The experimental procedure used in the present study was approved by the Ethics Committee for Animal Experimentation of Nanjing University of Chinese Medicine (Nanjing, China) and all procedures were performed in accordance with the National Institutes of Health Guidelines for Animal Research (20). Male Sprague-Dawley rats weighing between 280 and 300 g were provided by the Experimental Animal Center of Nanjing Medical University. The rats were housed in cages under controlled conditions, with a 12 h light/dark cycle, temperature at 22±2°C and humidity at 60–70% for at least one week prior to surgery, and received food and water ad libitum. The rats were fasted 12 h prior to surgery, but were allowed free access to water. A total of 140 rats were randomly divided into five groups: Sham group (S); 3 h post-I/R group (I/R3 h), 24 h post-I/R group (I/R24 h); EA pretreated 3 h post-I/R group (EA+I/R3 h); and EA pretreated 24 h post-I/R group (EA+I/R24 h). Each group contained 28 rats All the rats were assessed between the point of ischemia and 24 h post-reperfusion.

Transient focal cerebral ischemia was induced by MCAO. Anesthetization was induced using pentobarbital sodium (40 mg/kg body weight; Sigma-Aldrich, St. Louis, MO, USA) intraperitoneally. The right common carotid artery and the right external carotid artery were exposed through a ventral midline neck incision, and were ligated proximally and temporarily. A 2.0 monofilament nylon suture (Doccol Corporation, Redlands, CA, USA), with its tip rounded through heating in a flame, was inserted into the common carotid artery through an arteriectomy, just beneath the carotid bifurcation, and was advanced into the internal carotid artery ~18-20 mm distal to the carotid bifurcation, until mild resistance indicated occlusion of the origin of the anterior cerebral artery and the MCA Following 2 h of ischemia, reperfusion was accomplished by withdrawing the suture. The incision was closed and the animals were allowed to recover from the surgery. In the sham group, the same surgery was performed, but without occlusion of the MCA. The rectal temperature was maintained at 38±0.5°C throughout the procedure using a thermostat-controlled heating pad (Doccol Corporation).

Cerebral blood flow (CBF) of the MCA was measured using laser Doppler flowmetry. A flexible fiber-optic probe was affixed to the skull over the cortex supplied by the proximal part of the MCA (2 mm caudal to the bregma and 6 mm lateral to the middle). Rats exhibiting <80% reduction in CBF in the core of the MCA area were excluded from further investigation. Besides the sham group (S), the other four groups all had rats excluded (I/R3 h, n=3; I/R24 h, n=2; EA+I/R3 h, n=2; EA+I/R24 h, n=2).

According to the Experimental Animals Meridians Mapping, Baihui (GV20), which is located at the vertex of the parietal bone, that is, the midpoint of the connecting line between the auricular apices, was selected (18,19). The rats in EA pretreatment groups were anesthetized and stimulated using a 1 mA current, with a density-sparse wave of 2/15 Hz, for 30 min/day for five consecutive days. This was performed using a Hwato Electronic Acupuncture Treatment instrument (SDZ-V; Suzhou Medical Applicances Co., Ltd., Suzhou, China).

A modified neurologic deficit score, described by Longa et al (21) was used for neurological assessment and was scored as follows: 0, no deficit; 1, failure to extend left forepaw fully; 2, circling to the left; 3, falling to the left; 4, no spontaneous walking with a depressed level of consciousness. Rats scoring between 2 and 3 were included in the subsequent experiments.

Following neurobehavioral evaluation, the rats (n=5) were sacrificed and decapitated after an intraperitoneal injection of 40 mg/kg pentobarbital sodium, followed by removal of the brains. The brains were rapidly separated into left and right cerebral hemispheres, and the right cerebral hemispheres were weighed (wet weight) using an electronic balance (Ohaus Corporation, Parisippany, NY, USA). Subsequently, the brains were dried for 24 h at 100°C in order to obtain the dry weight. The brain water content was calculated as follows: Brain water content = [(wet weight - dry weight) / wet weight] × 100. This was used as an index for brain edema (22).

Following decapitation, the brains (n=8) were removed and frozen at −20°C for 15 min. The brains were were then sliced using a plastic matrix (2 mm thickness; Sigma-Aldrich) and stained using 2% 2,3,5-triphenyltetrazolium chloride (TTC; Sigma-Aldrich) in 0.1 mol/l phosphate buffer (Sigma-Aldrich) for 30 min at 37°C to evaluate the infarct volume (23). The infarcted tissue remained unstained (white), whereas normal tissues was stained red. The infarct volume was calculated as follows: Infarct volume = contralateral hemisphere volume - non-infarcted volume of the ipsilateral hemisphere) / contralateral hemisphere volume (24).

The rats (n=5) were anesthetized with 40 mg/kg intraperitoneal pentobarbital sodium, then were perfused with pre-cooled phosphate-buffered saline (PBS; pH 7.4), followed by PBS containing 4% para-formaldehyde (Sigma-Aldrich) and 0.25% glutaraldehyde (Sigma-Aldrich). The brain was sectioned into three slices, starting 3 mm from the anterior tip of the frontal lobe in the coronal plane. The slices were 3, 4, and 3 mm thick from front to back, respectively. The middle slice was cut longitudinally in the ischemic hemisphere 2 mm from the midline then a transverse diagonal cut was made at the 2 o’clock position to separate the core from the penumbra. A 1 mm thick coronal slice from the cortex penumbra area was removed. The slice was placed in fresh prepared 2.5% glutaraldehyde overnight at 4°C. Following rinsing with 0.1 mol/l PBS three times, the slice was post-fixed in 1% osmium tetroxide (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) for 1 h, dehydrated in graded ethanol (Sigma-Aldrich), and embedded in epoxy resin (Sigma-Aldrich). Polymerization was performed at 80°C for 24 h. Blocks were cut from the slice using a Reichert/Leica Ultracut S Ultramicrotome (Leica Microsystems GmbH, Wetzlar, Germany) into ultrathin sections (60–70 nm), which were then post-stained with uranyl acetate (Sigma-Aldrich) and lead citrate (Sigma-Aldrich), and examined using a Hitachi 7100 electron microscope (Nikon, Corporation, Toyko, Japan).

To analyze the expression of proteins, the rats (n=5) were decapitated and brains were removed. The ischemic cortices, and corresponding cortices of the sham rats, were rapidly dissected and frozen on dry ice (Sigma-Aldrich). Total protein samples were extracted using radioimmunoprecipitation assay (RIPA; Sigma-Aldrich) buffer for determination of the expression of claudin-5, occludin, caveolin-1 and Akt. To determine the expression of p-caveolin-1 and p-Akt, protease inhibitor cocktail tablets (Sigma-Aldrich) were added, at a ratio of 1 tablet/10 ml RIPA buffer. The protein concentrations were determined using a spectrophotometer (UV-2540; Shimadzu Corporation, Kyoto, Japan). Equivalent quantities of proteins (70 µg) from each sample were separated using 10% SDS-PAGE (Sigma-Aldrich) and subsequently transferred onto a nitrocellulose membrane (Sigma-Aldrich). The membranes were then incubated overnight at 4°C with the following rabbit primary antibodies: Claudin-5 (1:1,000; cat.no. ab53765; Abcam), occludin (1:1,000; cat.no. ab31721, Abcam), caveolin-1 (1:1,000; cat. no. 3238; Cell signaling Technology, Inc., Danvers, MA, USA), Akt (1:1,000; cat. no. 9272; Cell signaling Technology, Inc.), p-caveolin-1 (1:1,000; cat. no. 3251; Cell signaling Technology, Inc.) and p-Akt (1:1,000; cat. no. 9275; Cell signaling Technology, Inc.), followed by incubation with the respective horseradish peroxidase-conjugated anti-rabbit secondary antibodies (1:5,000; Jackson Immunoresearch Laboratories, Inc., West Grove, PA, USA) for 60 min at room temperature. Immunoreactivity was detected using an enhanced chemiluminescent autoradiography system (sc-2048 Western Blotting Luminol Reagent; Santa Cruz Biotechnology, Inc.). Each blot was reprobed with β-actin (1:5,000; cat. no. sc-130657; Santa Cruz Biotechnology, Inc.), following stripping by heat and detergent to remove the antibodies from the membrane, to provide a control for the load variations between the samples. Autoradiographic films (Santa Cruz Biotechnology, Inc.) were used for the final determination of protein expression using SigmaScan 5.0 (Sigma-Aldrich) and normalized to the relative optical density obtained for β-actin.

SPSS 19.0 for Windows (IBM SPSS, Armonk, NY, USA) was used to performed statistical analysis. All values are expressed as the mean ±standard deviation. Data were analyzed using one-way analysis of variance, and inter-group differences were detected using Student-Newman-Keuls post-hoc analysis. P<0.05 was considered to indicate a statistically significant difference.

A modified neurologic deficit score, described by Longa et al (21), was used for neurological assessment of the different groups. The rats in the I/R3 h group exhibited severe neurological deficits compared with the sham group. Following 24 h reperfusion without any pretreatment, the neurological deficits remained unchanged, while EA pretreatment significantly improved the neurological deficits induced by I/R.

TTC staining was used to reveal cerebral infarcts, with normal brain tissues stained red, and infarct lesions remaining unstained (white). Compared with the sham group, the brain infarct volume increased significantly in the I/R3 h and I/R24 h groups. In the EA+I/R groups, these I/R-induced cerebral infarcts were reduced significantly and dose-dependently (Fig. 1).

The brain water content was significantly increased in the I/R groups, compared with the sham group. In addition, EA pretreatment significantly reduced the brain water content, compared with the I/R3 h and I/R24 h groups. (P<0.05; Fig. 2).

Taken together, these findings demonstrated that EA pretreatment attenuated the I/R-induced increase in BBB permeability.

In order to determine the mechanisms underlying the role of EA pretreatment in maintaining BBB permeability, transmission electron microscopy was used to identify the morphological changes of cerebral microvasculature in the cortex in all the groups. In the sham group, the cerebral microvasculature was relatively normal, with normal caveolae, and the TJs localized between endothelial cells as continuous lines. By contrast, in the I/R groups, the caveolae in the endothelial cells were increased, and the continuous lines of the TJs became ill-defined, appearing as dotted lines, indicating degradation of the TJ proteins. However, these observations were improved by EA pretreatment. These results were also confirmed using western blotting (Fig. 3).

Western blotting was used to evaluate the expression levels of claudin-5 and occludin. Compared with the sham group, I/R induced a significant decrease in claudin-5 and occludin. Following EA pretreatment, the expression of these two TJ proteins increased significantly. No significant differences were observed between the EA+I/R3 h group and the EA+I/R24 h group (Fig. 4).

Western blotting was used to evaluate the expression of caveolin-1 and p-caveolin-1. In terms of caveolin-1, no significant differences were identified among the treatment groups. However, the expression of p-caveolin-1 was significantly increased following I/R compared with that of the sham group, while EA pretreatment significantly relieved this effect (Fig. 5).

Western blotting was used to evaluate the expression of Akt and p-Akt. Similar to caveolin-1, no significant differences were observed among the groups for Akt. However, the expression of p-Akt was significantly increased following I/R compared with that of the sham group, while EA pretreatment significantly relieved this effect (Fig. 6).