Unidentified neuroglia potentials during propagated seizures in neocortex

Cortical surface and intracellular recording of silent cells (neuroglia) was carried out in the pericruciate cortex of cats during propagated seizures produced by repetitive stimulation of the surface of the opposite homotopic neocortex. The membrane characteristics of these cells were similar to neuroglial cells studied in leech, amphibian, and rat optic nerves, tissue culture, and mammalian cerebral cortex. By varying the parameters of transcallosal stimulation, it was possible to obtain either minor or major propagated seizures. All cells with resting membrane potentials (RMP) greater than 30 mv recorded during minor propagated seizures exhibited a depolarizing response (5–14 mv) during the seizure episode followed by a postictal hyperpolarizing response (1–9 mv) and a slow return to the original resting level. The peak amplitude of the depolarizing response was proportional to the cell's RMP and the amplitude of the seizure waves in the EEG. During major propagated seizures, an augmentation of the depolarizing response to 16–30 mv and the hyperpolarizing response to 10–15 mv was noted. A membrane conductance change during these events was not observed. During major propagated seizures, an increase in [K⁺]_o over the resting [K⁺]_o was calculated to be 10 meq/liter. However, the level of [K⁺]_o reached in the extracellular clefts was probably much higher than this calculated value for reasons which are discussed. A model for seizure propagation is presented. The postictal hyperpolarization most likely represents the effect of a K⁺-sensitive electrogenic pump in the glial membrane.     

Synaptic impairment induced by paroxysmal ionic conditions in neocortex

Purpose: Seizures are associated with a reduction in extracellular Ca²⁺ concentration ([Ca²⁺]_o) and an increase in extracellular K⁺ concentration ([K⁺]_o). The long‐range synchrony observed between distant electrodes during seizures is weak. We hypothesized that changes in extracellular ionic conditions during seizures are sufficient to alter synaptic neuronal responses and synchrony in the neocortex. Methods: We obtained in vivo and in vitro electrophysiologic recordings combined with microstimulation from cat/rat neocortical neurons during seizures and seizure‐like ionic conditions. In vitro the [K⁺]_o was 2.8, 6.25, 8.0, and 12 mm and the [Ca²⁺]_o was 1.2 and 0.6 mm . Key Findings: During seizures recorded in vivo, we observed abolition of evoked synaptic responses. In vitro, the membrane potential of both regular‐spiking and fast‐spiking neurons was depolarized in high [K⁺]_o conditions and hyperpolarized in high [Ca²⁺]_o conditions. During high [K⁺]_o conditions, changes in [Ca²⁺]_o did not affect membrane potential. The synaptic responsiveness of both regular‐spiking and fast‐spiking neurons was reduced during seizure‐like ionic conditions. A reduction in [Ca²⁺]_o to 0.6 mm increased failure rates but did not abolish responses. However, an increase in [K⁺]_o to 12 mm abolished postsynaptic responses, which depended on a blockade in axonal spike propagation. Significance: We conclude that concomitant changes in [K⁺]_o and [Ca²⁺]_o observed during seizures contribute largely to the alterations of synaptic neuronal responses and to the decrease in long‐range synchrony during neocortical seizures.     

Potassium activity and changes in glial and neuronal membrane potentials during initiation and spread of afterdischarge in cerebral cortex of cat

Trains of electrical stimuli of increasing intensity were applied to the surface of the anterior suprasylvian gyrus to produce afterdischarges (AD) that remained localized or spread to a recording site 2 cm posteriorly in the same gyrus. The local afterdischarge was associated with a negative steady potential (SP) shift and increases in [K⁺]₀ that were maximal at or near the surface and gradually decreased in magnitude at deeper layers of the cortex.During spreading AD, recordings at the stimulus site showed a secondary increase in both the SP shift and [K⁺]₀ at about 400 to 1400 μm below the cortical surface. As the AD invaded the distant recording site it was associated with a comparable negative SP shift and increase in [K+]o. Neither the appearance of local AD nor its spread to the distant recording site were contingent upon critical elevations of [K⁺]₀.During secondary increases in [K+]o glial depolarizations were less than would be predicted if the membrane potential were determined solely by changes in the ratio of intra- to extracellular [K⁺]. Smaller deviations from the Nernst equation also occurred during the repolarizing phase of glial depolarization produced by weaker stimuli that did not produce a secondary increase in [K⁺]₀. Only immediately after the stimulus train did the relationship between glial depolarization and [K⁺]₀ approach the expected slope of a K⁺ electrode.Simultaneous intracellular recording of neurons and [K⁺]₀ did not show an increase in neuronal firing rate or membrane depolarization to account for the additional increase in [K⁺]₀ during spreading AD.The possible sources of the secondary increase in [K⁺]₀ and the significance of the failure of glia to depolarize to levels predicted by the Nernst equation are discussed.     

Changes in extracellular potassium activity during neocortical propagated seizures

Direct measurements of extracellular potassium concentration [(K⁺)o] using a potassium-sensitive microelectrode were carried out in the pericruciate cortex of cats during propagated seizures initiated by repetitive stimulation of the surface of the opposite homotopic neocortex. (K⁺)o increases during either subthreshold stimulus trains or ictal episodes were dependent on cortical depth and distance from the homotopic point. Seizures could be classified by typical increases in (K⁺)o: 0.5± 0.06mm/1 with minor and 5.9 ± 0.8 mm/1 above the steady state value with major seizures. The upper limit for (K⁺)o during ictal events was 10.2 mm/1. A postictal undershoot of (K⁺)o below steady state was observed. The elevation of (K⁺)o to a critical value during stimulation was associated with a “threshold” for the initiation of propagated seizures. That value was 0.5 mm/1 for minor seizures and 1.8 mm/1 for major seizures. Both minor and major seizures were found to act as “all or none” phenomena of variable morphology. The role of (K⁺)o in the modulation of CNS activity and the pathogenesis of ictal phenomena is discussed.     

The Role of Extracellular Potassium in Early Epilepsy

Clinical studies indicate that early epilepsy after head injury may be associated with some transient and reversible pathophysical processes of the brain. It has been proposed that epileptogenesis in the neocortex and hippocampus may be related to potassium ion accumulation in extracellular spaces. To investigate this hypothesis, we measured [K⁺]₀ using potassium-sensitive microelectrodes in the sensorimotor cortex of cats during early seizures induced by trauma.

The [K⁺]₀ increases associated with seizure activity ranged from 14.6 to 25.1 mM, and these were significantly higher than those unassociated with spikes or seizure discharges. Moreover, high K⁺ solutions (15 mM or more) directly applied to the cortex produced spiking and seizures. These results seem to support the hypothesis that accumulation of [K⁺]₀ is related to development of early epilepsy.     

Effects of serotonin on extracellular potassium concentration in the rat hippocampal slice

The effects of serotonin (5-HT) on extracellular potassium concentration ([K⁺]₀) were measured with ion-selective microelectrodes in rat hippocampal slices. Electrical stimulation of an excitatory afferent system, the Schaffer collateral commissural pathway, caused a 2–4 mM rise in [K⁺]₀ in the stratum pyramidale of area CA1. 5-HT caused a 0.6–1.1 mM rise in [K⁺]₀. This rise was associated with hyperpolarization of neurons and cessation of their spontaneous spike discharge. Methysergide, a 5-HT antagonist, reduced the 5-HT effect. The change in [K⁺]₀ was highest in stratum moleculare and lowest in stratum pyramidale, the opposite gradient to that found with excitatory electrical stimulation. The 5-HT-induced [K⁺]₀ changes were maximal in CA1 stratum moleculare, intermediate in the dentate stratum granulare and almost non-existent in the CA3 stratum pyramidale.GABA, but not norepinephrine, produced a small (up to 0.5 mM) rise in [K⁺]₀ in stratum pyramidale. Extracellular calcium concentration measured with a Ca²⁺-sensitive microelectrode was reduced by electrical stimulation but unchanged by 5-HT or norepinephrine. It is suggested that 5-HT hyperpolarizes hippocampal cells by activation of sodium- and calcium-independent potassium channels, which cause a rise in [K⁺]₀.     

Extracellular potassium activity during epileptogenesis

Direct measurements of extracellular potassium concentration ([K⁺]₀) changes in penicillin epileptogenic foci of cat cortex were made using potassium-sensitive microelectrodes. Interictal EEG events were accompanied by increases in [K⁺]₀ lasting several seconds. The amplitudes and rise rates of these increases varied with cortical depth, distance from the center of the focus, and the [K⁺]₀ at which they occurred. During ictal events, [K⁺]₀ consistently reached 9–10 mm. The patterns of ictal [K⁺]₀ changes also showed variations with depth in the cortex and distance from the focus. There was an upper limit for [K⁺]₀ during both interictal and ictal epileptiform activity at about 10 mm. The role of these [K⁺]₀ changes in modulating neuronal excitability in the epileptogenic focus is discussed. The [K⁺]₀ level did not appear to be a critical factor in initiation or termination of ictal episodes.     

Light-evoked changes in extracellular potassium concentration in mudpuppy retina

Light-evoked changes in extracellular potassium concentration ([K⁺]₀) and field potentials were recorded simultaneously in response to a wide variety of stimuli and at various depths within the retina of Necturus. At both light onset and offset, small diameter flashed stimuli elicit a large increase in [K⁺]₀ in the proximal retina. The depth profile of this K⁺ increase is nearly identical to that of the proximal negative response (PNR), and both responses exhibit similar behavior to a number of other stimulus parameters. This suggests that the same neurons which generate the PNR may be the source of the observed K⁺ flux. Large diameter flashed stimuli elicit a slow decrease in [K⁺]₀ in the distal retina and a small increase proximally. The K⁺ increase occurs at a depth where the b-wave of the electroretinogram is positive going, and where its current source lies. Increasing background light intensity decreases [K⁺]₀ in the distal retina, and generally increases [K⁺]₀ proximally.     

Potassium activity in rabbit cortex

Measurements of extracellular potassium concentration ([K⁺]₀) were made in rabbit cortex using potassium-sensitive microelectrodes with a rapid response time. Resting levels of [K⁺]₀ averaged about 3 mM and transient increases occurred following direct stimulation of normal cortex, during spreading depression (SD), and during interictal epileptiform discharges.Interictal discharges were associated with increases in [K⁺]₀ to maximum levels of 9–10 mM. These changes in [K⁺]₀ had a laminar distribution through the cortex being maximal in deeper layers. During ictal episodes [K⁺]₀ peaked at 9–10 mM and appeared ‘clamped’ at this level. Neither seizure onset nor termination appeared to occur at particular threshold levels of [K⁺]₀. SDs were associated with increases in [K⁺]₀ to about 40 mM. During SDs, ictal activity might be generated even though local [K⁺]₀ was 25–30 mM. DC shifts in EEG activity recorded adjacent to the K⁺ electrode usually preceded the [K⁺]₀ changes and returned more quickly to baseline levels, but the envelope of DC changes closely mirrored the alterations in [K⁺]₀.     

Secretion from rat neurohypophysial nerve terminals (neurosecretosomes) rapidly inactivates despite continued elevation of intracellular Ca

Cytoplasmic calcium concentration was measured in neurosecretory nerve terminals (neurosecretosomes) isolated from rat neurohypophyses by fura-2 fluorescence measurements and digital video microscopy. Hormone release and cytoplasmic calcium concentration were measured during depolarizations induced by elevated extracellular potassium concentration. During prolonged depolarizations with 55 mM [K⁺]₀, the cytoplasmic calcium concentration remained elevated as long as depolarization persisted, while secretion inactivated after the initial sharp rise. The amplitude and duration of the increase in [Ca²⁺]_i was dependent on the degree of depolarization such that upon low levels of depolarizations (12.5 mM or 25 mM [K⁺]₀), the calcium responses were smaller and relatively transient, and with higher levels of depolarization (55 mM [K⁺]₀) the responses were sustained and were higher in amplitude. Responses to low levels of depolarization were less sensitive to the dihydropyridine calcium channel blocker, nimodipine, while the increase in [Ca²⁺]_i induced by 55 mM [K⁺]₀ became transient, and was significantly smaller. These observations suggest that these peptidergic nerve terminals possess at least two different types of voltage-gated calcium channels. Removal of extracellular sodium resulted in a significant increase in [Ca²⁺]_i and secretion in the absence of depolarizing stimulus, suggesting that sodium-calcium exchange mechanism is operative in these nerve terminals. Although the [Ca²⁺]_i increase was of similar magnitude to the depolarization-induced changes, the resultant secretion was 10-fold lower, but the rate of inactivation of secretion, however, was comparable.     

Neurosteroids modulate the reaction of astroglia to high extracellular potassium levels

The extracellular concentration of potassium ([K⁺]₀) in brain tissue is modified by neuronal activity and is increased under several pathological conditions. The influence of neurosteroids on the astroglia response to high [K⁺]₀ was assessed on cultured slices from rat hippocampus. Exposure to [K⁺]₀ above physiological (3 mM) levels resulted in the progressive appearance of cell processes immunoreactive for glial fibrillary acidic protein (GFAP). The maximal effect was observed at 50 mM [K⁺]₀, and further increases of [K⁺]₀ did not increase the extension of GFAP-immunoreactive processes. The effect was observed as early as 10 min after increasing [K⁺]₀, was independent of new protein synthesis, and was reversible, reaching control conditions by 15 h after resetting [K⁺]₀ to physiological levels. Gonadal hormones and neurosteroids had prominent and variable effects on the stimulatory influence of high [K⁺]₀ on astroglia morphology. At physiological [K⁺]₀, 17β-estradiol and pregnenolone, as well as its sulfate derivative, increased the extension of GFAP-immunoreactive processes. However, at high [K⁺]₀, testosterone, pregnenolone, and dehydroepiandrosterone and its sulfate derivative decreased the extension of GFAP-immunoreactive processes. Effects of gonadal hormones and neurosteroids were blocked by the protein synthesis inhibitor cycloheximide. These results suggest that non-genomic effects of high [K⁺]₀ on glial cells interact with genomic effects of steroids to modulate astroglia morphology. © 1996 Wiley-Liss, Inc.     

Potassium activity in immature cortex

Extracellular potassium concentration ([K⁺]₀) was determined in neocortex of rabbits aged 1–19 days using potassium-sensitive microelectrodes. Baseline [K⁺]₀ levels of 3 mM were obtained in some animals of all ages but levels as high as 35 mM were found in some newborn rabbits, presumably due to iatrogenic factors. This suggests that a fragility of K⁺ control mechanisms exists in immature cortex. Increases in [K⁺]₀ of up to 2.6 mM occured during each interictal penicillin epileptiform discharge, were maximal in the oldest animals, and varied at different depths, being largest in deeper cortical layers. There was an inverse relationship between the baseline [K⁺]₀ and the amount of increase during each interictal event. At resting [K⁺]₀ levels of 10 mM or above, no detectable increase occurred during interictal or ictal epileptogenesis. Increases in [K⁺]₀ during ictal episodes ranged from 3 to 7 mM and were larger in older animals. The rate of rise and fall of [K⁺]₀ during interictal and ictal epileptogenesis was slower in younger animals. Ictal activity occurred in the cortex even when baseline [K⁺]₀ levels were at leasr 20 mM.     

Characteristics of activity-dependent potassium accumulation in mammalian peripheral nerve in vitro

Ion-sensitive microelectrodes were used to study the behavior of extracellular ions in rat sciatic nerve during and following activity. Nerve stimulation produced increases in [K⁺]_o that were dependent upon the frequency and duration of stimulation; no change in extracellular pH occurred with stimulation. Increases in [K⁺]_o dependent on axonal discharge since they were blocked by inhibiting sodium channels with tetrodotoxin. At 22°C, stimulation could induce increases in [K⁺]_o of several mM; at 36°C, stimulation rarely produced increases in [K⁺]_o greater than 1mM. Stimulated increases in [K⁺]_o dissipated very slowly (i.e. t_1/2 = 50–100s) and the rate of dissipation was not significantly affected by anoxia, changes in temperature, changes in extracellular pH, or the application of a blocker of Na⁺, K⁺-ATPase (ouabain) or a K⁺ channel blocker (Ba²⁺). In comparison to the central nervous system, neural activity in rat sciatic nerve produced smaller increases in [K⁺]_o and these increases dissipated much more slowly. The primary mechanism of K⁺ dissipation appeared to be diffusion, probably facilitated by the larger extracellular space in peripheral nerve compared to the central nervous system, but impeded by diffusion barriers imposed by the blood-nerve barrier.     

Aquaporin‐4 regulates the velocity and frequency of cortical spreading depression in mice

The astrocyte water channel aquaporin‐4 (AQP4) regulates extracellular space (ECS) K⁺ concentration ([K⁺]_e) and volume dynamics following neuronal activation. Here, we investigated how AQP4‐mediated changes in [K⁺]_e and ECS volume affect the velocity, frequency, and amplitude of cortical spreading depression (CSD) depolarizations produced by surface KCl application in wild‐type (AQP4^+/+) and AQP4‐deficient (AQP4^?/?) mice. In contrast to initial expectations, both the velocity and the frequency of CSD were significantly reduced in AQP4^?/? mice when compared with AQP4^+/+ mice, by 22% and 32%, respectively. Measurement of [K⁺]_e with K⁺‐selective microelectrodes demonstrated an increase to ～35 mM during spreading depolarizations in both AQP4^+/+ and AQP4^?/? mice, but the rates of [K⁺]_e increase (3.5 vs. 1.5 mM/s) and reuptake (t_1/2 33 vs. 61 s) were significantly reduced in AQP4^?/? mice. ECS volume fraction measured by tetramethylammonium iontophoresis was greatly reduced during depolarizations from 0.18 to 0.053 in AQP4^+/+ mice, and 0.23 to 0.063 in AQP4^?/? mice. Analysis of the experimental data using a mathematical model of CSD propagation suggested that the reduced velocity of CSD depolarizations in AQP4^?/? mice was primarily a consequence of the slowed increase in [K⁺]_e during neuronal depolarization. These results demonstrate that AQP4 effects on [K⁺]_e and ECS volume dynamics accelerate CSD propagation. GLIA 2015;63:1860–1869     

The increase in extracellular potassium concentration in the ischemic brain in relation to the preischemic functional activity and cerebral metabolic rate

The course of ischemic increase of extracellular potassium concentration ([K⁺]_e) was studied in rat cerebral cortex with potassium selective microelectrodes and correlated to the preischemic functional and metabolic state. Complete cerebral ischemia was induced in artificially ventilated rats by cardiac arrest. Seven different functional states including conditions with cerebral hypermetabolism (seizures, amphetamine intoxication, hyperthermia) and hypometabolism (barbiturate anesthesia, hypothermia) were chosen in order to cover a wide range of cerebral metabolic rates (CMR_o2 : 28.7-2.4 ml O₂/(100 g)/min). The ischemic increase of [K⁺]_e was delayed in conditions with low CMR_o2 and accelerated in conditions with high CMR_o2; the time interval to the terminal steep rise in extracellular potassium concentration varied within the extremes of35 ± 5 and 365 ± 12sec (means ± S.E.M.), the control state (N₂O-analgesia) being116 ± 5sec. In groups with high CMR_o2 electrocortical activity ceased within 15 sec and in groups with low CMR_o2 within 22 sec. The rates of the ischemic [K⁺]_e increase, measured as rate of change in the potassium electrode potential (mV/sec), remained high in conditions with high preischemic CMR_o2 and low in conditions with low CMR_o2, indicating a remaining influence of the preischemic metabolism on membrane ion permeability. These results support previous metabolic data indicating that the rate of consumption of high energy phosphates during ischemia mirrors the preischemic cerebral metabolic rate. Phenobarbital anesthesia did not change the initial rate of [K⁺]_e increase but reduced the rate of [K⁺]_e increase later during ischemia, suggesting a special effect of barbiturates on partly depolarized membranes.     

Importance of astrocytes for potassium ion (K+) homeostasis in brain and glial effects of K+ and its transporters on learning

Initial clearance of extracellular K⁺ ([K⁺]_o) following neuronal excitation occurs by astrocytic uptake, because elevated [K⁺]_o activates astrocytic but not neuronal Na⁺,K⁺-ATPases. Subsequently, astrocytic K⁺ is re-released via Kir4.1 channels after distribution in the astrocytic functional syncytium via gap junctions. The dispersal ensures widespread release, preventing renewed [K⁺]_o increase and allowing neuronal Na⁺,K⁺-ATPase-mediated re-uptake. Na⁺,K⁺-ATPase operation creates extracellular hypertonicity and cell shrinkage which is reversed by the astrocytic cotransporter NKCC1. Inhibition of Kir channels by activation of specific PKC isotypes may decrease syncytial distribution and enable physiologically occurring [K⁺]_o increases to open L-channels for Ca²⁺, activating [K⁺]_o-stimulated gliotransmitter release and regulating gap junctions. Learning is impaired when [K⁺]_o is decreased to levels mainly affecting astrocytic membrane potential or Na⁺,K⁺-ATPase or by abnormalities in its α2 subunit. It is enhanced by NKCC1-mediated ion and water uptake during the undershoot, reversing neuronal inactivity, but impaired in migraine with aura in which [K⁺]_o is highly increased. Vasopressin augments NKCC1 effects and facilitates learning. Enhanced myelination, facilitated by astrocytic-oligodendrocytic gap junctions also promotes learning.     

Contributions of the Na+/K+‐ATPase,NKCC1, and Kir4.1 to hippocampal K+ clearance and volume responses

Network activity in the brain is associated with a transient increase in extracellular K⁺ concentration. The excess K⁺ is removed from the extracellular space by mechanisms proposed to involve Kir4.1‐mediated spatial buffering, the Na⁺/K⁺/2Cl^? cotransporter 1 (NKCC1), and/or Na⁺/K⁺‐ATPase activity. Their individual contribution to [K⁺]_o management has been of extended controversy. This study aimed, by several complementary approaches, to delineate the transport characteristics of Kir4.1, NKCC1, and Na⁺/K⁺‐ATPase and to resolve their involvement in clearance of extracellular K⁺ transients. Primary cultures of rat astrocytes displayed robust NKCC1 activity with [K⁺]_o increases above basal levels. Increased [K⁺]_o produced NKCC1‐mediated swelling of cultured astrocytes and NKCC1 could thereby potentially act as a mechanism of K⁺ clearance while concomitantly mediate the associated shrinkage of the extracellular space. In rat hippocampal slices, inhibition of NKCC1 failed to affect the rate of K⁺ removal from the extracellular space while Kir4.1 enacted its spatial buffering only during a local [K⁺]_o increase. In contrast, inhibition of the different isoforms of Na⁺/K⁺‐ATPase reduced post‐stimulus clearance of K⁺ transients. The astrocyte‐characteristic α2β2 subunit composition of Na⁺/K⁺‐ATPase, when expressed in Xenopus oocytes, displayed a K⁺ affinity and voltage‐sensitivity that would render this subunit composition specifically geared for controlling [K⁺]_o during neuronal activity. In rat hippocampal slices, simultaneous measurements of the extracellular space volume revealed that neither Kir4.1, NKCC1, nor Na⁺/K⁺‐ATPase accounted for the stimulus‐induced shrinkage of the extracellular space. Thus, NKCC1 plays no role in activity‐induced extracellular K⁺ recovery in native hippocampal tissue while Kir4.1 and Na⁺/K⁺‐ATPase serve temporally distinct roles. GLIA 2014;62:608–622     

Glial membrane potential and extracellular potassium concentration in cultured glial cells.

The relationship between membrane potential and extracellular potassium concentration [K⁺]₀ of primary cultured glial cells showed a complete Nernstian slope (10-fold change of [K⁺]₀ corresponding to 58 mV). The C₆ cell line glia, however, showed a more gradual membrane potential/[K⁺]₀ slope (10-fold [K⁺]₀ change corresponding to 40 mV).     

Interstitial potassium concentration, slow depolarization and focal potential responses in the dorsal horn of the rat spinal slice

Extracellular potassium concentration ([K⁺]_o) was measured, and intra- and extracellular recordings made, in the dorsal horn of rat spinal cord slices maintained in vitro during repetitive dorsal root stimulation. In about half of the dorsal horn neurons, the stimulation evoked a possibly substance P-mediated slow depolarization. [K⁺]_o increased during stimulation, reaching its highest values approximately 150 μm from the dorsal surface. The time course of Δ[K⁺]_o was different from that of the slow depolarization. Substance P itself evoked a much smaller Δ[K⁺]_o (0.4 mM) in the dorsal horn. It is concluded that the slow depolarization is not mediated by elevated [K⁺]_o.     

Focal gliosis and potassium movement in mammalian cortex.

Potassium concentration in the extracellular spaces of the brain ([K⁺]₀) has been proposed to play a role in cortical excitability. It has further been suggested that extracellular potassium concentrations are not regulated normally in damaged (gliotic) brain, and that this defect might lead to initiation or development of epileptic seizures. In order to evaluate this hypothesis we elevated [K⁺]₀ iatrogenically by superfusion of high-potassium artificial cerebrospinal fluid (CSF) through a special chamber placed over focally injured cortex of cats who had recovered from freeze lesions. Standard mathematical techniques were used to calculate from the data: the diffusion coefficient, D, of potassium in cortex; the magnitude, h, of the pial-glial surface barrier to penetration of K⁺ into the brain; and the extent of active uptake of potassium from the intercellular clefts. In comparison with measurements previously made in normal cortex, no significant differences were found in resting K⁺ levels (normal 3.15 ± 0.15 mm, gliotic = 3.24 ± 0.21 mm), in diffusion coefficients (normal = 1.03 ± 0.16 mm²/hr, gliotic = 0.93 ± 0.15 mm²/hr), or in the gross degree of active uptake into cells and blood vessels. However, a more substantial barrier to penetration of potassium was found at the surface of gliotic cortex. At no time were spontaneous seizures observed. We concluded that seizure susceptibility of chronically injured cortex is not related to, or is at best a very complicated function of, altered potassium kinetics.