Disease-Modifying Effects of a Glial-targeted Inducible Nitric Oxide Synthase Inhibitor (1400W) in Mixed-sex Cohorts of a Rat Soman (GD) Model of Epilepsy

Background Acute exposure to seizurogenic organophosphate (OP) nerve agents (OPNA) such as diisopropylfluorophosphate (DFP) or soman (GD), at high concentrations, induce immediate status epilepticus (SE), reactive gliosis, neurodegeneration, and epileptogenesis as a consequence. Medical countermeasures (MCMs-atropine, oximes, benzodiazepines), if administered in < 20 minutes of OPNA exposure, can control acute symptoms and mortality. However, MCMs alone are inadequate to prevent OPNA-induced brain injury and behavioral dysfunction in survivors. We have previously shown that OPNA exposure-induced SE increases the production of inducible nitric oxide synthase (iNOS) in glial cells in both short- and long-terms. Treating with a water soluble and highly selective iNOS inhibitor, 1400W, for three days significantly reduced OPNA-induced brain changes in those animals that had mild-moderate SE in the rat DFP model. However, such mitigating effects and the mechanisms of 1400W are unknown in a highly volatile nerve agent GD exposure. Methods Mixed-sex cohort of adult Sprague Dawley rats were exposed to GD (132μg/kg, s.c.) and immediately treated with atropine (2mg/kg, i.m) and HI-6 (125mg/kg, i.m.). Severity of seizures were quantified for an hour and treated with midazolam (3mg/kg, i.m.). An hour post-midazolam, 1400W (20mg/kg, i.m.) or vehicle was administered daily for two weeks. After behavioral testing and EEG acquisition, animals were euthanized at 3.5 months post-GD. Brains were processed for neuroinflammatory and neurodegeneration markers. Serum and CSF were used for nitrooxidative and proinflammatory cytokines assays. Results We demonstrate a significant long-term (3.5 months post-soman) disease-modifying effect of 1400W in animals that had severe SE for > 20min of continuous convulsive seizures. 1400W significantly reduced GD-induced motor and cognitive dysfunction; nitrooxidative stress (nitrite, ROS; increased GSH: GSSG); proinflammatory cytokines in the serum and some in the cerebrospinal fluid (CSF); epileptiform spikes and spontaneously recurring seizures (SRS) in males; reactive gliosis (GFAP + C3 and IBA1 + CD68 positive glia) as a measure of neuroinflammation, and neurodegeneration (including parvalbumin positive neurons) in some brain regions. Conclusion These findings demonstrate the long-term disease-modifying effects of a glial-targeted iNOS inhibitor, 1400W, in a rat GD model by modulating reactive gliosis, neurodegeneration, and neuronal hyperexcitability.

Choline acts on the muscarinic and nicotinic receptors and induces classic parasympathetic (cholinergic) toxicity symptoms such as salivation, lacrimation, urination, and defecation (SLUD) and seizures [3,4]. Medical countermeasures (MCMs), if administered immediately, control cholinergic symptoms and SE, to some extent, and prevent mortality. MCMs for OPNA toxicity include atropine sulfate, an oxime HI-6 dimethanesulfonate (DMS), and a benzodiazepine-derivative diazepam or midazolam. Atropine sulfate is a competitive, reversible antagonist of muscarinic receptors that prevents excessive glandular secretions, smooth muscle relaxation, and increases heart rate. HI-6 DMS is an asymmetric bis-pyridinium aldoxime and it is required for the reactivation of AChE before the enzyme ages. Therefore, both atropine and HI-6 should be administered immediately after an OPNA exposure. Experimental models suggest that diazepam or midazolam are effective in controlling behavioral SE and can reduce mortality if it is given in < 40 minutes of OPNA exposure [5][6][7][8][9]. Therefore, MCMs e cacy in OPNA toxicity is time sensitive.
Although MCMs control the acute symptoms of OPNA toxicity, they do not prevent SE-induced long-term brain injury [10][11][12][13]. Therefore, our hypothesis was that a combination of MCMs and a diseasemodifying agent as an adjunct therapy that targets brain pathology will prevent SE-induced long-term brain injury in the soman model. Our previous study demonstrated a signi cant increase in the inducible nitric oxide synthase (iNOS) in the brains of OPNA exposed adult rats, which persisted for long-term in glial cells [14]. Treatment with a highly selective iNOS inhibitor, 1400W hydrochloride, suppressed serum nitrite production and also reduced iNOS levels in the brain [14,15]. In addition, we found a signi cant increase in neuronal nitric oxide synthase (nNOS) post-SE, but unlike iNOS, it was only transiently upregulated [16,17]. Excessive production of NO is toxic [18,19]. Our previous studies in both rat kainate and DFP models of epilepsy demonstrated a signi cant disease-modifying effect when the animals were treated with 1400W for the rst three days of post-SE [14,20]. Although DFP is considered a surrogate for soman and structurally similar, the latter is a highly volatile agent. Like DFP, the onset of severe SE (convulsive seizures) in soman model is rapid and persistent [1,2]. Our recent ndings suggest that the initial SE severity determines the extent of brain injury in the OPNA model [21]. Therefore, rigorous SE quanti cation (behavioral and EEG based), grouping based on the SE severity, and randomization will minimize bias while testing the e cacy of disease-modifying drugs. We have thoroughly characterized and quanti ed SE severity in both DFP and soman models in adult male and female rats with or without telemetry devices [1,2]. In this study, we demonstrate signi cant disease-modifying effect of 1400W in both males and females when mixed-sex cohorts were treated for two weeks post MCM treatment following acute exposure to soman (GD). We demonstrate sex differences in EEG, behavioral outcomes, and histological changes (reactive gliosis and neurodegeneration) in the brain of animals exposed to soman and treated with 1400W or vehicle.
2. Methods 2.1.1 Animal source, care, and ethics statement The adult male and female Sprague-Dawley rats (7-8 weeks old; 250-300 g; Charles River, USA) were used in the study. Males and females were housed in the same room but in individual cages with a 12-h light: dark cycle in an enriched environment. The experiments were conducted after 2-3 days of acclimatization as per the approved protocols by the Institutional Animal Care and Use Committees (IACUC protocol: 20-090) and complied with the NIH ARRIVE Guidelines for the Care and Use of Laboratory Animals. Soman exposure was done at MRI Global, Kansas City, MO, and all other experiments were conducted at the Iowa State University. All animals were euthanized at the end of the study with pentobarbital sodium and phenytoin sodium (100 mg/kg, i.p.) as per the American Veterinary Medical Associations Guidelines for euthanasia.

Chemicals and reagents
The MRI Global, Kansas City, MO, United States acquired and administered the soman (> 95% pure by GC-MS) to the animals as per the approved IACUC protocols. Soman was diluted in cold 0.1M PBS just before administration. HI-6 (99.9% pure-by LC/MS, Kalexsyn, Kalamazo, MI, United States) and Atropine sulfate monohydrate (ATS, 99.9% pure by LC-MS, Tokyo Chemical Industry, USA) was diluted in saline at 250 mg/ml and 5 mg/ml, respectively. All key chemicals, except soman were validated by LC/GC-MS/MS, the chemicals purity and identity were authenticated before testing in animals. Midazolam (MDZ, prepared as 5 mg/ml stock solution) was supplied by the MRI Global. 1400W dihydrochloride (≥ 99% pure by HPLC, 98% by LC-MS, Tocris Bioscience, Bristol, UK) was diluted in distilled water at 20 mg/ml. Euthanasia solution (pentobarbital sodium and phenytoin sodium) was purchased from the Pharmacy at the ISU Lloyd Veterinary Medical Center. The antibodies used in this study are listed in the supplementary table 1.

Experimental Design
The experimental designs are illustrated in Fig. 1A and 4A.
2.2.2. Telemetry device implantation prior to soman exposure 36 rats (18/sex) were implanted with CTA-F40 telemetry devices (Data Science International, MN, USA) for integrated video-EEG acquisition. The surgical procedure was carried out about two weeks before challenging with soman. Four animals were excluded from the study due to transmitter device-related issues during the four months study. The devices were implanted as described in our previous publications [2,14,21]. The animals were administered an analgesic buprenorphine (0.3 mg/kg, s.c.) before anesthetizing with 3.0% iso urane (at 1litre/min O 2 ) and maintained at 1.0-1.5% during surgery.
SomnoFlo small animal anesthetic machine by Kent Scienti c (Torrington, CT, USA) was used for anesthesia. Arti cial tears ointment was applied to the eyes to protect the cornea. A midline-incision was made on the skin of the head and the underlying tissue was resected to expose the parietal bone. Electrodes were implanted bilaterally on the dura mater of the cortical hemisphere by drilling holes into the skull. The telemetry device was implanted in a subcutaneous pouch in the ank region. The electrodes were secured to the skull with a dental cement (A-M Systems, Carlsberg, WA, USA) and the wound was closed with sterile clips. A triple antibiotic ointment (Vetropolycin) was applied to the wound, Baytril (Enro oxacin, 5mg/kg, s.c., Bayer Pharmaceuticals, Pittsburgh, PA, USA) and 1ml of 5% dextrose saline were administered subcutaneously. After full recovery, the animals were individually housed in a cage on PhysioTel receiver pads (RPC-1) which were connected to the Data Exchange Matrix 2.0 (DSI) for continuous video-EEG acquisition with Ponemah Software (DSI). Baseline video-EEG was recorded for two day-night cycles before exposure to soman.

Soman exposure and medical countermeasures
36 rats with telemetry devices and 60 rats without telemetry devices were exposed to soman as described in our recent publication [1]. Rats were treated with vehicle (phosphate buffered saline; PBS) or soman (GD, 132 µg/kg, s.c. 1.2 x LD 50 ), followed by HI-6 (125 mg/kg, i.m.) and atropine sulfate (2 mg/kg, i.m.) within 1 min of soman exposure to control for peripheral cholinergic effects and mortality. The anticonvulsant midazolam (3 mg/kg, i.m.) was administered one-hour post soman exposure to control behavioral SE.

2.2.4
Behavioral and EEG-based seizures quanti cation, SE severity-based grouping and coding, and 1400W or the vehicle treatment Behavioral seizures were scored immediately after the animals were exposed to soman until the midazolam was administered, which is typically 60 minutes. Two experimenters (per animal) scored the seizures during SE directly. For telemetry animals, both behavioral and EEG-based seizure quanti cation were considered. Animals of both sexes were further grouped based on the similar SE severity and coded. The stages of seizures (stage 1 to stage 5) during SE in soman model, the stage-speci c symptoms, and SE severity quanti cation are described in our recent publication [1]. Brie y, stage 1-excessive salivation, lacrimation, urination and defecation (SLUD), mastication, chewing; stage 2-the stage 1 symptoms progressed-tremors, wet-dog shakes, head nodding/neck jerks, kyphosis, and opisthotonos; stage 3forelimb clonus, Straub tail, rearing and rigid extension of forelimbs; stage 4-repeated rearing, continuous forelimb clonus and loss of righting re ex; and stage 5-abducted limbs clonus and generalized seizures.
One group received 1400W (20 mg/kg, i.m) and the other group received the vehicle (distilled water, i.m).
1400W or the vehicle was given twice daily for the rst three days followed by a single injection per day for the next 11 days. The vehicle group received equal number and volumes of vehicle injections (i.m.). Animals that had < 20 minutes of convulsive seizures (stages [3][4][5] were considered under mild SE group.

Epileptiform spike rate and SRS quanti cation
Baseline EEG prior to soman exposure was used to normalize the post-soman EEG traces for accurate seizure and epileptiform spikes detection. Artifacts from electrical noise, grooming and exploratory behavior were identi ed and excluded from the epileptiform spike analysis as described in our publications [2,20,21]. We used NeuroScore version 3.4.0 software (DSI) for seizure detection and epileptiform spike quanti cation. Convulsive spontaneous recurrent seizures (SRS) were manually marked and staged. Epileptiform spikes were automatically calculated by NeuroScore by baseline amplitude subtraction from a cleaned EEG trace after excluding the artefacts. All EEG events were cross veri ed with the integrated video (behavioral seizure) and the power spectral changes. The epileptiform spikes and seizure reports were generated for each animal for graphing and statistical analyses.

Behavioral testing
Four groups from the non-telemetry cohort were used for all behavioral tests: two control groups (control and vehicle + 1400W) and two soman exposed groups (soman + vehicle and soman + 1400W). All experiments were conducted by the same experimenters and started at the same time each day in a distraction free environment. The room temperature and lighting were kept the same for all animals and for all experiments. All animals were handled gently to minimize stress. The animals from all four groups were randomized irrespective of sex and treatment. Equipment was thoroughly cleaned with 70% ethanol to remove urine or feces smell from the previous animal and dried before a new animal was introduced. Animals were given 2-3 days of rest between each new behavioral test.

Accelerated rotarod test
Animals were tested for motor learning and motor coordination using an accelerating rotarod apparatus (AccuRotor 4-channel, Omnitech Electronics Inc. and Start Fusion v6.2 AccuRotor Edition Software). We used a modi ed protocol from our previous publications [14,21]. The acceleration speed was set from 5 rpm to 60 rpm with a duration of 180s. The test was conducted for three days with two days of training and a day of testing with a 24-hour rest between training and testing. Evaluation of motor coordination and learning was done in three trials on both training and testing days, and the mean time spent by the rats on the rotating rod (latency to fall) was measured and compared between the groups and sex.

Novel Object Recognition (NOR) and Open Field Tests
The NOR test is a behavioral assay to measure learning and memory in rodents. The Open Field test, using the same apparatus, measures anxiety-like behavior. The test was conducted for 3-days: habituation day/open eld test, testing day, and probing day. A low-wattage red light was used to illuminate the arena [22]. On the habituation day/open eld test, animals were allowed to explore the arena (40 X 40 inches) for 10 minutes with no objects. The time spent at the center versus periphery of the arena was used to interpret the Open Field test data. On the day of testing, two similar objects (Stoelting Co., USA) were placed diagonally in the arena and the rat was released in the corner opposite to that of the objects and left to explore both objects for 5 minutes. Three hours or twenty-four hours after testing (probing day), one of the two objects were replaced with a novel object and the rat was allowed to explore the objects for 5 minutes. The movement of the rat was tracked by the camera operated by ANY-Maze software (Stoelting Co., USA). The time spent by the rat in the center versus periphery during the habituation phase and between the novel versus the familiar objects (discrimination index, DI= (t novelt familiar )/(t novel +t familiar )) were analyzed and compared between the groups and sexes.

Elevated Zero/Circular Maze Test
The elevated zero/circular maze test assesses anxiety-related behavior. In contrast to the open eld, the elevated circular platform (47-inch diameter) was set at a height of 20 inches off the ground. The circular platform (4-inch-wide) was divided into two enclosed arenas, with 12-inch-high walls, opposite to each other and two open arenas between the enclosed arenas. The rats were placed in at the junction of open and closed arenas and allowed to explore the maze for 5 min. Movement of the rat was monitored remotely via an overhead camera operated by ANY-Maze software (Stoelting Co., USA) connected to a computer in a separate room. The percentage time spent by the rat in open versus closed arms between each group and sex were quanti ed and analyzed.

Contextual and Cued Fear Conditioning test
The rats were tested for contextual and cued fear conditioning to assess their ability to recall an association between a conditioned stimulus (a tone) and an aversive unconditioned stimulus (an electric foot shock). The test consisted of conditioning (day 1) and probing (day 2). On the day of conditioning, the rats were allowed to explore the apparatus for 2 minutes, followed by an auditory tone (70-80 dB) for 20 seconds. At the last 2 seconds of the auditory tone, a 0.7 mA foot shock was delivered continuously. The conditioning paradigm was repeated four times per session with an inter-trial interval of 80 seconds, followed by a rest for 80 seconds. The rats were probed 24 hours after the conditioning session. The probing protocol was the same as conditioning test but without a foot shock. The movement of the rat was tracked by the camera operated by ANY-maze software (Stoelting Co., USA). The freezing time during conditioning and probing were recorded and analyzed and compared between the groups and sexes. The freezing time (%) increase was calculated by averaging the freezing times (%) from all four intertrial intervals (ITIs) and subtracting these from the freezing time (%) during habituation phase of the probe test.

Euthanasia and tissue collection
All animals were euthanized with a lethal dose of pentobarbital sodium and phenytoin sodium (100 mg/kg, i.p.) at the end of the study. Some animals were perfuse-xed with 4% paraformaldehyde (PFA) and the brains were isolated and processed for immunohistochemistry (IHC). The blood and CSF were collected for nitrooxidative markers and cytokine assays. The blood was centrifuged at 1,000-1,500 x g for 10 minutes to separate serum. Both serum and CSF were aliquoted and stored at -80 0 C until further analysis.

Tissue processing, immunohistochemistry, imaging, and cell quanti cation
After perfusion, the brains were isolated and stored in 4% PFA for 24 h and transferred to 25% sucrose in PBS for at least 3 days at 4 0 C. The brains were then embedded in gelatin (15% Porcine gelatin Type A, 7.5% sucrose, and 0.1% sodium azide in PBS, Sigma, MO, USA) at 37 0 C for 3 h and stored overnight at 4 0 C. The gelatin blocks were then rapidly frozen in liquid nitrogen-cooled 2-methyl butane. The brains were cut on a cryostat at 16 µm thickness from rostral to caudal using a cryostat (ThermoFisher) as described in our previous publication [20]. Five sections per slide were collected on chrome alum gelatin (Pfalz and Bauer, CT, USA) coated slides and stored at -20 0 C until they were processed for IHC. Each slide contained representative sections from rostral to the caudal end of the brain. The brain sections were subjected to antigen retrieval by treating with citrate buffer (10 mM citric acid and 0.05% tween-20, pH 6.0) at 95 0 C for 20 min. After cooling, the slides were covered with cover plates and transferred to Shandon racks for IHC. The sections were washed with PBS 3 times and incubated in blocking buffer for an hour (10% donkey serum, 0.05% TritonX-100 in PBS) at room temperature. The sections were incubated with primary antibodies (GFAP-mouse/IBA1-Goat/NeuN-rabbit/C3-species/CD68species/PVB-species) overnight at 4 0 C. The next day, slides were washed 3 times in PBS and incubated with appropriate secondary antibodies (AlexaFluor 488 anti-mouse/ Rhodamine Red X anti-goat/ Rhodamine Red X anti-rabbit/ Biotinylated anti-rat) for an hour at room temperature. The dilutions for antibodies, the source, and the species in which the antibodies were raised are included in the supplementary table 1. Slides were again washed 3 times in PBS, rinsed in water to remove salts, and mounted with VectaShield® mounting medium containing DAPI nuclear stain (Vector Laboratories Inc., CA, USA). The slides were stored at -4 0 C until imaging.
To investigate the extent of neurodegeneration, the brain sections were processed for NeuN-Fluro Jade B (FJB) dual staining. The sections were rst stained for NeuN using the procedure described above. The sections were then washed with PBS three times and distilled water before immersing in freshly prepared 0.006% potassium permanganate for 5 min. The slides were washed in distilled water two times for two minutes to remove potassium permanganate residues. The slides were incubated in freshly prepared 0.0003% FJB solution in 0.1% acetic acid for 10 min in the dark, followed by three washes in distilled water for one minute. The slides were air dried overnight in the dark and mounted with Surgipath acrytol (Leica Biosystems, IL, USA).
The sections were imaged using the Leica DMi8 inverted uorescence microscope (Wetzlar, Germany) tted with Leica K5 passive cooled sCMOS camera system. Images from 3 coronal sections per brain from each group, representing the same regions of the brain from rostral to caudal were used for cell quanti cation. IBA1, GFAP, FJB positive cells and C3 + GFAP colocalized cells were quanti ed manually using the multipoint tool in ImageJ software as described in our previous publication [14,16]. Reactive astrocytes and microglia were also quanti ed manually. The morphology-based parameters were large cell bodies, short and thick processes in reactive phenotypes in contrast to a small cell body, thin and long processes in homeostatic -state controls [23][24][25][26]. CD68 + IBA1 colocalized foci were quanti ed using the analyze particles function in ImageJ.

Diaminobenzidine (DAB)-NeuN staining and stereology
The DAB-NeuN immunostaining was performed to identify the neurons in the hilus of the DG, CA1, and AMY + PC regions for stereological counting. After processing the brain sections for NeuN primary and secondary biotinylated antibodies, the slides were washed with PBS (3x). The horse radish peroxidaseavidin solution (HRP, Vectastain) was prepared fresh and the sections were incubated at RT for an hour. After PBS washing (3x), the slides were placed in a DAB solution (25 mg DAB, 50 ml PBS, and 250 µl hydrogen peroxide) until desired stain intensity was observed. The slides were washed in tap water for 10 min, air-dried, and mounted with acrytol. A Zeiss Axio Imager 2 light microscope was used to image the sections and neuronal counts were performed using the Micro Bright Field Bioscience Stereo Investigator v9 software (Williston, VT, USA). The neurons were quanti ed at a 40x objective with the following parameters-counting frame 70 X 70, SRS grid 150 X 150, and de ned detector 0-7-0.

Cytokines and chemokine assays
We used RayBiotech® ELISA kits (Norcross, GA, USA) for key proin ammatory cytokine and chemokine assays (IL1-β, IL-6, MCP-1, TNFα). Pilot testing of previously con rmed positive samples reassured the quality of the kits. The serum was used without dilution, but the CSF was used at 1:2 dilution. The assay was conducted according to the manufacturer's protocol in duplicates along with the positive and negative controls. Using the standard reagents, the standard curves for all the cytokines were obtained which were used to determine the concentrations of cytokines/chemokine in the test samples with the absorbance read at 450nm and expressed as pg/mL.

Glutathione assay
The GSH and GSSG (oxidized GSH) were determined in the serum using a ThermoScienti c® Glutathione Colorimetric Detection Kit (Carlsbad, CA, USA). The assay was performed according to the manufacturer's protocol. GSH and GSSG assays were done in duplicates and previously known positive and negative controls were used simultaneously alongside the test samples. To measure oxidized glutathione (GSSG), 2-vinylpyridine (2VP) treated standards and samples were used. The concentration of GSSG was half of the GSH concentration read from the curve (i.e., 1 GSSG = 2 GSH). Free/reduced glutathione (GSH) concentrations were determined by subtracting the GSSG concentrations from the values obtained from the samples and standards that were treated with 2VP. Absorbance was read at 405nm. Concentrations were expressed as µM.

Nitrite assay
The serum nitrite assay was determined using a Sigma-Aldrich® Nitrite Assay or Griess Reagent Kit (St. Louis, MO, USA) according to the manufacturer's protocol. The assay was done in duplicates and previously known positive and negative controls were included simultaneously. A standard curve was prepared using the standard reagents. The standard curve was used to derive the concentrations of nitrite in the test samples with the absorbance read at 540nm and the nitrite concentrations were expressed as µM.

ROS Assay
We used OxiSelect™ Intracellular ROS Assay Kit (Green Fluorescence) to determine ROS levels in the serum. The assay was performed as per the manufacturer's protocol. The ROS content in unknown samples was analyzed by comparing with the predetermined DCF standard curve with the absorbance read at 480nm excitation/530nm emission. The ROS concentration was expressed in relative uorescence units (RFUs).

Statistical analysis
We used GraphPad Prism 9.0 for statistical analysis. Normality tests were performed using the Shapiro-Wilk test, where applicable. Based on the normality of the data, t-test or Mann-Whitney test was applied for comparing data between two groups. One-way ANOVA with appropriate post-hoc test was used to compare multiple groups with one factor, while sex-differences were tested using the two-way ANOVA to detect signi cant interactions between sex and treatment effects [27] with appropriate post-hoc test for multiple comparisons across groups. If the interaction effects were found to be signi cant, then the data was presented as separate sex cohorts with relevant post-hoc test for multiple comparisons. A mixedeffects model or repeated measures two-way ANOVA was performed to determine overall main treatment effects when comparing across different brain regions in immunohistochemistry panels. Further statistical details are included in the gure legends of the corresponding gures. The p values of the region-wise two-way ANOVA interaction tests for sex differences are listed in supplementary table 2.

Rigor, sample size, and inclusion or exclusion criteria
All animals used in this study were purchased from the same source (Charles River, USA). Animals were rst randomized ignoring sex, stage of the estrous cycle, and bodyweight. Age and sex matched controls and the soman exposed groups treated with vehicle or 1400W were blinded. The sample size and power were determined based on our previous OPNA studies outcome from gliosis and neurodegeneration parameters [21] at 95% con dence interval.

Results
3.1. Comparison of the initial SE severity in soman exposed animals and the change in bodyweights during the treatment period: Vehicle vs. 1400W groups Experimental design is illustrated in Fig. 1A. The soman-exposed animals were divided into two groups and the initial SE severity was compared before assigning the animals for vehicle or 1400W treatment. There were no signi cant differences in the SE severity between the vehicle and 1400W groups when they were analyzed as mixed-sex cohorts or as male or female groups independently (Fig. 1B). SE severity is the duration of convulsive seizures during the 60 minutes between soman exposure and midazolam injection. All soman exposed animals, irrespective of sex or treated with vehicle or 1400W, signi cantly lost weight in the rst 3-4 days of post-exposure, while there was no weight loss in both control groups (control; Vehicle + 1400W) (Fig. 1C). However, soman exposed animals irrespective of sex, recovered their bodyweight and showed a similar trajectory as their respective control animals (Fig. 1C).
Acute exposure to soman in the long-term caused a signi cant reduction in accelerated motor coordination in the vehicle-treated group, compared to controls, and 1400W mitigated the effect ( Fig. 2A).
When mild versus severe SE groups were compared, the soman and 1400W effects were observed in the severe SE groups (Fig. 2B). No sex interaction was detected.

1400W mitigated the soman-induced long-term memory de cit
A schematic diagram illustrating the experimental design for the Novel Object Recognition test (Fig. 2C) and representative heatmaps are shown in Fig. 2D. There were no signi cant differences in memory retention in any of the four groups when the animals were tested 3h post-familiarization of the objects (Fig. 2E). However, when tested 24h post-familiarization, the soman vehicle group showed a signi cant reduction in the discrimination index which was mitigated by 1400W (Fig. 2F). The signi cant differences were observed in those animals that had severe SE after soman exposure (Fig. 2G) and no sex differences were observed.

1400W mitigated the soman-induced fear extinctionlike behavior
We have previously shown that acute exposure to DFP, an analog of soman, causes injury to the amygdala [21], the part of the brain involved in regulating fear/anxiety-like behavior. In the Elevated Zero Maze test (Fig. 3A), the control and soman exposed 1400W-treated groups spent signi cantly less time in open arms compared to closed arms, while the soman exposed vehicle treated group spent relatively more time in open arms (Fig. 3B, 3C). However, in the Open Field test (Fig. 2D), soman exposed groups, irrespective of vehicle or 1400W, in contrast to their respective controls, spent more time at the center of the open eld chamber or less time in the periphery (Fig. 2D, 2E). No sex interaction was observed in either Elevated Zero Maze or Open Filed test.

Impact of soman and 1400W on Contextual Fear Conditioning
When the animals were placed in the fear conditioning chamber (habituation), soman exposed animals irrespective of vehicle or 1400W treated froze signi cantly longer than their respective control groups (Fig. 3F, 3G). Interestingly, after habituation, all four groups responded similarly to the conditioning test that involved a tone for 20 seconds followed by foot shock for 2 seconds at the end of the tone (Fig. 3F).
During the probe test habituation, 24h after conditioning, soman exposed animals also froze signi cantly longer than the controls (Fig. 3G). However, during the rst inter-trial interval (i.e., between the introduction of the rst tone without foot shock and the second tone), control animals froze for longer time than the soman exposed groups (Fig. 3G). The freezing time decreased in subsequent inter-trial intervals in both control groups, more so in the control than the vehicle + 1400W group. Amongst the soman exposed groups, the freezing time decreased after the last trial (Fig. 3G). Overall, soman exposed animals showed a signi cant reduction in freezing time increase than their respective controls suggesting the compromised fear pathway (Fig. 3H). No sex interaction was observed.

1400W signi cantly reduced soman-induced epileptiform spiking and SRS in males
The experimental design for telemetry device implanted animals is illustrated in Fig. 4A. Representative showing the frequency of SRS in the soman exposed animals, treated with vehicle or 1400W, during the four months is illustrated in Fig. 5B. The EEG analysis revealed a signi cant and progressive increase in epileptiform spikes and SRS in soman exposed vehicle-treated males and 1400W signi cantly reduced the epileptiform spikes and SRS in males but not in females (Fig. 5D, 5F). Interestingly, in females, 1400W increased the epileptiform spikes during the last 4 weeks of the acquisition period, however, the differences were not statistically signi cant (Fig. 5E).
3.3.6. 1400W signi cantly reduced the soman-induced oxidative and nitrooxidative stress Serum samples from 3.5 months post-soman animals, treated with or without 1400W, and age matched vehicle control animals were used for nitrite, ROS, and glutathione assays. Soman exposure caused a signi cant increase in serum nitrite and ROS concentrations and 1400W signi cantly reduced the soman-induced effects (Fig. 6A, 6B). The ratio of reduced and oxidized glutathione (GSH: GSSG) was signi cantly reduced in the soman exposed vehicle-treated group and 1400W mitigated the somaninduced effects (Fig. 6C).

1400W signi cantly reduced soman-induced proin ammatory cytokines in the serum and CSF
The serum and CSF samples were collected under terminal anesthesia, before the animals were perfused for brain histology, and used for the assay. Soman exposure signi cantly increased IL-1β, TNFα, and MCP1 in both serum and CSF, and 1400W signi cantly reduced soman-induced effects (Fig. 6D, 6E).
Similar effects were observed in IL-6 levels in the serum (Fig. 6D).
3.3.8. 1400W reduced the soman-induced reactive gliosis in several brain regions There was an overall increase in microgliosis in the extrahippocampal and the thalamic nuclei in soman exposed animals (Fig. 7A, 7C, 7D). Interestingly, when region-wise comparison was made for each sex, the increase in the IBA1 positive cells were observed in both AMY and PC in females, but not in males ( Fig. 7E, 7F). When the reactive microglia with large cell bodies and thick cytoplasmic processes (Fig. 7B) were counted, there was an overall effect, i.e., signi cant increase in reactive microglia in soman exposed vehicle group (Fig. 7G). When regional differences were analyzed, signi cant differences were observed in all regions (Fig. 7H). In females, the signi cant differences between soman exposed vehicle and 1400W groups was observed in the amygdala and all three thalamic regions (LDT, MDT, and CMT) (Fig. 7J), but not in males (Fig. 7I). Reactive microglia more frequently contain CD68 [28,29]. Therefore, we further analyzed CD68 colocalized IBA1 cells (Fig. 7L). 1400W signi cantly reduced soman-induced effects in AMY and PC, while it was not signi cantly reduced in the thalamic nuclei (Fig. 7M). Interestingly, no sex difference was observed in CD68 + IBA1 cells in these regions.
Representative GFAP positive astrocytes in piriform cortex from each treatment group are presented in Fig. 8A. There were no signi cant differences in astrogliosis per se in soman exposed vehicle or 1400W treated groups in mixed-sex cohorts when analyzed using the repeated measures two-way ANOVA (overall treatment effect) or in the region-wise two-way ANOVA (Fig. 8C, 8D). There were no signi cant sex interactions in any region. However, when the reactiveastrocytes with thick cytoplasmic process and large cell bodies with intense GFAP stained cells (Fig. 8B) were compared, there was a signi cant increase in the soman + vehicle group across all brain regions and 1400W signi cantly reduced the soman-induced effects (Fig. 8E, 8F). Sex interaction was observed only in CMT region (Fig. 8G). Since reactive astrocytes upregulate complement 3 (C3) [30, 31], we investigated C3 positive GFAP cells (Fig. 8H-I). There was a signi cant increase in such cells in soman + vehicle group across all regions, and 1400W signi cantly reduced C3 colocalized astrocytes in CA1, subiculum, AMY, MDT and CMT regions (Fig. 8J). There was no sex interaction.
3.3.9. Soman exposure caused signi cant neurodegeneration and 1400W reduced the effect in some brain regions Soman exposure has been known to induce long-lasting neurodegeneration in various parts of the brain [1,32,33]. Representative images of AMY and MDT are shown in Fig. 9A. FJB positive NeuN cells were signi cantly increased in the soman-exposed vehicle-treated group (Fig. 9B). Regional analysis indicated a signi cant increase in all regions except dentate gyrus and CA1 (Fig. 9C). Sex interaction was observed in subiculum, AMY, MDT and CMT regions ( Fig. 9D-E). 1400W signi cantly reduced the soman-induced neurodegeneration in MDT in females and CMT in males (Fig. 9E). We further quanti ed the absolute number of neurons in the hilus of the dentate gyrus, CA1, AMY and PC. Low magni cation (4x) of whole half brain images revealed enlarged lateral ventricle in vehicle-treated soman exposed animal compared to 1400W-treated and control groups (Fig. 10A). Representative histological images are shown in Fig. 10A and 10B. We observed a signi cant reduction in the number of neurons in the hilus of the dentate gyrus, CA1 hippocampus, amygdala, and the piriform cortex in the soman exposed animals (Fig. 10D). In the hilus, there was ~ 55% reduction in neurons in soman vehicle group while in the 1400W group there was 35% loss compared to the control (Fig. 10D). In CA1, there was ~ 50% reduction in the number of NeuN positive neurons while in the 1400W-treated group, there was ~ 25% reduction compared to control ( Fig. 10D). In the amygdala and piriform cortex, there was > 50% neuronal loss in soman exposed animals and the 1400W had no effect compared to control (Fig. 10D). We further analyzed a subset of inhibitory neurons, the parvalbumin (PVB) positive neurons in the most vulnerable brain region, the amygdala (Fig. 10C). 1400W signi cantly reduced soman-induced reduction of PVB neurons (Fig. 10E).
No sex interaction was observed in stereological counts or in PVB neurons.

Discussion
Chemoconvulsant-induced SE increases the production of reactive oxygen and nitrogen species in the brain [34][35][36]. The role of nitrooxidative stress in SE-induced epilepsy is emerging as a potential target for disease modi cation [14,20,37,38]. Our previous short-and long-term studies in the rat DFP model revealed a signi cant upregulation of iNOS in the hippocampus and piriform cortex at 48 hours and 7 days post-SE [14]. Immunohistochemistry of brain sections at 7d and 12 weeks post-SE con rmed the increased iNOS expression, predominantly in microglia, while nitrosylated proteins (3-NT) were observed in neurons in both the hippocampus and piriform cortex [14]. Treating with a highly selective iNOS inhibitor, 1400W, twice a day for the rst three days of DFP exposure signi cantly reduced DFP-induced changes in the brain including the spontaneously recurring seizures in the three months study [14].
However, in this study, animals with severe SE were less responsive to the 1400W treatment in the DFPinduced behavioral dysfunction implying the dosing regimen used in the previous study was inadequate to mitigate behavioral dysfunction across all severity groups. Furthermore, the previous DFP study did not address sex differences. In the current soman study, we demonstrate the differential effects of soman and 1400W in the brains of both sexes and its impact on behavioral, EEG, and brain histological outcomes. We also demonstrate the long-term changes in the peripheral biomarkers of nitrooxidative stress and proin ammatory cytokines in the serum and CSF in response to acute exposure to soman and the effects of two weeks of 1400W treatment after soman exposure.
As per the NIH-NINDS guidelines for rigor [39], we designed and conducted experiments in age-matched mixed-sex cohorts. In SE-induced epilepsy studies, the initial seizure severity is a critical variable [1,2,21,40]. Therefore, based on our previous studies in different models of epilepsy, a reliable and reproducible SE quanti cation method was employed in this study. Irrespective of estrous stages in females at the time of exposure to soman, all females had SE severity for > 30min while most males had > 20 min of continuous convulsive seizures. Based on the SE severity, the animals were grouped, blinded, and treated with vehicle or 1400W.
SE-induced long-term effects on the brain structure and functions in experimental models of epilepsy are well known [21, 33,41]. Soman-induced long-term effects, especially the changes in the hippocampus and amygdala, have been demonstrated in rat models [33,[42][43][44][45]. Some interventional studies such as targeting the GluK1/AMPA receptors with LY293558, or muscarinic receptors with caramiphen, have demonstrated the mitigating effects in soman-induced brain pathology and dysfunction in a 4-week study [46,47]. A combination of antiseizure medications such as phenobarbital, ketamine, and midazolam demonstrated a reduction in soman-induced epileptogenesis and brain pathology in a rat model in a two-week study [48,49]. In another study, treating guinea pigs with galantamine, a glutamate antagonist, signi cantly reduced soman-induced lethality and mitigated hippocampal dependent memory de cits [50]. In all these models, atropine, an anticholinergic and an oxime (HI-6 and/or 2-PAM) were administered immediately after soman exposure to control for mortality. Atropine has a narrow window of action, therefore, should be administered no later than 20 min of exposure. The oximes do not effectively penetrate the brain to protect it from SE-induced effects [9,51]. In some of these models, pyridostigmine, a reversible AChE inhibitor, was administered before exposure to soman to control mortality, which is not feasible in real-world scenarios of nerve agent exposure. Therefore, in our soman model, animals were not pretreated with any medical countermeasures [1]. In addition to the ion channels-targeted drugs, our recent studies demonstrate the role of nitrooxidative stress and iNOS as a potential target for disease modi cation in DFP-induced models of epilepsy [14,52]. Antioxidant drug therapy as a neuroprotective countermeasure for nerve agent toxicity has been proposed in another study [53].
Soman-induced SE in experimental models, depending on the severity, leads to the development of epilepsy and behavioral dysfunction [43,[54][55][56]. In soman exposed animals in this study, we demonstrated signi cant locomotor ( Fig. 2A) and long-term memory de cits (Fig. 2F) and compromised anxiety/fear-like behavior (Fig. 3B, 2C), which was reduced by 1400W implying the role of iNOS/nitrooxidative stress-mediated pathway. Interestingly, no sex interaction was observed in any of the behavioral studies. The role of nitrergic stress in neurodegenerative diseases is well known [35,[57][58][59]. NO produced in excess following stress can nitrosylate proteins' cysteine residues that can lead to protein dysfunction and facilitate disease progression [60-62]. The neuroin ammatory role of NO signaling that promotes glycation and neuronal dysfunction has been demonstrated in a mouse prion disease model [63]. We demonstrated a signi cant upregulation of both iNOS and 3-NT and glutathiolated proteins in the brains of animals exposed to DFP, a surrogate for soman [14,52]. Therefore, it is likely that treating with a highly selective iNOS inhibitor, 1400W, mitigated the soman-induced motor and memory dysfunction in this study.
The role of amygdala, the bed nucleus of stria terminalis (BNST), the ventral hippocampus ("anxiety cells") and their projections to hypothalamus and brain stem mediate fear memory and anxiety-like behaviors [64-67]. The amygdala is highly susceptible to OPNA exposure and causes loss of inhibitory neurons and seizures, which, in the long-term, induces neurodegeneration and the development of anxiety-like behavior in rodent models [8,46,[68][69][70][71]. The occurrence of SRS in soman model in this study could be due to the loss of inhibitory neurons in the amygdala and its consequence of hyperexcitability [44,[72][73][74]. Stereological cell count con rmed a signi cant reduction in neuronal population in amygdala and the piriform cortex in the soman exposed animals (Fig. 10D). 1400W treatment did not completely prevent neuronal loss in these regions. However, when parvalbumin (PVB) positive neurons were counted from the amygdala, we found a signi cant protection by 1400W (Fig. 10E). We observed a signi cant reduction of compromized anxiety-like behavior in the Elevated zero/circular Maze test compared to the vehicle treated group in the soman exposed animals (Fig. 3B-E) implying the survival of certain subsets of inhibitory neurons, such as PVB, in the amygdala by 1400W may have mitigated the soman-induced effects. However, the impact of soman on BNST and its projection bers to the other brain regions are not investigated in this study.
Freezing behavior is a physiological response to a stimulus and a measure of fear in experimental models [56, 75,76]. Injury to amygdala in rodent model prevented fear-potentiated freezing/startle response to foot-shock when paired with exposure to bright light or a tone [77][78][79]. The contextual fear conditioning response reveals functional network connectivity between the hippocampus (contextual recollection), amygdala (fear/anxiety-like behavior), BNST, brain stem and hypothalamus [65,67]. If the tone is introduced as a context followed by foot shock, as in this study, the auditory pathway will be activated upon the introduction of the tone in a normal individual. Injury to any of these structures will yield a different response which is measured as the increased duration of freezing following a tone as in this study. Initial SE severity determines the extent of injury to various brain regions. For example, hippocampus is less likely be affected compared to the basolateral amygdala and the piriform cortex if initial SE persists for ~ 20min after acute exposure in OPNA model. If SE severity is > 20 minutes, widespread brain pathology occurs [14,21]. In this study, SE severity was > 20 min. Therefore, it was anticipated that the animals exposed to soman and 1400W would behave differently. All animals responded similarly to the tone and the foot shock during the conditioning phase (Fig. 3F) suggesting the response pathway to noxious stimulus (fear learning) was unaffected in soman exposed animals. However, the obvious differences between groups emerged when they were probed 24h post-conditioning without foot shock implying the disruption in the connectivity between amygdala and the other brain regions. The freezing time increase in control groups was higher in response to a tone, without shock, and it signi cantly decreased in soman exposed animals suggesting the disrupted brain connectivity after exposure to soman. In an another soman exposure study male rats revealed impaired auditory and contextual fear conditioning due to neurodegeneration in the piriform cortex, amygdala, thalamus, and the hippocampus [56]. Our present study con rmed pathology in some of these brain regions. 1400W reduced neurodegeneration in the centromedial (in males) and mediodorsal (in females) thalamic nuclei, and amygdala (increased PVB neurons) (Fig. 9D, 9E, and 10E). However, we did not nd sex differences in behavioral outcomes and in the effects of 1400W on PVB neurons.
Loss of inhibitory neurons or functional connectivity due to soman exposure is known to cause hyperexcitability in the network and onset of seizures [43,44]. The epileptiform spiking and SRS frequency in soman exposed vehicle-treated males was signi cantly reduced by 1400W in soman exposed animals suggesting the disease-modifying potential of the iNOS inhibitor. We have previously demonstrated similar disease-modifying effects of 1400W in both DFP and KA rat models of chronic epilepsy [14,20]. In a rat neuropathic pain model, 1400W signi cantly reduced pain sensitivity [80]. In an in vitro brain slice model too, pre-treating with 1400W signi cantly reduced KA-induced hyperexcitability [81]. In contrast, in females exposed to soman, no signi cant effects of 1400W on EEG parameters were observed in this study. This could be due to the secondary effects of the subcutaneously implanted transmitter devices that caused skin ulcers (bed sores) in chronic studies in females due to the thin skin.
This resulted in repeated repositioning of the transmitter device under brief general anesthesia (iso urane). Repeated iso urane exposures in the rat model impaired long-term potentiation (LTP) at both 1 day and 1 week after the last exposure, which lasted for about a month [82]. In both KA and paraoxon rat model, SE induction in the presence of iso urane did not affect the severity of SE but impacted epileptogenesis [83]. In this study, since female rats were exposed more often to iso urane than males, there was an overall reduction in SRS frequency in both vehicle and 1400W treated groups (Fig. 5G), therefore, iso urane exposure may have confounded the SRS frequency and the real effect of 1400W in females.
Nitrooxidative stress and proin ammatory cytokines and chemokines production are well known mediators of epileptogenesis that follows chemoconvulsant-induced SE [14,34,53,84,85]. Neuronal hyperexcitability activates glial cells [35], which produce ROS/RNS and cytokines and mediate neurodegeneration if not controlled at an early stage of SE induction. Several studies have demonstrated that intervening RNS/ROS and/or cytokines using antioxidants or anti-in ammatory agents such as NOS inhibitors, Src kinase inhibitors or cyclin D kinase inhibitors or the drugs that transiently suppress microglial activation can modify epileptogenesis [14,20,[85][86][87]. In OPNA toxicity models, we also demonstrated the roles of two RNS/ROS generating enzymes, the iNOS and the NADPH oxidase 2 (NOX2). These enzymes and the nitrosylated (3NT) or glutathiolated (anti-GSH) proteins were signi cantly upregulated in the rat brain following exposure to an OPNA. Additionally, we observed a signi cant increase in reactive gliosis (C3 positive astrocytes and CD68 positive microglia) and neurodegeneration in several brain regions in DFP or soman exposed animals [1,14,21]. Reactive glial cells are known to produce proin ammatory cytokines and iNOS [14,88,89]. Treating with an iNOS inhibitor (1400W) or a NOX2 inhibitor (Diapocynin) mitigated the DFP-induced nitrooxidative stress, neuroin ammation, neurodegeneration, and neuronal hyperexcitability [14,52]. These ndings suggest that the antioxidant mechanism is a potential therapeutic target for disease-modi cation in OPNA models.
It is interesting to note that there are regional variation and sex differences in reactive microglia that contained CD68 and reactive astrocytes with C3, which may have had impacted the behavioral outcome.
Signi cant reduction of reactive microglia by 1400W in soman exposed group was observed in amygdala and thalamus of females but not males. However, no sex interaction was observed in the suppression of CD68 positive reactive microglia by 1400W in these regions. Reactive microglia more frequently contain CD68 [28,29]. CD68 is a lysosomal transmembrane gycoprotein expressed in high levels in phagocytic microglia. In contrast, reactive astrocytes express C3 [30,31]. In the brain, C3 is a mediator of innate immunity and communicate with microglia for synaptic pruning during development [90][91][92]. In injured brain, reactive astrocytes release C3, which we found in the extra cellular spaces, and could reactive microglia, which express receptors for C3 [31]. Although we did not investigate the extent of C3a receptors expression in microglia, increased expression of CD68 in vulnerable brain regions such as AMY, PC, and thalamic nuclei suggest the ongoing phagocytosis. Furthermore, in these regions there was a signi cant increase in neurodegeneration, especially the parvalbumin positive inhibitory neurons in amygdala, in soman exposed animals that was reduced by 1400W. Loss of such inhibitory neurons, reactive gliosis and a concurrent increase in cytokines and nitrooxidative stress may have caused neuronal hyperexcitability and spontaneously recurring seizures in soman exposed animals. Treating with 1400W reduced soman-induced effects.
This study encompasses a rigorous investigation of sex differences across multiple analyses by incorporating a factorial design of both sex and treatment effects [27]. We examined the interaction effects of sex on treatment outcomes, rather than disaggregating the data by sex and then examining for treatment effects, as has traditionally been the norm in previous studies. Such robust studies yield meaningful interpretations of true sex differences in translational outcomes for preclinical studies.
In summary, the MCM treatment (atropine, oxime, midazolam) alone had no long-term protective effect in soman induced SE in both males and females. 1400W treatment for two weeks mitigated soman-induced motor and memory de cits, and reduced anxiety/fear like-behavior. No sex differences were observed in behavioral tests. 1400W also signi cantly reduced epileptiform spike rate (a measure of neuronal hyperexcitability) and SRS frequency in males but not in females. Nitrooxidative markers such as nitrite, ROS, and oxidized glutathione, and proin ammatory cytokines and chemokine in the serum and CSF were signi cantly reduced in 1400W treated group. Additionally, the soman-induced reactive gliosis and neurodegeneration were reduced by 1400W treatment in some brain regions in both sexes. Parvalbumin positive neurons were protected by 1400W in amygdala in soman exposed animals. Overall, these ndings suggest the disease-modifying effects of a highly selective iNOS inhibitor, 1400W, in an experimental model of nerve agent neurotoxicity.

Availability of data and materials
All the raw data and processed data are stored in University's long-term data storage facility and can be accessed by requesting tswamy@iastate.edu. This project did not generate any materials to share. All the materials are available for purchase through commercial companies which is stated in the manuscript.  1400W signi cantly reduced soman-induced motor de cit, measured by rotarod, in mixed sex cohort (A).
The impact of mild and severe SE (B). Experimental design for the Novel Object Recognition (NOR) test (C) and representative heatmap from each group during 24 hours probe testing (D). 1400W signi cantly rescued soman-induced long-term memory de cits at 24h post-familiarization (F) but not at 3h (E). The effects of soman /vehicle or 1400W in mild and severe SE groups were compared (G The Elevated Zero/Circular Maze (A-C) and Open Field tests (D, E). Representative heatmaps from each group is shown (A). Control and soman+1400W groups signi cantly spent more time in closed arms (B, C). Soman exposed 1400W treated group spent signi cantly more time at the center (D) or less time in the periphery (E) in an open eld (illustrated in the schematic diagram) than the vehicle 1400W group.
Contextual Fear Conditioning Test (F-H). During habituation before conditioning (F) and probing (G), soman exposed groups froze longer than the controls. There were no signi cant differences in freezing time during the inter-trial intervals (ITI) between any groups during the conditioning phase (F). Differences   SRS frequency in males and females compared between vehicle and 1400W treated soman exposed animals; (C. Males, D) Unpaired t-test; (C. Females, E) Mann-Whitney test; (D) Mixed-effects analysis (Šídák's multiple comparisons test); for all, n=12-13 for mixed sex cohort or 5-8/sex; *p<0.05, **p<0.01.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. Tables1and2.docx