CDDO-Me Attenuates Clasmatodendrosis in CA1 Astrocyte by Inhibiting HSP25-AKT Mediated DRP1-S637 Phosphorylation in Chronic Epilepsy Rats

Clasmatodendrosis is one of the irreversible astroglial degeneration, which is involved in seizure duration and its progression in the epileptic hippocampus. Although sustained heat shock protein 25 (HSP25) induction leads to this autophagic astroglial death, dysregulation of mitochondrial dynamics (aberrant mitochondrial elongation) is also involved in the pathogenesis in clasmatodendrosis. However, the underlying molecular mechanisms of accumulation of elongated mitochondria in clasmatodendritic astrocytes are elusive. In the present study, we found that clasmatodendritic astrocytes showed up-regulations of HSP25 expression, AKT serine (S) 473 and dynamin-related protein 1 (DRP1) S637 phosphorylations in the hippocampus of chronic epilepsy rats. 2-Cyano-3,12-dioxo-oleana-1,9(11)-dien-28-oic acid methyl ester (CDDO-Me; bardoxolone methyl or RTA 402) abrogated abnormal mitochondrial elongation by reducing HSP25 upregulation, AKT S473- and DRP1 S637 phosphorylations. Furthermore, HSP25 siRNA and 3-chloroacetyl-indole (3CAI, an AKT inhibitor) abolished AKT-DRP1-mediated mitochondrial elongation and attenuated clasmatodendrosis in CA1 astrocytes. These findings indicate that HSP25-AKT-mediated DRP1 S637 hyper-phosphorylation may lead to aberrant mitochondrial elongation, which may result in autophagic astroglial degeneration. Therefore, our findings suggest that the dysregulation of HSP25-AKT-DRP1-mediated mitochondrial dynamics may play an important role in clasmatodendrosis, which would have implications for the development of novel therapies against various neurological diseases related to astroglial degeneration.


Introduction
Temporal lobe epilepsy (TLE) is one of the common neurological diseases, which is characterized by presence of spontaneous episodes of abnormal excessive neuronal discharges [1,2]. Neuronal loss including γ-aminobutyric acid (GABA)-ergic interneurons and synaptic rearrangement lead to seizure generation (ictogenesis) and the development of epilepsy (epileptogenesis). Together with the dysfunctions of neurons, aberrant astroglial functionality also contributes to pathogenesis of TLE, since astrocytes control the homeostasis of synaptic transmission and blood brain barrier (BBB), and glia-induced inflammation [2][3][4][5]. Indeed, astroglial dysfunctions such as disturbance of astrocyte gap junction coupling and K + buffering are involved in the etiology of TLE [6]. Although astrocytes are believed to be resistant to harmful stresses [7,8], astroglial degeneration is also induced by various pathological conditions [5,[9][10][11]. In particular, clasmatodendrosis (an irreversible autophagic astroglial death) has been reported by Alzheimer and Cajal more than 100 years ago [12]. Clasmatodendritic astrocytes show extensive swollen cell bodies with lysosome-derived vacuoles indicating ubiquitin proteasome system (UPS)-mediated autophagocytosis and disintegrated/beaded processes [12][13][14][15][16]. Furthermore, this astroglial degeneration may be involved in the synchronous epileptiform discharges and regulate seizure duration, but not seizure on-set or its severity, in chronic epilepsy rats [11,17].
Here, we demonstrate that CDDO-Me ameliorated clasmatodendrosis in CA1 astrocytes in the hippocampus of chronic epilepsy rats. In addition, CDDO-Me decreased accumulation of elongated mitochondria in CA1 astrocytes, concomitant with the HSP25 downregulation and the reduced DRP1 S637 and AKT S473 phosphorylation levels, which increased the DRP1-S616/S637 phosphorylation ratio. HSP25 knockdown showed the similar effect on clasmatodendritic CA1 astrocytes. 3-Chloroacetyl-indole (3CAI, an AKT inhibitor [43]) also mitigated clasmatodendrosis without altering prolonged HSP25 upregulation in CA1 astrocytes. Therefore, our findings suggest that sustained HSP25 induction may trigger the impaired mitochondrial fission in CA1 astrocytes during clasmatodendrosis by enhancing AKT-mediated DRP1 S637 phosphorylation, which was mitigated by CDDO-Me.

CDDO-Me Reduces AKT S473 Phosphoprylation and Mitochondrial Length in CA1 Astrocytes
The dysregulation of mitochondrial dynamics is one of the causes for clasmatodendrosis in CA1 astrocytes. Briefly, aberrant mitochondrial elongation evokes clasmatodendritic (autophagic) degeneration in CA1 astrocytes [28]. Considering the inhibitory effect of CDDO-Me on clasmatodendrosis in CA1 astrocytes [17] and the AKT-mediated regulation of mitochondrial dynamics [39][40][41], it is likely that CDDO-Me may rescue aberrant mitochondrial elongation (fusion) by inhibiting AKT activity. To confirm this, we evaluate the effect of CDDO-Me on AKT activity (phosphorylation). In chronic epilepsy rats, clasmatodendritic CA1 astrocytes showed the upregulated AKT S473 fluorescent intensity and the accumulation of elongated mitochondria (Figure 2A,B). The AKT S473 fluorescent intensity was 3.61-fold of control level (t (12) = 15.5, p < 0.001 vs. control animals, Student t-test, n = 7, respectively; Figure 2A,C).

Figure 9.
Effects of 3CAI on DRP1 S616-and S637 phosphorylations in CA1 astrocytes of control and epileptic rats. As compared to control animals (Cont), DRP1 S616 phosphorylation is reduced in CA1 astrocytes in epileptic rats, while DRP1 S637 phosphorylation is increased. 3CAI reduces the increased DRP1 S637 phosphorylation in CA1 astrocytes without affecting DRP1 S616 phosphorylation, as compared to vehicle (Veh).
Although HSP25 plays a protective role against harmful stress, prolonged HSP25 translation results in high-energy consumption and ER stress and finally triggers astroglial autophagy [18,20,22]. Compatible with a previous study [17], the present data demonstrate that CDDO-Me may ameliorate clasmatodendrosis by inhibiting dysregulation of HSP25-AKT signaling pathway. HSP25 modulates AKT enzyme activity, since HSP25 acts as a chaperone to retain AKT S473 phosphorylation by abrogating the pleckstrin homology domain and leucine-rich repeat protein phosphatase (PHLPP) 1-and 2-binding to AKT. Therefore, sustained HSP25 induction is sufficient for AKT-mediated astroglial autophagy, although HSP25 is not an indispensable factor for AKT activation [21,23,51,52]. Indeed, the present study shows that both CDDO-Me and HSP25 siRNA attenuated clasmatodendrosis, accompanied by the reduced AKT S473 phosphorylation. Therefore, our findings suggest that HSP25-AKT signaling pathway may play a key role in clasmatodendritic astroglial degeneration.
CDDO-Me prevents prolonged HSP25 induction by enhancing ERK1/2 activity that would also phosphorylate DRP1 S616 site [54,55]. Indeed, DRP1 activation through S616 phosphorylation is regulated by ERK/AKT [56]. In addition, amyloid-β (Aβ) sustains AKT activation that induces DRP1 S616 phosphorylation, and facilitates mitochondrial fission in neurons [57]. Ca 2+ influx induced by Aβ activates Ca 2+ /calmodulin-dependent protein kinase II (CaMKII)-AKT signaling pathway, which facilitates DRP1-mediated mitochondrial fragmentations and suppresses mammalian target of rapamycin (mTOR)-dependent autophagy in neurons [57]. However, clasmatodendrosis is mTOR-independent astroglial autophagy that is regulated by HSP25-mediated AKT activation [21,23]. Furthermore, the present data reveal that CDDO-Me did not affect DRP1 S616 phosphorylation, and that clasmatodendritic astrocytes showed the accumulation of elongated mitochondria and the AKT S473 hyper-phosphorylation. These findings indicate that CDDO-Me-induced ERK1/2 activation may not be involved in DRP1-mediated mitochondrial fission, but may inhibit the prolonged HSP25 induction in clasmatodendritic astrocytes. In addition, the increased AKT activity in clasmatodendritic astrocytes may elongate mitochondrial length by enhancing DRP1 phosphorylation at S637 rather than S616 site. Therefore, it is plausible that the distinct signaling pathways in response to the disparate stimuli would cause the different DRP1 regulations.
The dysregulation of mitochondrial fission (aberrant mitochondrial elongation) leads to oxidative stress and further elevates reactive oxygen species, which triggers autophagic cell death [58][59][60][61]. In the present study, CDDO-Me abrogated the aberrant mitochondrial elongation in clasmatodendritic CA1 astrocytes by increasing DRP1 S616/S637 phosphorylation ratio. CDDO-Me is an activator of nuclear factor-erythroid 2-related factor 2 (Nrf2, a redox-sensitive transcription factor) that maintains redox homeostasis by regulating antioxidant-response element (ARE)-dependent transcription and the expression of antioxidant defense enzymes including heme oxygenase-1 (HO-1), which protect neurons and astrocytes from various harmful stresses [17,38,55,62,63]. Therefore, it is presumable that the antioxidant effect of CDDO-Me would restore the abnormal mitochondrial elongation, and subsequently inhibit clasmatodendrosis in CA1 astrocytes. However, the present study shows that CDDO-Me inhibited AKT that promotes HO-1 gene expression in rat astrocytes [64]. Furthermore, AKT inactivation by 3CAI abrogated accumulation of elongated mitochondria by reducing DRP1 S637 phosphorylation. Given the deterioration of autophagic astroglial death induced by Mdivi-1 [28], therefore, our findings indicate that CDDO-Me may attenuate clasmatodendrosis by direct regulation of AKT-mediated DRP1 phosphorylation as well as its antioxidant properties.
Astrocytes play an important role in delays clearance of K + and glutamate from extracellular space [11,65,66]. Therefore, astroglial dysfunction or death is involved in spontaneous seizure activity and epileptogenesis [5,11,17,67,68]. Clasmatodendrosis is a consequence of spontaneous recurrent seizures due to over-activation of the temporoammonic path, which is involved in the duration and propagation of synchronous discharges (but not its frequency and severity) in the epileptic hippocampus [11,17]. Furthermore, conventional anti-epileptic drugs prevent clasmatodendrosis in chronic epilepsy rats [11]. Although CDDO-Me and 3CAI have not anti-epileptic properties [54,69], furthermore, CDDO-Me attenuates reduces seizure duration and its progression, accompanied by abrogating clasmatodendritic degeneration [17], and 3CAI improves the efficacies of α-Amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor (AMPAR) antagonists on spontaneous seizure activities [43]. Considering that α-aminoadipic acid (an astroglial toxin) and 4-aminopyridine (a K + channel blocker) synchronizes reverberant epileptiform discharges [11,17,67], clasmatodendrosis may be one of considering factors leading to prolonged seizure activity and its propagation in the epileptic hippocampus through impairments of inwardly K + channel as well as GJ, although it may not be a primary cause of ictogenesis. Therefore, the prevention of clasmatodendrosis may attenuate the duration and propagation of synchronous discharges in the epileptic hippocampus by exerting clearance of K + and glutamate from extracellular space in ictal stage.
On the other hand, inflammation-related disturbance and dysregulation of astroglial gap junction connexin 43 (Cx43) contribute to the seizure generation, because the uncoupling of Cx43 results in gliotransmitter release and the accumulation of K + and glutamate in the extracellular space [75,76]. Interestingly, carbenoxolone (a gap junction blocker) attenuated astroglial swelling, rupture of astroglial nuclear membrane and vacuolization of the astroglial cytoplasm at post-SE 60 days [76]. Therefore, it is possible that CDDO-Me would ameliorate clasmatodendrosis by affecting Cx43 functionality. Further studies are needed to elucidate the CDDO-Me-induced Cx43 regulation.

Experimental Animals and Chemicals
Adult male Sprague-Dawley (SD) rats (7 weeks old) were used in the present study. Animals were kept under controlled environmental conditions (23-25 • C, 12 h light/dark cycle) to access freely to water and food throughout the experiments. All experimental protocols described below were approved by the Institutional Animal Care and Use Committee of Hallym University (Chuncheon, South Korea, Code number: Hallym 2018-3, approval date: 30 April 2018). All reagents were obtained from Sigma-Aldrich (St. Louis, MO, USA), except as noted.

Tissue Processing and Immunohistochemistry
Seven days after infusion (8 weeks after SE), animals were transcardially perfused with 4% paraformaldehyde under urethane anesthesia (1.5 g/kg i.p.), and after additional fixation for overnight at 4 • C. The brains were rinsed in PB containing 30% sucrose at 4 • C for 2 days. Thereafter, coronal sections (30 µm) were cut with a cryostat. Age-matched control (normal) animals were also perfused by the same method. Then sections were incubated in 0.1% bovine serum albumin and subsequently primary antibody (Table 1). Tissue sections visualized with appropriate Cy2-and Cy3-conjugated secondary antibodies. Immunofluorescence was observed using an Axio Scope microscope (Carl Zeiss Korea, Seoul, South Korea). Negative control test was performed with normal rabbit serum (#31883, Ther-moFisher Korea, Seoul, South Korea), mouse IgG1 isotype control (#02-6100, ThermoFisher Korea, Seoul, South Korea), and mouse IgG2a isotype control (#02-6200, ThermoFisher Korea, Seoul, South Korea), instead of the primary antibodies. No immunoreactivity was observed for the negative control in any structures [17,22,43].

Western Blots
Animals were sacrificed via decapitation. The brains were quickly removed and coronally cut to 1 mm thickness using rodent brain matrix (World Precision Instruments, Sarasota, FL, USA) on ice. Thereafter, the stratum radiatum of the CA1 region of the dorsal hippocampus were dissected out in cold artificial cerebrospinal fluid (4 • C) under stere-omicroscope. The tissues were were homogenized and protein concentration determined using a Micro BCA Protein Assay Kit (Pierce Chemical, Rockford, IL, USA). Following electrophoresis, proteins were transferred to nitrocellulose membranes that were blocked overnight with 2% bovine serum albumin in Tris-buffered saline (in mM 10 Tris, 150 NaCl, pH 7.5, and 0.05% Tween 20) and then incubated overnight at 4 • C in blocking solution containing primary antibodies (Table 1). After washing, membranes were incubated for 1 h at room temperature in a solution containing horseradish peroxidase-conjugated secondary antibodies. A chemiluminescence signal was detected by luminol substrate reaction (ECL Western Blotting System, GE Healthcare Korea, Seoul, South Korea). The bands were detected and quantified on an ImageQuant LAS4000 system (GE Healthcare Korea, Seoul, South Korea). 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 level [17,22,43].

Cell Count, Measurement of Fluorescent Intensity and Mitochondrial Morphometry
The hippocampal tissues were captured (10 sections per each animal), and areas of interest (1 × 10 5 µm 2 ) were selected from the striatum radiatum of the CA1 region. Thereafter, clasmatodendritic astrocytes were counted on 20× images using AxioVision Rel. 4.8 Software. In addition, five brain sections from each animal were randomly selected at different rostro-caudal hippocampal levels. One randomly selected CA1 astrocytes (naïve astrocytes in control animals, and clasmatodendritic and reactive astrocytes in epileptic animals) from each slice (total 35 cells in each group, respectively) were used for quantification of mitochondrial morphometry using ImageJ software. Mitochondria were analyzed for perimeter and area. Mitochondrial parameters were calculated as followed: Area-weighted form factor = perimeter 2 /4π (an indicative of mitochondrial elongation); Form factor = perimeter 2 /4π × area (indicating the transition from punctiform to elongated, complex shaped, but still isolated mitochondria); Cumulative area:perimeter ratio = Σarea/Σperimeter (indicating the transition from elongated, isolated mitochondria to a reticular network or aggregation of interconnected mitochondria) [44][45][46]. For measurement of fluorescent intensity, 30 areas/rat (400 µm 2 /area) were randomly selected within the stratum radiatum of CA1 region (15 sections from each animal, n = 7 in each group). Mean intensity was measured using AxioVision Rel. 4.8 software (Carl Zeiss Korea, Seoul, South Korea). Fluorescent intensity was normalized by setting the mean background [44][45][46]. The investigators were blinded to experimental groups in performing cell counts and morphological analysis.

Data Analysis
Comparisons of data among groups were performed using Student t-test or oneway ANOVA followed by Bonferroni's post hoc comparisons after evaluating the values on normality using Shapiro-Wilk W-test. A p-value less than 0.05 was considered to be significant.

Conclusions
The present data demonstrate for the first time that HSP25-mediated AKT activation impaired mitochondrial fission by DRP1 S637 hyper-phosphorylation and led to autophagic astroglial degeneration, which was abrogated by CDDO-Me, HSP25 siRNA and 3CAI. These new findings may have implications for the development of novel therapies against various neurological diseases by regulating astroglial degeneration and mitochondrial dynamics.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.

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.