Anxa2 knock-in mice (ANXA2+/+, C57BL/6 background), which overexpress ANXA2, were commercially generated by standard homologous recombination at Cyagen (Suzhou, China). Male adult C57BL/6 mice (8-10 weeks) (from Shanghai Slac Laboratory Animal) were used. The environmental conditions which maintain the mouse were: 12-hour dark/12-hour light cycle, humidity range of 30–50%, temperature range of 20–25 °C. All behavior experiments were performed between 9:00 and 19:00. All animal experiments were conducted in accordance with the ethical guidelines of the Zhejiang University Animal Experimentation Committee and Institutional Animal Care. Every Effort has been made to minimize animal suffering, and a minimum number of animals were used.
There are three special regions on ANXA2+/+ mice knock-in allele: region 1 is located at the border between homology arm and CAG promoter sequences; region 2 is located at the border between CAG promoter and Anxa2-cDNA sequences; region 3 is located at the border between Anxa2-cDNA and homology arm sequences (Supplementary Fig. 1A). For genotyping of ANXA2+/+ mice, four pairs of primers were applied. The primers sequences are listed in Supplementary Table. For primer pair of ANXA2-region 1F and ANXA2-region 1R, the PCR products were 463 bp (region 1 of the knock-in allele). For primer pair of ANXA2-region 2F and ANXA2-region 2R, the PCR products were 383 bp (region 2 of the knock-in allele). For primer pair of ANXA2-region 3F and ANXA2-region 3R, the PCR products were 437 bp (region 3 of the knock-in allele). For the pair of Internal-control F and Internal-control R the PCR products were 689 bp (Internal control) (Supplementary Fig. 1B). The PCR program was 94 °C for 2 mins (1×); 94 °C for 15 s, 60 °C for 15 s, and 72 °C for 30 s (30×); and 72 °C for 1 min (1×).
Growth Factor Reduced Matrigel was purchased from Corning (356230, USA). Recombinant human ANXA2 protein was purchased from R&D System (D9409-AN-050, USA). Anti-BrdU antibody was purchased from Abcam (ab1893, USA). Anti-Lectin was purchased from Vectorlabs, (B-1415-2, USA). Anti-Tubulin monoclonal antibody was purchased from Abcam (ab210797, USA). Anti-A2R polyclonal antibody was purchased from Novus Biologicals (NBP-53077, USA). Anti-AKT polyclonal antibody was purchased from Cell Signaling Technology (9272s, USA). Anti-p-AKT polyclonal antibody was purchased from Cell Signaling Technology (4060s, USA). Anti-ERK monoclonal antibody was purchased from Santa Cruz Biotechnology (sc-81457, USA). Anti-phospho-ERK monoclonal antibody was purchased from R&D System (DYC1483-5, USA). Alexa Fluor® 488 AffiniPure donkey anti-goat IgG were purchased from Jackson ImmunoResearch Laboratories (705-545-147, USA). Alexa Fluor® 488 IgG fraction monoclonal mouse anti-Biotin were purchased from Jackson ImmunoResearch Laboratories (200-542-211, USA). Alexa Fluor® 594 IgG fraction monoclonal mouse anti-Biotin were purchased from Jackson ImmunoResearch Laboratories (200-582-211, USA).
Brains from mice were sliced in 20 μm thickness using a Thermo Fisher® freezing microtome. After being washed 3 times with PBS for 5 mins each, the brain slices were placed into the glass Petri dish with the 10 mM citric acid (pH=6) antigen retrieval buffer which was preheated to 90-100˚C. The glass Petri dish was heated in a microwave oven for 10 mins. Then it was cooled to room temperature. Next, the brain slices were permeabilized with 0.3% (w/v) Triton X-100 in PBS at room temperature for 30 mins. After blocking in PBS containing 5% normal donkey serum at room temperature for 2 h, the brain slices were first incubated with anti-BrdU antibody and anti-Lectin antibody at 4 °C overnight and then with Alexa Fluor-conjugated secondary antibodies at room temperature. FluoroshieldTM with DAPI (Sigma-Aldrich, USA) was used to label the nuclei. Before permeabilizing with 0.3% (w/v) Triton X-100 in PBS, the sections used for BrdU staining were pretreated with 1 M hydrochloric acid at 37 °C for 30 mins before the above steps, washed with 0.1 M sodium tetraborate for 3 times, and neutralized for 5 mins each time. Then the primary and secondary antibodies were incubated according to the above process. Images were taken by a laser confocal microscope (Leica SP8, Germany). All the image quantification analyses were done by imageJ-win64. All the data were collected and analyzed by the rater who was blind to experiments design.
Aortic ring angiogenesis assay
The assay developed by Nicosia and Ottinetti was modified . Aortas of 2 weeks WT control or ANXA2+/+ littermate mice were sliced into 1mm diameter rings. Aortic rings were then added in DMEM with 2.5% fetal bovine serum (FBS) to a 24-well plate that were pre-coated with a rat tail collagen I gel (BD Biosciences, UK). Medium was changed every 2 days. At the sixth day of culture, the explants were stained by anti-Lectin (B-1415-2, Vectorlabs, USA) and visualized under a Leica SP8 laser confocal microscope.
Photothrombotic model of stroke
Photothrombotic model was performed as previously described to induce focal cerebral ischemia [13-15]. In brief, mice were anesthetized with isoflurane and the heads were fixed on a stereotaxic apparatus. A cold light source (17000 lux; diameter 2.0 mm) was placed on the mouse skull surface (1.5 mm lateral to bregma). Rose Bengal solution (Sigma, USA) was dissolved in saline and injected into mice intraperitoneally, at 100 mg/kg. The head of the mouse was illuminated for 10 mins through the intact skull 5 mins after drug injections. Sham mouse was injected with Rose Bengal solution at same dosage and did not subject to illumination.
The grid-walking test was performed on a metal grid area of 20 × 32 × 50 cm of width, length and height, with a 12 mm square wire mesh, as previously described [14,15]. A video camera was placed under the apparatus to record the gait of mouse. Mouse was allowed to freely move on the apparatus for 5 mins. The number of foot errors and the normal steps of left forelimb were counted (the left forelimb was counted for 100 steps per video). The ratio of forelimb errors was calculated as follows: total number of foot errors / (total number of foot errors + total number of the normal steps) *100. Foot errors were considered to occur when the forelimb was not providing support and the foot went through the grid hole, or animal was resting with the grid at the level of the wrist. The video analysis was performed offline by the rater who was blind to experiments design.
The cylinder test was performed as previously described [16,15]. In short, mouse was allowed to freely explore for 5 min in a Plexiglas cylinder (15 cm in height with a diameter of 10 cm). A camera was positioned above the apparatus to record the exploratory rear of mouse. Exploratory rear was considered to occur when the mouse spontaneously reared by touching the wall of cylinder with its forelimbs. The time (in seconds) spent by the both forelimbs or each forelimb of mouse pressing the wall of cylinder was counted. The percentage of time spent on each forelimb were used to derive an asymmetry index which was calculated as follows: (the time spent on ipsilateral forelimb - the time spent on contralateral forelimb) / total time spend on forelimb. The video was analyzed by the rater who was blind to experiments design.
Human brain microvascular endothelial cell (HBMEC) culture
HBMECs were obtained from Cell Systems Corporation (Kirkland, WA, USA). The method of HBMECs culture has been described previously . In brief, the cells of passage 6th to 12th were grown with EBM-2 Basal Medium (Lonza, USA) containing Endothelial Cell Growth Medium-2 (Lonza, USA). HBMECs were cultured with RPMI-1640 containing 10% FBS. All cells were kept in a 37 °C humidified incubator with 5% CO2. HBMEC was treated with indicated concentrations (from 0.1 to 1μg/ml) of recombinant ANXA2. OGD model was used to mimic ischemic stroke in vitro. The cells were cultured with glucose-free RPMI-1640 in a sealed chamber (Billups-Rothenberg, USA) loaded with mixed gas containing 95% N2 and5% CO2 for 3 hrs. 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenylte- trazolium bromide (MTT) assay was used to evaluate cell viability. Cell viability was expressed as a percentage relative to the normal cells. The data represented 6-12 separate wells assayed per data point, with about 500–1000 cells counted per well.
In Vitro Angiogenesis Assays
In vitro, angiogenesis was assessed by scratch migration and tube formation assays. The ability of endothelial cells to form capillary-like structures was evaluated by the Matrigel tube formation assay. In short, HBMECs (4×104 cells/well) were inoculated in 48-well culture plates which were pre-coated with growth factor-reduced Matrigel. Then cells were maintained with different treatments at 37 °C for 24 hrs. The number of the branch point and tubes in 3 random fields from each well were counted.
To measure the migration abilities of HBMECs, scratch migration assay was performed. After being seeded in 12-well plates, HBMECs were scratched by a 200 µl pipette tip. Then, the cells were incubated with different treatments for 2 days. HBMEC migration was monitored every day. Three different visual fields were randomly selected per well for observation and photograph under a phase contrast microscopy. The percentage of areas occupied by migrated cells were used to estimate the migration ability of HBMECs.
Cell or tissue samples were homogenized in RIPA buffer and the protein concentration was measured using a BCA assay. Samples of the same weight (50 µg per sample) were loaded on 12% SDS-PAGE gel, followed by transfer to PVDF membrane (Millipore, USA). The PVDF membranes were probed with the following primary antibodies: Tubulin (1: 5000), A2R (1: 1000), AKT (1: 1000), p-AKT (1: 1000), ERK (1: 1000) and p-ERK (1: 1000) at 4 °C overnight. HRP-linked secondary antibodies were then applied accordingly at room temperature for 2 hrs. Chemiluminescence was used to detect the signals of protein.
Real-time PCR was performed with MonAmp™ SYBR® Green qPCR Mix (MQ10101S, Monad, Shanghai), using a CFX-96 real-time PCR Detection system (Bio-Rad, USA). The relative quantities of total cDNA of ANXA2+/+ mice and wild-type mice were quantified using the ∆∆CT method. The primer sequences used were as follows: mouse Anxa2 (Fw: 5′- ATGTCTACTGTCCACGAAATCCT -3′; Rv: 5′- CGAAGTTGGTGTAGGGTTTGACT -3′), mouse Actb (Fw: 5′- GGCTGTATTCCCCTCCATCG -3′; Rv: 5′- CCAGTTGGTAACAATGCCATGT -3′).
All statistical data are represented as mean±s.e. All experiments were reproduced three – four independent times. For experiments with only two groups, unpaired two-tailed t-tests were used. For multiple comparisons, one-way or two-way ANOVA (with repeated measures when appropriate) followed by Tukey and Bonfferroni corrections were used. Data were considered statistically significant when the p values were less than 0.05.