Blast exposure causes redistribution of phosphorylated neurofilament subunits in neurons of the adult rat brain.

There is little information on threshold levels and critical time factors for blast exposures, although brain damage after a blast has been established both clinically and experimentally. Moreover, the cellular pathophysiology of the brain response is poorly characterized. This study employs a rat model for blast exposure to investigate effects on the neuronal cytoskeleton. Exposure in the range of 154 kPa/198 dB or 240 kPa/202 dB has previously been shown neither to cause visual damage to the brain, nor to affect the neuronal populations, as revealed with routine histology. Here, the brains were investigated immunohistochemically from 2 h to 21 days after blast exposure. A monoclonal antibody was used which detects only the phosphorylated epitope of the heavy subunit of the neurofilament proteins (p-NFH). This epitope is normally restricted to axons, that is, not demonstrable in the perikarya. Eighteen hours after exposure in the 240-kPa/202-dB range, p-NFH immunoreactivity accumulated in neuronal perikarya in layers II-IV of the temporal cortex and of the cingulate and the piriform cortices, the dentate gyrus and the CA1 region of the hippocampus. At the same time, the p-NFH immunoreactivity disappeared from the axons and dendrites of cerebral cortex neurons. The most pronounced immunostaining of neuronal perikarya was found in the hemisphere, which faced the blast source. The perikaryal accumulation of p-NFH was present also at 7 days but the neuronal perikarya had become negative at 21 days, at which time the axons again displayed p-NFH immunoreactivity. Exposure in the range of 154 kPa/198 dB caused similar, although less marked accumulation of p-NFH immunoreactivity in the neuronal perikarya. The findings are interpreted to show a dephosphorylation of NFHs in axons and dendrites and a piling up of p-NFHs in the perikarya due to disturbed axonal transport.

Quantitative immunochemistry on neuronal loss, reactive gliosis and BBB damage in cortex/striatum and hippocampus/amygdala after systemic kainic acid administration.

Cell specific markers were quantified in the hippocampus, the amygdala/pyriform cortex, the frontal cerebral cortex and the striatum of the rat brain after systemic administration of kainic acid. Neuron specific enolase (NSE) reflects loss of neurons, glial fibrillary acidic protein (GFAP) reflects reactive gliosis, and brain levels of serum proteins measures blood-brain-barrier permeability. While the concentration of NSE remained unaffected in the frontal cerebral cortex and the striatum, their GFAP content increased during the first three days. In the hippocampus and amygdala, NSE levels decreased significantly. GFAP levels in the hippocampus were unaffected after one day and decreased in the amygdala/pyriform cortex. After that, GFAP increased strikingly until day 9 or, in the case of amygdala/pyriform cortex, even longer. This biphasic time course for GFAP was accompanied by a decrease of S-100 during days 1-9 followed by a significant increase at day 27 above the initial level. The regional differences in GFAP and S-100 could result from the degree of neuronal degeneration, the astrocytic receptor set-up and/or effects on the blood-brain barrier.

Neuronal glycolytic pathway impairment induced by HIV envelope glycoprotein gp120.

Neurological impairment is a common feature of Acquired Immunodeficiency Syndrome (AIDS); functional alterations have been reported both in central and peripheral nervous system and the Human Immunodeficiency Virus (HIV) envelope glycoprotein gp120 has been proposed as a neurotoxin acting through a calcium-dependent mechanism. On the other hand it has been reported that gp120 treatment also induce about a 20% decrease in the cerebral glucose utilization and in the cellular ATP levels. The reported observations were performed on experimental system where also non-neuronal cells where present; in order to evaluate whether a direct interaction between HIV proteins and neuronal cells takes place, we used a neuroblastoma cultures where only neuronal cells are present. We analysed the effects of gp120 on the N18TG2 neuroblastoma clone. Treatments were performed both on growing and confluent cultures. Short time treatment with gp120 of confluent cultures causes a 25% reduction in the level of neuron-specific enolase, resulting in a similar decrease of oxygen consumption. Long time exposure of growing cells also causes a reduction in cell survival. Furthermore, using a membrane-specific fluorescent probe we observed that gp120 produces an increase of membrane trafficking. These observations suggest a direct interaction between the viral envelope protein and neuronal cells, which results in an alteration of glycolytic metabolism. This alteration may be related to the neurologic impairments observed in AIDS patients.

Elevated S100 blood level as an early indicator of intraventricular hemorrhage in preterm infants. Correlation with cerebral Doppler velocimetry.

The aim of this study was to assess the use of S100 protein in blood as a means of identifying preterm infants at risk of intraventricular hemorrhage. In 25 preterm newborns, S100 blood concentrations were measured by an immunoradiometric assay during the first 48 h. Cerebral Doppler velocimetry waveform patterns were also tested at the time the blood sample was taken, when clinical and cerebral ultrasound scanning were still normal. Of the 25 newborns studied, 14 were controls and 11 developed intraventricular hemorrhage as revealed by ultrasound scanning more than 72 h after birth, and clinically confirmed by neurological examination on the seventh day of follow-up. S100 blood concentrations were significantly higher (P<0.002) in infants with intraventricular hemorrhage than in control infants and also correlated significantly (r=0.81, P<0.003) with the grade of hemorrhage. A significant correlation (r=0.70, P<0.05) between the S100 blood concentration and the middle cerebral artery pulsatility index was also observed. The present data show that S100 blood concentrations offer a measurable parameter of brain lesion in preterm infants before a radiological assessment of hemorrhage can be performed, when clinical symptoms may be silent and preventive/therapeutic action could be especially useful.

Bcl-2 expression regulates cell sensitivity to S100beta-mediated apoptosis.

The S100beta protein is overexpressed in the brain of patients with Alzheimer's disease and Down's syndrome and is able to induce apoptosis in neurons at high concentrations. The intracellular events that regulate the apoptotic effect are largely unknown. This study investigates the roles of the bcl-2 proto-oncogene, one of the best-defined apoptotic genes, on cell death induced by S100beta. Human neuronal precursor NT2/D1 cells showed a high degree of cell death by apoptosis after exposure to 2 microM S100beta in serum-free medium. Death was preceded by a down-regulation of the Bcl-2 protein. Gene transfer with a full-length bcl-2 cDNA under the control of a constitutive promoter in NT2 cells elevated Bcl-2 protein levels and repressed S100beta-mediated cell death. When exposed to retinoic acid, the NT2/D1 cells differentiated into a neuronal phenotype. The differentiated cells up-regulated their levels of Bcl-2 and became resistant to S100beta-induced cell death. Downregulation of Bcl-2 by an antisense oligonucleotide in the differentiated cells, however, increased their susceptibility to S100beta-related cytotoxicity. Therefore, apoptosis induced through S100beta signaling is subject to regulation by Bcl-2. A combined alteration such as up-regulation of S100beta together with down-regulation of Bcl-2 may be important in the pathogenesis of Alzheimer's disease and Down's syndrome.

An acquired sensitivity to H2O2-induced apoptosis during neuronal differentiation of NT2/D1 cells.

Brief exposure of neuronally differentiated human NT2/D1 cells to hydrogen peroxide induced cell death by apoptosis with an ED50 of 30 microM, whereas a 70-fold higher concentration was required to obtain an ED50 effect in undifferentiated NT2/D1 neuronal precursor cells. This enhanced sensitivity in NT2/D1 neurons was correlated with an 8-fold lower level of intracellular glutathione. Pretreatment with N-acetyl-L-cysteine, an agent that is able to raise the levels of intracellular glutathione, promoted the survival of hydrogen peroxide-treated NT2/D1 neurons. Thus, the low glutathione level may contribute to the high sensitivity of NT2/D1 neurones to hydrogen peroxide-induced apoptosis. This study indicates that neuronal susceptibility to oxidative damage is developmentally regulated.

S-100beta stimulates neurite outgrowth in the rat sciatic nerve grafted with acellular muscle transplants.

S-100beta promotes neurite extension in vitro and motoneuron survival in the chicken embryo. We demonstrate here that local administration of S-100beta stimulates the sciatic nerve regeneration into acellular muscle grafts. Normally there is a 8-10 day delay in the regeneration of axons into such grafts. Local administration of S-100beta (0.5-1.0 microg/h) significantly stimulated regeneration into the grafts. In S-100beta treated grafts, the regeneration distance was increased with a factor of about 2.3 times as compared to vehicle treated grafts. The distance of regeneration was monitored with pinch test which detects sensory axons. Regenerating axons were growing outside the necrotic muscle cells as revealed with immunohistochemistry for the neurofilament light weight polypeptide. S-100beta was demonstrated immunocytochemically in motor neurons of the rat lumbar spinal cord and in large and medium sized neurons of the dorsal root ganglia. The results suggest that S-100beta is a physiological growth factor for peripheral nerve axons.

Delayed decrease of calbindin immunoreactivity in the granule cell-mossy fibers after kainic acid-induced seizures.

Kainic acid (KA) administration induces an abnormal excitation and spontaneous recurrent seizures. Alterations of granule cell properties may be potential mechanisms. In this study, dynamic alterations of calbindin, a calcium binding protein particularly abundant in the granule cells, have been investigated immunocytochemically in the rat hippocampus after the KA-induced seizures. The calbindin immunoreactivity decreased slightly in the CA1/CA2 fields already after 1 and 3 days, and was lost partly or completely in the pyramidal layer after 10 days. From day 21, the calbindin immunoreactivity decreased in dendrites and soma of the granule cells and mossy fibers. The alterations remained at least to day 90, while no evident neuronal loss occurred in the granule cells. This may reflect a disturbance of calcium homostasis in the granule cells after seizures. The delayed decrease of calbindin has a time course similar to the occurrence of spontaneous recurrent seizures, suggesting a possible correlation between the two events.

Presence of S-100 beta in cholinergic neurones of the rat hindbrain.

The double staining of S-100 beta and choline acetyltransferase (ChAT) revealed that S-100 beta immunoreactivity was localized in most, but not in all, cholinergic neurones in the somatomotor nuclei of the cranial nerves and in the ambiguus nucleus. S-100 beta was present in almost all cholinergic neurones in the brain stem reticular, red, vestibular (excluding medial), mesencephalic trigeminal and cerebellar nuclei. However, S-100 beta immunoreactivity was lacking in cholinergic neurones in the parabrachial complex, the dorsal motor nucleus of the vagal nerve and most sensory nuclei. No S-100 beta-positive neurones lacked ChAT immunoreactivity. Taken together with the fact that the vulnerability of motoneurones to axotomy is markedly reduced in the first 3 postnatal weeks, during which period neuronal S-100 beta appears and increases, a possible effect of S-100 beta on the survival of cholinergic motoneurones may be suggested.

Eliprodil prevents expression of the 70 kDa heat shock protein in MK801-injured neurones.

The present study examined whether eliprodil (SL 82.0715), an N-methyl-D-aspartate (NMDA) receptor antagonist acting on the polyamine sites induced expression of the 70 kDa heat shock protein (HSP70) in the rat brain. Whereas the NMDA channel blocker MK801 consistently induced HSP70 in posterior cingulate and retrosplenial cortices, eliprodil had no such effects even at the highest dose (50 mg/kg, intraperitoneally), supporting the idea that injury to the cerebrocortical neurones by NMDA receptor antagonists is probably related to specific sites of the receptor. Furthermore, eliprodil, given immediately after injection of MK801, blocked the effects of MK801 on HSP70. The result is discussed in terms of high affinity of eliprodil for the sigma receptor.

Plasticity of granule cell-mossy fiber system following kainic acid induced seizures: an immunocytochemical study on neurofilament proteins.

Abnormal reestablishment of mossy fibers with the CA3 pyramidal cells and granule cells is an important aspect of postlesional plasticity in epilepsy. However, basis for the structural reorganisation and functional consequences of the event remain uncertain. Therefore we have investigated alterations of neurofilaments, major cytoskeletal components of neurons, in the rat hippocampus after the kainic acid (KA) administration, an experimental model for the temporal lobe epilepsy. The immunoreactivity for phosphorylated heavy weight neurofilament (pNFH) and non-phosphorylated heavy weight neurofilament (npNFH), in particular the pNFH, decreased in the CA1 field and inner molecular layer of the dentate gyrus during 3 and 10 days after the KA administration. After 10 days, npNFH immunoreactivity appeared in the mossy fibers, in which it is normally absent, meanwhile the pNFH staining in the mossy fibers did not decrease. From day 21, the immunoreactivity of pNFH and npNFH was normal or above normal in the CA1 stratum lacunosum-moleculare, mossy fibers, hilus and inner molecular layer of the dentate gyrus. These alterations in the later phase remained at least to day 90. The reappearance and increase of the neurofilament immunoreactivity in the inner molecular layer of the dentate gyrus probably reflects a collateral extension of the granule cell axons known as mossy fiber sprouting. The results suggest that neurofilament changes in the granule cell-mossy fiber system may be a morphological basis for the structural reconstruction of granule cell axons, and neurofilaments are involved in the plasticity after the KA induced seizures.

Appearance of neuronal S-100 beta during development of the rat brain.

In addition to being an astroglial protein, S-100 beta is localised in distinct populations of neurons in the adult rat hindbrain. We report, here, the expression of S-100 beta in both neurons and glia of the rat brain during development. Prenatally, S-100 beta immunoreactivity was confined to glial cells close to the germinal zone. After birth, S-100 beta positive glial cells were seen mainly in the brainstem and cerebellum, while only a few were detected in cerebral cortex and hippocampus. The number of S-100 beta containing glial cells increased steadily during the first 2 postnatal weeks after which the adult pattern was attained. No S-100 beta containing neurons were present prenatally. The first S-100 beta containing neurons were seen in the mesencephalic trigeminal nucleus at postnatal day 1 (P1), and in the motor trigeminal nucleus at P3. Neuronal S-100 beta immunoreactivity in other nuclei was mostly attained from the 10th to the 21st postnatal day. The neuronal S-100 beta immunoreactivity was first detected in the cell nuclei during development, then increased in the cytoplasm with ages. A nuclear staining in many immunoreactive neurons persisted until the adult. It usually took 1 to 2 weeks for neuronal S-100 beta to attain the adult staining pattern, i.e., heavy staining of the cytoplasm and processes, after its appearance. The forebrain never contained S-100 beta positive neurons. The S-100 beta is first expressed in glial cells, suggesting it is primarily of the glial origin. Coupled with neurotrophic effects of the protein, the time course of neuronal S-100 beta expression during the critical period of neuronal development implies that it may be involved in neuronal differentiation and maturation.

S-100 beta immunoreactivity in neurones of the rat peripheral sensory ganglia.

S-100 beta, which is capable of exerting neurotrophic effects on cultured neurones and promoting the survival of motor neurones in vivo, has recently been found in distinct neurones of the rat hindbrain. Here we report that S-100 beta, as well as being present in satellite and Schwann cells, is also present in neurones of sensory ganglia (dorsal root ganglion, trigeminal, petrosal, jugular and nodose ganglia) but absent from neurones of the superior cervical ganglion. In the sensory ganglia, many neurones were immunoreactive, while the staining intensity varied among the neurones. Neuronal S-100 beta appeared in developing rats as early as postnatal day 1. No immunoreactive neurones were observed in the superior cervical ganglion during development. The results are suggestive of selective neurotrophic effects of S-100 beta.

S-100 beta has a neuronal localisation in the rat hindbrain revealed by an antigen retrieval method.

The localisation of S-100 in mammalian CNS neurons has been under debate for more than two decades. We address the question with two polyclonal and two new monoclonal antibodies. The specificity and the distribution in rat brain is based on an antigen retrieval method. We present evidence that aldehyde fixatives mask S-100 beta in neurons, and that the immunoreactivity is retrieved after trypsinisation. Neuronal S-100 beta is also detected in unfixed and ethanol fixed sections. The neuronal immunoreactivity is partly solubilised from unfixed tissue sections with 2.5 mM EDTA and is completely extracted with 2.5 mM EDTA and 1% Triton X-100. Most of the glial S-100 beta is washed out from unfixed tissue sections with saline. S-100 beta has distinct distribution in neurons of the hindbrain, i.e., the brainstem and cerebellum, but is not observed in the forebrain. One of the monoclonal antibodies immunostained neither neurons nor glia when it had been absorbed with S-100 crosslinked to nitrocellulose membranes. The distribution of neuronal S-100 beta differed from that of other neuronal calcium binding proteins, such as calbindin and parvalbumin. It was confined mainly to cholinergic neurons of the hindbrain. The presence of S-100 beta in distinct neuronal populations may indicate neurotrophic effects of S-100 beta. The notion is supported by the capability of S-100 to cause neurite outgrowth in vitro.

Phosphorylated and non-phosphorylated neurofilament proteins: distribution in the rat hippocampus and early changes after kainic acid induced seizures.

The regional distribution of neurofilament proteins in the rat hippocampus and their early changes after kainic acid induced seizures were investigated immunocytochemically with antibodies against light weight neurofilament, phosphorylated and non-phosphorylated heavy weight neurofilament. The light weight and non-phosphorylated heavy weight neurofilaments were distributed more unevenly than the phosphorylated neurofilament. The perikarya and processes of pyramidal cells in the CA3 field contained the highest light weight and non-phosphorylated heavy weight neurofilaments, while the perikarya of granule cells contained only few light weight neurofilament and the perikarya of CA1 pyramidal cells were even devoid of immunoreactivity of both light and heavy weight neurofilaments. The fiber staining of the light weight and non-phosphorylated heavy weight neurofilaments, especially the former, was less in the CA1 field and molecular layer of dentate gyrus. The phosphorylated neurofilament immunoreactivity was identified only in axons. Mossy fibers, the axons of granule cells, contained the light weight and phosphorylated heavy weight neurofilaments, but not the non-phosphorylated neurofilament. Seven days after the kainic acid induced seizures, the phosphorylated neurofilament staining was greatly reduced in the CA1 and inner molecular layer of the dentate gyrus, probably resulting from the axonal degeneration of the Schaffer collaterals and the commissural/associational fibers. Furthermore, the nonphosphorylated neurofilament appeared in the mossy fibers of the CA3 stratum lucidum, which normally do not express such immunoreactivity. The results indicate that the neurofilaments are altered following the neuronal degeneration and postlesional plasticity caused by the kainic acid administration. Therefore, the examination of various phosphorylated neurofilaments may offer a comprehensive understanding of major hippocampal pathways, axonal plasticity and the possible roles of neurofilaments in the hippocampus following excitotoxic insults.

Excitotoxicity. Experimental correlates to human epilepsy.

Neurochemical observations on cortical biopsies form 48 patients under surgical treatment for pharmacoresistant partial epilepsy showed a 70-80% increase in glutamate concentration when expressed in relation to neuron specific enolase. Intraperitoneal administration of one of its receptor agonists, kainic acid (KA), to the rat led to increased epileptogenic activity of the limbic type in a dose-dependent fashion. The KA injection also led to a neuronal cell death and a gliosis, closely correlated to the extent of seizure activity. In biopsies from human epileptogenic cortex, the concentration of neuron specific enolase correlated inversely to that of glial fibrillary acidic protein, a marker for astrocytic glial cells. Stimulation of the KA receptor decreased the extent of phosphorylation of the largest subunit of neurofilaments (NF-H) that have consequences for structural stability and axonal transport. Phosphorylated NF-H decreased also in human epileptic cortex, indicating either an overactivity of excitatory neurotransmitters or a loss of axonal compartments.

Changes in neurofilament protein NF-L and NF-H immunoreactivity following kainic acid-induced seizures.

Postlesion plasticity of neuronal processes might contribute to secondary spontaneous seizures after kainic acid administration. In this study, neurofilament (NF) proteins were examined following intraperitoneal injection of kainic acid, and special reference was given to temporal changes in quantity and quality of the NF light (NF-L) and heavy (NF-H) subunits. A pronounced decrease in phosphorylation-related immunoreactivity of NF-H occurred as early as 1 day after the injection in the amygdala/pyriform cortex, hippocampus, striatum, and dorsal cerebral cortex. A shift of NF-H from the phosphorylated to nonphosphorylated form was evident in immunoblots, suggesting dephosphorylation contributed to the decrease. Decreases in NF-L and phosphorylated NF-H contents in the limbic structure at 3 days were correlated with the increasing kainic acid doses from 2.5 to 10 mg/kg. The degradation pattern in immunoblots with antibodies against NF-L indicated that the decrease in NF-L was probably due to calcium-activated proteolysis. NF-L and phosphorylated NF-H contents secondarily increased from 9 days onward, with approximately 20% above the control level of phosphorylated NF-H immunoreactivity at 27 days in the amygdala/pyriform cortex and ventral hippocampus. Immunohistochemical examination of the hippocampus revealed that an increase of NF staining in the mossy fiber system may contribute to the NF recovery in this region. Furthermore, the temporal changes of NF-L and phosphorylated NF-H contents were positively correlated with those of the neuronal cell adhesion molecule, a neuritic growth cone marker, substantiating postlesion regenerative reactions of NF proteins. Functional consequences of the NF plasticity remain to be identified.

Induction of glucose regulated protein (grp78) and inducible heat shock protein (hsp70) mRNAs in rat brain after kainic acid seizures and focal ischemia.

Specific probes were obtained using PCR cloning from rat brain for the 78 kDa glucose regulated (grp78), inducible 72 kDa (hsp70) as well as constitutive 73 kDa (hsc73) heat shock mRNAs. Grp78 and hsc73 were expressed in normal rat brain whereas hsp70 was not. Subcutaneous injection kainic acid (10 mg/kg) produced seizures and induced all three mRNAs. The induction of grp78 and hsp70 mRNAs occurred within 2 h, peaked between 6-8 h, persisted for 48 h, and returned to control levels by 72 h. Expression of the grp78 and hsp70 mRNAs after focal ischemia progressively increased with occlusion durations from 15-120 min in the cerebral cortex. Though grp78 and hsp70 mRNAs were induced modestly in the striatum by 15 min of ischemia, longer durations of ischemia were characterized by little change in the grp78 mRNA levels and relatively lower levels of hsp70 expression. This result indicates that progressive increases in the duration of ischemia in brain, prior to infarction, may produce proportional increases in transcription of the heat shock genes. However, once the duration of ischemia is long enough to produce infarction, this severely limits the availability of ATP which blocks transcription of the heat shock genes. In conclusion, concurrent induction of the heat shock genes suggests that kainic acid seizures and focal ischemia induce several different stress responses in brain cells caused by denaturation of proteins, changes of protein synthesis, and changes of protein glycosylation.

Neuronal and glial marker proteins in encephalopathy associated with acute liver failure and acute hyperammonemia in the rabbit.

Neuronal and glial cell marker proteins were quantified in order to evaluate the possibility of increased proteolysis in the brain of rabbits with acute liver failure and acute hyperammonemia. Acute liver failure was induced by a two-stage devascularization procedure. Acute hyperammonemia was induced by a prolonged infusion of ammonium acetate, which simulates the plasma ammonia level in acute liver failure. Control animals received an infusion of sodium/potassium acetate. After development of severe encephalopathy, the animals were sacrificed (13.7 +/- 1.3 hours for rabbits with acute liver failure and 20.2 +/- 0.8 hours for rabbits with hyperammonemia) (x +/- S.E.M./n = 6) and their brains were dissected into cerebral cortex, hippocampus, cerebellum and brain stem. The total protein content and the concentrations of the neuronal cell marker proteins NSE (neuron specific enolase), NF68 and NF200 (68 kD and 200 kD neurofilament polypeptides) and the glial cell marker proteins GFAP (glial fibrillary acidic protein) and S-100 were determined. Total protein content was decreased in the brain stem in acute hyperammonemia only. The content of neuronal and glial cell markers was not affected in either of the two conditions. However, low molecular weight proteolytic fragments of the NF 68 kD polypeptide were observed in the hippocampus of three out of six animals in both experimental groups. No proteolytic degradation of GFAP was observed. The results show that, in experimental encephalopathy due to acute liver failure and acute hyperammonemia, no major changes occur in the marker proteins. The finding of proteolytic fragments of the NF68 polypeptide indicates that the neuronal population is affected prior to glial alterations. These findings are in agreement with the concept that acute hepatic encephalopathy is reversible and induces only slight structural changes.

Biochemical correlates to cortical dysplasia, gliosis, and astrocytoma infiltration in human epileptogenic cortex.

The study provides detailed biochemical correlates to the common histopathological diagnoses in epilepsy. A dot immunobinding procedure was used for quantification of NSE, GFA, S-100, NCAM, NF 68 and NF 200. The material consisted of samples from 48 patients either selected for surgical treatment of partial epilepsy or for disorders not related to epilepsy. The histopathological diagnosis of the epileptic cases was: MCD (mild cortical dysplasia, microdysgenesis), gliosis, astrocytoma, ganglioglioma, oligodendroglioma and single cases. The concentration in non-epileptic white matter, in per cent of that in grey matter was: NSE, 85; GFA, 175; S-100, 117; NCAM, 43; NF 68,227 and NF 200, 173. The concentration of NSE as well as of GFA was close to normal in the specimens of the MCD and gliosis groups and of one subgroup of the astrocytomas. There was a striking inverse relationship of the GFA vs the NSE concentrations in the whole material. The concentrations of S-100 showed no such inverse relationship to NSE levels. In all the epileptic groups, total NCAM was lower than 50% of that of the non-epileptic group. The mean NF 68 and NF 200 concentration in the gliosis and astrocytoma groups was 75% of that of the non-epileptic group while the corresponding value for the MCD group was 50%. There was a positive correlation of immunochemically determined GFA and the histopathological gliosis score in the samples of epileptogenic cortex. There was no correlation between the concentration of GFA in the samples and the duration of epilepsy.(ABSTRACT TRUNCATED AT 250 WORDS)