Tyrosine Phosphorylation of the Kv2.1 Channel Contributes to Injury in Brain Ischemia

In brain ischemia, oxidative stress induces neuronal apoptosis, which is mediated by increased activity of the voltage-gated K+ channel Kv2.1 and results in an efflux of intracellular K+. The molecular mechanisms underlying the regulation of Kv2.1 and its activity during brain ischemia are not yet fully understood. Here this study provides evidence that oxidant-induced apoptosis resulting from brain ischemia promotes rapid tyrosine phosphorylation of Kv2.1. When the tyrosine phosphorylation sites Y124, Y686, and Y810 on the Kv2.1 channel are mutated to non-phosphorylatable residues, PARP-1 cleavage levels decrease, indicating suppression of neuronal cell death. The tyrosine residue Y810 on Kv2.1 was a major phosphorylation site. In fact, cells mutated Y810 were more viable in our study than were wild-type cells, suggesting an important role for this site during ischemic neuronal injury. In an animal model, tyrosine phosphorylation of Kv2.1 increased after ischemic brain injury, with an observable sustained increase for at least 2 h after reperfusion. These results demonstrate that tyrosine phosphorylation of the Kv2.1 channel in the brain may play a critical role in regulating neuronal ischemia and is therefore a potential therapeutic target in patients with brain ischemia.


Introduction
Oxidative stress has been shown to be a major risk factor for neuronal apoptosis and neurodegenerative diseases, as seen with Alzheimer's and brain ischemia [1][2][3]. Neuronal cell death is mainly attributable to apoptosis, although cell death mechanisms are complicated and diverse [4]. The oxidative stress seen in ischemia in mammalian neurons is accompanied by enhancement of the K + current [5]. This enhancement results in an efflux of K + and is an essential factor for neuronal apoptosis [6]. Additionally, a number of studies have reported that caspase activation, the mitochondrial membrane potential, and overall cellular volume are regulated by the necessary abundance of intracellular K + levels, in addition to cell osmolality [7]. It is understandable then that K + efflux initiates the apoptotic cascade, indicated by cell shrinkage and mitochondrial cytochrome-c release [8]. Further K + loss after any apoptotic stimulus can be caused by increased activation of voltage-gated K + (K v ) channels [6]. Blocking K v channels effectively attenuates cell death in many apoptotic models, which can be pharmacologically accomplished through the administration of

Oxidative Stress Induces Tyrosine Phosphorylation of K v 2.1
In an effort to determine whether oxidative stress is associated with K v 2.1 channel tyrosine phosphorylation, we observed the phosphorylation changes in HEK293 cells transiently expressing the rat K v 2.1 channel. The cells were treated with increasing concentrations of 2,2 -dithiodipyridine (DTDP), a sulfhydryl oxidizing agent. The channel proteins were immunopurified with an anti-phosphotyrosine (PY20) and analyzed by immunoblotting with an anti-K v 2.1 (K89/34) mAb in order to detect the tyrosine phosphorylated K v 2.1 channel. We found that the administration of DTDP treatment in a concentration-dependent manner decreased the expression levels of the K v 2.1 channel protein, whereas the tyrosine phosphorylation levels of the K v 2.1 channel were shown to be significantly increased ( Figure 1A). We next examined the viability of cells under oxidative stress due to the presence or absence of the K v 2.1 channel. The vulnerability of these cells to DTDP-induced death significantly increased in the K v 2.1-expressing cells ( Figure 1B).
We used mass spectrometry in previous studies to identify two novel tyrosine phosphorylation sites (Y686 and Y810) on the K v 2.1 channel, also showing that the Y124, Y686, and Y810 residues on K v 2.1 are directly phosphorylated by Src kinase and are involved in the K v 2.1 channel activity [16]. Moreover, Src kinase activity has been proven to rise after DTDP-induced cell death, while tyrosine phosphorylation of the K v 2.1 channel is induced after oxidative stress through activated Src kinases [19]. Here, we performed a comparative analysis of the role of the WT K v 2.1 channel with that of the K v 2.1 channel with non-phosphorylatable mutations in response to DTDP-induced cell death. We transfected pEGFP-C1, K v 2.1-WT-GFP, and three non-phosphorylatable mutants (K v 2.1-Y124F-GFP, K v 2.1-Y686F-GFP, and K v 2.1-Y810F-GFP) into HEK293 cells. Cells were then treated with DTDP (200 µM, 10 min). The tyrosine phosphorylated K v 2.1 protein was then immunopurified with an anti-PY20 mAb, before undergoing immunoblotting with an anti-K v 2.1 (K89/34) mAb. The DTDP-induced tyrosine phosphorylation levels were not different between the K v 2.1-WT, K v 2.1-Y124F, and K v 2.1-Y686F channels, whereas the tyrosine phosphorylation levels of the K v 2.1-Y810F mutation decreased significantly compared to levels seen in the K v 2.1-WT and other channel mutations. However, despite replacing each tyrosine residues (Y124, Y686, or Y810) with phenylalanine, the overall expression levels of the K v 2.1 mutations were not different from that of the K v 2.1-WT ( Figure 1C). Cells were treated with the oxidant DTDP (200 µM, 10 min) and then subjected to cell viability analyses with WST-1 reagent. The cell viability analysis of the DTDP-treated cells showed that viability decreased in cells expressing K v 2.1-WT, K v 2.1-Y124F, and K v 2.1-Y686F compared to GFP-expressing cells, whereas viability was increased in cells expressing the K v 2.1-Y810F mutation ( Figure 1D). The results of the cell viability and immunochemical assays were similar for viability and K v 2.1 tyrosine phosphorylation levels in DTDP-treated cells. These results suggest that increased tyrosine phosphorylation levels of K v 2.1 may be a major contributor to DTDP-induced cell death, and the Y810 residue is a key regulatory phosphorylation site. significantly compared to levels seen in the Kv2.1-WT and other channel mutations. However, despite replacing each tyrosine residues (Y124, Y686, or Y810) with phenylalanine, the overall expression levels of the Kv2.1 mutations were not different from that of the Kv2.1-WT ( Figure 1C). Cells were treated with the oxidant DTDP (200 µM, 10 min) and then subjected to cell viability analyses with WST-1 reagent. The cell viability analysis of the DTDP-treated cells showed that viability decreased in cells expressing Kv2.1-WT, Kv2.1-Y124F, and Kv2.1-Y686F compared to GFP-expressing cells, whereas viability was increased in cells expressing the Kv2.1-Y810F mutation ( Figure 1D). The results of the cell viability and immunochemical assays were similar for viability and Kv2.1 tyrosine phosphorylation levels in DTDP-treated cells. These results suggest that increased tyrosine phosphorylation levels of Kv2.1 may be a major contributor to DTDP-induced cell death, and the Y810 residue is a key regulatory phosphorylation site.

Regulation of Apoptotic Cell Death by Tyrosine Phosphorylation of Kv2.1
In order to identify the tyrosine phosphorylation sites (Y124, Y686, or Y810) on the Kv2.1 channel that are associated with DTDP-induced apoptotic cell death, we treated cells expressing the mutations Kv2.1-Y124F, Kv2.1-Y686F, and Kv2.1-Y810F with the oxidizing agent DTDP (200 µM, 10 min). We observed the expression of cleaved PARP-1 protein as an apoptotic marker. All of the DTDP-treated cells showed a striking decrease in the expression of the cleaved PARP-1 protein

Regulation of Apoptotic Cell Death by Tyrosine Phosphorylation of K v 2.1
In order to identify the tyrosine phosphorylation sites (Y124, Y686, or Y810) on the K v 2.1 channel that are associated with DTDP-induced apoptotic cell death, we treated cells expressing the mutations K v 2.1-Y124F, K v 2.1-Y686F, and K v 2.1-Y810F with the oxidizing agent DTDP (200 µM, 10 min). We observed the expression of cleaved PARP-1 protein as an apoptotic marker. All of the DTDP-treated cells showed a striking decrease in the expression of the cleaved PARP-1 protein ( Figure 2A). Similar to the results of the viability tests, the tyrosine mutants of K v 2.1-expressing cells showed a decrease in cleaved PARP-1 protein, when compared with the K v 2.1-WT-expressing cells. Particularly, K v 2.1-Y810F-expressing cells showed approximately a 35% decrease in cleaved PARP-1 protein ( Figure 2B). These results indicate that the Y810 residues of the K v 2.1 channel may play an important role in apoptotic cell death caused by oxidative stress.
Int. J. Mol. Sci. 2020, 21, x 4 of 12 ( Figure 2A). Similar to the results of the viability tests, the tyrosine mutants of Kv2.1-expressing cells showed a decrease in cleaved PARP-1 protein, when compared with the Kv2.1-WT-expressing cells. Particularly, Kv2.1-Y810F-expressing cells showed approximately a 35% decrease in cleaved PARP-1 protein ( Figure 2B). These results indicate that the Y810 residues of the Kv2.1 channel may play an important role in apoptotic cell death caused by oxidative stress.

Brain Ischemia Results in Tyrosine Phosphorylation of the Kv2.1 Channel
We generated an anti-pY810-Kv2.1 to detect the phospho-Y810 residue of Kv2.1, thus elucidating whether the previously identified tyrosine phosphorylation sites on the HEK293 Kv2.1 channel [16] are phosphorylated on the native Kv2.1 channel expressed in the ischemic brain ( Figure 3A). To that end, we analyzed the brain sections of two-vessel occlusion mice killed 45 min or 3 days after 20 min of global ischemia. After 3 days of reperfusion, brains were perfusion-fixed with paraformaldehyde, and NeuN or Fluoro-Jade B staining was performed in order to discover the

Brain Ischemia Results in Tyrosine Phosphorylation of the K v 2.1 Channel
We generated an anti-pY810-K v 2.1 to detect the phospho-Y810 residue of K v 2.1, thus elucidating whether the previously identified tyrosine phosphorylation sites on the HEK293 K v 2.1 channel [16] are phosphorylated on the native K v 2.1 channel expressed in the ischemic brain ( Figure 3A). Int. J. Mol. Sci. 2020, 21, x 4 of 12 ( Figure 2A). Similar to the results of the viability tests, the tyrosine mutants of Kv2.1-expressing cells showed a decrease in cleaved PARP-1 protein, when compared with the Kv2.1-WT-expressing cells. Particularly, Kv2.1-Y810F-expressing cells showed approximately a 35% decrease in cleaved PARP-1 protein ( Figure 2B). These results indicate that the Y810 residues of the Kv2.1 channel may play an important role in apoptotic cell death caused by oxidative stress.

Brain Ischemia Results in Tyrosine Phosphorylation of the Kv2.1 Channel
We generated an anti-pY810-Kv2.1 to detect the phospho-Y810 residue of Kv2.1, thus elucidating whether the previously identified tyrosine phosphorylation sites on the HEK293 Kv2.1 channel [16] are phosphorylated on the native Kv2.1 channel expressed in the ischemic brain ( Figure 3A). To that end, we analyzed the brain sections of two-vessel occlusion mice killed 45 min or 3 days after 20 min of global ischemia. After 3 days of reperfusion, brains were perfusion-fixed with paraformaldehyde, and NeuN or Fluoro-Jade B staining was performed in order to discover the To that end, we analyzed the brain sections of two-vessel occlusion mice killed 45 min or 3 days after 20 min of global ischemia. After 3 days of reperfusion, brains were perfusion-fixed with paraformaldehyde, and NeuN or Fluoro-Jade B staining was performed in order to discover the surviving neurons and apoptotic cells in the hippocampus CA1 region. In the ischemia group, the number of NeuN-positive neurons was shown to have decreased compared with the control, whereas Fluoro-Jade B-positive apoptotic cells had significantly increased ( Figure 4A). Previous studies demonstrated that K v 2.1 channels form distinct clusters that are restricted to the neuronal cell membrane of the somatodendritic regions. In addition, pathogenic conditions, such as brain ischemia and spinal cord injury, lead to K v 2.1 dispersion [12,20]. Thus, we went on to determine whether K v 2.1 tended to cluster in the hippocampal CA1 region. We found that the K v 2.1 channel was entirely dispersed over the somatodendritic membrane of neurons 45 min after brain ischemia. However, clustering of the K v 2.1 channel was observed three days after ischemia ( Figure 4B).
clustering of the Kv2.1 channel was observed three days after ischemia ( Figure 4B).
In order to determine whether ischemic brain injury mediates tyrosine phosphorylation on the Kv2.1 channel, we performed immunoblotting and immunoprecipitation analyses on the ischemiainjured brain tissue of rats that had undergone four-vessel occlusion using tyrosine phospho-specific Kv2.1 channel antibodies. After 20 min of ischemia and the subsequent reperfusion, the Kv2.1 channel protein was immunopurified with PY20 mAb from the brain tissue of the control animals, as well as from brains 2 and 20 h after undergoing experimentally induced ischemia. The total protein levels of the Kv2.1 channel decreased 2 h after ischemia but were already restored to levels seen in the control by 20 h after ischemia. Importantly, although Kv2.1 protein expression had significantly decreased 2 h after ischemia, the tyrosine phosphorylation level of the Kv2.1 channel was markedly increased at the same time, whereas the tyrosine phosphorylation of the channel could not be detected 20 h after brain ischemia ( Figure 4C). We then investigated whether the Y810 residue of Kv2.1 was phosphorylated as a result of acute ischemic brain injury in two-vessel occlusion mice. After reperfusion, the total Kv2.1 protein level decreased, in rat brain membranes whereas the tyrosine phosphorylation level markedly increased ( Figure 4D). These results were consistent with our immunostained brain tissue findings, demonstrating that acute ischemic stress enhanced Kv2.1 tyrosine phosphorylation, irrespective of the methods used.  In order to determine whether ischemic brain injury mediates tyrosine phosphorylation on the K v 2.1 channel, we performed immunoblotting and immunoprecipitation analyses on the ischemia-injured brain tissue of rats that had undergone four-vessel occlusion using tyrosine phospho-specific K v 2.1 channel antibodies. After 20 min of ischemia and the subsequent reperfusion, the K v 2.1 channel protein was immunopurified with PY20 mAb from the brain tissue of the control animals, as well as from brains 2 and 20 h after undergoing experimentally induced ischemia. The total protein levels of the K v 2.1 channel decreased 2 h after ischemia but were already restored to levels seen in the control by 20 h after ischemia. Importantly, although K v 2.1 protein expression had significantly decreased 2 h after ischemia, the tyrosine phosphorylation level of the K v 2.1 channel was markedly increased at the same time, whereas the tyrosine phosphorylation of the channel could not be detected 20 h after brain ischemia ( Figure 4C). We then investigated whether the Y810 residue of K v 2.1 was phosphorylated as a result of acute ischemic brain injury in two-vessel occlusion mice. After reperfusion, the total K v 2.1 protein level decreased, in rat brain membranes whereas the tyrosine phosphorylation level markedly increased ( Figure 4D). These results were consistent with our immunostained brain tissue findings, demonstrating that acute ischemic stress enhanced K v 2.1 tyrosine phosphorylation, irrespective of the methods used.

Y810 Phosphorylation Affects p38-Mediated Phosphorylation of K v 2.1 at S800
It has been reported in previous studies that the S800 residue of the K v 2.1 channel is directly phosphorylated by p38 MAPK during apoptosis and that Src directly influences the phosphorylation of this residue [9,18]. In order to examine whether Y810 phosphorylation influences the phosphorylation of S800, we generated an anti-pS800 K v 2.1 ( Figure 3B) and then expressed K v 2.1-Y124F, K v 2.1-Y686, and K v 2.1-Y810 mutations in HEK293 cells. We then analyzed the mutants by immunoblotting with anti-K v 2.1 (K89/34) mAb and anti-pS800-K v 2.1. The S800 v2.1 residue was shown to be clearly phosphorylated following oxidant treatment in the K v 2.1-WT, whereas S800 was only weakly phosphorylated in cells expressing the K v 2.1-Y810F mutation ( Figure 5A,B). These findings indicate that the Y810 phosphorylation site on the K v 2.1 channel is a key regulatory site and may affect the function of the S800 residue in oxidative stress-induced apoptosis.

Y810 Phosphorylation Affects p38-Mediated Phosphorylation of Kv2.1 at S800
It has been reported in previous studies that the S800 residue of the Kv2.1 channel is directly phosphorylated by p38 MAPK during apoptosis and that Src directly influences the phosphorylation of this residue [9,18]. In order to examine whether Y810 phosphorylation influences the phosphorylation of S800, we generated an anti-pS800 Kv2.1 ( Figure 3B) and then expressed Kv2.1-Y124F, Kv2.1-Y686, and Kv2.1-Y810 mutations in HEK293 cells. We then analyzed the mutants by immunoblotting with anti-Kv2.1 (K89/34) mAb and anti-pS800-Kv2.1. The S800 v2.1 residue was shown to be clearly phosphorylated following oxidant treatment in the Kv2.1-WT, whereas S800 was only weakly phosphorylated in cells expressing the Kv2.1-Y810F mutation ( Figure 5A,B). These findings indicate that the Y810 phosphorylation site on the Kv2.1 channel is a key regulatory site and may affect the function of the S800 residue in oxidative stress-induced apoptosis.

Discussion
K v 2.1 channel dephosphorylation and cluster dispersion as a result of altered neuronal activity are elicited by excitatory stimuli like ischemia, spinal cord injury, seizures in vivo, or by glutamate treatment, serum deprivation, or oxidative stress, in vitro [5,[20][21][22]. A role for K v 2.1 has been reported as a pro-apoptotic neuronal trigger, but the mechanism for this is not fully understood. In the present work, we explored the role of K v 2.1 tyrosine phosphorylation in the neuronal apoptosis induced by brain ischemia.
We found that the K v 2.1 channel undergoes rapid tyrosine phosphorylation after oxidative stress in transfected HEK293 cells and in neurons after global brain ischemia (Figures 1A and 4C,D). It has previously been shown that ischemia regulates the rapid tyrosine phosphorylation of the K v 1.2 channel and that persistent neuronal depolarization and enhanced intracellular calcium and zinc concentrations are induced by ischemia [23]. Convergent calcium and zinc signaling regulate this apoptotic K v 2.1 channel, and its tyrosine phosphorylation and activity are increased with increasing intracellular zinc concentrations [18,22,24]. Moreover, K v 2.1 is functionally modulated by zinc and calcium in response to ischemia [13,25]. Therefore, the elevation of intracellular calcium and zinc concentrations induced by ischemic injury in neurons after brain ischemia [26] may lead to tyrosine phosphorylation of K v 2.1 within a short time ( Figure 4C,D). We also found that tyrosine phosphorylation of K v 2.1 was no longer present 20 h after ischemia ( Figure 4C). In other words, neuronal K v 2.1 was transiently tyrosine phosphorylated after brain ischemia. This observation is supported by previous studies reporting that intracellular calcium levels and the activity of the Src family kinases (SFKs) are increased by brain ischemia [27]. The SFK pathway is also involved in neuronal cell apoptosis in response to oxidative stress conditions [22,28]. We had previously observed that the K v 2.1 channel is tyrosine phosphorylated and that the K + efflux is induced by Src kinase [16]. Therefore, it is reasonable to assume that tyrosine phosphorylation of the K v 2.1 channel after brain ischemia is regulated by Src kinase.
Several studies have shown that mutating specific tyrosine phosphorylation residues bestows in neuroprotective and anti-apoptotic effects [29,30]. Src kinase activity is induced by oxidative stress [31] and we previously reported that the Y124, Y686, and Y810 residues of the K v 2.1 channel are directly phosphorylated by Src kinase [16]. In the present work, we found that a mutation of Y810F in the K v 2.1 channel showed a significant decrease in tyrosine phosphorylation during oxidative stress ( Figure 1C). Moreover, the cells expressing K v 2.1-Y810F showed a higher survival rate during oxidative stress than cells with other mutations ( Figure 1D). Thus, we can conclude that Y810 phosphorylation on K v 2.1 may be a major contributor to oxidative stress-induced apoptosis.
We also showed that K v 2.1 clustering was dispersed after brain ischemia ( Figure 4B). It has already been shown that the clustering of the K v 2.1 channel is indicative of a restricted localization in the somatodendritic plasma membrane [21], although the specific role of K v 2.1 declustering and its association with neuronal apoptosis in brain ischemia is not fully understood. The K v 2.1 channel is also localized to lipid drafts in the brain, and the current density and location of the K v 2.1 channels in lipid rafts are altered by cholesterol depletion of the cell membrane [32]. We previously reported that K v 2.1 channel activity is dynamically changed by Src-mediated tyrosine phosphorylation [16]. Thus, it will be important to conduct future studies that can help clarify whether tyrosine phosphorylation of K v 2.1 can regulate the channel clustering in the plasma membrane.
The present work showed that the Y124, Y686, and Y810 residues of the K v 2.1 channel are involved after they are phosphorylated during oxidative stress-induced neuronal apoptosis. The coordination of Y124 and S800 residues has been reported to regulate the channel activity of K v 2.1 in oxidative stress-induced apoptosis [22]. The Y124 and S800 residues of K v 2.1 are phosphorylated by Src and p38 kinase, respectively. Additionally, p38 kinase is activated via apoptosis signal-regulating kinase 1 in oxidant-stimulated zinc release [22,33]. Oxidative stress-induced phosphorylation of S800 increases K v 2.1 currents, blocking toxicity through p38 kinase inhibition [9]. Src kinase-mediated Y124 phosphorylation is inhibited by the cytoplasmic protein tyrosine phosphatase ε (Cyt-PTPε) [15]. Inhibition of Src kinase activity blocks the apoptotic K + current surge and overexpression of Cyt-PTPε inhibits K + current, thus performing a neuroprotective function [15,22]. Src and p38 kinase-mediated K v 2.1 phosphorylation have been suggested as regulators of K + current and cell survival during apoptosis [18]. In the present study, in agreement with previous results [18,22], Y124, Y686, and Y810 K v 2.1 mutations decreased the phosphorylation of S800 under the pathological conditions induced by oxidant treatment, with the Y810 mutation serving as the predominant blocker of S800 phosphorylation ( Figure 5). Therefore, coordinating the phosphorylation of Y810 may regulate the activity of the K v 2.1 channel after ischemia, and thus neuronal apoptosis. Our findings also suggest that the K v 2.1 channel Y810 residue can be added to the list of ischemic regulatory factors mediated by Src kinase.
In summary, we demonstrated that oxidative stress-induced neuronal apoptosis promoted the K v 2.1-Src kinase signaling pathway, decreased the expression of the K v 2.1 channel protein, and increased the tyrosine phosphorylation of K v 2.1. Most importantly, the Y810 residue of the K v 2.1 channel was found to be a critical site for phosphorylation in oxidative stress-induced neuronal apoptosis. Our findings support the potential pharmacological targeting of the K v 2.1 channel, which will likely be beneficial in combatting brain damage. Therefore, inhibiting the K v 2.1 tyrosine phosphorylation induced by ischemia may be a therapeutic target during early neuronal apoptotic conditions.

Animals
Adult 8-week-old male C57BL/6 mice or SD rats (Daehan Biolink, Chungbuk, Korea) were used for the animal experiments. All animals were housed in groups in temperature-controlled (20 ± 2 • C) housing with free access to food and water and exposed to a 12-h light/dark cycle. The surgical interventions and postoperative animal care were performed in accordance with the Guidelines and Policies for Rodent Survival Surgery from the Animal Care Committee of Kyung Hee University (Approval Number KHSASP-19-087).

Phospho-Specific Antibodies
Synthetic phosphopeptides containing phosphoserine amino acids (S800, with amino acids 773-808, ESSPLPT(pS)PKFLRPNC and Y810, with amino acids 805-818, RPNCV(pY)SSEGLTGK) or non-phosphopeptides containing serine at the corresponding positions were synthesized by Quality Controlled Biochemicals (AbClon, Seoul, Korea). Phosphopeptides were conjugated to keyhole limpet hemocyanin (1 milligram peptide per milligram of carrier protein) and injected into rabbits for the production of antisera (AbClon, Seoul, Korea). In order to achieve affinity purification, both the phosphopeptides and the non-phosphopeptides were conjugated to SulfoLink coupling gel (Pierce) via their terminal cysteine residues, and phospho-specific antibodies were affinity-purified with a two-step modification of the standard procedures [12]. Briefly, polyclonal sera were passed over the respective phosphopeptide beads and bound and eluted antibodies were immunoadsorbed against the respective non-phosphopeptide beads, in order to remove non-phospho-specific antibodies. Next, phospho-specificity was verified by immunoblot analyses against extracts from HEK293 cells expressing the WT K v 2.1, as well as the respective phosphorylation site mutants, after verifying the comparable immunoreactivity of the immunoblot samples using a general anti-K v 2.1 monoclonal antibody (K89/34, Neuromab, CA, USA).

Immunoprecipitation
A crude brain membrane fraction was prepared, as previously described [21]. Animals were killed by rapid decapitation, whereupon the brains were collected and homogenized in buffer (5 mM sodium phosphate, pH 7.4, 320 mM sucrose, 100 mM NaF, and a protease inhibitor cocktail containing 2 µg/mL aprotinin, 2 µg/mL antipain, 1 µg/mL leupeptin, and 10 µg/mL benzamidine). Homogenates were then centrifuged at 800× g for 10 min at 4 • C. The supernatants were centrifuged at 38,000× g for 90 min at 4 • C. Total membrane protein (4 mg) was solubilized in a lysis buffer (1% Triton X-100, 150 mM NaCl, 1 mM EDTA, 50 mM Tris-HCl, pH 7.4), 1 mM activated sodium orthovanadate, 5 mM NaF, and the protease inhibitor cocktail for 2 h at 4 • C. This was followed by centrifugation at 13,200 rpm for 40 min at 4 • C and the supernatants were incubated with PY20 antibody overnight at 4 • C. Next, 50 µL of protein G-Sepharose beads were added for 2 h at 4 • C. The beads were washed three times in lysis buffer, and the immunopurified proteins were eluted by boiling in an SDS sample buffer.

Two-Vessel Occlusion
A midline incision was made between the neck and sternum, in order to expose the trachea and the right and left common carotid arteries were carefully separated. Cerebral ischemia was induced by clamping both arteries with two miniature artery clips for 20 min. The clips were then removed to allow reperfusion of blood through the carotid arteries. The control mice underwent the same surgical procedure without the artery occlusion. During the surgery, body temperature was monitored with a rectal probe and was maintained at 37.0 • C, using a temperature-controlled Homeothermic Blanket System (Harvard Apparatus, Holliston, MA, USA).

Four-Vessel Occlusion
Both vertebral arteries were heat-cauterized at each alar foramen using a soldering iron (Change-A-tip™ cautery, Bovie Medical Corporation, Purchase, NY, USA) and both common carotid arteries were occluded with vascular clamps for 20 min the following day. During surgery, body temperature was monitored with a rectal probe and was maintained at 37 • C using a temperature-controlled Homeothermic Blanket System (Harvard Apparatus, Holliston, MA, USA).

Immunohistochemistry
In order to analyze the time course of the neurodegeneration in the hippocampal CA1 layer, the animals were allowed to survive for 45 min or 3 days after global ischemia. They were then anesthetized with Zoletil 50 (Virbac, Carros, France) and transcardially perfused with 4% paraformaldehyde in PBS. The fixed brain was subsequently equilibrated in 30% sucrose solution in order to ensure cryoprotection and was sectioned into 30-µm-thick coronal sections using a cryostat (CM1850, Leica, Germany). The tissue sections were washed in PBS and treated for 30 min in a 1% H2O2 PBS solution. The sections were then incubated with anti-NeuN (1:1000, Millipore) to prepare for the analysis of neuronal cell death. They were washed before biotinylated goat anti-mouse IgG (1:200, Vector Laboratories) was used as the secondary antibody. Signals were amplified using an avidin-peroxidase complex (ABC) kit (1:100, Vector Laboratories). After being allowed to react with 3,3 -diaminobenzidine (Sigma-Aldrich Korea), the sections were dehydrated using ethanol and xylene. They were then mounted with neutral resin. For the analysis of neuronal apoptosis, the sections were incubated for Fluoro-Jade B (Histo-Chem, Inc., AR, USA), following the manufacturer's instructions [34]. In order to analyze K v 2.1phosphorylation and clustering, the sections were incubated with anti-K v 2.1 (K89/34, NeuroMab, CA, USA) and in-house phospho-specific antibodies (anti-pan-pY-K v 2.1). These sections were washed and were probed with Alexa 594-conjugated goat anti-mouse and Alexa 488-conjugated goat anti-rabbit (1:500, Invitrogen) antibodies for 1 h at room temperature. After being washed again, the sections were mounted onto glass slides and cover slipped for microscopic analysis. All images were acquired with the Zeiss LSM 710 confocal microscope (Carl Zeiss, Oberkochen, Germany).

Cell Viability
Transfected cells were grown to confluence in 96-well plates and treated with a stress-inducing media of 2,2 -dithiodipyridine (DTDP) at a concentration of 200 µM for 10 min (37 • C, 95% air, 5% CO2). This was subsequently replaced with fresh media (110 µL) containing the Ez-Cytox reagent (WST-1, Daeil Lab, Korea) and cells were then incubated for 4 h under normal cell culture conditions. After this incubation, the absorbance of each well on the plate was measured at a wavelength of 450 nm using a VersaMax multiwall plate spectrophotometer (Molecular Devices, Sunnyvale, CA, USA).

Statistical Analysis
Data are expressed as mean ± S.E.M. for all three independent experiments, unless otherwise specified for each figure. The Student's t test was used to determine significant difference between means and p < 0.05 was considered statistically significant.

Conflicts of Interest:
The authors declare that they have no competing interests.