Background
During pathological brain insults such as cerebral ischemia, a rapid increase of intracellular calcium initiates apoptotic and necrotic cell death and reactive gliosis. Several lines of evidence suggest that [Ca
2+]
i oscillations evoked by focal ischemia spread through the astrocytes and cause damage in distal regions of the central nervous system (CNS) [
1]. KCa3.1, a Ca
2+-activated potassium channel, is involved in regulating membrane potential during activation of non-excitable inflammatory and structural cells, including astrocytes, in response to Ca
2+ influx [
2‐
4]. As such, KCa3.1 is a promising target to ameliorate the phenotype switch of astrocytes and microglia from resting to astrogliosis and microglia activation in ischemia, traumatic brain injury, Alzheimer’s disease (AD) as well as spinal cord injury [
5,
6].
Most recently, we reported that KCa3.1 was increased in reactive astrocytes as well as neurons in the brains of both mouse models of AD and AD patients. Furthermore, the blockade of KCa3.1 resulted in a decrease in astrogliosis and an attenuation of memory deficits in the AD mouse model [
7]. After ischemic stroke in rats, blood-brain barrier endothelial cells exhibited KCa3.1 protein and activity, and pharmacological blockade of KCa3.1 significantly reduced Na
+ uptake and cytotoxic edema [
8]. Moreover, in a mouse model of ischemia, genetic deletion and wild type mice treated with the KCa3.1 blocker TRAM-34 resulted in a decrease in infarct areas and improved neurological deficits [
9]. These findings suggest that KCa3.1 channels are important in the process of stroke, and that their blockade might prove useful as therapy in stroke with upregulated KCa3.1 expression.
Over the past few years, a large number of studies suggested that astroglial calcium influx after ischemia could be mediated by the activation of Ca
2+ permeable cation channels such as transient receptor potential (TRP) channels. The transient receptor potential vanilloid 4 (TRPV4) channel is widely expressed in the CNS, with demonstrated function not only in neurons and astrocytes but also in endothelial cells of cerebral arteries [
10]. During cerebral ischemia, TRPV4 was over-activated by cytotoxic edema or the metabolites of arachidonic acid (AA), and TRPV4-mediated Ca
2+ entry likely played a role in the intracellular Ca
2+ overload [
11]. Blockade of the TRPV4 channel reduced the damage to hippocampal pyramidal neurons and astrocytes upon oxidative stress [
12] or oxygen-glucose deprivation [
13] in vitro and reduced brain infarction in a mouse model of focal cerebral ischemia [
14]. At present, the signal transduction pathways used by TRPV4 to induce astrogliosis are not well defined but seem mostly related to calcium overload of the cells, as TRPV4 channels are involved in ischemia-induced calcium entry in reactive astrocytes and thus, might participate in the pathogenic mechanisms of astroglial reactivity following ischemic insult [
15].
The present study aims to investigate the mechanisms by which KCa3.1 regulates Ca2+ entry via TRPV4 channels, leading to reactive astrogliosis during ischemia stroke. We show, for the first time, a potential co-localization between KCa3.1 and TRPV4 channels in astrocytes. This co-localization allows KCa3.1 to maintain a negative membrane potential during astrogliosis, thus increasing the driving membrane potential for Ca2+ influx through TRPV4 channels. These strongly suggest that KCa3.1 is involved in reactive astrogliosis in the process of stroke, making it a promising target for the development of novel therapies.
Methods
Permanent focal cerebral ischemia
The study (ethics protocol number: A-2015-010) was approved by the Animal Care and Use Committee of the Shanghai Jiao Tong University School of Medicine, Shanghai, China. KCa3.1
−/− mice were obtained from the Jackson Laboratory as described precisely [
16‐
18]. 10–12 week old male C57BL/6 wild type mice and KCa3.1
−/− mice were housed in a specific pathogen-free animal facility. A permanent focal cerebral ischemia model was prepared in accordance with the guidelines as described previously [
19]. In brief, adult male mice weighing 23 to 27 g were anesthetized with 2% chloral hydrate and body temperature was maintained at 37 ± 0.5 °C throughout the surgery using a heating pad and lamp (ALC-HTP, Shanghai Alcott Biotech Co. Ltd). To induce permanent focal cerebral ischemia, a silicon rubbed-counted suture with a tip diameter of 0.22 mm (L2000, AAA, Guangzhou Jialing Biiotech Co, Ltd.) was inserted into the left external carotid artery to block the middle cerebral artery. Transcranial laser Doppler (moorVMS-LDF2) was used to monitor cerebral blood flow (CBF) to assure reduction of CBF through the surgery (Additional file
1: Figure S1). The sham-operated mice experienced the same surgical operations except for the silicon rubbed-counted suture inserted.
Stroke study population and quality control
The operators were not involved in data analysis and acquisition. The observers performed the surgeries, and parameters evaluation was unaware of the group to which each mouse belonged. The following conditions excluded mice from end-point analyses (exclusion criteria): (1) < 80% reduction in CBF; (2) subarachnoid hemorrhage or the brain parenchyma bleeding (as macroscopically assessed during brain sampling); (3) neurological score = 0 (6 h, 24 h after pMCAO); and (4) operation time > 10 min. In total, 158 mice (86 C57BL/6 WT, 72 KCa3.1−/−) were used in this study. Of the 120 mice subjected to pMCAO, 12 mice (10%) met at least one exclusion criterion after randomization and, therefore, were withdrawn from the study.
Measurement of neurological deficits
Mice were studied for neurological deficits at 3, 6, and 24 h after pMCAO as described previously [
20]. Briefly, neurological findings were scored on a 5-point scale: 0, no observable neurological deficits (normal); (1) failure to extend the right forepaw (mild); (2) circling to the contralateral side (moderate); (3) loss of walking or righting reflex (severe); (4) dead. The observers were unaware of the group to which each mouse belonged.
Determination of infarct volume
Groups of mice were euthanized at 3, 6, and 24 h after MCAO. Brains were quickly removed and chilled at − 20 °C for 20 min to slightly harden the tissue. Then brains were sectioned into five 1 mm-thick coronal slices starting from the frontal pole. All sections were stained with 2% 2,3,5-triphenyltetrazolium hydrochloride in the dark for 20 min at 37°C and flipped every 5 min for staining of anterior and posterior faces [
21]. Finally slices were fixed in 4% paraformaldehyde overnight at 4 °C. ImageJ was used to measure the infarct area of each brain.
Immunohistochemistry
For immunofluorescence staining of serial brain coronal sections (12 μm) and cultured cells, the tissues and cells were blocked with 1% bovine serum albumin and 1% goat normal serum 1 h at room temperature. Sections and cells were incubated at 4 °C overnight with primary antibodies: mouse anti-KCa3.1 (1:100; Alomone Labs), rabbit anti-GFAP (1:500; Dako); rabbit anti-Iba1 (1:500; Abcam); rabbit anti-NeuN antibody (1:500; Millipore), rabbit anti-TRPV4 (1:200; Alomone Labs). The sections and cells were incubated with following secondary antibodies: Alexa Fluor 555 goat anti-rabbit IgG (1:500; Invitrogen), Alexa Fluor 488 goat anti-mouse IgG (1:500; Invitrogen) for 1 h at room temperature. Then washed with PBS and stained with DAPI (4′, 6-diamidino-2-phenylindole).
Data collection and statistics of immunofluorescence
Twelve-micrometer-thick brain slices were collected from mice and four slices at 120 μm intervals from each brain were used to examine GFAP, Iba-1, and NeuN positive cells.
The average optical density of GFAP. At least three microscopic photographs of vision were selected in hippocampus or cortex of each immunofluorescence hemisphere slice. Leica·TCS·SP8 Laser scanning confocal microscope was used to capture photographs, which were obtained under the same confocal settings. The area of the AOI (area of interest, area) and the integral optical density (IOD) value was measured using the Image-Pro Plus 6.0 software as in previous study [
22], and the IOD/area was calculated to obtain the mean IOD in each image. At last, the final mean IOD of each slice was determined by the average of the mean IOD in each image.
Cell counting. At least three fields were captured in each slice with the same reference position for quantification. The numbers of Iba1+ and NeuN+ cells per 1 mm2 in each slice were counted using the Image-Pro Plus 6.0 software. The data from at least three photographs of each hemisphere slice were averaged as one value and values from three slices were calculated.
Analysis of co-localization. For analysis of TRPV4 and KCa3.1 co-localization, images were processed and analyzed using Leica LAS AF Lite software (Leica, Germany). Pearson correlation was used to express the degree of co-localization as described early [
23,
24]. The co-localization of KCa3.1 and TRPV4 in the overlap of the two channels was assessed using the co-localization tool in Leica LAS AF Lite software. The Pearson correlation values range from − 1 to + 1. A correlation of 1 indicates complete co-localization between the two proteins. A correlation of − 1 indicates a negative interaction, and a correlation of 0 indicates no co-localization between the two proteins [
24].
Primary culture of astrocytes
Primary astrocyte cultures were prepared from neonatal (0–2 days old) C57BL/6 wild type or KCa3.1
−/− mouse brains as described previously [
17]. Briefly, the cerebral cortices were dissected out and dissociated into a single cell suspension. When cells grew to confluence (10-14 days later), flasks were shaken overnight (200 rpm, 37 °C) and the medium exchanged to remove adherent microglia and oligodendrocytes. The purified astrocytes were plated into plates in serum-containing DMEM. After once again reaching confluence, the medium was exchanged for serum-free DMEM for 24 h before treatments. In some cases, the cells were pretreated with the blockers 30 min before oxygen-glucose deprivation.
Oxygen-glucose deprivation and drug exposure
Confluent astrocytes were grown in serum-free DMEM for 24 h before OGD, at which time the culture medium was replaced with glucose/glutamine-free DMEM medium after a gentle cell washing with the same medium. The serum-free and glucose/glutamine-free DMEM medium was balanced for 30 min with 95% (
v/
v) N
2, 5% (
v/v) CO
2 at 37°C before OGD. Then cells were exposed to hypoxia for different time points in a small anaerobic chamber filled with 95% (
v/v) N
2 and 5% (
v/v) CO
2 at 37 °C. Drugs and inhibitors were pretreated to the cells 30 min before OGD. A Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Kumamoto, Japan) was used to measure cell viability [
25].
Western blotting
Mouse brain tissues or astrocytes were homogenized in RIPA buffer (25 mM Tris pH 7.4, 150 mM NaCl, 1% NP-40) containing 0.1% sodium dodecyl sulfate (SDS) and 4% protease inhibitor (complete protease inhibitor cocktail, Roche). Tissue lysates were centrifuged at 13500 rpm for 30 min at 4 °C, and supernatants were collected. Total lysates were diluted in 2 × SDS sample buffer (120 mM Tris/HCl, 10% SDS, 20% glycerine, 20% 2-mercaptoethanol, pH 6.8) to a final concentration of 2 μg/μl and were used for western blotting. The following primary antibodies were used: anti-KCa3.1 (1:500; Abcam), anti-TRPV4 (1:500; Alomone Labs), anti-GFAP (1:2000; Z0334, Dako), and anti-β-actin (1:1000; Santa Cruz). HRP-conjugated anti-rabbit or anti-mouse IgG secondary antibodies (1:3000; Amersham Biosciences) were used for 1 h at room temperature. ImageJ software was used to quantify the protein bands and normalized to the actin band, which served as loading control [
26].
Cell viability
Cell viability was assessed using a Cell Counting Kit-8 (CCK-8, Dojindo Molecular Technologies) as previously described [
4]. Briefly, the confluent astrocytes were cultured in serum-free media for 24 h and were then treated with OGD for different time points (1, 3, 4 6, 12 h). Ten microliter CCK-8 was added to each well of the 96-well plate and then was placed in a CO
2 incubator for 2 h. Measure the absorbance at 450 nm with a microplate reader.
Membrane potential measurement
Bis-(1,3-dibutylbarbituric acid) trimethine oxonol [DiBAC4(3)], the potentiometric fluorescent dye, was used to measure membrane potential as described previously [
27]. Briefly, primary cultured astrocytes were loaded with 100 nmol/l DiBAC4 (3) for 20 min to ensure dye distribution across the cell membrane at 37 °C in an incubator. 1-EBIO (200 μM) or RN1747 (10 μM) was added to the cells at 60 s. A TCS SP8 confocal laser-scanning microscope (Leica, Germany) was used to evaluate relative changes in membrane potential by monitoring DiBAC4(3) fluorescence. TRAM-34 (1 and 10 μM) or HC 067047 (10 μM) was added 1 h before the experiment. DiBAC4(3) fluorescence was measured at 530 nm with excitation at 488 nm. Confocal images were taken and stored every 1 s for 1800 s.
[Ca2+]i measurement
Cytosolic Ca
2+ ([Ca
2+]
i) was measured as described previously [
27]. Briefly, astrocytes were loaded with 5 μM Fluo-4 AM (MAIBIO, Shanghai, China) for 30 min at 37 °C in an incubator, rinsed, and incubated in DMEM with the appropriate test reagents. Baseline fluorescence was measured for the first 60 s, and then 1-EBIO (200 μM) or RN1747 (10 μM) was added to the cell plate. A TCS SP8 confocal laser-scanning microscope (Leica, Germany) was used to evaluate relative changes in intracellular calcium concentration ([Ca
2+]
i) by monitoring Fluo-4 fluorescence. TRAM-34 (1 and 10 μM) or HC 067047 (10 μM) was added 1 h before the experiment. Fluo-4 fluorescence was measured at 510 nm, with excitation at 488 nm. Confocal images were taken and stored every 1 s for 360 s.
Statistical analysis
All data are presented as means ± SEM. Statistical analyses were performed using Prism software (GraphPad Software, Inc., La Jolla, CA, USA). Data were tested for Gaussian distribution with the Kolmogorov–Smirnov normality test and then analyzed by one-way ANOVA and Dunnett’s post hoc tests. Data were analyzed with unpaired, two-tailed Student’s t test when comparing between two groups, or the non-parametric Mann–Whitney test was applied. Statistical significance was set at p < 0.05.
Discussion
The major findings of this study are that the KCa3.1 channel contributes to reactive astrogliosis following ischemia, as shown in both OGD-treated astrocytes in vitro and the brains of mice subject to pMCAO in vivo, that KCa3.1 regulated Ca2+ influx and membrane potential by functional cooperation with the TRPV4 channel in OGD-treated astrocytes. The expression of both channels was increased during astrogliosis and blockade either KCa3.1 or TRPV4 attenuated the astrogliosis process. Additionally, double-labeled staining experiments showed the co-localization between TRPV4 and KCa3.1 in mouse brain and primary astrocytes, where they might cooperate to regulate the reactive astrogliosis.
We provided genetic target validation of the role of KCa3.1 by demonstrating that mice with a gene deletion of KCa3.1 exhibited significantly reduced pathology after pMCAO. This was manifested as both reduced infarct area after the insult as well as reduced astrogliosis, microglial activation and neuron loss. This is consistent with recent studies that demonstrated that pharmacological blockade of KCa3.1 or gene silence reduces infarct size and other neurological deficits in rats or mice [
34]. Moreover, in the present study, we extended this hypothesis to show that KCa3.1 acted as an endogenous sensor of Ca
2+ influx caused by activation of TRPV4. We found that blockade of either channel could reduce astrogliosis. In addition, we found that these channels could be co-localization.
The lack of voltage-gated Ca
2+ channels in non-excitable cells allows KCa3.1 to act as Ca
2+ detectors and Ca
2+ amplifiers. Activation of KCa3.1 induced K
+ efflux and membrane hyperpolarization, which ultimately upregulated the driving force for Ca
2+ influx [
35,
36], and further KCa3.1 channels will be activated in a positive feedback loop. KCa3.1-Ca
2+ channels were involved in non-excitable cells activation process [
35]. An interaction between KCa3.1 and TRPV4 channels has been found in several diseases, during which Ca
2+ dynamics induced by KCa3.1 were dependent on Ca
2+ influx via TRPV4 channels [
37,
38]. There was a functional co7upling between KCa3.1 and TRPV4 to regulate Ca
2+ levels leading to pulmonary circulatory collapse and hemorrhage [
39].
An interaction of TRPV4 and KCa channels has been found in several other brain functions. In retinal ganglion cells, activation of TRPV4 induced apoptosis due to Ca
2+ overload [
40]. Shi et al. [
41] reported that Ca
2+ influx via TRPV4 channels involved in infrasound-induced activation of astrocytes and microglia, and then neuronal death. Astrocytes volume was regulated by TRPV4/AQP4 complex through water transport and calcium homeostasis. Over activation of the interaction of TRPV4/AQP4 complex would trigger the pathological swelling and reactive gliosis [
42]. Ca
2+ influx through TRPV4 channels activated the large Ca
2+-activated K
+ channel (BK), and TRPV4/BK functional coupling regulated bladder contractility in the storage phase [
43]. Gene deletion of KCa3.1 channels attenuated lung damage and pulmonary circulatory collapse caused by TRPV4 activation. In addition, these data are consistent with recent studies that suggested that the KCa channels might serve as amplifiers of the Ca
2+ signaling through TRP channels [
44] and data demonstrating a coupling of the two types of channels in osmosensors in the paraventricular nucleus [
45]. Recently, it was reported that calcium-gated K
+ channels of the KCa1.1- and KCa3.1-type couple intracellular Ca
2+ signals to membrane hyperpolarization in mesenchymal stromal cells from the human adipose tissue (Tarasov et al. 2017). It might explain the reason that stimulation of TRPV4 agonist RN1747 still caused membrane hyperpolarization in the KCa3.1
−/− astrocytes as shown in Fig.
6e.
Reactive astrocytes are involved in many pathological processes of CNS diseases, such as stroke, traumatic brain injury, and AD [
46]. During the process of neuroinflammation, activated microglia induce the phenotypic switch of astrocytes from a quiescent to a reactive phenotype. Neurotoxic reactive astrocytes, termed A1 astrocytes, are induced by activated microglia, which lose the ability to support neurons and oligodendrocyte survival and synaptogenesis [
47]. In our previous studies, we showed that a similar phenotype can be induced in culture using TGF-β [
17], and that blockade or deletion of the KCa3.1 channel prevented the emergence of the reactive phenotype. Whether this mechanism of astrocyte activation depends upon the astrocytic TRPV4 channels is a subject for future investigation.
KCa3.1 is thought to regulate microglial activation such as migration and neurotoxicity induced by activated microglia in vitro [
48,
49]. Higher densities of Kv1.3, KCa3.1, and Kir2.1 currents were detected in microglia from the infarcted area of reversible MCAO than that from non-infarcted control brains. Similarly, strong KCa3.1 immunoreactivity was also found on activated microglia/macrophages of human infarcts [
34]. However, we did not find any obvious co-localization between KCa3.1 and microglia in the brains of pMCAO mice by immunofluorescence staining. It may be that the different model of ischemic stroke (reversible MCAO vs pMCAO) as well as the different detection method (patch clamp vs immunofluorescence staining) might be the reason.
Acknowledgements
We thank Dr. Tianle Xu and Dr. Jingjing Wang (Department of Anatomy, Histology and Embryology, Collaborative Innovation Center for Brain Science, Shanghai Key Laboratory for Tumor Microenvironment and Inflammation, Shanghai Jiao Tong University School of Medicine, Shanghai, China.) for kindly providing the Transcranial laser Doppler.