Background
The blood-brain barrier (BBB) is a physical and functional barrier between the central nervous system (CNS) and blood that maintains CNS homeostasis [
1]. Astrocytes, pericytes, microglia, brain capillary endothelial cells, and even some neurons are critical components of the BBB [
1,
2]. These components are activated in response to acute ischemic stroke (AIS) and then secrete inflammatory factors that directly or indirectly disrupt the integrity of the BBB. In fact, the loss of tight junction proteins (TJPs) between brain capillary endothelial cells is an indicator of BBB destruction [
3,
4]. The role of astrocytes in maintaining the integrity of BBB under physiological and pathological conditions is indispensable. The integrity of in vitro BBB models comprising brain endothelial cells co-cultured with astrocytes and/or pericytes is greater than that of one comprising an endothelial cell monolayer [
5]. However, under pathological conditions, such as traumatic and ischemic injury, astrocytes secrete inflammatory factors, such as vascular endothelial growth factor (VEGF-A), matrix metalloproteinases (MMPs), chemokines, and cytokines; these factors directly or indirectly aggravate brain damage and BBB disruption [
6‐
8]. VEGF-A secreted by astrocytes directly destroys the BBB during cerebral ischemia [
9]. The levels of MMPs, mainly MMP-2 and MMP-9, are increased during stroke and are related to the degradation of the basal lamina and loss of BBB integrity [
6,
10]. Although microglia and neutrophils are the main sources of MMPs, astrocytes have also been reported to express MMPs under pathological conditions [
11,
12]. In addition, astrocytes are a source of the chemokine monocyte chemoattractant protein-1 (MCP-1) and chemokine C-X-C motif ligand 1 (CXCL-1) [
7,
13]. MCP-1 not only influences the cytokines (IL-1β, IL-6 and TNF-α) secreted by astrocytes but also plays a pivotal role in post-ischemic microglial activation [
14]. On the other hand, as a neutrophil chemoattractant, a reduction in CXCL-1 levels impairs the recruitment of neutrophils after cerebral ischemia [
7]. Therefore, astrocytes participate in the complex formation of inflammatory environment and the mechanism of BBB disruption following AIS.
The glucagon-like peptide-1 receptor (GLP-1R) agonist exendin-4 (Ex-4) is a long-acting analog of the endogenous insulinotropic peptide, which has been approved as a treatment for type 2 diabetes mellitus [
15]. As a small molecule, this peptide diffuses across the BBB [
16]. GLP-1R agonists have recently been reported to protect against ischemic stroke by reducing the degradation of TJPs between endothelial cells to maintain BBB stability in both diabetic and non-diabetic middle cerebral artery occlusion (MCAO) models and a model of traumatic brain injury [
15,
16]. On the other hand, GLP-1R agonists reduce microglial activation and neutrophil infiltration in the warfarin-associated hemorrhagic transformation model, thus contributing to maintaining the integrity of the BBB [
17]. However, the mechanism of the protective effect of GLP-1R agonists during AIS is still very uncertain, and no study has assessed the direct effects of GLP-1R agonists on astrocytes during AIS, although astrocytes express GLP-1Rs and contribute to both BBB disruption and inflammation activation [
18]. In the current study, we investigated whether Ex-4 impacted the secretion of astrocytes and how this effect influenced the BBB integrity to provide additional evidence supporting the use of GLP-1R agonists as a treatment for AIS.
Methods
Cell culture
Primary cultures of cortical astrocytes were prepared from the cortices of neonatal 1-day-old C57BL/6 mice using methods described in a previous stuciefly, cortices were dissected from the brains, and the meninges were carefully removed [
19]. Then, the cortices were dissociated into a cell suspension by digestion with 0.25% trypsin for 10 min at 37 °C, and the digestion was stopped with Dulbecco’s Modified Eagle’s Medium containing 4.5 g/L glucose (DMEM; Gibco, New York, NY, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, New York, NY, USA), 2 mmol/L glutamine, 100 units/mL penicillin, and 100 μg/mL streptomycin. After washes with 0.01 M phosphate-buffered saline (PBS), the dissociated cells were seeded at a density of hemispheres/24-well plate or 1.5 hemispheres/6-well plate in 4.5 g/L glucose DMEM and cultured at 37 °C in a humidified atmosphere of 5% CO
2/95% air. After 10–12 days of culture, the cells reached 60–70% confluence and were shaken (250 rpm) for 18 h to minimize microglial contamination. Adherent astrocytes in one of the 6-well plate were detached with trypsin/EDTA, and the recovered cells were plated on new 24-well plate. The cells in the new and original 24-well plates were used to confirm the purity of astrocytes using immunofluorescence staining for glial fibrillary acidic protein (GFAP). Ninety-five percent of the cultured cells were identified as astrocytes, and the adherent cells (primary astrocytes) never subjected to cell passaging in 6- or 24-well plates were used in the following experiments. The mouse brain endothelial cell line bEnd.3 was obtained from American Type Culture Collection (ATCC, Manassas, VA, USA) and cultured in 4.5 g/L glucose DMEM supplemented with 10% FBS, 2 mmol/L glutamine, 100 units/mL penicillin, and 100 μg/mL streptomycin after thawing. Then, the cell medium was replaced with new culture medium every 2–3 days, and the cells were subcultured after trypsinization at a ratio of 1:3 every 3 days until use in the experiments The bEnd.3 cells used in experiments are 3rd-5th passage.
Exposure of astrocytes to the oxygen-glucose deprivation plus reoxygenation treatment
Control cultures were incubated with serum-free 4.5 g/L glucose DMEM at 37 °C in a 5% CO2/95% air atmosphere for 4 h and then refresh normal medium with (Ex4 group) or without Ex4(Medium group). Astrocytes were washed twice with PBS and incubated with glucose-free DMEM [oxygen-glucose deprivation (OGD) medium]. Then, cells were transferred to an anaerobic chamber filled with a gas mixture of 95% N2 and 5% CO2 at 37 °C for 2 h or 4 h as the OGD treatment (OGD+RO group). After OGD treatment, a germ-free glucose solution and FBS were added to the OGD medium to yield 4.5 g/L glucose DMEM supplemented with 10% FBS and cells were cultured for an additional 24 h with or without the corresponding treatment [including 10/50/100/200 nM GLP-1R activator Ex-4 (OGD+RO+Ex4 group) and/or 0.1 μM GLP-1R antagonist exendin-(9-39) (Ex[9–39] (OGD+RO+Ex4+Ex[9–39] group)] (MedChemExpress, USA) and/or 10 μM AG490 (OGD+RO+AG490 group) and 20 μM LY294002 (OGD+RO+Ex4+ LY294002 group) at 37 °C in a humidified atmosphere of 5% CO2/95% air.
Culture of bEnd.3 cells with different types of astrocyte-conditioned media
After culture under normal culture conditions to 70–80% confluence, bEnd.3 cells were cultured with different types of astrocyte-conditioned medium (ACM) for another 24 h before being subjected to different experiments [
20]. The grouping method in experiments involved with ACM was based on astrocyte treatment. The sources of ACM included medium from the Medium group (ACM-Medium), from the Ex-4 group (ACM-Ex4) for 24 h, from the OGD+RO group (ACM-OGD+RO), from the OGD+RO+Ex4 group (ACM-OGD+RO+Ex-4), and from astrocytes treated with OGD for 4 h plus RO for 24 h in the presence of Ex-4 and Ex(9-39) [ACM-OGD+RO+Ex-4+Ex(9-39)]. All bEnd.3 cell monolayers were pretreated with Ex(9-39) before culture with different ACMs to avoid the direct influence of Ex-4 present in the ACM on bEnd.3 cells. Cells were treated with a 10-ng/mL dose of a VEGFR2 inhibitor (VI) (R&D Systems, USA) (ACM-OGD+RO+VI) or a 50-ng/mL dose of the MMP inhibitor batimastat (MI) (MedChemExpress, USA) (ACM-OGD+RO+MI) to block VEGFR2 or MMP-9, respectively, as indicated.
Measurement of transendothelial electric resistance
After the administration of different treatments, the transendothelial electric resistance (TEER) values of confluent bEnd3 cell monolayers were measured using a voltage/ohm meter (Millipore, Billerica, MA, USA). The TEER of blank inserts was subtracted from the measured TEER of each group to reflect that of cell monolayers themselves. The values are expressed as Ω cm2.
Paracellular permeability assay
Briefly, bEnd.3 cells were seeded onto polycarbonate 12-well transwell inserts with a 0.4-μm mean pore size and a 0.33-cm
2 surface area (Corning, USA) at a density of 4 × 10
4 cells/cm
2, and the growth medium was refreshed every other day. The seeded cells were allowed to grow at 37 °C in a 5% CO
2/95% air atmosphere until they reached 70–80% confluence, which was confirmed using phase contrast microscopy. Then, cells were cultured with different ACMs for 24 h. After two washes with PBS, the media in the transwell inserts were replaced with media supplemented with 500 μL of 1 mg/mL sodium fluorescein (NaF), the lower chamber was filled with 1000 μL of normal media, and cells cultured for 1 h at 37 °C under normoxic conditions. Relative fluorescence passing through the chamber was measured as follows: 100 μL of medium in the lower chamber was assayed in triplicate in black 96-well plates (Corning, USA). The fluorescence intensity was measured using an EnSpire Manager (PerkinElmer, USA) multimode plate reader at an excitation wavelength of 460 nm and an emission wavelength of 515 nm [
11]. NaF permeability (μg/cm
2) = total NaF quantity in the lower chamber/the surface area of insert (0.33 cm
2).
Western blotting
Confluent astrocytes or bEnd.3 cells at 70–80% confluence cultured in 6-well plates were exposed to the corresponding culture conditions. After different treatments, total protein was collected by lysing the cells with RIPA buffer (Beyotime Biotechnology, China). The protein content in each well was determined using a BCA protein assay kit (Beyotime Biotechnology, China). Lysates containing equal amounts of total protein were separated on polyacrylamide gels (6% or 10%) and then transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore, USA). Membranes were blocked with 3% bovine serum albumin (BSA) for 1 h at room temperature (RT) and incubated with the following primary antibodies: GAPDH (Beyotime Biotechnology, China); claudin-5, occludin, and zonula occludens-1 (ZO-1) (Invitrogen, USA); endothelial nitric oxide synthase (eNOS) and VEGF-A (Abcam, UK); and phosphorylated phosphatidylinositol 3-kinase (p-PI3K), total PI3K, p-protein kinase B (p-Akt), total Akt, p-phospholipase Cγ (p-PLCγ), total PLCγ, p-protein kinase Cα (p-PKCα), total PKCα, p-Janus kinase 2 (p-JAK2), total JAK2, p-signal transducer and activator of transcription 3 (p-STAT3), and total STAT3 [all from Cell Signaling Technology (CST), USA] at 4 °C overnight. Membranes were washed with Tris-buffered saline containing Tween 20 (TBST) (10 mM Tris, pH 8.0, 150 mM NaCl, and 0.1% Tween 20) before the addition of horseradish peroxidase-conjugated secondary antibodies (CST, USA) for 1 h at RT. After three washes with TBST, the signals were visualized with a Tanon Imaging System (China), and the density of each band was quantified using ImageJ software (National Institutes of Health, USA).
Immunofluorescence staining
Twenty-four hours after MCAO, the rats were anesthetized and first perfused with saline followed by fixation with a 4% buffered paraformaldehyde solution. The brains were removed and postfixed with 4% paraformaldehyde; the paraformaldehyde was then removed and replaced with a 30% sucrose solution overnight. Then, 8-μm coronal sections were obtained using a cryostat. Sections were incubated with 0.3% Triton X-100 at RT. After three washes with PBS three, brain slices were blocked with PBS containing 5% BSA for 1 h at RT. In vitro cell specimens, after different treatment, confluent bEnd.3 cells and astrocytes growing on collagen-coated coverslips were subjected to the aforementioned treatments and washed three times with PBS before 4% paraformaldehyde fixation. Fixed cells were permeabilized with 0.3% Triton X-100 for 15 min and blocked with 10% goat serum in PBS for 1 h at RT. After that, sections or fixed cells were incubated at 4 °C overnight with the following primary antibodies: GFAP (CST); VEGF-A (Abcam); MMP9 (Invitrogen); claudin-5 and and ZO-1 (Invitrogen); occludin (Abcam); and GLP-1R (Bioss). After three washes with PBS, an Alexa Fluor 488- or 555-labeled secondary antibody (Invitrogen) was added and incubated for 1 h at RT. Tissue sections or cells specimens were washed twice with PBS and rinsed with a DAPI solution. Tissue sections were covered with a coverslip that contained a drop of antifade mounting medium. Images of both brain slices and cells on coverslips were acquired with a fluorescence microscope or with an LSM800 confocal microscope.
Immunohistochemistry
Twenty-four hours after MCAO, the rats were anesthetized and first perfused with saline followed by fixation with ice-cold 4% paraformaldehyde. Brain tissues were removed and fixed overnight in 4% paraformaldehyde at 4 °C and then immersed in 30% sucrose. The tissues were cut to 2-mm-thick coronal sections and embedded in paraffin and further sliced at a thickness of 5 μm. The slides were deparaffinized, sequentially rehydrated in graded alcohol, and then immersed in PBS (pH 7.4). The slides were then microwaved for 2 min in antigen-unmasking solution, cooled, and washed three times for 2 min in PBS. Sections were immersed for 25 min in 3% hydrogen peroxide in distilled water to eliminate endogenous peroxidase activity, then blocked in immunohistochemical grade 1% bovine serum albumin in PBS for 1 h and diluted goat serum for 30 min to reduce nonspecific staining. Sections were incubated overnight with primary antibodies anti-CXCL1 (Invitrogen) or anti-MCP1 (Abcam). Peroxidase-conjugated anti-rabbit IgG (BD Biosciences) was used as the secondary antibody and was incubated for 30 min at 37 °C. Antibodies were detected using the DAB kit (Beyotime Biotechnology, China) following the manufacturer’s instructions. Finally, the slides were observed using a light microscope. The positive cells were expressed as the number of immunopositive cells/mm2.
Lactate dehydrogenase assay
The viability of bEnd.3 cells was evaluated by quantifying plasma membrane damage, which resulted in the release of LDH. The level of Lactate dehydrogenase (LDH) released in the cell culture supernatant was detected using an LDH cytotoxicity assay detection kit (Beyotime, China) according to the manufacturer’s instructions.
Enzyme-linked immunosorbent assay
The cell culture supernatants of astrocytes subjected to different treatments were harvested and assayed with Enzyme-linked immunosorbent assays (ELISAs) to determine the levels of tumor necrosis factor-α (TNF-α), interleukin (IL)-1β, IL-6, transforming growth factor-β (TGF-β), CXCL-1, VEGF-A (R&D Systems, USA), and MCP-1 (Sigma-Aldrich, USA) according to the manufacturer’s protocols. The results were calculated by measuring the absorbance at a wavelength of 450 nm using a multiplate reader (Bio-Tek, Winooski, VT, USA). The experiments were performed in triplicate.
Real-time quantitative PCR
Total RNA was extracted from astrocytes using TRIzol (Invitrogen, USA) according to the manufacturer’s instructions. The cDNA templates were synthesized from total RNA using a Transcriptor First Strand cDNA Synthesis kit (TaKaRa Biotechnology). TaqMan primers and probes used for testing were obtained from TaKaRa Biotechnology, Japan. GAPDH was used as an endogenous reference. Quantitative PCR using SYBR Green II (TaKaRa Biotechnology) was performed with an ABI PRISM 7900 Sequence Detector system (Applied Biosystems). Target gene expression was normalized to GAPDH expression, and the values were calculated relative to control values using the ΔΔCT method. The following PCR primers were used (5′ to 3′): IL-1β-F, GACCTGTTCTTTGAGGCTGACA and IL-1β-R, CTCATCTGGACAGCCCAAGTC; IL-6-F, TAGTCCTTCCTACCCCAACTTCC and IL-6-R, TTGGTCCTTAGCCACTCCTTC; TGF-β-F, CATTGCTGTCCCGTGCAGA and TGF-β-R, AGGTAACGCCAGGAATTGTTGCTA; TNF-α-F, TTCCAATGGGCTTTCGGAAC and TNF-α-R, AGGTAACGCCAGGAATTGTTGCTA; VEGF-A-F, TCCTGCAGCATAGCAGATGTGA and VEGF-A-R, CCAGGATTTAAACCGGGATTTC; MMP-2-F, TCCCGAGATCTGCAAGCAAG and MMP-2-R, AGAATGTGGCCACCAGCAAG; MMP-9-F, GGGAACGTATCTGGAAATTCGAC and MMP-9-R, CCGGTTGTGGAAACTCACAC; MCP-1-F, TGTCTCAGCCAGATGCAGTTAAT and MCP-1-R, CCGACTCATTGGGATCATCTT; CXCL-1-F, CAATGAGCTGCGCTGTCAGT and CXCL-1-R, TTGAAGTGAATCCCTGCCACT; and GAPDH-F, GGCACAGTCAAGGCTGAGAATG and GAPDH-R, ATGGTGGTGAAGACGCCAGTA.
Gelatin zymography
The same number of astrocytes was seeded in each well of a 6-well plate and grown to confluence. After the administration of different treatments (all in serum-free media for 24 h), equal amounts of proteins from ACM were collected and concentrated using centrifugal filters. After electrophoresis on an 8% SDS-PAGE gel containing 1 mg/mL gelatin, gels were washed with 2.5% Triton X-100 for 30 min. Then, the gel was incubated in a developing buffer (20 mM Tris-HCl, pH 7.8, 1% Triton X-100, 10 mM CaCl2, 5 μM ZnCl2) for 24 h at 37 °C. Thereafter, the gel was stained with 1% Coomassie Brilliant Blue R-250. The gelatinolytic activities of MMP-9 were detected as transparent bands.
Experimental ischemic stroke model and treatment
All animal experiments were approved by the Ethics Committee of Sun Yat-sen University. Male SD rats weighing 220–250 g were randomly assigned to different groups using a random number table. The rats were anesthetized with 2% pentobarbital (50 mg/kg, i.p.). The MCAO model of ischemic stroke was induced by a left side occlusion of the middle cerebral artery with silicone-coated sutures as previously described [
21]. After 90 min of occlusion, the filament was removed to allow reperfusion. The rats in the treatment group were administered Ex-4 and/or 3000 μg/kg Ex[(9-39)] (in saline at a total volume of 1 mL) or saline by intraperitoneal injections immediately after reperfusion. The doses of Ex-4 were 50, 150, and 300 μg/kg to determine the effective dose that suitable for subsequent experiments.
Neurological deficit score
Neurological function was evaluated at the time rats recovered from anesthesia (approximately 3 h) and 24 h after MCAO. The deficits were scored on a modified scoring system based on that developed by Longa et al. [
21] as follows: 0, no neural defective symptoms; 1, rats cannot stretch the contralateral front paws; 2, rats circle to the contralateral side while crawling; 3, rats tumble to the contralateral side while crawling; 4, rats cannot walk independently or lose consciousness or death. A first score of 2–3 points was considered a successful model, and these rats were included in the experimental group according to the treatment.
Quantification of infarct area
The brain was removed and immediately sliced into 1-mm-thick sections. The slices were then stained with a 2% solution of 2,3,5-triphenyltetrazolium chloride (TTC) at 37 °C for 30 min. The infarcted area of each brain slice was measured by ImageJ analysis software, and the percentage hemisphere lesion area was calculated as follows: {[total infarct area-(left hemisphere area-right hemisphere area)]/right hemisphere area} × 100% [
22].
Leakage of Evans blue (EB, Sigma-Aldrich, USA) dye in the ischemic brain tissue, which is indicative of BBB disruption, was evaluated after the 90-min MCAO procedure (with or without Ex-4/Ex(9-39) treatment) or after the sham operation (sham group). EB (2%) in normal saline (6 mL/kg) was intravenously injected into the animals and allowed to circulate for 3 h before sacrifice. After sacrifice, every 100 mg of the collected ischemic-side brain tissue was added with 1 mL of 50% trichloroacetic acid solution to extract the EB. After homogenization of the brain tissue, the mixture was centrifuged at 15000g for 15 min, and the supernatant was harvested. The supernatant was diluted 4-fold with ethanol, and the mixture was allowed to stand at room temperature for 30 min. The amount of EB in the ischemic tissue was quantified at 610 nm by a spectrophotometer according to a standard curve of optical density (OD) obtained according to different concentrations of EB.
Statistical analysis
All data are presented as the mean ± SD. The data were analyzed using a two-tailed t test or one-way analysis of variance (ANOVA) followed by the post hoc Student-Newman-Keuls (SNK) t test for multiple comparisons. Differences were considered significant at P < 0.05. Statistical analyses were performed using SPSS 18.0 software.
Discussion
Astrocytes are the most abundant cell type within the CNS, and they provide structural support, promote the formation of the BBB, and release beneficial factors that maintain brain cell development and homeostasis of the extracellular environment [
24]. The function of brain endothelial monolayer cells, ATP-binding cassette transporters, and TJP expression for example, is enhanced when the cells are co-cultured with astrocytes or ACM for a period of time [
5,
25]. In the present study, ACM did not alter TJP expression or the integrity of bEnd.3 cell monolayer, potentially due to the relatively short culture period, which was consistent with the results from previous studies [
11,
25]. Although astrocytes are generally more resistant to the effects of AIS than other cells [
26], AIS- or OGD-induced-astrocytes aggravate BBB disruption by inflammatory factors [
27], consistent with our results that ACM from astrocytes treated with OGD+RO destroyed the integrity of the bEnd.3 cell monolayer.
In the mammalian brain, GLP-1R expression has been detected in endothelial cells, neurons, astrocytes, and microglia [
16,
17,
28]. The administration of a GLP-1R agonist in vitro was recently shown to strengthen the integrity of brain endothelial cells and improve cortical neuron survival under pathological conditions [
16,
28] and reduce the permeability of the BBB, microglia activity, and neutrophil infiltration after MCAO in vivo [
17,
29]. However, whether the effect of Ex-4 on BBB integrity (or protecting TJPs) is related to the viability of endothelial cells remains unknown because the effect is accompanied by PI3K/Akt signaling activation, a pathway related to cell viability [
17]. On the other hand, although Ex-4 reduces the inflammatory factors in stroke [
17], Wu et al. and Darsalia et al. found that GLP-1R activation did not inhibit LPS-induced or ischemia-induced spinal or brain expression of microglia-derived pro-inflammatory cytokines [
30,
31]. Our experiment also found that Ex-4 did not reduce the inflammatory factors (IL-1β, IL-6, TNF-α) (Additional file
1: Figure S1) secreted by BV-2 cells after OGD, suggesting that Ex-4 does not directly affect the microglia secretion of inflammatory factors. The above data cannot fully explain the effect and the mechanism of Ex4 on ischemic stroke. However, researchers have not determined whether the effect of Ex-4 on AIS is related to astrocytes. The TJPs between brain endothelial cells are the physical barrier of the BBB that prevent peripheral harmful and superfluous substances from entering the brain to maintain the homeostasis of the intracerebral environment [
1]. In the present study, ACM from astrocytes treated with OGD+RO destroyed the integrity of the bEnd.3 cell monolayer, but this change was reversed by Ex-4 through GLP-1R expressed on astrocytes and was not related to cell viability. Restored TJPs account for the effect of ACM from astrocytes treated with Ex-4, indicating that the components of ACM that were changed by Ex-4 ultimately affected TJPs on bEnd.3 cells to protect BBB integrity.
Astrocytes are a major source of inflammatory factors in lesions after AIS [
6,
7,
10,
13,
14,
32], which contribute greatly to peripheral immune cell recruitment, microglial activation, brain damage aggravation, and BBB disruption [
7,
14]. According to our data, the levels of astrocyte-secreted inflammatory factors were significantly increased by the OGD+RO treatment, consistent with the results of previous studies [
6,
7,
10,
13,
14]. However, only the secretion of MCP-1, CXCL-1, VEGF-A, and MMP-9 was reduced by the Ex-4 treatment after OGD, and all 4 factors were reduced by Ex-4 treatment after MCAO. Further experiments investigating the effects of MCP-1 and CXCL-1 in an MCAO model have not performed, but it was clear that MCP-1 is involved in recruiting monocytes/macrophages and activating microglia [
33,
34], and CXCL-1 is an important chemokine that recruits neutrophils to the area of the brain affected by AIS [
7]. The reduction in the levels of these two chemokines in astrocytes by Ex-4 in vitro and in vivo may partially explain why Ex-4 ameliorated the microglial activation and neutrophil recruitment in previous in vivo experiments [
16,
17].
Further data obtained from in vitro and in vivo experiments confirmed that Ex-4 ameliorated ischemic anoxic attack-induced astrocyte-derived VEGF-A and MMP-9 secretion. VEGF-A was found to induce vessel formation [
35]. However, after AIS, VEGF-A expression is upregulated in astrocytes in the mammalian brain for up to 2 weeks [
36], and VEGF-A increases BBB permeability through VEGFR2 expressed on brain endothelial cells and induces the degradation of TJPs at the beginning of AIS [
9,
37]. MMP-9 levels are increased in subjects with AIS and degrade the neurovascular basal lamina or TJPs to disrupt the BBB [
38,
39]. Astrocytes produce more MMPs in both the MCAO and OGD+RO models [
39,
40]. Our study supported the points and illustrated that Ex-4 protected the TJPs between bEnd.3 cells via an astrocyte-dependent manner.
Ex-4 preserves the BBB integrity after MCAO by increasing PI3K/Akt signaling [
17]. However, VEGF-A activates VEGFR on endothelial cells and mainly signals through the PLCγ/PKCα/eNOS or PI3K/Akt/eNOS pathways to degrade TJPs, which may contradict the results reported by Chen [
17]. eNOS has been shown to participate in increasing the permeability of BBB under pathological conditions [
8]. In the present study, Ex-4 ameliorated eNOS expression in bEnd.3 cells cultured with ACM from OGD-treated astrocytes, and the levels of phosphorylated PLCγ and PKCα showed the same trend, which was consistent with the effect of the eNOS inhibitor on brain endothelial cells [
8]. No ACM altered the PI3K/Akt pathway in bEnd.3 cells, although PI3K/Akt is one of the downstream targets of VEGF-A-VEGFR2. This discrepancy might be attributed to the involvement of the PI3K/Akt pathway in the cell survival/death pathway through the activation of different downstream targets [
41] and demonstrated that the manner of how Ex-4 improved BBB integrity through astrocytes was not related to bEnd.3 cell viability. However, brain endothelial cells are more sensitive to ischemic and anoxic attack than astrocytes [
26]; thus, the direct effect of Ex-4 on brain endothelial cells may also be involved in their increased vitality.
PI3K/Akt is the most common downstream pathway of GLP-1R activation and was found to be activated by Ex-4 in many in vivo stroke models [
7,
42]. However, our previous data showed that Ex-4-induced reduction in VEGF-A was not related to PI3K activation. It is known that HIF-1α plays an important role in BBB breakdown following hypoxic/ischemic injury [
43]. Exposure to OGD increases HIF-1α expression in astrocytes, and one of the downstream proteins of HIF-1α is VEGF-A [
23,
26]. On the other hand, our and other laboratories have found that JAK2/STAT3 or IL-6/JAK2/STAT3 activation leads to increased VEGF-A expression in different cells, including astrocytes [
44,
45]. AG490, a JAK2 inhibitor, has been reported to suppress pathological processes in different diseases by inactivating the JAK2/STAT3 pathway [
45,
46]. In the present study, both Ex-4 and AG490 deactivated the JAK2/STAT3 pathway in astrocytes in vitro and in vivo, accompanied by decreased levels of astrocyte-derived VEGF-A but not the levels of other pro-inflammatory cytokines and HIF-1α after OGD treatment. Based on these results, Ex-4 affected OGD+RO-treated astrocytes by decreasing the level of the phosphorylated JAK2/STAT3 protein, which is a new signaling pathway that has not previously been reported.
Our study has some limitations that should be acknowledged. First, we used only a relevant inhibitor to assess the effects of GLP-1R agonists on astrocytes in vitro and did not knock out the related genes. Similarly, GLP-1R knockout in astrocytes in vivo would make the data more convincing. Second, we only chose some familiar astrocyte-derived factors that are related to the inflammation, BBB disruption, microglia activation, and neutrophil recruitment, which may overlook some factors that are also influenced by Ex-4. Third, we focused on the mechanism by which Ex-4 treatment of astrocytes influences BBB integrity. Although OGD+RO-induced increases in the levels of astrocyte-secreted chemokines were influenced by Ex-4 treatment, we did not further confirm their possible roles in Ex-4-related microglial activation and neutrophil recruitment in vitro or in vivo. Finally, the immediate administration of Ex-4 after reperfusion through an intraperitoneal injection is not an appropriate treatment for patients in the clinic. The route of administration and the dose of this treatment require further study.
Nevertheless, our study presented evidence suggesting a vital effect of GLP-1R agonists on astrocytes in AIS. Because Ex-4 crosses the BBB, pretreatment or treatment with Ex-4 immediately after AIS may improve patient outcomes, particularly patients suffering from diabetes. Clinical trials of GLP-1R agonists and related beneficial inhibitors are needed and will focus on the safety, efficacy, and mechanism of these therapeutics in patients with AIS.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.