Electroacupuncture enhances rehabilitation through miR-181b targeting PirB after ischemic stroke

Recent studies have demonstrated microRNAs (miRNAs) and proteins are beneficial to axon regeneration, which may be involved in Electroacupuncture (EA) therapy against stroke. In this study, we aimed to determine the pivotal role of PirB in EA-produced rehabilitation against ischemic stroke; and to screen and investigate the potential miRNAs directly regulating PirB expression. The results showed EA treatment enhanced axon regeneration and new projections from the corticospinal tract at 28 d after cerebral ischemic reperfusion injury of rats. Then, we found EA decreased pirb mRNA and PirB protein expression in the penumbra within 28 days after reperfusion. The reduction of PirB expression facilitated neurite outgrowth after oxygen-glucose deprivation injury. The miRNA microarray showed the level of twenty kinds of miRNAs changed in the penumbra after EA administration. The bioinformatics study and luciferase assay verified miR-181b directly regulated pirb mRNA expression. EA increased miR-181b levels in the penumbras, and improved neurobehavioral function rehabilitation through miR-181b direct targeting of pirb mRNA to regulate the expression of PirB, RhoA and GAP43. In conclusion, we provide the first evidence that EA enhances rehabilitation against stroke by regulating epigenetic changes to directly act on its targets, such as the miR-181b/PirB/RhoA/GAP43 axis, which is a novel mechanism of EA therapy.

might be closely associated with the down-regulation of Nogo-A and its receptor, NgR1, in the central nervous system (CNS) [20][21][22] . However, whether EA enhances neurobehavioral functional recovery via regulating PirB expression remains to be explored.
Most importantly, recent studies have revealed that a group of microRNAs (miRNAs) are involved in the mechanism employed by EA in the regulation of its target molecules after stoke 23,24 . The miRNAs are a class of endogenous, short (18 to 25 nucleotides) non-coding RNAs that bind to the 3′ untranslated regions (UTRs) of target mRNAs by complementary base pairing and prevent translation or destabilize the mRNAs to mediate their degradation or inhibit their translation 25,26 . The miRNAs can be activated in response to neuronal activity, and therefore, they offer a highly effective means of and play crucial roles in controlling the expression of proteins involved in both the developing and mature brain, specifically during neuronal differentiation 27,28 , axon regeneration 29 and synaptic plasticity 30 . Thus, among the miRNAs with high levels of expression in the brain, those that regulate axon growth and target pirb mRNA in response to EA treatment for ischemic stroke need to be further explored.
To address these questions, the present study aimed to determine the pivotal role of PirB in EA-produced rehabilitation against ischemic stroke, and to screen and investigate the potential miRNAs, which can directly regulate PirB expression in vivo and in vitro study. This study is the first to reveal that EA exerts therapeutic effects through inducing epigenetic changes to regulate its targets, such as the miR-181b/PirB/RhoA/GAP43 axis.

Results
EA enhanced axon regeneration and CST projection after stroke. The axonal tracer biotionylated dextran amine (BDA) was injected into the intact (undamaged hemisphere) motor cortex at 14 d post-MCAO to label the descending axons in the intact hemisphere (Fig. 1b). As shown in Fig. 1a, the Sham group exhibited unilateral CST innervation at 28 d post-MCAO. In the MCAO group, scarce BDA-labeled midline-crossing CST axons extended toward the ventral horn of the spinal gray matter on the denervated side of the cervical cord, which served the affected limb. ANOVA showed a difference among the treatment groups [F(2,15) = 26.93, P < 0.001]. Bonferroni post hoc analyses showed that the amounts of BDA-positive CST axons on the middle and left sides of the spinal cord gray matter (C3-5) were significantly increased in the MCAO + EA group compared with the Sham and MCAO groups (P < 0.05, Fig. 1a and c). ANOVA showed a difference among the treatment groups [F(2,15) = 19.23, P < 0.001]. Post hoc analyses showed EA treatment significantly increased the number of these fibers detected ipsilateral to the injection site compared with that in the MCAO group (P < 0.05, Fig. 1a and d).
Axon damage and regrowth is typically assessed by immunostaining using an anti-NF-200 antibody. In the Sham group, long neuronal processes were detected that were immunoreactive for NF-200. The morphology of the NF-200-immunoreactive neuronal fibers was irregular and fragmented. ANOVA showed a difference among the treatment groups [F(2,15) = 45.38, P < 0.001]. Post hoc analyses showed NF-200 expression was less in the MCAO group compared with that in the Sham group at 28 d after reperfusion (P < 0.05). In addition, the NF-200 expression was significantly increased in the MCAO + EA group compared with the MCAO group at 28 d after reperfusion (P < 0.05, Fig. 1e and f).
The expression of axon growth-associated proteins was also evaluated. RhoA signaling plays an important role under pathological conditions, such as cerebral ischemia reperfusion injury and axon remodeling 30 . ANOVA showed a difference among the treatment groups [F(2,15) = 94.40, P < 0.001]. Post hoc analyses showed RhoA expression was increased in the MCAO group (P< 0.05). However, its expression was inhibited after EA treatment at 28 d post-MCAO (P < 0.05, Fig. 1g and h). Growth-associated protein-43 (GAP43) is a pre-synaptic protein that regulates axon growth and guidance 31 . ANOVA showed a difference among the treatment groups [F(2,15) = 101.7, P < 0.001]. Post hoc analyses showed GAP43 protein expression was decreased in the MCAO group compared with the Sham group at 28 d post-MCAO (P < 0.05). EA treatment increased GAP43 protein expression compared with that in the MCAO group at 28 d post-MCAO (P < 0.05, Fig. 1g and i). These results demonstrated that EA treatment enhanced axon regeneration and the formation of new projections from the CST after stroke.
EA inhibited PirB up-regulation in the ischemic penumbra after stroke. Expression of the pirb gene in the ischemic penumbra was examined via qRT-PCR, as Fig. 2b shown. There was a significant difference among the groups [F(8,75) = 11.72, P < 0.0001; F(2,75) = 161.26, P < 0.0001; F(4,75) = 40.12, P < 0.0001]. Post hoc analyses showed pirb mRNA level had no difference in every group at 1 h before MCAO surgery. The pirb mRNA level was higher in the MCAO groups than that in the Sham groups at 7, 14, 21 and 28 d post-MCAO (P < 0.05). In contrast, the pirb mRNA level was lower in the MCAO + EA groups than that in the MCAO groups at 7, 14, 21 and 28 d post-MCAO (P < 0.05).
Then, Western blot analysis revealed a low level of PirB protein expression in every group at 1 h before MCAO surgery. There was a significant difference among the groups [F(8,75) = 16.94, P < 0.0001; F(2,75) = 209.24, P < 0.0001; F(4,75) = 58.80, P < 0.0001]. Post hoc analyses showed PirB protein expression was higher in the MCAO groups than that in the Sham groups at 7, 14, 21 and 28 d after MCAO (P < 0.05). The result also showed the expression of PirB protein was lower in the MCAO + EA groups than that in the MCAO groups at 7, 14, 21 and 28 d after reperfusion (P < 0.05). (P < 0.05; Fig. 2a and c).
Next, we examined whether neurons in the ischemic penumbra region expressed PirB. We detected the presence of PirB/NeuN double-labeled neurons in this region at 28 d after reperfusion (Fig. 2d); ANOVA showed a difference among the treatment groups [F(2,15) = 32.57, P < 0.001]. Post hoc analyses showed the number of PirB-positive neurons in the MCAO group was greater than that in the Sham group at 28 d after reperfusion. In addition, the number of PirB-positive neurons in the MCAO + EA group was lower than that in the MCAO group pirb mRNA down-regulation facilitated neurite growth in cortical neurons exposed to oxygen-glucose deprivation (OGD) in vitro. We used HEK293 cells to determine the efficacy of the RNA interference (RNAi)-mediated reduction in PirB expression. At 24 h after transient transfection, Western blot analysis indicated that the GFP expression levels were similar between the control RNAi and pirb RNAi groups. However, the PirB expression level in the pirb RNAi group was significantly lower than that in the control RNAi group (Fig. 3a). An in vitro OGD model was used to evaluate the efficacy of pirb RNAi following ischemic injury. At 24 h post-OGD, the Western blot analysis revealed that PirB expression in the OGD and control RNAi groups was increased compared with that in the uninjured group (P < 0.05). However, its expression was significantly decreased in the pirb RNAi group compared with the OGD and control RNAi groups (P < 0.05, Fig. 3b,c).
The neurites of cortical neurons were then examined to assess the effects of pirb RNAi on neurite growth. Under normal culture conditions, cortical neurons contained a single long axon in vitro. ANOVA showed a difference among the treatment groups [F(3,20) = 21.69, P < 0.001]. Post hoc analyses showed the extension of cultured cortical neuronal axons was significantly reduced at 72 h after exposure to OGD (P < 0.05). However, the mean axon length in the pirb RNAi group was longer than those in the OGD and control RNAi groups (P < 0.05, Fig. 3d and e). These results indicated that the reduction in PirB expression enhanced neurite growth after OGD injury in vitro.
EA increased the miR-181b level in the rat ischemic penumbra after stroke. Differential expression of miRNAs was evaluated in the rat ischemic penumbras in the Sham, MCAO and MCAO + EA groups at 28 d post-MCAO. The miRNA microarray analysis resulted in the significant change of 20 differential miR-NAs in the ischemic penumbra of the rats in each group. Among these 20 miRNAs, miR-181b expression was markedly down-regulated in the MCAO group compared with the Sham group ( Fig. 4a and b). In contrast, its expression was increased dramatically in the MCAO + EA group compared with the MCAO group ( Fig. 4c and d). Intriguingly, using computational miRNA TargetScan analysis (http://www.targetscan.org) to search for targets of miR-181b, we found that the 3′ -UTR of pirb mRNA contains two putative miR-181b target sites (Fig. 4e). To determine whether the pirb mRNA is potentially regulated by miR-181b, we constructed a reporter plasmid by inserting the cDNA corresponding to the 3′ -UTR of pirb mRNA into the 3′ region of the firefly luciferase gene. Cells were co-transfected with the firefly luciferase target reporter plasmid containing the pirb 3′ -UTR (with either wild-type or mutated miR-181b binding sites), a Renilla luciferase control reporter, and miR-181b mimics or a negative control. ANOVA showed a difference among the treatment groups [F(3,20) = 7.614, P < 0.001]. Post hoc analyses showed miR-181b caused a significant decrease in luciferase activity (P < 0.05; Fig. 4f), whereas they did not alter the luciferase activity in cells transfected with reporter constructs containing a mutated miR-181b 3′ -UTR. These results suggest that miR-181b may directly regulate pirb mRNA expression and have a very important function in the treatment of cerebral ischemic-reperfusion injury using EA.
Furthermore, mimics or inhibitors of miR-181b were transfected into the primary neuron cultures. ANOVA showed a difference among the treatment groups [F(2,15) = 415.7, P < 0.001]. Post hoc analyses showed miR-181b level in the miR-181b mimics group was higher than that in the Normal group (P < 0.05; Fig. 5h). In contrast, its level was lower in the miR-181b inhibitors group compared with that in the Normal group (P < 0.05; Fig. 5h). Next, miR-181b mimics or miR-181b inhibitors were transfected into cerebral cortex neurons in the ischemic penumbra. ANOVA showed a difference among the treatment groups [F(4,25) = 42.31, P < 0.001]. Post hoc analyses showed that EA treatment or miR-181b mimic transfection increased the miR-181b level at 28 d post-MCAO (P< 0.05; Fig. 5i); however, this effect was reversed by miR-181b inhibitor transfection (P < 0.05; Fig. 5i). These results demonstrate that EA can regulate the miR-181b level in the ischemic penumbra after stroke.
Moreover, neurobehavioral function was evaluated. ANOVA showed a difference among the treatment groups for mNSS test [F(4,25) = 130.0, P < 0.001], for rota-rod test [F(4,25) = 69.40, P < 0.001] and for grip strengths test [F(4,25) = 29.33, P < 0.001]. The mNSS of the MCAO + EA group and MCAO+ miR-181b mimics group were significantly lower than that of the MCAO group at 28 d after reperfusion (P < 0.05); this effect was reversed by miR-181b inhibitor transfection (P < 0.05; Fig. 5d). In addition, the times on the rota-rod and grip strengths were increased in the MCAO + EA group and MCAO+ miR-181b mimics group compared with MCAO group at 28 d after reperfusion (P < 0.05). However, miR-181b inhibitor transfection altered the effects of EA or the miR-181b mimics on the rota-rod and grip strength performances (P < 0.05; Fig. 5e and f). These results suggested that EA enhanced neurobehavioral functional recovery by promoting the targeting of pirb mRNA by miR-181b after ischemic stroke.
MiR-181b targeted pirb mRNA to regulate PirB and its downstream signaling molecules after ischemic-reperfusion injury. To determine whether miR-181b is associated with the expression of PirB and its downstream signaling molecules after ischemic-reperfusion injury, the effects of miR-181b mimics and miR-181b inhibitors on miR-181b expression in a model of OGD were detected in vitro. ANOVA showed a difference among the treatment groups [F(3,20) = 118.3, P < 0.001]. Post hoc analyses showed the miR-181b level was lower in the OGD group compared with the Normal group at 72 h after OGD injury (P < 0.05; Fig. 6a). In addition, significantly increased miR-181b expression was observed in the OGD+ miR-181b mimics group compared with the OGD group (P < 0.05; Fig. 6a). In contrast, decreased miR-181b expression was detected in the OGD+ miR-181b inhibitors group compared with the Normal group (P < 0.05; Fig. 6a).
Next, we assessed the effects of miR-181b mimics and miR-181b inhibitors on pirb mRNA in the model of OGD. ANOVA showed a difference among the treatment groups for pirb mRNA test [F(3,20) = 37.07, P < 0.001]. Decreased pirb mRNA was observed in the OGD+ miR-181b mimics group compared with the OGD group at 72 h after OGD injury (P < 0.05; Fig. 6b and c). In contrast, the expression was higher in the OGD group and OGD+ miR-181b inhibitors group compared with the Normal group at 72 h after OGD injury (P < 0.05; Fig. 6b and c). Moreover, the PirB level and neurite outgrowth were evaluated at 72 h after OGD injury. As shown in the Fig. 6d and e, ANOVA showed a difference among the treatment groups for PirB protein test [F (3,20) = 86.14, P < 0.001]. The analyses showed that transfection of miR-181b mimics into neurons led to a decrease in PirB expression (P < 0.05); in contrast, transfection of miR-181b inhibitors significantly increased its expression (P < 0.05). In addition, ANOVA showed a difference among the treatment groups [F (3,20) = 126.2, P < 0.001].
Post hoc analyses showed the longest neurite length was significantly inhibited in the OGD group compared with the Normal group (P < 0.05). However, the longest neurite length was almost completely restored in the OGD+ miR-181b mimics group compared with the OGD group (P < 0.05). In contrast, the neurite length was significantly reduced in the OGD+ miR-181b inhibitors group compared with the OGD+ miR-181b mimics group (P < 0.05). These findings suggested that miR-181b up-regulation enhanced neurite growth by inhibiting PirB expression after OGD injury.
Furthermore, Western blot analysis tested RhoA expression. ANOVA showed a difference among the treatment groups [F(3,20) = 23.49, P < 0.001]. Post hoc analyses revealed that RhoA expression increased in the OGD group compared with the Normal group (P < 0.05; Fig. 6f and g). However, its expression decreased in the OGD+ miR-181b mimics group compared with the OGD group (P < 0.05; Fig. 6f and g). In contrast, RhoA expression increased in the OGD+ miR-181b inhibitors group compared with the OGD+ miR-181b mimics group (P < 0.05; Fig. 6f and g). Furthermore, Western blot analysis tested GAP43 expression. ANOVA showed a difference among the treatment groups [F(3,20) = 10.08, P < 0.001]. Post hoc analyses revealed GAP43 expression was reduced in the OGD group compared with the Normal group (P < 0.05; Fig. 7h and i). In contrast, GAP43 expression increased in the OGD+ miR-181b mimics group compared with the OGD group (P < 0.05; Fig. 6h and i), and decreased in the OGD+ miR-181b inhibitors group compared with the OGD+ miR-181b mimics group (P < 0.05; Fig. 6h and i). These results suggested that miR-181b regulated the expression of PirB and its downstream signaling molecules, thereby restricting neurite growth after ischemic injury. This may be the mechanism underlying the EA-produced rehabilitation of axon regeneration, CST projection and neurobehavioral functional recovery, as shown in Fig. 7.

Discussion
Acupuncture has been used for stroke recovery in East Asia for centuries. The positive effects of acupuncture have been reported in studies of its application to treat neurological disorders 7,32 . Specifically, the use of acupuncture or EA for the treatment of stroke has been shown to be effective for improving brain damage and motor function 8,33 . However, its effect and mechanism on motor rehabilitation remain to be clearly elucidated in animal studies 34,35 . After transient focal cerebral ischemia injury, the extents of brain and behavioral recovery are influenced by neural circuit plasticity, which includes axon regeneration and remodeling, particularly rewiring and the formation of new projections from the CST after stroke 12,13 . In the present study, we demonstrated that EA treatment increased the number of BDA-positive CST axons projecting from the undamaged CST to the middle or left side of the spinal cord gray matter (C3-5) at 28 d post-MCO. We also found that EA treatment enhanced NF-200 and GAP43 expression, whereas it inhibited RhoA expression, at 28 d post-MCAO. Previous studies have demonstrated that neurons in the intact cortex can extend their axons into damaged regions, which are amenable to motor functional recovery 13 . Furthermore, the sprouting and growth of intracortical axons plays important roles after cerebral ischemia-reperfusion injury 15,16 . Other researchers have suggested that neurofilaments expressed in axons are among the most important structural substrates of the axonal cytoskeleton, which is necessary for axonal maintenance and regeneration 36,37 . These results support the view that EA treatment exerts rehabilitation effect by enhancing axon regeneration and CST projection after stroke.
PirB is a type I transmembrane glycoprotein consisting of an extracellular portion that contains six Ig-like domains, a hydrophobic transmembrane segment and a cytoplasmic portion that contains three immunoreceptor tyrosine-based inhibitory motifs (ITIMs) and one ITIM-like sequence 17 . Atwal et al. first found that the inhibition of PirB activity using a neutralizing antibody alleviated the suppression of axon growth and growth cone collapse and demonstrated that PirB is a novel functional receptor for Nogo-66, MAG, and Omgp for inhibiting axon growth and regeneration 17 . Fujita et al. further found that the binding of PirB to Nogo, OMgp or MAG induced downstream signals, for example, activation and delivery of the inhibitory signal for neurite outgrowth 18 . Our previous study has revealed that neuronal PirB expression is up-regulated in the ischemic penumbras of mice from 24 h to 7 d after reperfusion 19 . In this study, we found that the pirb mRNA and PirB protein expression were decreased following EA treatment in the ischemic penumbra from 7 to 28 d after stroke. The results provided the evidence that EA administration can effectively inhibit PirB up-regulation in the ischemic penumbra after stroke. Recent studies have demonstrated that PirB contributes to stabilization of the adult neuronal circuitry and motor function by inhibiting axon regeneration and synaptic plasticity 38,39 . Moreover, pirb knockout mice exhibit increased crossed projections of CST axons and improved motor performance 15 . The down-regulation of PirB expression in cortical neurons has been shown to result in increased neurite outgrowth 40 . These previous findings are consistent with our current study. Thus, it is plausible to deduce that the significant up-regulation of PirB expression in the ischemic penumbra may be one of crucial "troublemaker" that exacerbates brain damage and restricts motor recovery following cerebral ischemia-reperfusion injury. Moreover, EA enhanced axon regeneration via the down-regulation of PirB expression, indicating that PirB is a novel therapeutic target of EA for promoting neurobehavioral functional recovery following ischemic stroke.
Recently, a large number of reports have been published on the involvement of miRNAs in the response to EA therapy against cerebral ischemia reperfusion injury 40 . In particular, the relevance of miR-181b function to the organization of neural connections has been reported and identified. MiR-181b, a member of the miR-181 family, is expressed at intriguingly high levels in the retina and brain areas associated with motor function 29 . Furthermore, increasing evidence shows that miR-181b governs axon specification and growth by regulating its target signaling molecules and plays an important role in the organization of neural connections 29,41 . However, the role of miR-181b in the ischemic penumbra in association with EA treatment for stroke remains to be elucidated. Interestingly, miRNA microarray analysis revealed that miR-181b expression was significantly decreased in the ischemic penumbra at 28 d post-MCAO, whereas EA treatment enhanced its expression. A similar changing trend was identified by qRT-PCR after reperfusion. Furthermore, these results provide the first evidence that EA directly regulates miR-181b expression following administration of miR-181b mimics or miR-181b inhibitors. These results suggest that an increase in the brain-specific miR-181b level in the ischemic penumbra may be directly responsible for the mechanical repair of brain tissue, and they also imply that the rehabilitative effects of EA administration may include promoting miR-181b expression after cerebral ischemic injury.
To further elucidate the underlying molecular mechanisms of EA-produced neurobehavioral functional recovery, we searched for a possible relationship between miR-181b and pirb mRNA, which may be the miR-181b target that participates in this process. Computational miRNA target analysis revealed that the pirb 3′ -UTR was a predicted target of miR-181b. Luciferase assay further demonstrated that PirB was likely a direct target of miR-181b. The results of these investigations increase the understanding of the mechanisms underlying the Figure 7. Hypothetical model depicting epigenetic mechanisms underlying the therapeutic effects of EA against ischemic stroke. EA enhances neurobehavioral functional recovery from ischemic stroke through a series of complex processes, which may include increasing the miR-181b level, decreasing PirB expression, and regulating the levels of PirB downstream signaling molecules (such as RhoA and GAP43). The findings of this study indicate that EA exerts its therapeutic effects by inducing epigenetic changes to directly regulate its targets, such as the miR-181b/PirB/RhoA/GAP43 axis.
up-regulation of miR-181b expression and down-regulation of PirB expression by EA in the ischemic penumbra. Our results also showed that EA treatment and miR-181b mimic administration had similar effects on pirb mRNA and PirB protein expression. In contrast, the miR-181b inhibitors increased the pirb mRNA and PirB protein levels. These findings verify that PirB is a direct target of miR-181b and demonstrate that EA enhances neurobehavioral performance recovery through miR-181b-mediatedpirb gene silencing after ischemic stroke.
To explore the mechanisms of the miR-181b-mediated regulation of PirB expression and neurite growth, an OGD model was used. To the best of our knowledge, our study is the first to demonstrate a decrease in the miR-181b level at 72 h after OGD injury. Moreover, the decreased miR-181b expression inhibited neurite growth after OGD injury, which could be rescued by miR-181b mimics. One possible explanation for these findings is that miR-181b increased pirb mRNA and PirB protein expression to regulate the expression of downstream signaling molecules, such as RhoA and GAP43. These results are consistent with those of the above in vivo study, in addition to a previous report demonstrating that PirB inhibition promotes the regeneration of hypoxic-ischemic-damaged axons and that this inhibitory signal may be transduced via the Rho-ROCK signaling pathway 29,41 . Moreover, GAP43 is another downstream molecule of PirB that facilitates axonal sprouting and outgrowth 16 . These results indicate that miR-181b targets pirb mRNA to regulate the expression of PirB and its downstream signaling molecules after ischemic-reperfusion injury.
In conclusion, as shown in Fig. 7, this study first provides and verifies the solid evidences that EA treatment enhances neurobehavioral functional recovery through miR-181b targeting PirB after ischemic stroke, suggesting that miR-181b and PirB act as key regulators of axon regeneration and CST projection after cerebral ischemic-reperfusion injury. These findings indicate that EA exerts its therapeutic effects on stroke by regulating epigenetic changes to directly act on its targets, such as the miR-181b/PirB/RhoA/GAP43 axis, which may be a novel mechanism underlying EA treatment against ischemic stroke.

Materials and Methods
Animals. Male Sprague-Dawley rats (280-300 g) were provided by the Experimental Animal Center of the Fourth Military Medical University and were housed under controlled conditions, including a 12-h light/dark cycle, a temperature of 21 ± 2 °C and humidity of 60-70%, for at least one week prior to receiving drug treatment or surgery. The animals were allowed free access to standard rodent food and tap water. The experimental protocol used in this study was conducted according to the Guidelines for Animal Experimentation of the Fourth Military Medical University and was approved by the Ethics Committee for Animal Experimentation of the Fourth Military Medical University (Xi'an, Shaanxi province, China).

Middle cerebral artery occlusion (MCAO) model.
The focal cerebral ischemic model of rat was induced by MCAO as described previously 42 . Briefly, the rats were anesthetized with 2% chloral hydrate. The right middle cerebral artery of the rats was occluded using a 3-0 nylon monofilament suture (Ethicon Nylon Suture, Ethicon Inc., Sukagawa, Japan). Regional cerebral blood flow (rCBF) was monitored using a disposable microtip fiberoptic probe (diameter, 0.5 mm) connected through a master probe to a computerized laser Doppler main unit (PeriFlux 5000, Perimed AB, Sweden). Rats retaining > 20% of baseline perfusion during ischemia were excluded. The temporalis muscle temperature was monitored and was maintained at 37.0 °C-37.5 °C by surface heating or cooling during surgery. After 120 min of ischemia, reperfusion was done. At 120 min after reperfusion, the rats received EA treatment. After that, rats were then allowed to recover before being returned to their home cages. EA treatment. The animals were anesthetized with 2% chloral hydrate, and then the "Baihui (GV 20)" acupoint 42 , which is located at the intersection of the sagittal midline and the line linking the two rat ears, was stimulated electrically at an intensity of 1-2 mA and dense-disperse frequency of 2/10 Hz. A fine needle (0.5 mm in diameter) placed at GV 20 was connected to one electrode of a HWATO electronic acupuncture treatment instrument (model No. SDZ-V, Suzhou Medical Appliances Co., Ltd., Suzhou, China). Individual EA sessions were administered for 30 min per day for 5 successive days, followed by 2 days of rest, for a period of 4 weeks. The core temperatures of the rats receiving the anesthesia and EA treatment were maintained at 37.0 ± 0.5 °C by surface heating or cooling (Spacelabs Medical Inc.).

Retrograde tracing of the CST.
Rats which had been operated successfully were randomly divided into the following groups: Sham, MCAO and MCAO + EA group (n = 6 for each group). After reperfusion, the rats received EA treatment. At 14 d after reperfusion, the rats were anesthetized using 2% chloral hydrate, and a large craniotomy was performed to expose the left hemisphere of the brain. Surgeries were performed using a Kopf stereotaxic setup and procedures adapted from previous study 13 . The cortex overlying the injection area was exposed by removing the skull using a dental drill. Then, BDA (NeuroTrace Neuronal Tracer Kit, Molecular Probes, Eugene, OR) was injected using a Nanoject injector (Drummond Scientific, Broomall, PA) at different sites, with the delivery of 100 nl BDA at three depths per site (0.5, 1.0, and 1.5 mm deep from the cortical surface). The injection site coordinates were as follows (in mm): 3.8 L/1.2A, 3.4 L/0.7A, 3.4 L/0.2A, 3.6 L/0.3P, 3.4 L/1.3P, 2.2 L/1/3P, 2.2 L/2.3P, 2.2 L/2.8P, and 3.0 L/3.3P. These coordinates are described with respect to bregma. Each 100-nl injection was performed over a 1-min period, such that the time required to inject each site was approximately 4-5 min. After completion of the injections, the scalp was sutured. The body temperature was maintained at 37 °C using a warming blanket.
After 28 d after reperfusion, the rats were deeply anesthetized and perfused, and coronal brain sections were obtained. Selected sections through the cervical region of the spinal cord were prepared as described by Oswald Steward. BDA was stained using Streptavidin-FITC (1:100, eBioscience). BDA-labeled axons in the dorsal CST ipsilateral to the cortex of origin (left hemisphere) were counted in four sections in the cervical enlargement.

Quantitative real-time-polymerase chain reaction (qRT-PCR). Rats which had been operated
successfully were randomly divided into the following groups: Sham, MCAO, MCAO + EA group (n = 6 for each group). After reperfusion, the rats received EA treatment. At 1 h before MCAO, 7 d post-MCAO, 14 d post-MCAO, 21 d post-MCAO, 28 d post-MCAO, the rats were deeply anesthetized and obtained brain tissue. The pirb mRNA and miR-181b expression were tested by qRT-PCR analysis. The ischemic penumbra of the cerebral cortex at the Bregma level + 2.0 to − 5.0 mm was harvested as previously described 42 . The total RNA was isolated from each sample using RNAiso Plus (TaKaRa) according to a standard protocol and subsequently quantified. cDNA was synthesized using the Superscript First-Strand Synthesis Kit (Promega, Madison, WI) according to the manufacturer's instructions. PCR was performed under the following thermal cycling conditions: one cycle at 94 °C for 5 min; 25 cycles at 94 °C for 30 s, 55 °C for 60 s, 72 °C for 30 s; and one cycle at 72 °C for 10 min. The following primers were designed by TaKaRa Corp.: PirB (Sense 5′ -CATGTCATGGGGACTGCCCT-3′ , Antisense 5′ -CCAGAGTCATTCCAAGACAGAGC-3′ ); β -actin (Sense 5′ -TCAGGTCATCACTATCGGCAAT3-′ , Antisense 5′ -AAAGAAAGGGTGTAAAACGCA-3′ ). To perform qRT-PCR of miR-181b, a small RNA fraction was isolated with a mirVana miRNA Isolation Kit (Genetimes Technology, Inc.), and 200 pg miRNA were used for each reaction. qRT-PCR analysis was performed using an Agilent Technologies Stratagene Mx3000P Real-Time PCR system (Genetimes Technology, Inc.) with real-time SYBR Green PCR technology. The threshold cycle (Ct) value was determined using the automatic baseline determination feature of Stratagene MX3000. Each sample was tested in triplicate. The samples were obtained from three independent experiments and were analyzed for relative gene expression using the 2-Δ Δ CT method. The efficiency of each primer was determined and used to quantify the expression level as follows: Expression constructs and RNAi. The short interfering RNA (siRNA) expression vector pLenR-GIP-GFP was used, and the PirB (Lilrb3, NM_031713.1) siRNA sequence was 5′ GCTGGAAAGTTATGTGAATGC3′ . Western blot analysis confirmed the knockdown of endogenous PirB in the transfected HEK293 cells and primary cortical neurons. Briefly, HEK293 cells were transiently transfected with a control RNAi construct or PirB expression construct, which also expressed GFP. The expression of PirB, GFP and β -actin in cell was determined by Western blot analysis using the corresponding specific antibodies (anti-PirB rabbit monoclonal antibody, 1:1000, Abcam; anti-GFP rabbit monoclonal antibody, 1:1000, Abcam; and anti-β -actin rabbit monoclonal antibody, 1:2000, Abcam, respectively) at 24 h after transfection. The secondary antibodies were used as previously described.
Primary cortical neurons culture and oxygen glucose deprivation (OGD) model. Primary cortical neurons were cultured as previously described 43 . In brief, neurons were isolated from SD rat embryos (E18.5 d), washed with D-Hank's solution three times under sterile conditions and seeded at a density of 1 × 10 5 cells/cm 2 on plates coated with poly-L-lysine (50 mg/mL) (Sigma, USA). The cells were cultured in Neurobasal medium (Gibco, Invitrogen Corp., USA), which was supplemented with 2% B27, 1% glutamine, and 1% penicillin/streptomycin (Sigma, USA), at 37 °C in a humidified incubator in air that contained 5% CO 2 . The purity of the neurons was determined via immunocytochemistry for β -III-tubulin (1:250; Millipore) at 7 d after seeding, which indicated that 95% of the culture cells were positive for β -III-tubulin (data not shown).
The OGD model was used as previously described 44 . In brief, cortical neurons were cultured in Neurobasal medium that lacked glucose (Invitrogen, Carlsbad, CA, USA) in a humidified incubator which contained an anoxic gas mixture (5% CO 2 and 95% N 2 ) at 37 °C for 1 h; the neurons were subsequently cultured in normal medium. The cells were transiently transfected with the control RNAi construct or the pirb RNAi construct. The cultures were divided into groups as follows: Normal, OGD, OGD+ control RNAi and OGD+ pirb RNAi group.
To evaluate the effect of pirb RNAi on PirB expression and neurite growth after OGD injury, the expression of PirB, GFP and β -actin in the extracts was determined via Western blot analysis at 24 h after transfection. Then, cells were subjected to immunocytochemistry staining at 72 h after transfection. The cells were washed after fixation in 4% PFA and subsequently stained with a rabbit monoclonal anti-Tau antibody (1:1000, Abcam). The measurement of neurite length was performed as follows. Five randomly selected fields of view on the slides were photographed using a phase-contrast Olympus IMT2 microscope and an F-View camera. The average length of the longest neurites was measured as the staining for Tau (1:500; Millipore), which labels axons. The average lengths of the longest neurites from three independent experiments were analyzed using Student's t-test, as previously described 44 .
MicroRNA microarray analysis. Rats which had been operated successfully were randomly divided into the following groups: Sham, MCAO, MCAO + EA group (n = 6 for each group). After reperfusion, the rats received EA treatment. At 28 d post-MCAO, RNA was isolated from each cerebral ischemic penumbra using the mirVana miRNA Isolation Kit (Ambion, USA). The miRNA microarray was implemented according to MicroRNA Expression Profiling Assay Guide (Illumina Inc., USA). After adding a stretch of poly A tail to the 3′ -end of each sequence in the 1000 ng intact total RNA sample. Then the biotinylated cDNAs were linked with miRNA specific oligos in an assay specific extension plate. The oligos were hybridized to the cDNA, mis-hybridized and excess oligos were washed away. In this process, the BeadChips were hybridized using the hybridization chamber. The BeadChips were hybridized overnight in the Illumina hybridization oven, with a temperature ramp from 60 °C to 45 °C. The Illumina BeadArray Reader (Illumina Inc.) used a laser to excite the fluor of the single-base extension product on the beads of the BeadChip sections.
Dual-luciferase reporter assay. HEK293T cells were plated at a density of 5-6 × 10 4 cells/well in 96-well plates one day before transfection. The cells were co-transfected with 0.2 μ g of a firefly luciferase target reporter plasmid, the pMIR-REPORT PirB 3′ -UTR plasmid (with either wild-type or mutated miR-181b-binding sites), 0.01 μ g of a Renilla luciferase control reporter, and 100 nM miRNA mimics or negative control using Lipofectamine2000 (Invitrogen, California, USA), according to the manufacturer's instructions. At 6 h post-transfection, the medium in each well was replaced with fresh culture medium. Cells were harvested at 48 h post-transfection and were assayed using a Dual Luciferase System (E1910, Promega). The results are expressed relative to luciferase activity; the values were first normalized to the Renilla luciferase transfection control and then to the firefly/Renilla value of the empty control vector and finally to the corresponding seed mutant reporter control.

Drug administration.
To test whether EA could enhance neurobehavioral functional recovery through regulating miR-181b expression and whether miR-181b could target pirb gene silencing after ischemic stroke, the rats received a subcutaneous Alzet osmotic minipump (Alzet, Cupertino, CA, USA) 45,46 . The osmotic minipump was implanted beneath the skin via mid-scapular incision. Each pump contained sterile 0.9% saline with miR-181b mimics or miR-181b inhibitors in dimethylsulfoxide (DMSO), which were administered to the animals continuously until the study end (delivering 2 ng/day for 28 days). The miR-181b mimics or miR-181b inhibitors were purchased from GenePharma. After minipump implantation, the rats received EA administration as above method. Rats which had been operated successfully were randomly divided into the following groups: Sham, MCAO, MCAO + EA group, MCAO + EA group+ miR-181b inhibitors, MCAO+ miR-181b mimics (n = 6 for each group). After reperfusion, the rats received EA treatment. At 28 d post-MCAO, the mNSS, Rota-rod test and Grip strength were tested. The expression of miR-181b, pirb mRNA and PirB protein were also performed at 28 d after reperfusion.
Neurobehavioral assessments. Neurobehavioral tests were performed by experienced testers who were blind to the experimental groups. Performance on the neurobehavioral tests was assessed at baseline and on different days post-stroke. The following behavioral assessments were used in the present study: ① calculation of the mNSS 47 ; ② the rota-rod test 48 ; and ③ the grip strengthtest 49 . The mNSS system 47 , a composite of motor, sensory, reflex and balance tests, is the most commonly used neurological scoring system in animal studies of focal cerebral ischemia. Neurological deficit was graded on a scale ranging from 0 to 18, with a higher score indicating worse recovery. Baseline performance was determined at 1 day before ischemia. The mNSSs were determined at 28 d after reperfusion. Rota-rod test. The rota-rod test was used to evaluate the motor performances of the rats 48 . Three days before MCAO, the rats were trained on an accelerating (4-35 rpm) rota-rod (7750 Ugo Basile, Italy). All animals received a 3-day training program with 3 sessions per day, each lasting for 5 min. On the experimental day, the time spent walking on the rota-rod without falling was measured twice per animal. The interval between each trial was 15 min. The mean time from two trials was calculated for each rat. Baseline performance was determined at 1 day before ischemia. The motor performance was evaluated at 28 days post-MCAO. Grip strength. To test the rats' forelimb grip strengths, an electronic digital force gauge grip strength meter (Columbus Instruments) was used as previously described 49 . Baseline performance was determined at 1 day before ischemia.