Perampanel Affects Up-Stream Regulatory Signaling Pathways of GluA1 Phosphorylation in Normal and Epileptic Rats

To elucidate the pharmacological properties of perampanel [2-(2-oxo-1-phenyl-5-pyridin-2-yl-1,2-dihydropyridin-3-yl)benzonitrile, a novel non-competitive antagonist of AMPA receptor], we investigated its effects on the up-stream regulatory pathways of GluA1 phosphorylation including protein kinase C (PKC), Ca2+-calmodulin-dependent protein kinase II (CAMKII), protein kinase A (PKA), extracellular signal-regulated kinase 1/2 (ERK1/2), c-Jun N-terminal kinase (JNK), protein phosphatase (PP) 1, PP2A, and PP2B in normal and pilocarpine-induced epileptic rat model using Western blot analysis. In normal animals, perampanel affected GluA1 expression/phosphorylation, PKC, CAMKII, PKA, ERK1/2, JNK, and PPs activities. In epileptic rats, perampanel effectively inhibited spontaneous seizure activities. Perampanel enhanced phospho (p)-GluA1-S831 and -S845 ratios (phosphoprotein/total protein), while it reduced GluA1 expression. Perampanel also increased pCAMKII and pPKA ratios, which phosphorylate GluA1-S831 and -S845 site, respectively. Perampanel elevated pJNK and pPP2B ratios, which phosphorylates and dephosphorylates both GluA1-S831 and -S845 sits. Perampanel also increased pERK1/2 ratio in epileptic animals, while U0126 (an ERK1/2 inhibitor) did not affect pGluA1 ratios. Perampanel did not influence PKC, PP1, and PP2A expression levels and their phosphorylation ratios. In addition, perampanel did not have a detrimental impact on cognitive abilities of epileptic and normal rats in Morris water maze test. These findings suggest that perampanel may regulate AMPA receptor functionality via not only blockade of AMPA receptor but also the regulations of multiple molecules (CAMKII, PKA, JNK, and pPP2B)-mediated GluA1 phosphorylations without negative effects on cognition, although the effects of perampanel on PKC, PP1, and PP2A activities were different between normal and epileptic rats.


INTRODUCTION
Glutamate is one of the excitatory neurotransmitters, which is involved in various physiological functions of the brain, including synaptic plasticity. The glutamate action is mediated by ionotropic and metabotropic receptors (Greenamyre and Porter, 1994;Michaelis, 1998). α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor is one of ligand-gated ion channels for glutamate (Traynelis et al., 2010). AMPA receptors have four subunits (GluA1-GluA4), which compose the pentameric structures of the receptor (Borges and Dingledine, 1998;Dingledine et al., 1999). AMPA receptors influx Na + ion and, to a lesser extent, Ca 2+ (Hollmann et al., 1991). AMPA receptor plays a role in fast postsynaptic depolarization, and is critical to seizure initiation, epileptic synchronization and the spread of seizure activity. Thus, pharmacological inhibitors of AMPA receptors may have potentials as a therapeutic approach for the treatment of epilepsy (Rogawski, 2013).
The phosphorylation of GluA1 subunits increases the conductance of AMPA receptor and potentiates rapid excitatory neurotransmission (Derkach et al., 1999). The carboxyl terminus of GluA1 subunit at serine (S) residues 831 (S831) and 845 (S845) are phosphorylated by protein kinases . S831 site is phosphorylated by protein kinase C (PKC), Ca 2+ -calmodulin-dependent protein kinase II (CAMKII) and c-Jun N-terminal kinase (JNK). In contrast, phosphorylation of S845 is regulated by protein kinase A (PKA), extracellular signal-regulated kinase 1/2 (ERK1/2), and JNK (Banke et al., 2000;Wang et al., 2006;Ahn and Choe, 2009). In addition, various protein phosphatases (PPs) also regulate AMPA receptor functionality by dephosphorylating these sites (Snyder et al., 2003;Ahn and Choe, 2009). With respect to these previous reports, it is likely that perampanel may modulate GluA1 phosphorylation via regulating activities of protein kinases and phosphatases. Therefore, the objective in the present study is to investigate the intracellular mechanisms of perampanel involving GluA1 phosphorylation in normal and epileptic rats.

Experimental Animals, Chemicals and Experimental Design
In the present study, we used male Sprague-Dawley (SD) rats (7 weeks old) obtained from the Daehan Biolink, South Korea. Animals were given a commercial diet and water ad libitum under controlled conditions (22 ± 2 • C, 55 ± 5% and a 12:12 light/dark cycle with lights). Animal protocols were approved by the Institutional Animal Care and Use Committee of Hallym University (Chuncheon, South Korea). The number of animals used and their suffering were minimized in all cases. All reagents were obtained from Sigma-Aldrich (St. Louis, MO, United States), except as noted. Figure 1 illustrates the scheme of the experimental design of methodology and the animal numbers used in the present study (Figure 1).

SE Induction
Rats were given LiCl (127 mg/kg, i.p.) 24 h before the pilocarpine treatment. Animals were treated with pilocarpine (30 mg/kg, i.p.) 20 min after atropine methylbromide (5 mg/kg i.p.). Two hours after SE onset, diazepam (Valium; Hoffmannla Roche, Neuilly-sur-Seine, France; 10 mg/kg, i.p.) was administered to terminate SE and repeated, as needed. Control animals received saline in place of pilocarpine. Animals were video-monitored 8 h a day for selecting chronic epileptic rats showing spontaneous recurrent seizures (Ko and Kang, 2015). Behavioral seizure severity was evaluated according to Racine's scale: 1, immobility, eye closure, twitching of vibrissae, sniffing, facial clonus; 2, head nodding associated with more severe facial clonus; 3, clonus of one forelimb; 4, rearing, often accompanied by bilateral forelimb clonus; and 5, rearing with loss of balance and falling accompanied by generalized clonic seizures. We classified epileptic rats that showed behavioral seizure activities with seizure score ≥3 more than once.

Electrode Implantation
Control and epileptic rats were implanted with monopolar stainless steel electrodes (Plastics One, Roanoke, VA, United States) in the right hippocampus under Isoflurane anesthesia (3% induction, 1.5-2% for surgery, and 1.5% maintenance in a 65:35 mixture of N 2 O:O 2 ) using the following coordinates: −3.8 mm posterior; 2.0 mm lateral; −2.6 mm depth. Throughout surgery, core temperature of each rat was maintained 37-38 • C. Electrode was secured to the exposed skull with dental acrylic (Ko and Kang, 2015).

Perampanel Trials and Quantification of Seizure Activity
After baseline seizure activity was determined over 3 days, perampanel (8 mg/kg, i.p, Eisai Korea Inc.) or saline (vehicle) was daily administered at a certain time of the day (PM 6:00) over a 3 days or a 1 week period. EEG was recorded 2 h a day at the same time (Figure 1). EEG signals were recorded with a DAM 80 differential amplifier (0.1-3000 Hz bandpass; World Precision Instruments, Sarasota, FL, United States) and the data were digitized (1000 Hz) and analyzed using LabChart Pro v7 (ADInstruments, NSW, Australia). Behavioral seizure severity was also evaluated as aforementioned. After recording (18 h after the last treatment), animals were used for western blot study. Some animals (n = 4) were anesthetized with urethane (1.5 g/kg, i.p.) and then transcardially perfused with 4% paraformaldehyde (pH 7.4). The brains were removed and post-fixed overnight in the same solution then sequentially placed in 30% sucrose at 4 • C. Coronal sections were cut at a thickness of 30 µm on a cryostat, and used for Cresyl violet staining to further confirm epileptic animals.

Western Blot
After animals were sacrificed via decapitation, the left hippocampus was obtained. The hippocampal tissues were homogenized, and determined protein concentration using a Micro BCA Protein Assay Kit (Pierce Chemical, Rockford, IL, United States). Western blot was performed by the standard protocol. The primary antibodies used in the present study were listed in Supplementary Table 1. The bands were detected and quantified on ImageQuant LAS4000 system (GE Healthcare, Piscataway, NJ, United States). Since ERK1/2 and pERK1/2, but not others, antibodies clearly showed two (p42 and p44) bands and were changed to the same degree, we quantified both bands. As an internal reference, rabbit anti-β-actin primary antibody (1:5000) was used. The values of each sample were normalized with the corresponding amount of β-actin. The ratio of phosphoprotein to total protein was described as phosphorylation ratio.

Morris Water Maze
Spatial learning and memory were tested by the Morris water maze hidden platform task using the same maze and protocol as previous reported (Postina et al., 2004;Kim et al., 2016). Briefly, rats received five consecutive days of hidden platform training. The animals were allowed to search for the hidden platform for a period of 120 s. In the last day, animals were tested in a probe trial in which the platform was removed from the pool and allowed to search for a period of 120 s. Swimming time and path length in target quadrant, where the platform had been placed, were recorded.

Data Analysis
One-way ANOVA with post hoc Bonferroni's multiple comparison test (Biochemical data), Kruskal-Wallis test with Dunn's multiple comparison test (seizure frequency and seizure severity), two-way ANOVA with post hoc Bonferroni's multiple comparison test (probe trials of Morris water maze test), repeated measures two-way ANOVA with the least significant difference test (Morris water maze test), χ 2 test (comparison of seizure frequency between responder and non-responder) and Student's t-test (comparison of seizure duration between responder and non-responder) were used to determine statistical significance of data. A p-value below 0.05 was considered statistically significant.
To confirm this, we evaluated the effects of BIM (a PKC inhibitor) and KN-93 (a CAMKII inhibitor) on GluA1  ; Figures 5B,D]. Therefore, these findings also indicate that perampanel may regulate GluA1 expression and pGluA1-S831 ratio via CAMKII signaling pathway.

Effect of Perampanel on Spatial Memory in Normal and Epileptic Rats
To evaluate the cognitive effects of perampanel, we employed the Morris water maze test (Figure 1). Over 5 days of training, normal rats improved their ability to find the submerged platform, which exhibited decreasing escape time [F (1,48) = 6.652; p < 0.05 vs. epileptic rats, n = 5, respectively; Figures 12A,B]. In probe trials, time spent in goal quadrant in normal rats was higher than that in epileptic rats [F (1,8) = 16.853; p < 0.05 vs. epileptic rats, n = 5, respectively; Figures 12A,C]. The percentage of the time spent and the path length in goal quadrant were also higher than those in epileptic rats [F (1,8) = 21.14 and 14.136, respectively; p < 0.05 vs. epileptic rats, n = 5, respectively; Figures 12A,D,E]. Perampanel did not affect escape duration between both groups [F (1, 48) = 0.781; n = 5, respectively; acquisition, consolidation and retention of memory in normal and epileptic rats.

DISCUSSION
The major findings in the present study are that perampanel reduced GluA1 expression and regulated GluA1 phosphorylations by multiple signaling molecules in epileptic rats. Briefly, perampanel increased pCAMKII and pPKA ratios, which phosphorylate GluA1-S831 and -S845 site, respectively. Perampanel also enhanced pJNK and pPP2B ratios, which phosphorylates and dephosphorylates both GluA1-S831 and -S845 sits (Figure 13). Furthermore, these perampanel-induced changes did not lead to a detrimental impact on cognitive ability in normal and epileptic rats.
In the present study, perampanel effectively reduced seizure activities in epileptic rats, although some animals (30% in treated animals) were non-responders. Furthermore, perampanel reduced GluA1 expression in epileptic rats that showed the lower GluA1 expression than control animals. These findings indicate that the anti-epileptic effect of perampanel may be relevant to inhibition of GluA1 expression as well as blockaded of AMPA receptor-mediated currents as previously reported (Hanada et al., 2011;Ceolin et al., 2012;Krauss, 2013;Steinhoff et al., 2013;Chen et al., 2014). The effects of AMPA receptor antagonist on GluA1 expression have been controversial. NBQX, a competitive AMPA receptor antagonist, increases GluA1 expression (Thiagarajan et al., 2005) or not (Maeng et al., 2008). Although we could not explain the underlying mechanisms concerning perampanel-mediated down-regulation of GluA1 expression, it is plausible that the properties of perampanel as a noncompetitive (allosteric) AMPA receptor antagonist would result in the distinct effect on GluA1 expression as compared to NBQX. Furthermore, it is considerable that perampanel affected CAMKII and JNK phosphorylations (activities). CAMKII inhibition and gene deletion reduce GluA1 expression level (Ota et al., 2010;Hagihara et al., 2011). In contrast, JNK activation diminishes The quantitative analyses of the Morris water maze. Over 5 days of training, normal rats improve their ability to find the submerged platform, which exhibits decreasing escape time, while epileptic rats do not. Perampanel (PER) does not affect escape duration in normal and epileptic rats. * ,# p < 0.05 vs. first day and epileptic rats, respectively; n = 5, respectively. Error bars in graphs indicates SEM. (C-E) The quantitative analyses of the probe trials. Perampanel (PER) does not affect the time spent in goal quadrant (C) and the percentage of the time spent (D) and the path length (E) in goal quadrant in normal and epileptic rats. * ,# p < 0.05 vs. normal animals and non-goal quadrants, respectively; n = 5, respectively. Error bars in graphs indicates SEM.
GluA1 expression level (Lin and Lee, 2012). Indeed, the present data reveal that KN-93 (a CAMKII inhibitor) reduced GluA1 expression, while SP600125 (a JNK inhibitor) increased it in normal and epileptic rats. Therefore, it is likely that perampanel may reduce GluA1 expression in normal and epileptic rats through CAMKII and JNK signaling pathways. Further studies are needed to elucidate the pharmacological mechanisms of perampanel in the regulation of GluA1 expression.
perampanel could theoretically result in negative effects on cognition. However, recent clinical studies demonstrate that perampanel does not negatively affect cognition in epilepsy patients (Villanueva et al., 2016;Meschede et al., 2018;Piña-Garza et al., 2018). Similar to human studies, the present study demonstrates that perampanel did not influence spatial learning in normal and epileptic rats during Morris water maze test. Furthermore, perampanel enhanced pGluA-S831 and -S845 ratios, although it reduced GluA1 expression. Regulation of the phosphorylation of serine residues in GluA1 by kinases (PKA, PKC, CaMKII, ERK1/2, and JNK) and phosphatases (PP1, PP2A, and PP2B) plays an important role in governing the conductance and trafficking of AMPA receptor in and out of the synaptic membranes (Lee et al., 1998Shepherd and Huganir, 2007). In the present study, perampanel increased pCAMKII and pPKA ratios, which phosphorylate GluA1-S831 and -S845 site, respectively (Banke et al., 2000;Wang et al., 2006;Ahn and Choe, 2009). In addition, perampanel increased pJNK and pPP2B ratios (indicating JNK activation and PP2B inhibition, respectively), which phosphorylates and dephosphorylates both GluA1-S831 and -S845 sits, respectively Kim et al., 2009;Kam et al., 2010). Indeed, KN-93, H-89, and SP600125 diminished pGluA1-S831 and -S845 ratios, while CsA increased them. However, BIM and U0126 did not affect GluA1 phosphorylation ratios. With respect to the negative feedback-regulation of AMPA receptor activity in response to AMPA (Snyder et al., 2003), our findings suggest that perampanel may enhance GluA1 phosphorylations via regulating CAMKII, PKA, JNK and PP2B activities in epileptic rats, which may be one of adaptive responses for the diminished GluA1 expression or AMPA receptor-mediated currents. Thus, our findings also provide the possible underlying mechanisms answering why perampanel does not have a detrimental impact on cognitive ability, in spite of blockade of AMPA receptor.
In the present study, pGluA1-S831 ratio in epileptic rats was lower than that in normal animals. Furthermore, pPP1 ratio was lower than that in normal animals, while its expression level was similar to that in normal animals. These findings indicate that PP1 activity in the epileptic hippocampus may be higher than that in the normal one. Since PP1, but not PP2A and PP2B, dephosphorylates GluA1-S831 site (Huang et al., 2001), it is likely that PP1 activation would reduce GluA1-S831 phosphorylation ratio in epileptic animals. Indeed, okadaic acid (a PP1/PP2A inhibitor) increased pGluA1-S831 ratio in the present study. However, perampanel did not affect PP1 expression and pPP1 ratio in epileptic animals. Thus, our findings suggest that perampanel may regulate GluA1-S831 phosphorylation independent of PP1 activity in epileptic hippocampus.
Unlike PP1, the present study shows that PP2A and PP2B expressions and their phosphorylation ratios in epileptic animals were lower than those in normal animals. Similar to the case of pGluA1 ratios, it is likely that the reduced pPP2A and pPP2B ratios may be a compensatory response for maintenance of their activities against down-regulation of expressions. However, perampanel increased pGluA1 ratio in epileptic animals, concomitant with the elevated pPP2B, but not pPP2A, ratio, although okadaic acid and CsA increased pGluA1 ratios in epileptic animals. Therefore, our findings indicate that PP2B may be the potential pharmacological target of perampanel.
In the present study, pERK1/2 ratio in epileptic animals was lower than that in normal animals. Because both PP2A and PP2B deactivates ERK1/2 kinase activity through dephosphorylation of the threonine and tyrosine residues (Waskiewicz and Cooper, 1995;Gabryel et al., 2006;Min et al., 2017), it is likely that PP2A and PP2B activations would affect pGluA1 ratios in epileptic rats via reducing ERK1/2 activity. However, U0126 did not affect pGluA1 ratios, although perampanel elevated them epileptic animals. Therefore, these findings indicate that perampanelmediated ERK1/2 phosphorylation may not be relevant to GluA1 phosphorylation.
In the present study, there was no difference in the effects of inhibitors of kinases and PPs on GluA1 expressions and its phosphorylation ratios between normal and epileptic rats. Unexpectedly, the present data demonstrate that the responses to perampanel in normal animals were different from those in epileptic rats. Similar to the case of epileptic animals, perampanel reduced GluA1 expression in 3-days and 1-week treated normal animals. In 3-days treated groups, pGluA1-S831 and -S845 ratios were decreased, concomitant with reductions in pPKC, pCAMKII, pPKA, pJNK, pPP1, pPP2A, and pPP2B ratios. In 1week treated group, however, pGluA1-S831 ratio was rebounded more than vehicle-treated animals, although pPKC, pCAMKII, and pJNK ratios were reduced. Furthermore, pGluA1-S845 ratio was increased to vehicle level, accompanied by recovery of pPKA ratio. Perampanel also reduced pPP1, pPP2A, and pPP2B ratios in 1-week treated groups. Although we could not explain the exact mechanisms in the present study, it can be speculated that these phenomena may be adaptive responses to prolonged inhibition of AMPA receptor functionality in normal hippocampus. Conversely, the down-regulated GluA1 expression or the distinct neurochemical characteristics of epileptic hippocampus would lead to these discrepancies of perampanel actions between normal and epileptic animals. Further studies are needed to elucidate the pharmacological actions of perampanel in normal animals.
In conclusion, to the best of our knowledge, the present data demonstrate previously unreported pharmacological properties of perampanel concerning GluA1 phosphorylation and its upstream regulatory signaling pathways in normal and epileptic rats. Therefore, our findings suggest that perampanel may regulate AMPA receptor functionality via not only blockade of AMPA receptor but also modulation of GluA1 phosphorylations.

AUTHOR CONTRIBUTIONS
T-CK designed and supervised the project. All authors performed the experiments described in the manuscript and analyzed the data. J-EK and T-CK wrote the manuscript.

FUNDING
This work was supported by Hallym University (HRF-201806-014). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.