A rat model of organophosphate-induced status epilepticus and the beneficial effects of EP2 receptor inhibition

Abstract

This review describes an adult rat model of status epilepticus (SE) induced by diisopropyl fluorophosphate (DFP), and the beneficial outcomes of transient inhibition of the prostaglandin-E₂ receptor EP2 with a small molecule antagonist, delayed by 2-4h after SE onset. Administration of six doses of the selective EP2 antagonist TG6-10-1 over a 2-3 day period accelerates functional recovery, attenuates hippocampal neurodegeneration, neuroinflammation, gliosis and blood-brain barrier leakage, and prevents long-term cognitive deficits without blocking SE itself or altering acute seizure characteristics. This work has provided important information regarding organophosphate-induced seizure related pathologies in adults and revealed the effectiveness of delayed EP2 inhibition to combat these pathologies.

Graphical Abstract

INTRODUCTION:

Exposure to organophosphorus (OP) based agents poses a major public health concern. Recently, there has been a burgeoning number of threats by terrorist organizations and radicals to induce mass casualties by the release of OP nerve agents. Examples of the devastation caused by the use of such an agent amid a civilian population are evident in the Ghouta chemical attack in Syria in 2013 and again in the Khan Shaykhun chemical attack in 2017. Death or injury incurred from OP based agents can also result from an industrial accident or by ingestion as a means of suicide. Organophosphorus based agents potently inhibit acetylcholinesterase (AChE), leading to an acute cholinergic crisis. Seizures can be evident within minutes of OP exposure and often progress to status epilepticus (SE). The current treatment to combat OP poisoning in humans is administration of three drugs (atropine, pralidoxime and diazepam), which must be given soon after OP exposure for optimal effectiveness (Jett, 2007, 2016). It is not logistically possible to reach all victims exposed to OP agents within the short therapeutic window needed for current antidotes, and thus therapies are needed that are effective hours to days after OP exposure to mitigate the ensuing brain injury and development of cognitive deficits in survivors. In this review we discuss the rat diisopropyl fluorophosphate (DFP) model of SE and summarize the efficacy of a prostaglandin-E₂ (PGE₂) receptor 2 (EP2) antagonist as an anti-inflammatory strategy to combat OP-induced neuroinflammation, neurodegeneration, gliosis, blood-brain barrier leakage and long-term cognitive deficits.

CYCLOOXYGENASE SIGNALING AND THE EP2 RECEPTOR

Cyclooxygenase-1 (COX-1; prostaglandin G/H synthase 1, PTGS1) is constitutively expressed throughout the body and involved in many physiological functions including cellular homeostasis as well as maintenance of gastrointestinal integrity (Davies, 1995; Murray & Brater, 1993), whereas cycloxygenase-2 (COX-2; prostaglandin-endoperoxide synthase 2, PTGS2) is upregulated in response to an injury or seizures. In the brain, COX-2 is induced in response to a variety of injuries including those produced by stroke and SE (Kaufmann et al., 1996; Liu et al., 2005; McCullough et al., 2004). The prostanoids made by COX-2 are lipid molecules that mediate a variety of physiological and pathological processes, activating a total of nine G protein-coupled receptors (Figure 1A). The various molecules made by the activity of COX-2 and the numerous cellular processes involving COX-2 signaling underpin the ardent interest in COX-2 as a therapeutic target.

Figure 1. Cyclooxygenase and EP2 signaling cascades.

A. Schematic of the COX-2 signaling cascade following initiation of status epilepticus induced by an organophosphorus based agent. Both COX-1 and COX-2 enzymes convert arachidonic acid to prostaglandin H2 (PGH₂). Cell specific prostanoid syntheses convert PGH₂ into the various prostaglandin ligands, which bind and activate specific G protein-coupled receptors. COX-2 also metabolizes endocannabinoids to prostaglandin analogs (glycerol esters and ethanolamides (not shown for simplicity). Four receptors, prostaglandin D2 receptor 1 (DP1), prostaglandin E2 receptor 2 (EP2), prostaglandin E2 receptor 4 (EP4) and prostacyclin receptor (IP), promote cyclic adenosine monophosphate (cAMP) production upon activation. Receptor activation resulting in cAMP production and Ca²⁺ mobilization in conjunction with other cellular responses like changes in gene expression lead to the various neuropathologies that develop as a result of OP-poisoning. ERK, extracellular signal-regulated kinase; p38, mitogen-activated protein kinase; cPLA2, calcium-dependent phospholipase A2. The boxed in receptors and signaling molecules in A indicates that prostanoid receptors may function in an autocrine manner by activating receptors expressed on the same cells that synthesize the prostanoids or in a paracrine manner by activating receptors on neighboring cells. B. Schematic of the EP2 signaling cascade following activation by PGE₂. Adenylyl cyclase (AC) promoted cAMP production which activates protein kinase A (PKA) and/or exchange protein directly activated by cAMP (EPAC) signaling. RAP, Ras GTPase quinine nucleotide exchange factor; CREB, cAMP response element-binding protein; GSK, glycogen synthase kinase 3; ERK, extracellular signal–regulated kinase. C, chemical structures of organophosphorus (OP) based agents and EP2 receptor antagonist (PF-04418948, Pfizer; TG6-10-1, Emory University).

Due to the adverse effects of selective COX-2 inhibitors and the inconsistencies of the efficacy of these COX-2 inhibitors to alter seizure related pathologies, summarized in two reviews (Rojas et al., 2014; Rojas et al., 2019), an alternative therapeutic strategy has been to target the prostanoid receptors downstream of COX-2. Activation of the PGE₂ receptors EP2 and EP4 has divergent effects on inflammatory signaling in myeloid cells, elevating expression of interleukin 6 (IL-6) and interleukin 1 beta (IL-1β) but depressing that of tumor necrosis factor alpha (TNFα) and chemokine (C-C motif) ligand 2 (CCL2) (Jiang et al., 2013; Rojas et al., 2016; Woodling et al., 2014). The EP2 and EP4 receptors are Gαs-coupled and stimulate adenylate cyclase when activated by PGE₂, resulting in the elevation of cAMP (Figure 1B). EP2 plays an important role in neuroinflammatory conditions in the brain (Andreasson, 2010; Bilak et al., 2004; Jiang et al., 2012; Jiang et al., 2013; Johansson et al., 2013; Liang et al., 2005; Liu et al., 2005; McCullough et al., 2004; Taniguchi et al., 2011). Antagonism of the EP2 receptor with small molecule inhibitors (i.e., PF-04418948 and TG6-10-1, Figure 1B) could be an alternative therapeutic strategy to the generic block of COX-2 for anti-inflammatory therapy following exposure to OP based agents. Thus far, Inhibition of the EP2 receptor with TG6-10-1 has been found to accelerate functional recovery after DFP-induced SE in adult rats, affords hippocampal neuroprotection and reduces neuroinflammation, microgliosis, blood-brain barrier leakage and cognitive deficits (Rojas et al., 2015; Rojas et al., 2016). An effective therapeutic agent that can be delivered hours to days after OP exposure is not available thus far. Thus, inhibition of the EP2 receptor appears promising as a novel strategy for neuroprotection and functional restoration after OP-induced brain injury in adults.

DFP-INDUCED SE IN ADULT RATS

Organophosphorus based agents like DFP (Figure 1C) are chemiconvulsants that at high doses can induce SE in humans and experimental animals. The structure and mechanism of action of DFP is similar to that of the nerve agent sarin (GB), and so the rat DFP model (Zhu et al., 2010) has been used to replicate the sequelae of lethal or sublethal effects following exposure to sarin. The OP agents are inhibitors of acetylcholinesterase, which is the major enzyme that breaks down the neurotransmitter acetylcholine at cholinergic synapses and the neuromuscular junction (King & Aaron, 2015). Inhibition of acetylcholinesterase leads to an accumulation of acetylcholine at muscarinic and nicotinic synapses. The excess acetylcholine overstimulates the cholinergic system, inducing a cholinergic crisis and subsequent increase in seizure activity, SE and eventually death if left unresolved. A seizure is a sudden, synchronous electrical disturbance in the brain that can be detected with electroencephalography (EEG). Status epilepticus (SE) is a medical emergency resulting from a seizure that lasts longer than 5 minutes in humans, or multiple seizures that occur consecutively without full regain of consciousness.

In rats, DFP-induced seizure behaviors begin within a few minutes of DFP exposure and consist of distinct motor behaviors that include forelimb clonus, tail extension, and whole body clonic seizures. These behaviors are observed and scored using a modified Racine scale (Table 1). Adult rats exposed to a high dose of DFP often experience SE (Rojas et al., 2015; Rojas et al., 2016; Rojas et al., 2018). For example, nearly 80% of 386 adults rats experienced SE when DFP was administered at 9.5 mg/kg intraperitoneally. A similar percentage of rats experienced SE when DFP was administered at 5 mg/kg subcutaneously (Rojas et al., 2018). These two doses and routes of DFP result in a similar latency to SE onset and prolonged SE lasting >5 h without pharmacological intervention, consistent with a previous study demonstrating that a dose of DFP equal to or greater than 4 mg/kg administered subcutaneously results in consistent SE in adult rats (Pouliot et al., 2016). In our studies involving EP2 receptor inhibition all rats were administered DFP at 9.5mg/kg intraperitoneally (Rojas et al., 2015, 2016). In the study comparing urethane to diazepam most of the rats were administered DFP at 5 mg/kg subcutaneously (Rojas et al., 2018). Only a small cohort of normal adult male rats (from the uninterrupted group) were administered DFP at 9.5 mg/kg intraperitoneally in the study comparing urethane to diazepam. We found no difference in the level of neuroinflammation, neurodegeneration and gliosis observed on day 4 in rats that received DFP at 5 mg/kg subcutaneously compared to rats that received DFP at 9.5 mg/kg intraperitoneally (unpublished data). Behavioral scoring using a modified Racine scale also revealed an identical temporal evolution of status epilepticus between rats given DFP at 5 mg/kg subcutaneously and those that received DFP at 9.5 mg/kg intraperitoneally (unpublished data).

Table 1.

Modified Racine scale for behaviors after DFP exposure

Behavioral Score		Observed Motor Behavior
0	Normal Behavior:	walking, exploring, sniffing, grooming
1	Freeze Behavior:	immobile, staring, heightened startle, curled-up posture
2	Repetitive Behavior: washing, whisker twitching	blinking, chewing, head bobbing, scratching, face
3	Early Seizure Behavior:	myoclonic jerks, partial body clonus
4	Advance Seizure Behavior:	whole body clonus
5	Status Epilepticus (SE):	repeated seizure activity (≥2 events in stages 3, 4 or 6 within a 5-minute window)
6	Intense Seizure Behavior:	repetitive jumping or bouncing, wild running, tonic seizures
7	Death

The success rate of SE in rats observed with such high doses of DFP can be partially attributed to the administration of supporting agents (i.e., pyridostigmine and quaternary atropine analogs) that allow more rats to survive the acute effects of DFP. Pyridostigmine bromide (0.1 mg/kg in 0.9% saline, sc) and methylatropine bromide, methylatropine nitrate or ethylatropine bromide (20 mg/kg in 0.9% saline, sc) were administered prior to DFP to protect the animal’s peripheral nervous system, based on their inability to cross the blood-brain barrier and capacity to increase survival of rats following DFP exposure without altering the development of seizures. All rats were given the supporting agents prior to DFP. Pyridostigmine bromide was administered at time zero followed by methylatropine bromide, methylatropine nitrate or ethylatropine bromide twenty minutes later. We recently demonstrated that ethylatropine bromide, though less potent than methylatropine bromide at antagonizing the M1 muscarinic receptor, was highly effective at blocking fluid secretions induced by pilocarpine in mice without affecting seizures (Rojas et al., 2017). Ethylatropine bromide is inexpensive to synthesize and the synthesis is environmentally safe compared to methylatropine bromide and methylatropine nitrate. The synthesis of methylatropine bromide was halted due to environmental concerns and is no longer manufactured in the United States. This caused an increase demand for methylatropine nitrate and thus the price ballooned. Therefore, recently ethylatropine bromide has replaced methylatropine bromide and methylatropine nitrate as a supporting agent prior to DFP to increase survival following DFP exposure (Rojas et al., 2018). Ten minutes after administration of methylatropine bromide, methylatropine nitrate or ethylatropine bromide, the rats were injected with DFP diluted in sterile distilled water (9.5 mg/kg, ip or 5 mg/kg, sc). DFP was always prepared fresh within 5 min of administration with thorough mixing. In rodents, SE results in high mortality if it persists for many hours. Unfortunately, administration of supporting agents like pyridostigmine bromide and atropine methyl nitrate or ethylatropine bromide did not protect all rats from the acute effects of DFP poisoning (Rojas et al., 2015; Rojas et al., 2016; Rojas et al., 2018). Following DFP exposure, a large fraction of the rats that die succumb to the acute effects of DFP, dying before SE or within 8 hours of the onset of SE. The rats that die before SE tend to do so within minutes of the DFP administration mainly by respiratory cessation, heart failure or the multifaceted effects of DFP on other systems. This acute mortality was seen with both dosing paradigms of DFP (i.e., 5 mg/kg, sc and 9.5 mg/kg, ip). A smaller fraction of rats that experience SE survive the first 24 hours, but die within a 7 day period. This delayed mortality appears to be related to the duration of the initial SE as it was higher in rats that experienced uninterrupted SE (Rojas et al., 2015; Rojas et al., 2016; Rojas et al., 2018).

Electrographic SE was terminated in rats administered a subanesthetic subcutaneous dose of urethane 1 h after DFP-exposure, but the synchronous discharges of SE returned overnight in rats administered diazepam (Figure 2A). Only a few rats experienced SE and died after 24 hours when SE was interrupted by diazepam or terminated by urethane (Rojas et al., 2018). Although diazepam was not as effective as urethane at terminating SE electrographically, diazepam treatment resulted in a high survival following SE (Rojas et al., 2018) consistent with an improved survival following diazepam administration in other models of SE (Apland et al., 2014; Mello et al., 1993; Trandafir et al., 2015). Taken together, the high success rate of rats experiencing SE coupled with the relatively low mortality allows for investigation into neuropathologies that develop in survivors of DFP-induced SE.

Figure 2. Urethane but not diazepam suppresses the return of seizure activity.

All rats were administered pyridostigmine bromide (0.1 mg/kg in 0.9% saline, sc) at time zero and methylatropine bromide, methylatropine nitrate or ethylatropine bromide (20 mg/kg in 0.9% saline, sc) 20 minutes later followed by DFP in sterile water (5 mg/kg sc or 9.5 mg/kg ip) 10 minutes later to induce SE. Non-seizure control rats (“controls”) received the same supporting agents and DFP was replaced by sterile water. A, the EEG power in the 20-70 Hz bandwidth averaged over 300 sec epochs during the 24 hour period for 6 diazepam-treated and 6 urethane-treated rats. The dashed line indicates baseline power before DFP. Diazepam (10 mg/kg, ip) was administered 60 minutes after SE onset. A subset of rats were exposed for 5-7 minutes to isoflurane by inhalation followed by administration of a subanaesthetic does of urethane (0.8 g/kg, sc). B and C, are images of the hippocampus (25x total magnification) of a non-seizure control rat stained for Nissl (B) and AChE (C). The brain of a non-seizure control rat was rapidly removed and bisected longitudinally. One hemisphere was fixed in 4% paraformaldehyde overnight and transferred to 30% sucrose the next day until it sank. The brain was embedded in tissue freeze medium and sectioned coronally on a cryostat at 40 μm. Coronal sections were stained for Nissl and acetylcholinesterase activity using the protocol described by Paxinos et al. (1980). The dark purple indicates the presence of the Nissl bodies in B and the brown precipitate in C is indicative of areas of high activity of AChE. D, the brain of a rat that experienced uninterrupted SE induced by DFP 24 hours earlier was rapidly removed, bisected longitudinally, coronally sectioned and stained as described above. Shown is an image of the hippocampus (25x total magnification) of a section stained for AChE activity. The brown precipitate although present in some cells is much lower overall indicating reduced AChE activity. The scale bar in B, C and D = 500 μm. E, AChE inhibition in rat brain measured by an acetylcholinesterase assay (Rojas et al., 2015) is significantly reduced 1 h after DFP exposure and is prominent by 24 h after DFP-exposure. ** = p < .01, one-way ANOVA with posthoc Dunnett’s. The number inside the bar represent the number of rats in each group. F, immunohistochemistry was performed on rat coronal hippocampal sections for COX-2 and NeuN as described by Rojas et al. (2015). The NeuN antibody was diluted 1:2000 (MAB377, Millipore). Alexa Fluor594 goat anti-mouse was diluted 1:1000 (ThermoFisher Scientific). Fluorescent images taken from the CA3 region in the hippocampus (200x total magnification) reveals basal expression of neuronal COX-2 in rats that did not experience status epilepticus (No SE, left insert). Neuronal COX-2 in the CA3 region is greatly induced 24 hours after DFP-induced SE (DFP-SE, middle insert). Red fluorescent images of the CA3 region in the hippocampus reveals expression of the neuronal nuclei marker NeuN. Overlapping the green COX-2 stain, the red NeuN stain and the Hoechst revealed COX-2 induction in the same neurons positively stained for NeuN. Examples are indicated by the white arrows. The images shown are representative of five sections each from three or more rats. Scale bar, 30 μm.

EARLY CONSEQUENCES AFTER DFP EXPOSURE

Early consequences represent features of the acute crisis that occur within a few hours of OP exposure (Chen, 2012; Vale & Lotti, 2015). The early signs of a cholinergic crisis induced by OP agents in humans include pupil dilation, respiratory distress, decreased heart rate and seizure activity (Hulse Clutton, et al., 2014; Hulse, Davies, et al., 2014; Peduto et al., 1996; Vale & Lotti, 2015). In rats, DFP exposure also results in early consequences of cholinesterase inhibition such as whole body motor convulsions, muscle weakness and the initial development and return of SE (refractory SE) (Deshpande et al., 2010; Li et al., 2011; Liang et al., 2018; Pessah et al., 2016; Pouliot et al., 2016; Wu et al., 2018). Another common early consequence seen in humans and animals, which manifests within minutes of OP exposure, is fluid secretions that include salivation, lacrimation, urination and defecation (SLUD). These SLUD symptoms are the result of AChE inhibition in the periphery and are reduced by the use of supporting agents such as ethylatropine bromide. Protein lysates obtained from half-brains lacking the cerebellum were used to measure AChE activity with an acetylcholinesterase assay kit (Abcam). Brain acetylcholinesterase levels were significantly inhibited when measured at 1 hour post-DFP exposure (Figure 2B-E) and remained inhibited for several days (Rojas et al., 2018). Interestingly, inhibition of AChE in the brain measured 5 hours after DFP was similar for rats that entered or did not enter SE. A previous study also showed a similar reduction of AChE in the brains of rats that experienced SE induced by soman and rats that were administered soman but did not experience SE (Prager et al., 2013). However, in the same study by Prager et al. (2013) further examination of AChE inhibition in the basolateral amygdala revealed that it was different for rats that entered or did not enter SE, suggesting that the degree of AChE inhibition in certain brain structures and not the whole the brain may be associated with SE.

INTERMEDIATE CONSEQUENCES AND EP2 ANTAGONISM WEIGHT LOSS AND REGAIN

Intermediate consequences are delayed effects such as neurodegeneration, neuroinflammation, gliosis, blood-brain barrier (BBB) breakdown, weight loss/regain, muscle weakness, and disruption of GI function that manifest days after exposure (Chen, 2012; Vale & Lotti, 2015). Rats exposed to DFP that enter SE which is interrupted after 60-90 min typically lose body weight over the first two days and then start to regain weight from day three onward. The amount of weight the animal loses appears to correlate with the initial SE insult. For example, rats that are administered DFP but do not experience SE only lose weight between day 0 and day 1 and begin to regain weight by day 2, whereas rats that experience uninterrupted SE often continue to lose weight through day 3 (Figure 3A). The weight loss is usually temporary as survivors tend to regain weight beginning on day 4 and may reach or surpass their original weight by day 10 and recover to a normal growth rate (Rojas et al., 2016). A small fraction of rats continue to lose weight and never recover from the SE experience, reaching a euthanasia endpoint. These rats were not included in the data analysis. All rats exposed to DFP lost a similar amount of body weight by day 1 regardless of whether or not they experienced SE. However, the weight change between day 1 and day 4 has been shown to correlate with functional measures of recovery in mice after pilocarpine induced SE, and recovery could be hastened by inhibiting EP2 after SE was terminated (Jiang et al., 2013). We asked whether EP2 inhibition alters weight loss/regain in the DFP-treated rat. We have demonstrated that the efficacy of the EP2 antagonist to alter weight loss after DFP-induced SE depends on the dose and timing of administration of EP2 antagonist, TG6-10-1 (Rojas et al., 2015; Rojas et al., 2016). DFP-induced SE was not terminated pharmacologically in studies involving inhibition of the EP2 receptor. Administration of a single dose of TG6-10-1 prior to DFP or administration of two doses 4 hours and 21 hours after DFP-SE onset did not facilitate weight regain. On the other hand, rats began to regain weight significantly faster when TG6-10-1 was administered 6 times over a 48-hour period in a manner that produced consistently high plasma levels of TG6-10-1 (Figure 3A). Pharmacological termination of SE with urethane can also facilitate weight regain and reduce variation in the percent of weight loss (Rojas et al., 2018).

Figure 3. Beneficial effects of TG6-10-1 after DFP-induced status epilepticus.

All rats were administered pyridostigmine bromide (0.1 mg/kg in 0.9% saline, sc) at time zero and methylatropine bromide (20 mg/kg in 0.9% saline, sc) 20 minutes later followed by DFP in sterile water (9.5 mg/kg ip) 10 minutes later to induce SE. Non-seizure control rats (“controls”) received the same supporting agents and DFP was replaced by sterile water. A, six injections of TG6-10-1 (n = 8 rats) beginning 80-150 min after SE onset significantly accelerated weight regain compared to vehicle (n = 8 rats) administration (p < .0001 by one-way ANOVA with posthoc Bonferroni). Rats administered DFP that did not enter status epilepticus (DFP No SE, n = 7) returned to the initial weight just prior to DFP exposure by day four. B, change in abundance of 10 inflammatory mediator mRNAs from the forebrain of rats 4 d after injection with water or DFP to induce status epilepticus. Post-DFP treatment was 6 doses of TG6-10-1 (n = 8 rats) or vehicle (n = 9 rats). Following DFP induced status epilepticus, the mRNA fold change for 8 mediators as a group was significantly reduced by TG6-10-1 compared to vehicle (p = .019, paired t test). C, the average number of injured neurons per section in three hippocampal regions of rats treated with 6 doses of vehicle (n = 6 rats) and rats injected with 6 doses of TG6-10-1 (n = 7 rats) four days after DFP-induced status epilepticus. (* p < .05 in CA1, one-way ANOVA with posthoc Bonferroni). D, induction of GFAP and Iba1 mRNA in the forebrain four days following DFP status epilepticus in vehicle treated (n = 7 rats) and TG6-10-1 treated rats (n = 7 rats) (one-way ANOVA with posthoc Bonferroni).

NEUROINFLAMMATION IN DFP-TREATED RATS

Neuroinflammation is observed in the days following DFP-induced SE in rats (Rojas et al., 2015; Rojas et al., 2018). The inflammation is characterized by the transient upregulation of a small subset of inflammatory mediators including cytokines (interleukin 1 beta, IL-1β; tumor necrosis factor alpha, TNFα; interleukin 6, IL-6), chemokines (chemokine (C-C motif) ligand 2, CCL2; CCL3; CCL4; C-X-C motif chemokine 10, CXCL10) and pro-inflammatory enzymes (COX-2 and NADPH oxidase 2, NOX2). Within hours of DFP exposure the levels of these inflammatory mediators are significantly upregulated in the brain and this increase persists for several days. For example, COX-2 induction in hippocampal neurons is very prominent 24 hours following DFP exposure (Figure 2F). The temporal pattern of upregulation is quite different for individual inflammatory mediators. IL-1β and TNFα protein levels returned to baseline by day 4 following uninterrupted SE, whereas COX-2 remained elevated in neurons on day 4 (Rojas et al., 2015), consistent with the prolonged upregulated level of COX-2 following soman (GD) exposure in rats (Angoa-Perez et al., 2010). On the other hand, cyclooxygenase 1 (COX-1) and interleukin 10 (IL-10) protein levels remained unchanged through the first four days following DFP-induced SE (Rojas et al., 2015). The mRNA levels of all inflammatory mediators tested were elevated above the control level on day 4 in DFP-treated rats that had experienced uninterrupted SE (Figure 3B). This robust inflammatory mediator burst in the brain is of particular interest as the EP2 receptor is an effector molecule that sits in the lower portion of an inflammatory cascade mediated by COX-2 (Figure 1A). COX-2 ablation in principal forebrain neurons resulted in a reduced hippocampal inflammatory burst following pilocarpine induced SE in mice (Serrano et al., 2011). Similarly, EP2 antagonism by TG6-10-1 dampened the inflammatory burst in the hippocampus of mice that endured SE induced by pilocarpine (Jiang et al., 2013). Therefore, we determined whether EP2 inhibition by TG6-10-1 opposes the inflammatory burst in the brain of rats following DFP exposure. Administration of 6 doses of TG6-10-1 strongly attenuated the induction of the eight inflammatory mediators assayed in rats following DFP (Figure 3B). Based on the plasma half-life of 2.5h for TG6-10-1 in rats, we believe that multiple injections were needed for a sustained inhibition of EP2 receptor over a 2 day period, in order to reduce inflammation. In a more recent study, interruption of SE with diazepam 1 hour after SE onset resulted in a similar level of inflammatory mediators compared to rats that experienced uninterrupted SE as measured on day 4 (Rojas et al., 2018). However, rats administered urethane 1 hour after SE onset displayed no evidence of the inflammatory burst as mRNA for all mediators were at the basal level on day 4 (Rojas et al., 2018). It is likely that urethane prevents the overnight return of SE, thus the observed inflammation on day 4 may largely be triggered by a second bout of SE in already metabolically challenged animals (Rojas et al., 2018).

DFP-INDUCED NEURODEGENERATION

COX-2 activation in the brain has been shown to promote delayed neuronal damage in rodent models of excitotoxicity or temporal lobe epilepsy (Kawaguchi et al., 2005; Manabe et al., 2004; Polascheck et al., 2010; Serrano et al., 2011; Takemiya et al., 2006). Rats that experience DFP-induced SE display neurodegeneration in multiple brain regions including the hippocampus, amygdala, thalamus, piriform cortex, parietal cortex, and entorhinal cortex (Rojas et al., 2015; Rojas et al., 2018 and unpublished data). Neurodegeneration can be determined using FluoroJade B (FJB), a dye that enters dead or dying neurons and emits green fluorescence (Schmued et al., 1997). Following DFP-induced SE, hippocampal neurodegeneration determined by FJB staining was seen in hilar interneurons at 5 hours. These data are consistent with a study demonstrating near complete hilar neuron destruction by 6 hours after pilocarpine-induced SE in mice (Borges et al., 2003). In the principal CA1 and CA3 neuronal areas, FJB staining was undetectable at 5 hours but obvious at 24 hours and more prominent at 4 days post SE. Administration of 6 doses of TG6-10-1 over a 48h period resulted in neuroprotection of pyramidal cells in the cornu ammonis 1 (CA1) and CA3 regions (Figure 3C). However, TG6-10-1 did not protect hilar neurons following DFP-induced SE, which was similar to results obtained following pilocarpine-induced SE in mice (Jiang et al., 2013; Rojas et al., 2015). On the other hand, hilar neurons were partially protected following pilocarpine-induced SE in the neuronal COX-2 conditional knockout mice (Serrano et al., 2011). It is likely that SE-induced hilar neuron injury became irreversible before a therapeutic level of TG6-10-1 was reached. Limiting the duration of SE to 1 hour by termination with urethane did not protect hilar neurons from injury (Rojas et al., 2018). Although significant neuroprotection of hippocampal pyramidal neurons was found in rats administered six doses of TG6-10-1 over 48h, the highest degree of hippocampal neuroprotection observed thus far was a result of complete termination of SE by urethane injection (Rojas et al., 2018). Because subanesthetic urethane can prevent the overnight re-entry into SE (Figure 2A), we suspect that the most injurious phase of DFP occurs during the re-entry of already-vulnerable rodents into SE on day zero, rather than the initial more intense bout of SE. It is unclear whether early neurodegeneration observed in the days following DFP-induced SE continues throughout the epileptogenesis period. Taken together, these experiments provide evidence of neurodegeneration in the hippocampus and reveal a neurodegenerative role of EP2 in the setting of brain injury incurred by OP-induced SE in adult rats. In the future, it will be important to determine whether the neuroprotection of EP2 antagonism is shared by other seizure sensitive brain regions such as the amygdala, thalamus and piriform cortex.

DFP-INDUCED GLIOSIS

Gliosis, the morphological and metabolic reaction of glial cells to injury, can be detected in the brain of rats following OP agents (Angoa-Perez et al., 2010; Kuruba et al., 2018; Liu et al., 2012; Rojas et al., 2015; Rojas et al., 2018; Wu et al., 2018). Astrogliosis (astrocyte activation) and microgliosis (microglia activation) can be revealed by more intense immunohistochemical staining of glial fibrillary acidic protein (GFAP) and ionized calcium-binding adapter molecule 1 (IBA1) respectively after DFP-induced SE (Ferchmin et al., 2014; Kuruba et al., 2018; Li et al., 2015; Li et al., 2011; Liu et al., 2012; Rojas et al., 2015; Siso et al., 2017; Wu et al., 2018). Alternatively, the mRNA levels of these two glial markers can be quantified to determine the amount of gliosis that occurs following SE in rodents (Jiang et al., 2013; Rojas et al., 2015;Rojas et al., 2018; Serrano et al., 2011). Following DFP-induced SE there is an early astrogliosis response in the brain that occurs within hours of SE and a more robust and delayed response that is observed within days (Liu et al., 2012). Astrogliosis and microgliosis were prominent in the hippocampus 4 days after DFP-induced SE that had not been interrupted pharmacologically (Rojas et al., 2015). Rats administered six doses of the EP2 antagonist, TG6-10-1, over 48h displayed a more pronounced reduction of microgliosis than astrogliosis (Figure 3D), suggesting that the EP2 receptor contributes to the robust brain inflammation following DFP-induced SE and microglia expressing EP2 partially mediate this inflammatory response. This is coherent with the demonstration that microglia cells express functional EP2 receptors (Bonfill-Teixidor et al., 2017; Fu et al., 2015; Johansson et al., 2013; Johansson et al., 2015; P. Li et al., 2009; Li et al., 2015; Quan et al., 2013). TG6-10-1 reduced microgliosis in rats exposed to DFP and mice that experienced pilocarpine-induced SE (Jiang et al., 2013). Although the expression pattern of the EP2 receptor has not been investigated following DFP-induced SE, it was shown that EP2 expression level was only minimally changed by pilocarpine-induced SE in mouse brains measured up to 4 days (Jiang et al., 2015). Termination of SE by urethane, but not interruption by diazepam, 1h after SE reduced astrogliosis but not microgliosis (Rojas et al., 2018). This is consistent with the idea that the return of SE or a prolonged initial SE exacerbates gliosis.

BLOOD-BRAIN BARRIER INTEGRITY

The blood-brain barrier (BBB) plays a critical role in homeostasis in the brain. It is comprised of endothelial cells, astrocytes, pericytes and microglia (Liebner et al., 2018; Sa-Pereira et al., 2012). Neuronal COX-2 induction and subsequent EP2 signaling coupled with BBB breakdown may promote the epileptogenic process (Bankstahl et al., 2018; Bauer et al., 2008; Broekaart et al., 2018; Jiang et al., 2013; Serrano et al., 2011; Zibell et al., 2009). Epileptic seizures can also lead to upregulation of the efflux drug transporter P-glycoprotein via a COX-2 mediated signaling pathway (van Vliet et al., 2010). The integrity of the BBB can be determined by perfusing animals with phosphate buffered saline to remove blood from the brain, and subsequently measuring the extravasation of serum albumin in the brain parenchyma by western blot. Serum albumin is low or absent in the brain parenchyma when the BBB is intact. In mice, elevated levels of serum albumin were found in the cortex four days after pilocarpine-induced SE (Jiang et al., 2013; Serrano et al., 2011; Varvel et al., 2016), which is consistent with breakdown of the BBB. Elevated albumin was not detected in the brain of mice lacking COX-2 in principal forebrain neurons following pilocarpine-induced SE (Serrano et al., 2011). Administration of the EP2 antagonist, TG6-10-1, likewise prevented extravasation of albumin into the cortex four days after pilocarpine-induced SE in mice (Jiang et al., 2013). Serum albumin was increased about four fold in the brain of rats that experienced uninterrupted SE by DFP (Figure 4). However, albumin levels were reduced in the brain parenchyma of rats that received two subcutaneous doses of TG6-10-1 delivered 4 and 24 hours after SE onset (Figure 4). These results indicate that neuronal COX-2 signaling involving activation of the EP2 receptor initiates breakdown of the BBB after SE. Preventing activation of the COX-2/EP2 signaling pathway could be a promising approach to control the integrity of the BBB.

Figure 4. EP2 receptor antagonist TG6-10-1 aids in maintaining the integrity of the blood-brain barrier after DFP-induced SE.

All rats were administered pyridostigmine bromide (0.1 mg/kg in 0.9% saline, sc) at time zero and methylatropine bromide (20 mg/kg in 0.9% saline, sc) 20 minutes later followed by DFP in sterile water (9.5 mg/kg ip) 10 minutes later to induce SE. Non-seizure control rats received the same supporting agents and DFP was replaced by sterile water. A, The amount of serum albumin in the cortex 4 days after SE was used to assess the integrity of the blood–brain barrier as all rats were perfused with sterile saline to completely remove blood from all tissues. The albumin levels (green) in the cortex of non-seizure controls or rats that experienced uninterrupted SE induced by DFP that received TG6-10-1 (5, 50 mg/kg, sc) or the vehicle (olive oil) at 4 and 24 hours after SE onset were measured by western blot with GAPDH (red) used as a loading control. B, the band intensity of the albumin was normalized to the housekeeping GAPDH. The bar represents the mean ± SEM. The number in the white box within the bar represents the total number of rats within each group. p < 0.05, p < 0.01, one-way ANOVA and post hoc Bonferroni test with selected pairs.

LONG-TERM CONSEQUENCES OF DFP AND EP2 INHIBITION

Long-term consequences are effects that manifest weeks, months or even years following OP exposure such as cognitive deficits, altered anxiety behavior and the development of spontaneous recurrent seizures (SRS) (Chen, 2012; de Araujo Furtado et al., 2012; Vale & Lotti, 2015). Status epilepticus induced by nerve agents leads to the development of SRS and long-term cognitive deficits in animals and humans (Chen, 2012; de Araujo Furtado et al., 2010; de Araujo Furtado et al., 2012; Joosen et al., 2009; Miyaki et al., 2005; Nishiwaki et al., 2001). In animal studies, these long-term cognitive deficits are correlated with the degree of early neuronal injury and subsequent neuronal plasticity that occurs in the brain after SE (Chen, 2012; Filliat et al., 1999; Joosen et al., 2009; McDonough et al., 1987; McDonough et al., 1986; Myhrer et al., 2005). Treatments that can ameliorate long-term consequences of OP poisoning in survivors are needed, and therefore whether early treatment with the EP2 antagonist alters long-term consequences of DFP-induced SE was investigated. Rats that survive DFP-induced SE tend to survive long-term, however 5-12 weeks after DFP exposure the rats developed reduced anxiety behaviors, memory deficits and spontaneous recurrent seizures (Rojas et al., 2016; Rojas et al., 2018). A technical confound of performing behavioral assays in epileptic rats is that the rats may experience seizures during testing in stress-inducing paradigms. To limit this confound, we used behavioral assays that limit stress responses in rats, namely the light-dark preference task for anxiety related behaviors and novel object recognition (NOR) that investigates recognition memory. Reduced anxiety behavior was observed in the light-dark preference task as well as in an open field paradigm for NOR >5 weeks after SE in adult rats (Figure 5A-C). The anxiety behaviors in rats that experienced SE were very different from nonseizure control rats (Figure 5A-C). Reduced anxiety behaviors developed as a result of the DFP-SE experience (Figure 5A-B). This is consistent with other studies that demonstrated a trend towards decreased anxiety behaviors (Schultz et al., 1990; Wright et al., 2010) or no effect on anxiety (Nieminen et al., 1991; Valvassori et al., 2007) when rats were exposed to sub-SE doses of other organophosphorus agents. The mechanisms underlying the reduced anxiety behaviors in rats that survive DFP-induced SE is unknown. On the other hand, rats that experience SE induced by soman displayed increase anxiety behavior (Prager et al., 2014; Prager et al., 2015). We believe the net effect of organophosphorus agents on anxiety-related behaviors could depend on the agent and its dose. Nevertheless, transient early inhibition of EP2 receptors did not affect anxiety behaviors in rats that experienced SE in either the light-dark preference test or in the open field test (Figure 5A-C). However, rats treated with six doses of TG6-10-1 performed similar to control rats in the NOR memory test 8-12 weeks after DFP-induced SE, whereas vehicle treated rats could not differentiate the familiar from the novel object as the discrimination index was almost zero (Figure 5D). These studies suggests that early activation of the EP2 receptor after DFP promotes the development of long-term cognitive deficits such as an ability to form new memories. Rats that experienced one hour of SE terminated by urethane also displayed reduced anxiety behaviors that were very similar to rats that experienced longer durations of SE (Rojas et al., 2018).

Figure 5. TG6-10-1 alters memory retention but not anxiety behavior following DFP-induced SE.

All rats were administered pyridostigmine bromide (0.1 mg/kg in 0.9% saline, sc) at time zero and methylatropine bromide (20 mg/kg in 0.9% saline, sc) 20 minutes later followed by DFP in sterile water (9.5 mg/kg ip) 10 minutes later to induce SE. Non-seizure control rats received the same supporting agents and DFP was replaced by sterile water. The time spent in the light compartment (A) and latency to enter the light compartment (B) are shown for the four groups of rats (sham treated controls, DFP-no-SE, DFP-SE followed by vehicle, DFP-SE followed by TG6-10-1). The bars show the mean of the group and the number in the white box within the bar represent the total number of rats in each group. The error bars represent the standard error of the mean. The “+/−” symbol next to TG6-10-1 denotes sham treated control rats and DFP-no-SE rats that received TG6-10-1 (n=6) or vehicle (n=6). These rats were combined into one group as they were not different in any measure. ns = p > .05 by One-way ANOVA with Bonferroni posthoc. C, time spent in the center of the box during habituation in the NOR task is shown for the three groups tested. p < .01, One-way ANOVA, Dunnett’s posthoc. D, a discrimination index was used as a measure of memory retention. The individual groups were compared to zero by a 1-sample t test. The number in the white box within the bar represents the total number of rats in each group. The “+/−” symbol next to TG6-10-1 denotes sham treated control rats that received TG6-10-1 (n=6) or vehicle (n=6). These rats were combined into one group as they were not different in any measure. The horizontal dashed line at 0 indicates the point at which there is no discrimination between the novel and familiar objects. ns = p > .05. E, cortical EEG activity was recorded 4-6 weeks after SE induction by exposure to DFP. Shown are heat maps with the number of seizures detected per day from each rat (rows). A difference was detected between the two treatment groups [diazepam (n = 10) and urethane (n = 11)] in the percentage of rats that experienced at least one spontaneous recurrent seizure during the 6-12 days of EEG recording.

Currently, no strategy exists to prevent epilepsy in those at risk. Long-term monitoring of rats that experienced DFP-induced SE revealed that epileptic rats have a shorter life span than normal rats and this is unaltered by early administration of TG6-10-1 (Rojas et al., 2016). Utilizing cortical EEG recordings and the seizure detection criteria in Table 2, the number of SRS in rats was quantified 4-6 weeks following DFP-induced SE. All rats displayed normal cortical activity before and after ictal events as determined by the low amplitude and low power of the waveforms in the 20- to 70-Hz band. Each SRS was defined by the appearance of a gradually intensifying burst of large-amplitude (>3 X the baseline) and high-frequency spikes that persisted for >20 seconds followed by a post-ictal depression (a quieting of electrical activity) (Table 2). Both male and female rats develop SRS 4-6 weeks following DFP-induced SE (Rojas et al., 2018). The development of SRS appears to be correlated with the duration of the initial SE. Although SRS were found in both urethane and diazepam treated rats, a low proportion of rats (3 of 11) displayed SRS when SE was limited to 60 minutes by urethane administration (Figure 5E). By contrast, all 10 rats administered diazepam displayed at least one SRS 4-6 weeks following DFP-induced SE (Figure 5E). Furthermore, the number of SRS per day was much higher in diazepam treated epileptic rats compared to urethane. It appears that an overnight return of SE after interruption with diazepam, rather than the initially intense SE experience itself or prolonged SE in the case of rats that experience uninterrupted SE, is responsible for much of the morbidity associated with DFP-induced SE. In the future, it will be important to determine whether the beneficial effect of TG6-10-1 extends to the development of spontaneous recurrent seizures.

Table 2.

Spontaneous recurrent seizure detection criteria

1.	Normal baseline activity for a minimum of 5 minutes prior to an event
2.	Gradual (not abrupt) ramp up of spiking, typically over several seconds
3.	Continuous high frequency spiking (3x greater amplitude than baseline) that lasts >20 seconds
4.	Abrupt shutoff of spiking (noticeable stop to high frequency spiking)
5.	Clear post-ictal suppression that lasts ≥ 30 seconds (even if there is after discharge)

SUMMARY AND FUTURE DIRECTIONS

DFP causes a rapid induction of COX-2 in hippocampal neurons, rapid synthesis of PGE₂ and consequent activation of the EP2 receptor on neurons and microglia. The competitive EP2 antagonist, TG6-10-1, reduced morbidity in rats exposed to DFP, as manifested by reduced neuroinflammation, neurodegeneration, blood-brain barrier leakage, microgliosis, and accelerated weight regain within 4 days after DFP exposure. Inhibition of EP2 receptors in rats exposed to DFP also blocks long-term memory impairment without affecting anxiety behaviors or the intensity of the acute SE experience. Termination of SE after 1-2h by a subanesthetic dose of urethane reduced neuroinflammation, neurodegeneration and astrogliosis, accelerated weight regain within four days after DFP exposure, and reduced the incidence of epilepsy as well as the frequency of spontaneous recurrent seizures in epileptic rats. These beneficial effects of urethane are all attributed to the effective termination of electrographic SE. However, TG6-10-1 does not alter seizure characteristics and thus a somewhat trivial explanation for the beneficial effect of TG6-10-1 in the adult rat DFP model, that the compound simply aborts or truncates SE, has been ruled out. Future experiments should determine whether early urethane treatment ameliorates the consequences of SE by other OP-based agents. In the future, experiments may also determine whether a combination therapy of terminating electrographic SE with urethane and subsequent inhibition of the EP2 receptor completely eliminates consequences of OP-induced SE.

CONCLUSION

Delayed inhibition of EP2 receptors by TG6-10-1, administered beginning ≥ 80 min after DFP exposure, produces a number of benefits on intermediate and long-term consequences of SE. Together, these studies give insight into the consequences of OP-induces SE and therapeutic modalities for the inhibition of EP2 in OP-agent induced pathologies.

ABBREVIATIONS:

PGE₂

prostaglandin-E₂

EP2

prostaglandin-E₂ receptor 2

SE

status epilepticus

GFAP

glial fibrillary acidic protein

Iba1

ionized calcium binding adaptor molecule 1

CA1

Cornu Ammonis 1

CA3

Cornu Ammonis 3

CT

cycle threshold

Con

control

COX-2

cyclooxygenase 2

COX-1

cyclooxygenase 1

DFP

diisopropyl fluorophosphate

OP

organophosphorus compound

ip

intraperitoneal

sc

subcutaneous

FJB

FluoroJade B

AchE

acetylcholinesterase

qRT-PCR

quantitative real time polymerase chain reaction

GAPDH

Glyceraldehyde 3-phosphate dehydrogenase

IL-1β

interleukin -1β

TNFα

tumor necrosis factor alpha

CXCL10

C-X-C motif chemokine 10

CCL2

chemokine (C-C motif) ligand 2

CCL3

chemokine (C-C motif) ligand 3

CCL4

chemokine (C-C motif) ligand 4

IL-6

interleukin 6

HPRT1

hypoxanthine phosphoribosyltransferase 1

Veh

vehicle

mpk

mg/kg

ELISA

enzyme-linked immunosorbent assay

TG6-10-1

potent and selective EP2 receptor antagonist

SRS

spontaneous recurrent seizure

PID

post-ictal depression

AD

after discharge

Pgp

P-glycoprotein

NOR

novel object recognition

AEDs

anti-epileptic drugs

cAMP

cyclic adenosine monophosphate

BBB

blood-brain barrier

PF-04418948

Pfizer EP2 antagonist