The welfare of all animals was ensured, and all experimental procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals established by the National Institutes of Health. The Animal Ethics Committee of China Pharmaceutical University (Nanjing, China) approved all protocols [No. SYXK(Su)2018-0008].

Male C57BL/6J mice were purchased from the Animal Center of Yangzhou University (Yangzhou, China). Adequate food and water were provided. Animals were housed in cage in an environment maintained at a constant temperature (22–24°C) with a normal circadian rhythm.

bEnd.3 cells were purchased from Bioleaf Biotech (Shanghai, P.R. China) and cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Billings, MT, United States) supplemented with 15% fetal bovine serum (FBS; Gibco), 100 U/mL penicillin and 100 U/mL streptomycin (Ameresco, Framingham, MA, United States) at 37°C in a humidified atmosphere of 5% CO₂ and 95% air. Cells were plated onto cell culture dishes and grown to 80–90% confluence before experimentations.

Mice were anesthetized in an induction chamber using 3–4% isoflurane in 30% O₂/70% N₂. Anesthetization was confirmed after approximately 2 min when respiration slowed to one breathe per second. Animals were removed from the induction chamber and placed in an anesthesia mask, which maintained an isoflurane concentration of 1–1.5%. Middle cerebral artery occlusion/reperfusion (MCAO/R) was induced using a method based on intraluminal filaments with slight modification as described previously (Gong et al., 2019). Briefly, the right middle cerebral artery of mouse was occluded by inserting a blunt-tip 4-0 nylon monofilament for 1 h followed with reperfusion for 24 h.

bEnd.3 cells were placed in a 37°C anaerobic chamber (0.2% O₂, 5% CO₂, 95% N₂) and cultured in glucose-free medium for 6 h. After the oxygen-glucose deprivation, the cells were placed in glucose-containing DMEM with 15% FBS and incubated under normoxic conditions for hours in order to imitate I/R-like conditions Cao et al. (2016).

The protective effects of verteporfin (VP, CAS No. 129497-78-5) were examined in vitro. Cells were divided into four groups (n = 3): Control, OGD/R, VP (1 μM, Supplementary Figure S1) treatment after 6 h OGD, and edaravone (Edara, 1 μM, CAS No. 89-25-8). Edara is a commonly used drug for the clinical treatment of ischemic stroke and is often used as a positive drug in basic research (Bao et al., 2018). bEnd.3 cells were cultured on top of gelatin-coated transwell inserts in 24-well plates for 7 days. The Transendothelial electrical resistance (TEER) of the EC monolayer was monitored daily using a Millicell-ERS voltohmmeter (Millipore, United States). The results obtained from the experimental groups were measured after subtracting the value of a blank, cell-free filter.

The protective effects of VP were examined in mice that were randomly divided into four groups (n = 6 per group): sham, MCAO/R, VP (10 mg/kg, i. p.) after 1 h MCAO, and Edara (5 mg/kg, i. p.). After 24 h of reperfusion, the mice brains were quickly removed and frozen at −70°C. The frozen brain was coronally cut into five slices, then incubated individually using a 24-well culture plate with 1% triphenyl tetrazolium chloride (TTC) solution at 37°C for 15 min. The infarct area was measured by computerized planimetry after photographing with a digital camera. The infarct volume is calculated by summing infarct areas on each slice and multiplying by slice thickness. The personnel conducting the TTC staining was blinded to the study group assignment in order to avoid subjective factors affecting the experimental results Tsubokawa et al. (2017).

The mice were sacrificed after MCAO/R induction. The brains were taken out and the wet weight of tissue was accurately measured. After dried in an oven at 100°C for 48 h, the lung tissues were weighed again, recording as dry weight. The content of water in brain were calculated to determine the degree of brain edema. The neurological deficits of the experimental animals were graded on an 18-point scale, as previously described Wu et al. (2019). The evaluation indicators include body symmetry, gait, climbing, circling behavior, forelimb symmetry, compulsory circling and whisker response. The index scores are added together as the final score. The higher the score, the more severe the neurological deficit.

Animal brains were removed 24 h after reperfusion, mice were euthanized, the brains excised rapidly, and dipped in 4% paraformaldehyde. Examination was completed in the Pathology Department of the Jiangsu Center for Safety Evaluation of Drugs (Jiangsu, P.R. China) and the brain slices were observed by a digital scanner (NanoZoomer 2.0 RS, Hamamatsu, Japan).

After anesthesia with 3% pentobarbital sodium, an incision of about 1–2 cm was made in the abdominal cavity of mice. Cerebral blood flow (CBF) in the mesentery was measured using a laser Doppler flowmeter Laser. Images were acquired at ischemia onset and during reperfusion (n = 6 per group).

Evans Blue (EB) extravasation was used to determine BBB integrity as described previously Wu et al. (2019). At 22 h after reperfusion, 2% EB dye (3 ml/kg, Sigma, United States) was injected via the tail vein. The mice were euthanized at 2 h after injection of EB and then perfused with saline. The brains were rapidly taken out and imaged. Then the right hemisphere of brain tissue was weighed, homogenized in formamide (0.1 g/ml) and centrifuged at 5,000 g for 30 min after incubated at 60°C for 18 h. The supernatants were collected to determine the quantity of EB, the absorbance at 620 nm was measured spectrophotometrically using an Infinite M200 Pro plate reader (Tecan, NC, United States). EB leakage into the brain tissue was assessed with a standard curve and expressed as micrograms per Gram of wet brain tissue.

Cells were incubated in the Millicell™ cell culture inserts in a humidified atmosphere of 5% CO₂ and 95% air for 7 days. After exposure to OGD/R conditions and drugs, the medium was removed. 200 μL of EB solution (0.67 mg evans blue powder dissolved in 4% BSA solution) were added into the Millicell cell culture inserts and 600 μL of 4% BSA solution were added into the external chamber. The cells were continuously incubated for another 1 h and then the external solution was collected and the absorbance at 620 nm was measured spectrophotometrically using an Infinite M200 Pro plate reader (Tecan, NC, United States). The EB leakage of each group was calculated according to the standard curve and expressed as a percentage of the values of control group.

Culture medium containing 5 mg/ml 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) solution replaced the complete medium Four hours after incubation at 37°C, the reaction solution was removed, and 150 µL DMSO was added to each well. A microplate reader (Epoch, Bio Tek, Winooski, VT, United States) was used to record the absorbance with dual waves at 570 and 650 nm after 10 min of shaking.

The cells or brain tissue samples (n = 6, for each group) were decapitated and rapidly collected. The prepared cells or tissues (brain tissues from the ischemic penumbra) were homogenized in 1:10 (w/v) ice-cold protein extraction buffer in glass homogenizers. To detect the levels of YAP and phospho- (p)-YAP in the nucleus, a Nuclear Extraction Kit (Solarbio, Cat: SN0020) was used to isolate and purify nuclear and cytoplasmic fractions. To examine the levels of zonula occludens-1 (ZO-1), occludin, matrix metalloproteinase (MMP)-2 and MMP-9, soluble proteins were extracted from cell lysates by centrifugation at 12,000 × g for 10 min at 4°C and collecting the supernatant. The membranes were blocked with phosphate-buffered saline containing Tween20 (PBST) containing 5% skim milk for 2 h at room temperature and then incubated with primary rabbit monoclonal antibody overnight at 4°C (YAP, p-YAP, 1:500; Proteintech Group, United States ZO-1, occludin 1:500; Abcam, United Kingdom MMP-2, MMP-9, 1:800; CST, United States). The membranes were then washed and incubated with secondary antibody (anti-rabbit IgG, 1:3,000; Proteintech Group, United States) for 1.5 h at room temperature. The anti-actin antibody (1:1,000; Proteintech Group, United States) served as a loading control. The protein bands were visualized with enhanced chemiluminescence reagents (ECL), and the signal densitometry was quantified using a western blotting detection system (Quantity One, Bio-Rad Laboratories, United States) by an observer blinded to the groups of animals or cells being examined.

Specimens were sectioned at thickness 10 µm to adhesive slides and bEnd.3 cells were cultured on laser confocal dishes. Specimens were treated with blocking buffer (5% bovine serum albumin, 0.2%Triton-100) for 1 h at 4 °C and then incubated overnight at 4°C with primary antibody against YAP, p-YAP, and ZO-1 (ZO-1, 1:200; Proteintech Group, United States; YAP, 1:300; p-YAP, 1:100; Abcam, United Kingdom), followed by incubation with an Alexa Fluor 488-conjugated donkey anti-rabbit IgG (HþL) antibody (Invitrogen, Carlsbad, CA, United States) and 4′,6-Diamidino-2-phenylindole (Beyotime Biotechnology). Fluorescent images were observed with a confocal laser scanning microscope (LSM700; Zeiss, Jena, Germany) and processed using ZEN imaging software. Regarding the IF brain slice, the location of the studied brain area was showed as an illustration figure in Supplementary Figure S1.

Data are expressed as the mean ± SEM. Statistical analyses were carried out using the Student’s t-test (two-tailed) for comparison between two groups and one-way analysis of variance (ANOVA) followed by Dunnett’s test if the data involved three or more groups. Tests were considered significant at p < 0.05. Analyses were carried out using Prism v5.01 (GraphPad, San Diego, CA, United States).

VP (10 mg/kg) significantly reduced cerebral infarct volume, cerebral edema, and neurological deficits after MCAO/R injury (Figures 3A,B, Supplementary Figures S8 and 9). The damaged area of the brain was reduced significantly in MCAO/R model mice after the administration of VP compared with the untreated MCAO/R mice (Figure 3C). CBF improved in the ischemic hemisphere of MCAO/R mice after the administration of VP compared with untreated MCAO/R mice (Figure 3D). The efficacy of VP was similar to that observed for Edara, which was used as a positive control, which indicated that VP treatment induced improvements following ischemic brain injury in mice.

Compared with the untreated MCAO/R group, the administration of VP significantly reduced EB leakage, increased the expression levels of ZO-1 and occludin, and decreased the expression levels of matrix metallopeptidase (MMP)-2 and MMP-9 in brain tissues (Figure 4). The efficacy of VP was similar to Edara, which indicated that VP could improve BBB integrity following MCAO/R injury.

To evaluate changes in YAP expression in an in vitro model of cerebral I/R injury, YAP/p-YAP expression levels were examined in the brain-derived EC line bEnd.3 following the induction of an OGD/R model. YAP/p-YAP expression levels in the cytoplasm reduced gradually after 6 h of OGD and 6 h of reoxygenation, whereas YAP expression increased gradually in the nucleus (Figure 5A). IF analysis also showed changes in YAP/p-YAP expression occurred after 6 h of OGD followed by 6 h of reoxygenation (Figure 5B), which was consistent with the results of in vivo studies.

Whether VP can inhibit the expression of YAP in the nucleus under OGD/R injury conditions remains to be elucidated. The expression levels of the Hippo pathway target kinase YAP and p-YAP in the Hippo pathway were measured by western blotting and IF analyses in vitro after the administration of VP. The results showed that VP (1 µM) significantly increased the expression levels of p-YAP and YAP in the cytoplasm and decreased YAP expression levels in the nucleus (Figures 6A–C). Further analysis showed that under OGD/R conditions, the ratio of translocated YAP protein into the nucleus was approximately 85%, and after VP administration, the ratio of translocated YAP into the nucleus was approximately 30% (Supplementary Figure S10).

To further investigate the protective effects of VP in vitro, bEnd.3 cells exposed to OGD/R injury were utilized. Compared with the control group, the cell viability of the OGD/R group decreased, based on the results of an MTT assay (Figure 7A, Supplementary Figure S11). Treatment with VP (1 µM) and Edara (1 µM) significantly increased cell survival following OGD/R injury. TEER was lower after OGD/R injury compared with the control group and increased significantly following treatment with VP and Edara (Figure 7B). EB leakage increased in the OGD/R group compared with that in the control group, and VP significantly inhibited OGD/R-induced EB leakage to an equivalent level as observed for Edara (Figure 7C). Compared with the OGD/R group, the protein expression levels of ZO-1 and occludin were increased, and MMP−2 and −9 expression levels were decreased significantly after VP administration (Figures 7D,E).

To further evaluate the effects of YAP on TJs between ECs, we knocked down YAP expression in bEnd.3 cells using YAP-siRNA in vitro (Supplementary Figures S11 and 12). Compared with the control group, the TEER of cells subjected to OGD/R. Treatment with YAP-siRNA induced a significant increase in TEER (Figure 8A). EB leakage increased in the OGD/R group compared with the control group. YAP-siRNA significantly inhibited OGD/R-induced EB leakage to an equivalent level (Figure 8B). Compared with the OGD/R group, the protein expressions levels of ZO-1 and occludin increased significantly after the administration of YAP-siRNA (Figure 8C). The IF results also showed that the fluorescence intensity of ZO-1 increased significantly after the administration of YAP-siRNA compared with that in the OGD/R group (Figure 8D).