Cutaneous challenge with chemical warfare agents in the SKH-1 hairless mouse (II): effects of some currently used skin decontaminants (RSDL and Fuller's earth) against liquid sulphur mustard and VX exposure.

Using the hairless mouse screening model presented in the companion paper(1) the aim of this study was to assess two skin decontaminating systems: Fuller's earth (FE) and Reactive Skin Decontamination Lotion (RSDL) against two extremely toxic chemical warfare agents that represent a special percutaneous hazard, sulphur mustard (SM) and O-ethyl-S-(2[di-isopropylamino]ethyl)methyl-phosphonothioate (VX). Five minutes after being exposed on the back to either 2 µL of neat sulphur mustard or 50 µg.kg(-1) of diluted VX, mice were decontaminated. Both systems were able to reduce blisters 3 days after SM exposure. However, RSDL was found to be more efficient than FE in reducing the necrosis of the epidermis and erosion. In the case of VX exposure, RSDL, whatever the ratio of decontaminant to toxicant used (RSDL 10, 20, 50), was not able to sufficiently prevent the inhibition of plasma cholinesterases taken as a surrogate marker of exposure and toxicity. Only FE reduced significantly the ChE inhibition. Some of these observations are different from our previous results obtained in domestic swine and these changes are thus discussed in the perspective of using SKH-1 hairless mice for the initial in vivo screening of decontaminants.

Cutaneous challenge with chemical warfare agents in the SKH-1 hairless mouse. (I) Development of a model for screening studies in skin decontamination and protection.

Exposure to lethal chemical warfare agents (CWAs) is no longer only a military issue due to the terrorist threat. Among the CWAs of concern are the organophosphorus nerve agent O-ethyl-S-(2[di-isopropylamino]ethyl)methyl-phosphonothioate (VX) and the vesicant sulfur mustard (SM). Although efficient means of decontamination are available, most of them lose their efficacy when decontamination is delayed after exposure of the bare skin. Alternatively, CWA skin penetration can be prevented by topical skin protectants. Active research in skin protection and decontamination is thus paramount. In vivo screening of decontaminants or skin protectants is usually time consuming and may be expensive depending on the animal species used. We were thus looking for a suitable, scientifically sound and cost-effective model, which is easy to handle. The euthymic hairless mouse Crl: SKH-1 (hr/hr) BR is widely used in some skin studies and has previously been described to be suitable for some experiments involving SM or SM analogs. To evaluate the response of this species, we studied the consequences of exposing male anaesthetized SKH-1 mice to either liquid VX or to SM, the latter being used in liquid form or as saturated vapours. Long-term effects of SM burn were also evaluated. The model was then used in the companion paper (Taysse et al.(1)).

Changes in mouse brain metabolism following a convulsive dose of soman: a proton HRMAS NMR study.

Soman, an irreversible organophosphorus cholinesterase inhibitor, induces status epilepticus and, in sensitive brain areas, seizure-related brain damage (e.g. brain edema and neuronal loss). The brain metabolic disturbances associated with these events are ill known. In the present study, we thus evaluated these changes in a murine model of soman-induced status epilepticus up to 7 days after intoxication. Mice, protected by HI-6 and atropine methyl nitrate, were poisoned with soman (172 microg/kg) and then sacrificed at set time points, from 1 h to 7 days. Brain biopsies from the piriform cortex (Pir) and cerebellum (Cer) were analyzed by 1H HRMAS NMR spectroscopy. Spectra were then analyzed using both a supervised multivariate analysis and the QUEST procedure of jMRUI for the quantification of 17 metabolites. The multivariate analysis clearly showed the metabolic differences between a damaged structure (Pir) and a structure with less prominent changes (cerebellum) and helped to globally assess the time course of metabolic changes. Analysis of the individual metabolites showed that the major changes took place in the piriform cortex but that cerebellum was not change-free. The most prominent changes in the former were an early (1-4 h) increase in alanine and acetate, a delayed increase in lactate, glycerophosphocholine and glutamine as well as a delayed decrease in myo-inositol and N-acetylaspartate. A week after poisoning, some metabolic disturbances were still present. Further research will be necessary to clarify what could be the involvement of these metabolites in physiological processes and how they might become useful surrogate markers of brain damage and repair.

Intrahippocampal cholinesterase inhibition induces epileptogenesis in mice without evidence of neurodegenerative events.

The mechanisms of epileptogenesis remain largely unknown and are probably diverse. The aim of this study was to investigate the role of focal cholinergic imbalance in epileptogenesis. To address this question, we monitored electroencephalogram (EEG) activity up to 12 weeks after the injection of a potent cholinesterase (ChE) inhibitor (soman) at different doses (0.53, 0.75, 1, 2, 2.8, 4 and 11 nmol) into the right dorsal hippocampus of C57BL/6 mice. Different parameters were used to choose the dose for a focal model of epileptogenesis (mainly electrographic patterns and peripheral ChE inhibition). The pattern of neuronal activation was studied by Fos immunohistochemistry (IHC). Brain damage was evaluated by hemalun-phloxin, neuronal nuclei antigen IHC and silver staining. Glial fibrillary acidic protein IHC was used to evaluate astroglial reaction. Finally, long-term behavioral consequences were characterized. At the highest dose (11 nmol), soman quickly evoked severe signs, including initial seizures and promoted epileptogenesis in the absence of tissue damage. With lower doses, late-onset seizures were evidenced, after 1-4 weeks depending on the dose, despite the absence of initial overt seizures and of brain damage. Only a weak astroglial reaction was observed. Following injection of 1 nmol, Fos changes were first evidenced in the ipsilateral hippocampus and then spread to extrahippocampal areas. A selective deficit in contextual fear conditioning was also evidenced two months after injection. Our data show that focal hypercholinergy may be a sufficient initial event to promote epilepsy and that major brain tissue changes (cellular damage, edema, neuroinflammation) are not necessary conditions.

Quantitation with QUEST of brain HRMAS-NMR signals: application to metabolic disorders in experimental epileptic seizures.

Quantitation of High Resolution Magic Angle Spinning (HRMAS) Nuclear Magnetic Resonance (NMR) signals enables establishing reference metabolite profiles of ex vivo tissues. Signals are often contaminated by a background signal originating mainly from macromolecules and lipids and by residual water which hampers proper quantitation. We show that automatic quantitation of HRMAS signals, even in the presence of a background, can be achieved by the semi-parametric algorithm QUEST based on prior knowledge of a metabolite basis-set. The latter was quantum-mechanically simulated with NMR-SCOPE and requires accurate spin parameters. The region of interest of spectra is a small part of the full spectral bandwidth. Reducing the computation time inherent to the large number of data-points is possible by using ER-Filter in a preprocessing step. Through Monte-Carlo studies, we analyze the performances of quantitation without and with ER-Filtering. Applications of QUEST to quantitation of 1H ex vivo HRMAS-NMR data of mouse brains after intoxication with soman, are demonstrated. Metabolic profiles obtained during status epilepticus and later when neuronal lesions are installed, are established. Acetate, Alanine, Choline and gamma-amino-butyric acid concentrations increase in the piriform cortex during the initial status epilepticus, when seizures are maximum; Lactate and Glutamine concentrations increase while myo-Inositol and N-acetylaspartate concentrations decrease when neuronal lesions are clearly installed.

An unexpected plasma cholinesterase activity rebound after challenge with a high dose of the nerve agent VX.

Organophosphorus chemical warfare agents (nerve agents) are to be feared in military operations as well as in terrorist attacks. Among them, VX (O-ethyl-S-[2-(diisopropylamino)ethyl] methylphosphonothioate) is a low volatility liquid that represents a percutaneous as well as an inhalation hazard if aerosolized. It is a potent irreversible cholinesterase (ChE) inhibitor that causes severe signs and symptoms, including respiratory dysfunction that stems from different mechanisms. VX-induced pulmonary oedema was previously reported in dogs but mechanisms involved are not well understood, and its clinical significance remains to be assessed. An experimental model was thus developed to study VX-induced cardiovascular changes and pulmonary oedema in isoflurane-anaesthetized swine. In the course of this study, we observed a fast and unexpected rebound of plasma ChE activity following inhibition provoked by the intravenous injection of 6 and 12 microg kg(-1) of VX. In whole blood ChE activity, the rebound could stay unnoticed. Further investigations showed that the rebound of plasma esterase activity was neither related to spontaneous reactivation of ChE nor to VX-induced increase in paraoxonase/carboxylesterase activities. A bias in Ellman assay, haemoconcentration or severe liver cytolysis were also ruled out. All in all, these results suggest that the rebound was likely due to the release of butyrylcholinesterase into the blood stream from ChE producing organs. Nature of the organ(s) and mechanisms involved in enzyme release will need further investigations as it may represent a mechanism of defence, i.e. VX scavenging, that could advantageously be exploited.

Delta activity as an early indicator for soman-induced brain damage: a review.

The organophosphorus (OP) compound soman is known to produce long-lasting epileptic seizure activity and associated brain damage. The present paper reviews the findings of five recent studies that tentatively established correlations between the development of soman-induced neuropathology and some subtle changes in the electrocortigraphic (ECoG) power spectrum. It is important to note that the reported experiments have been performed independently by three different teams (France, The Netherlands, USA) in various animal models (rat, guinea-pig, cynomolgus monkey) through different protocols of intoxication, pharmacological environments, and methods for ECoG spectral analysis. Despite these disparities, the five studies show that a suistained shift of ECoG power toward the lowest frequency range, i.e. the delta band, occurs within the first hours of soman-induced seizures. This early ECoG spectral change is concurrent with the first neuropathological changes in brain and is almost constantly followed, days or weeks later, by at least minimal neuropathology. Moreover the relative contribution of delta activity to the ECoG power spectrum still remains abnormally high for 1-3 days after seizure onset, i.e. within the phase of damage maturation. On the other hand, somnan-induced neuropathology was not observed in non-seizuring animals in which the delta activity was not increased above the pre-soman baseline. Similarly, no brain damage was ever shown in seizuring subjects in which the initial delta change eventually normalized after the curative administration of efficient anticonvulsant drugs such as the non-competitive antagonists of the NMDA receptor. These results, in agreement with previously published observations, strongly suggest that an increase of the relative power in the delta band might be a real-time marker of the ongoing development of soman-induced, seizure-related cerebral lesions and a reliable predictor for the final neuronal losses to come. Therefore, the monitoring of delta activity during the 24-72 h period that follows soman exposure may potentially be a useful tool to follow "on-line" the progression of brain damage and to control the neuroprotective activity of'a medication. Moreover since the method is non-invasive in man and since the above-presented results have been partly found in primates, the applicability of spectral analysis as a prognostic means in human OP poisoning ought to be seriously considered.

Subchronic administration of various pretreatments of nerve agent poisoning. I. Protection of blood and central cholinesterases, innocuousness towards blood-brain barrier permeability.

PYR, a reversible AChE inhibitor, is the current pretreatment against OP intoxication. However, PHY in the presence or absence of SCO on one side, and HUP on the other side, could be considered as potential substitutes for PYR. In the present study, the effects of the subchronic administration of these different current or potential pretreatments on the BBB permeability for blood-borne albumin and on the activity of the blood and central cholinesterases are comparatively evaluated in guinea-pigs. Altogether, although some marginal disruptions of BBB are detected, the different current or potential pretreatments studied seem to have a total innocuousness on the permeability of the BBB for proteins. Finally, at the light of its particular inhibitory effects on blood and central cholinesterases, HUP, compared to the other drugs, seems to be the optimal candidate to be used as pretreatment against OP poisoning.

Subchronic administration of various pretreatments of nerve agent poisoning. II. Compared efficacy against soman toxicity.

OP nerve agents, such as soman, are potent irreversible inhibitors of central and peripheral acetylcholinesterases. Pretreatment of OP poisoning relies on the subchronic administration of a reversible acetylcholinesterase inhibitor. In the present study, the protective effects against soman toxicity of such compounds i.e. pyridostigmine, physostigmine (alone or associated with scopolamine) or huperzine are compared in guinea-pigs instrumented for EEG recording. Each medication is given via a subcutaneous mini-osmotic pump for 6 days at a delivery rate providing about 30% maximal inhibition of red cell acetylcholinesterase activity. The animals then receive iterative injections of soman (1/3 LD50) every 10 min. With pyridostigmine, reflecting a decreased overall tolerance to the poisoning, the cumulative doses of soman producing either tremors and convulsions or seizures are lower than those found in non-pretreated intoxicated controls. On the other hand, physostigmine does not afford satisfactory protection against the early mortality after intoxication. On this specific point, physostigmine + scopolamine and huperzine, although they do not prevent the appearance of seizures, give best results. The effects of each pretreatment on acetylcholinesterase, butyrylcholinesterase and carboxylesterase (these two latter enzymes may act as endogenous scavengers of OP compounds) are also examined in vitro and in the blood of each animal during subchronic administration. Huperzine appears as a selective inhibitor of red cell acetylcholinesterase activity while pyridostigmine or physostigmine additionally inhibit plasmatic butyrylcholinesterase. Considerations about huperzine or physostigmine + scopolamine as the most appropriate candidate for the pretreatment of OP poisoning are given.

Effects of thienylphencyclidine (TCP) on seizure activity and brain damage produced by soman in guinea-pigs: ECoG correlates of neurotoxicity.

The capacity of thienylcyclohexylpiperidine (TCP), a non-competitive blocker of the N-methyl-D-aspartate (NMDA) receptor, to counteract the convulsant, lethal, and neuropathological effects of 2 x LD50 of soman (an irreversible inhibitor of cholinesterase) was investigated in guinea-pigs treated by pyridostigmine and atropine sulphate. The effects of a weak dose of TCP (1 mg/kg) used in the present study globally reproduced those previously obtained with a higher dose (2.5 mg/kg; [Neurotoxicology 15 (1994) 837]): TCP was again most protective when given curatively within the first hour of soman-induced seizures. In this condition, (a) paroxysmal activity ceased in 10-20 min, (b) all the animals survived, (c) the majority of them recovered remarkably well and did not show any brain damage 24 h after the intoxication, and (d) the minimal duration of seizure activity normally required for producing soman-induced brain damage in other pharmacological environments was increased from 10 to 40 min to 80 min. Strikingly, when TCP was given 120 min after seizure onset, it failed to show any anticonvulsant activity but still provided neuroprotection in the hippocampus. The present study also gives additional evidence (see [Neurotoxicology 21 (4) (2000) 521]) that in soman poisoning, (a) the development of brain damage depends on the occurrence of ECoG seizures, (b) the topographical distribution of lesions depends on seizure duration, and (c) an increase of the relative power in the lowest (delta) frequency band might be a reliable marker of neuronal degradation. All these findings confirm that (a) glutamatergic NMDA receptors are involved in the mechanisms of soman-induced seizures and brain damage, (b) non-competitive antagonists of NMDA receptors might be promising candidates for post-treatment of soman poisoning, and (c) ECoG parameters from ECoG tracings and power spectrum might serve as useful external predictors for soman-induced neuropathological changes.

Effects of atropine sulphate on seizure activity and brain damage produced by soman in guinea-pigs: ECoG correlates of neuropathology.

The present study describes the effects of pyridostigmine (PYR; 0.2 mg/kg) and atropine sulphate (AS; 5 mg/kg) on guinea-pigs intoxicated by a high dose (2xLD50) of the organophosphate compound, soman, an irreversible inhibitor of acetylcholinesterase. The medication was shown to counteract the acute respiratory distress and lethality normally produced by the intoxication. Moreover, due to the central activity of AS, soman-induced electrocorticographic (ECoG) seizure activity was either totally prevented, or reduced in duration and overall intensity. In addition, as established in the 24-hr survivors, seizure-related neuropathology was either prevented, or reduced in topographical extent and severity. An attempt to correlate our electrographic and morphological findings gives evidence that (a), the occurrence of seizure activity is the primary factor necessary for the development of acute neuropathology; (b), the duration of ECoG seizures is a secondary factor, on which the topographical distribution of brain damage finally depends; (c), the minimal duration of seizures necessary to produce 24 hr-damage in the most sensitive areas (e.g. the amygdala) is less than 70 min; (d), the overall intensity/power of epileptiform discharges is a tertiary factor which influences the severity of damage; (e), in addition, ECoG power spectral analysis suggested that an acute increase of relative power in the lower (delta) frequency band might be a real-time external marker of the starting cerebral lesions and is thus predictive for their future installation. All these data confirm the tight relationships which exist between seizure activity and neuropathology in soman poisoning, and suggest that refined, standardized analysis of electrographic parameters drawn from ECoG tracings and power spectrum might serve as a useful tool to predict the presence, localization, and severity of soman-induced brain damage.

Memory impairment after soman intoxication in rat: correlation with central neuropathology. Improvement with anticholinergic and antiglutamatergic therapeutics.

The effects of soman, a potent irreversible inhibitor of acetylcholinesterase, on central neuropathology in rats were studied in relation with subsequent spatial memory impairments. In a first step, it was found that, without treatment, neuropathology and learning impairment were observed only in rats which experienced convulsions. Then, treatment consisting of atropine sulfate, and/or TCP and/or NBQX was administered to intoxicated animals at infraanticonvulsant doses to obtain a graded subsequent neuropathology and to appreciate an eventual relation between neuropathology and spatial memory impairment. Thus, a correlation between neuropathology in the hippocampal CA1 region and spatial learning performance was found, the degradation of performance of rat being directly related to the amplitude of their neural damage. A threshold was emphasized : below a certain degree of neural loss, no memory impairment was found. Only treatment with tritherapy (atropine + TCP + NBQX) was able to improve the different parameters of spatial learning, despite no effect on the convulsions of the animals.

Seizure-related opening of the blood-brain barrier produced by the anticholinesterase compound, soman: new ultrastructural observations.

Previous macroscopic and light microscopic observations established that the organophosphate soman, an irreversible inhibitor of cholinesterases, produces seizure-related opening of the blood-brain barrier (BBB) to proteins. In Wistar rats, this BBB alteration was found to be reversible. This alteration was greatest during the first hour of seizures, and was topographically limited to sensitive areas such as the thalamus. In contrast, the hippocampus remained free of any vascular leakage. The present study is an attempt to elucidate, in rat thalamus, the subcellular mechanisms involved in soman-induced BBB alteration. A combination of three ultrastructural approaches was used: examination of ultra-thin sections, freeze-fracture, and post-embedding protein A-gold immunocytochemistry of the endogenous, normally exclusively blood-borne, albumin. Our findings show that soman-induced seizure activity produced no discernible structural change in the endothelial tight junctions, whereas it unambiguously increased the number of endothelial vesicles. Finally, immunolabelled albumin clearly crossed the endothelium, but was not systematically found inside the endothelial vesicles. Altogether, the present ultrastructural study confirms that soman can alter the integrity of the BBB, and demonstrates that the blood-to-brain passage of proteins does not mainly derive from the opening of tight junctions. Although transcytosis is clearly increased through the cerebral endothelium, there is little evidence that blood-borne proteins penetrate the brain in this way. The actual mechanisms of transport thus remain to be clarified.

Heat stress, even extreme, does not induce penetration of pyridostigmine into the brain of guinea pigs.

Stress due to forced swimming was recently shown to allow penetration of pyridostigmine (PYR) into the brain of mice. Accordingly, it was suggested that in troops exposed to emotional stress under conditions of war, as during the Gulf War, the BBB may have unexpectedly become permeable to PYR thus leading to an increased frequency of CNS symptoms. In this study, the entry of PYR into the brain was investigated in guinea pigs subjected to different heat stress levels. In a first group, guinea pigs were maintained at room temperature for 2 hours, their core temperature remaining stable at about 39.8 degrees C. In a second group, animals were placed in a climatic chamber in order to keep their core temperature at 41.5 degrees C for 2 hours. In a third group, animals were subjected to a high ambient temperature (42.6 degrees C) during about 2 hours and developed heatstroke symptoms, their core temperature progressively increasing and reaching around 44.3 degrees C. In each group, the stress of the animals was assessed by measuring the increase of plasma cortisol level. PYR (0.2 mg/kg, s.c.) was injected 90 minutes after beginning the experiment. Penetration of the drug into the brain was examined by measurement of acetylcholinesterase (AChE) activity in the cortex, the striatum and the hippocampus of the animals 30 minutes after PYR administration. A passage of this drug into the brain was also evaluated autoradiographically after i.v. injection of tritiated PYR 90 minutes after the beginning of the experiment (100 microCi/animal). Whatever the group examined, no entry of PYR into the CNS could be detected. Exposure to an ambient temperature at 42.6 degrees C for 2 hours resulted by itself in a partial inhibition of cerebral AChE activity. Our results, which agree with previous data obtained in humans exposed to heat stress, are opposite to the recent research showing a central passage of PYR in mice following a forced swim stress test. This demonstrated that the penetration of PYR into the brain of rodents under stress depends on the experimental conditions used (animal species, nature of the stressor, etc.). Extrapolations to humans of results primarily obtained in rodents about central passage of a drug under stress must thus be done very carefully.

c-fos antisense oligonucleotide prevents delayed induction of hsp70 mRNA after soman-induced seizures.

Previous studies have shown the successive expression of c-fos and hsp70 genes, especially in the hippocampal formation, during soman-induced seizures. In order to detect a possible link between the induction of these two genes, antisense strategies have been used. First, the ability of unilateral intrahippocampal infusion of c-fos antisense oligonucleotides to inhibit ipsilateral, seizure-related, c-FOS-like immunoreactivity, was verified. Second, induction of hsp70 mRNA was investigated using in situ hybridization. Unilateral inhibition of c-fos induction clearly reduced levels of hsp70 mRNA in the c-fos antisense-infused hippocampus relative to the non-infused contralateral side. Infusion of c-fos sense probe or vehicle did not affect bsp70 mRNA induction. This study suggests a role of c-FOS in regulating bsp70 mRNA induction.

Time course and regional expression of C-FOS and HSP70 in hippocampus and piriform cortex following soman-induced seizures.

The time course of induction of the proto-oncogene c-fos and the inducible heat shock hsp70 gene was studied from 5 minutes to 24 hours at both transcriptional (c-fos and hsp70 mRNA) and translational levels (C-FOS and HSP72 proteins) in the rat hippocampus and piriform cortex (Pir) after soman-induced seizures. Induction of c-fos was noticed as early as 5 minutes after seizures onset in all fields of hippocampal formation (CA1, CA3, CA4, and dentate gyrus) and in piriform cortex. The most intense induction was observed in piriform cortex. A sustained activation of c-fos occurred in Pir and in CA1, CA3, and CA4 areas of hippocampus. Nevertheless, histological analysis showed rare affected neurons in CA4, whereas damage was severe in Pir and in CA1 and CA3 hippocampal subfields. Induction of hsp70 mRNA occurred but was delayed in all areas previously exhibiting c-fos expression. Nevertheless HSP72 protein was never expressed in the structures where injury was high (i.e., CA1 and piriform cortex) and mainly occurred in the less damaged structure (i.e., CA4 area of hippocampus). Regional expression of glial fibrillary acidic protein mRNA was also studied in order to exclude an astroglial origin of the c-fos and hsp70 gene inductions. Our results demonstrated that after soman induced-seizures 1) there was no strict correlation between time course or intensity of neuronal c-fos induction and subsequent neuropathology, and 2) the most lesioned areas did not express HSP72 protein in spite of intense mRNA induction, suggesting that transcriptional and translational events for hsp70 gene might vary according to the severity of seizure insult.

The role of nitric oxide in soman-induced seizures, neuropathology, and lethality.

The effects of the inhibitors of endothelial and neuronal nitric oxide (NO) synthases, N-nitro-L-arginine methyl ester (L-NAME) and 7-nitroindazole (7-NI), respectively, and the precursor of NO, glyceryl trinitrate, on soman-induced seizures, lethality, and neuropathology were studied in rats. It was found that pretreatment of rats with L-NAME and 7-NI potentiated the severity of motor convulsions and enhanced lethality produced by soman. On the other hand, glyceryl trinitrate, administered transdermally at doses ranging from 2.5-5 mg/day 1 day before soman, decreased seizure susceptibility and lethality in soman-intoxicated animals. This was accompanied by a subsequent reduction of central neuronal damage 24 h after soman treatment. Pretreatment with glyceryl trinitrate also reversed seizure latency produced by 7-NI treatment during soman intoxication. These results indicate that neuronal NO may play a prominent role in seizures by acting as an anticonvulsant and neuroprotectant in soman intoxication.

Early regional changes of GFAP mRNA in rat hippocampus and dentate gyrus during soman-induced seizures.

We investigated the time course of GFAP levels in the hippocampal formation during the first 24 h following soman intoxication in rats. GFAP mRNA expression was detected by in situ hybridization. Intense GFAP mRNA expression was present in the molecular layer of the dentate gyrus as early as 6 h after intoxication. This expression was comparatively lower in other dentate gyrus layers and hippocampal CA1, CA3 and CA4 areas and seemed to be related to excessive neuronal activity. Histopathological examination demonstrated that GFAP expression in dentate gyrus is not correlated with lesions. The high astrocytic reactivity in the molecular layer of the dentate gyrus is discussed in relation to the maintenance of the homeostasis of glutamate and of synaptic plasticity in this area during soman intoxication.