Sex Differences and Olfactory Bulb Alterations Contribute to Limbic Circuit Dysfunction and Behavioral Disturbances in Pilocarpine Model of Epilepsy

The olfactory bulb at the sensory and circuit level transmits information to the limbic and cortical systems for behavioral outputs, and disruption of such circuits induces behavioral disturbances in rodents. Previously, data from our laboratory showed the occurrence of behavioral disturbances in Wistar rats submitted to the pilocarpine model of epilepsy (PME) and that these alterations were sex related. Here we deepen our ndings that sex-linked differences are present in PME and that male epileptic rats exhibit profound recurrent seizure patterns, namely seizure duration, severity, and distribution along the light/dark cycle different from that observed in epileptic female rats. Further, using isotropic fractionator we observed signicant alterations in the number of neuronal and non-neuronal cells of the olfactory bulb, amygdala, and hippocampus following 3 months of spontaneous recurrent seizures in epileptic male and female rats. Altogether, our study suggests that neuronal and non-neuronal cell death in olfactory bulb may interfere with sex-related differential recurrent seizure patterns, limbic circuit dysfunction, and behavioral disturbances in PME. Lastly, the pilocarpine epilepsy model provides an evidence-based tool to study mechanisms of behavioral disturbances in epileptogenesis that may provide future therapeutic insights in our quest to improve the life of people with epilepsy.


Introduction
Epilepsy is a brain disorder with electroclinical characteristics presenting recurrent non-evoked seizures, often associated with signi cant psychological morbidity and complications 1,2 . Comorbidities in epilepsy share complex and common pathogenic mechanisms whose pathophysiological mechanism is poorly understood and need to be studied as an essential part of epilepsy research 3,4 . Several studies on comorbidities related to epilepsy indicate that there are gender differences in these manifestations, for example, major depression is greater in female patients, while anxiety disorders are more common in men 5,6 . Behavioral disturbances have also been studied in experimental epilepsy. Our group have observed that male and female rats submitted to the pilocarpine model of epilepsy present behavioral changes, such as: maternal behavior is disrupted in female epileptic rats that can even cannibalize their pups while male epileptic rats are aggressive and present hyposexuality 7,8 . Other behavioral abnormalities found in mice and rats submitted to the pilocarpine model include anxiety 9,10 , and depressive like behavior 11,12 . Interestingly, similar behavioral disturbances and morphological alterations in the limbic circuits have also been observed in rats with olfactory bulbectomy 13,14 . The olfactory bulb of the rat is a crucial part of the neuronal network, whose sensory modality determines the immediate survival of an organism through activities such as the search for food, recognition of predators, social interaction, reproduction and several other aspects of behavior 15 . Neurons present in the olfactory bulb form diverse connections with various structures of the brain 16 . Via efferent connections to the amygdala and hippocampus, the olfactory bulb sends information to limbic structures where modulation and orchestration of emotional and behavioral response take place 15 . It is impressive to note the importance of olfactory stimuli to the relationship between rodent mothers and their offspring. In fact, extensive circuit reorganization and neurogenesis occur in the olfactory bulb and associated structures during pregnancy, delivery and puerperium of female rats and mice 17,18 .
Epileptic circuits, regardless of the underlying mechanism lead to an imbalance between glutamate and GABA activity 19 . This imbalance may be associated with circuit alterations that directly or indirectly contribute to limbic dysfunction and sex-linked behavioral disturbances observed in the pilocarpine epilepsy model.
Although the occurrence of psychiatric and behavioral comorbidities associated with epilepsy are well recognized 20 , the mechanisms underlying these comorbidities in humans and in animals with temporal lobe epilepsy are poorly understood.
Recurrent seizures are the hallmark of behavioral manifestation of temporal lobe epilepsy 21 . Studies in the pilocarpine model of epilepsy showed that limbic seizures commit to neuronal damage and circuit disruption 22,23 . Different lines of evidence reported that mechanistic characterization of recurrent seizures is poorly understood and lacking in sex difference studies of epilepsy 24,25 . Furthermore, we clearly do not understand, how neuronal death and injury-stimulated circuitry reorganization present in the epileptic brain, may induce behavioral disturbances in male and female rats.
The aim of the present study was to understand whether the occurrence of behavioral disturbances in male and female rats submitted to the pilocarpine model is related to different patterns of brain injuries involving the limbic system and to verify whether changes in the olfactory bulb are also sex related during epileptogenesis. Understanding modi cations that occur in the olfactory bulb, amygdala and hippocampus underlying long-term recurrent seizures is important in the pathophysiology of temporal lobe epilepsy and could advance our understanding of epileptic brain pathways involved in associated comorbidities.

Results
Initiation of status epilepticus in all animals was observed at an average of 25 min after pilocarpine administration (range 15-37.8 minutes). Both male and female Wistar rats were exposed to 5hr of status epilepticus (SE) and were able to sustain seizure stage score of 3 or higher on the Racine scale. Unpaired t-test showed no statistical signi cance between the mean latent period (time in days between SE induction and expression of rst spontaneous seizure) of epileptic males (15.1 days) and female rats (14.6 days), p > 0.627. Graph not shown.
We examined male and female Wistar rats submitted to pilocarpine after the rst spontaneous seizure using continuous video monitoring and quanti ed for recurrent seizure patterns using the Racine score scale. Repeated measures two-way ANOVA revealed signi cant main effect of sex (F (1, 8) = 91.35; p < 0.0001), time (F (1.301, 10.41) = 88.03; p < 0.0001) and interaction between sex and time (F (2, 16) = 31.60; p < 0.0001) on seizure duration over a 3-month period. Sidak's post-hoc tests showed that the duration of each single seizure was higher in male in comparison to female epileptic rats in the1st (mean gap = 109.8 sec, p < 0.003), in the 2nd (mean gap = 274.8 sec, p < 0.001), and in the 3rd month (mean gap = 610.4 sec, p < 0.0008) of observation (Fig. 1A).
On the other hand, and as shown in Fig. 1B, the number of spontaneous seizures in female epileptic rats was signi cantly higher than that observed in male epileptic rats during the 2 rst months of observation  Sidak's post-hoc tests revealed that the severity of spontaneous seizures based on Racine scale 26 in male epileptic rats was higher than that observed in female epileptic rats in the 1st (mean range = 1, p < 0.004), in the 2nd (mean range = 1.4, p < 0.007), and in 3rd month (mean range = 2, p < 0.0003).
The occurrence of cluster of seizures during the observation period showed also to be in uenced by the sex (F (1, 8) = 43.20; P < 0.0002) and by the time (F (2, 16) = 4.098; p < 0.03). Sidak's post-hoc tests showed that male epileptic rats had less seizure clusters during the rst month than the female epileptic rats (mean range = 4 seizure groups, p < 0.0001). However, no difference was observed in the 2nd and 3rd month (p > 0.05) of observation (Fig. 1D).
Repeated measures two-way ANOVA demonstrated signi cant main effects of sex F (1, 8) = 27075; p < 0.0001), time (F (1.393, 11.14) = 7411; p < 0.0001) and interaction between sex and time (F (11, 88) = 37.23; p < 0.0001) in pattern of distribution of spontaneous recurrent seizures recorded during the light/dark cycle for a period of 3 months total. Sidak's post-hoc tests showed epileptic male rats had a signi cant increase in the number of seizures during light/dark cycles when compared with female epileptic rats throughout the 24 hours, p < 0.0001. Two-way ANOVA demonstrated a signi cant main effect of sex (F (1, 16) = 30.48; p < 0.0001) but not group (F (1, 16) = 1.976; p = 0.1789) and interaction between sex and group (F (1, 16) = 0.4245; p = 0.5240) in brain mass between control and male and female epileptic rats. Tukey's post-hoc tests showed signi cant decrease in brain mass between control male and epileptic male rats (mean gap = 0.14g, p < 0.0159), control female and epileptic female rats (mean gap = 0.18g, p < 0.0024) but no other signi cant group differences were observed p > 0.05 (Fig. 3A). Figure 3B shows that the hippocampal mass signi cantly decreased (F (1, 16) = 27.38; p < 0.0001) in epileptic animals (males or females) compared to control animals (males or females). However, the decrease in mass related to the presence of epilepsy was similar for animals of both sexes (F (1, 16) = 2420; p = 0.1394), that is, the decrease in the epileptic hippocampus was 32 % for males (p < 0.0173) and 33.3% for females (p < 0.0051) in relation to the respective controls. Figure 3C shows the amygdala mass in control animals and rats with epilepsy. It can be observed that the amygdala mass was signi cantly reduced (F (1, 16) = 3.556; p = 0.0002) in both male and female animals with epilepsy when compared to controls. However, when analyzing the loss of the amygdala mass between female and male rats with epilepsy, it is noted that the loss of mass of the amygdala in female rats with epilepsy (57.1%) was greater than that observed in male rats with epilepsy (41.8 %).
In relation to the olfactory bulb (Fig. 3D), we could notice that a signi cant difference in the mass of this structure was already present between females and control males. The olfactory bulb in female rats is approximately 19.5% heavier than that of male rats (p < 0.02). In both groups, the presence of epilepsy signi cantly decreased olfactory bulb mass in animals of both sexes and, interestingly, mass loss was greater in male rats with epilepsy (F (1, 16) = 44.31; p < 0.0001) than in female rats with epilepsy ((F (1, 16) = 37.23; p < 0.0001).
Isotropic fractionator method was used to examine for the neuronal and non-neuronal cells in the hippocampus, amygdala and olfactory bulb respectively.   Tukey's post-hoc tests showed signi cant difference in the number of olfactory bulb neuronal cells between groups (control males and epileptic males with 27% reduction, p < 0.0001), (control females and epileptic females with 19.5% reduction, p < 0.0001), (epileptic males and epileptic females with 7.5% reduction, p < 0.0002) and (control males and females with 12.5% increase, p < 0.004).
Regarding the non-neuronal cells present in the three analyzed structures (Fig. 5A), the two-way ANOVA test showed a signi cant variation in the number of these cells depending on the group (control versus epilepsy; F (1, 16) = 72.38; p < 0.0001), of the interaction between sex and group (F (1, 16) = 7.66; p = 0.0137), but not sex (F (1, 16) = 2.35; P = 0.1443). Tukey's post-hoc tests showed a signi cant increase in the number of non-neuronal cells between groups (control female rats versus female rats with epilepsy with 10% increase, p < 0.03), (control male rats versus female control rats with 17% increase, p < 0.004), (epileptic male rats versus epileptic female rats with 36.1% increase, p < 0.0001), but no differences were observed between the other groups.
The number of non-neuronal cells in the amygdala (Fig. 5B)

Discussion
The present study provides two main novel ndings in the epileptic brain. First, signi cant differences in the spontaneous recurrent seizure patterns exist among male and female rats submitted to the pilocarpine model of epilepsy. Second, the topographic distribution of neuronal and non-neuronal cell death differs in the brains of male and female epileptic rats. These ndings provide one possible explanation for limbic circuit dysfunction and behavioral disturbances observed in the pilocarpine epilepsy model. There is scanty data on chronic epileptic seizure patterns in female rats, most of the information is tailored towards S.E which does not meet the comparison criteria.
The present evidence summarizes signi cant sex-linked differences in main structures and long term recurrent seizure patterns: male and female rats demonstrated differences in amygdala and olfactory bulb mass reduction, hippocampus, amygdala and olfactory neuronal loss and non-neuronal cell death.
Additionally, epileptic male rats exhibited increased seizure duration, severity and more seizures during light/dark cycles per day while epileptic females showed increased seizure frequency and seizure clusters when compared during the 3 months period.
It is important to note the pilocarpine epilepsy model is an appropriate paradigm to study extensive extra hippocampal lesions or adjacent structure involvement in temporal lobe epilepsy 27,28 . Pilocarpineinduced neuronal death in limbic circuitry follows a cascade of events: from the activation of muscarinic receptors to the over activation of AMPA and NMDA receptors culminating in massive permeability to calcium ions leading to neuronal injury and death. Insights on how glutamatergic pathways promote hyperexcitable circuits associated with neuronal alterations and damage to adjacent structures has been provided by kindling studies 29 . For instance, olfactory bulb kindling facilitates spread of seizure activity via extensive connections to the amygdala and to the hippocampus leading to permanent structural changes in these structures 30 . Additionally, repeated amygdaloid kindling results into profound seizure activity and synaptic reorganization of the olfactory bulb-subventricular zone 31 .
The different distribution of neuronal and non-neuronal cell death in the limbic system of male and female epileptic rats may re ect the differential distribution of receptors to sex hormones in these structures as well as the connections formed between them throughout life and other factors such as the circadian cycle. The fact that female rats express higher density and differential distribution of estrogen receptors in the limbic structures compared to male rats and the in uence of the estrous cycle and circadian cycle in the number and function of these receptors is well known 32 .
In relation to experimental epilepsy, our laboratory has previously shown that epileptic female rats release increased estradiol and reduced progesterone levels 33 . Epileptic female rats are susceptible to progesterone withdrawal predisposing to increased seizure frequency and decreased allopregnanolone levels have been implicated in this phenomenon 25 . The mechanisms by which these two hormones promote or hinder seizure activity in limbic circuits are not fully understood in the brain 33 . In brief within brain circuits, estradiol is bidirectional ligand that exerts its proconvulsant effects through NMDA receptor and anticonvulsant action by Neuropeptide Y within the brain 34,35 . Progesterone mediates its anticonvulsant through allopregnanolone 36,37 . Testosterone is converted to estradiol promoting intense tonic clonic seizures observed in male epileptic rats 36,38 . Furthermore, progesterone has shown to modulate neuronal and non-neuronal cell death, neuroin ammation, neurogenesis in epileptic female rats 39 .
In support with our ndings (Figs. 1 and 2), kainic and pilocarpine model of epilepsy agree that epileptic male rats gradually intensify seizures characterized with increase in duration, severity and frequency with more seizure occurring during the light cycles of day over a period of time [40][41][42] . Seizure clusters have been identi ed in sub population of animals as seen in humans with TLE but exact mechanisms by which dysfunctional microcircuits trigger, propagate and terminate while expressing these different seizure patterns remain elusive 41 . We have earlier demonstrated that epileptic male rats continue to lose neurons in the hippocampus as they also exhibited recurrent seizures throughout their life-time 43 .
Brought together, sex differences and broad mechanistic effects of gonadal hormones on limbic excitability and seizure vulnerability may be in uenced by various factors like dimorphic disparities in epileptogenic regions of the brain that initiate and terminate seizures, intrinsic brain connectivity, receptor distribution and signaling pathways 44 . These changes may contribute to the differences in recurrent seizure patterns including seizure frequency, duration, clusters and severity.
Sex-linked differences of neuronal and non-neuronal cell death seen in the olfactory bulb, amygdala and hippocampus happens in a very important pathway that modulate behavioral outputs and also form a sensory window to the outside environment in rats. The long term recurrent seizures promote gradual neuron loss disrupting limbic circuits throughout epilepsy 43 . For instance, several lines of evidences in experimental epilepsy reported GABAergic circuit disruption of parvalbumin cell death in the olfactory bulb 45 , basolateral amygdala 29 , and hippocampus 46 . Via neural pathways, ventral CA1-hippocampal region and basolateral amygdala heavily rely on the olfactory sensory inputs to modulate fear, mood and anxiety-like behaviors 47 . In addition, these regions also express the greatest cell loss and damage from epileptic seizures 29,46 , and contain anxiety circuit pathways to the prefrontal cortex and hypothalamus 48, 49 . In the face of temporal lobe epilepsy, olfactory glutamatergic dysfunction via pathways may promote hyperexcitability in epileptogenic regions of the amygdala and hippocampus eliciting more recurrent seizures, damaging microcircuits that control behavior within limbic system and cortical connections.
Therefore, we conclude that sex-linked differences and recurrent seizure patterns in temporal lobe epilepsy signify the extent of brain damage and these microcircuit damages through neuronal death within the limbic system may contribute behavioral disturbances observed in the pilocarpine epilepsy model. Our results suggest that neuronal and non-neuronal cell death with differential recurrent seizure patterns among epileptic male and female rats may contribute to limbic circuit dysfunction and behavioral disturbances in the pilocarpine model of epilepsy. We have begun to explore genetic manipulations and special cell types in the olfactory-limbic pathways involved in circuit-level dysfunction of behavior disturbances in TLE. We hope that our studies will help identify potential molecular mechanisms and targets for behavioral disturbances in order to bridge our understanding of epilepsy within brain circuits.

Ethical declaration
This study was approved by the Ethic Committee on Animal Use of the Federal University of Sao Paulo (CEUA/UNIFESP) under the reference number 8624210118. All animal experiments were in accordance to international guidelines of animal care 50 and ARRIVE guidelines for conducting invivo experiments of animal research 51 .

Animals and Experimental groups
A total of 30 Wistar rats were used, consisting of males and female rats weighing 210 to 245g that were grouped into 4 groups namely: control males (n=5), control females (n=5), epileptic males (n=10) and epileptic females (n=10) respectively. Animals were housed in groups of four in controlled climatic conditions (light-dark cycle of 12 hours, lights on between 7 AM and 7 PM, and constant temperature of (19-23 °C), with free access to water and food. All experiments were randomly grouped, and a blinded manner approach used until data was analyzed.

Vaginal Cytology
Female Wistar rats were introduced with noninvasive dropper to obtain vaginal contents for examination under the microscope. The vaginal specimen was collected every 8 AM and 10 AM to observe for estrous cycle pattern following an interval of 3 to 4 days. Only animals with estrous phase pattern were included in the study and submitted to pilocarpine administration 52 .

Pilocarpine Epilepsy Model
The pilocarpine model of epilepsy represents a suitable paradigm to study pathophysiological mechanisms of temporal lobe epilepsy and has been used in numerous studies to the phenomenon 27 . Shortly, female Wistar rats in the estrus stage and males were pretreated with methyl scopolamine (Sigma Co.) at dose of 1 mg/kg subcutaneously to inhibit systemic cholinergic effects of pilocarpine 30 minutes prior 33 . Thereafter, pilocarpine hydrochloride (Sigma Co.) at a dose of 330 mg/kg was injected intraperitoneally to induce progressive sequential behavioral and electrographic changes leading to status epilepticus of long duration 22,33 . After ve hours of status epilepticus graded on the Racine scale, diazepam (Merck) at a dose of 10 mg/kg was given subcutaneously to limit behavioral limbic seizures 53 . The surviving animals (7 males and 9 females) were nursed with a special fractionated diet till recovery and transferred to the video lab for monitoring. Control male and female rats were injected with methyl scopolamine (1 mg/kg, s.c.) followed 30 minutes later by saline solution (1 ml/kg, i.p.) instead pilocarpine.

Seizure Video Monitoring
A high-de nition video system (Intelbras VT 4 S 120 HD) was used to continuously monitor the pilocarpine treated animals for recurrent seizure activity for 24h every day 43 . Limbic seizure behavior was monitored for three months after the rst spontaneous seizure (SRS) under the following parameters: number, duration, and severity of recurrent seizures. Animals surviving the pilocarpineinduced status epilepticus evolved into the period of spontaneous recurrent seizures of the epilepsy model. Seizure distribution in 24h was recorded for light (07:00-18:59h) and dark (19:00-06:59h) periods. All seizure parameters were analyzed according to the progressive scale proposed by Racine for the amygdala kindling 26 , as follows: 1. mouth and facial movements; 2. head nodding; 3. forelimb clonus; 4. forelimb clonus with rearing; 5. rearing and falling. Seizure cluster was considered if an animal had ≥2 seizures in 24 h 54 .

Isotropic fractionator method
This method is used to quantify the number of neuronal and non-neuronal cells in the brain or dissected structure and has been compared in several studies with unbiased stereological techniques, producing reliable and reproducible results faster. This technique does not need special software but follows the similar principles used in stereological analysis to produce about 100% of the number of cells in the brain 55 .
Brie y, at the end of the 3-month observation period, all animals from each group were thoroughly anesthetized with sodium thiopental (80 mg/kg, sc -Thiopentax) and perfused transcardially with phosphate-buffered solution (0.01 M -pH7.4), and thereafter, paraformaldehyde at 4% in PBS. Their brains were carefully removed and post xed in paraformaldehyde at 4% for 24 hours. Then brains were separated into the three structures of interest: olfactory bulb, amygdala and hippocampal formation using rat brain matrices (1 mm, Kent) xed on crushed ice, and consistent anatomical landmarks under a Leica stereomicroscope. Olfactory bulbs were separated by cutting away from the frontal poles. Hippocampal formation was carefully dissected from each hemisphere by detaching it from the striatum and cortical tissue. Amygdala were identi ed lateral to both hypothalamus and ventral portion of the rhinal sulcus, by cutting away 8 mm of loose cortex allowed isolation of amygdala consisting of basolateral and centromedial portions of the structure 56 . Each brain structure was mechanically dissociated in a saline solution with 1% Triton X-100 and turned into an isotropic suspension of isolated nuclei, kept homogeneous by agitation. The total number of cells were estimated by determining the number of nuclei in small aliquots stained with uorescent DNA marker 4′-6-diamidino-2-phenylindole dihydrochloride (DAPI, 1:1000; Sigma-Aldrich, D9542) utilizing a Zeiss Axiovert 100 microscope with a 40x objective, using a hemocytometer for quanti cation (Neubauer chamber, Loptik Labor). To determine neuronal and non-neuronal cells number, the samples were then incubated with the primary antibody against the neuron-speci c nuclear protein (NeuN, 1:100; Abcam, ab104225) at 4 °C overnight and, subsequently, with secondary antibody conjugated to AlexaFluor®488 (Abcam; ab150077) diluted in PBS (1:200) and 10% normal goat serum (Vector labs, S-1000-20) for 2 h. The neuronal fraction in each sample was estimated by counting NeuN-labeled nuclei in at least 500 DAPI-stained nuclei and the number of non-neuronal nuclei were obtained by subtraction 57 .

Statistical Analysis
Shapiro-Wilk test was done to ascertain data normality using Graph Pad Prism version 8.4.2 (Graph Pad Software, La Jolla California USA). Unpaired t-test was used to analyze for the duration of latent period in pilocarpine treated animals. Two-ANOVA (repeated measures) and Sidak's post-hoc test was performed for seizure dependent variables. Independent two-way ANOVA and Tukey's post-hoc test were carried out on dependent variables: mass, number of neurons and non-neuronal cell of brain structures. Data was presented as means and standard deviation with level of signi cance set at 0.05.

Declarations
Characterization of recurrent seizure patterns in both epileptic male and female rats submitted to 5h of status epilepticus (SE) and monitored continuously 24h for a period of 3 months after expression of the  Recurrent seizure distribution in 2h pattern for 24h recordings over a period 3 months during the light (07:00-18:59h) and dark (19:00-06:59h) cycle in both male and female epileptic rats. Data expressed as mean± standard deviation. *p < 0.05 for signi cance. Mass changes (in grams) in the total brain (A), hippocampal formation (B), amygdala (C) and olfactory bulb (D) of male and female epileptic Wistar rats in comparison to respective control groups.*p < 0.05 for signi cance.