Mitochondrial fission as a target for supressing aberrant neuroplasticity and degeneration in the hippocampus

Full Text

Introduction

The balance of mitochondrial fusion and fission is crucial for maintaining cellular viability. Impairments in mitochondrial dynamics have been described in various neurological disorders [1], making the study of molecular mechanisms regulating mitochondrial dynamics in normal and abnormal conditions a promising direction for developing modern pharmacological agents. Modulation of mitochondrial dynamics, in particular, represents a potential therapeutic approach for mesial temporal lobe epilepsy with hippocampal sclerosis, which is considered the most common cause of drug-resistant epilepsy [2]. Effective pharmacological treatment strategies for this condition are currently lacking.

A well-characterized experimental model replicating the abnormalities observed in mesial temporal epilepsy with hippocampal sclerosis is the administration of kainic acid (KA) — an agonist of ionotropic kainate receptors — to animals. KA administration induces status epilepticus, excitotoxic damage to hippocampal neurons, and promotes spontaneous seizure activity [3, 4].

KA-induced excitotoxic neuronal death is mediated by Ca²⁺ overload, oxidative stress, and consequent mitochondrial dysfunction. Damage to mitochondrial membranes facilitates the release of apoptosis-inducing factors from mitochondria, leading to apoptotic cell death [3].

Neurons in the CA3 region (cornu ammonis) of the hippocampus are most vulnerable to KA, which is associated with their high expression of GluK4 and GluK5 kainate receptor subunits that exhibit maximal affinity for KA [3]. Furthermore, high-affinity GluK2 receptors are localized at the postsynaptic terminals of mossy fibers in the CA3. Their stimulation leads to rapid activation of detonator synapses via recurrent glutamatergic collaterals, promoting neuronal synchronization and induction of epileptiform activity [4]. KA administration also induces pro-inflammatory glial activation, causing disruptions in glutamate metabolism, water-ion homeostasis, and reorganization of gliovascular interactions [5].

KA-induced damage to CA1 and CA3 neurons triggers secondary changes in the hippocampal dentate gyrus (DG): reduced hilar neuron density, dispersion and ectopia of granular layer neurons, along with reorganization of neuronal processes through aberrant basal dendrite formation and axonal sprouting into the inner molecular layer. The latter is considered a pathological form of neuroplasticity leading to multiple excitatory circuits in the hippocampal DG, which contributes to epileptogenic focus development [4, 6, 7]. Thus, the KA-induced model reproduces the pathomorphological changes observed in hippocampal sclerosis and mesial temporal lobe epilepsy, making it valuable for developing neuroprotective strategies.

Mitochondrial dysfunction and dysregulation of mitochondrial dynamics are associated with both neuronal damage during status epilepticus and other pathogenetic aspects of epileptogenesis [1, 8]. Mitochondrial fission is a staged energy-dependent process requiring conformational changes in Drp1 (dynamin-related protein) — a cytoplasmic GTPase of the dynamin family [9]. Mitochondrial fusion requires coordinated activity of mitofusins (Mfn1, Mfn2) and OPA1 protein, which mediate the merging of outer and inner mitochondrial membranes [10]. Fusion proteins also participate in mitochondrial bioenergetics and degradation of dysfunctional organelles. Additionally, Mfn2 interacts with endoplasmic reticulum membranes (mitochondria-associated membranes, MAMs), influencing Ca²⁺ homeostasis and reactive oxygen species production [10].

The Drp1 protein facilitates mitochondrial fission and participates in quality control of these organelles and regulated apoptosis [9]. The choice of Drp1-dependent division mechanism depends on Drp1 interaction with adapter proteins: Fis, MFF, MiD49, and MiD51, where the latter three ensure equatorial mitochondrial fission, while Fis1 promotes organelle fragmentation and subsequent utilization through mitophagy [9]. An additional regulatory level involves post-translational modifications of Drp1 protein. For instance, phosphorylation of Drp1 at serine (Ser) 585 and Ser 616, as well as dephosphorylation at Ser 637, promote Drp1 translocation to mitochondria, activating fission [9]. Conversely, phosphorylation at Ser 637 inhibits it, protecting mitochondria from degradation via autophagy [9].

Activation of Drp1-dependent mitochondrial fission is known to accompany various pathological conditions of the central nervous system [1], yet remains essential for maintaining normal neuronal functions. Specifically, Drp1 participates in neurogenesis regulation, controls neuronal survival, contributes to synaptic contact organization, and mediates vesicular neurotransmitter transport [11]. Cases of epilepsy associated with mutations in the Drp1-encoding gene have been documented [12]. Notably, Drp1 inhibition has been shown to reduce seizure frequency in epilepsy models and exert neuroprotective effects [13].

In this context, investigation of the mitochondrial division inhibitor mdivi-1 — an inhibitor of Drp1-mediated mitochondrial division — appears promising. Studies confirm that mdivi-1 crosses the blood-brain barrier [14] and demonstrates several neuroprotective properties in experiments: anti-apoptotic effects, reduction of glutamate excitotoxicity and reactive oxygen species production [15, 16], as well as potential modulation of pathological neuroplasticity through regulation of neurogenesis and synapse formation [17].

The anti-apoptotic mdivi-1 effect is associated with reduced release of apoptosis-inducing factor and its impact on caspase-dependent apoptotic factors [15]. The impact of mdivi-1 on excitotoxicity mechanisms stems from its ability to regulate calpain activation by modifying Ca²⁺ homeostasis [16], and to moderately inhibit mitochondrial respiration [18], which however determines a range of off-target effects that can both expand and limit the application spectrum of this inhibitor.

Mdivi-1 has been reported to suppress neuroinflammation and reduce production of pro-inflammatory factors by activated astrocytic glia and microglia [19, 20], though its effect on oligodendroglia proved negative by increasing their susceptibility to Ca²⁺ overload [21].

Despite numerous studies describing neuroprotective properties of mdivi-1 in various neurodegeneration models, its effects on neurons and other cellular populations, including cells of the neurogenic niche in the hippocampal subgranular zone (SGZ), remain insufficiently studied [17, 22]. The pleiotropy of mdivi-1 effects should be considered when evaluating its impact on stem and differentiating cells, as neurogenesis involves stage-specific restructuring of bioenergetics and mitochondrial morphology [23], which may significantly alter expected outcomes of agents regulating mitochondrial dynamics.

The study aim is to conduct comprehensive analysis of possibilities for mitochondrial dynamics modulation in a kainate-induced hippocampal damage and summarize data from a series of experimental studies with mitochondrial fission inhibitor mdivi-1.

Materials and methods

Substance administration and experimental groups

The study was approved by the Ethics Committee of Russian Center of Neurology and Neurosciences (Protocol 5-3/22 dated June 06, 2022). Experiments were conducted on male Wistar rats. For intracerebral substance administration, an RWD Stereotaxic_Auto stereotaxic instrument (RWD) was used. Anesthesia was maintained using Zoletil-xylazine combination. Kainic acid injections (0.2 µg) were performed in the CA1 region of the right hippocampus, while 0.9% NaCl was administered contralaterally. Mdivi-1 (Ambeed) was injected into the cerebral ventricles (2.33 µg in 0.1% dimethyl sulfoxide into both ventricles). The methodology of stereotaxic surgeries, animal handling and maintenance procedures have been described in detail in our previous publications [17, 22]. Proliferating cell labeling was performed using 5-bromo-2’-deoxyuridine (BrdU, Ambeed) administered intraperitoneally at a dose of 40 mg/kg.

For each experiment, appropriate animal groups were formed (n = 5–9, minimum 5 in control groups, 6 or more in experimental groups), including a control group receiving only 0.9% NaCl injections, and experimental groups receiving either mdivi-1 alone, KA alone, or their combination. The timing and frequency of substance administration in the experiments are shown in Fig. 1. Specific instructions and references are provided for the discussed experiments with mdivi-1 administration, whose results have been partially published previously (Fig. 1, C) [17].

 

Fig. 1. Schematic diagram of a series of experiments to assess the effect of mdivi-1 on hippocampal DG neurons and the stages of neurogenesis therein.

А, B — experiment with simultaneous intrahippocampal KA administration and intraventricular mdivi-1 injection (animal removal from experiment after 5 (A) and 14 days (B)); С — experiment with mdivi-1 administration 14 days after KA injection.

 

Behavioral testing

Animal testing began 2 days prior to removal from the experiment. The novel object recognition test apparatus consisted of a square field (78 × 78 cm) with 42 cm high walls and 600 lux illumination. Animals were tested in 2 sessions. On day 1, two objects (cubes) of identical color and size were placed in the arena. Both objects in the first session were equivalent, therefore investigation time (sniffing) for each object normally showed no difference. On day 2, one object was replaced with a new one (cone), and the time spent investigating each object was recorded.

To assess photostimulation effects on behavior, the Opto-Varimex 4 system (Columbus Instruments) was used to record animal movement parameters. In a 42 × 42 cm square field, freezing time and bursts of stereotypical movements were recorded during 2-minute sessions under low illumination (172 lux) and stroboscopic arena lighting (5000 lux). Photostimulation at 10–20 Hz has been described for inducing photosensitivity in rats and eliciting spike-wave discharges on cortical EEG [24].

Immunomorphological study

Immunomorphological evaluation and morphometry were performed using established protocols [17, 22]. Serial cryosections from the hippocampal anterior third were immunostained with antibodies to doublecortin (DCX) and polysialylated neural cell adhesion molecule (PSA-NCAM) to identify neuronal precursors. Mature neurons were labeled with neuronal nuclear protein NeuN. Bromodeoxyuridine (BrdU) incorporation was detected using Bu20a antibodies to assess proliferative activity. Mitochondrial markers included succinate dehydrogenase B (SDHB) and Translocase of Outer Mitochondrial Membrane 20 (TOMM20). Mitochondrial dynamics were analyzed using antibodies against mitofusin 2 (Mfn2) and dynamin-related protein 1 phosphorylated at Ser616 (pDrp1). Microglial (IBA1+) and astrocytic (GFAP+) populations were identified using cell-specific markers. Apoptosis was assessed via activated caspase-3 immunolocalization. Long-term experiments (28 days; Fig. 1, C) included antibodies targeting mitochondrial complex I subunit NDUFS3, synaptophysin, and postsynaptic density protein 95 (PSD95).

Image analysis utilized ImageJ (NIH) and Nikon NIS-Elements platforms. Fluorescence intensity was quantified through threshold-based segmentation (≥ 250 cells/group) or manual analysis at 40× magnification. Cell counting involved operator-guided region-of-interest demarcation and nuclear classification. DCX⁺ cell density was normalized to DG length along its inferior margin. Quantitative methodologies were implemented as previously described [17].

Statistical data processing

Statistical data processing was performed using GraphPad Prism software (GraphPad Software). Data were presented as mean ± SD (standart deviation). One-way or two-way ANOVA with Tukey’s post hoc test were used for group comparisons. Differences were considered statistically significant at p<0.05.

Results

Changes in mitochondrial dynamics indicators under mdivi-1 exposure

In a short-term experiment (5 days) with concurrent administration of mdivi-1 and KA, we observed significant decrease in pDrp1 to TOMM20 (outer mitochondrial membrane marker) ratio in granular layer cells compared to controls across all experimental groups (Fig. 2, A). The combined administration of KA and mdivi-1 caused the most pronounced decrease in this parameter. For Mfn2 to SDHB (mitochondrial marker) ratio, we detected statistically significant increase in both mdivi-1 treatment groups and decrease in KA-only group compared to controls (Fig. 2, B). We suggest that using Drp1 to TOMM20 staining level ratio as an indicator of mitochondrial dynamics most comprehensively reflects Drp1 involvement in mitochondrial changes, since Drp1 levels should be assessed relative to total mitochondrial content. Notably, phosphorylated Drp1 levels in granular neurons were generally lower than in polymorphic layer neurons.

 

Fig. 2. Changes in fluorescence intensity ratios for fission protein (pDrp1) and fusion protein (Mfn2) to mitochondrial markers TOMM20 and SDHB in DG hippocampal granular layer neurons under mdivi-1 and KA exposure (5-day duration).
А — intracellular pDrp1/TOMM20 ratio; B — intracellular Mfn2/SDHB ratio; С — example of immunofluorescence detection of pDrp1 (green) and outer mitochondrial membrane marker TOMM20 (red) in DG hippocampal neurons (mdivi-1 treated animal). Nuclear staining with DAPI shown in blue. Upper left inset shows channel overlay. *p < 0.05 vs control group (NaCl); ^#p < 0.05 vs KA-treated animals.

 

Regarding changes in Drp1 under KA exposure in long-term experiments, we obtained data differing from this result. Specifically, 14 and 28 days after KA administration, the levels of phosphorylated and total Drp1 in granule layer neurons were increased [22], indicating the phasic nature of mitochondrial dynamics changes under KA exposure. As for the direct effects of mdivi-1, our previous study demonstrated that the significant increase in mitochondrial size induced by this division inhibitor persists for up to 14 days post-administration [17], which aligns with the present findings. We also previously observed that mdivi-1 did not affect immunofluorescence intensity when detecting proteins of respiratory complexes I (NDUFS3) and II (SDHB) [17], though literature reports describe its non-selective suppression of NADH dehydrogenase activity [25].

Notably, mdivi-1 reduced pDrp1 levels in DG granule neurons while increasing relative Mfn2 content in both mdivi-1-only and KA-treated groups during short-term experiments, with similar changes observed in the 14-day post-inhibitor administration experiment.

Effect of mdivi-1 on neurodegeneration and kainate-induced neuroinflammation

Five days after KA administration, a significant decrease in neuronal density (Fig. 3, A) was observed in the CA3 of the hippocampus (by 40%), and to a lesser extent in the granular DG layer. Concurrently, an increase in immunofluorescence staining for GFAP and the microglial marker IBA1 (Fig. 3, В, С, Е) was noted in the polymorphic DG layer, consistent with our previous findings in the kainate model [5].

 

Fig. 3. Indicators of the neurodegenerative process and proinflammatory changes in glia during kainate-induced hippocampal injury and mdivi-1 administration (duration - 5 days).
А — number of NeuN⁺ neurons per field of view; B — changes in immunofluorescence intensity for GFAP in DG hippocampus, % relative to control; C — changes in immunofluorescence intensity for IBA1 in DG hippocampus, % relative to control; D — mean immunofluorescence intensity of caspase-3 in CA3 hippocampal field neurons, arbitrary units; Е — detection of neurons (NeuN, red) and microglia (IBA1, green) in granular (GrL) and polymorphic (Pol) layers of DG hippocampus; F — detection of neuronal marker NeuN (red) and activated caspase-3 (green) in CA3 hippocampal field.

 

The level of activated caspase-3 (Fig. 3, D, F) in CA3 field neurons was also significantly increased in KA-treated animals. Combined administration of mdivi-1 and KA in the short-term experiment did not significantly affect the studied neurodegeneration parameters, except for reducing the microglial marker level, suggesting a targeted action of mdivi-1 on microglial activation.

Effect of mdivi-1 on proliferation and differentiation of SGZ cells in the hippocampus

Assessment of neural progenitor proliferation in SGZ on day 5 of the experiment revealed that mdivi-1 did not reduce the number of dividing cells. In contrast, animals receiving KA showed suppressed cell division in SGZ, while combined administration of KA and mdivi-1 increased the number of BrdU-labeled nuclei to control levels (Fig. 4). As mitochondrial fusion promotes proliferation and self-renewal of neural progenitors during neurogenesis, mdivi-1 administration in the KA+mdivi-1 group likely affected proliferating cells by altering their mitochondrial dynamics and enhancing mitochondrial fusion, ultimately leading to increased BrdU⁺ cell numbers in SGZ.

The number of DCX⁺ neuronal precursors in the SGZ on day 5 after combined administration of KA and mdivi-1 was decreased, with a further decline observed in all experimental groups by day 14, reaching the lowest values in the mdivi-1 group (Fig. 4, C). In animals receiving KA alone, this parameter was lower than in the control group but increased compared to the mdivi-1 group, while combined administration of KA and mdivi-1 did not result in a synergistic reduction of precursor numbers. By day 28, an increase in DCX⁺ cells was observed in KA-treated animals compared to the previous time point, while in the mdivi-1 group their numbers increased but did not reach control values. The KA+mdivi-1 group showed uniformly reduced DCX⁺ cell counts at all time points (Fig. 4, C).

 

Fig. 4. Effects of mdivi-1 on proliferation and differentiation of SGZ cells in hippocampal DG of control animals and in KA administration experiment.
А — changes in percentage ratio of DCX⁺ cells at different maturation stages (according to Plumpe classification, 2006) in the granular layer of hippocampus 14 days after mdivi-1 administration; B — number of BrdU⁺ cells per SGZ unit length in short-term (5 days after BrdU and KA administration) experiment and number of BrdU⁺ neurons in long-term (28 days) experiment; C — changes in DCX⁺ cell counts in hippocampal SGZ on days 5, 14, and 28 after KA and mdivi-1 administration (% relative to corresponding control); D — differences in morphology of DCX⁺ cells (green) in hippocampal SGZ between control animals and KA+mdivi-1 group; Е — example of BrdU detection (red) in SGZ in short-term experiment (5 days) and BrdU⁺ neurons (NeuN staining — green) in long-term experiment (28 days). Arrows indicate nuclear (DAPI staining — blue) localization of the label, ×100.

 

Analysis of DCX⁺ cell morphological type ratios according to the classification proposed by T. Plümpe et al. [26] revealed that mdivi-1 increases the proportion of DCX⁺ cells at early maturation stages (Fig. 4, A) while reducing the fraction of cells in the postmitotic phase with branched processes extending to the molecular layer (Fig. 4, A, D). We previously reported altered morphology and distribution patterns of DCX⁺ cells in the kainate-induced injury model, where most DCX⁺ cells lacked long processes and were localized in the polymorphic layer of DG [22].

Analysis of BrdU⁺/NeuN⁺ cell numbers in the long-term experiment demonstrated that by day 28 mdivi-1 reduced the number of mature NeuN⁺ neurons with incorporated label, while KA administration increased their numbers, and the KA+mdivi-1 group maintained values close to controls, confirming the inhibitory effect of mdivi-1 on neuronal precursor maturation (Fig. 4, B, E).

We suggest that KA initially suppresses proliferation in SGZ, but the developing epileptogenic focus subsequently promotes accelerated maturation of granular neurons, most of which exhibit abnormal process organization, resulting in aberrant neurogenesis. Conversely, mdivi-1 appears to restrain this accelerated differentiation.

Assessment of mdivi-1 effects on animal behavior and synaptic parameters

In the novel object recognition test on day 1, no significant differences in exploration time of identical objects were found among rats from all groups. On day 2, when replacing one familiar object with a novel one, animals receiving only mdivi-1 showed increased exploration time of the novel object, representing a normal response to new environment, while animals receiving KA alone or in combination with mdivi-1 showed no difference in interaction time between both objects (Fig. 5, A), indicating memory impairment. In the KA+mdivi-1 group, the total object exploration time was significantly higher, which might be associated with prolonged interaction episodes.

 

Fig. 5. Behavioral changes in animals 14 days after combined administration of KA and mdivi-1 in novel object recognition and photostimulation tests.
А — exploration time of familiar (white bars) and novel (gray bars) objects on day 2 of testing; B — duration of freezing episodes in actometer arena under normal (white bars) and strobe (gray bars) lighting conditions, dashed line — mean for control animals.
*p < 0.05 between indicators within one group; ^#p < 0.05 compared to KA group.

 

Moreover, mdivi-1 exacerbated cognitive impairments 14 days after KA administration by increasing total exploration time without differentiation between novel and familiar objects, which could be considered as stereotypy and disruption of goal-directed behavior. These changes might be caused by hippocampal damage, particularly in CA1-CA3 regions involved in contextual recognition formation.

Our previous study demonstrated that 14 days after mdivi-1 administration, no significant changes in motor activity compared to sham-operated controls were observed [17]. In the current work, when assessing motor activity using Opto-Varimex actometer under photostimulation (Fig. 5, B), no significant differences were found in groups receiving only mdivi-1 or only KA (though KA group showed higher immobility duration). However, animals receiving both mdivi-1 and KA exhibited significantly increased freezing duration in response to photostimulation. Furthermore, the KA+mdivi-1 group showed significant increase in stereotypical movements during photostimulation (18.6 ± 2.0) compared to KA-only group (12.5 ± 1.8).

Additionally, previous reports indicated that 14 days after intracerebroventricular administration of mdivi-1, synaptophysin staining intensity in DG molecular layer significantly decreased compared to sham-operated animals, while PSD95 content in granular layer increased [17]. The observed behavioral changes and previously obtained data on mdivi-1 effects on synaptic proteins in hippocampal DG suggest synaptic influence imbalance and allow us to conclude that mdivi-1 aggravates KA-induced behavioral impairments at investigated time points.

Discussion

Based on literature data [13, 16–19] and experimental results, the following processes potentially affected by the mitochondrial fission inhibitor mdivi-1 in the hippocampal DG can be proposed (Fig. 6):

 

Fig. 6. Neurogenesis in the granule cell layer of the hippocampal DG and effects of mitochondrial dynamics inhibitor mdivi-1 under abnormal conditions.
1 — apoptotic cell death; 2 — gliogenesis and pro-inflammatory glial activation; 3 — impaired migration and differentiation of neuronal precursors, dispersion of granule cell layer; 4 — synaptic dysfunction and aberrant plasticity.
CA — cornu ammonis; SGZ — subgranular zone; GCL — granule cell layer; MCL — molecular layer; NSC — neural stem cell; SYP — synaptophysin; Nestin — neural stem cell marker; NeuN — mature neuron marker protein.

 

It is suggested that reduced mitochondrial fission under mdivi-1 action normalizes mitochondrial functions, decreases cell apoptosis and suppresses neurodegeneration, as demonstrated in various experimental models [19]. However, we did not detect increased levels of active Drp1 in DG granule layer neurons at early stages after KA administration. Increased Drp1 phosphorylation under similar conditions has been described for CA3 field neurons [27]. These differences may be attributed to regional specificity of Drp1 phosphorylation and features of mitochondrial fission regulation in neurons of different hippocampal regions [28]. Additionally, selective activation of kainate receptors, compared to glutamate action in actual abnormal conditions, causes less pronounced mitochondrial membrane depolarization and Ca²⁺ overload [29], which are required for induction of mitochondrial fission [30]. Drp1-dependent mitochondrial fission activation at later stages, demonstrated in our previous studies [22], is most likely associated with pathological synaptic reorganization in DG [6] and reflects delayed effects of KA administration.

The excessive inhibition of Drp1 by mdivi-1 observed in the short-term experiment in KA-treated animals further destabilizes mitochondrial dynamics. Cells with high basal pDrp1-S616/pDrp1-S637 ratio show increased sensitivity to inhibition of mitochondrial division, which exacerbates KA-induced damage, including through autophagy impairments [27].

The observed increase in Mfn2/SDHB ratio in the KA+mdivi-1 group may be associated with mdivi-1’s ability to modulate Ca²⁺ signaling independently of Drp1 [16]. Mfn2, besides regulating mitochondrial fusion, also controls mitochondrial interaction with endoplasmic reticulum membranes [10]. This elevation of Mfn2 in the KA+mdivi-1 group may reflect enhanced mitochondrial-ER tethering and improved Ca²⁺ buffering capacity.

In our experiment, granule neurons of the DG showed less damage compared to CA3 neurons, consistent with reports of low apoptosis levels in granule neurons following status epilepticus induction [31]. Mdivi-1 administration did not demonstrate immediate neuroprotective effects in the short-term experiment, likely due to predominant necrotic excitotoxic cell death. Other studies have reported mdivi-1-mediated reduction of KA-induced CA3 neuronal damage, suppression of Drp1-dependent mitochondrial fission, and PARKIN-mediated degradation [32]. An anti-apoptotic effect of mdivi-1 has also been proposed through reduced cytochrome c release and caspase-3 inhibition [28]. Conversely, supporting our findings, the pilocarpine epilepsy model revealed no neuroprotective effects of mdivi-1 [33]. In the same study, ¹⁸F-FDG PET demonstrated mdivi-1 exacerbation of status epilepticus-induced cortical hypometabolism, attributed by authors to astrocytic dysfunction [33]. These literature discrepancies likely relate to mdivi-1 administration protocols, as studies showing neuroprotection [28, 32] administered the drug prior to KA.

KA administration induced a pro-inflammatory glial response, correlating with previously described glial changes [5, 32]. According to our data, mdivi-1 attenuated microglial activation but did not affect astrocytes. Mitochondrial fission activation is known to be associated with microglial release of pro-inflammatory factors and reactive oxygen species [34], as well as linked to NLRP3 inflammasome signaling cascades [35]. In astrocytes, mitochondrial fission through the NF-κB signaling pathway enhances the expression of interleukin-1β, tumor necrosis factor-α, and inducible nitric oxide synthase [20], which exacerbate glutamate excitotoxicity by suppressing astrocytic glutamate metabolism, leading to neuronal hyperexcitability [36]. The present findings are consistent with reported reductions in microglia-induced neuroinflammation in the kainate model [32], but contradict aforementioned experiments in the pilocarpine epilepsy model [33] where mdivi-1 reduced reactive astrogliosis while enhancing microglial pro-inflammatory response. The role of microglia in epilepsy remains ambiguous, as it controls neuronal survival in the DG, while its activation is necessary for eliminating ectopic neurons that promote aberrant excitation in the DG [37].

Aberrant neurogenesis in the DG is considered to contribute to neuronal network hypersynchronization in temporal lobe epilepsy. The proliferation, differentiation, maturation, and migration of neuronal precursors are known to involve modifications in mitochondrial dynamics and metabolic reprogramming [23], suggesting that regulators of mitochondrial fission may be relevant for addressing abnormal plasticity.

Our studies and others using the KA-induced hippocampal sclerosis model have demonstrated morphological and distributional abnormalities in DCX⁺ neuronal precursors [22, 38]. Ectopic migration and dendritic reorganization of granule cells have been described in human hippocampal sclerosis and experimental epilepsy models [39, 40]. According to D.M.S. Moura et al., DCX⁺ cell maturation depends on network activity in hippocampal structures, while DCX⁺ cells themselves can regress to earlier developmental stages with restored astrogliogenic potential [41].

The present study demonstrated suppression of neurogenesis in the SGZ under mdivi-1 influence. Moreover, we previously observed formation of DCX⁺ cell clusters, reduction of PSA-NCAM^+/DCX⁺ cells with branched processes in the molecular layer, and impaired maturation of granular neurons [17]. Similar changes have been described in the subventricular zone and, according to H.J. Kim et al., are associated with altered intracellular mitochondrial localization necessary for cytoskeletal remodeling in differentiating neurons [42].

The observed decrease in SGZ proliferation on day 5 post-KA injection appears related to direct damage to the hippocampal neurogenic niche [43], and may also be explained by tumor necrosis factor-α secretion from reactive microglia capable of suppressing neural progenitor proliferation in SGZ [44].

Previous research has demonstrated the role of mitochondrial dynamics in differentiation and proliferation of neural stem cells [45] — mitochondrial fusion enhanced their renewal, while activation of mitochondrial fission promoted further differentiation. However, another study showed that mdivi-1 exposure in vitro reduced proliferation of neuronal precursors and HeLa cells [25]. In our study, we found no significant effect of mdivi-1 on SGZ cell proliferation in control animals at early timepoints after administration, but in the KA+mdivi-1 group, the number of BrdU-labeled cells in SGZ increased, which may be associated with mdivi-1 anti-apoptotic effects and could depend on agent concentrations and the balance of mitochondrial fusion/fission processes.

KA administration in this study reduced DCX⁺ cell number by day 14, indicating decreased neurogenesis. However, by day 28, there was an increase in DCX⁺ neuronal precursors compared to earlier timepoints, accompanied by elevated numbers of mature neurons with incorporated labels. These findings align with the described biphasic action of KA showing early inhibitory and delayed stimulatory effects on neurogenesis [46]. It should be noted that despite increased neuron formation in the DG granular layer in epileptic models, most demonstrate distribution and morphology abnormalities [39, 47], indicating pathological neurogenesis that promotes epileptiform activity. Early administration of mdivi-1 in the kainate model normalized dividing cell counts in SGZ but did not increase DCX⁺ precursors and reduced BrdU-labeled mature neurons, likely reflecting slowed maturation and shift toward gliogenic differentiation, correlating with findings by D.M.S. Moura et al. [38].

Therefore, in our experiments mdivi-1 demonstrated an effect on hippocampal neurogenesis, which could be considered as a potential opportunity for modulating aberrant neurogenesis in epilepsy. However, several studies have reported that suppression of neurogenesis increases sensitivity to KA action. Mice with reduced neurogenesis showed higher frequency, severity and duration of seizures induced by KA administration [48], with similar results observed in the pilocarpine model of temporal lobe epilepsy [49]. These findings are consistent with our observed exacerbation of KA-induced behavioral impairments in animals following mdivi-1 administration, which suppressed neurogenesis and neuronal differentiation in DG. In hippocampal lesions and epilepsy modeling, animals demonstrate a tendency for repetitive behavioral strategies, including repeated returns to the same arena zones or objects [50–52], which parallels our results indicating impaired recognition mechanisms and increased stereotypical movements in animals receiving mdivi-1 after KA-induced hippocampal damage.

Conclusion

The mitochondrial division inhibitor mdivi-1 in our experiments did not demonstrate neuroprotective properties against the acute excitotoxic effect of kainic acid on mature hippocampal neurons, although reduced microglial activation in DG was noted. At the same time, we demonstrated a pronounced inhibitory effect of mdivi-1 on maturation and differentiation of granular layer neurons in the hippocampus both in control animals and in the KA administration model, indicating the possibility of modulating aberrant neurogenesis through inhibition of mitochondrial fission. However, no positive behavioral effects were detected when using mdivi-1 at short intervals following kainate-induced hippocampal damage.

Despite numerous promising experimental data, the use of mitochondrial dynamics inhibitors for correcting pathological processes in the hippocampus requires differentiated assessment of their effects on various cell types. The effects of mdivi-1 should be investigated at different time points after hippocampal injury reproduction to determine its true therapeutic potential. In most studies characterizing the neuroprotective properties of mdivi-1, it was used preventively, which prevents full translation of the established protective effects to real clinical conditions.

About the authors

Dmitry N. Voronkov

Russian Center of Neurology and Neurosciences

Author for correspondence.
Email: voronkov@neurology.ru
ORCID iD: 0000-0001-5222-5322

Cand. Sci. (Med.), senior researcher, Laboratory of neuromorphology

Russian Federation, Moscow

Evgenia N. Fedorova

Russian Center of Neurology and Neurosciences; Pirogov Russian National Research Medical University

Email: ewgenia.feodorowa2011@yandex.ru
ORCID iD: 0000-0002-2128-9056

junior researcher, Laboratory of neuromorphology, Russian Center of Neurology and Neuroscience, assistant, Department of morphology, Institute of Anatomy and Morphology named after Acad.Yu. M. Lopukhin

Russian Federation, Moscow; Moscow

Anastasia K. Pavlova

Russian Center of Neurology and Neurosciences

Email: pav_nastasya@mail.ru
ORCID iD: 0009-0006-5653-5524

research assistant, Laboratory of experimental pathology of nervous system and neuropharmacology

Russian Federation, Moscow

Maria S. Ryabova

Russian Center of Neurology and Neurosciences

Email: voronkov@neurology.ru
ORCID iD: 0009-0003-5596-7630

research assistant, Laboratory of neuromorphology

Russian Federation, Moscow

Anna V. Egorova

Russian Center of Neurology and Neurosciences; Pirogov Russian National Research Medical University

Email: av_egorova@bk.ru
ORCID iD: 0000-0001-7112-2556

Cand. Sci. (Med.), researcher, Laboratory of neuromorphology, Associate Professor, Department of morphology, Institute of Anatomy and Morphology named after Acad. Yu. M. Lopukhin

Russian Federation, Moscow; Moscow

Alla V. Stavrovskaya

Russian Center of Neurology and Neurosciences

Email: alla_stav@mail.ru
ORCID iD: 0000-0002-8689-0934

Cand. Sci. (Biol.), Head, Laboratory of experimental pathology of nervous system and neuropharmacology Brain Institute

Russian Federation, Moscow

Ivan A. Potapov

Russian Center of Neurology and Neurosciences

Email: potapov.i.a@neurology.ru
ORCID iD: 0000-0002-7471-3738

junior researcher, Laboratory of experimental pathology of nervous system and neuropharmacology Brain Institute

Russian Federation, Moscow

Vladimir S. Sukhorukov

Russian Center of Neurology and Neurosciences; Pirogov Russian National Research Medical University

Email: voronkov@neurology.ru
ORCID iD: 0000-0002-0552-6939

Dr. Sci. (Med.), Professor, Head, Laboratory of neuromorphology, Department of morphology, Institute of Anatomy and Morphology named after Acad. Yu. M. Lopukhin

Russian Federation, Moscow; Moscow

References

Supplementary files