Soman-induced seizures: limbic activity, oxidative stress and neuroprotective proteins.

Soman, a potent acetylcholinesterase inhibitor, induces status epilepticus in rats followed by conspicuous neuropathology, most prominent in piriform cortex and the CA3 region of the hippocampus. Cholinergic seizures originate in striatal-nigral pathways and with fast-acting agents (soman) rapidly spread to limbic related areas and finally culminate in a full-blown status epilepticus. This leads to neurochemical changes, some of which may be neuroprotective whereas others may cause brain damage. Pretreatment with lithium sensitizes the brain to cholinergic seizures. Likewise, other agents that increase limbic hyperactivity may sensitize the brain to cholinergic agents. The hyperactivity associated with the seizure state leads to an increase in intracellular calcium, cellular edema and metal delocalization producing an oxidative stress. These changes induce the synthesis of stress-related proteins such as heat shock proteins, metallothioneins and heme oxygenases. We show that soman-induced seizures cause a depletion in tissue glutathione and an increase in tissue 'catalytic' iron, metallothioneins and heme oxygenase-1. The oxidative stress induces the synthesis of stress-related proteins, which are indicators of 'stress' and possibly provide neuroprotection. These findings suggest that delocalization of iron may catalyze Fenton-like reactions, causing progressive cellular damage via free radical products.

Effects of hypoxia preconditioning on expression of metallothionein-1,2 and heme oxygenase-1 before and after kainic acid-induced seizures.

Global hypoxia preconditioning provides neuroprotection against a subsequent, normally damaging challenge. While the mechanistic pathways are unknown, changes in the expression of stress-related proteins are implicated. Hypoxia preconditioning attenuates the brain edema and neuropathology associated with kainic acid-induced status epilepticus in a protein synthesis-dependent manner when a kainic acid challenge is given up to one week post-preconditioning. Kainic acid initiates a glutamate-driven status epilepticus causing a Ca2+ and oxidative stress, resulting in injury to the piriform cortex and hippocampus. Stress-related gene expression [e.g. metallothioneins (MTs), heme oxygenase-1 (HO-1)] is enhanced during seizures in vulnerable brain areas, (e.g. piriform cortex). This study explores the effects of hypoxia preconditioning on expression of MT-1, MT-2 and HO-1 before and after kainic acid-induced seizures. Analysis of MT-1, MT-2 and HO-1 expression, through Western and Northern blotting, indicates that there is a variable pattern of induction and suppression of these two genes following hypoxia preconditioning alone as well as after kainic acid-induced seizures compared to non-preconditioned animals. These findings suggest that hypoxia preconditioning induces an adaptive response that prevents kainic acid seizure-associated neuropathology even when robust seizures occur. This may involve a variety of stress-related proteins, working in concert, each with their own individual expression profiles. Induction of this type of neuroprotection pharmacologically, or through preconditioning, will provide a better understanding of the stress response in brain.

Measurement of loosely-bound iron in brain regions using redox cycling and salicylate.

A sensitive iron assay was developed for measuring non-heme and loosely bound iron in regions of rat brain. The method is based on the salicylate trapping of hydroxyl radicals generated from ascorbate-driven redox cycling of Fe3+-EDTA. This assay has high sensitivity (about 20 nM) because of amplification obtained with redox-cycling and fluorescent detection of the salicylate hydroxylation product, 2,5-dihydroxybenzoate. The assay detects iron as Fe2+ and Fe3+ combined. Values of non-heme and loosely bound iron are given for three areas of cortex, caudate, hippocampus, thalamus and brainstem of the rat brain.

A global hypoxia preconditioning model: neuroprotection against seizure-induced specific gravity changes (edema) and brain damage in rats.

Hypoxia preconditioning states that a sublethal hypoxia episode will afford neuroprotection against a second challenge in the near future. We describe and discuss a procedure for the development of global hypoxia preconditioning in adult male Wistar rats, using a mildly hypoxic (9% O(2), 91% N(2)) atmospheric exposure of 8 h. The persistence of neuroprotection was analyzed using a kainic acid (KA) model of brain injury. Rats were challenged with KA (14 mg/kg, i.p.) on 1-14 days post-hypoxia. The effects of hypoxia preconditioning on seizure score, weight loss, brain edema and histopathology were assessed. Brain edema, predominantly of vasogenic origin, was measured 24 h after KA administration using a reproducible and quantitative method based on the specific gravities of tissue samples. A density gradient column (1.0250-1.0650 g/cm(3)) comprised of kerosene and bromobenzene was used to assess the presence of edema in regions involved in seizure initiation and propagation that are normally extensively damaged (i.e., piriform cortex and hippocampus). Specific gravities of tissues were calculated through extrapolation with known NaCl standards. We found that hypoxia preconditioning prevented the formation of edema in these brain regions when KA challenge was given 1, 3, and 7, but not 14 days post-hypoxia exposure. Furthermore, neuroprotection was observed in animals that had robust seizures. The described procedure may be used to examine the neuroprotective mechanisms induced by global hypoxia preconditioning against many subsequent challenges reflecting a variety of experimental models of brain injury, and will provide a better understanding of the brain response to hypoxia and stress.

Hypoxia preconditioning attenuates brain edema associated with kainic acid-induced status epilepticus in rats.

Kainic acid (KA)-induced seizures elicit edema associated with necrosis in susceptible brain regions (e.g., piriform cortex and hippocampal CA1 and CA3 regions). To test the hypothesis that hypoxia preconditioning protects against KA-induced edema formation, adult male rats were exposed to a 9% O2, 91% N2 atmosphere for 8 h. KA (14 mg/kg, i.p.) was administered 1, 3, 7, or 14 days later. Regional analysis of edema indicated that hypoxia exposure attenuated edema formation in piriform and frontal cortices and hippocampus when KA was given 1, 3, or 7 days later but not 14 days after hypoxia. Cycloheximide (2 mg/kg s.c.) given 1 h prior to hypoxia prevented the protective effect of hypoxia on KA-induced edema attenuation in the piriform cortex and hippocampus. Thus, hypoxic challenge induces a general adaptive response that protects against the seizure-associated pathophysiology, with no direct relationship to seizure intensity. This response may involve stress-related transcription factors and effector proteins.

Reactive oxidant species in piriform cortex extracellular fluid during seizures induced by systemic kainic acid in rats.

Kainic acid (KA) administered systemically to rats produces seizures and brain damage. We measured an increase in reactive oxidant species (ROS) during KA-induced seizures in the extracellular fluid (ECF) of the piriform cortex, a brain region known to be subsequently damaged. Intracerebral microdialysis samples were collected and assayed for isoluminol-dependent chemiluminescence before and after injection of KA (16 mg/kg, i.p.). Hydrogen peroxide (H2O2) concentrations were calculated from catalase-sensitive chemiluminescence, the difference between total and catalase-resistant chemiluminescence. During generalized tonic-clonic seizures, both total and catalase-resistant chemiluminescence increased significantly in samples from brain ECF. Catalase-resistant chemiluminescence, most likely produced by ascorbic acid, increased for a full hour during sustained seizure activity. H2O2 concentrations showed a trend towards elevation during seizures. Increased ROS suggest that oxidative stress occurs in brain ECF during sustained seizure activity.

Alterations in local cerebral glucose utilization produced by D3 dopamine receptor-selective doses of 7-OH-DPAT and nafadotride.

The D3 dopamine receptor, localized primarily in limbic brain areas, is a potential antipsychotic site. The effects of D3 receptor stimulation or blockade on neuronal activity were determined using the [14C]-2-deoxyglucose method. Freely-moving, adult, male, Sprague-Dawley rats were injected s.c. with saline, agonist 7-hydroxy-diphenylaminotetralin (7-OH-DPAT) (0.1 mg/kg), or antagonist l-nafadotride (1 mg/kg). These doses of 7-OH-DPAT and l-nafadotride are behaviorally active and are 10-fold lower than a dose producing significant in vivo occupancy of D2 receptors. The [14C]-2-deoxyglucose procedure was initiated 30 min after the administration of the test compound. The rate of local cerebral glucose utilization (LCGU) was determined by quantitative autoradiography. 7-OH-DPAT produced a significant increase in LCGU in the substantia nigra. l-Nafadotride produced significant increases in LCGU in several brain areas including the lateral preoptic area, lateral habenula, caudate, septal area, entorhinal cortex, and some thalamic and hypothalamic areas. These observations indicate that stimulation or blockade of D3 receptors alters LCGU and that produces a unique pattern of alterations in LCGU suggestive of potential antipsychotic activity.

Kainic acid causes redox changes in cerebral cortex extracellular fluid: NMDA receptor activity increases ascorbic acid whereas seizure activity increases uric acid.

Kainic acid (KA) causes seizures and extensive brain damage in rats. To study the effects of KA on the redox state in cerebral cortex extracellular fluid (ECF), ascorbic and uric acid concentrations were measured in intracerebral microdialysis samples before and after systemic KA administration (ip). During seizures, concentrations of ascorbic and uric acid increased 500 and 100%, respectively. When midazolam was given with KA to prevent seizures, ascorbic acid still increased 400%, but uric acid increased only transiently. When the NMDA receptor antagonist aminophosphonovaleric acid (APV) was included in the microdialysis perfusion media, ascorbic acid levels decreased during baseline perfusion in a concentration-dependent manner. APV then suppressed the KA-induced increase in ascorbic acid levels, without blocking seizure activity. In summary, increased uric acid levels in brain ECF activity after KA administration are related to the induced seizure, but ascorbic acid levels are associated with NMDA receptor activity.

Recovery of metabolic activity in retinofugal targets after traumatic optic nerve injury is independent of retinofugal input.

Traumatic injury of the adult optic nerve causes a progressive degeneration of retinal ganglion cells. Despite this ongoing degeneration, a partial recovery of visual behavioral function and of local cerebral glucose use (LCGU) has been observed. To evaluate whether this partial recovery of LCGU is due to a recovery of visual conductance (extrinsic) or intrinsic neuronal activity, visual stimulation alone and combined with physostigmine,an acetylcholinesterase inhibitor, were used to activate the retinofugal pathway. LCGU was determined in 30 male adult rats with or without physostigmine treatment 2 or 9 days after crush or 8 days after cut of the right optic nerve. Analysis of LCGU in contralateral first-order projection areas revealed no differences 8 days after cut and 9 days after optic nerve crush. Furthermore, LCGU in the contralateral areas could not be stimulated by the treatment with physostigmine. We therefore conclude that the increase in LCGU from 2 to 9 days after crush is not due to a recovery in the conductance of visual input. We hypothesize a relief of an injury-dependent active suppression (diaschisis) of LCGU. This reversal of diaschisis may, in part, account for the return of visual functions after mild optic nerve injury.

Cerebral penetration injury leads to H2O2 generation in microdialysis samples.

Delayed tissue damage is proposed to be caused by reactive oxygen species. We investigated the effects of microdialysis probe penetration into rat piriform cortex on hydrogen peroxide (H2O2) in brain extracellular fluid (ECF). H2O2 decreased immediately after probe insertion into the brain, but increased over 300% in samples within minutes after collection. We assessed H2O2 changes in vitro in microdialysis perfusion media containing various ascorbic acid concentrations and confirmed ascorbic acid is a source of H2O2. We conclude that decreased H2O2 concentrations in perfusion media as it passes through the brain reflect an extracellular antioxidant effect, whereas the increase in H2O2 with time after sample collection indicates that H2O2 generating substances are present in ECF. Thus, the potential for producing reactive oxygen species in brain ECF exists following penetration injury, especially if transition metals are released into the neuronal microenvironment.

Redox changes in perfusates following intracerebral penetration of microdialysis probes.

Microdialysis probe insertion into rat cerebral cortex significantly affects the levels of redox-active substances in brain extracellular fluid. Ascorbic acid levels are high immediately after probe insertion, decline rapidly, and then rise as the rat recovers from anesthesia 5-8 hours after surgery. Uric acid is at a low level for 5 hours and then rapidly increases in parallel with ascorbic acid. High ascorbic acid levels immediately after probe insertion are likely due to a shift from intracellular to extracellular fluids, whereas the delayed increase in uric acid may be due to increased enzymatic formation. After removal from the brain, hydrogen peroxide (H2O2) in microdialysis samples produces catalase-sensitive oxidative chemiluminescence. Microdialysis samples also produce high level catalase-resistant chemiluminescence associated with ascorbic acid levels after penetration injury. Although ascorbic acid is likely an antioxidant at concentrations estimated to be in brain extracellular fluid, it may have prooxidant effects when complexed with transition metals released into the neuronal microenvironment during traumatic brain injury.

Nitric oxide detection with intracerebral microdialysis: important considerations in the application of the hemoglobin-trapping technique.

Nitric oxide (NO.) is involved in processes such as neurotransmission, memory, brain injury, vessel relaxation, etc. To study the functional and pathological roles of NO. in the brain, a reliable method to monitor NO. directly is needed. Since oxyhemoglobin (Hb) has a high affinity for NO. and upon binding is converted quantitatively to methemoglobin (MetHb), spectrophotometry of Hb conversion to MetHb can give a credible measurement of NO. concentration. Although this method is especially promising for in vivo microdialysis, factors can influence the reproducibility and stability, making it difficult to obtain reliable results at low NO. levels. Evaluation of the diffusion rates of NO. and sodium nitroprusside across the microdialysis membrane indicates that NO. readily diffuses through the membrane. By taking into account protein degradation and Hb autoxidation as well as integrating the difference spectra, this assay has a practical differential detection limit of about 7 nM (0.4 pmol) in vivo. We evaluated this method in anesthetized and awake rats by measuring the release of NO. induced by the excitotoxin kainic acid (13 mg/kg, i.p.). A protocol with detailed analytical parameters for NO. monitoring in neurobiological research is given.

Loss and subsequent recovery of local cerebral glucose use in visual targets after controlled optic nerve crush in adult rats.

A mild crush of the adult rat optic nerve serves as a model to study the restoration of function after traumatic brain injury. It causes a progressive degeneration of retinal ganglion cells, but visually guided behavior is partially restored within 2-3 weeks. The purpose of this study was to determine to what extent local cerebral glucose use (LCGU) decreases and if it recovers in retinofugal targets following unilateral optic nerve crush. At intervals of 2, 9, and 22 days after crush, LCGU was monitored in rats in which the visual system was stimulated by a strobe-light and pattern. In the ipsilateral retinofugal targets there was only a minimal loss of LCGU use, but in the contralateral retinofugal targets, LCGU was reduced at Postlesion Day 2: to 50% in the superior colliculus (SC), to 60% in the lateral geniculate nucleus of the thalamus (LGN), and to 87% in the visual cortex. On Postoperative Days 9 and 22 we observed a partial restoration of LCGU in the contralateral SC and LGN to 68 and 79%, respectively. As recovery of visual performance is known to follow a similar time course, we conclude that restoration of metabolic activity in target structures may contribute to the restoration of vision after optic nerve crush.

Temporalspatial patterns of expression of metallothionein-I and -III and other stress related genes in rat brain after kainic acid-induced seizures.

Kainic acid-induced seizures in the rat brain cause severe brain damage that is thought to result, in part, from oxidative stress. In this study, we examine the consequences of systemic administration of kainic acid on expression of several genes that encode proteins thought to play roles in protection from oxidative stress, including metallothionein-I, and -III. Kainic acid causes an increase in metallothionein-I and heme oxygenase-I mRNAs, as well as an increase in c-fos, heat shock protein-70, and interleukin-1 beta mRNAs. The induction of these mRNAs is seizure dependent, and is greater in brain areas with extensive damage (e.g. piriform cortex) than in areas with minimal damage (e.g. frontal cortex and cerebellum). In contrast, little or no change in mRNA for metallothionein-III, manganese superoxide dismutase, copper-zinc superoxide dismutase, glutathione-s-transferase ya subunit or glutathione peroxidase occur. The prolonged and robust concordant induction of the metallothionein-I and heme oxygenase-I genes may reflect the oxidative stress produced by kainic acid-induced seizures. In addition, the induction of interleukin-1 beta gene expression suggests an inflammatory response in brain regions damaged by kainic acid-induced seizures. Delineating the regulation of genes associated with oxidative and inflammatory responses can contribute to a fuller understanding of seizures and associated brain damage.

Patterns of glucose use after bicuculline-induced convulsions in relationship to gamma-aminobutyric acid and mu-opioid receptors in the ventral pallidum--functional markers for the ventral pallidum.

Bicuculline-induced convulsions increased glucose use throughout the brain and sharply demarcated the ventral pallidum and globus pallidus. Glucose use in the nucleus accumbens also increased after bicuculline-induced convulsions, except for a circumscribed region in the dorsomedial shell. Since the projection from the nucleus accumbens to the ventral pallidum contains gamma-aminobutyric acid (GABA) and the opioid peptide, enkephalin, the pattern of increased glucose use in the ventral pallidum and nucleus accumbens after bicuculline-induced convulsions was compared to the topography of GABAA and mu-opioid receptors. The pattern of glucose use in the nucleus accumbens and ventral pallidum resembled the topography of GABAA, but differed from that of mu-opioid receptors. Bicuculline may disinhibit GABAergic efferents to the ventral pallidum resulting in a dramatic increase in glucose use within striatopallidal synaptic terminals as well as in local terminals of the pallidal projection neurons.

The osmotic/calcium stress theory of brain damage: are free radicals involved?

This overview presents data showing that glucose use increases and that excitatory amino acids (i.e., glutamate, aspartate), taurine and ascorbate increase in the extracellular fluid during seizures. During the cellular hyperactive state taurine appears to serve as an osmoregulator and ascorbate may serve as either an antioxidant or as a pro-oxidant. Finally, a unifying hypothesis is given for seizure-induced brain damage. This unifying hypothesis states that during seizures there is a release of excitatory amino acids which act on glutamatergic receptors, increasing neuronal activity and thereby increasing glucose use. This hyperactivity of cells causes an influx of calcium (i.e., calcium stress) and water movements (i.e., osmotic stress) into the cells that culminate in brain damage mediated by reactive oxygen species.

Lithium modifies convulsions and brain phosphoinositide turnover induced by organophosphates.

Inositol-1-phosphate (Ins1P), an index of phosphoinositide (PI) turnover, was measured in frontal and piriform cortices, caudate, thalamus, hippocampus and cerebellum in saline or LiCl (5 m Eq./kg) pretreated rats 60 min. after graded doses of DFP, paraoxon, or soman. DFP only produced bursts of convulsive activity whereas both paraoxon and soman produced prolonged tonic-clonic convulsions. All three organophosphates (OP) produced convulsions at a lower dose in LiCl than in saline pretreated rats. Regional Ins1P correlated better with the presence or absence of convulsions than with the dose of paraoxon or soman. This was true both in saline and LiCl pretreated rats. In saline pretreated non-convulsing rats, there was a cholinergic increase (1.5-2.0 X) in Ins1P in all brain regions except cerebellum after OP injection. In saline pretreated convulsing rats, there was a marked seizurogenic further increase in Ins1P; highest in caudate (8 X) and cortex (6 X). In LiCl pretreated nonconvulsing rats, the OP-induced cholinergic increase in Ins1P was significant only in caudate, thalamus and hippocampus. In LiCl pretreated convulsing rats, the further seizurogenic increase in Ins1P was less than in saline pretreated rats except in thalamus and hippocampus. Thus, OP produce both a cholinergic and a seizurogenic increase in PI turnover. These data suggest that increased PI turnover in the hippocampus may indicate a lithium-induced lowering of the seizure threshold for OP in limbic regions.

Changes in cerebral inositol-1-phosphate concentrations in LiCl-treated rats: regional and strain differences.

LiCl-induced (5 mEq/kg) regional differences in the cerebral phosphoinositide (PI) cycle were studied by measuring inositol-1-phosphate (Ins-1-P), an intermediate in the PI cycle, in male Sprague Dawley and Han/Wistar rats by gas chromatography/mass spectrometry. Control Ins-1-P levels were higher frontally than caudally in both rat strains. LiCl increased Ins-1-P levels 1.8 to 7.4 fold in different regions of brain of Sprague Dawley rats but only 1.2 to 1.8 fold in Han/Wistar rats. This strain difference offers a way to compare the effects of lithium on PI metabolism versus receptor-G protein-phospholipase C coupling mechanisms.

Effects of microdialysis on brain metabolism in normal and seizure states.

The effect of intracranial microdialysis on brain glucose metabolism in control and kainic acid-treated rats was assessed by semi-quantitative [14C]2-deoxyglucose autoradiography. A dialysis fiber loop was implanted into the piriform cortex or a horizontal Vita fiber into the hippocampus, and 24 h later, fibers were perfused with Krebs-Ringer bicarbonate solution before and after injection of kainic acid (16 mg/kg, i.p.) [14C]2-Deoxyglucose was injected i.p. 3 h after the injection of kainic acid. Rats injected with kainic acid were initially lethargic and then proceeded through behavioral phases of staring, "wet-dog shakes", Straub tail, rearing, forepaw clonus, and, in some cases, tonic-clonic convulsions. Three hours after kainic acid, the fiber presence in the piriform cortex enhanced kainic acid-induced metabolic activity in areas adjacent to the fiber assembly, whereas the fiber in hippocampus attenuated kainic acid-induced metabolic activity in areas adjacent to the fiber assembly. The results indicate that intracranial microdialysis alters the already abnormal brain metabolism in a kainic acid-induced seizure state, but has no significant effect in the non-seizure control state.