New uses of LFPs: Pathway-specific threads obtained through spatial discrimination.

Local field potentials (LFPs) reflect the coordinated firing of functional neural assemblies during information coding and transfer across neural networks. As such, it was proposed that the extraordinary variety of cytoarchitectonic elements in the brain is responsible for the wide range of amplitudes and for the coverage of field potentials, which in most cases receive contributions from multiple pathways and populations. The influence of spatial factors overrides the bold interpretations of customary measurements, such as the amplitude and polarity, to the point that their cellular interpretation is one of the hardest tasks in Neurophysiology. Temporal patterns and frequency bands are not exclusive to pathways but rather, the spatial configuration of the voltage gradients created by each pathway is highly specific and may be used advantageously. Recent technical and analytical advances now make it possible to separate and then reconstruct activity for specific pathways. In this review, we discuss how spatial features specific to cells and populations define the amplitude and extension of LFPs, why they become virtually indecipherable when several pathways are co-activated, and then we present the recent advances regarding their disentanglement using spatial discrimination techniques. The pathway-specific threads of LFPs have a simple cellular interpretation, and the temporal fluctuations obtained can be applied to a variety of new experimental objectives and improve existing approaches. Among others, they facilitate the parallel readout of activity in several populations over multiple time scales correlating them with behavior. Also, they access information contained in irregular fluctuations, facilitating the testing of ongoing plasticity. In addition, they open the way to unravel the synaptic nature of rhythmic oscillations, as well as the dynamic relationships between multiple oscillatory activities. The challenge of understanding which waves belong to which populations, and the pathways that provoke them, may soon be overcome.

Spatial modules of coherent activity in pathway-specific LFPs in the hippocampus reflect topology and different modes of presynaptic synchronization.

Ongoing network activity often manifests as irregular fluctuations in local field potentials (LFPs), a complex mixture of multicellular synaptic currents of varying locations and extensions. Among other conditions, for synchronously firing presynaptic units to generate sizable postsynaptic LFPs, their axonal territories should overlap. We have taken advantage of anatomical regularity of the rat hippocampus and combined multiple linear recordings and spatial discrimination techniques to separate pathway-specific LFPs with enough spatial resolution to discriminate postsynaptic regions of varying activation, and to investigate their presynaptic origin, chemical nature, and spatial extension. We identified 6 main excitatory and inhibitory LFP generators with different synaptic territories in principal cells and hippocampal subfields matching anatomical pathways. Some recognized pathways did not contribute notably to LFPs. Each showed different septo-temporal spatial modules over which the field potential fluctuations were synchronous. These modules were explained by either the strong overlap of synaptic territories of coactivated afferent neurons (e.g., CA3 clusters for CA1 Schaffer LFPs), or widespread coalescence of postsynaptic territories (granule cell somatic inhibition). We also show evidence that distinct modes of afferent synchronization generate stereotyped spatial patterns of synchronous LFPs in one pathway. Thus, studying spatial coherence of pathway-specific LFPs provides remote access to the dynamics of afferent populations.

Minor contribution of principal excitatory pathways to hippocampal LFPs in the anesthetized rat: a combined independent component and current source density study.

Analysis of local field potentials (LFPs) helps understand the function of the converging neuronal populations that produce the mixed synaptic activity in principal cells. Recently, using independent component analysis (ICA), we resolved ongoing hippocampal activity into several major contributions of stratified LFP-generators. Here, using pathway-specific LFP reconstruction, we isolated LFP-generators that describe the activity of Schaffer-CA1 and Perforant-Dentate excitatory inputs in the anesthetized rat. First, we applied ICA and current source density analysis to LFPs evoked by electrical subthreshold stimulation of the pathways. The results showed that pathway specific activity is selectively captured by individual components or LFP-generators. Each generator matches the known distribution of axonal terminal fields in the hippocampus and recovers the specific time course of their activation. Second, we use sparse weak electrical stimulation to prime ongoing LFPs with activity of a known origin. Decomposition of ongoing LFPs yields a few significant LFP-generators with distinct spatiotemporal characteristics for the Schaffer and Perforant inputs. Both pathways convey an irregular temporal pattern in bouts of population activity of varying amplitude. Importantly, the contribution of Schaffer and Perforant inputs to the power of raw LFPs in the hippocampus is minor (7 and 5%, respectively). The results support the hypothesis on a sparse population code used by excitatory populations in the entorhino-hippocampal system, and they validate the separation of LFP-generators as a powerful tool to explore the computational function of neuronal circuits in real time.

A role for glutamate and glia in the fast network oscillations preceding spreading depression.

The mechanism of the propagation of spreading depression is unclear. Classical theories proposed a self-maintained cycle fed by elevated potassium and/or glutamate in the extracellular space. Earlier we found in vivo a characteristic oscillatory field activity that is synchronous in a strip of tissue ahead of the oncoming wave of neuron depolarization and that occurs before the extracellular potassium level begins to rise [Herreras O, Largo C, Ibarz JM, Somjen GG, Marrín del Río R (1994) Role of neuronal synchronizing mechanisms in the propagation of spreading depression in the in vivo hippocampus. J Neurosci 14:7087-7098]. We investigated here the possible participation of glutamate and the role of glia in the prodromal field oscillations using extra and intracellular recordings and pharmacological manipulations in rat hippocampal slices. As earlier shown in vivo, field oscillations propagated ahead of the negative potential shift covering distances of up to 1 mm. The oscillatory prodromals were initially subthreshold but then each wave became crowned by a population spike. The frequency of the oscillatory prodromals was variable among slices (80-115 Hz), but constant in individual slices. The blockade of ionotropic glutamate receptors decreased the frequency of prodromal oscillations, retarded spreading depression propagation, and shortened the duration of depolarization. Blocking the glutamate membrane transport increased the oscillatory frequency. The selective metabolic poisoning of astrocytes led to gradual disorganization of prodromal oscillations whose frequency first increased and then decreased. Also, the amplitude of the population spikes within the burst diminished as individual cells fired fewer action potentials, although still phase-locked with population spikes. The effects of glial metabolic impairment were observed within the period when neuron electrical properties were still normal, and were blocked by glutamate receptor antagonists. These data suggest that glutamate released from glial cells and possibly also from neurons has a role in the generation of oscillations and neuron firing synchronization that precede the spreading depression-related depolarization, but additional mechanisms are required to fully explain the onset and propagation of spreading depression.

Longitudinal depolarization gradients along the somatodendritic axis of CA1 pyramidal cells: a novel feature of spreading depression.

We studied the subcellular correlates of spreading depression (SD) in the CA1 rat hippocampus by combining intrasomatic and intradendritic recordings of pyramidal cells with extracellular DC and evoked field and unitary activity. The results demonstrate that during SD only specific parts of the dendritic membranes are deeply depolarized and electrically shunted. Somatic impalements yielded near-zero membrane potential (V(m)) and maximum decrease of input resistance (R(in)) whether the accompanying extracellular negative potential (V(o)) moved along the basal, the apical or both dendritic arbors. However, apical intradendritic recordings showed a different course of local V(m) that is hardly detected from the soma. A decreasing depolarization gradient was observed from the edge of SD-affected fully depolarized subcellular regions toward distal dendrites. Within apical dendrites, the depolarizing front moved toward and stopped at proximal dendrites during the time course of SD so that distal dendrites had repolarized in part or in full by the end of the wave. The drop of local R(in) was initially maximal at any somatodendritic loci and also recovered partially before the end of SD. This recovery was stronger and faster in far dendrites and is best explained by a wave-like somatopetal closure of membrane conductances. Cell subregions far from SD-affected membranes remain electrically excitable and show evoked unitary and field activity. We propose that neuronal depolarization during SD is caused by current flow through extended but discrete patches of shunted membranes driven by uneven longitudinal depolarization.

Synaptically recruited apical currents are required to initiate axonal and apical spikes in hippocampal pyramidal cells: modulation by inhibition.

Dendritic voltage-dependent currents and inhibition modulate the information flow between synaptic and decision areas. Subthreshold and spike currents are sequentially recruited by synaptic potentials in the apical shaft of pyramidal cells, which may also decide cell output. We studied the global role of proximal apical recruited currents on cell output in vitro and in the anesthetized rat after local blockade of Na+ currents in the axon initial segment (AIS) or the proximal apical shaft and their modulation by inhibition. Microejection of TTX, field potentials, and intrasomatic and intradendritic recordings were employed. Dendritic population spikes (PSs) were much smaller in vitro, but the gross relations between synaptic and active currents are similar to in vivo. Activation of Schaffer collaterals triggered PSs and action potentials (APs) in the apical shaft that fully propagated to the axon. However, the specific blockade of proximal Na+ currents avoided cell firing, although antidromic PSs and APs readily invaded somata. The somatic depolarization of subthreshold excitatory postsynaptic potentials (EPSPs) also decreased to about 50%. These results were not due to decreased excitatory input by TTX. However, when GABA(A) inhibition was locally removed, Schaffer synaptic currents skipped the proximal dendrite and fired somatic PSs, although initiated at the AIS. It is concluded that apical currents recruited en passant by Schaffer synaptic potentials in the apical shaft constitute a necessary amplifier for this input to cause output decision. Local inhibition decides when and where an AP will initiate, constituting an efficient mechanism to discriminate and weight different inputs.

Structural inhomogeneities differentially modulate action currents and population spikes initiated in the axon or dendrites.

Action potentials (APs) in CA1 pyramidal cells propagate in different directions along the somatodendritic axis depending on the activation mode (synaptic or axonal). We studied how the geometrical inhomogeneities along the apical shaft, soma, and initial axon modulate the transmembrane current (I(m)) flow underlying APs, using model and experimental techniques. The computations obtained at the subcellular level during forward- and backpropagation were extrapolated to macroscopic level (field potentials) and compared with the basic in vivo features of the ortho- and antidromic population spike (PS) that reflects the sum total of all elementary currents from synchronously firing cells. The matching of theoretical and experimental results supports the following conclusions. Because the charge carried by axonal APs is almost entirely drained into dendrites, the soma invasion is preceded by little capacitive currents (I(cap)), the ionic currents (I(ion)) dominating I(m) and the depolarizing phase. The subsequent invasion of the tapering apical shaft is preceded, however, by significant I(cap), while I(ion) decayed gradually. A similar pattern occurred during backpropagation of spikes synaptically initiated in the axon. On the contrary, when the AP was apically initiated, the dendritic I(ion) was boosted by the apical flare, it was preceded by weak I(cap) and spread forwardly at a slower velocity. Soma invasion is reliable once the AP reached the main apical shaft but less so distal to the primary bifurcation, where it may be upheld by concurrent synaptic activity. The decreasing internal resistance of the apical shaft guided most axial current into the soma, causing its fast charging. There, I(ion) began later in the depolarizing phase of the AP and the reduced driving force made it smaller. This, in addition to a poor temporal overlapping of somatodendritic inward currents within individual cells, built a smaller extracellular sink, i.e., a smaller PS. In both experiment and model, the antidromic (axon-initiated) PS in the soma layer is approximately 30% larger than an orthodromic (apical shaft-initiated) PS contributed by the same number of firing cells. We conclude that the dominance of capacitive or ionic current components on I(m) is a distinguishing feature of forward and backward APs that is predictable from the geometric inhomogeneities between conducting subregions. Correspondingly, experimental and model APs have a faster rising slope during ortho than antidromic activation. The moderate flare of the apical shaft makes forward AP conduction quite safe. This alternative trigger zone enables two different processing modes for apical inputs.

[Tracking along the path toward ischemic neuronal death]. / Las huellas eléctricas en el camino hacia la muerte neuronal isquémica.

INTRODUCTION AND DEVELOPMENT: During episodes of ischemia/anoxia, the neurochemical and environmental changes considered toxic for nervous tissue lie behind the characteristic abrupt massive cell depolarization (MCD). A strong resemblance with other pathologic events enable us to postulate that MCD is a different state of the tissue that includes among others the anoxic depolarization and Le o s spreading depression. MCD is an active event. Neurons enter and leave MCD suddenly and synchronously, and contrary to current belief, their membrane integrity is preserved and ion gradients are only reduced. Biophysical membrane properties are not compatible with some postulates based on endotoxines. There is a direct relation between MCD susceptibility of the different neuron types/nuclei and their vulnerability to ischemia/anoxia. Two different substates can be distinguished in the associated interstitial potentials that are likely related to neuronal and glial dysfunction, respectively. The different modes and timings of anoxic neuronal death depend on the duration of MCD, the functional integrity of the glial network, and the history of previous insults. CONCLUSIONS: MCD is a cellular state of risk bridging life and death. Neurons die if they cannot exit, but may recover if they do promptly, although still have to face subtle changes as well initiated during MCD that will eventually lead them to a delayed death. Avoiding MCD is escaping death. From a clinical point of view, the relevant point is that manipulating MCD entails the simultaneous control of all toxic neurochemical concomitants. Reinforcing vulnerable neurons to avoid their falling into MCD is possible

Las huellas eléctricas en el caminohacia la muerte neuronal isquémica / Tracking along the path toward ischemic neuronal death

Introducción y desarrollo. Durante episodios de isquemia/anoxia, las alteraciones neuroquímicas y ambientales consideradas neurotóxicas se inician tras una despolarización celular masiva (DCM) abrupta y característica. Su similitud con varios procesos fisiopatológicos permite postular que la DCM es un estado tisular/ celular propio que comprende, entre otros, la despolarización anóxica y las ondas de Leão (spreading depression). La DCM es un proceso activo. Las neuronas entran y salen sincrónica y súbitamente, y contra la creencia general, conservan la integridad de membrana, y los gradientes iónicos sólo se reducen. Las propiedades biofísicas de membrana son incompatibles con algunas interpretaciones `toxicistas'. La propensión de diversos tipos/núcleos neuronales a desarrollar DCM y su vulnerabilidad a isquemia/anoxia se relacionan directamente. En los potenciales extracelulares se observan dos subestados, posiblemente asociados a disfunción neuronal y glial. La duración de este `coma' celular, el estado funcional del sincitio glial y el historial de `agresiones' previas determinan la muerte neuronal en sus distintas modalidades y tempos. Conclusiones. La DCM es un estado de riesgo que se proyecta como un puente entre vida y muerte celular. Si las neuronas no consiguen salir de él, mueren, pero si lo hacen, pueden recuperarse, aunque algunas todavía tendrán que hacer frente a cambios más sutiles iniciados también en ese período y que les llevará a una muerte retrasada. Evitar la DCM es evitar la muerte neuronal. Clínicamente, lo más relevante es que la manipulación de la DCM implica el control simultáneo de todas sus concomitantes neuroquímicas `tóxicas'. Es posible reforzar las neuronas vulnerables para impedir su entrada en DCM (AU)

Activity-dependent changes of tissue resistivity in the CA1 region in vivo are layer-specific: modulation of evoked potentials.

We have investigated the spatial map of tissue resistivity across CA1 layers in vivo, its modifications during repetitive orthodromic activity, and the influence of this factor on the shaping of population spikes. Measurement of tissue resistance was made by a high-spatial-resolution three-electrode method. A computer network of equivalent resistors aided theoretical analysis. Tissue resistivity was homogeneous within the basal and apical dendritic trees (260+/-4.5 and 287+/-2.6 Omega cm, respectively). In the stratum pyramidale we found a sharply delimited high-resistivity (643+/-35 Omega cm) band approximately 20 microm wide. Resistivity in slices was approximately 30% higher than in vivo. Computer analysis indicated that the high-resistance somatic layer has a strong influence on the somatic and proximal dendritic contribution to the shaping of population spikes, and reduces volume propagation of currents between dendritic trees. Repetitive orthodromic activation at the theta frequency (4-5 Hz, 20-30 s) caused a stereotyped cycle of field potentials and layer-specific changes of resistivity. Initially (approximately 10 s), long-lasting field excitatory postsynaptic potentials and multiple somatodendritic population spikes developed, and resistivity gradually increased in all layers at a similar rate (period average: 11%). Subsequently, the long-lasting field excitatory synaptic potential subsided and dendritic spike generators were strongly reduced, but multiple somatic spikes remained. Concurrently, the resistivity reached a plateau in all dendritic layers but continued to increase in the somatic layer for about 10-15 s (20% average and up to 50% maximum). Recovery required approximately 60 s. The orthodromic somatic population spike increased variably during stimulation (up to 60%). Using local resistivity changes for correction, supernormal increments of the population spikes were offset, but not totally, uncovering several sub- and supernormal phases that were partially related to changes in dendritic population spike. These resistivity-independent modulations of the somatic population spike are caused by variable volume spread from dendritic spike currents and changed somatic contribution of firing units. This report demonstrates that the strong heterogeneity in the stratum pyramidale is an important factor shaping and modulating the population spike. The different regional rates of resistivity variation force the independent correction of local evoked potentials. We show that not doing so may cause bulk errors in the interpretation of, for instance, field potential ratios widely used to measure the population excitability. The present results underscore the importance of checking variations in recording conditions, which are inherent in most experimental protocols.

Macroscopic and subcellular factors shaping population spikes.

Population spikes (PS) are built by the extracellular summation of action currents during synchronous action potential (AP) firing. In the hippocampal CA1, active dendritic invasion of APs ensures mixed contribution of somatic and dendritic currents to any extracellular location. We investigated the macroscopic and subcellular factors shaping the antidromic PS by fitting its spatiotemporal map with a multineuronal CA1 model in a volume conductor. Decreased summation by temporal scatter of APs reduced less than expected the PS peak in the stratum pyramidale (st. pyr.) but strongly increased the relative contribution of far dendritic currents. Increasing the number of firing cells also augmented the relative dendritic contribution to the somatic PS, an effect caused by the different waveform of somatic and dendritic unitary transmembrane currents (I(m)). Those from somata are short-lasting and spiky, having smaller temporal summation than those from dendrites, which are smoother and longer. The different shape of compartmental I(m)s is imposed by the fitting of backpropagating APs, which are large and fast at the soma and smaller and longer in dendrites. The maximum sodium conductance ((Na)) strongly affects the unitary APs at the soma, but barely the PS at the stratum pyramidale (st. pyr.). This occurred because somatic I(m) saturated at low (Na) due to the strong reduction of driving force during somatic APs, limiting the current contribution to the extracellular space. On the contrary, (Na) effectively defined the PS amplitude in the st. radiatum. The relative contribution of dendritic currents to the st. pyr. increases during the time span of the PS, from approximately 30-40% at the peak up to 100% at its end, a pattern resultant from the timing of active inward currents along the somatodendritic axis, which delay during backpropagation. Extreme changes imposed on dendritic currents caused only moderate effects on the st. pyr. due to reciprocal shunting of active soma and dendrites that partially counterbalance the net amount of instant current. The amplitude of the PS follows an inverse relation to the internal resistance (R(i)), which turned out to be a most critical factor. Low R(i) facilitated the spread of APs into dendrites and accelerated their speed, increasing temporal overlapping of inward currents along the somatodendritic axis and yielding the best PS reproductions. Model reconstruction of field potentials is a powerful tool to understand the interactions between different levels of complexity. The potential use of this approach to restrain the variability of some experimental measurements is discussed.

Modulation of dendritic action currents decreases the reliability of population spikes.

During synchronous action potential (AP) firing of CA1 pyramidal cells, a population spike (PS) is recorded in the extracellular space, the amplitude of which is considered a reliable quantitative index of the population output. Because the AP can be actively conducted and differentially modulated along the soma and dendrites, the extracellular part of the dendritic inward currents variably contributes to the somatic PS by spreading in the volume conductor to adjacent strata. This contribution has been studied by current-source density analysis and intracellular recordings in vivo during repetitive backpropagation that induces their selective fading. Both the PS and the ensemble action currents declined during high-frequency activation, although at different rates and timings. The decline was much stronger in dendrites than in the somatic region. At specific frequencies and for a short number of impulses the decrease of the somatic PS was neither due to fewer firing cells nor to decreased somatic action currents but to the blockade of dendritic action currents. The dendritic contribution to the peak of the somatic antidromic PS was estimated at approximately 30-40% and up to 100% at later times in the positive-going limb. The blockade of AP dendritic invasion was in part due to a decreased transfer of current from the soma that underwent a cumulative increase of conductance and slow depolarization during the train that eventually extended into the axon. The possibility of differential modulation of soma and dendritic action currents during APs should be checked when using the PS as a quantitative parameter.

Kainate receptors presynaptically downregulate GABAergic inhibition in the rat hippocampus.

Using microcultured neurons and hippocampal slices, we found that under conditions that completely block AMPA receptors, kainate induces a reduction in the effectiveness of GABAergic synaptic inhibition. Evoked inhibitory postsynaptic currents (IPSCs) were decreased by kainate by up to 90%, showing a bell-shaped dose-response curve similar to that of native kainate-selective receptors. The down-regulation of GABAergic inhibition was not affected by antagonism of metabotropic receptors, while it was attenuated by CNQX. Kainate increased synaptic failures and reduced the frequency of miniature IPSCs, indicating a presynaptic locus of action. In vivo experiments using brain dialysis demonstrated that kainate reversibly abolished recurrent inhibition and induced an epileptic-like electroencephalogram (EEG) activity. These results indicate that kainate receptor activation down-regulates GABAergic inhibition by modulating the reliability of GABA synapses.

Effects of the gliotoxin fluorocitrate on spreading depression and glial membrane potential in rat brain in situ.

DC extracellular potential shifts (deltaVo) associated with spreading depression (SD) reflect massive cell depolarization, but their cellular generators remain obscure. We have recently reported that the glial specific metabolic poison fluorocitrate (FC) delivered by microdialysis in situ caused a rapid impairment of glial function followed some hours later by loss of neuronal electrogenic activity and neuron death. We have used the time windows for selective decay of cell types so created to study the relative participation of glia and neurons in SD, and we report a detailed analysis of the effects of FC on evoked SD waves and glial membrane potential (Vm). Extracellular potential (Vo), interstitial potassium concentration ([K+]o), evoked potentials, and transmembrane glial potentials were monitored in the CA1 area before, during, and after administration of FC with or without elevated K+ concentration in the dialysate. SD waves propagated faster and lasted longer during FC treatment. DeltaVo in stratum pyramidale, which normally are much shorter and of smaller amplitude than those in stratum radiatum, expanded during FC treatment to match those in stratum radiatum. The coalescing SD waves that develop late during prolonged high-K+ dialysis and are typically limited to stratum radiatum, also expanded into stratum pyramidale under the influence of FC. SD provoked in neocortex normally does not spread to the CA1, but during FC treatment it readily reached CA1 via entorhinal cortex. Once neuronal function began to deteriorate, SD waves became smaller and slower, and eventually failed to enter the region around the FC source. Slow, moderately negative deltaVo that mirrored [K+]o increments could still be recorded well after neuronal function and SD-associated Vo had disappeared. Glial cell Vm gradually depolarized during FC administration, beginning much before depression of neuronal antidromic action potentials. Calculations based on the results predict a large decrease in glial potassium content during FC treatment. The results are compatible with neurons being the major generator of the deltaVo associated with SD. We conclude that energy shortage in glial cells makes brain tissue more susceptible to SD and therefore it may increase the risk of neuron damage.

Glutamate receptors of the kainate type and synaptic transmission.

Glutamic acid is an important excitatory neurotransmitter in the mammalian CNS. It has been established that synaptic transmission is mediated mostly by the ionotropic glutamate receptors AMPA and NMDA, with fast and slow kinetics, respectively. The recent demonstration in hippocampal neurones of a class of glutamate receptors that are activated by kainate and not by AMPA (that is, kainate-selective receptors) opens the possibility that receptors, others than those of the AMPA type, might also be involved in fast neurotransmission. The lack of specific pharmacological tools to dissect out AMPA from kainate receptors has hampered the functional study of kainate receptors. However, the recent finding that a 2,3-benzodiazepine (GYK153655) behaves as a selective antagonist of AMPA receptors allows us to address the question of the role of rapidly inactivating kainate receptors in synaptic transmission.

Heptanol but not fluoroacetate prevents the propagation of spreading depression in rat hippocampal slices.

We investigated whether heptanol and other long-chain alcohols that are known to block gap junctions interfere with the generation or the propagation of spreading depression (SD). Waves of SD were triggered by micro-injection of concentrated KCl solution in stratum (s.) radiatum of CA1 of rat hippocampal tissue slices. DC-coupled recordings of extracellular potential (V0) were made at the injection and at a second site approximately 1 mm distant in st. radiatum and sometimes also in st. pyramidale. Extracellular excitatory postsynaptic potentials (fEPSPs) were evoked by stimulation of the Schaffer collateral bundle; in some experiments, antidromic population spikes were evoked by stimulation of the alveus. Bath application of 3 mM heptanol or 5 mM hexanol completely and reversibly prevented the propagation of the SD-related potential shift (delta V0) without abolishing the delta V0 at the injection site. Octanol (1 mM) had a similar but less reliably reversible effect. fEPSPs were depressed by approximately 30% by heptanol and octanol, 65% by hexanol. Antidromic population spikes were depressed by 30%. In isolated, patchclamped CA1 pyramidal neurons, heptanol partially and reversibly depressed voltage-dependent Na currents possibly explaining the slight depression of antidromic spikes and, by acting on presynaptic action potentials, also the depression of fEPSPs. Fluoroacetate (FAc), a putative selective blocker of glial metabolism, first induced multiple spike firing in response to single afferent volleys and then severely suppressed synaptic transmission (confirming earlier reports) without depressing the antidromic population spike. FAc did not inhibit SD propagation. The effect of alkyl alcohols is compatible with the idea that the opening of normally closed neuronal gap junctions is required for SD propagation. Alternative possible explanations include interference with the lipid phase of neuron membranes. The absence of SD inhibition by FAc confirms that synaptic transmission is not necessary for the propagation of SD, and it suggests that normally functioning glial cells are not essential for SD generation or propagation.

Is glia disfunction the initial cause of neuronal death in ischemic penumbra?

The supportive role of glial cells on neuronal function and survival has been studied in anesthetized rats by using the selective gliotoxin fluorocitrate. Disabling glia operation reproduced many features of ischemic penumbra. An initial mild acidosis and increased interstitial potassium but not glutamate was followed after 3-4 h by repetitive spreading depression waves. These gradually provoked higher levels of acidosis, potassium and glutamate, gradual neuronal function decay and finally, neuron death. We conclude that neurons become highly vulnerable to spreading depression waves only in absence of normal glia operation. Our findings directly associate early glial disfunction to neuronal loss and lead to new insights for the understanding of ischemic pathology.

The effect of depressing glial function in rat brain in situ on ion homeostasis, synaptic transmission, and neuron survival.

The supporting role of glial cells in maintaining neurons and in ion homeostasis has been studied in situ by perfusing the gliotoxin fluorocitrate (FC) through a microdialysis fiber in the CA1 area of urethane-anesthetized rats. Extracellular direct current potential, extracellular potassium concentration ([K+]o) and amino acid levels, extracellular pH (pHo), and evoked field activity were studied. Histology verified the swelling of glial cells after 4 hr of FC treatment. Massive neuron damage was evident after 8 hr. FC dialysis caused the rapid decrease of glutamine, pHo became progressively more acid, and [K+]o moderately elevated. Orthodromic transmission was variably blocked within 30 min to 4 hr. After 4 hr, spreading depression (SD) waves that originated from the neocortex invaded hippocampal CA1, [K+]o increased to higher levels, pHo became very acid, and there were steep increases in taurine, glutamate, and GABA levels. Simultaneously, the antidromic population spike (a-PS) became depressed and eventually disappeared. When a shorter dialysis probe that spared cortex was used to sample CA1, no SD was seen, a-PS was not abolished, and ion homeostasis was altered less markedly. Repeated SD provoked in hippocampus in the absence of FC caused only mild depression of a-PS. Dialysis of high-K+ solution in healthy neocortex or hippocampus caused only slight elevation of [K+]o at distances of 200-400 microns from the dialysis membrane. After treatment with FC, similar high-K+ dialysis raised [K+]o much more. We conclude the following: (1) recurrent SD waves injure neurons if and only if glial function has failed; (2) neurons can regulate [K+]o, albeit imperfectly; (3) glia is required for the normal fine tuning of [K+]o and particularly for the recovery of pathologically elevated [K+]o; and (4) glia are required for the regulation of pHo. The similarities between glial poisoning by FC and the reported changes in the penumbra of ischemic infarcts suggest that the extension of neuron loss into the penumbral region might depend on failure of glial protection.

Role of neuronal synchronizing mechanisms in the propagation of spreading depression in the in vivo hippocampus.

To detect what initiates spreading depression (SD), the early prodromal events were investigated in hippocampal CA1 of urethane-anesthetized rats. SD was provoked by microdialysis or focal microinjection of high-K+ solution. Extracellular DC potentials and extracellular potassium concentration ([K+]o) were recorded, and spontaneous and evoked potentials analyzed for current source-density (CSD). In the front of an approaching SD wave, several seconds before the onset of the typical sustained negative potential shift (delta Vo) and the increased [K+]o, fast electrical activity was detected. This consisted initially of small rhythmic (60-70 Hz) sawtooth wavelets, which then gave way to a shower of population spikes (PSs) of identical frequency. Prodromal wavelets and PSs were synchronized over considerable distances in the tissue. Sawtooth wavelets were identified as pacemakers of the prodromal PS burst. Simultaneous recording at three depths revealed that the spontaneous prodromal PSs occurred exactly in phase in dendrites and somata whereas synaptically transmitted PSs arose first in the proximal dendrites and were conducted from there into the soma membrane. During a spike burst, stratum (st.) pyramidale served as current sink, while in the proximal sublayer of st. radiatum spike-sinks gave way to spike sources that grew larger as the sinks in st. pyramidale began to subside. Blocking synaptic transmission did not abolish the prodromal spike burst, yet repetitive orthodromic activation inhibited it without altering the subsequent SD waveform. Complex changes in cell excitability were detected even before fast spontaneous activities. We concluded that, in the initial evolution of SD, changes in neuron function precede the regenerating depolarization by several seconds. We propose that the opening of normally closed electric junctions among neurons can best explain the long-distance synchronization and the flow current that occurs in the leading edge of a propagating wave of SD.

Neuroprotective effect of acidic fibroblast growth factor on seizure-associated brain damage.

The neuroprotective effect of acidic fibroblast growth factor (aFGF) has been analysed in a rat model of seizures-associated brain damage. We report that after treatment with a convulsivant dose of Kainic acid, systemically administered aFGF prevents neuronal degeneration in specific brain areas, mainly in the hippocampal formation. Our findings extend the potential pharmacological use of fibroblast growth factors and afford new data to understand the neurophysiology of these proteins.