Claudin 5

To image monolayers, the gels were fixed and stained, then inverted on a glass slide. Four days after seeding, gels were washed with PBS then fixed in 3.7% paraformaldehyde for 10 min. Cells were then permeablized for 1 h in 0.1% Triton-X100 and blocked for 1 h in 10% donkey serum in PBS. Primary antibodies were incubated overnight at 4 °C in 10% donkey serum followed by 1 h incubation with secondary antibodies in 10% donkey serum. Cells were stained for ZO-1 (Life Technologies), claudin-5 (Life Technologies), and DAPI as described previously [8 (link)]. Imaging was performed on a Nikon TiE confocal microscope using 40× objective and NIS Advanced Research software (Nikon, Minato, Tokyo, Japan). The images are maximum intensity projections of confocal z-stacks with a z spacing of 0.4 µm.

The GECs on insert filters were fixed in 4% paraformaldehyde for 20 min and blocked by 5% BSA for 2 h at room temperature. After washed with PBS for 3 times, GECs were incubated by primary antibodies of ZO-1, occludin, and claudin-5 (1:50; Life Technologies) at 4°C overnight. The nuclei were stained by DAPI (0.5 mg/ml, Beyotime Institute of Biotechnology) for 8 min. The fluorescence was visualized by DP71 immunofluorescence microscope (Olympus, Tokyo, Japan), and merged by Chemi Imager 5500 V2.03 software.

Protein (4.1 µg) and Laemmli buffer (Bio-Rad Laboratories GmbH, Munich, Germany) were mixed and loaded onto a 12.5% polyacrylamide gel. Electrophoresis was allowed to run for 5 min at 200 mV, followed by 60 min at 100 mV. Afterwards, the samples were transferred onto a PVDF membrane, which was blocked for 60 min in 5% milk (in Tris-buffered saline with 0.1% Tween 20) and later incubated with the antibodies for claudin-1, claudin-3, claudin-5, tricellulin, or beta-actin (Life Technologies, Carlsbad, CA, USA) overnight at 4 °C. Antibodies were used as recommended by the manufacturer (1 µg/mL). To enable visual detection, peroxidase-conjugated goat anti-mouse or anti-rabbit antibodies (Cell Signaling Technology, Danvers, MA, USA) were employed to bind the primary antibodies. Signals were then visualized by employing the chemiluminescence detection system of Clarity Western ECL Blotting Substrate (Bio-Rad Laboratories GmbH, Munich, Germany) and a luminescence imager (ChemiDoc MP, Munich, Germany).
The density of individual immunoblotting bands was analyzed by using the imager-associated software (ImageLab, BioRad, Hercules, CA, USA). The measured values were divided by the coherent beta-actin band signal. Finally, the signal for the caprate challenge was expressed in relation to the control values for VE and PP, which were set to 100%.

Brain slices were fixed with methanol at 4 °C for 10 min and blocked with diluted donkey serum (Jackson ImmunoResearch, West Grove, PA) for 60 min at room temperature. Slides were incubated with primary antibodies of CD31 (1:200, R&D system, Minneapolis, MN), occludin (1:200, Life Technologies, Carlsbad, CA), claudin-5 (1:200, Life Technologies), β-catenin (1:100, Abcam), GFAP (1:500, Millipore), NeuN (1:50, Millipore), and Wnt7a (1:100, Abcam) overnight at 4 °C. After rinsing three times with PBS, brain slices were incubated with the fluorescence conjugated second antibodies for 1 h at room temperature. Immunofluorescence photos were collected by a confocal microscope (Leica, Solms, Germany). For claudin-5 and β-catenin quantification, mean integrated density was measured by ImageJ software (National Institutes of Health). For occludin staining, the gap length was presented as percentage (%) of whole tight junction staining^{26 (link),27 (link)}. We randomly chose at least four vessels in the peri-focal region per brain section, and total eight sections per animal. Then we measured the length of vessel and gap by ImageJ software (National Institutes of Health). Gap length was presented as percentage (%) of gap length in whole vessel.