Acupuncture and neuroregeneration in ischemic stroke.

Acupuncture is potentially beneficial for post-stroke rehabilitation and is considered a promising preventive strategy for stroke. Electroacupuncture pretreatment or treatment after ischemic stroke by using appropriate electroacupuncture parameters generates neuroprotective and neuroregenerative effects that increase cerebral blood flow, regulate oxidative stress, attenuate glutamate excitotoxicity, maintain blood-brain barrier integrity, inhibit apoptosis, increase growth factor production, and induce cerebral ischemic tolerance.

Introduction

Ischemic stroke is a major cause of mortality and disability worldwide. Acupuncture, one of the most important components of traditional Chinese medicine, has been shown to activate relevant brain regions, modulate cerebral blood flow (CBF), and regulate multiple molecules and signaling pathways that lead to excitotoxicity, oxidative stress, inflammation, neuronal death, and survival after interruption of blood supply. A large number of animal experiments revealed the neuroprotective effects of acupuncture on ischemic stroke (Feng and Zhang, 2014). Furthermore, acupuncture promotes neurogenesis, angiogenesis, and neural plasticity, in addition to inhibiting apoptosis after ischemic damage. Clinical and laboratory evidence suggests that acupuncture induces multilevel regulation through complex mechanisms against cerebral ischemia (Zhu et al., 2017).

Acupuncture serves not only as a complementary and alternative therapy for poststroke rehabilitation but also as a promising preventive strategy in stroke, which could induce cerebral ischemic tolerance, especially when combined with modern electrotherapy (Li and Wang, 2013). Electroacupuncture (EA) possesses additional characteristics such as acupoint specificity and stimulation parameters, which produce different effects against cerebral ischemia.

EA Pretreatment Induces Tolerance to Cerebral Ischemia

EA pretreatment has been shown to induce ischemic tolerance and may be a promising preventive strategy for patients with high risk of acute ischemic stroke. Many studies have shown that the protective mechanisms of EA pretreatment may involve a series of regulatory molecular pathways including enhancement of antioxidants, regulation of the endocannabinoid system, and inhibition of apoptosis (Li et al., 2012).

Several studies have demonstrated the involvement of different cannabinoid receptors in different types of ischemic tolerance. Pretreatment with EA induces rapid (2 hours after EA) and delayed (24 hours after EA) tolerance to focal cerebral ischemia induced by middle cerebral artery occlusion (MCAo) in rats (Zhang et al., 2003). Cannabinoid receptor type 1 (CB1) was thought to be involved in rapid ischemic tolerance, whereas cannabinoid receptor type 2 (CB2) was considered to contribute to the delayed neuroprotective effect (Ma et al., 2011). Pretreatment with EA in an animal model of focal cerebral ischemia reduced infarct size, improved neurological outcome, and inhibited neuronal apoptosis. EA pretreatment upregulated the neuronal expression of CB1 and increased the production of endocannabinoid 2-arachidonylglycerol and N-arach-idonoylethanolamine-anandamide in rat brains, thereby inducing rapid tolerance to focal cerebral ischemia (Wang et al., 2009). EA pretreatment protects against cerebral ischemia/reperfusion (I/R) injury induced by MCAo in rats through CB1 receptor (CB1R)-mediated phosphorylation of glycogen synthase kinase 3 (Wei et al., 2014). In addition, pretreatment with EA may activate endogenous epsilon protein kinase C-mediated antiapoptosis to protect against ischemic damage after focal cerebral ischemia through CB1-induced rapid tolerance to focal cerebral ischemia in rats (Wang et al., 2011). EA pretreatment (15 Hz) at the Baihui (GV20) acupoint could induce rapid tolerance to focal cerebral ischemia (Wang et al., 2005). Additionally, repeated EA pretreatment at GV20 stimulates the release of enkephalins, which may bind to delta- and micro-opioid receptors and induce delayed cerebral ischemic tolerance (Xiong et al., 2007).

EA Pretreatment Regulates Oxidative Stress, Maintains the Integrity of the Blood-Brain Barrier (BBB), and Inhibits Apoptosis

Nitric oxide biosynthesis is a key factor in the pathophysiological response of the brain to hypoxia-ischemia. Brain ischemia activates Ca2+-dependent nitric oxide synthase (NOS) isoforms, namely neuronal NOS (nNOS) and endothelial NOS (eNOS). Although eNOS appears to have neuroprotective properties, nNOS may have neurotoxic effects (Bolaños and Almeida, 1999). Furthermore, delayed ischemia or reperfusion after an ischemic episode induces the expression of Ca2+-independent inducible NOS (iNOS), which may have neurotoxic effects, mainly in glial cells (Garry et al., 2015).

Abrupt reperfusion after ischemia results in overproduction of reactive oxygen species (ROS), which leads to brain injury. Preischemia EA therapy at either Fengchi (GB20) or Zusanli (ST36), similarly, attenuates lipid peroxidation and reduces ROS production, consequently improving the function of the respiratory chain and the antioxidant capacity in the ischemic penumbra (Siu et al., 2004b; Zhong et al., 2009).

EA pretreatment at GV20 in diabetic mice with cerebral I/R injury reduced infarct size and improved neurological outcomes. EA attenuated cerebral ischemic injury by inhibiting nicotinamide adenine dinucleotide phosphate (NADPH) oxidase-mediated oxidative damage (Guo et al., 2014).

EA preconditioning (2 Hz/15 Hz) at GV20 and Dazhui (GV14) can protect the ischemic cerebral cortex tissue from injury in cerebral I/R rats; this may be related to its effects in downregulating the expression of nNOS and iNOS and upregulating the expression of glial fibrillary acidic protein (GFAP) (Zhang et al., 2015c).

EA pretreatment at GV20 significantly reduced BBB permeability and brain edema (Zou et al., 2015). Furthermore, EA pretreatment could alleviate brain edema and BBB dysfunction caused by cerebral ischemia by reducing matrix metalloproteinase-9 (MMP-9) levels (Dong et al., 2009). EA pretreatment significantly attenuated neuronal apoptosis, inhibited caspase-3 activity in the hippocampal CA (Cornu Ammonis) 1 region, and ameliorated learning and memory function in rats exposed to high-sustained positive acceleration (Feng et al., 2010). Moreover, excessive activation of N-methyl-D-aspartate (NMDA) glutamate receptors contributes to neuronal death after stroke. EA pretreatment (1.7 Hz, 1 mA) at GV20, Shenshu (BL23), and ST36 can suppress the increase in hippocampal glutamate content and downregulate NMDA receptor subunit 1 (NR1) mRNA expression in rats with vascular dementia established by MCAo (Meng et al., 2008). EA treatment also reversed the high expression of NR1 and upregulated the level of tropomyosin receptor kinase A (TrkA) in a MCAo rat model; this effect was mediated by the stimulation of the phosphatidylinositol 3-kinase (PI3K) pathway but not the extracellular signal-regulated kinases (ERKs)/mitogen-activated protein kinase (MAPK) pathway. Therefore, EA pretreatment attenuates glutamate excitotoxicity by modulating PI3K pathway (Sun et al., 2005).

EA Pretreatment Influences Growth Factors Following Cerebral Injury

Stem cells from the subventricular and subgranular zones or those recruited from the bone marrow (BM) through peripheral circulation possess a definitive role in neuronal regeneration following cerebral injury. Some of the most commonly explored growth factors include vascular endothelial growth factor (VEGF), nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), and insulin growth factor-1 (IGF-1), each of which may activate ischemic brain endogenous repair through specific mechanisms (Dailey et al., 2013).

Among the neurotrophic factors, BDNF and stromal cell-derived factor-1α (SDF-1α) are considered to be potent candidates for the recovery from cerebral ischemia. SDF-1α resulted in neuroprotection against neurotoxic insult, and it induced BM-derived cell targeting in the ischemic brain, thereby reducing the volume of cerebral infarction and improving neural plasticity (Shyu et al., 2008).

EA preconditioning (2 Hz, at GV20 and GV14, for 20 minutes) reduced infarct volume in mice with cortical ischemia, leading to prominent improvement of neurological function. Pretreatment with EA also increased the production of BDNF and SDF-1α, which elicited protective effects against focal cerebral ischemia (Kim et al., 2013a).

Acupuncture therapy was reported to produce alterations in the levels of growth factors such as glial cell line derived neurotrophic factor (GDNF), BDNF, and VEGF (Peplow and Martinez, 2016).

Neuroglobin (NgB) is involved in cellular oxygen homeostasis. It increases oxygen availability to brain tissue and provides protection under hypoxic or ischemic conditions, potentially limiting brain damage. In other words, NgB enhances cell viability under hypoxia and under various types of oxidative stress, and it is beneficial for neurons (Burmester and Hankeln, 2009). Applying EA preconditioning to GV20 has a significant neuroprotective effect on cerebral ischemia-reperfusion, which is closely related to the upregulation of cerebral NgB expression (Xie et al., 2012).

EA pretreatment at GV20 enhanced the neuronal expression of hypoxia-inducible factor-1α (HIF-1α), reduced infarct volume, improved neurological outcome, inhibited neuronal apoptosis, upregulated the expression of B-cell lymphoma 2 (Bcl-2), and downregulated the expression of Bcl-2-associated X protein (Bax) after reperfusion in the penumbra (Zhao et al., 2015).

Neuroprotective Effect of EA after Ischemic Stroke

EA was applied to Renzhong (GV26) and Neiguan (PC6) acupoints for 30 minutes, starting immediately after the onset of reperfusion in a MCAo rat model, leading to a significant reduction in ischemic infarction and neurological deficits, upregulation of delta-opioid receptor expression, and protection of the brain from I/R injury (Tian et al., 2008).

EA at GV20 and GV26 reduced neurological deficit, brain swelling, and infarct area; it also increased the percentages of residual cells in the ipsilateral striatum and cortex, and facilitated electroencephalogram recovery following MCAo in monkeys (Gao et al., 2002). EA was found to promote the recovery of neurological function in patients with acute ischemic stroke and somatosensory evoked potential in rats with MCAo (Si et al., 1998).

A study assessed the hypothesis that EA can enhance cerebral glucose metabolism by using 18F fluorodeoxyglucose/positron emission tomography (PET) imaging in rats subjected to I/R injury. After EA treatment at the Quchi (LI11) and ST36 acupoints, T2 weighted imaging revealed a significant reduction in infarct volume, PET imaging of glucose metabolism in the caudate putamen, motor cortex, and somatosensory cortex regions was promoted by EA, accompanied by functional recovery in Catwalk and Rota-rod performance; moreover, EA could promote adenosine monophosphate (AMP) activated protein kinase α (AMPKα) phosphorylation of these regions to enhance neural activity and motor functional recovery after ischemic stroke (Wu et al., 2017).

EA stimulation applied to GV20 and GV14 significantly reduced infarct volume, brain water content, and neuronal injury in rats with MCAo. EA exerts neuroprotective effects against I/R injury by attenuating inflammatory cytokines, upregulating antioxidant systems, and reducing excitotoxicity (Shen et al., 2016).

EA Increases Cerebral Blood Flow in Cerebral Ischemia

In our previous study, we found that 2 or 15 Hz EA at ST36 (bilaterally) could increase CBF in rats with and without cerebral ischemia (Hsieh et al., 2006). In addition, we observed that neither 2 nor 15 Hz EA influenced the expression levels of nitric oxide (NO) in peripheral blood and calcitonin gene-related peptide (CGRP) in the cerebral cortex and thalamus.

EA stimulation at GV26 could promote the proliferation of vascular endothelial cells and increase regional CBF, indicating that EA might promote angiogenesis under cerebral ischemic conditions (Du et al., 2011).

Influence of EA on ROS after Ischemic Stroke

EA treatment at GV26 and GV20 might markedly reduce the neurological deficit score, promote respiratory enzyme activity, and reduce ROS generation, consequently improving respiratory chain function and the antioxidative capability of brain tissues in the infarct penumbra zone (Zhong et al., 2009). These mechanisms may account for the anti-injury effect of EA on brain function in rats with MCAo.

EA treatment at GB20 was suggested to alleviate lipid peroxidation after cerebral I/R injury by promoting the activities of superoxide dismutase and glutathione peroxidase (Siu et al., 2004a). In a previous study, the effect of EA at GV26 and Chengjiang (CV24) on NOS in rats with cerebral I/R injury was explored. The study revealed that the abnormally increased expression levels of nNOS and iNOS were reversed partly through the TrkA/PI3K-mediated signal transduction pathway (Chen et al., 2011).

EA stimulation at GV20 and GV14 in mice after MCAo significantly reduced infarct volume and increased cerebral perfusion in the cerebral cortex, consistent with a prominent improvement in neurological function and vestibule-motor function. EA stimulation after moderate focal cerebral ischemia, but not severe ischemia, improves tissue and functional recovery. Acetylcholine (Ach)/eNOS-mediated perfusion augmentation might be related to these beneficial effects of EA by interventions in acute ischemic injury (Kim et al., 2013b).

Cholinergic Anti-Inflammatory Pathway in Attenuation of Neuroinflammation Suggestions and Advises for the Most Promising Further Direction of the Study

EA stimulation in an experimental stroke model improved cerebral perfusion, thus reducing infarct volume and hindering apoptosis, neuronal and peripheral inflammation, and oxidative stress. Furthermore, a dramatically lower reduction in mRNA levels of choline acetyltransferase and α7 nicotinic Ach receptors (α7nAChR) was detected, suggesting the inhibition of central cholinergic system impairment. EA also activated the dorsal motor nucleus of the vagus, thereby implying its role as an alternative modality of parasympathetic nervous system activation for stroke therapy (Chi et al., 2017). A similar result was reported wherein EA at the GV20 and Shenting (GV24) acupoints activated the expression of α7nAChR in the hippocampus; EA treatment also led to decreased production of the inflammatory cytokines tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β), leading to a reduction in neuroinflammatory response (Liu et al., 2017). Our previous study investigated the effect of electric stimulation (ES) of the ears; 2 Hz ES of the ears ameliorated learning and memory impairment and increased the number of α4nAChR in the hippocampus in rats with I/R injury (Kuo et al., 2016). These findings suggest that ES (2 Hz) of the ears exerts neuroprotective effects that are related to acetylcholine release.

Stimulation of acupuncture points on the extremities has been demonstrated to result in stimulation of the vagus nerve. In our previous study, 2 Hz EA stimulation at bilateral ST36 and Shangjuxu (ST37) acupoints reduced pulse rate, which indicated that such an application might enhance parasympathetic activity (Hsieh et al., 1999). The auricular branch of the vagus nerve is the only peripheral branch of the vagus nerve. Auricular electrical stimulation on the concha of the ear induced an increase in vagal activity (La Marca et al., 2010). Cholinergic anti-inflammatory pathway is a physiological mechanism by which central nervous system regulates immune response and controls inflammation (Duris et al., 2017). Stimulation of the vagus nerve may represent a new way to prevent pathological inflammation, which has shown potential as a strategy to lessen the inflammatory response and facilitate functional recovery in stroke patients (Neumann et al., 2015). Vagus nerve stimulation is neuroprotective in acute cerebral I/R injury by suppressing inflammation and apoptosis via activation of cholinergic and α7 nAChR/Akt pathways (Mravec, 2010; Jiang et al., 2014). A short period of transcutaneous stimulation in the external ear initiated 30 minutes after contralateral transient MCAo reduces infarct volume by 28% in rats and leads to an improvement in neurological outcome that is sustained at 24 hours (Ay et al., 2015). This convenient technological development deserves further study on the role of vagus nerve stimulation as a neuromodulator in acute and chronic stroke and as a potential secondary preventive option (Cai et al., 2014). In addition, Acupuncture stimulation apply to forepaw in rat can increase the release of acetylcholine in cerebral cortex and increase cortical cerebral blood flow. The afferent pathway involves to group III and IV afferent nerves, whereas intrinsic cholinergic vasodilators originating in the nucleus basalis of Meynert participates in the efferent pathway (Uchida et al., 2000). The brain mediates via vagal secretion of ACh to suppress peripheral inflammation and the acetylcholinesterase mRNA-targeting microRNA-132 may play a functional regulator role in brain to body resolution of inflammation (Shaked et al., 2009). Taken together, acupuncture stimulation induces vagal activity and cholinergic anti-inflammatory pathway in attenuation of neuroinflammation possibly are a further direction of the study.

Influence of EA on the BBB after Ischemic Stroke

EA was presumed to improve the function of the BBB, which is generally disrupted after cerebral ischemia (Wu et al., 2001). Applying EA (20 Hz/80 Hz, 1–3 mA) to GV20 and GV14 can reduce ischemic injury of the cerebral cortical neurons and BBB in rats with cerebral I/R injury (Shen et al., 2009). Matrix metalloproteinase (MMPs) are neutral proteases that disrupt the BBB and degrade myelin basic proteins under conditions of neuroinflammation, resulting in brain edema (Candelario-Jalil et al., 2011).

Our previous study indicated that EA at GV20 and GV14 during the subacute phase of cerebral I/R injury significantly reduced cerebral infarct and neurological deficit scores. EA also downregulated the expression of nuclear factor kappa B (NF-κB) p50, and TNF-α, in addition to reducing iNOS and apoptosis levels in the ischemic cortical penumbra to provide neuroprotection (Cheng et al., 2014a). Additionally, EA at GV26 and GV20 in a rat model of cerebral I/R injury markedly reduced neurological deficits and the aquaporin-4 protein and mRNA expression levels of corpus striatum, thereby relieving damage to the BBB (Peng et al., 2012).

EA at GV20 and ST36 can improve neurological outcomes in a cerebral I/R injury rat model and reduce inflammation as well as MMP-9 expression in the brain (Xu et al., 2010; Chen et al., 2012b). Acupuncture and EA administered at the GV20 and ST36 acupoints significantly reduced infarct size and improved neurological function. Furthermore, inflammatory cell infiltration and MMP2, aquaporin 4, and aquaporin 9 expression levels were significantly reduced. Acupuncture and EA can exert similar neuroprotective actions in a MCAo rat model (Xu et al., 2014). EA stimulation (2 Hz, 1 mA) at the GV20 and Siguan acupoints (bilateral Hegu (LI4) and Taichong (LR3)) significantly reduced cerebral infarct area and neurological deficit scores, reduced the number of apoptotic cells, upregulated Bcl-2 protein expression, and downregulated Bax protein expression in rats with MCAo. Additionally, EA stimulation significantly downregulated the expression levels of MMP-9 and simultaneously upregulated the mRNA and protein expression levels of tissue inhibitor of metalloproteinases-1 (TIMP-1) (Ma et al., 2016).

Bloodletting puncture at Jing-well points has been demonstrated to alleviate cerebral edema, which mainly results from the disruption of the BBB. A study reported that bloodletting puncture at 12 Jing-well points of the hand once a day could reduce water content in the brain and the permeability of the BBB in a MCAo rat model, in addition to ameliorating tight junctions, as observed under electron microscopy; the expression levels of occludin and claudin-5 were also upregulated, whereas intercellular adhesion molecule-1 (ICAM-1) and VEGF were downregulated (Yu et al., 2017). Another study observed the effects of bloodletting puncture at Jing-well points at the distal ends of the finger and toe on survival rate, survival time, and brain edema in rats with cerebral ischemia; the study revealed that bloodletting puncture prolonged the survival time of the rats and improved ischemic brain edema (Gao et al., 2012).

Antiapoptotic Effect of EA after Ischemic Stroke

EA (2 Hz, 1 mA) applied to a MCAo rat model resulted in a marked reduction of infarct area after stroke and a reduction in the number of apoptotic cells. Moreover, EA enhanced the expression of the antiapoptotic markers Bcl-2, Bcl-xL, and both cIAP-1 and -2. The activities of caspase-3, -8, and -9 were also markedly inhibited by the antiapoptotic effects of EA treatment (Kim et al., 2013c).

Neuroregenerative Effect of EA after Ischemic Stroke

EA could upregulate GDNF expression and extend the duration of this upregulated expression after ischemic insult (Wei et al., 2000). The results of our previous study indicated that 5 Hz EA at GV20 and GV14 upregulated BDNF expression and thus provided BDNF-mediated neuroprotection against caspase-3-dependent neuronal apoptosis by activating the Raf-1/MEK1/2/ERK1/2/p90RSK/Bad signaling cascade in mild MCAo rats (Cheng et al., 2014b).

A rat model of cerebral I/R was established by suture occlusion of the left middle cerebral artery. Low-frequency continuous-wave EA (frequency, 2–6 Hz; current intensity, 2 mA) stimulation of the brachial plexus trunk on the right side increased the amount of BDNF and alleviated neurological function deficits (Guo and Wang, 2012). EA at GV20 improved motor recovery and stimulated BDNF/TrkB expression in rats with cerebral ischemia (Kim et al., 2012). Applying EA stimulation (2 Hz) to GV20 and GV14 after ischemic stroke in a MCAo mouse model may promote poststroke functional recovery by enhancing the proliferation and differentiation of neural stem cells via the BDNF and VEGF signaling pathways (Kim et al., 2014). SDF-1α plays a crucial role in regulating the mobilization, migration, and homing of endothelial progenitor cells (EPCs). The administration of EA at the GV20 and Siguan acupoints in a rat model after focal cerebral I/R could accelerate and increase the formation of an SDF-1α concentration gradient to further induce the mobilization of EPCs and angiogenesis in the ischemic brain and improve neurological function recovery (Xie et al., 2016). EA at GV20 and GV26 with dense-sparse waveforms was effective in attenuating cerebral ischemic injuries and upregulating endogenous IGF-1 expression following MCAo in monkeys; this might be an important mechanism by which EA exerts its neuroprotective effects against cerebral ischemia (Gao et al., 2006). EA at GV20 and GV26 (with a “disperse-dense” wave at 2 and 150 Hz, alternately, and a constant intensity of 3 mA) increased the serum levels of transforming growth factor beta (TGF-β1) in rats with acute cerebral I/R injury, thereby exerting its neuroprotective effects (Wang et al., 2016).

Inflammation after stroke is the main cause of cerebral I/R injury. EA (“disperse-dense” wave at 2 to 100 Hz, alternately; 2 mA) at GV20 and Qubin (GB7) could attenuate cerebral I/R injury and suppress leukocyte infiltration by reducing cyclooxygenas-2 (COX-2) and NF-κB levels and enhancing TGF-β1 expression in brain tissues of rats with MCAo (Zhang et al., 2009). EA treatment (60 Hz, 1 second and 2 Hz, 3 seconds alternately at an intensity of 10 mA) at Fengfu (GV16) and Jinsuo (GV8) activates cell proliferation and facilitates neurogenesis as well as maturation of newly generated neurons; thus, in addition to neuronal regeneration, EA can improve newborn neuron migration and its maturation in the striatum of adult rat brains after stroke (Yang et al., 2005). Blood-derived leukocytes and resident microglia are the most activated inflammatory cells that accumulate in brain tissue after cerebral ischemia, leading to inflammatory injury. Microglia, the major source of cytokines and other immune molecules of the central nervous system (CNS), are the first nonneuronal cells that respond to CNS injury, and they become phagocytic when fully activated by neuronal death.

Cerebral ischemia might promote the proliferation of neural stem cells, and some of them could differentiate into astroglia or neurons. EA could enhance the proliferation and differentiation of neural cells into mature neurons, which might be one of the key reasons why EA could improve neurological dysfunction (Tao et al., 2010). EA was stated to maintain the structural integrity of astrocyte (Xiao et al., 2013). Applying EA at GV20 and GV14 has the potential to activate astrocytes in the peri-infarct region and to avoid excess reactive gliosis in a MCAo rat model; moreover, it can facilitate the recovery of postischemic behavioral dysfunction (Han et al., 2010). EA treatment significantly promoted the recovery of neurological deficits in MCAo rats, which correlated with enhanced lactate energy metabolism in the resident astrocytes around the ischemic area and upregulated the expression of the lactate transporter monocarboxylate transporter 1 (MCT1) in these astrocytes (Lu et al., 2015). Applying EA (4/20 Hz, 1–3 mA) at GV20 and GV14 in MCAo rats is useful for synaptic reorganization, which may be associated with its effect on intervening the activation state of astrocytes; thus, acupuncture could initiate the adjustment of neuron–glial networks and thus improve synaptic reorganization, which may be key for the treatment of cerebral ischemia (Luo et al., 2011). Reactive astrogliosis is a common phenomenon in CNS injury such as ischemic stroke. EA treatment at the LI11 and ST36 acupoints attenuated neurological deficits and cerebral infarct volume in I/R injured rats. This treatment also exerted neuroprotection through the proliferation of GFAP/vimentin/nestin-positive reactive astrocytes and the potential secretion of reactive astrocytes-derived BDNF (Tao et al., 2016).

EA at GV20 and GV14 in mice using bilateral common carotid artery stenosis attenuated spatial and short-term memory impairments and enhanced oligodendrocyte differentiation from oligodendrocyte precursor cells. EA stimulation promotes the recovery of memory function following white matter injury through a mechanism that promotes oligodendrocyte regeneration and involves neurotrophin (NT) 4/5-TrkB signaling (Ahn et al., 2016). A study compared the effects of EA (15 Hz, 1–3 mA) applied bilaterally at Sanyinjiao (SP6) and Fenglong (ST40) with those of manual acupuncture at GV20 and GV26 (punctured and stimulated manually for 1 min) in rats with hyperlipidemia plus cerebral ischemia (Ren et al., 2010). Manual acupuncture and EA significantly upregulated the expression levels of vimentin and beta-Tubulin (Tju-1) in the subependymal ventricular zone of the ischemic side, suppressing the reduction of proliferation and differentiation of neural stem cells (Ren et al., 2010). Our previous study investigated whether acupuncture enhances neuronal regeneration in rats with ischemic stroke. We observed that EA (2 Hz) at both ST36 and ST37 acupoints reduced the infarction/hemisphere ratio, reduced the modified neurological severity score, and increased the rotarod test time. Additionally, EA (2 Hz) reduced nestin immunoreactive cells in the penumbra and ischemic core areas; it also reduced Ki67 and increased GFAP immunoreactive cells in the penumbra area. Our findings suggest that EA (2 Hz) at the ST36 and ST37 acupoints exerts a neuroprotective role (Liao et al., 2017).

The PI3K/Akt pathway, a critical mediator of cell survival, is suppressed during cerebral I/R injury. Administering EA at the LI11 and ST36 acupoints on the contralateral paralyzed limb significantly improved neurological deficits and cerebral infarction and activated PI3K/Akt signaling in ischemic cerebral tissues, resulting in the inhibition of cerebral cell apoptosis; this treatment also increased serum secretion levels of the PI3K activators BDNF and GDNF, in addition to upregulating the antiapoptotic Bcl-2/Bax ratio in the ischemic cerebrum (Chen et al., 2012a). Furthermore, EA treatment at ST36 and LI11 in cerebral I/R injury rats improved neurological deficit and cerebral infarction and profoundly activated PI3K/Akt signaling, resulting in the inhibition of cerebral cell apoptosis in the ischemic penumbra; this treatment increased PI3K, p-Akt, p-Bad, and Bcl-2 expression at the protein level but inhibited Bax and cleaved Caspase-3-positive expression (Xue et al., 2014). The ERK pathway, a critical mediator of cell proliferation, is activated in cerebral I/R injury. A study revealed that EA at the LI11 and ST36 acupoints significantly ameliorated neurological deficits and cerebral infarction in cerebral I/R-injured rats, increased the phosphorylation levels of ERK, and increased the protein expression levels of Ras, cyclin D1, and cyclin-dependent kinase (CDK)4; EA-mediated activation of the ERK pathway resulted in the stimulation of cerebral cell proliferation (Xie et al., 2013). EA probably protects cerebral I/R injury by alleviating the damage to the ultrastructure of brain cells and downregulating the expression levels of neurite outgrowth inhibitor-A (Nogo-A) (Liang et al., 2012). Administering EA at bilateral PC6, SP6, GV26, and GV20 could effectively improve neurological function in rats with cerebral infarction; this may be attributed to its effects on the upregulation of cerebral VEGF, nerve growth-associated protein-43 (GAP-43), synaptophysin (SYN), and myelin basic protein (MBP) expression and the downregulation of Nogo-A protein, thus indicating its protective effects on the neurovascular unit (Han et al., 2013). EA (50 Hz, 3 mA) could enhance the recovery of neurological function, reduce cerebral infarction volume, and increase HIF-1α expression in ischemic rats (Li et al., 2017). Intracellular protein denaturation is a significant pathological step in acute conditions such as stroke and myocardial infarction. Heat shock proteins (HSPs) are fundamental for intracellular protein repair and work, because they prevent protein aggregation and assist the refolding of denaturated proteins (Cakmak, 2009). HSP70 and other endogenous injury-signaling molecules are released by damaged cells. One of the molecular mechanisms involved in EA treatment was revealed to be the promotion of the expression of inducible HSP70 (Sun et al., 2003). Conversely, another study that administered EA (2–100 Hz, 2 mA) at the GV20 and ST36 acupoints in rat models of cerebral I/R injury lowered the peak levels of adrenocorticotrophic hormone and HSP70, suggesting that EA may inhibit excessive stress, reduce inflammation, and promote neural repair, thus facilitating healing in ischemic stroke (Shi et al., 2017).

EA Attenuates Glutamate Excitotoxicity via NMDA Receptors after Ischemic Stroke

Ischemia impairs brain function and networks. The loss of GABAergic neurons disrupts the balance of excitation and inhibition, which causes further neural excitotoxicity and nerve cell death. Acupuncture at GV20 improves ischemic stroke by preventing the impairment of cortical GABAergic neurons (Zhang et al., 2011). Cerebral ischemia induces excessive glutamate release and excitotoxicity. NMDA receptors are responsible for glutamate-induced excitotoxicity in the postischemic brain. Both hyperemia and glutamate overrelease after ischemia were reported to be vital factors for brain damage due to reperfusion injury, and EA treatment (7 Hz, 6 mA) at GV16 and Shendao (GV11) for 30 minutes might inhibit the overrelease of glutamate and thus protect neurons against I/R injury (Pang et al., 2003).

EA could regulate the content of Ca2+ in the ischemic area of the brain and inhibit Ca2+ overload in order to protect the neurons in rats with focal cerebral ischemia (Xu et al., 2002). Our previous study indicated that 2 Hz EA at GV20 reduced the expression levels of NR1 and transient receptor potential vanilloid subtype 1 (TRPV1) receptors in the hippocampal CA1 areas in a vascular dementia model induced by the MCAo technique. Additionally, our previous study indicated that MCAo-induced behavior and long-term potentiation impairment were improved (Lin and Hsieh, 2010).

Glutamate-NMDA receptor excitotoxicity and oxidative stress are two common mechanisms associated with most neurodegenerative diseases, and neurotoxicity through these two mechanisms is dependent upon cAMP response element-binding protein (CREB) and NF-κB DNA transcription that regulates the vitality of neurons (Zou and Crews, 2006). A study investigated the mechanisms by which EA ameliorates learning and memory in rats following MCAo surgery. EA at GV24 and GV20 significantly ameliorated neurological deficits and reduced cerebral infarct volume. In addition, EA improved learning and memory abilities in the rats and markedly activated the CREB signaling pathway, resulting in the inhibition of cerebral cell apoptosis in the ischemic penumbra. Furthermore, EA increased the activity of superoxide dismutase and glutathione peroxidase, the protein expression levels of phosphorylated CREB and Bcl 2, and the mRNA expression levels of Bcl 2. Conversely, it reduced the levels of malondialdehyde and inhibited the expression levels of Bax (Lin et al., 2015).

EA Parameters for Treating Cerebral Ischemic Injury

Research demonstrated that the appropriate electrical stimulation parameters of EA pretreatment to induce cerebral ischemic tolerance in rats are a density-sparse wave of 2/15 Hz and a current intensity of 1 mA applied for 30 minutes per day for 5 consecutive days. The density-sparse wave had the most obvious neuroprotective effect, followed by the intermittent wave, whereas the neuroprotective effect of the continuous wave was relatively poor (Yang et al., 2004). EA pretreatment at GV20 could induce more robust neuroprotection against cerebral I/R injury than did stimulation 1 cm lateral to GV20 or the nonmeridian points of the distal limbs; this thus indicates the acupoint specificity of EA pretreatment (Lu et al., 2002; Li et al., 2012).

Previously, experiments were performed on rats subjected to MCAo using different intensities and frequencies of EA at GV26 and GV20 to optimize the stimulation parameters. The results showed that EA at 1.0–1.2 mA and 5–20 Hz remarkably increased blood flow and reduced ischemic infarction, neurological deficits, and death rates. The “nonoptimal” parameters of EA (e.g., < 0.6 mA or > 40 Hz) could not improve the blood flow or reduce ischemic injury. In addition, the same EA treatment with optimal parameters could not increase blood flow in naive brains (Zhou et al., 2011).

EA at GV20 and GV26 by using a sparse-dense wave (5 Hz/20 Hz) at 1.0 mA for 30 minutes greatly increased CBF, reduced infarction, significantly improved neurological deficit, and reduced the death rate in a rat model of cerebral ischemia caused by right MCAo. A similar result was observed with EA at the left (contralateral to the ischemic side) forelimb (LI11 and PC6). By contrast, EA at the right LI11 and PC6 and at the acupoints in the hindlimb (GB34 and SP6) had no such effect. These results imply that EA protection against cerebral ischemia is relatively acupoint specific (Zhou et al., 2013).

Acupuncture stimulation of the cheek, forepaw, upper arm, and hindpaw, but not the chest, back, lower leg, or perineum, produced significant increases in CBF in anesthetized rats (Uchida et al., 2000). A study was designed to investigate the effects of different frequencies of EA on the latent period and wave amplitude of motor-evoked potentials in rats with focal cerebral infarction. EA was applied to GV26 at a frequency of 2 Hz, 50 Hz, or 100 Hz (intensity 1 mA) for 10 min twice daily for 3 days. The latency on the affected side in the 2 Hz group was significantly shortened (P < 0.05), whereas the amplitude was significantly increased when compared with the model group, indicating that low-frequency EA at GV26 can promote recovery of motor function after focal cerebral ischemic injury in rats (Yao et al., 2012).

Another study investigated optimal EA frequencies for maintaining the structural integrity of ischemic brain tissue. In a rat model of MCAo, EA (15 and 30 Hz) at bilateral LI11 and ST36 reduced neurological deficit, increased GFAP expression, and alleviated ultrastructural damage of astrocytes at the edge of the infarct, compared with the group with no treatment or EA treatment at 100 Hz. EA interventions at frequencies of 15 and 30 Hz can favorably maintain the structural integrity of astrocytes and play a protective role in cerebral ischemic injury (Xiao et al., 2013).

In another study, the effects of different intensities of EA on neuroprotection were observed in rats with cerebral I/R injury. EA stimulations at GV20, Mingmen (GV4), and ST36 with the same stimulation waveform (30 to 50 Hz) and different electric current intensities (5, 3, and 1 mA) were performed. The 3 mA EA was found to be the most effective in strengthening aerobic metabolism, maintaining the ionic equilibrium in the exterior and interior brain cells, and relieving cellular edema by reinforcing the activities of Na+-K+-ATPase (Tian et al., 2015).

EA-induced neuroprotection against cerebral I/R injury depends on an optimal EA duration. A study applied EA (5 Hz/20 Hz at 1 mA) at the GV26 and GV20 acupoints for 5, 15, 30, and 45 minutes after 24 hours reperfusion in rats exposed to right MCAo for 60 minutes. The study revealed that 30 minutes of EA, starting at 5 minutes after the onset of MCAo (EA during MCAO) or 5 minutes after reperfusion (EA after MCAO), significantly reduced infarct volume, improved neurological deficit, and reduced the death rate. Moreover, the protective effect was proportional to the increase in the duration of stimulation in the EA group with MCAo, with the maximum increase observed between 5 and 30 minutes of stimulation. EA for 45 minutes did not cause a reduction in infarct volume or neurological deficits; instead, it showed an increase in death rate in this group (Zhou et al., 2013). In another study, the PC6 acupoint in rats with MCAo was needled at a fixed frequency (3 Hz) with different durations (5, 60, and 180 seconds) under a twisting-rotating acupuncture method. Results showed that different durations of acupuncture had different therapeutic effects, with the 60 seconds duration yielding a more favorable therapeutic effect than did the other two durations for ischemic stroke (Zhang et al., 2015a). To observe the effects of different needle-retaining durations on hemorheology in patients with ischemic stroke, EA was applied at several acupoints including the Jianyu (LI15), LI11, Waiguan (TW5), Hegu (LI4), Futu (ST32), and ST36 acupoints by using a frequency of 2 Hz at 2–6 mA for 20, 40, and 60 minutes separately; the treatment was performed once daily for 10 times. The therapeutic effects on various parameters of hemorheology in the 60 minutes group were more favorable than those in the 20 and 40 minutes groups (He et al., 2007).

To investigate the differential effects between multiple EA and single-time EA stimulation on ischemic injury, a previous study found that both methods significantly reduced MCAo-induced ischemic infarction; however, only multiple EA attenuated sensorimotor dysfunctions. The short-term effect of single-time EA stimulation differs from the cumulative effect of multiple EA, which possibly depends on their differential modulation of the expression of neurotrophic signaling molecules (Wang et al., 2014).

A multicenter, single-blinded, randomized controlled trial was performed in China. In this trial, 862 patients with limb paralysis between 3 and 10 days after ischemic stroke onset were allocated to acupuncture plus standard care or standard care alone. The acupuncture was applied 5 times per week for 3 to 4 weeks. Fewer patients appeared to die or become dependent in the acupuncture group when compared with the control group at 6 months, particularly in the subgroup receiving ≥ 10 sessions of the treatment (Zhang et al., 2015b).

Effectiveness of Acupuncture in Acute Ischemic Stroke Patients

A systematic review showed that scalp acupuncture in patients with acute ischemic stroke appears to improve the neurological deficit score and clinical effective rate when compared with Western conventional medicine (Wang et al., 2012). Studies have examined the additional therapeutic effects of EA in patients with first-ever ischemic stroke. In one study, the study and control groups underwent a conventional rehabilitation program, with the study group receiving an additional eight courses of EA over a period of 1 month. The study results revealed that EA could improve motor function, especially in the upper limbs (Hsieh et al., 2007). In another study, 290 patients aged 40–75 years with a first onset of acute ischemic stroke (more than 24 hours but within 14 days) were treated using standard treatment methods, and they were randomly allocated to an intervention group (treated with resuscitating acupuncture) or a control group (treated using sham-acupoints). The results of this clinical trial showed a clinically relevant reduction in relapse in patients treated with resuscitating acupuncture intervention by the end of 6 months, compared with those who underwent needling at the sham-acupoints. Resuscitating acupuncture intervention could also improve self-care ability and quality of life in these patients (Shen et al., 2012). In our previous, randomized, single-blinded, controlled study, 30 first-time ischemic stroke patients underwent acupuncture treatment along with the manual twisting of needles at GV20 and four spirit acupoints (1.5 cun anterior, posterior, left, and right from GV20) for 20 minutes. The displacement area from the center of gravity decreased in the experimental group, suggesting that acupuncture stimulation may induce an immediate effect that improves balance function in stroke patients (Liu et al., 2009).

Increased CBF was observed in single-photon emission computed tomography (SPECT) brain perfusion images obtained in six patients with MCAo following acupuncture at acupoints LI4, Shousanli (LI10), LI11, LI15, and Jugu (LI16), and at TW5, in the affected arm. Acupuncture stimulation after stroke appears to activate the perilesional sites and may aid in brain reorganization (Lee et al., 2003).

PET is used to observe cerebral function. A study applied this method to observe cerebral functioning in six patients suffering from ischemic stroke after receiving EA treatment at GV20 and right GB7. The results revealed that glucose metabolism had changed significantly in the primary motor area, premotor cortex, and superior parietal lobule bilaterally, as well as in the supplementary motor area on the unaffected hemisphere immediately after the first EA treatment. Similarly, significant changes in glucose metabolism were noted in other areas such as the insula, putamen, and cerebellum. This thus demonstrates that EA was very useful for cerebral motor plasticity after the ischemic stroke (Fang et al., 2012).

Conclusion

In recent years, results from numerous animal experiments and clinical research have uncovered some of the resulting molecular and biophysical correlates of acupuncture or EA in alleviating cerebral ischemic injury. Acupuncture intervention can significantly reduce the size of the infarcted area, improve cerebral blood circulation to promote regional energy metabolism, regulate blood lipid metabolism to resist cerebral free radical damage, inhibit cerebral cortical apoptosis, reduce the amount of excitatory amino acids to lower neurogenic toxicity, reduce calcium overloading, ease cerebral vascular immune-inflammatory reactions, and upregulate the expression of antiapoptosis genes and neurotrophic factors, thereby promoting the proliferation and differentiation of neural stem cells in the focal cerebral cortex and hippocampus. Acupuncture stimulation induces vagal activity and cholinergic anti-inflammatory pathway in attenuation of neuroinflammation possibly are a further direction of the study. In addition, the acupoint that is in common use for neuroprotection or neuroregeneration in ischemic stroke in .

The acupoint that is in common use for neuroprotection or neuroregeneration in ischemic stroke.

LI4: Hegu; LI10: Shousanli; LI11: Quchi; LI15: Jianyu; LI16: Jugu; ST32: Futu; ST36: Zusanli; ST37: Shangjuxu; ST40: Fenglung; SP6: Sanyinjiao; BL23: Shenshu; PC6: Neiguan; TE5: Waiguan; GB7: Qubin; GB20: Fengchi; LV3: Taichong; GV24: Chengjiang; GV4: Mingmen; GV8: Jinsuo; GV11: Shendao; GV16: Fengfu; GV20: Baihui; GV24: Shenting; GV26: Renzhong.