR2 were markedly increased in HUVECs under hypoxic conditions, whereas this promotional effect was attenuated following treatment with miR-451 mimics (Fig. 2F), which indicated that the overexpression of miR-451 could inhibit the VEGF/VEGFR2 pathway.

Next, to assess the effect of miR-145 inhibition on HUVECs migration and tubulogenesis under normoxic conditions, HUVECs were transfected with miR-451 inhibitor to decrease miR-451 expression and then under normoxic conditions for 6 h. Following miR-451 inhibitor transfection, miR-451 expression was markedly decreased when compared with inhibitor NC (Fig. 3A). Then, the angiogenic properties of HUVECs were detected using tube formation, Transwell and wound healing assays. The tube formation assays revealed that knockdown of miR-451 promoted the ability of HUVECs to form tubular structures during normoxia (Fig. 3B). The Transwell assay revealed that the number of invading cells was significantly increased in the miR-451 inhibitor group when compared with the inhibitor NC group (Fig. 3C). In addition, miR-451 inhibitor also increased the cell migration distance when compared with the inhibitor NC group (Fig. 3D). All of these results indicated that low level miR-451 is beneficial to angiogenesis in HUVECs.

To explore the molecular mechanisms by which miR-451 regulates the tube formation, invasion and migration of HUVECs, candidate target genes of miR-451 were computationally screened using TargetScan and PicTar algorithms. Among several predicted target genes, MIF, which has been reported to improve angiogenesis, was of interest due to its high scores in the two algorithms. As shown in Fig. 4A, miR-451 contained a sequence that was complementary to MIF. In addition, previous research has shown that MIF was a direct target of miR-451 in human osteosarcoma (30). However, the association between miR-451 and MIF in CII remains unclear. Firstly, the present study determined the expression of MIF in HUVECs under hypoxic conditions by RT-qPCR. As shown in Fig. 4B, hypoxia treatment caused a significant increase in the expression of MIF in HUVECs, which reached a peak at 6 h and declined to near basal levels by 12 h. The expression levels of MIF in all serum samples were also determined and it was revealed that the expression of MIF was significantly increased in the ischemia group when compared with the normal group (Fig. 4C). To further confirm that MIF was negatively regulated by miR-451, the present study performed western blot analysis to determine the protein level of MIF. The expression of MIF at the protein level was significantly downregulated following the overexpression of miR-451 in HUVECs cells under hypoxic conditions, but upregulated following knockdown of miR-451 in HUVECs cells under normoxic conditions (Fig. 4D and E). To verify whether MIF is a direct target of miR-451, the 3′-UTR of MIF containing the WT or Mut miR-451 target sequences was cloned into the pmirGLO vector. Co-transfection was conducted with these reporter plasmids and miR-451 mimics, mimics NC, miR-451 inhibitor or inhibitor-NC into HUVECs, then luciferase activities were analyzed 48 h post-transfection. Luciferase reporter gene assays demonstrated that overexpression of miR-451 markedly repressed, while knockdown of miR-451 increased, the relative luciferase activity of constructs containing the WT MIF 3′-UTR. However, the luciferase activity of the reporter containing the mutant binding site was not altered (Fig. 4F). These data indicated that miR-451 regulates the expression of MIF in HUVECs.

To investigate whether MIF was involved in the anti-angiogenesis effects of miR-451, the present study transfected pcDNA-MIF plasmids into HUVECs cells, followed by exposure to hypoxic conditions. Western blotting was conducted to examine whether MIF protein expression was effectively enhanced. As shown in Fig. 5A, pcDNA-MIF transfection significantly overexpressed MIF protein expression in HUVECs. In addition, miR-451 mimics inhibited tube formation capacity, invasive ability and migration distance, and its inhibitory effect was attenuated when MIF was overexpressed (Fig. 5B-D). To inhibit MIF protein expression, the present study transfected MIF siRNA into HUVECs, followed by exposure to normoxic conditions. Western blotting was also performed to examine whether MIF protein expression was effectively altered. As shown in Fig. 5E, MIF siRNA transfection significantly inhibited MIF protein expression in HUVECs. Furthermore, the miR-451 inhibitor promoted tube formation capacity, invasive ability and migration distance, and its promotional effect was reduced when MIF was knocked down (Fig. 5F-H). These data suggested that the miR-451/MIF axis may serve an important role in the regulation of angiogenesis in HUVECs.