Fentanyl treatment reduces GABAergic inhibition in the CA1 area of the hippocampus 24 h after acute exposure to the drug

The effect of in vivo fentanyl treatment on synaptic transmission was studied in the CA1 area of the rat hippocampus. Animals were treated either with saline or fentanyl (4 x 80 microg/kg, s.c./15 min). Intracellular in vitro recordings were obtained, 24 h after treatment, from CA1 pyramidal neurons. No difference in pyramidal neuron basic membrane properties or postsynaptic membrane excitability was observed between neurons from saline- and fentanyl-treated animals. The peak amplitude of fast (f-) and slow (s-) components of IPSPs elicited in standard ACSF and the peak amplitude and rate of rise of isolated f- and s-IPSPs elicited in the presence of antagonists (CNQX, 10 microM; AP-5, 10 microM; CGP 55845, 1 microM; and bicuculline methochloride, 10 microM), in response to various stimulus intensities, was smaller in fentanyl-treated animals. Conversely, the rising slope of excitatory responses was similar in neurons from saline- and fentanyl-treated animals. Furthermore, in fentanyl-treated animals, lower stimulus strengths were required to elicit subthreshold excitatory responses of the same amplitude suggesting that acute exposure to fentanyl increases susceptibility of pyramidal neurons to presynaptic stimulation. GABA immunohistochemistry revealed lower GABA content in processes and neuronal somata suggesting diminished GABA release onto pyramidal neurons. We conclude that acute in vivo exposure to fentanyl is sufficient to induce long-lasting reduction in GABA-mediated transmission, rather, than enhanced excitatory transmission or modulation of the intrinsic excitability of pyramidal neurons. These findings provide evidence regarding the mechanisms involved in the early stages of tolerance development towards the analgesic effects of opioids.

The GABAA receptor-mediated recurrent inhibition in ventral compared with dorsal CA1 hippocampal region is weaker, decays faster and lasts less

Hippocampal functions appear to be segregated along the dorso-ventral axis of the structure. Differences at the cellular and local neuronal network level may be involved in this functional segregation. In this study the characteristics of CA1 recurrent inhibition (RI) were measured and compared between dorsal (DH, n = 95) and ventral (VH, n = 60) hippocampal slices, using recordings of suprathreshold field potentials. RI strength was estimated as the percentile decrease of the population spike (PS) amplitude evoked with an orthodromic stimulus (at the Schaffer collaterals) when preceded by an antidromic stimulus (at the alveus). Varying the interpulse interval (IPI) between the two stimuli, we estimated RI duration. Alvear stimulation produced significant PS suppression in both VH and DH at every IPI tested, from 10 to 270 ms. Moreover, gradually more oblique DH (but not VH) slices displayed increasing RI, which at IPIs < or = 125 ms was reversibly abolished by the GABAA receptor antagonist picrotoxin (10 microM). The GABAA-mediated RI, measured under the blockade of GABAB receptors, was weaker, decayed faster and lasted less in VH compared to DH slices, regardless of the slice orientation. Specifically, in VH compared to DH, the PS suppression at 20 ms was 34.4 +/- 4.5% versus 69.9 +/- 6.5% (P < 0.001), the time constant of RI decay was 29 +/- 2.4 versus 87.5 +/- 13.6 ms (P < 0.01) and the duration was 50 versus 125 ms (P < 0.001). Thus, GABAA-mediated RI may control the CA1 excitatory output less effectively in VH compared to DH. The observed dorso-ventral differences in RI contribute to the longitudinal diversification of the structure and may underlie to some extent the region-specificity of hippocampal functions.

Long-term potentiation of high-frequency oscillation and synaptic transmission characterize in vitro NMDA receptor-dependent epileptogenesis in the hippocampus

The implication of high-frequency network oscillations (HFOs) in brain pathology resides in as yet unclear mechanisms. Employing field recordings from ventral hippocampal slices and two models of epileptogenesis (i.e. establishment of interictal-like persistent bursts), we found that HFOs associated with epileptiform bursts and excitatory synaptic transmission were co-modulated during epileptogenesis. NMDA receptor-dependent epileptogenesis in CA3 was consistently accompanied by long-lasting strengthening in synaptic transmission (by 94+/-17%, n=5) and HFOs (frequency, power and duration increased by 24+/-8%, 57+/-18% and 33+/-10%, respectively). Co-modulation of synaptic transmission and HFOs was also observed in NMDA receptor-independent epileptogenesis, although in individual experiments either enhancement or depression of both phenomena was observed. Pathological HFOs >200 Hz were unequivocally present in persistent bursts induced by NMDA receptor-dependent but not NMDA receptor-independent mechanisms. The duration of pathological HFOs associated with persistent bursts but not of HFOs associated with bursts before the establishment of epileptogenesis was linearly and strongly correlated with the duration of bursts (r=0.58, P<0.0001). We propose that interplay between spontaneous synchronous bursting and long-lasting synaptic potentiation accompanying certain forms of epileptogenesis may underlie long-lasting potentiation of HFOs, whose quantitative aspects may reliably signal the degree of network changes involved in epileptogenesis.

Primary pleural hemangiopericytoma-like tumor: an unusual localized fibrous tumor of the pleura (2007: 4b)

In this report, the case of a 65-year-old woman with cough and dyspnoea is described. The chest radiographs showed a sharply demarcated mass located in the left paravertebral hemithorax and a left pleural effusion. Computed tomography after intravenous contrast material showed a heterogeneous, mainly cystic mass with peripheral enhancement and enhancing intratumoral foci and septa. Magnetic resonance imaging confirmed the location and enhancement pattern of the lesion. Histology following the surgical excision of the tumor showed a localized fibrous tumor of the pleura and specifically the hemangiopericytoma-like type.

Greater contribution of N-methyl-D-aspartic acid receptors in ventral compared to dorsal hippocampal slices in the expression and long-term maintenance of epileptiform activity

Functional segregation along the dorso-ventral axis of the hippocampus is a developing concept. The higher susceptibility of the ventral hippocampus to epileptic activity compared with dorsal hippocampus is one of the main features, which still has obscure mechanisms. Using the model of magnesium-free medium and field recordings, single epileptiform discharges displayed higher incidence (77% vs 57%), rate (41.7+/-3.1 vs 13.5+/-0.7 events/min), duration (173.9+/-17.7 vs 116.8+/-13.6 ms) and intensity (coastline, 25.4+/-2.5 vs 9.5+/-1.8) in ventral compared with dorsal rat hippocampal slices. In addition, the decay phase of the evoked synaptic potentials was 110% slower in ventral slices. The N-methyl-D-aspartate (NMDA) receptor antagonist d-(-)-2-amino-5-phosphonopentanoic acid (50-100 microM) decreased the discharge rate and coastline similarly in ventral and dorsal slices, but it shortened the discharges in ventral slices (by 40%) only. The NMDA receptor antagonist 3-((R)-2-carboxypiperazin-4-yl)-propyl-1-phosphonic acid (10 microM) decreased the rate in both groups and additionally shortened discharges in both kinds of slices, an effect which was greater in ventral ones (31% vs 13%). Furthermore, both drugs shortened the evoked potentials more in ventral (77%) than in dorsal slices (52%). On the other hand, 1 microM of 3-((R)-2-carboxypiperazin-4-yl)-propyl-1-phosphonic acid shortened the discharges and evoked synaptic potentials only in ventral slices, and slowed down the discharge rate only in dorsal slices. Addition of NMDA, in the magnesium-free medium, enhanced activity in both kinds of slices. At 5 and 10 microM of NMDA 51% of the ventral but only 9% of the dorsal slices displayed persistent epileptiform discharges, which were recorded for at least one hour after reintroduction of magnesium in the medium. At 10-20 microM the enhancement of activity was transient, followed by suppression of discharges in 40% and 76% of the ventral and dorsal slices, respectively. Most of the slices having experienced suppression did not develop persistent activity. We propose that the NMDA receptors contribute to the higher susceptibility of the ventral hippocampus to expression and long-term maintenance of epileptiform discharges. This diversification may be related to other aspects of hippocampal dorso-ventral functional segregation.

Up-regulation of adenosine A1 receptor binding in pentylenetetrazol kindling in mice: effects of angiotensin IV

The effects of the hexapeptide angiotensin II (3-8) ANG IV, the selective A(1) receptor agonist cyclohexyladenosine (CHA) and the combination of ANG IV + CHA on pentylenetetrazol (PTZ)-generalized seizures; kindling development and maintenance were studied. By using in vitro quantitative receptor autoradiography, the regulation of adenosine A(1) receptor density at different time points during the kindling procedure and postkindling period was determined. ANG IV and CHA effectively reduced clonic seizures in PTZ-generalized seizure model, in PTZ-kindled mice as well as during kindling development and a week later by rechallenge with PTZ. Furthermore, coadministration of ANG IV and CHA had a strong anticonvulsant effect, both compounds acting synergistically. A significant increase of adenosine A(1) receptor density was detected in somatosensory cortex, hippocampus, amygdala and geniculate nuclei early in the kindling procedure (after the 3rd injection), which persisted at least 1 month after the end of kindling procedure. In addition, a delayed up-regulation of adenosine A(1) receptor binding was observed a week after kindling in the mamillary bodies and a month later in the motor cortex. The pretreatment with ANG IV caused a down-regulation of adenosine A(1) receptor density to the control level in most time points and brain areas. In conclusion, PTZ kindling-induced increase of adenosine A(1) receptor binding at different time points and in specific brain structures might represent an adaptive mechanism for coping with the hyperexcitability typical for this phenomenon. The antiepileptogenic effect of ANG IV could be realized partly through an adenosine-dependent mechanism.

A computer model of field potential responses for the study of short-term plasticity in hippocampus

Activity-dependent synaptic plasticity has important implications for network function. The previously developed model of the hippocampal CA1 area, which contained pyramidal cells (PC) and two types of interneurons involved in feed-forward and recurrent inhibition, respectively, and received synaptic inputs from CA3 neurons via the Schaffer collaterals, was enhanced by incorporating dynamic synaptic connections capable of changing their weights depending on presynaptic activation history. The model output was presented as field potentials, which were compared with those derived experimentally. The parameters of Schaffer collateral-PC excitatory model synapse were determined, with which the model successfully reproduced the complicated dynamics of train-stimulation sequential potentiation/depression observed in experimentally recorded field responses. It was found that the model better reproduces the time course of experimental field potentials if the inhibitory synapses on PC are also made dynamic, with expressed properties of frequency-dependent depression. This finding supports experimental evidence that these synapses are subject to activity-dependent depression. The model field potentials in response to various randomly generated and real (derived from recorded CA3 unit activity) long stimulating trains were calculated, illustrating that short-term plasticity with the observed characteristics could play specific roles in frequency processing in hippocampus and thus providing a new tool for the theoretical study of activity-dependent synaptic plasticity.

Vascular network of the rat hippocampus is not homogeneous along the septotemporal axis

Qualitative and quantitative image analysis of hippocampal vascular bed, after transcardial perfusion of India ink, reveals significant differences among hippocampal subfields and along the septotemporal axis of the rat hippocampus. Ventral hippocampus exhibits significantly higher levels of vascularization compared to dorsal hippocampus, which, however, is characterized by significantly higher capillary density. These results may explain the selective ischemia vulnerability of hippocampus along its septotemporal axis.

Weaker synaptic inhibition in CA1 region of ventral compared to dorsal rat hippocampal slices

Extracellular and intracellular recordings were made from slices taken from the dorsal (DH) and ventral (VH) part of rat hippocampus. Using paired-pulse stimulation of Schaffer collaterals, at different interpulse intervals (IPIs), and records of the population spike (PS) we found that the strength and duration of paired-pulse inhibition was much weaker in VH compared to DH slices: at the IPI of 10 ms the decrease of PS in VH (40%) was significantly smaller compared to that in DH slices (76%), while at 20 ms the decrease of PS in DH slices (60%) corresponded to facilitation in VH slices. Moreover, the amplitude and duration of intracellularly recorded fast inhibitory postsynaptic potentials (fast-IPSPs) were found significantly smaller in VH (5.2+/-0.6 mV, 54.8+/-5.8 ms) than in DH (11.2+/-1.1 mV, 105+/-10 ms) neurons. The smaller and shorter fast-IPSP recorded in VH neurons may at least in part explain the results in paired-pulse inhibition. The demonstrated weaker inhibition may underlie the higher propensity of the ventral hippocampus for epileptiform activity.

Bronchogenic cyst infected by Salmonella enteritidis followed gastroenteritis

Congenital bronchogenic cysts of the lung and mediastinum develop from the ventral foregut during embryogenesis. Bronchogenic cysts are seldom seen in the adults and most are thought to be asymptomatic and free of complications unless they become infected or are large enough to cause pressure on contiguous vital structures such as the tracheal carina, the lung or the esophagus. We present the unique case of a 24-year-old man who developed respiratory symptoms after Salmonella enteritidis infected bronchogenic cyst following Salmonella gastroenteritis.

Spontaneous GABA(A)-dependent synchronous periodic activity in adult rat ventral hippocampal slices

The present study shows that adult rat transverse slices from the ventral hippocampus perfused with standard medium persistently generate spontaneous synchronous field potentials. In CA1 st. pyramidale this regular ventral hippocampus spontaneous synchronous activity (VHSSA) was positive with mean amplitude 0.18 +/- 0.02 mV (n=80 slices) and occurred every 0.48 +/- 0.02 s. Simultaneous intracellular recordings from CA1 pyramidal neurons demonstrated that concomitant hyperpolarizations invariably occurred in association to this field activity and could thus constitute its electrical generators. These hyperpolarizations, had mean amplitude 2.7 +/- 0.6 mV, duration at half amplitude 44.8 +/- 6.6 ms, they reversed at -72.6 +/- 1.5 mV (n=10 cells), they effectively suspended the depolarization-induced tonic neuronal firing of all ten pyramidal neurons and they were reversibly abolished, together with field potentials, by the GABA(A) receptor antagonist bicuculline (5 microM, n=4). VHSSA was also dependent on fast glutamatergic transmission, since it was blocked by the antagonist of AMPA/kainate receptors 6-Cyano-7-nitroquinoxaline-2,3-dione disodium (10 microM, n=3). We propose that, under standard in vitro conditions, synchronous GABA(A)-mediated hyperpolarizing potentials are spontaneously generated in pyramidal neurons presumably resulting from the phasic quasi-rhythmic discharge of a local interneuronal network of ventral hippocampus.

Spontaneous, low frequency (approximately 2-3 Hz) field activity generated in rat ventral hippocampal slices perfused with normal medium

This study demonstrates that transverse slices taken from the ventral hippocampus of adult rats perfused with a medium of normal ionic composition sustain spontaneous periodic field potentials due to the synchronous activity of a population of neurons. This ventral hippocampus spontaneous synchronous activity (VHSSA) in CA1 stratum pyramidale consisted of positive potentials (approximately 0.12 mV, 55 ms) occurring at a frequency of 2.8 +/- 0.2 Hz for hours without interruption. VHSSA was most frequently observed in slices taken 1-3 mm from the ventral end of hippocampus, and was absent in slices taken from tissue more than 4.5 mm away from it. Stimulation of Schaffer collaterals primed the appearance of potentials, which were similar to VHSSA and clearly distinguishable from excitatory postsynaptic potentials. In view of the known relative proneness of ventral hippocampus to epilepsy, we perfused ventral slices with high-[K(+)](o) medium (8 mM). Albeit reduced in amplitude, VHSSA persisted during the high-[K(+)](o) induced interictal-like epileptiform activity. We could not document any temporal relationship between the two phenomena. Low concentrations of the antagonist of gamma-amino-butyric acid receptors, type A, bicuculline (2-3 microM), which enhanced the high-[K(+)](o) induced epileptiform activity, reversibly blocked the VHSSA. We conclude that under standard in vitro conditions small circuits in the ventral hippocampus are most often and for long periods of time engaged in synchronous quasi-rhythmic low-frequency activity, generated locally by mechanisms substantially differing from those supporting epileptiform discharges.

Dorsal-ventral differentiation of short-term synaptic plasticity in rat CA1 hippocampal region

Two forms of short-term synaptic plasticity (STP), paired-pulse facilitation (PPF) and frequency potentiation (FP) of CA1 field excitatory postsynaptic potentials (EPSP) to afferent stimulation were compared in slices taken from the dorsal and ventral parts of rat hippocampus. While dorsal slices showed significant PPF at all interpulse intervals (20-1400 ms, 80% at 40 ms), PPF in ventral slices was substantially weaker at intervals shorter than 100 ms (19% at 40 ms) and nil at longer intervals. While dorsal slices showed substantial FP at frequencies 1-40 Hz and frequency depression at 50-100 Hz, ventral slices showed only a much smaller potentiation at 1 Hz and substantial depression at 20-100 Hz. Decreasing [Ca(2+)](o) from 2 to 1 and 0.5 mM substantially reduced the baseline EPSPs in both groups of slices but its effect on PPF was greater in ventral slices. On the contrary when [Ca(2+)](o) was increased to 5 mM only dorsal slices showed an enhancement of baseline EPSP. It is concluded that ventral hippocampus CA1 area has a specific deficit in STP, which is related to the important presynaptic role of calcium and is consistent with a relatively higher transmitter release probability.

Reduction of A1 adenosine receptors in rat hippocampus after kainic acid-induced limbic seizures

In a temporal lobe epilepsy (TLE) model induced by kainic acid (KA), we examined the effect of limbic seizures on A1 adenosine receptor distribution in hippocampus and cortex. By using quantitative autoradiography, we determined a progressive decrease in A1 receptor density in CA1 and CA3 regions of hippocampus, which coincided in time with the degenerating process of hippocampal pyramidal cells. This result indicates that a great amount of A1 receptors are located postsynaptically on pyramidal cell dendrites. No difference in A1 receptor density was observed in the inner compared to the outer molecular layer of dentate gyrus, or in the infrapyramidal band compared to the outer layer of stratum oriens of CA3. This could indicate that the newly sprouted mossy fiber glutamatergic terminals do not contain A1 receptors, thus lacking a restrain in the release of glutamate.

Decreased ability of rat temporal hippocampal CA1 region to produce long-term potentiation

Tetanic stimulation of Schaffer collaterals in the CA1 region of transverse slices, taken from the septal (dorsal) part of young rat hippocampus, produced N-Methyl-D-aspartate-dependent long-term potentiation (LTP) of the rising slope of excitatory postsynaptic potential (mean 38%). Under identical conditions of stimulation (100 Hz, 1 s) slices taken from the temporal (ventral) third of hippocampus presented a substantially reduced ability for LTP (mean 5%). The defect appeared to lie with the induction rather than the maintenance phase of LTP. These results suggest that a significant functional differentiation at the local synaptic plasticity level occurs between the two poles of hippocampus, which together with the substantial differences in their extrinsic connections, may help explain the reported differential participation of neurons in these parts of hippocampus during animal memory tests.

Lower density of A1 adenosine receptors in nucleus reticularis thalami in rats with genetic absence epilepsy

The possible involvement of the adenosinergic modulatory system in the pathogenesis of absence seizures was investigated in genetic absence epilepsy rats from Strasbourg (GAERS). Using in vitro quantitative autoradiography, the distribution of A1 adenosine receptors and adenosine uptake sites in the brain of GAERS was studied and compared to that of control animals. An area-specific lower density of A1 receptors (15% decrease) was detected in reticular (nRT) and anterior ventral (AV) thalamic nuclei as well as basal ganglia in the brains of GAERS animals compared with control animals. Since adenosine exerts an anti-oscillatory effect on the thalamic nuclei by suppressing (via A1 receptors) excitatory as well as inhibitory neurotransmitter release, the impairment in A1 receptor density seen here, especially in nRT, could be implicated in the thalamic rhythmicity underlying spike and wave discharges present in this absence epilepsy model.

Development of a transient increase in recurrent inhibition and paired-pulse facilitation in hippocampal CA1 region

Paired-pulse recurrent inhibition (RI) of population spike (PS) and facilitation (PPF) of field excitatory postsynaptic potential (EPSP) were studied in the CA1 region of hippocampal slices taken from Wistar rats aged from 9 days to 16 months. The comparison of three different paired-pulse protocols revealed the antidromic-orthodromic (A-O) stimulation as the most reliable in quantifying the strength of fast (peaking at 10 ms) and slow (peaking at 200 ms) components of recurrent inhibition. Fast RI, present but weak at 9 days, progressively increased to reach its maximal strength at 30 days, declining in adult (2 m) and middle-aged (16 m) animals. Slow RI was replaced by facilitation at 9 days while it was absent at 15 days. It reached adult values at 30 days. A reduction of the test response at interpulse interval (IPI) of 2-4 ms was strong in developing and adult animals, but was significantly decreased in 16 m. At maximal stimulation PPF was expressed as an enhancement of the slow rather than the fast phase of the EPSP and was particularly strong with a prominent N-methyl-D-aspartate dependent component. A very characteristic selectivity for a prominent PPF at stimulation frequency of 5 Hz appeared first at the 18th day and increased gradually to reach a maximum at the 30th day, after which it declined to very low values in middle-aged animals. A similar developmental pattern was observed in slices taken from rats reared in complete darkness, suggesting a strong innate origin. The ability of hippocampal circuits for plastic gating of information appears to be transiently enhanced at the completion of the first postnatal month as it can be exercised at a wider part of the frequency spectrum, with maximal inhibition and potentiation especially at the frequency of theta rhythm.

Age-related changes in excitability and recurrent inhibition in the rat CA1 hippocampal region

In hippocampal slices from male Wistar rats aged 1-34 months, we recorded the synaptic field potential responses of the CA1 neurons to stimulation of Schaffer collaterals. Eight electrophysiological indexes were extracted from input/output curves and compared in 11 age groups from 1 to 30 months. Neuronal excitability presented a U-shaped curve of development with a minimum at approximately 7-8 months of age. There was a significant continuous increase in neuronal excitability, i.e. a decrease in excitatory postsynaptic potential (EPSP) producing both the threshold and half-maximal population spike from middle age (8-10 months) to senescence (30 months). Synaptic efficiency also increased in old rats to reach a maximum during senescence, i.e. both the current for threshold EPSP and that for half-maximal EPSP reached a minimum in senescence, although the earlier developmental patterns of these two indexes were non-linear. The duration of the field EPSP elicited with maximal stimulation presented an abrupt decay after the first month. Aged animals presented a relatively small maximal population spike. Recurrent inhibition was most prominent on neuronal excitability rather than synaptic strength. Measured as the percentage change in the half-maximal EPSP and half-maximal population spike, recurrent inhibition was found to decrease during the first 7-10 months of life and remained small in later development.

Simulation study for the transition from spindles to spike and wave epileptogenesis

A mathematical model is presented, based on existing anatomical and physiological data, which simulates the behaviour of representative types of cortical cells. It is used to test whether a set of synaptic connections of these cells exists, which, paced by the same rhythmical thalamic input, could produce spindles under normal conditions and spike and wave discharges (SW) under conditions of cortical hyperexcitability. This is possible if the interneurons do not provide recurrent excitatory or inhibitory input on themselves, if the thalamic afferents contact the cortical projecting pyramidal cells through local excitatory neurons, and if the inhibitory interneurons receive input only from the pyramidal cells. The results suggest that an increase of all cortical synaptic actions (both excitatory and inhibitory) is sufficient for the transition from spindles to the first stages in the development of SW discharges in the cortex, whereas the thalamus can be driven to the SW characteristic frequency at the immediate next stages.

Pentylenetetrazol-induced seizures decrease gamma-aminobutyric acid-mediated recurrent inhibition and enhance adenosine-mediated depression

To elucidate the consequences of convulsions, we examined biochemically and electrophysiologically the brains of mice that had sustained two complete tonic-clonic convulsions after administration of pentylenetetrazol (PTZ 50 mg/kg intraperitoneally, i.p.), 48 and 24 h before decapitation. Control mice were injected with saline. Input/output curves of the extracellular synaptic responses in the CA1 area of hippocampal slices showed that PTZ-induced seizures do not establish the persistent change in hippocampal excitability itself that can be detected in vitro. However, use of the paired-pulse stimulation paradigm showed that gamma-aminobutyric acid A (GABAA)-mediated recurrent inhibition was significantly weaker (by 19-25%) in the CA1 area of slices from PTZ-treated mice (PTZ slices) as compared with slices from control mice (control slices). The density of GABAA receptors (high-affinity component) was also lower in hippocampus (by 19%) and cortex (by 14%) of PTZ-treated mice. A GABA-related disinhibitory mechanism underlying PTZ seizures may thus persist for 1 day after the seizure, predisposing the brain to subsequent seizures. On the other hand, the depressant effect of a single dose of adenosine 10 microM on the CA1 synaptic response was stronger (by 35% on population spikes) and longer lasting in PTZ slices as compared with controls. This could be attributed to significantly higher adenosine A1 receptor density in hippocampus (Bmax of [3H]CHA was higher by 34%) as well as cortex and cerebellum of these animals. The phenomenon may reflect an adenosine A1-mediated adaptive mechanism that offers protection from subsequent seizures.

Upregulation of A1 adenosine receptors in human temporal lobe epilepsy: a quantitative autoradiographic study

A significant increase of A1 adenosine receptor binding (48% increase of mean) was detected in human neocortex obtained from patients suffering from temporal lobe epilepsy as compared to control neocortex from non-epileptic patients. Such increase was equally distributed in the six cortical layers and reached similar levels in each of the five specimens tested independently of age, sex and pharmacological treatment of the patient. Since adenosine exerts a depressant effect on neocortical neurons in slices obtained from epileptic patients, this upregulation of A1 receptor binding may constitute a protective mechanism against subsequent seizures, which is exerted by elevating the depressant response of the brain to endogenous adenosine.

Effect of pentylentetrazol-induced seizures on A1 adenosine receptor regional density in the mouse brain: a quantitative autoradiographic study

Adenosine has been shown to be a major regulator of neuronal activity in convulsive disorders, exerting its anticonvulsant effect through central A1 adenosine receptors. The aim of the present study was to investigate the effect of generalized tonic-clonic seizures induced by pentylentetrazol on regional changes in A1 adenosine receptor density and distribution in the mouse brain by in vitro quantitative autoradiography. As radioligand the specific agonist of A1 receptors [3H]cyclohexyladenosine was used. After two consecutive (once daily) pentylentetrazol-induced convulsions a widespread upregulation of A1 receptor density was detected with a marked enhancement in structures that mediate seizure activity like hippocampus, mamillary bodies, septum, substantia nigra, thalamic nuclei and cerebral cortices. On the contrary, in basal ganglia a significant downregulation of A1 receptors was observed. These results indicate that: (i) the observed increases or decreases in A1 receptor density are organized in selective anatomical structures related to seizure development rather than uniform in the brain; and (ii) since the upregulation of A1 receptors is sufficient to enhance the physiological depressive response of adenosine, the overall evoked increases seen here may lead to a stronger inhibitory tone and accordingly to a more efficient anticonvulsant effect of endogenous adenosine.

Comparison of the effects of increased potassium and of adenosine on hippocampal neurons from normal and genetically epileptic tg/tg mice

Tottering mice are an experimental model of genetically determined generalized epilepsy of the absence type. We investigated possible mechanisms underlying epileptogenic hyperexcitability in these mice by studying input/output (I/O) curves of the extracellular response of CA1 neurons to stratum radiatum stimulation in hippocampal slices maintained in vitro. Increases in extracellular potassium are considered to contribute to epileptogenesis, whereas adenosine has been proposed to be an endogenous antiepileptic agent. Moderate elevations (+2 mM) of extracellular K+ concentrations induced a significantly smaller increase of this response (leftward shift of the input/output curves) in slices from epileptic mice as compared with controls. Perfusion of slices with adenosine 10 microM decreased excitability in both groups of slices, especially with regard to response threshold. Adenosine more effectively decreased the responses elicited by low-intensity stimulation than those elicited by high intensity. No significant difference between the groups of slices was observed. On the basis of the present data, it is unlikely that the previously observed hyperexcitability of hippocampal neurons of tottering mice results from a genetically altered sensitivity to moderate increases in [K+]o or to adenosine.

Long-term increase in excitability induced by Mg(2+)-free medium in the absence of afferent stimulation in CA1 area of mouse hippocampal slices

Omission of Mg ions from the perfusion fluid of hippocampal slices unblocks the N-methyl-D-aspartate (NMDA) type of glutamate receptor/channel, and induces long term enhancement of synaptic responses. In order to test the role of afferent activation in induction of long term potentiation in CA1 area by this process, we switched off stimulation during the time of perfusing the slices with Mg(2+)-free medium (30 min). In addition to a short lasting increase in synaptic activation we observed a long term increase in population spike amplitude in all slices tested (n = 5), which lasted for at least 2 h. This process was antagonized by 50-100 microM DL-2-amino-5-phosphonovalerate, a specific NMDA receptor antagonist (n = 8), but not by isolating CA1-CA3 areas prior to the testing (n = 5). These results suggest that the resting levels of the endogenous excitatory neurotransmitter(s) can induce long term increase in firing probability of CA1 pyramidal cells, when NMDA channels are unblocked, in the absence of afferent stimulation and irrespective of CA3 area prior excitation.

Changes in seizure latency correlate with alterations in A1 adenosine receptor binding during daily repeated pentylentetrazol-induced convulsions in different mouse brain areas

The seizure latency changed during daily pentylentetrazol (PTZ) induced convulsions showing an increase between days 2 and 4, a rapid decrease between days 5 and 10 and a slight increase again between days 11 and 14. At the respective timepoints, [3H]CHA binding, in cortex and cerebellum of PTZ treated animals followed exactly the same pattern, suggesting that the alterations in A1 receptors in these areas may partly determine the PTZ seizure latency curve. On the contrary, the changes of [3H]CHA binding in hippocampus (sustained increase) and striatum (sustained decrease) didn't follow the latency curve pattern. These results suggest that changes in A1 receptor density in specific brain areas may be involved in the modulation of seizure susceptibility.

Electrophysiological analysis of human neocortex in vitro: experimental techniques and methodological approaches

In this review we summarize a number of technical and methodological approaches that have been used in our laboratory to study human brain slices maintained in vitro. The findings obtained in the course of these studies appear to be relevant in establishing the mechanisms that underlie physiological phenomena of the human brain such as synaptic plasticity or responses to neuroactive drugs. Moreover, these data are important for understanding certain fundamental mechanisms of epilepsy. In this respect, however, we caution that the mechanisms that apply to different forms of clinical epilepsy might be difficult to find given the variability present in the pathogenesis of human epilepsy.

A comparison of recurrent inhibition and of paired-pulse facilitation in hippocampal slices from normal and genetically epileptic mice

Tottering mice exhibit inherited generalized epilepsy of the 'absence' type. In hippocampal slices from these mutant mice studied in vitro, pairing an alvear antidromic stimulus to an orthodromic one revealed a strong recurrent inhibition (RI) of CA1 pyramidal neurons. RI was maximal at 10 ms inter-pulse interval (IPI 70% decrease of population spike, PS) gradually decreasing to 15% at 320 ms IPI. At 10 ms IPI it shifted the input/output curves to the right and decreased maximum PS. In the group of slices from epileptic mice the early part of RI (2.5-60 ms) was indistinguishable from that of normal mice, with respect to both its strength and its liability to activity-dependent decrement induced by a train of antidromic stimuli (8 s, 5 Hz). However, the delayed part (80-320 ms) was slightly stronger in the epileptic group. Also in this group only the train of antidromic pulses caused a significant and lasting decrease in the unconditioned orthodromic PS. Paired-pulse facilitation was equally strong in the 2 groups of slices. It is concluded that mechanisms underlying epileptogenic hyperexcitability in the tottering mutant may not include a failure of inhibition, at least in the CA1 area of the hippocampus. On the contrary some inhibitory mechanisms may be stronger.

Alterations of A1 adenosine receptors in different mouse brain areas after pentylentetrazol-induced seizures, but not in the epileptic mutant mouse 'tottering'

Single and repeated Pentylentetrazol (PTZ)-induced convulsions are associated with significant changes of A1 adenosine receptors (detected using the radioligand [3H]cyclohexyladenosine, [3H]CHA) in 4 different brain areas of the mouse, namely cortex, hippocampus, cerebellum and striatum. In hippocampus and cerebellum, a rapid increase in [3H]CHA binding, by 26% and 30% respectively, was observed 1 h after a single PTZ convulsion. In striatum, on the contrary, a significant decrease by 30% in [3H]CHA binding was seen, whereas in cortex no significant change could be detected. After daily repeated PTZ convulsions, a significant increase of A1 receptors by 26% appeared also in cortex, while the changes of A1 receptors observed in the other brain areas after a single PTZ convulsion were maintained in almost the same range. All the alterations observed were due to changes of the total number of A1 receptors (Bmax) without changes in receptor affinity (Kd). A significant increase in the latency of PTZ seizure (time between the PTZ-injection and the beginning of the seizure) was also observed after repeated PTZ-induced convulsions at the time when the changes in A1 adenosine receptors were noted. Considered together, these results provide further evidence for an A1 receptor-mediated modulation of seizure susceptibility and indicate that specific brain areas may play different roles in this modulation. The binding of [3H]CHA to membranes from different cortical and subcortical areas of the epileptic mutant mouse 'tottering' was not different from that in control animals.

An electrophysiological study of the ontogenesis of adenosine receptors in the CA1 area of rat hippocampus

The depressant effect of adenosine (Ad) was studied electrophysiologically in hippocampal slices from 5-, 10-, 15-, 20-, 30- and 120-day-old rats. Ad (10 microM) depressed the field EPSP in CA1 to the same extent in all age groups. Caffeine (Caf), an Ad receptor antagonist, enhanced and nitrobenzylthioinosine (NBI), an Ad uptake blocker, depressed the field EPSP. Both these effects were, however, less prominent in slices from younger animals, a finding consistent with lower extracellular levels of endogenous Ad in neonatal rats.

Increased postsynaptic excitability in hippocampal slices from the tottering epileptic mutant mouse

The tottering mouse exhibits an inherited form of generalized epilepsy, which can be characterized by electroencephalographic, behavioral and pharmacological criteria as belonging to the 'absence' type. In vitro electrophysiological experiments in hippocampal slices revealed a higher than normal postsynaptic excitability in slices from epileptic mice. Upon stimulation of Schaffer collaterals, we obtained input/output curves from the CA1 pyramidal cell layer and determined several indices of synaptic activation and postsynaptic excitability. Only the latter were found to be statistically different: population spikes were elicited by relatively smaller field EPSPs (P less than 0.001) in the slices from epileptic mice. However, their maximum population spike was significantly smaller, which indicated that fewer neurons were available for firing. In the normal but not in the epileptic mice in vitro postsynaptic excitability was correlated to the age of the animal.

Long-term enhancement of postsynaptic excitability after brief exposure to Mg2(+)-free medium in normal and epileptic mice

Brief exposure to Mg2(+)-free medium (MFM) enhanced the population response of CA1 neurons to stratum radiatum stimulation in hippocampal slices from normal (+/?) and epileptic tottering (tg/tg) mice. The enhancement was maintained in both groups for at least 2 h following reperfusion with normal medium (NM). Excitability curves obtained from the extracellular records suggest that, while both synaptic activation and postsynaptic excitability are enhanced during MFM perfusion, only the latter enhancement is maintained at significant levels after reperfusion with NM. The long-term increase in postsynaptic excitability was comparable in strength to that produced by long-term potentiation (LTP) inducing tetanic stimuli, was accompanied by an increase in the slope of the population spike/field excitatory postsynaptic potential (PS/fEPSP) curve and did not appear to depend on the induction of epileptiform activity by MFM. Both the short- and the long-term effects of MFM on synaptic activation and postsynaptic excitability were qualitatively similar in normal and epileptic mice and any quantitative differences were not statistically significant. Thus, epileptogenesis in the tottering mutant may not involve a change in the NMDA receptor-mediated control of excitability, at least in the CA1 area of hippocampus.

Endogenous adenosine can reduce epileptiform activity in the human epileptogenic cortex maintained in vitro

The effects induced by adenosine and some related compounds upon Mg2+-free epileptogenesis were studied in slices of human epileptogenic neocortex maintained in vitro. Extracellular recordings revealed stimulus-induced and spontaneous epileptiform activity within 1-2 h of perfusion with Mg2+-free medium. A 30-90% decrease of the frequency of occurrence of spontaneous epileptiform discharges was induced by 40-50 microM adenosine while the analog 2-Cl-adenosine exerted a depressant effect (greater than 75% reduction in frequency of occurrence) at 0.3-3 microM. 2-Cl-adenosine also depressed stimulus-induced epileptiform responses and often blocked spontaneous epileptiform activity. Similar effects were seen during bath application of the adenosine uptake inhibitor nitrobenzylthioinosine (10-50 microM) indicating that endogenous adenosine can by itself influence epileptogenicity. Our data demonstrate that in the human epileptogenic neocortex a purinergic mechanism can control Mg2+-free epileptiform activity.

Membrane properties, response to amines and to tetanic stimulation of hippocampal neurons in the genetically epileptic mutant mouse tottering

The petit-mal seizures of the "tottering" mutant mouse (tg) have been attributed to an exaggerated noradrenergic projection from locus coeruleus to the telencephalon (Noebels 1984). In order to investigate the possible epileptogenic mechanisms involved, we have compared hippocampal slices from epileptic (tg/tg) and phenotypically healthy (tg/+) mice. Resting potentials, action potentials and afterpotentials, membrane impedances and time constants were not significantly different in 11 neurons from each group. Bath application of noradrenaline, isoproterenol and histamine or a transient exposure to Mg++-free medium caused a long lasting increase in extracellularly recorded population spikes induced in CA1 by electrical stimulation of stratum radiatum. Isoproterenol blocked the calcium dependent afterhyperpolarization and accommodation of firing. Tetanization of afferent fibres evoked post-tetanic potentiation and long-term potentiation. All these results are qualitatively similar to those previously described in rats and guinea pigs and have revealed no significant difference between tg/tg and tg/+ mice.

Caffeine blocks absence seizures in the tottering mutant mouse

The neurological tottering mutant mouse is characterized by frequent "absence" seizures accompanied by bilateral synchronous spike and wave EEG bursts. Under anesthesia, adult homozygous tottering mice were implanted with permanent epidural electrodes, and at least 7 days elapsed before electrocorticograms in unrestrained mice were scored for seizure incidence and duration. Caffeine (5, 10, 15 mg/kg, n = 8) injected intraperitoneally (i.p.) at the fourth hour of 8-h recording sessions significantly (p less than 0.001 for 10 and 15 mg) decreased seizure incidence as compared with control saline injections. Spike and wave bursts were eliminated during the 30 min after injection and reached 50% preinjection levels between the first and the second hour after injection. Another central nervous system (CNS) stimulating drug, amphetamine (1 mg/kg; n = 5), under identical conditions failed to decrease seizure incidence in this mutant.

Absence of modification in GABA and benzodiazepine binding and in choline acetyltransferase activity in brain areas of the epileptic mutant mouse tottering

1. In the tottering mutant mouse, which suffers from epilepsy and cerebellar ataxia, we examined whether possible changes in GABA, benzodiazepine receptors and choline acetyltransferase (ChAT) activity are implicated in the pathophysiology of these animals. 2. No alteration in GABAA and GABAB binding could be detected in cerebellar membranes of epileptic mice as compared to normal mice. 3. Benzodiazepine receptor density and affinity showed no statistical difference in cerebellar membranes of epileptic and normal mice. 4. The activity of ChAT determined in the cortices of epileptic and normal mice did not differ significantly between the two groups.

Neuronal sensitivity to GABA and glutamate in generalized epilepsy with spike and wave discharges

In awake but painlessly immobilized cats the extracellular activity of the same cortical neurons was recorded before and for 2 to 5 h after the injection of penicillin G (350,000 IU/kg, i.m.) during the development of generalized epilepsy with bilaterally synchronous spike and wave discharges. Possible changes in their sensitivity to microiontophoretically applied glutamate and GABA during this period were searched for using computer-generated periejection histograms at intervals of about 30 min. In contrast to reported studies in other models of epilepsy, glutamate excited and GABA depressed virtually all neurons tested during fully developed spike and wave epilepsy. Spike height was not apparently affected either by the amino acids or by the development of epilepsy. Comparison of relative thresholds for the above effects on rhythmical neuronal activity associated with spike and wave discharge versus effects on random neuronal activity during the interburst periods, supported the idea that spikes and waves result from strong excitatory and inhibitory synaptic drives of the neurons. In all neurons until the appearance of spike and wave discharges, changes in the effect of amino acids, if observed, were small and statistically nonsignificant. This suggests that the hyperexcitability of cortical neurons which reportedly leads to the appearance of spike and wave discharges depends on mechanisms other than an increase in sensitivity to glutamate or a desensitization to GABA. Sometimes the sensitivity to GABA decreased later in this experimental model when the very frequent appearance of spike and wave discharges eventually led to EEG tonic-clonic seizures.

Intracortical inhibitory mechanisms are preserved in feline generalized penicillin epilepsy

Intracortical inhibition elicited by direct cortical stimulation or by stimulation of the cerebral peduncle, the latter inducing recurrent inhibition of cortical neurons, is not significantly affected by intramuscular injection of penicillin sufficient for inducing the syndrome of feline generalized penicillin epilepsy characterized by generalized spike and wave (SW) discharges in the EEG. This raises to four the number of paradigms of presumably postsynaptic inhibition resistant to penicillin concentrations sufficient to produce generalized SW discharges, a form of epileptic discharge which thus cannot be attributed to blockage of the forms of intracortical postsynaptic inhibition so far tested.

The mechanisms underlying potentiation of recruiting responses may include GABA-disinhibition

Potentiation of recruiting responses has been demonstrated to occur in the course of development of generalized epilepsy in the cat (Kostopoulos and Avoli 1983). A similar potentiation of RR of cortical neurons was demonstrated here after microiontophoretic application of bicuculline. The release of cortical neurons from a GABA-mediated inhibition or another mechanism more sensitive to bicuculline may underly both these phenomena.

Computer assisted analysis of relations between single-unit activity and spontaneous EEG

Two mutually complementary computer methods are described which can be used for the study of unit-EEG relationships during spontaneous EEG waves. The first one consists of using the unit activity to trigger the averaging of sections of EEG preceding and following each unit; the same unit activity is used for building a histogram of unit firing from another cell. Sections of data subjected to this analysis need not be continuous; they may be chosen interactively on the computer terminal, thus allowing to analyze intermittent phenomena. The second method consists of using a particular point of an EEG wave to trigger EEG averages from other channels as well as unit histograms. Here again the waves are chosen interactively. The unit-triggered EEG averages are more objective and less time consuming. However, they do not describe accurately the characteristics of the individual wave to which a unit firing is associated and also they give no information about inhibitory phenomena. Both these drawbacks are corrected by the wave-triggered unit histograms where the experimenter interactively selects and stores for analysis EEG waves with the appropriate characteristics. Several examples are given from the utilization of these programs in neurophysiological and neuropharmacological experiments, with special emphasis on generalized epilepsy.

Enhanced response of cortical neurons to thalamic stimuli precedes the appearance of spike and wave discharges in feline generalized penicillin epilepsy

Peristimulus time histograms of extracellularly recorded action potential discharges of cortical neurons in response to single shock and/or repetitive stimulation of 'specific' and 'non-specific' nuclei of the thalamus were studied after i.m. penicillin injection during a period corresponding to that of the development of spike and wave (SW) discharges of feline generalized penicillin epilepsy (FGPE). After i.m. penicillin cortical neurons displayed an enhancement of both the excitatory and 'inhibitory' phases of their responses to single shock stimulation of n. centralis medialis (NCM). This increase was even more pronounced for responses induced by repetitive stimulation of NCM at the frequencies inducing typical recruiting responses. These changes always preceded the appearance of SW discharges. Changes of the responses of cortical neurons to single shock and repetitive stimulation of 'specific' thalamic nuclei after penicillin were weak and inconsistent, although when observed were characterized by an enhancement of both excitatory and 'inhibitory' phases. The latter appeared not to decrease after i.m. penicillin. These data suggest that the appearance of SW discharges of FGPE is closely related to an increased responsiveness of cortical neurons to thalamocortical volleys arising from the so-called 'non-specific' nuclei. This facilitation of the recruiting process is accompanied by an increase of both excitatory and 'inhibitory' phases of the cortical neuronal responses induced by the volleys.

An analysis of penicillin-induced generalized spike and wave discharges using simultaneous recordings of cortical and thalamic single neurons

To study the relationship between cortical and thalamic single-neuron activity during spike and wave (SW) discharge of feline generalized penicillin epilepsy (FGPE), extracellular single-unit and local electroencephalogram (EEG) activity were recorded simultaneously from pairs of neurons, one located in the cortex of the middle suprasylvian gyrus (MSS), the other in the dorsal thalamic nuclei (n. lateralis posterior or pulvinar). These two areas are anatomically and functionally closely interrelated. Computer-generated EEG averages and histograms of single-unit activity triggered by either peaks of cortical or thalamic EEG transients or by cortical or thalamic action potentials (aps) showed that cortical neurons in the MSS fired at the time of the spike of the SW complex, while at the time of the wave they became silent. Two populations of thalamic neurons also fired maximally during the spike of SW discharge, but they differed in the precise timing of their firing in relation to that of the simultaneously recorded cortical neuron. The first group of thalamic neurons tended to fire 5-45 ms before the cortical neuron. Of these 28 neurons, 9 were antidromically and 2 orthodromically activated by cortical stimulation. The neurons of the second group tended to fire 0-45 ms after the cortical neuron. Cortical stimulation activated 15 of these 19 neurons orthodromically and 2 antidromically. A third and smaller population of thalamic neurons (n = 8) increased its firing probability during the wave of the SW complex and decreased it during the spike. In 74% of the pairs of neurons the cyclic alternation of excitation and "inhibition" associated with SW activity appeared in the cortex by 1-3 cycles earlier than in the thalamus. This was most common when the thalamic neuron of the pair reached its peak firing probability before the simultaneously recorded cortical neuron. In 11 pairs of neurons the same rhythmic alternation of excitation and "inhibition" of neuronal firing was seen in both the cortex and thalamus during SW discharges evoked by single-shock stimulation of nucleus centralis medialis. These data demonstrate that both cortical and thalamic neurons participate in the SW firing pattern of FGPE by undergoing periods of mutually phase-locked cyclic alternations of excitation and "inhibition" at the frequency of the EEG SW rhythm. Although the initial steps leading to generalized SW discharge in FGPE take place in the cortex, the thalamus soon becomes entrained in the SW rhythm.(ABSTRACT TRUNCATED AT 400 WORDS)

Participation of cortical recurrent inhibition in the genesis of spike and wave discharges in feline generalized penicillin epilepsy

Cortical recurrent inhibition (RI) evoked in pericruciate cortex by antidromic stimulation of the cerebral peduncle (CP) was studied in normal cats and in cats exhibiting the signs of feline generalized penicillin epilepsy (FGPE) following the i.m. injection of penicillin. Two measures of RI evoked by antidromic CP stimulation were used: (i) the averaged focal potential in the pericruciate gyrus; and (ii) the duration of the suppression or diminution of extracellularly recorded action potential (ap) discharge of antidromically activated pericruciate neurons measured in peristimulus time histograms (PSTHs). After i.m. injection of 350,000 IU/kg of penicillin RI remained preserved as long as only generalized spike and wave (SW) discharges appeared in the EEG, although in 5/17 neurons a modest to moderate reduction in the duration of RI occurred once SW discharges had appeared in the EEG. This inconstant reduction was probably not caused by a direct anti-inhibitory action of penicillin, but is a consequence of the increased number of ap discharges curtailing RI. At the small concentrations of penicillin existing in brain in FGPE its anti-inhibitory action evident with larger concentrations cannot be demonstrated. When focal or generalized tonic-clonic (T-C) seizures occurred, RI was reduced in slightly more than half of the instances for a few minutes before the onset of these seizures. This suggests that the transition from SW discharge to T-C seizure may be caused by a breakdown of RI.

Participation of corticothalamic cells in penicillin-induced generalized spike and wave discharges

Single unit extracellular recordings were performed in the cortex of awake painlessly immobilized unanesthetized cats during generalized spike and wave discharges (SW) induced by i.m. penicillin. Corticothalamic cells were identified in cortical areas 3a and 4 gamma by stimulating n. ventralis lateralis (VL) and in cortical areas 5 and 7 by stimulating n. lateralis posterior (LP). Twelve of 24 neurons antidromically invaded from VL were also pyramidal tract cells. Two of 11 neurons antidromically invaded from LP also displayed orthodromic responses. Corticothalamic cells fired bursts of action potentials in association with the 'spike' whereas a period of inhibition was associated with the 'wave' of the SW complex. The data suggest that in this experimental model the appearance of SW in the thalamus is due to secondary activation of thalamic neurons by volleys arising from the cortex and mediated through corticothalamic connections.

Potentiation and modification of recruiting responses precedes the appearance of spike and wave discharges in feline generalized penicillin epilepsy

Recruiting responses (RR) were evoked by stimulation of nucleus centralis medialis in awake and painlessly immobilized cats. Following the administration of sodium penicillin G (350,000 IU/kg i.m.) and at a time preceding the development of generalized spike-and-wave discharge we observed a strong potentiation of RR in 10 out of 15 experiments (50-200% increase in amplitude). The major feature of wave form modification consisted mainly of a development or increase of positive phases of individual recruiting waves. In between such large amplitude negative-positive recruiting waves a slow negative wave developed. One of every two recruiting waves was often diminished when the preceding recruiting wave had reached considerable amplitude. The changes in the RR were antagonized by barbiturates and by caffeine. In conjunction with previous evidence these results support the hypothesis that spikes of spike-and-wave discharges in FGPE are generated by similar thalamocortical volleys as those creating RR and spindles. They further suggest that the crucial neuronal mechanism underlying this effect of penicillin is a shift in the emphasis from distal apical dendritic thalamocortical synapses on cortical pyramidal neurons to more proximal ones.