Inhibition of miRNA‐27b enhances neurogenesis via AMPK activation in a mouse ischemic stroke model

Stroke is a leading cause of death and disability, but treatment options remain limited. Recent studies have suggested that cerebral ischemia‐induced neurogenesis plays a vital role in post‐stroke repair. Overactivation of AMP‐activated protein kinase (AMPK), a master sensor of energy balance, has been reported to exacerbate neuron apoptosis, but the role of chronic AMPK stimulus in post‐stroke recovery remains unclear. MicroRNAs have emerged as regulators of neurogenesis and have been reported to be involved in neurological function. In this study, we verified that miR‐27b directly targets AMPK and inhibits AMPK expression. In cultured neural stem cells, miR‐27b inhibitor improved proliferation and differentiation via the AMPK signaling pathway, but did not have an obvious effect on cell viability under oxygen and glucose deprivation conditions. In a mouse middle cerebral artery occlusion model, administration of miR‐27b inhibitor significantly enhanced behavioral function recovery and spatial memory. Up‐regulation of neurogenesis was observed both in the subventricular zone and in the hippocampal dentate gyrus. Collectively, our data suggest that miR‐27b inhibition promotes recovery after ischemic stroke by regulating AMPK activity. These findings may facilitate the development of novel therapeutic strategies for stroke.

Stroke is a leading cause of death and disability, but treatment options remain limited. Recent studies have suggested that cerebral ischemiainduced neurogenesis plays a vital role in post-stroke repair. Overactivation of AMP-activated protein kinase (AMPK), a master sensor of energy balance, has been reported to exacerbate neuron apoptosis, but the role of chronic AMPK stimulus in post-stroke recovery remains unclear. Micro-RNAs have emerged as regulators of neurogenesis and have been reported to be involved in neurological function. In this study, we verified that miR-27b directly targets AMPK and inhibits AMPK expression. In cultured neural stem cells, miR-27b inhibitor improved proliferation and differentiation via the AMPK signaling pathway, but did not have an obvious effect on cell viability under oxygen and glucose deprivation conditions. In a mouse middle cerebral artery occlusion model, administration of miR-27b inhibitor significantly enhanced behavioral function recovery and spatial memory. Up-regulation of neurogenesis was observed both in the subventricular zone and in the hippocampal dentate gyrus. Collectively, our data suggest that miR-27b inhibition promotes recovery after ischemic stroke by regulating AMPK activity. These findings may facilitate the development of novel therapeutic strategies for stroke.
Stroke is one of leading causes of death and long-term disability worldwide [1]. Ischemic stroke accounts for more than 80% of stroke, local ischemia and a hypoxic environment induce neuronal cell apoptosis, necrosis and other metabolism-related disorders. At present, recombinant tissue plasminogen activator is the only FDA-approved treatment for acute ischemic stroke, but a narrow therapeutic window limits its clinic usage [2]. Meanwhile, there is no neuroprotective agent that has shown prospective results in clinical trials.
Brain repair processes after ischemic stroke are attracting more attention. Accumulating studies have suggested that cerebral ischemia-induced neurogenesis [3][4][5], which occurs mainly in the subventricular zone (SVZ) and subgranular zone in hippocampal dentate gyrus (DG) [6], might give rise to a cure strategy for recovery of brain function [7,8]. In DG, newborn neurons are continuously produced throughout adulthood [9][10][11]. A manipulated increase of neurogenesis in DG was reported to improve synaptic plasticity and memory [12,13]. For now, it remains a challenge to amplify this process to enhance neurological function after injury.
The role that AMP-activated protein kinase (AMPK) plays in ischemic brain remains a matter of debate. Several studies reported that AMPK activation is detrimental, since it enhances energy-consumptive pathways that further contribute to the death of injured neuronal cells [14][15][16]. Meanwhile, Kuramoto et al. [17] suggested that AMPK leads to neuroprotection through functional modulation of the GABA B receptor. Besides the effects on neural apoptosis and mortality, AMPK has been reported to play a dual role in generation of new neurons; the divergent effects result from the duration and extent of AMPK activation [18,19]. The use of low dose 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR; an AMPK direct activator) enhanced neurogenesis and cognition, while high dose AICAR impaired cognition and induce cell apoptosis [20].
MicroRNAs (miRNAs) are small non-coding RNAs of~20 nucleotides that act as post-transcriptional regulators of their targets mRNAs [21]. The studies of other groups have indicated that miRNAs play an important role in neuroprotection and post-stroke recovery [22][23][24][25][26][27]. It has been demonstrated that miR-27b was markedly reduced in numerous types of cancer and exerted anti-tumor effects by suppressing cell proliferation [28], but its effects on neurogenesis after ischemic stroke have not been investigated.
In the present study, we confirmed the direct targeting of miR-27b on AMPKa2, the main effective AMPKa isoform in ischemic brain [14]. We down-regulated miR-27b expression in cultured neural stem cells (NSCs) and observed mild stimulation of AMPK. In conditions of ischemia and hypoxia, the enhancement of AMPK activity by anti-miR-27b improved neurogenesis and poststroke neurological function recovery.

Cortical neuronal culture
Cortical neurons were isolated from 18-day-old male C57BL/ 6 mice as described [32]. Neurons were dissociated with 0.125% trypsin and pre-purified by density gradient. Neurons were seeded at a density of~2500 cellsÁmm À2 and cultured using DMEM containing 20% FBS. On the next day, the medium was changed to neuron-basal medium supplemented with 2% B27 (R&D Systems). Neurons were cultured for another 3 days and used for cell viability assay.

Oxygen-glucose deprivation and cell viability assay
Cortical neurons and NSCs were washed with PBS three times, then placed in glucose-free Earle's Balanced Salt Solution (EBSS) buffer and stored in an anoxic chamber filled with 95% N 2 and 5% CO 2 for 6 h [33]. For the control, cells remained in normoxic conditions. The cell survival rates were determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) assay. MTT was added into the medium at a final concentration of 0.5 mgÁmL À1 and incubated at 37°C for 4 h. The MTT formazan was dissolved in DMSO and the absorbance was measured at 490 nm. Caspase-3 activity was tested using a Caspase-3 Colorimetric Assay kit (Abcam, Cambridge, MA, USA).

Immunofluorescence staining and quantification
Neural stem cells were transfected with anti-miR-27b or anti-miR-NC and cultured for another 5 days. To label the dividing cells, 5 lM 5-bromo-2 0 -deoxyuridine (BrdU; Sigma-Aldrich) was added to the proliferation medium and the cells were cultured for another 8 h. The cells were then fixed and incubated with rabbit anti-BrdU (1 : 2000; Abcam) overnight at 4°C and then incubated with goat anti-rabbit IgG (FITC, 1 : 300; Abcam) for 2 h. The slides were washed and mounted using mounting medium (Vector Laboratories, Burlingame, CA, USA) containing 4 0 ,6-diamidino-2-phenylindole for nucleus staining. To detect production of neurons, NSCs were transfected as described above and cultured in differentiation medium (the proliferation medium supplemented with 5% FBS, without any growth factors) for 5 days, then cells were fixed and blocked and incubated with mouse anti-neuron-specific class III b-tubulin (Tuj1, 1 : 300; Abcam) overnight at 4°C. The next day, cells were washed and incubated with goat anti-mouse IgG (FITC, 1 : 300; Abcam) secondary antibody for 2 h. The slides were washed and mounted as above.
For quantification, BrdU + or Tuj1 + and total cell numbers were counted in 10 random view fields/well and three wells/group under a 920 objective lens. The results were presented as the percentage of immunoreactive cells.
To determine neurogenesis in vivo, coronal sections of brains were stained with the following markers: rabbit antidoublecortin (DCX; 1 : 100; Abcam), rabbit anti-neuronal nuclei (NeuN; 1 : 100; Abcam). Images were taken under a 940 objective lens, and five random view fields were chosen for each section. One-in-ten series of sections were stained for each animal.

Animal experimental procedures
All experimental procedures were carried out in accordance with the Care and Use of Laboratory Animals Guide from the Ministry of Public Health of China and approved by the medical ethics committee of Weifang Medical University and Shandong University. Male C57BL/6J mice (~10 weeks, 18-22 g, specific pathogen free) were purchased from Animal Centre of Shandong University. The animals were acclimatized for a week before surgery. On day 0, all mice underwent functional tests and performed permanent middle cerebral artery occlusion (MCAo; sham group only exposed arteries, no sutures were inserted) [31,34,35]. Briefly, animals were anesthetized and the common carotid artery, external carotid artery and internal carotid artery were exposed on the right side. An 8-0 nylon filament was inserted from the external carotid artery into the lumen of the internal carotid artery. Two hours later, the motor tests were re-done in a modified neurological severity score (mNSS) test (Table 1). The animals were randomly divided and received scores of 3-5 in the following three groups: MCAo, antagomir negative control (200 lgÁkg À1 , antagomir-NC; GenePharma, Shanghai, China) and antagomir-27b (200 lgÁkg À1 ; Gene-Pharma); n = 15 for each group. Antagomir negative control or antagomir-27b was i.v. injected and the same dose of antagomir was injected on days 7, 14 and 28. Cell proliferation was assessed using BrdU incorporation; five mice from each group were i.p. injected with 100 mgÁkg À1 BrdU from day 21 to day 28 after MCAo. At the end of the study, these animals were perfused with saline and 4% paraformaldehyde, and then brains were removed for immunostaining. Five mice from each group were sacrificed and their brains removed for 2,3,5-triphenyltetrazolium chloride (TTC) staining. The other five animals were tested for cerebral blood flow (CBF) using laser Doppler flowmetry (LSFG-ANW, Softcare, Ltd., Fukuoka, Japan). Using the same filter and image solution setting, relative blood flow in regions of interest from both the contralateral and the ipsilateral half brain was recorded and analyzed, and then mice were sacrificed and the brains removed for protein extraction and western blot analysis.

Behavioral tests
A modified neurological severity score test was performed before MCAo, and on days 1, 4, 7, 14 and 28 after MCAo according to a previous report [31,36]. All animals were tested for motor and sensory abilities, balance, reflex and abnormal movements. The evaluation standard is shown in Table 1; higher scores represent severer injury and the maximum score is 14.
The Morris water maze was used as described [31,37]. A white swimming pool 150 cm in diameter was filled with water to a depth of 30 cm. Pool temperature was maintained at 25 AE 0.5°C. The escape platform was placed in the center of one quadrant of the pool, 1 cm below the surface of the water. Mice were allowed to find the submerged platform within 90 s. If they failed, the mice were placed on the platform for another 15 s. On each training day, each mouse was released from four different start points. The training was repeated for five consecutive days. The escape latency and the time spent in the quadrant with the platform were recorded .

Statistical analysis
All data are presented as the mean AE standard error (SE). Data analysis was performed using PRISM 6 (Graph-Pad Software, San Diego, CA, USA). ANCOVA was used to analyze functional test data; non-parametric oneway ANOVA followed by a post hoc Tukey's test was used to compare the difference in cell viability data, immunofluorescence staining data, and western blot data among groups. Statistical significance was set at P < 0.05.

MiR-27b directly targets AMPKa2
MiR-27b was predicted to match the 3 0 UTR of PRKAA2 (the gene encoding AMPKa2) mRNA according to the gene sequence database at targetscan.org (Fig. 1A). We employed a luciferase assay to confirm the targeting relation of miR27b and AMPKa2. In cells co-transfected with WT AMPKa2 3 0 UTR (AMPKa2 WT) and miR-27b mimic, luciferase activity was significantly decreased compared to miR control or mutant AMPKa2 3 0 UTR (AMPKa2 Mu), which indicated that AMPKa2 is directly targeted by miR-27b (Fig. 1A).

The effect of AMPK activation by anti-miR-27b on NSCs
To test the effect of anti-miR-27b AMPK activation, AICAR (1 mM), Compound C (an AMPK inhibitor, 1 lM), and anti-miR-27b were administrated at the indicated dose and the NSC culture was exposed to oxygen-glucose deprivation (OGD). We compared the expression of p-AMPKa2 in AICAR-treated and anti-miR-27b-transfected NSCs. AICAR enhanced p-AMPKa2 expression starting at 0.5 h (2.1-fold) but AMPKa2 expression was not affected. The effect of anti-miR-27b was detected at 6 h, and both p-AMPKa2 and AMPKa2 were significantly increased (3.4-fold and 2.3-fold, respectively; Fig. 1B). At 6 h, anti-miR-27b increased the expression of p-AMPKa2 and p-ACC (a down-stream target of AMPK in which p-ACC level generally reflects the AMPK activity) (Fig. 1C). The AMPK-mTOR axis appears to have a role in neurodegenerative processes. The expression pattern of p-mTOR was the converse of that of p-AMPK, while total mTOR expression was not affected (Fig. 1C).

The effect of AMPK activation on cell viability in OGD treatment
AMP-activated protein kinase is known to be activated by energy deficiency. To determine the effect of activated AMPK on apoptosis, we examined the time course of AMPK activity and cell viability in OGD treatment. Both in cortical neurons and in NSCs, p-AMPKa2 expression was significantly increased 0.5 h after OGD treatment and further enhanced at 6 h ( Fig. 1D,E). OGD decreased neuron viability to 51% at 6 h (Fig. 1D). In NSCs, the time course for p-AMPKa2 expression was similar to that in neurons, but cell viability was slightly decreased to 89% with OGD treatment (Fig. 1E). These data indicate that in a hypoxic environment, AMPK stimuli might play different roles in different cell types. Next, we prolonged the OGD to 24 h, and tested the NSC viability. In OGD-treated groups, cell numbers were slightly increased compared to control group; anti-miR-NC and anti-miR-27b treatments exerted no obvious effects on apoptosis (Fig. 1F,G).

AMPK activation by anti-miR-27b improved proliferation and differentiation of NSCs
NSCs cultured in growth medium were transfected with anti-miR-NC or anti-miR-27b, and another group of cells were transfected with miR-27b inhibitor with Compound C added (1 lM). Immunofluorescence staining showed that OGD increased the percentage of BrdU-positive cells compared with control group (28.1 AE 3.0% vs 17.2 AE 2.1%, P < 0.05) ( Fig. 2A,B). Anti-miR-27b further up-regulated the BrdU + cell numbers by 55.4 AE 6.3%. But the effect of anti-miR-27b was blocked by Compound C (%BrdU + cells was 15.3 AE 2.2%). Next, we tested differentiation of NSCs in all groups treated the same way as above. Except the control group, NSCs were cultured in an OGD incubator for 24 h followed by 4 days in normal conditions. The percentage of Tuj1-positive cells was increased after OGD treatment (3.8 AE 0.4% vs 5.9 AE 0.5%, P < 0.05). Anti-miR-27b further enhanced the production of new neurons to 13.5 AE 2.1%. Compound C reduced the percentage Tuj1 + cells to 4.6 AE 0.8% (Fig. 2C,D). These data indicated that, AMPK activation by anti-miR-27b transfection induced NSCs proliferation and differentiated into new neurons.

Antagomir-27b improved neurological recovery in MCAo mice
To verify the effects of miR-27b inhibition in neurological outcomes after ischemic stroke. We treated mice with antagomir-27b to silence endogenous miR-27b after MCAo surgery, and observed neurological functional recovery. As shown in Fig. 3, the antagomir-27b group showed a better neurological functional recovery outcome compared to the MCAo group from day 4 and this lasted through the entire study period (P < 0.05, Fig. 3A). From day 28, we employed the Morris water maze test to assess the spatial memory of all groups. The escape latency time for finding a hidden platform in the group that went through MCAo surgery took much longer than in the sham group. Although the escape latency time of all groups showed decreased tendency during the training period of five consecutive days, antagomir-27b-treated mice took a shorter time to find the platform from the third to fifth days compared with the MCAo group (P < 0.05, Fig. 3B). In the probe trial session, swimming time within the target quadrant was low in the MCAo group, which indicated that spatial memory of these mice was impaired after stroke. The antagomir-27b group showed a longer swimming time in the target quadrant than the MCAo group (P < 0.05, Fig. 3C, D). These data strongly indicated that antagomir-27b improved neurological function recovery of ischemic stroke mice.
During the study period, the mortality rate of the MCAo group was 17%, and treatment with antagomir-27b showed no obvious effect (Fig. 3E). Before sacrifice, the CBF in both the contralateral and the ipsilateral half brain was measured. The ipsilateral/ contralateral flow ratio of the antagomir-27b group was significantly higher than that of the MCAo group (78.4% vs 62.7%, P < 0.05, Fig. S1). TTC staining revealed that the brain infarct volume was obviously decreased after antagomir-27b (Fig. 3E,F).

Antagomir-27b enhanced neurogenesis in MCAo mice via activating AMPK
To explore the mechanism beneath the therapeutic effects of antagomir-27b on ischemic stroke mice, we further tested whether the regulation of AMPK activity would affect neurogenesis in vivo. Immunostaining of hippocampal sections demonstrated that cell proliferation in the subgranular zone (revealed by BrdU + cell populations) showed newborn neurons were enhanced in the MCAo group. The treatment of antagomir-27b significantly increased the number of BrdU + cells by 44.5% compared with the MCAo group ( Fig. 4A-D). Consistent with this, we observed an increase in cells expressing DCX, a marker of newly generated neurons, in DG (2.01-fold, Fig. 4E-H), CA3 (2.16-fold, Fig. 4I-L) and SVZ (1.93-fold, Fig. 4M-P) in the antagomir-27b-treated group. These data indicate that neurogenesis was improved and this process could at least partially explain the improvement of neurological function in antagomir-27b group. Western blotting demonstrated the expression of p-AMPKa2 and p-PKCf/k was significantly enhanced in DG and SVZ from the antagomir-27b group (Fig. 4S,T).

Discussion
In the present study, we report that the inhibition of miR-27b improved neurogenesis in DG and SVZ via activating AMPK, and enhanced post-stroke recovery in a mouse MCAo model, which may become a potential therapeutic option for ischemic stroke.
Ischemic stroke, accounting for~80% of stroke, leads to a high mortality rate and causes irreversible disability. Nowadays, the only FDA-approved cure, recombinant tissue plasminogen activator, has a narrow therapeutic window, which means that only~4% patients receive the treatment in time [38]. Owing to their high metabolism rate, high oxygen consumption and low energy storage, neurons are vulnerable to damage under conditions of ischemia and hypoxia. AMPK is a known as a sensor and regulator of energy balance. When cellular energy supply is low, as signaled by increasing AMP/ATP, AMPK is activated and regulates metabolism by directly targeting metabolic enzymes and by regulating gene transcription.
Whether AMPK plays a beneficial or a detrimental role in ischemic brain has engendered considerable controversy. It is newly recognized that AMPK activation can lead to different outcomes depending on the physiological status of the organism, and AMPK activation can protect the heart, endothelial cells, and smooth muscle cells under certain stress conditions [39,40]. Other groups reported that, short-term AMPK activation improved cell survival since it induced nuclear translocation of forkhead box O3 (FOXO3), while longterm of AMPK activation leads to cell death via stimulating a second activation step of FOXO3 and increasing Bim expression [41]. In our study, we confirmed miR-27b directly targets AMPK mRNA. When expression of miR-27b was inhibited, AMPK activity was stimulated in a mild and gradual manner, without affecting the neuronal survival in the hypoxic condition but improved NSC proliferation and differentiation into neurons in vitro.
Neurogenesis in DG and SVZ is involved in neuronal repair after ischemic stroke. Neuroblasts from the SVZ migrate to the damaged areas and serve as a replacement of the dead neurons and help the restoration of cognition and motor skills [42]. Metformin, a first-line drug prescribed for type 2 diabetes and a well-established AMPK activator, has been proven to exert beneficial effects on stroke patients and to increase hippocampal neurogenesis in preclinical studies via the atypical protein kinase C (aPKC)-CREBbinding protein (CBP) pathway [43][44][45]. However, it has not been clarified whether the direct regulation of AMPK activity by miRNAs would affect neurogenesis. Increasing evidence has suggested that miRNAs participate in the cellular response to ischemic stroke injury [46][47][48]. In our study, inhibition of miR-27b enhanced both AMPK and p-AMPK expression levels, but did not change the p-AMPK/AMPK ratio, which differed from the mechanism of AICAR or metformin. To repair stroke-induced cell damage, a series of energyconsumptive pathways are activated, and an acute and high level of AMPK activation may further contribute to cell death. However, a chronic and low level of AMPK activation might accommodate the brain to metabolic stress and regulate related signaling pathways. In the present study, depressed miR-27b expression improved stroke outcome and spatial memory, and enhanced neurogenesis in DG and SVZ. In both area, the aPKC-CBP pathway was activated, which may explain the newborn neuron promotion, but the mechanism beneath this still needs to be explored.
In conclusion, we reported for the first time that a mild activation of AMPK by anti-miR-27b induced neurogenesis and promoted post-stroke recovery. This study provides a strong experimental basis for the use of AMPK activator after ischemic stroke, which may become a potential therapeutic option.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article. Fig. S1. Antagomir-27b improved CBF after MCAo. Quantitative data of CBF on day 28 after MCAo. The data are expressed as mean AE SEM. n = 5/group. ### P < 0.001 vs Sham; *P < 0.05 vs antagomir-27b (one-way ANOVA followed by post hoc Tukey's test).