CDDO-Me Attenuates Astroglial Autophagy via Nrf2-, ERK1/2-SP1- and Src-CK2-PTEN-PI3K/AKT-Mediated Signaling Pathways in the Hippocampus of Chronic Epilepsy Rats

Clasmatodendrosis is an autophagic astroglial death showing extensive swollen cell bodies with vacuoles and disintegrated/beaded processes. This astroglial degeneration is closely relevant to the synchronous epileptiform discharges. However, the underlying molecular mechanisms and the roles of clasmatodendrosis in spontaneous seizure activity are still unknown. The 2-cyano-3,12-dioxo-oleana-1,9(11)-dien-28-oic acid methyl ester (CDDO-Me; RTA 402) is one of the activators for nuclear factor-erythroid 2-related factor 2 (Nrf2) that is a redox-sensitive transcription factor. In the present study, we explored the effects of CDDO-Me on clasmatodendrosis in chronic epilepsy rats, which could prevent epilepsy-related complications. In the present study, clasmatodendritic astrocytes showed reduced Nrf2 expression and its nuclear accumulation, which were restored by CDDO-Me. CDDO-Me also abrogated heat shock protein 25 (HSP25) upregulation in clasmatodendritic astrocytes by regulating extracellular signal-related kinases 1/2 (ERK1/2)-specificity protein 1 (SP1)- and Src-casein kinase 2 (CK2)-phosphatase and tensin homolog deleted on chromosome 10 (PTEN)-phosphatidylinositol-3-kinase (PI3K)-AKT-glycogen synthase kinase 3β (GSK3β)-bax-interacting factor 1 (Bif-1)-mediated signaling pathways in chronic epilepsy rats. In addition, CDDO-Me ameliorated spontaneous seizure duration, but not seizure frequency and behavioral seizure severity. Therefore, our findings suggest that clasmatodendrosis may affect seizure duration in chronic epilepsy rats, and that CDDO-Me may attenuate autophagic astroglial degeneration by regulating various signaling pathways.


Introduction
Epilepsy is one of the most common chronic neurological diseases. The mean prevalence rate of active epilepsy is 8% worldwide. The etiology of epilepsy is unknown (idiopathic) or related to disease states including brain tumors and traumatic injury [1,2]. The main symptom of epilepsy is the presence of spontaneous episodes of abnormal excessive neuronal discharges. This seizure activity results in neuronal loss in the various brain regions, especially in the hippocampus (hippocampal sclerosis) [3,4]. Similar to other brain injuries, neuronal loss and synaptic rearrangement induce astroglial activation (reactive astrogliosis), which may contribute to epileptogenesis [5][6][7].
Astrocytes are key players in the regulation of extracellular glutamate concentration, ion homeostasis and neuronal functionality, and are believed to be resistant to harmful stresses [8,9]. More than 100 years ago, however, Alzheimer reported irreversible astroglial injury characterized by extensive swollen cell bodies with vacuoles and disintegrated/beaded processes, and Cajal termed it as "clasmatodendrosis" [10]. In addition,

Electrode Implantation, CDDO-Me Trials and Quantification of Seizure Activity
Control and epileptic rats were implanted with a monopolar stainless steel electrode (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. Animals were also implanted with a brain infusion kit 1 and an Alzet 1007D osmotic pump (Alzet, Cupertino, CA, USA) to infuse with vehicle or CDDO-Me (10 µM) into the right lateral ventricle [35,36]. The correct location of infusion needle into the ventricle was confirmed during brain sections and when sampling tissues for Western blot. Electrode and infusion needle were secured to the exposed skull with dental acrylic. Three days after surgery, an electroencephalogram (EEG) was recorded 2 h a day at the same time over 4 days [37][38][39]. Behavioral seizure severity was also evaluated as aforementioned [37,38]. After recording, animals were used for Western blot and immunofluorescent study.

Western Blots
Animals were sacrificed via decapitation. The brains were quickly removed and coronally cut to 1 mm thickness (approximately 3-4 mm posterior to the bregma) using rodent brain matrix (World Precision Instruments, Sarasota, FL, United States) on ice. Thereafter, the stratum radiatum of the CA1 region of the dorsal hippocampus were rapidly dissected out in cold (4 • C) artificial cerebrospinal fluid under stereomicroscope [14]. The CA1 tissues were homogenized and protein concentration determined using a Micro BCA Protein Assay Kit (Pierce Chemical, Rockford, IL, United States). Western blot was performed by the standard protocol as follows: following electrophoresis, proteins were transferred to nitrocellulose membranes. Membranes were incubated overnight at 4 • C with 2% bovine serum albumin (BSA) in Tris-buffered saline (TBS; in 10 mM Tris, 150 NaCl, pH 7.5 and 0.05% Tween 20) and then in primary antibodies (Supplementary  Table S1). Subsequently, the membranes were incubated for 1 h at room temperature in a solution containing horseradish peroxidase (HRP)-conjugated secondary antibodies. A chemiluminescence signal was detected by luminol substrate reaction (ECL Western Blotting System, GE Healthcare Korea, Seoul, Korea). The values of each sample were calculated with the corresponding amount of β-actin. The ratio of phosphoprotein to total protein was described as phosphorylation level [25,[37][38][39].

Immunohistochemistry, Cell Counts and Measurement of Fluorescent Intensity
Under urethane anesthesia (1.5 g/kg, i.p.), animals were transcardially and subsequently perfused with 0.9% saline followed by 4% paraformaldehyde in 0.1 M phosphate buffer (PB, pH 7.4). The brains were isolated and post-fixed in the same fixative overnight, and then stored in 30% sucrose/0.1 M PBS. Coronal sections were sliced at a 30-µm thickness with a freezing microtome. Then, sections were reacted with in 0.1% bovine serum albumin and successively primary antibody (Supplementary Table S1). After washing, sections were further incubated in appropriate Cy2-and Cy3-conjugated secondary antibodies. Immunofluorescence was observed using an AxioScope microscope (Carl Zeiss Korea, Seoul, Korea). To establish the specificity of the immunostaining, a negative control test was carried out with normal mouse serum (#31880, ThermoFisher Korea, Seoul, Korea), normal rabbit serum (#31883, ThermoFisher Korea, Seoul, Korea), mouse IgG1 isotype control (#02-6100, ThermoFisher Korea, Seoul, Korea) and mouse IgG2a isotype control (#02-6200, ThermoFisher Korea, Seoul, Korea), instead of the primary antibodies. No immunoreactivity was observed for the negative control in any structures ( Figure S1) [14,15]. All experimental procedures in this study were performed under the same conditions and in parallel. For cell counts, sections (10 sections per each animal) were captured and areas of interest (1 × 10 4 µm 2 ) were selected from the CA1 region using an AxioImage M2 microscope. Thereafter, cell counts were performed using AxioVision Rel. 4.8 Software. For measurement of fluorescent intensity, 30 areas/rat (400 µm 2 /area) were randomly selected within the stratum radiatum of the CA1 region (15 sections from each animal, n = 7 in each group). After capture, green or red channel was converted to a grayscale image, and mean intensity was measured using AxioVision Rel. 4.8 software (Carl Zeiss Korea, Seoul, Korea). Fluorescent intensity was normalized by setting the mean background. In addition, colocalization of Nrf2 with 4 ,6-diamidino-2-phenylindole (DAPI) was analyzed for nuclear Nrf2 intensity. Thereafter, the ratio of nuclear:cytosolic Nrf2 intensity was calculated. Cell counts and measurement of fluorescent intensities were performed by two different investigators who were blind to the classification of tissues [25,[37][38][39].

Data Analysis
Data were analyzed using Student t-test or one-way analysis of variance (ANOVA) followed by Bonferroni's post hoc comparisons after evaluating the values on normality using Shapiro-Wilk W-test. Mann-Whitney U-test was also used to determine statistical significance of data. A p-value less than 0.05 was considered to be significant [25,[37][38][39].
Antioxidants 2021, 10, x FOR PEER REVIEW 5 of 20 p < 0.00001, one-way ANOVA, n = 7, respectively; Figure 1A,B). Thus, CDDO-Me also increased the ratio of nuclear:cytosolic intensity in CA1 astrocytes (F(2,18) = 19.3, p = 0.00003, one-way ANOVA, n = 7, respectively; Figure 1A,D). These findings indicate that the reduced Nrf2 level may be involved in autophagic CA1 astroglial degeneration in epileptic animals. Figure 1. Effects of CDDO-Me on nuclear factor-erythroid 2-related factor 2 (Nrf2) expression and its nuclear accumulation in CA1 astrocytes of control and epileptic rats. As compared to control animals (Cont), Nrf2 expression and its nuclear accumulation were reduced in CA1 astrocytes in epileptic rats. CDDO-Me increased Nrf2 expression and its nuclear accumulation in CA1 astrocytes, as compared to vehicle (Veh). (A) Representative photos demonstrating astroglial Nrf2 expression in CA1 astrocytes. Arrows

CDDO-Me Inhibits Src Family Phosphorylation at Y416 Site in the Epileptic Hippocampus
As aforementioned, CK2 increased its catalytic activity by Src family protein tyrosine kinases-mediated phosphorylation at the Y255 residue [37,46]. The Src family of protein tyrosine kinase activities are regulated by tyrosine phosphorylation at two sites, but with opposing effects. Autophosphorylation of tyrosine (Y) 396 (equivalent to Y416 of Src), located in the catalytic domain, upregulates Src kinase activity. Y507 phosphorylation (equivalent to Y527 of Src) inactivates the kinase activity, while dephosphorylation of this site is not sufficient for full kinase activation [49,50]. Interestingly, CDDO-Me affects Src and AKT activation [51,52]. Thus, we evaluated the effect of CDDO-Me on Src family activities. As compared to the control hippocampus, Src family-Y416 phosphorylation was reduced in the epileptic hippocampus without changing Src family expression (t(12) = 19.4, p < 0.00001 vs. control animals, Student t-test, n = 7, respectively; Figure  6A-C, Figure S5). p-Src-Y416 family ratio was also decreased in the epileptic hippocam-

CDDO-Me Decreases Seizure Duration, But Not Seizure Frequency or Its Intensity
Since conventional anti-epileptic drugs protect CA1 astrocytes from clasmatodendritic degeneration [16], it seems that clasmatodendrosis is relevant to spontaneous seizure activity. Therefore, we evaluated the effect of CDDO-Me on spontaneous seizure activity in epileptic rats over a 4-day period. Under basal (vehicle-treated) conditions, the total seizure frequency was 7.9 ± 1.3/recording session and the total seizure duration was 874.4 ± 227.6 s. The seizure severity (behavioral seizure core) was 3.7 ± 0.5 ( Figure 7A-D). CDDO-Me did not affect total seizure frequency (7.6 ± 2.5; Figure 7B). However, the total seizure duration was reduced to 499.6 ± 178.7 s (t (6) = 3.27, p = 0.02 vs. vehicle, Student t-test, n = 7, respectively; Figure 7C). The seizure severity (3.4 ± 0.7) was unaffected by CDDO-Me ( Figure 7D). These findings indicate that CDDO-ME may decrease seizure duration, but not its frequency and severity, in chronic epileptic rats. activity in epileptic rats over a 4-day period. Under basal (vehicle-treated) conditions, the total seizure frequency was 7.9 ± 1.3/recording session and the total seizure duration was 874.4 ± 227.6 s. The seizure severity (behavioral seizure core) was 3.7 ± 0.5 ( Figure  7A-D). CDDO-Me did not affect total seizure frequency (7.6 ± 2.5; Figure 7B). However, the total seizure duration was reduced to 499.6 ± 178.7 s (t(6) = 3.27, p = 0.02 vs. vehicle, Student t-test, n = 7, respectively; Figure 7C). The seizure severity (3.4 ± 0.7) was unaffected by CDDO-Me ( Figure 7D). These findings indicate that CDDO-ME may decrease seizure duration, but not its frequency and severity, in chronic epileptic rats.

Discussion
The major findings in the present study were that CDDO-Me attenuated HSP25induced clasmatodendrosis through Nrf2-, ERK1/2-SP1-and Src-CK2-PTEN-PI3K-AKT-GSK3β-Bif-1-mediated signaling pathways in chronic epilepsy rats. In addition, CDDO-Me ameliorated spontaneous seizure duration, but not seizure frequency and behavioral seizure severity (Figure 8). Clasmatodendrosis was first described by Alzheimer and was postulated to reflect irreversible injury of astrocytes [10]. Clasmatodendrosis is a pathological substrate, linked to white matter hyperintensities, as seen on brain T2-weighted magnetic resonance imaging associated with stroke, Alzheimer's disease and vascular dementia [53,54]. In addition, influenza-associated encephalopathy [55], traumatic brain injury [56], methamphetamine abuse [57], neuromyelitis optica [58], explosive blasts [59] and osmotic demyelination syndrome [19] induce clasmatodendrosis in the brain gray matter as well as the white matter. Although recent reports have demonstrated that clasmatodendrosis is a corollary of senescence, autophagy, metabolic dysfunction, endoplasmic Clasmatodendrosis was first described by Alzheimer and was postulated to reflect irreversible injury of astrocytes [10]. Clasmatodendrosis is a pathological substrate, linked to white matter hyperintensities, as seen on brain T2-weighted magnetic resonance imaging associated with stroke, Alzheimer's disease and vascular dementia [53,54]. In addition, influenza-associated encephalopathy [55], traumatic brain injury [56], methamphetamine abuse [57], neuromyelitis optica [58], explosive blasts [59] and osmotic demyelination syndrome [19] induce clasmatodendrosis in the brain gray matter as well as the white matter. Although recent reports have demonstrated that clasmatodendrosis is a corollary of senescence, autophagy, metabolic dysfunction, endoplasmic reticulum (ER) stress or nuclear factor-κB (NFκB) activation [12,18,60,61], the role or the underlying mechanism of this pathological change are not well-defined.
In the present study, clasmatodendritic CA1 astrocytes showed a reduced Nrf2 level in chronic epilepsy rats, and CDDO-Me (an Nrf2 activator) attenuated this astroglial degeneration. Clasmatodendrosis is initiated by acidosis (~pH 5) and energy failure induced by mitochondrial inhibition [62], which are induced by oxidative stress [63][64][65]. This is because ROS reduce glycolysis, leading to intracellular acidosis by the inhibition of Na + -H + exchangers and Na + -HCO 3 − cotransporters [63], which subsequently impairs the enzymatic steps of glutathione (GSH, an endogenous antioxidant) synthesis [65]. Considering the Nrf2-mediated regulation of γ-glutamyl cysteine ligase and of cystine/glutamate transporters (xCT or SLC7a11), which facilities the up-take of cystine (a GSH precursor) [66], our findings indicate that the dysfunction of Nrf2-mediated antioxidant system may be involved in the initiation of clasmatodendrosis.
On the other hand, HSP25 binds and forms a new complex with AKT, securing AKT's natural conformation and enzymatic activity [80]. In addition, HSP25 increases AKT-S473 phosphorylation, which exerts a mechanistic target of rapamycin (mTOR)independent astroglial autophagy by GSK3β-mediated Bif-1 accumulation [25,81]. Indeed, HSP25 siRNA attenuates clasmatodendrosis by inhibiting the AKT-GSK3β-Bif-1 signaling pathway [24,25,82]. In addition, HSP25 over-expression sustains AKT-S473 phosphorylation by inhibiting the pleckstrin homology domain and leucine-rich repeat protein phosphatase (PHLPP) 1 and 2-binding to AKT (Lee and Kim, 2020). Since HSP25-AKT interaction is not required to promote AKT activation and HSP25 does not interact with PI3K [83,84], it is likely that HSP25 may function as a chaperone to maintain AKT activity rather than an indispensable factor for AKT activation. In the present study, CDDO-Me reduced p-HSP25, p-AKT and Bif-1 levels in the epileptic hippocampus. Regarding the aforementioned role of HSP25 in AKT phosphorylation [82][83][84][85][86], these findings suggest that CDDO-Me-induced HSP25 downregulation may prevent AKT hyper-phosphorylation by diminishing the HSP25 function as a chaperone for AKT.
In the canonical pathway, AKT is dephosphorylated by PTEN [87], which is downregulated by seizure activity [41,42]. Mutation or inactivation of PTEN contributes to seizures in human patients and animal models [88][89][90]. Indeed, PTEN mRNA and its protein expression are downregulated in the rat hippocampus following pentylenetetrazol or kainic acid injection [41,42]. We have also reported that PTEN expression level is lower in the rat epileptic hippocampus than that in the normal hippocampus [39]. In the present study, p-PTEN ratio was similarly observed between the control and the epileptic hippocampus, although PTEN phosphorylation level was lower in the epileptic hippocampus than the normal hippocampus. In addition, CDDO-Me decreased PTEN phosphorylation only in the epileptic hippocampus. Considering that PTEN phosphorylation represents its inactivation [43], these findings indicate that CDDO-Me may increase PTEN activity, and that CDDO-Me may mitigate clasmatodendrosis by facilitating PTEN-mediated AKT inhibition in the epileptic hippocampus.
On the other hand, Src family-Y416 phosphorylation level is significantly decreased in human symptomatic epileptic tissues, as compared to control tissues [91]. Consistent with this previous report, the present study reveals the reduced Src family-Y416 phosphorylation level in the epileptic hippocampus, accompanied by increased AKT phosphorylation. Furthermore, CDDO-Me diminished phosphorylation of Src family-Y416 and CK2-Y255, which are signaling molecules acting as PTEN kinases [44][45][46][47]. Thus, our findings indicate that CDDO-Me may increase PTEN activity by inhibiting Src and CK2 phosphorylation. ERK1/2 also phosphorylates CK2 primarily at T360/S362, subsequently enhancing CK2 activity. Indeed, the level of catenin (a substrate of CK2) phosphorylation correlates with levels of ERK1/2 activity in human glioblastoma [47]. Therefore, it is presumable that CDDO-Me-induced ERK1/2 activation is involved in CK2 activation. However, the present data demonstrate that CK2 T360/S362 phosphorylations were unaltered in the epileptic hippocampus, and unaffected by CDDO-Me. Therefore, it is likely that CK2-Y255 phosphorylation may be involved in PTEN phosphorylation. With respect to CDDO-Me-induced Src and AKT inactivations [51,52], thus, our findings suggest that CDDO-Me may mitigate clasmatodendrosis by regulating the Src-CK2-PTEN-AKT-GSK3β-Bif-1 signaling pathway.
In a previous study, we reported that clasmatodendritic changes might be a consequence of prolonged recurrent seizures induced by hyperexcitability of the temporoammonic path, and not a cause of epileptogenesis. This is because anti-epileptic drugs (valproate, carbamazepine and vigabatrin) prevent clasmatodendrosis in the epileptic hippocampus [16]. However, it could not be excluded the possibility that clasmatodendrosis could affect ictogenesis in the epileptic hippocampus. In the present study, we found that CDDO-Me diminished seizure duration, but not its frequency and severity, in chronic epileptic rats. Astrocytes play an important role in K + buffering [92]. In addition, αaminoadipic acid (an astroglial toxin) and 4-aminopyridine (a K + channel blocker) synchronize reverberant epileptiform discharges [16,93]. Thus, our findings indicate that clasmatodendrosis may influence seizure duration in the epileptic hippocampus. Indeed, 4,5,6,7tetrabromotriazole (TBB), a CK2 inhibitor, prevents acute epileptiform discharges [94]. Furthermore, we recently reported that α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor (AMPAR) antagonists decreased CK2 Y255 phosphorylation, but not its T360/S362 phosphorylation levels in chronic epilepsy rats, accompanied by reduced seizure activity [37]. However, they reduced Src-Y416 phosphorylation level, but increased Src-Y527 phosphorylation level [37], unlike CDDO-Me in the present study. Although the direct experimental evidence for a role of clasmatodendrosis in seizure activity is currently limited, it is plausible that clasmatodendrosis may be considered an epiphenomenon affecting seizure duration rather than a primary ictogenic factor in the epileptic hippocampus. In clasmatodendritic astrocytes, aquaporin-4 (AQP4; a water channel) expression is decreased and aggregated in dense peripheral cellular deposits at the edge of rounded or swollen astrocytes in the human brain [54,55]. Similarly, AQP4 expression is negligible in clasmatodendritic CA1 astrocytes in the hippocampus of chronic epileptic rats [17]. Interestingly, astrocytes lacking AQP4 increase seizure duration, since AQP4 deletion delays clearance of K + from extracellular space [95]. With respect to these previous reports, it is likely that clasmatodendritic astrocytes may lead to the slow clearance of K + and water from extracellular space in the ictal stage, which could be involved in the duration and propagation of synchronous discharges in the epileptic hippocampus. Further studies are needed to elucidate the underlying mechanisms concerning the role of clasmatodendrosis in ictogenesis.

Conclusions
In the present study, we demonstrated, for the first time, that CDDO-Me ameliorated HSP25-induced astroglial autophagy via Nrf2-, ERK1/2-SP1-and Src-CK2-PTEN-PI3K-AKT-GSK3β-Bif-1-mediated signaling pathways in chronic epilepsy rats. In addition, CDDO-Me shortened seizure duration. Therefore, our findings suggest that autophagic astroglial degeneration may play an important role in the maintenance of spontaneous seizure activity, which could be one of the therapeutic targets for TLE medications.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/antiox10050655/s1, Table S1: Primary antibodies used in the present study, Figure S1: Representative photos of negative control test, Figure S2: Representative full-gel images of Western blots in Figures 1E and 3A, Figure S3: Representative full-gel images of Western blots in Figure 4A, Figure S4: Representative full-gel images of Western blots in Figure 5A and Figure S5: Representative full-gel images of Western blots in Figure 6A.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.