Reversible osmotic opening of the blood-brain barrier.

Reversible breakdown of the blood-brain barrier is produced by a class of electrolytes and nonelectrolytes which have little or no lipid solubility but which difler in chemical and ionic properties. These agents may osmotically shrink barrier cells, possibly the vascular endothelium, and reversibly open spaces between them. Lipid-soluble nonelectrolytes damage the barrier irreversibly.

Oskar and Cécile Vogt, Lenin's brain and the bumble-bees of the Black Forest.

Oskar Vogt (1870-1955) was a prominent German neurologist and neuroanatomist with a strong interest in the pathogenesis of brain diseases. Together with his wife Cécile (1875-1962), he published landmark papers on the cyto- and myelo-architecture of the brain and the functional anatomy of the basal ganglia. He developed the concept of pathoclisis, i.e., the selective vulnerability of specific neuronal populations in the CNS. In the 1920's, Vogt created a multi-disciplinary brain research institute, the Kaiser-Wilhelm-Institut für Hirnforschung in Berlin-Buch. After Lenin's death in 1924, Oskar Vogt was called to Moscow where he formed a new brain research institute, with the main purpose to investigate the revolutionary's brain. After being dismissed from office by the Nazi government in 1937, the Vogts continued their work in a privately funded institute in Neustadt, the Black Forest.

Regional changes in cerebral blood flow and blood-brain barrier permeability during epileptiform seizures and in acute hypertension in rabbits.

Interrelationship between the breakdown of the blood-brain barrier (BBB) to Evans blue and elevations in the regional cerebral blood flow (rCBF) was studied in rabbits subjected to adrenaline- or metaraminol-induced systemic hypertension and also in bicuculline-induced seizures. The rCBF was assessed in small samples from various regions of the brain with the use of [3H]nicotine, and the permeability of the BBB was evaluated with an Evans blue tracer. In acute hypertension, Evans blue extravasations were observed in the occipital cortex and sometimes in the superior colliculus, i.e., the regions which also showed the highest elevations in rCBF. The breakdown of the BBB in acute hypertension was clearly related to the rate of mean arterial blood pressure rise, being much less pronounced in the metaraminol group, which showed a much slower blood pressure elevation rate. In bicuculline-induced seizures, there was no evident correlation between the amplitude of rCBF elevations and Evans blue extravasations. Preservation of BBB integrity was observed in areas showing high elevations in the rCBF.

Cerebral vascular volume after repeated ischemic insults in the gerbil: comparison with changes in CBF and brain edema.

The time course of changes in cerebral intravascular volume was evaluated during 24 h following a series of three 5-min carotid artery occlusions spaced at 1-h intervals and compared with the changes occurring after single 5- or 15-min occlusions. Quantitative estimates of cerebral red cell volume, plasma volume, and total blood volume were obtained from the distribution spaces of 51Cr-labeled erythrocytes and 125I-albumin infused prior to killing at varied recirculation intervals. Significant reductions in vascular volume occurred in all ischemic brain regions within 1 h following a single 5-min occlusion, which recovered to control values within 6 h. A similar time course was seen after repeated occlusions. The reductions in volume remained significant at 6 h after a single 15-min occlusion, but there was no difference from control by 24 h. Thus, the time course of total vascular volume correlates well with that of CBF changes previously described, and both blood flow and blood volume are at normal levels during the time of severe edema 24 h after repeated occlusions. Calculated cerebral hematocrit was 60-70% of that obtained from the femoral artery, but was identical in all brain regions and was constant throughout the postischemic recirculation period, with the exception of a transient reduction in both peripheral and cerebral hematocrit observed at 6-h recirculation following single 15-min occlusions. These results suggest that changes in CBF and blood volume reflect primarily the status of larger vessels and that values in the normal range may be observed even under conditions of severe edema and impaired perfusion at the capillary level.

Effect of acute electrode placement on regional CBF in the gerbil: a comparison of blood flow measured by hydrogen clearance, [3H]nicotine, and [14C]iodoantipyrine techniques.

Regional cerebral blood flow (rCBF) was compared in the gerbil by means of [3H]nicotine, [14C]-iodoantipyrine, and hydrogen clearance techniques. In agreement with other studies, nicotine and iodoantipyrine methods gave virtually identical results. With these methods, it was observed that a reduction in blood flow occurred shortly after insertion of an electrode into the striatum for hydrogen clearance measurement, affecting rCBF throughout the impaled hemisphere. The reduction was moderate (30%) in the striatum and hippocampus, but much greater (70%) in cortical regions. Identical deficits were observed following brief penetrations involving only cortex. Following chronic electrode placement in the striatum, regional blood flow values obtained with [3H]nicotine returned to the control range within 6 h. Blood flow estimates obtained in the striatum with the implanted electrode increased with a similar time course, so that by 6-24 h, hydrogen clearance gave values indistinguishable from control values obtained with [3H]nicotine. These results clearly demonstrate that reduction of CBF subsequent to electrode placement can account for the low values frequently obtained with the hydrogen clearance method in small animals. The distribution of the deficit and the time course of its recovery are similar to blood flow changes associated with spreading depression. While mechanisms responsible for this effect remain to be fully identified, chronic implantation is a practical solution that allows the continued use of hydrogen clearance as a convenient method for repeated measurement of blood flow in the same animal.

Experimental model for repetitive ischemic attacks in the gerbil: the cumulative effect of repeated ischemic insults.

An experimental model for repeated ischemic attacks, which allows easy induction of cerebral ischemia of any desired duration and frequency, has been developed in the gerbil. With this procedure, a pronounced cumulative effect on development of edema and tissue injury was observed using 3 separate, 5-min bilateral occlusions of the common carotid arteries spaced at various time intervals. This effect was most evident when the occlusions were carried out at 1-h intervals, i.e., during the period of marked postischemic hypoperfusion. Such animals, killed after 24 h of recirculation, showed significantly more severe edema and brain tissue injury in the areas exposed to ischemia than was observed in animals killed 24 h after single 5- or 15-min occlusions. The changes of regional CBF, assayed with a [3H]nicotine method, indicated a relatively rapid onset of hypoperfusion of similar degree after each release of arterial occlusion. The hypoperfusion recovered significantly within 6 h of recirculation following either single or multiple occlusions, and no residual hypoperfusion was observed in animals which, when killed at 24 h, showed severe edema and brain tissue injury. This model should prove useful in elucidating the pathophysiological mechanisms operative in repetitive cerebral ischemia.

Development of susceptibility to audiogenic seizures following cardiac arrest cerebral ischemia in rats.

Susceptibility to audiogenic seizures (AGS) was investigated in Sprague-Dawley rats subjected to cardiac arrest cerebral ischemia (CACI), produced by compression of the major cardiac vessels. The onset of AGS was regularly observed 1 day after CACI of > 5 min duration. The duration of postischemic susceptibility to AGS was directly related to the density of cerebral ischemia, with 50% of more severely ischemic animals still showing AGS susceptibility 8 weeks after CACI. Lesioning of the inferior colliculi (IC) abolished the onset of AGS; no such effect was observed after lesioning the medial geniculate (MG). Glutamic acid decarboxylase (GAD) immunochemistry revealed approximately 50% loss of GAD-positive neurons in the IC, which was similar in animals with various durations of AGS susceptibility. Otherwise, there was a conspicuous sprouting of gamma-aminobutyric acid (GABA)-ergic terminals in the ventral thalamic nuclei, which peaked approximately 1 month after the CACI. Evaluation of GABA-A inhibitory function in the hippocampus by the paired pulse stimulation revealed changes indicating loss of GABA-A inhibition coinciding with the onset of AGS, and its return in animals tested 2 months after CACI. Our observations suggest a potential role of GABA-ergic dysfunction in the postischemic development of AGS.

Global cerebral ischemia associated with cardiac arrest in the rat: I. Dynamics of early neuronal changes.

Light microscopic neuronal changes were studied in rats subjected to 10 min of global ischemia produced by compression of the major cardiac vessels. Observations of cresyl violet-stained sections revealed early changes involving predominantly GABAergic neurons in various locations. In rats killed 15 min after recirculation, the changes were characterized by the appearance of a clear peripheral zone with condensation of the remaining neuronal cytoplasm. After 1 h, these zones appeared to be compartmentalized into individual pearl-like vacuoles, especially prominent in the nucleus reticularis thalami. After 3 h, the cytoplasmic vacuoles disappeared and the neuronal changes, particularly in the cerebral cortex, striatum, hippocampus, and pars reticulata of the substantia nigra, consisted mainly of hyperchromasia or loss of Nissl substance. After 2 days, the cerebral cortex and thalamus contained occasional neurons with conspicuously large nucleoli. After 7 days, the hippocampus revealed an approximately 50% loss of CA1 pyramidal neurons, associated with intense microglial reactivity in the stratum radiatum, whereas the neuronal destruction was more complete in the nucleus reticularis thalami. Our observations suggest a possibility that early changes in GABAergic neurons may provide a period of neuronal disinhibition and thus contribute to an excitatory ischemic damage in regions connected by GABAergic circuitry.

Brain serotonin after experimental vascular occlusion.

Unilateral ligation of a common carotid artery in gerbils causes a decreased rate of serotonin synthesis and degradation but an increased release of this monoamine. In the brain, reduction of cerebral serotonin content during ischemia is followed by accumulation of its main metabolite, 5-hydroxyindolacetic acid. These data support the contention that serotonin plays an important role in the progression of cerebral infarction.

Cortical spreading depression induces long-term alterations of BDNF levels in cortex and hippocampus distinct from lesion effects: implications for ischemic tolerance.

Cortical spreading depression (CSD) protects hippocampal and cortical neurons from an otherwise lethal ischemic insult delivered days later. The present study was undertaken to evaluate changes in the expression of BDNF following CSD, distinct from lesion effects and its possible involvement in delayed ischemic tolerance. CSD was elicited by KCl application and a cortical lesion was made by hyperosmolar NaCl application. BDNF mRNA was examined by in situ hybridization and Northern blot up to 7 days post-CSD. BDNF protein content was measured by ELISA. In the cortex, BDNF protein was mildly elevated despite minimal increases of mRNA in the NaCl lesion group. CSD specifically up-regulated BDNF mRNA at 4 h, followed by a delayed secondary increase at 2-3 days. BDNF protein exhibited smaller biphasic increases at 24 h and 3-7 days post-CSD which were significantly higher than the NaCl lesion group. In the hippocampus, BDNF protein levels showed a delayed decrease in both groups independent of mRNA changes, but CSD specifically delayed this decrease. Thus, CSD can alter BDNF levels independent of lesion effects. The increased BDNF following CSD in the cortex is consistent with the involvement of BDNF in cortical ischemic tolerance. BDNF could not, however, be directly related to ischemic tolerance in the hippocampus.

Preservation of GABAergic perikarya and boutons after transient ischemia in the gerbil hippocampal CA1 field.

Using an antibody directed against the gamma-aminobutyric acid (GABA)-synthetizing enzyme glutamate decarboxylase (GAD) the fate of the GABAergic innervation was investigated in the hippocampal field CA1 of gerbils up to 14 days after a bilateral transient 5-min occlusion of carotid arteries. As described previously, the CA1 pyramidal cells were subject to the ischemia-induced delayed neuronal death, the first signs of which were detectable after 2 days and which was fully developed after 4 days. Local GAD-immunoreactive neurons and boutons, however, exhibited no changes in their distribution and morphology over the whole 14-day period investigated, as studied both at the light and electron microscopic level. Thus, it can be assumed that the increased excitation observed during the development of delayed neuronal death, is not due to a loss of GABAergic neuronal profiles. The resistance of the GABAergic neurons to the ischemic insult is discussed in relation to the presence of Ca2+-binding proteins in this class of neurons, and the long persistence of innervation in an area nearly devoid of postsynaptic targets is considered.

Evidence for increased calcium influx across the choroid plexus following brief ischemia of gerbil brain.

Following 5 min ischemia of gerbil brain, unidirectional transfer of calcium from plasma to cerebrospinal fluid was measured quantitatively with 45CaCl2 at 4 days postischemic recirculation. 45Ca influx across the choroid plexus increased significantly from 0.0101 +/- 0.001 min-1 measured in sham-operated animals (n = 15) to 0.0294 +/- 0.002 min-1 determined in ischemic animals (n = 21; P less than 0.05). Histological examination of choroid plexus was carried out in Cresyl violet-stained sections from gerbils subjected to 5 min ischemia followed by 4 days (n = 6) and 7 days (n = 7) postischemic recirculation. Increased calcium transfer to cerebrospinal fluid was associated with cell damage of choroid plexus observed at 4 days postischemia. Endothelial choroid plexus injury was still detectable at 7 days after the ischemic insult suggesting a long-lasting pathomorphological process. Postischemic alterations in choroid plexus functions apparently expose brain tissue to much higher calcium influx into cerebrospinal fluid which, in turn, may contribute to calcium-related cell damage of the central nervous system.

Regional patterns of blood-brain barrier breakdown during epileptiform seizures induced by various convulsive agents.

In unrestrained rabbits with generalized epileptic seizures induced by systemic application of convulsant drugs, regional changes in blood-brain barrier (BBB) permeability to macromolecules were investigated using Evans Blue (EB) as indicator. BBB leakage due to seizures was present only in animals in which the mean arterial blood pressure rose about 50 mm Hg with the onset of convulsive motor activity. However, a blood pressure increase was not necessarily associated with the occurrence of BBB opening. Pentylenetetrazole-induced seizures resulted in bilateral EB leakage mainly in the hypothalamus, with exception of the mammillary bodies, and the preoptic area, and they were associated, in most cases, with an intensive staining of the cerebellum and also of the midbrain tegmentum. In contrast, seizures due to the GABA receptor blocker bicuculline brought about a penetration of the dye in the region of the pallidum, whereas the GABA synthesis inhibitor methoxypyridoxine produced BBB breakdown in the hippocampus. Methionine-sulfoximine convulsions resulted in a selective stain of the corpora mammillaria, and kainic acid induced a diffuse leakage in neocortical brain areas. As a rule, BBB breakdown was bilateral and confined to anatomically limited brain areas, suggesting that BBB integrity was not only disturbed by abrupt increases in the intraluminal pressure, but was also influenced from the brain tissue. The fluorescence microscopic observations revealed that the tracer penetrated into the neuropil through larger vessels. It had the tendency to accumulate in neurons. In case of the hippocampus, CA2 pyramidal cells revealed more intense uptake of EB than those of the adjacent fields.

Specific gravity increases and brain water content decreases during short epileptiform seizures in discrete rabbit brain areas.

In rabbits subjected to bicuculline (BC)- or pentylenetetrazole (PTZ)-seizures of 3 or 20 min duration or to adrenaline-induced hypertension, specific gravity (SG) was measured bilaterally in 15 regional brain areas in order to detect possible associations between the regionally limited blood-brain barrier openings due to these insults (see Nitsch and Klatzo 1983) and the presence of brain edema. In controls, a large variation between regional SG levels became evident: between 1.0467 in cerebellum and 1.0417 in preoptic area. A seizure duration of only 3 min was not sufficient to change SG significantly. After 20 min of seizures independent from the inducing agent, SG increased in all brain areas. The degree of increase seemed to be unrelated to presence or absence of a blood-brain barrier opening. In an attempt to avoid the influence of blood impregnation on the SG value, blood was replaced by saline before measurement. In controls, saline perfusion decreased SG only in the two areas with the highest original levels, thus documenting the partial dependency of the regional SG on the hematisation of the tissue. After 20 min of PTZ-induced seizures, SG in saline-perfused rabbits increased, but no longer significantly in all brain areas. This suggests that a part of the seizure-induced SG increase can be attributed to the hyperemia of the brain during the convulsions. On the other hand, an increase in flow volume due to hypertension did not change SG with the exception of the septum, preoptic area and hypothalamus. The direct measurement of water content with the classical wet/dry method in 4 gross brain areas showed that early seizure periods are in fact associated with a dehydration of the brain. This phenomenon could be explained by a glucose- and lactate-induced rise in blood osmolarity which in turn might cause a dehydration of the brain tissue.

Incongruence of regional cerebral blood flow increase and blood-brain barrier opening in rabbits at the onset of seizures induced by bicuculline, methoxypyridoxine, and kainic acid.

Cerebral blood flow (CBF) was estimated using [3H]nicotine as tracer in 18 discrete rabbit brain regions at the onset of generalized seizures induced by bicuculline (BC), methoxypyridoxine (MP), and kainic acid (KA) and compared with the previously described regional pattern of blood-brain barrier (BBB) opening. Regional CBF increased 3- to 5-fold in BC-seizures, 2- to 3-fold (except in inferior colliculus) in MP-seizures, and twofold (no change in inferior colliculus) and periaqueductal gray) in KA-seizures. In general, CBF elevation was greater in areas with a high original CBF, and below average in regions with a low control value. However, under each condition a few areas did not follow this rule, viz. the occipital cortex and the basal part of the cerebellar vermis with low flow values in BC-seizures, the inferior colliculus with no change and the pallidum with a relatively high flow during MP-seizures, and limbic system areas with an exceptionally high flow during KA-seizures. The usual rise in arterial blood pressure (BP) at the onset of the seizure, necessary for the development of a BBB lesion, was absent in a few animals with BC- and MP-seizures, yet, their CBF was elevated in most regions although to a lesser extent. No relationship appeared to exist between the rise in CBF and the regional pattern of BBB opening. It is concluded that a loss of cerebral autoregulation is a prerequisite for BBB leakage, but the location of the opening is determined by neuronally derived factors which differ from convulsant to convulsant and from brain region to brain region.

Role of pressure gradients and bulk flow in dynamics of vasogenic brain edema.

The authors present the results of an investigation of the vasogenic type of brain edema using cold injury in cats as a model. Their findings indicate that bulk flow and not diffusion should be considered the main mechanism for the spread of edema through the white matter. This conclusion is based on: 1) comparison of the distances actually traveled by various substances during edema spread with those calculated theoretically for migration of the substances by diffusion; 2) coincidence in the speed of movement by two substances (sucrose and albumin) with widely different diffusion coefficients; 3) measurement of interstitial fluid pressure (IFP) at various distances from the lesion showing the presence of increased IFP in the lesion area and decreasing pressures along the edema pathway toward the normal tissue; and 4) the fact that spreading of edema can be significantly impeded by inducing before the cold lesion an intracellular type of brain edema that reduces the size of the extracellular space (ECS) and increases the resistance to flow of edema fluid. The pressure-volume curve of the brain ECS, as derived from determinations of IFP and tissue water content, indicates that initial steep slope in IFP probably represents the high resistance to fluid mobility through the small diameter extracellular channels and the counteracting resistance of the intermingled structures of brain parenchyma to be separated. Once the IFP exceeds these opposing forces, the ECS dilates, fluid mobility increases, and the edema front advances.

Role of cerebral endothelium in brain oedema.

The view that cerebral endothelial cells represent the cellular basis of the blood-brain barrier has been generally accepted. The regulation of transport processes operating in the cerebral endothelial cells is currently of great interest. Our knowledge of these regulatory mechanisms is briefly reviewed here with special regard to the molecular processes involved in the formation of brain oedema and emphasis on new therapeutic means for its prevention.

Repetitive ischaemia of cat brain: pathophysiological observations.

The cumulative effect of repetitive ischaemia on brain injury was studied in halothane-anaesthetized cats submitted to three episodes of global cerebro-circulatory arrest. Ischaemia of 5.0, 7.5 and 10.0 min duration was produced at hourly intervals by intrathoracic clamping of the innominate and subclavian arteries, and the resulting pathophysiological changes were evaluated by recordings on the electroencephalogram (EEG), blood flow and specific gravity. During each episode of ischaemia EEG flattened within 15 s. After ischaemia the latency of EEG recovery increased with the duration and with the number of repetitions of each ischaemic episode, indicating cumulation of electrophysiological impairment. The flow studies revealed a minor degree of hyperaemia after each ischaemic episode, followed by severe hypoperfusion in the caudate nucleus but not in the cerebral cortex. Brain oedema - as assessed by specific gravity measurements - developed in the hippocampus after three episodes of 5 min ischaemia, and in all grey matter structures investigated after three episodes of 10 min ischaemia. To evaluate the resistance of the ischaemically injured brain to respiratory hypoxia, total oxygen was repeatedly reduced to 5% for 5 min. During these episodes EEG activity progressively declined as a function of the length and the repetition of ischaemia. Parallel n.m.r. spectroscopic measurements in the same model have demonstrated that disturbances of brain energy state during the hypoxic episodes are minor even after three episodes of 10 min ischaemia. EEG suppression, in consequence, is an electrical shut-down phenomenon for the maintenance of cerebral energy state under critical conditions of oxygen delivery.

Protective effect of spreading depression against neuronal damage following cardiac arrest cerebral ischaemia.

Effect of spreading depression (SD) on survival of CA1 hippocampal neurons was studied in Sprague-Dawley rats subjected to cardiac arrest cerebral ischaemia (CACI). SD was induced either by the topical application of KCI to the cerebral cortex (CSD) or by KCI perfusion of the hippocampus via a microdialysis cannula (HSD). CACI was carried out by intrathoracic compression of major cardiac vessels at 1, 3 and 7 days after CSD induction and 3 days after HSD, following which neuronal loss in the CA1 sector of the hippocampus was determined morphometrically. Our study revealed the significant protective effect of SD on survival of the CA1 pyramidal neurons in rats which were subjected to CACI 3 days later. No significant indication of a protective effect was observed in animals with SD induced 1 or 7 days prior to CACI nor in rats in which KCI was substituted by NaCl of equivalent concentration which did not induce any SD waves. Microdialysis assay during HSD showed a significant elevation of extracellular glutamate. Our studies, demonstrating an induction of a transitory period of neuronal resistance to ischaemia by preceding spreading depression, open an opportunity for elucidation of endogeneous factors responsible for the development of such resistance.

Factors in the reproducibility of the gravimetric method for evaluation of edematous changes in the brain.

In the gravimetric method for estimation of water content in brain tissue, four factors that may affect specific gravity (SG) measurements are described and evaluated. These factors are: (1) the purity of kerosene (K) and monobromobenzene (MBB), (2) duration of interaction of MBB with tissue and/or standard solutions, (3) the size of tissue samples, and (4) the temperature. A water-soluble contaminant in K, which can be effectively extracted by water treatment, may significantly decrease SG values. Conversely, in MBB, the SG of tissue samples tends to increase in proportion with time. The SG of the tissue samples appears to be influenced by their size, the smaller pieces registering higher SG values. Also, the temperature of the column affects the SG, with higher temperatures resulting in higher SG values; a difference of 5 degrees C produces a significant difference in SG (p less than 0.05). A standardized procedure providing highly reproducible SG values is presented in which the size of the tissue samples, the recording time, and the temperature of the column consisting of water-treated K and MBB are kept strictly constant.