1 INTRODUCTION

Ischemic stroke, characterized by cerebral infarction-induced neurological impairment, represents a major cause of death and disability worldwide.1 The subsequent nutrient insufficiency induces an array of pathophysiological conditions, such as energy failure, oxidative stress, inflammation, and apoptosis, eventually leading to neuronal cell loss.2 Chronic or mild cerebral ischemia might be nonlethal and characterized by mild somatic symptoms; however, acute stroke can result in hemiplegia, cardiac dysfunction, and even death.3 Although continuous cell death is a widely accepted attribute of ischemic stroke,4 another type of cell death, autophagy, is also closely associated with the pathogenesis of ischemic stroke.5, 6 Identifying molecules responsible for cell death and autophagy in ischemic stroke might be helpful to develop novel therapeutic options for ischemic stroke.

MicroRNAs (miRNAs) are a representative class of noncoding RNA molecules that mediate functional alterations before, during, and after ischemic stroke.7 miRNAs play versatile functions mainly by binding to the sequences in mRNAs to suppress target gene expression.8 Dysregulation of miRNAs is frequently observed in neuronal diseases.9 Approximately 70% of all discovered miRNAs are abundant or specifically expressed in the brain and exert crucial functions in the development of the nervous system.10 miR-379-5p has been found to be poorly expressed in the plasma of patients with acute myocardial infarction and its overexpression suppressed proliferation of vascular smooth muscle cells,11 which is a critical event during the pathogenesis of cardiovascular diseases as well as cerebral ischemia.12, 13 In addition, miR-379-5p alleviated neurological dysfunction following spinal cord injury.14 These reports led us to examine its role in cerebral ischemic stroke as a typical neurovascular injury condition. In this paper, the integrated bioinformatics analyses suggested mitogen-activated protein kinase kinase kinase 2 (MAP3K2) as a target transcript of miR-379-5p. There are four major members of the mitogen-activated protein kinase (MAPK) family: the c-Jun N-terminal kinases (JNK), the extracellular signal-regulated kinases-1 and -2 (ERK1/2), p38 MAPKs, and extracellular signal-regulated kinase-5.15 MAP3K2 has been reported to trigger JNK phosphorylation.16 JNK is one of the representative MAPKs that have been reported to aggravate stroke injury by activating pro-apoptotic and pro-inflammatory cellular signaling.17 In addition, activation of the JNK/c-Jun pathway has been demonstrated to promote cell autophagy in several pathological conditions.18, 19 We therefore hypothesized that miR-379-5p may suppress MAP3K2 expression and the subsequent JNK/c-Jun activation, therefore reducing cell autophagy and apoptosis and alleviating stroke-induced injury.

2 MATERIALS AND METHODS 2.1 Ethics statement

This research was ratified by the Ethical Committee of Shanghai University of Medicine and Health Sciences Affiliated Zhoupu Hospital (approval no. 2019-A-001-E014) and adhered to the tenets of the Declaration of Helsinki. Signed informed consent was received from all subjects. All animal protocols were strictly according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The Animal Ethics Committee of Shanghai University of Medicine and Health Sciences Affiliated Zhoupu Hospital approved all protocols and experimental designs (approval no. 2020-ZPYY-KW-14-249X).

2.2 Collection of clinical samples

Thirty patients with ischemic stroke (48–79 years old, median age 65 years; 15 men and 15 women) treated at Shanghai University of Medicine and Health Sciences Affiliated Zhoupu Hospital were enrolled in this study. The patients were diagnosed within 5 days after onset using computed tomography (CT) or magnetic resonance imaging. Another 30 healthy individuals (45–71 years old, median age 65 years; 15 men and 15 women) who underwent routine physical examinations were recruited as controls. Pregnant patients and patients with malignancies or chronic diseases were excluded. The serum samples of patients were collected and frozen in liquid nitrogen until their subsequent use. The neurological function of patients at admission was assessed using the National Institutes of Health Stroke Scale (NIHSS).

2.3 Cell culture and treatment

A human cerebral cortical neuron cell line (HCN-2; CRL-10742) was acquired from ATCC (Manassas, VA, USA). Cells were incubated in Dulbecco's Modified Eagle Medium (DMEM; Gibco, Carlsbad, CA, USA) supplemented with 4 mM l-glutamine (G8540, Sigma-Aldrich, St Louis, MO, USA), 1.5 g/L sodium bicarbonate (S5761, Sigma-Aldrich), 4.5 g/L glucose (49163-100ML, Sigma-Aldrich), and 10% fetal bovine serum (12103C, Sigma-Aldrich) at 37°C with 5% CO2. After two phosphate-buffered saline washes, cells were cultured in glucose-free DMEM in an anoxic incubator (95% N2 and 5% CO2) for 3 h for oxygen/glucose deprivation (OGD) treatment.

An miR-379-5p mimic, pcDNA-MAP3K2, a short hairpin (sh) RNA of MAP3K2 1, 2, 3#, and the negative control (NC) mimic, a control pcDNA, and sh-NC were acquired from GenePharma Co., Ltd. (Shanghai, China). They were transfected into the HCN-2 cells using the Lipofectamine™ 2000 kit (Invitrogen, Thermo Fisher Scientific Inc., Waltham, MA, USA), according to the manufacturer's instructions. Rapamycin (HY-10219, purity 99.94%, CAS No. 53123-88-9) was acquired from MedChemExpress (Monmouth Junction, NJ, USA). The HCN-2 cells were treated with 20 nM rapamycin for 24 h to enhance the autophagy; dimethyl sulfoxide was used as the control.

2.4 Reverse transcription quantitative polymerase chain reaction

Total RNA from cells or tissues was extracted using the TRIzol reagent (10296010, Invitrogen). The RNA was reverse-transcribed to cDNA using a PrimeScript RT reagent kit (RR037Q, Takara, Kyoto, Japan). Next, real-time quantitative polymerase chain reaction (qPCR) was performed using a FastStart Universal SYBR Green Master (Rox) (4913850001, Roche Ltd., Basel, Switzerland) on a 7900 real-time PCR system (Applied Biosystems, Foster City, CA, USA). Relative gene expression was determined using the 2−ΔΔCt method. The primers are listed in Table 1. GAPDH served as the internal control for mRNAs, whereas U6 served as the control for miR-379-5p.

TABLE 1. Primer sequences for RT-qPCR Gene Primer sequence (5′–3′) hsa-miR-379-5p F: TGGTAGACTATGGAACGT R: GAACATGTCTGCGTATCTC mmu-miR-379-5p F: GGTAGACTATGGAACGTAG R: GAACATGTCTGCGTATCTC hsa-MAP3K2 F: TACACCCGTCAGATTCTGGAGG R: ATGGTCTGAAGCCGTTTGCTGG mmu-MAP3K2 F: CCCAGAGTATGACGACAGTCGA R: GGTAGACCCTACCAAAAGCTCC hsa-MTMR2 F: ACGAGGAACTCTGACTGTCACG R: TCACCTCGACTAGAAGCACCAC hsa-EIF4G2 F: CACGCACTCAAACACCACCTCT R: GGAGTTCTTCCTTTGACGGTGG hsa-ZBTB26 F: GAGCCACATTGTAGAACGGTGC R: GGAGAAGCACTCTGTGGTTCAC hsa-EDN1 F: CTACTTCTGCCACCTGGACATC R: TCACGGTCTGTTGCCTTTGTGG hsa-KCNJ1 F: GGCTACCGTTTTGCTCCCATAG R: CATAGCCTCTCTTCATCCTGGC hsa-HLCS F: ATCCAGCGTTCCAACCACACTC R: CGTTGAGCACTGGTGGGAAGAA hsa-LPGAT1 F: TACCACTTGGCTCTATCAGCGG R: CCACAAGTTGCTGAGGGTCATC hsa-NFATC3 F: AGACAGTCGCTACTGCAAGCCA R: GCGGAGTTTCAAAATACCTGCAC hsa-FIGN F: GCTACCTCATCCAACCACTCTG R: TTCCAGTCCACTGGAGGTCCTT GAPDH F: GTCTCCTCTGACTTCAACAGCG R: ACCACCCTGTTGCTGTAGCCAA U6 F: CTCGCTTCGGCAGCACAT R: TTTGCGTGTCATCCTTGCG Abbreviations: EDN1, endothelin 1; EIF4G2, eukaryotic translation initiation factor 4 gamma 2; F, forward; FIGN, fidgetin, microtubule severing factor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HLCS, holocarboxylase synthetase; hsa, homo sapiens; KCNJ1, potassium inwardly rectifying channel subfamily J member 1; LPGAT1, lysophosphatidylglycerol acyltransferase 1; MAP3K2, mitogen-activated protein kinase kinase kinase 2; miR, microRNA; mmu, mus musculus; MTMR2, myotubularin related protein 2; NFATC3, nuclear factor of activated T cells 3; R, reverse; RT-qPCR, reverse transcription quantitative polymerase chain reaction; ZBTB26, zinc finger and BTB domain containing 26. 2.5 Cell counting kit-8 method

A CCK-8 kit (C0037, Beyotime Biotechnology, Shanghai, China) was used to examine cell viability. The treated HCN-2 cells were cultured in 10 μl cell counting kit-8 (CCK-8) solution for 2 h. The optical density value was evaluated using a microplate reader (Tecan, Durham, NC, USA).

2.6 5-ethynyl-2′-deoxyuridine labeling assay

A 5-ethynyl-2′-deoxyuridine (EdU) labeling kit (C10310-1, RiboBio Co., Ltd., Guangdong, China) was used to measure the DNA replication ability of cells. In brief, the treated cells were exposed to 50 μmol/L EdU for 2 h. Subsequently, the cells were immobilized in 4% paraformaldehyde, incubated with 100 μl of Apollo reagent for 25 min, and then labeled with Hoechst 33342 solution. After staining, the cells were visualized using a fluorescence microscope (Olympus Optical Co., Ltd., Tokyo, Japan) with three fields included. The EdU positive rate was determined using the following equation: EdU positive rate = count of EdU-positive cells (red staining)/count of total cells (blue + red staining) × 100%.

2.7 Terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling

A terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) kit (C1088, Beyotime Biotechnology) was used to examine cell apoptosis. The transfected HCN-2 cells were fixed and permeated with 0.3% Triton X-100. The cells were incubated in the TUNEL solution in the dark for 60 min and then further stained with 4′,6-diamidino-2-phenylindole. Thereafter, the staining was observed under the fluorescence microscope with three random fields included. The apoptosis rate in cells was determined using the following equation: apoptosis rate = count of TUNEL-positive cells (red staining)/count of total cells (blue + red staining) × 100%.

2.8 Enzyme-linked immunosorbent assay

The concentration of the pro-inflammatory cytokine interleukin-6 in the cell supernatant was measured using an enzyme-linked immunosorbent assay (ELISA) kit (ab178013, Abcam Inc., Cambridge, MA, USA) according to the manufacturer's instructions.

2.9 Western blot analysis

Total protein from cells and tissues was isolated in phenylmethylsulfonyl fluoride- (10837091001, Roche) and cOmplete™ (11697498001, Roche) protease inhibitor-supplemented radio-immunoprecipitation assay cell lysis buffer (20-188, Upstate Biotechnology). A bicinchoninic acid kit (23227, Thermo Fisher Scientific) was utilized to examine protein concentration. Thereafter, the protein samples were run on a 10% sodium dodecyl sulfate-polyacrylamide gel for electrophoresis and transferred onto 0.22-μm nitrocellulose membranes (Millipore). The membranes were blocked in 5% nonfat milk and co-cultured with primary antibodies against Beclin1 (1:2000, ab207612, Abcam), LC3 (1:2000, ab192890, Abcam), MAP3K2 (1:10,000, ab33918, Abcam), JNK (1:2000, ab208035, Abcam), p-JNK (1:5000, ab76572, Abcam), c-Jun (1:1000, ab32137, Abcam), p-c-Jun (1:500, ab30620, Abcam), and GAPDH (1:1000, #5174, Cell Signaling Technology, Beverly, MA, USA) at 4°C overnight. Then, the membranes were further incubated with goat anti-rabbit IgG H&L (HRP) (1:10,000, ab205718, Abcam) at 20°C for 2 h. The protein blots were developed using an enhanced chemiluminescence reagent (32209, Pierce, Thermo Fisher Scientific). Relative protein expression was evaluated using Image J software.

2.10 Dual-luciferase reporter gene assay

The putative binding site between MAP3K2 and miR-379-5p was obtained using StarBase (http://starbase.sysu.edu.cn/). The MAP3K2 3′-UTR sequence containing the binding site with miR-379-5p was inserted into pmirGLO luciferase reporter vectors (Promega, Madison, WI, USA), denoted MAP3K2 wild-type (WT) vectors. The sequence containing a mutant binding site with miR-379-5p was designed, and the corresponding mutant type (MUT) vectors were constructed as well. Well-constructed vectors were co-transfected with either the miR-379-5p mimic or the mimic control into 293T cells (ATCC, USA). After 48 h, the luciferase activity was examined using a dual-luciferase reporter gene system (Promega).

2.11 Animal model with ischemic stroke

Male C57BL.6J mice (6–8 weeks old, 20–25 g) acquired from Vital River Laboratory Animal Technology Co., Ltd (Beijing, China) were used for animal experiments (Certificate No. SYXK [Beijing] 2017-0022). Ischemic stroke in mouse was induced by middle cerebral artery occlusion (MCAO). The mice were housed in an environment-controlled room in a 12 h light/dark cycle and were allowed free access to feed and water. The animals were numbered by weight and randomly allocated into four groups: sham group (n = 6); model group (n = 10); LV-NC group (LV refers to the lentiviral vector; these mice were injected with LV-NC 24 h prior to the MCAO; n = 15); and LV-mimic group (these mice were injected with LV-mimic 24 h prior to the MCAO; n = 15). There were no significant differences in the weights of the mice in each group. The discrepancy in the animal numbers between the groups was due to the modeling failure rate. At least six extra mice were required in each group to ensure the successful modeling.

The MCAO procedures were conducted as previously described.20 In brief, the animals were anesthetized via intraperitoneal injection of chloral hydrate (30 mg/kg). An incision was made on the midline of the neck. The left common carotid artery, external carotid artery, and internal carotid artery were isolated and ligated. A silicone rubber-coated monofilament was inserted into the internal carotid artery (9–10 nm) through the common carotid artery until mild resistance was felt, indicating an occlusion of all blood flow. After 1 h of ischemia, the ligation was removed to restore blood flow. The sham-operated mice were subjected to similar procedures except for the insertion of the monofilament. The body temperature of the mice was monitored and maintained at 37.0 ± 0.5°C throughout the operations. The mice were then housed in separate cages in a ventilated room at a constant temperature of 25 ± 3°C for recovery. After regaining consciousness, when the mice stood steadily, were paralytic, and turned to one side when their tails were lifted, the diseased model was considered to be successfully established. The MCAO procedure was performed 27 times to ensure six successfully modeled mice in each group: eight times in the model group, 10 times in the LV-NC group, and nine times in the LV-mimic group. The success rate of model establishment in each group was 75%, 60%, and 67%, respectively.

Injection of LV-NC and LV-mimic was performed 24 h prior to surgery as described in a previous report.21 The mice were anesthetized and fixed on a stereotaxic apparatus (RWD Life Science Co., Ltd., Shenzhen, China). The LV was incubated with cationic lipid polybrene (4 μg/μl, GenePharma) at 37°C for 15 min. Next, 5 μl of mixture was injected into the cortices of the mice for the MCAO operation through microliter syringes (Hamilton CO., Reno, NV, USA). The LV was acquired from GenePharma, and the injection titer was 109 TU/ml.

2.12 Scoring of neurological function

Twenty-four hours after the MCAO procedure, the neurological function of the mice was scored by two pathologists without any knowledge of the grouping details, according to the previous scoring standards.22 In brief, neurological function was scored based on six aspects: spontaneous activity (0~3), symmetry in the movement of four limbs (0~3), forepaw outstretching (0~3), climbing (1~3), body proprioception (1~3), and response to vibrissae touch (1~3). Total scores were calculated from the six tests, ranging from 3 to 18, with lower scores indicating worse neurological conditions and severer neurological impairment.

2.13 2,3,5-Triphenyltetrazolium chloride staining

The mice were euthanized by intraperitoneally administering an overdose of pentobarbital sodium (150 mg/kg). The mouse brain tissues were cut into 2-mm coronal sections. The sections were stained with 1% 2,3,5-triphenyltetrazolium chloride (TTC) solution (17779-10X10ML-F, Sigma-Aldrich), immersed in saline at 37°C for 20 min, fixed at 4°C overnight, and then visualized under a microscope. The noninfarct areas were stained in deep red, whereas infarct areas were not stained. Infarct size was evaluated using Image-Pro with the encephaledema excluded. The infarct area percentage was determined as follows: percentage of infarct area (%) = (total area of the contralateral hemisphere − noninfarct area of the ipsilateral hemisphere)/total area of the contralateral hemisphere × 100%.

2.14 Immunofluorescence staining

The HCN-2 cells were fixed with 3% glutaraldehyde and permeated with 0.3% Triton X-100. Subsequently, the cells were incubated with 5% normal goat serum (16210064, Gibco) for 1 h, the primary antibody against LC3 (1:200, ab192890, Abcam) at 4°C overnight, and the secondary antibody against IgG (Alexa Fluor® 647) (1:500, ab150079, Abcam) at 25°C for 1 h. The nuclei were counter-stained with 4′,6-diamidino-2-phenylindole (DAPI; C1005, Beyotime Biotechnology). The staining was observed under a fluorescence microscope (Olympus).

The cortical tissues of mice were cut into 4-μm sections, permeated with 0.2% Triton X-100 (93443-100ML, Sigma-Aldrich), treated with goat serum for 1 h, and incubated with the rabbit polyclonal antibody LC3 (1:200, ab192890, Abcam) and mouse monoclonal antibody NeuN (1:200, ab104224, Abcam) at 4°C overnight. Thereafter, the tissue slides were incubated with fluorescein-conjugated anti-rabbit IgG (Alexa Fluor® 647) (1:500, ab150079, Abcam) and anti-mouse IgG (Alexa Fluor® 488) (1:100, ab150113, Abcam) for 1 h. The staining was observed under a microscope. The staining intensity of LC3 (cytoplasm) surrounding the staining of the NeuN (nucleus) of 50 neurons was examined to evaluate autophagy in the neurons.

2.15 Statistical analysis

Statistical analysis was conducted using GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA). Measurement data were presented as the mean ± standard error of mean (SEM) from three independent experiments. Differences were analyzed using a t test (two groups) or one- or two-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test (over two groups). Correlation between variables was determined by Pearson's correlation analysis. Fisher's exact test was applied for association analysis in the contingency table. The significant difference was set at p < 0.05.

3 RESULTS 3.1 miR-379-5p is poorly expressed in patients with ischemic stroke and in OGD-treated HCN-2 cells

To evaluate the function of miR-379-5p in ischemic stroke (cerebral infarction), we determined the expression of miR-379-5p in the serum of patients and healthy individuals. The reverse transcription quantitative polymerase chain reaction (RT-qPCR) results suggested a notable downregulation of miR-375-5p in patients with ischemic stroke compared to control (Figure 1(A)). According to the mean value of miR-379-5p expression (0.405), the stroke patients were allocated into a high-miR-379-5p group (n = 13) and a low-miR-379-5p group (n = 17). It was found that low serum level of miR-379-5p was correlated with increased NIHSS scores in patients (Table 2).

miR-379-5p is poorly expressed in patients with ischemic stroke and in OGD-treated HCN-2 cells (n = 30 in each group). (A) Expression of miR-379-5p in the serum samples of patients with ischemic stroke or of healthy individuals; (B) viability of HCN-2 cells after OGD treatment determined by the CCK-8 method; (C) DNA replication ability of HCN-2 cells determined by the EdU labeling assay; (D) apoptosis rate of HCN-2 cells evaluated by the TUNEL assay; (E) miR-379-5p expression in HCN-2 cells after OGD treatment determined by RT-qPCR; (F) production of IL-6 in cell supernatant after OGD treatment measured by ELISA kits. Data were presented as mean ± SEM from three independent experiments. Differences were analyzed by the unpaired t test (A, B, C, D, E and F). **p < 0.01 versus Normal; #p < 0.05 versus control. miR-379-5p, microRNA-379-5p; OGD, oxygen/glucose deprivation; TUNEL, terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling

TABLE 2. Association between miR-379-5p expression and baseline characteristics of the patients enrolled in the study Characteristics miR-379-5p expression p-Value High (n = 13) Low (n = 17) Age (year) ≥65 (n = 15) 9 6 0.1394 <65 (n = 15) 4 11 Gender Female (n = 15) 5 10 0.4621 Male (n = 15) 8 7 Smoking Yes (n = 19) 9 10 0.7084 No (n = 11) 4 7 Drinking Yes (n = 18) 6 12 0.2641 No (n = 12) 7 5 NIHSS scores ≥15 (n = 21) 6 15 *0.0196 <15 (n = 9) 7 2 Old/new stroke Old (n = 17) 5 12 0.1376 New (n = 13) 8 5 Family history of stroke Yes (n = 18) 7 11 0.7106 No (n = 12) 6 6 Note: Fisher's exact test was applied for association analysis. The mean value of miR-379-5p expression was 0.405; *p < 0.05. Abbreviation: NIHSS, National Institutes of Health Stroke Scale.

HCN-2 cells were subjected to OGD treatment to mimic an in vitro condition of ischemic stroke. The results of the CCK-8 assay suggested that the viability of OGD-treated cells was significantly reduced (Figure 1(B)). Likewise, the EdU labeling assay indicated that the DNA replication activity of cells was significantly reduced after the OGD treatment (Figure 1(C)). However, the TUNEL assay results demonstrated that the number of apoptotic HCN-2 cells was significantly increased after the OGD treatment (Figure 1(D)). According to the RT-qPCR analysis, the miR-379-5p expression in cells was also significantly reduced following the OGD treatment (Figure 1(E)). In addition, the ELISA assay suggested that the OGD treatment significantly increased the production of IL-6 in cells (Figure 1(F)).

3.2 Upregulation of miR-379-5p attenuates OGD-induced damage on HCN-2 cells

To investigate the function of miR-379-5p in the neurons, overexpression of miR-379-5p was introduced in OGD-treated HCN-2 cells by transfecting them with an miR-379-5p mimic, and successful transfection was confirmed by RT-qPCR (Figure 2(A)). The viability of cells was found to be significantly increased after miR-379-5p overexpression (Figure 2(B)). Additionally, the rate of EdU-positive cells, indicating the number of proliferating cells, was increased by the miR-379-5p mimic (Figure 2(C)). The TUNEL assay suggested that the number of apoptotic HCN-2 cells induced by OGD was suppressed upon miR-379-5p upregulation (Figure 2(D)). The ELISA results suggested that miR-379-5p reduced the secretion of IL-6 in the cells (Figure 2(E)).

Upregulation of miR-379-5p attenuates OGD-induced damage on HCN-2 cells. (A) miR-379-5p expression in cells after transfection determined by RT-qPCR; (B) viability of HCN-2 cells determined by the CCK-8 method; (C) DNA replication ability of HCN-2 cells determined by the EdU labeling assay; (D) apoptosis rate of HCN-2 cells evaluated by the TUNEL assay; (E) production of IL-6 in cell supernatant after OGD treatment examined using ELISA kits. Data were presented as mean ± SEM from three independent experiments. Differences were analyzed by the unpaired t test (A, B, C, D and E). *p < 0.05 versus NC mimic. miR-379-5p, microRNA-379-5p; NC, negative control; OGD, oxygen/glucose deprivation; TUNEL, terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling

3.3 Rapamycin partially blocks the functions of miR-379-5p in HCN-2 cells exposed to OGD

Activation of autophagy has been reported to promote the pathogenesis of ischemic stroke.5, 6 Therefore, we determined the expression of the autophagy marker protein Beclin1 and the ratio of LC3II/I in cells after OGD treatment or miR-379-5p mimic transfection. It was found that the OGD treatment significantly increased the expression of Beclin1 and ratio of LC3II/I, whereas the overexpression of miR-379-5p successfully blocked Beclin1 expression and reduced the LC3II/I ratio in cells (Figure 3(A)). In addition, immunofluorescence staining of LC3 revealed that the staining intensity of LC3 in cells was significantly increased after the OGD treatment; however, this was reduced following miR-379-5p overexpression (Figure 3(B)).

Rapamycin partially blocks the functions of miR-379-5p on HCN-2 cells exposed to OGD. (A) Expression of Beclin1, LC3I, and LC3II in cells after OGD treatment and miR-379-5p mimic transfection determined by western blot analysis; (B) autophagy activity in cells examined by immunofluorescence staining; (C) expression of Beclin1, LC3I, and LC3II in cells after further rapamycin administration determined by western blot analysis; (D) autophagy activity in cells examined by immunofluorescence staining; (E) viability of HCN-2 cells detected by the CCK-8 method; (F) DNA replication ability of HCN-2 cells determined by the EdU labeling assay; (G) apoptosis rate in OGD- and rapamycin-treated HCN-2 cells evaluated by the TUNEL assay; (H) production of IL-6 in HCN-2 cells measured by ELISA assay; (I) apoptosis of HCN-2 cells treated with rapamycin alone examined by the TUNEL assay. Data were presented as mean ± SEM from three independent experiments. Differences were analyzed by the unpaired t test (C, D, E, F, G, and H) and one-way ANOVA (A and B). *p < 0.05 versus control; #p < 0.05 versus NC mimic; &p < 0.05 versus miR-379-5p mimic + DMSO; ns, no significance. DAPI, 4′,6-diamidino-2-phenylindole; DMSO, dimethyl sulfoxide; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; NC, negative control; OGD, oxygen/glucose deprivation; TUNEL, terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling

To confirm the involvement of autophagy inhibition in the neuronal protection mediated by miR-379-5p, cells transfected with the miR-379-5p mimic were further treated with an autophagy-specific activator, rapamycin. After 24 h, the suppressed expression of Beclin1 and the reduced LC3II/I ratio due to miR-379-5p were recovered after rapamycin treatment (Figure 3(C)). In addition, the fluorescence staining results suggested that the rapamycin treatment enhanced the staining intensity of LC3 in cells (Figure 3(D)). Furthermore, rapamycin suppressed the viability (Figure 3(E)) and inhibited the DNA replication abilities of HCN-2 cells, reversing the effects of miR-379-5p (Figure 3(F)). In addition, the TUNEL assay suggested that the apoptosis of cells was resumed and increased following the application of rapamycin (Figure 3(G)). In addition, the production of IL-6 in cells was increased after autophagy activation (Figure 3(H)). Moreover, the TUNEL assay suggested that using rapamycin alone at this dose did not significantly affect apoptosis of cells without OGD treatment (Figure 3(I)).

3.4 miR-379-5p directly targets MAP3K2

Following the findings above, we further probed for the potential downstream molecules involved. The target mRNAs of miR-379-5p were predicted using miRDB (http://mirdb.org/), TargetScan (http://www.targetscan.org/vert_72/), Starbase (http://starbase.sysu.edu.cn/), miRWalk (http://mirwalk.umm.uni-heidelberg.de/), and miRDIP (http://ophid.utoronto.ca/mirDIP/) (Figure 4(A)), and 10 common target genes were suggested to be intersected (Figure 4(B)). The expression of these mRNAs in cells after transfection with the miR-379-5p mimic was determined by RT-qPCR. We found that only MAP3K2 was significantly decreased in cells after miR-379-5p overexpression (Figure 4(C)). Moreover, the western blot analysis showed that the protein level of MAP3K2 in cells was increased by OGD but suppressed after miR-379-5p overexpression (Figure 4(D)).

miR-379-5p directly targets MAP3K2. (A) Targeting mRNAs of miR-379-5p predicted on several bioinformatic systems; (B) 10 common targeting genes of miR-379-5p; (C) expression of the 10 candidate genes in HCN-2 cells after miR-379-5p transfection determined by RT-qPCR; (D) protein level of MAP3K2 in HCN-2 cells after OGD treatment and miR-379-5p overexpression determined by western blot analysis; (E) expression of MAP3K2 in the serum of patients with ischemic stroke determined by RT-qPCR (n = 30); (F) correlation between miR-379-5p and MAP3K2 expression in the serum of patients; (G) putative binding site between MAP3K2 and miR-379-5p predicted on StarBase; (H) binding relationship between MAP3K2 and miR-379-5p validated through a luciferase reporter gene assay. Data were presented as mean ± SEM from three independent experiments. Differences were analyzed by the unpaired t test (D) and two-way ANOVA (C and H). Correlation between miR-379-5p and MAP3K2 (F) was analyzed by the Pearson's correlation analysis (r = −0.668, p < 0.001). *p < 0.05 versus NC mimic; #p < 0.05 versus Normal. MUT, mutant type; NC, negative control; WT, wild type

In addition, the serum level of MAP3K2 in patients with stroke and healthy individuals was examined by RT-qPCR. It was observed that MAP3K2 expression levels were increased in patients with ischemic stroke (Figure 4(E)), indicating a negative correlation with miR-379-5p (Figure 4(F)).

The putative binding site between MAP3K2 mRNA and miR-379-5p was obtained from StarBase (Figure 4(G)), and the binding relationship was validated through a dual-luciferase reporter gene assay. The MAP3K2-WT and MAP3K2-MUT luciferase reporter vectors were co-transfected with the NC mimic or miR-379-5p mimic into 293T cells. After 48 h, it was found that the luciferase activity in cells co-transfected with the miR-379-5p mimic and MAP3K2-WT vectors was reduced (Figure 4(H)), while transfection of MAP3K2-MUT or the NC mimic led to no changes in the luciferase activity of the cells.

3.5 Overexpression of MAP3K2 activates the JNK/c-Jun signaling pathway and blocks the protective roles of miR-379-5p on HCN-2 cells exposed to OGD

MAP3K2 has been found to be a positive regulator of JNK phosphorylation,16 and the JNK/c-Jun pathway has been reported to be a regulator of autophagy.18, 19 We speculated that MAP3K2 activates the JNK/c-Jun signaling pathway to increase autophagy in cells. Therefore, further overexpression of MAP3K2 was induced in HCN-2 cells using pcDNA-MAP3K2 after miR-379-5p mimic transfection, and the successful transfection was validated using RT-qPCR (Figure 5(A)). Importantly, the fluorescence staining results showed that miR-379-5p-induced reduction in autophagy was reversed due to MAP3K2 overexpression, presented as increased staining intensity of LC3 (Figure 5(B)). Further, the Beclin1 expression, LC3II/I ratio, and the JNK/c-Jun signaling activity in cells were detected by western blot analysis. The protein level of Beclin1, ratio of LC3II/I, and the phosphorylation of JNK and c-Jun were significantly suppressed by the addition of the miR-379-5p mimic; however, these increased following MAP3K2 overexpression (Figure 5(C)).

Overexpression of MAP3K2 activates the JNK/c-Jun signaling pathway and blocks the protective roles of miR-379-5p. (A) Transfection efficacy of pcDNA-MAP3K2 in cells determined by RT-qPCR; (B) autophagy activity in cells examined by immunofluorescence staining of LC3; (C) protein levels of Beclin1 and LC3-II/I and the phosphorylation of JNK and c-Jun in cells determined by western blot analysis; (D) viability of HCN-2 cells detected by the CCK-8 method; (E) DNA replication ability of HCN-2 cells determined by the EdU labeling assay; (F) apoptosis rate of HCN-2 cells evaluated by the TUNEL assay; (G) production of IL-6 in HCN-2 cells measured by ELISA assay. Data were presented as mean ± SEM from three independent experiments. Differences were analyzed by the unpaired t test (A, B, D, E, F, and G) and one-way ANOVA (C). #p < 0.05 versus NC mimic;*p < 0.05 versus miR-379-5p mimic + pcDNA. GAPDH, glyceraldehyde-3-phosphate dehydrogenase; JNK, Jun N-terminal kinase; MAP3K2, mitogen-activated protein kinase kinase kinase 2; miR-379-5p, microRNA-379-5p; NC, negative control; TUNEL, terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling

The CCK-8 assay results showed that the viability of cells enhanced by miR-379-5p was inhibited upon further MAP3K2 overexpression (Figure 5(D)). The EdU labeling assay results demonstrated that the DNA replication ability of cells was suppressed by MAP3K2 (Figure 5(E)). The TUNEL assay showed that the cell apoptosis suppressed by the miR-379-5p mimic was restored after MAP3K2 overexpression (Figure 5(F)). In addition, the ELISA assay suggested that the production of IL-6 in cells was increased after MAP3K2 overexpression (Figure 5(G)).

3.6 MAP3K2 mediates neuronal injury only in OGD-treated cells

The shRNAs of MAP3K2 (sh-MAP3K2 1, 2, 3#) were introduced into OGD-treated HCN-2 cells. The transfection efficacy was determined by RT-qPCR, and the sh-MAP3K2 1# with the best interfering efficacy was used in the subsequent experiments (Figure 6(A)). The western blot analysis suggested that silencing of MAP3K2 suppressed the expression of Beclin1, ratio of LC3II/I, and phosphorylation of the JNK /c-Jun signaling pathway (Figure 6(B)). The immunofluorescence staining also showed that the autophagy activity of cells was weakened by MAP3K2 silencing (Figure 6(C)). In addition, silencing of MAP3K2 restored the OGD-suppressed viability of cells (Figure 6(D)), enhanced the DNA replication of cells (Figure 6(E)), suppressed cell apoptosis (Figure 6(F)), and reduced the secretion of IL-6 in cells (Figure 6(G)). Moreover, pcDNA-MAP3K2 or the control pcDNA was administrated to HCN-2 cells without OGD treatment (Figure 6(H)). However, overexpression of MAP3K2 did not significantly affect the apoptosis of neurons (Figure 6(I)), indicating that MAP3K2-mediated neuronal damage is possibly only activated under hypoxic–ischemic injury conditions.

MAP3K2 mediates neuronal injury only in OGD-treated cells. (A) MAP3K2 expression in HCN-2 cells after sh-MAP3K2 1, 2, 3# transfection determined by RT-qPCR; (B) expression of Beclin1 and LC3 II/I, and the phosphorylation of JNK and c-Jun in the cells detected by western blot analysis; (C) LC3 expression in cells examined by immunofluorescence staining; (D) viability of HCN-2 cells determined by the CCK-8 method; (E) DNA replication ability of HCN-2 cells determined by the EdU labeling assay; (F) apoptosis of HCN-2 cells evaluated by the TUNEL assay; (G) production of IL-6 in cell supernatant after OGD treatment measured by ELISA kits; (H) transfection efficacy of pcDNA-MAP3K2 in non-OGD-treated cells examined by RT-qPCR; (I) apoptosis rate of non-OGD treated cells following MAP3K2 overexpression examined by the TUNEL assay. Data were presented as mean ± SEM from three independent experiments. Differences were analyzed by the unpaired t test (B, C, E, F, G, H, and I) or one-way ANOVA (A). *p < 0.05 versus sh-NC; #p < 0.05 versus pcDNA; ns, no significance. GAPDH, glyceraldehyde-3-phosphate dehydrogenase; JNK, Jun N-terminal kinase; MAP3K2, mitogen-activated protein kinase kinase kinase 2; sh-NC, short hairpin RNA-negative control; TUN