Glutamate receptor involvement in dentate granule cell epileptiform activity evoked by mossy fiber stimulation

In many persons with temporal lobe epilepsy, dentate granule cells form an interconnected synaptic network. This recurrent mossy fiber circuit mediates reverberating excitation that may facilitate seizure propagation by synchronizing granule cell discharge. The involvement of specific glutamate receptors in granule cell epileptiform activity evoked by stimulating the mossy fibers was investigated with use of rat hippocampal slices superfused with bicuculline, with or without increasing [K+](o) to 6 mM. The occurrence of short-latency mossy fiber-evoked granule cell epileptiform activity in slices from pilocarpine-treated rats correlated with the presence and extent of recurrent mossy fiber growth. Blockade of AMPA receptors nearly abolished the orthodromic component of the response; subsequent antagonism of kainate receptors as well appeared to have no further action. Antagonism of NMDA receptors reduced the duration of epileptiform discharge, but increased the amplitude of population spikes within the evoked burst. Thus AMPA and NMDA, but perhaps not kainate, receptors play an important role in this type of epileptiform activity. Activation of type II metabotropic glutamate receptors, which inhibits the release of glutamate from mossy fiber boutons, reduced the magnitude of epileptiform discharge. This action was reversed by a partial agonist of these receptors. However, neither an agonist nor an antagonist of type III metabotropic glutamate receptors significantly altered the response. Considering the importance of synchronous granule cell discharge for seizure propagation from the entorhinal cortex to the hippocampus, agonists of type II metabotropic glutamate receptors may be useful in suppressing such discharge both experimentally and clinically.

Ultrastructural features and synaptic connections of hilar ectopic granule cells in the rat dentate gyrus are different from those of granule cells in the granule cell layer

Several investigators have shown the existence of dentate granule cells in ectopic locations within the hilus and molecular layer using both Golgi and retrograde tracing studies but the ultrastructural features and synaptic connections of ectopic granule cells were not previously examined. In the present study, the biocytin retrograde tracing technique was used to label ectopic granule cells following injections into stratum lucidum of CA3b of hippocampal slices obtained from epileptic rats. Electron microscopy was used to study hilar ectopic granule cells that were located 20-40 microm from the granule cell layer (GCL). They had ultrastructural features similar to those of granule cells in the GCL but showed differences, including nuclei that often displayed infoldings and thicker apical dendrites. At their origin, these dendrites were 6 microm in diameter and they tapered down to 2 microm at the border with the GCL. Both biocytin-labeled and unlabeled axon terminals formed exclusively asymmetric synapses with the somata and proximal dendrites of hilar ectopic granule cells. The mean number of axosomatic synapses for these cells was three times that for granule cells in the GCL. Together, these data indicate that hilar ectopic granule cells are postsynaptic to mossy fibers and have less inhibitory input on their somata and proximal dendrites than granule cells in the GCL. This finding is consistent with recent physiological results showing that hilar ectopic granule cells from epileptic rats are more hyperexcitable than granule cells in the GCL.

Status epilepticus-induced hilar basal dendrites on rodent granule cells contribute to recurrent excitatory circuitry

Mossy fiber sprouting into the inner molecular layer of the dentate gyrus is an important neuroplastic change found in animal models of temporal lobe epilepsy and in humans with this type of epilepsy. Recently, we reported in the perforant path stimulation model another neuroplastic change for dentate granule cells following seizures: hilar basal dendrites (HBDs). The present study determined whether status epilepticus-induced HBDs on dentate granule cells occur in the pilocarpine model of temporal lobe epilepsy and whether these dendrites are targeted by mossy fibers. Retrograde transport of biocytin following its ejection into stratum lucidum of CA3 was used to label granule cells for both light and electron microscopy. Granule cells with a heterogeneous morphology, including recurrent basal dendrites, and locations outside the granule cell layer were observed in control preparations. Preparations from both pilocarpine and kainate models of temporal lobe epilepsy also showed granule cells with HBDs. These dendrites branched and extended into the hilus of the dentate gyrus and were shown to be present on 5% of the granule cells in pilocarpine-treated rats with status epilepticus, whereas control rats had virtually none. Electron microscopy was used to determine whether HBDs were postsynaptic to axon terminals in the hilus, a site where mossy fiber collaterals are prevalent. Labeled granule cell axon terminals were found to form asymmetric synapses with labeled HBDs. Also, unlabeled, large mossy fiber boutons were presynaptic to HBDs of granule cells. These results indicate that HBDs are present in the pilocarpine model of temporal lobe epilepsy, confirm the presence of HBDs in the kainate model, and show that HBDs are postsynaptic to mossy fibers. These new mossy fiber synapses with HBDs may contribute to additional recurrent excitatory circuitry for granule cells.

Modest increase in extracellular potassium unmasks effect of recurrent mossy fiber growth

The recurrent mossy fiber pathway of the dentate gyrus expands dramatically in many persons with temporal lobe epilepsy. The new connections among granule cells provide a novel mechanism of synchronization that could enhance the participation of these cells in seizures. Despite the presence of robust recurrent mossy fiber growth, orthodromic or antidromic activation of granule cells usually does not evoke repetitive discharge. This study tested the ability of modestly elevated [K(+)](o), reduced GABA(A) receptor-mediated inhibition and frequency facilitation to unmask the effect of recurrent excitation. Transverse slices of the caudal hippocampal formation were prepared from pilocarpine-treated rats that either had or had not developed status epilepticus with subsequent recurrent mossy fiber growth. During superfusion with standard medium (3.5 mM K(+)), antidromic stimulation of the mossy fibers evoked epileptiform activity in 14% of slices with recurrent mossy fiber growth. This value increased to approximately 50% when [K(+)](o) was raised to either 4.75 or 6 mM. Addition of bicuculline (3 or 30 microM) to the superfusion medium did not enhance the probability of evoking epileptiform activity but did increase the magnitude of epileptiform discharge if such activity was already present. (2S,2'R,3'R)-2-(2',3'-dicarboxycyclopropyl)glycine (1 microM), which selectively activates type II metabotropic glutamate receptors present on mossy fiber terminals, strongly depressed epileptiform responses. This result implies a critical role for the recurrent mossy fiber pathway. No enhancement of the epileptiform discharge occurred during repetitive antidromic stimulation at frequencies of 0.2, 1, or 10 Hz. In fact, antidromically evoked epileptiform activity became progressively attenuated during a 10-Hz train. Antidromic stimulation of the mossy fibers never evoked epileptiform activity in slices from control rats under any condition tested. These results indicate that even modest changes in [K(+)](o) dramatically affect granule cell epileptiform activity supported by the recurrent mossy fiber pathway. A small increase in [K(+)](o) reduces the amount of recurrent mossy fiber growth required to synchronize granule cell discharge. Block of GABA(A) receptor-mediated inhibition is less efficacious and frequency facilitation may not be a significant factor.

GABAB-Receptor-mediated currents in interneurons of the dentate-hilus border

GABA(B)-receptor-mediated inhibition was investigated in anatomically identified inhibitory interneurons located at the border between the dentate gyrus granule cell layer and hilus. Biocytin staining was used to visualize the morphology of recorded cells. A molecular layer stimulus evoked a pharmacologically isolated slow inhibitory postsynaptic current (IPSC), recorded with whole cell patch-clamp techniques, in 55 of 63 interneurons. Application of the GABA(B) receptor antagonists, CGP 35348 (400 microM) or CGP 55845 (1 microM) to a subset of 25 interneurons suppressed the slow IPSC by an amount ranging from 10 to 100%. In 56% of these cells, the slow IPSC was entirely GABA(B)-receptor-mediated. However, in the remaining interneurons, a component of the slow IPSC was resistant to GABA(B) antagonists. Subtraction of this antagonist resistant current from the slow IPSC isolated the GABA(B) component (IPSC(B)). This IPSC(B) had a similar onset and peak latency to that recorded from granule cells but a significantly shorter duration. The GABA(B) agonist, baclofen (10 microM), produced a CGP 55845-sensitive outward current in 19 of 27 interneurons. In the eight cells that lacked a baclofen current, strong or repetitive ML stimulation also failed to evoke an IPSC(B), indicating that these cells lacked functional GABA(B) receptor-activated potassium currents. In cells that expressed a baclofen current, the amplitude of this current was approximately 50% smaller in interneurons with axons that projected into the granule cell dendritic layer (22.2 +/- 5.3 pA; mean +/- SE) than in interneurons with axons that projected into or near the granule cell body layer (46.1 +/- 10.0 pA). Similarly, the IPSC(B) amplitude was smaller in interneurons projecting to dendritic (9.4 +/- 2.7 pA) than perisomatic regions (34.3 +/- 5.1 pA). These findings suggest that GABA(B) inhibition more strongly regulates interneurons with axons that project into perisomatic than dendritic regions. To determine the functional role of GABA(B) inhibition, we examined the effect of IPSP(B) on action potential firing and synaptic excitation of these interneurons. IPSP(B) and IPSP(A) both suppressed depolarization-induced neuronal firing. However, unlike IPSP(A), suppression of firing by IPSP(B) could be easily overcome with strong depolarization. IPSP(B) markedly suppressed N-methyl-D-aspartate but not AMPA EPSPs, suggesting that GABA(B) inhibition may play a role in regulating slow synaptic excitation of these interneurons. Heterogeneous expression of GABA(B) currents in hilar border interneurons therefore may provide a mechanism for the differential regulation of excitation of these cells and thereby exert an important role in shaping neuronal activity in the dentate gyrus.

Recurrent mossy fiber pathway in rat dentate gyrus: synaptic currents evoked in presence and absence of seizure-induced growth

A common feature of temporal lobe epilepsy and of animal models of epilepsy is the growth of hippocampal mossy fibers into the dentate molecular layer, where at least some of them innervate granule cells. Because the mossy fibers are axons of granule cells, the recurrent mossy fiber pathway provides monosynaptic excitatory feedback to these neurons that could facilitate seizure discharge. We used the pilocarpine model of temporal lobe epilepsy to study the synaptic responses evoked by activating this pathway. Whole cell patch-clamp recording demonstrated that antidromic stimulation of the mossy fibers evoked an excitatory postsynaptic current (EPSC) in approximately 74% of granule cells from rats that had survived >10 wk after pilocarpine-induced status epilepticus. Recurrent mossy fiber growth was demonstrated with the Timm stain in all instances. In contrast, antidromic stimulation of the mossy fibers evoked an EPSC in only 5% of granule cells studied 4-6 days after status epilepticus, before recurrent mossy fiber growth became detectable. Notably, antidromic mossy fiber stimulation also evoked an EPSC in many granule cells from control rats. Clusters of mossy fiber-like Timm staining normally were present in the inner third of the dentate molecular layer at the level of the hippocampal formation from which slices were prepared, and several considerations suggested that the recorded EPSCs depended mainly on activation of recurrent mossy fibers rather than associational fibers. In both status epilepticus and control groups, the antidromically evoked EPSC was glutamatergic and involved the activation of both AMPA/kainate and N-methyl-D-aspartate (NMDA) receptors. EPSCs recorded in granule cells from rats with recurrent mossy fiber growth differed in three respects from those recorded in control granule cells: they were much more frequently evoked, a number of them were unusually large, and the NMDA component of the response was generally much more prominent. In contrast to the antidromically evoked EPSC, the EPSC evoked by stimulation of the perforant path appeared to be unaffected by a prior episode of status epilepticus. These results support the hypothesis that recurrent mossy fiber growth and synapse formation increases the excitatory drive to dentate granule cells and thus facilitates repetitive synchronous discharge. Activation of NMDA receptors in the recurrent pathway may contribute to seizure propagation under depolarizing conditions. Mossy fiber-granule cell synapses also are present in normal rats, where they may contribute to repetitive granule cell discharge in regions of the dentate gyrus where their numbers are significant.

Interneurons of the dentate-hilus border of the rat dentate gyrus: morphological and electrophysiological heterogeneity

Interneurons located near the border of the dentate granule cell layer and the hilus were studied in hippocampal slices using whole-cell current clamp and biocytin staining. Because these interneurons exhibit both morphological and electrophysiological diversity, we asked whether passive electrotonic parameters or repetitive firing behavior correlated with axonal distribution. Each interneuron was distinguished by a preferred axonal distribution in the molecular layer or granule cell layer, and four groups could be discerned, the axons of which arborized in (1) the granule cell layer, (2) the inner molecular layer, (3) the outer molecular layer, and (4) diffusely in the molecular layer. In our sample, interneurons with axons arborizing diffusely in the molecular layer were most frequent, and those with axons restricted to the granule cell layer were least frequent. Resting potential, input resistance, time constant, electrotonic length, and spike frequency adaptation (SFA) were not significantly different among the four groups, and the variability in SFA between cells with similar axonal distributions was striking. Clear differences in action potential morphology and afterhyperpolarizations, however, emerged when nonadapting interneurons were compared with those exhibiting SFA. Interneurons exhibiting SFA had characteristically broader spikes, progressive slowing of action potential repolarization during repetitive firing, and slow afterhyperpolarizations that distinguished them from nonadapting interneurons. We propose that the variability in repetitive firing behavior and morphology exhibited by each of these interneurons makes each interneuron unique and may provide a high level of fine tuning of inhibitory control critical to information processing in the dentate.

Hippocampal mossy fiber sprouting and synapse formation after status epilepticus in rats: visualization after retrograde transport of biocytin

In complex partial epilepsy and in animal models of epilepsy, hippocampal mossy fibers appear to develop recurrent collaterals that invade the dentate molecular layer. Mossy fiber collaterals have been proposed to subserve recurrent excitation by forming granule cell-granule cell synapses. This hypothesis was tested by visualizing dentate granule cells and their mossy fibers after terminal uptake and retrograde transport of biocytin. Labeling studies were performed with transverse slices of the caudal rat hippocampal formation prepared 2.6-70.0 weeks after pilocarpine-induced or kainic acid-induced status epilepticus. Light microscopy demonstrated the progressive growth of recurrent mossy fibers into the molecular layer; the densest innervation was observed in slices from pilocarpine-treated rats that had survived 10 weeks or longer after status epilepticus. Thin mossy fiber collaterals originated predominantly from deep within the hilar region, crossed the granule cell body layer, and formed an axonal plexus oriented parallel to the cell body layer within the inner one-third of the molecular layer. When sprouting was most robust, some recurrent mossy fibers at the apex of the dentate gyrus reached the outer two-thirds of the molecular layer. The distribution and density of mossy fiber-like Timm staining correlated with the biocytin labeling. When viewed with the electron microscope, the inner one-third of the dentate molecular layer contained numerous mossy fiber boutons. In some instances, biocytin-labeled mossy fiber boutons were engaged in synaptic contact with biocytin-labeled granule cell dendrites. Granule cell dendrites did not develop large complex spines ("thorny excrescences") at the site of synapse formation, and they did not appear to have been permanently damaged by seizure activity. These results establish the validity of Timm staining as a marker for mossy fiber sprouting and support the view that status epilepticus provokes the formation of a novel recurrent excitatory circuit in the dentate gyrus. Retrograde labeling with biocytin showed that the recurrent mossy fiber projection often occupies a considerably greater fraction of the dendritic region than previous studies had suggested.

Increased AMPA-sensitive quisqualate receptor binding and reduced NMDA receptor binding in epileptic human hippocampus

Based on results from the kindling model of epilepsy, we hypothesized that enhanced binding of radioligands to the NMDA receptor and decreased binding to the alpha-amino-3-hydroxy-5-methyl-4- isoxazolepropionate (AMPA)-sensitive quisqualate (ASQ) receptor would be found within epileptic hippocampi of humans with complex partial epilepsy (CPE). To test these hypotheses, we used tissue that was surgically removed from patients with intractable CPE, and control tissue that was obtained at autopsy. We used autoradiographic techniques to measure ASQ receptor binding (with 3H-AMPA as the radioligand) and binding to 2 sites on the NMDA receptor/channel complex: the agonist recognition site (with 3H-glutamate) and the phencyclidine (PCP) binding site that resides within the NMDA channel [with 3H-N-(1-[thienyl]cyclohexyl) piperidine (TCP) in the presence of saturating concentrations of NMDA and glycine]. Measurements of receptor binding were corrected for pathologic alterations in neuronal density. Contrary to our expectations, ASQ receptor binding was significantly increased (100%; p less than 0.02) in the dentate gyrus stratum moleculare in patients with CPE (n = 8), and it was unchanged in other hippocampal regions. In nearby sections from the same specimens, binding was significantly decreased to the agonist recognition site of the NMDA receptor in the stratum oriens of area CA3 (46%; p less than 0.05) and was also decreased to the PCP site in the stratum radiatum and stratum oriens of CA3 (44% and 74%, respectively; p less than 0.05). The increase in ASQ receptor binding may contribute to hyperexcitability in these epileptic patients.(ABSTRACT TRUNCATED AT 250 WORDS)

Kainate and quisqualate receptor autoradiography in rat brain after angular bundle kindling

The kainate and quisqualate types of excitatory amino acid receptor were visualized autoradiographically in brain sections from rats kindled by stimulating the angular bundle. Kainate receptors were labeled with [3H]kainate and quisqualate receptors with L-[3H]glutamate. When assayed one day after the last evoked seizure, kainate receptor binding had declined by 24-29% in stratum lucidum of hippocampal area CA3 and by 12-14% in the inner third of the dentate molecular layer, but was unchanged in the neocortex and basolateral amygdala. Saturation binding curves revealed that, under the conditions of these experiments, [3H]kainate labeled a single class of binding sites with a KD of 33-36 nM. In stratum lucidum of area CA3, kindling reduced the density of kainate receptors without altering their affinity for kainate. At the same time, quisqualate receptor binding had declined by 20-35% in many layers of the hippocampal formation and neocortex, but remained unchanged in the basolateral amygdala. Repeated stimulation or repeated seizures were required to produce these effects, since both kainate and quisqualate receptor binding were unchanged one day after a single afterdischarge. These receptor changes largely or completely reversed during a 28-day period without further stimulation. Thus maintenance of the kindled state probably cannot be explained by a long-lasting change in the expression of kainate or quisqualate receptors. The transient, regionally-selective down-regulation of these receptors may represent a compensatory response of forebrain neurons to repeated stimulation or seizures.

N-methyl-D-aspartate receptor autoradiography in rat brain after angular bundle kindling

The specific binding of L-[3H]glutamate to N-methyl-D-aspartate (NMDA) receptors in brain regions of kindled rats was visualized autoradiographically and quantitated. When assayed 28 days after the last evoked seizure, NMDA receptor binding had declined by 7-11% in stratum radiatum of the dorsal hippocampal area CA1, in both deep and superficial layers of the motor cortex and in layers I-IV of the somatosensory cortex. No significant changes were detected in any other brain region nor in any region examined 1 day after the last evoked seizure. These findings suggest that the enhanced activation of NMDA receptors in kindled rats cannot be explained by an increased expression of these receptors. Rather, kindling leads to a regionally-selective down regulation of NMDA receptor binding.

Protective effects of mossy fiber lesions against kainic acid-induced seizures and neuronal degeneration

The effects of a hippocampal mossy fiber lesion have been determined on neuronal degeneration and limbic seizures provoked by the subsequent intracerebroventricular administration of kainic acid to unanesthetized rats. Mossy fiber lesions were made either by transecting this pathway unilaterally or by destroying the dentate granule cells unilaterally or bilaterally with colchicine. All control rats eventually developed status epilepticus and each temporally discrete seizure that preceded status epilepticus was recorded from the hippocampus ipsilateral to the kainic acid infusion before the contralateral hippocampus. A mossy fiber lesion of the ipsilateral hippocampus prevented the development of status epilepticus in 26% of subjects and in 52% of subjects seizures were recorded from the contralateral hippocampus before the ipsilateral hippocampus. Unlike electrographic records from other treatment groups, those from rats which had received a bilateral colchicine lesion exhibited no consistent pattern indicative of seizure propagation from one limbic region to another. A bilateral, but not a unilateral, mossy fiber lesion also dramatically attenuated the behavioral expression of the seizures. Regardless of its effects on kainic acid-induced electrographic and behavioral seizures, a mossy fiber lesion always substantially reduced or completely prevented the degeneration of ipsilateral hippocampal CA3-CA4 neurons. This protective effect was specific for those hippocampal neurons deprived of mossy fiber innervation. Neurons in other regions of the brain were protected from degeneration only when the mossy fiber lesion also prevented the development of electrographic status epilepticus. These results suggest that the hippocampal mossy fibers constitute an important, though probably not an obligatory, link in the circuit responsible for the spread of kainic acid seizures. Degeneration of CA3-CA4 neurons appears to depend upon (1) the duration of hippocampal seizure activity and (2) an as yet undefined influence of or interaction with the mossy fiber projection which enhances the neurodegenerative effect of the seizures.

Unchanged regional norepinephrine glycol metabolite levels in rat brain two months after amygdala kindling

Concentrations of the norepinephrine (NE) glycol metabolites MHPG (3-methoxy-4-hydroxyphenylethyleneglycol) and DHPG (3,4-dihydroxyphenylethyleneglycol) were examined in 13 brain regions of amygdala-kindled and yoked control rats. The subjects were killed 2 months after the kindled rats had exhibited their sixth 'stage 4-5' generalized seizure. In experiment 1, small but statistically significant decreases in total MHPG levels were found in the right hippocampus (91.6% of control) and hypothalamus (90.7% of control). When the study was repeated with 2 additional control groups, non-kindled, electrically stimulated controls and maximal electroshock convulsion controls, however, these small changes were not seen. The data suggest that since amygdala kindling does not produce any consistent, long-lasting alterations in brain regional NE glycol metabolite levels, there are no consistent, long-term changes in central NE neuronal activity.

Mossy fiber lesion reduces the probability that kainic acid will provoke CA3 hippocampal pyramidal cell bursting

Hippocampal slices prepared from rats which had received a mossy fiber lesion differed in their response to 50 nM kainic acid. Those slices in which the mossy fiber projection had been substantially destroyed were significantly less likely to develop epileptiform bursting in area CA3 than slices in which the mossy fiber projection was only modestly damaged. Similarly, mossy fiber lesions prevent the development of electrographic status epilepticus after intracerebroventricular administration of kainic acid in 26% of rats. Therefore mossy fiber lesions probably act, both in vivo and in vitro, by reducing the sensitivity of CA3 hippocampal pyramidal cells to the epileptogenic action of kainic acid.

Unchanged norepinephrine turnover and concentrations in amygdala-kindled rat brain regions 2 months postseizure

Norepinephrine turnover and concentrations were examined in eight brain regions of amygdala-kindled and yoked control rats. The subjects were killed 2 months after the kindled subjects had exhibited their sixth "stage 4-5" generalized seizure. No significant differences were found in norepinephrine turnover or concentrations between kindled and yoked control rats in any of the brain regions examined. Dopamine turnover and concentrations in brain regions were also unchanged in kindled rats compared with their yoked controls. The data suggest that amygdala kindling does not produce any long-lasting alterations in brain regional catecholamine turnover or concentrations.

Binding of [3H]flunitrazepam and [3H]Ro5-4864 to crude homogenates of amygdala-kindled rat brain: two months post-seizure

The binding of [3H]flunitrazepam and [3H]Ro5-4864 to crude homogenates of amygdala-kindled and 'yoked' control rat brains was evaluated in subjects sacrificed 60 days after the sixth generalized convulsion. Decreases in the Bmax for [3H]flunitrazepam binding to 'central-type' benzodiazepine receptors were observed in the hypothalamus and ipsilateral cortex of kindled animals. The binding of [3H]Ro5-4864 to 'peripheral-type' benzodiazepine receptors was unchanged.

Enhancement of 3H-phenytoin binding by diazepam and (+)bicuculline

3H-Phenytoin binding to the particulate fraction of rat wholebrain homogenate was studied using the filter assay technique. It was found that diazepam and (+)bicuculline methobromide caused a concentration-dependent enhancement of the total binding of 3H-phenytoin, whereas GABA and (-)bicuculline methobromide (the inactive bicuculline isomer) had no effect. In subsequent competition experiments (labelled versus unlabelled phenytoin), it was found that the presence of a potentiating concentration of diazepam transformed the biphasic phenytoin competition isotherm into a simple curve with a Hill coefficient of approximately one, and a Ki of 110 nM.

Saturable binding of [3H]phenytoin to rat brain membrane fraction

Preliminary studies indicate that [3H]phenytoin binds in a saturable and reversible fashion to at least two distinct sites in the membrane fraction of whole rat brain. One of these displays a high affinity (Kd = 6 nM) and a low maximal capacity (Bmax = 10 pmol/g protein). The other has a low affinity (Kd = 4.8 microM) and is estimated to have a very high maximal capacity. Phenytoin binding is reduced if the membrane fraction is preincubated with proteolytic enzymes and subcellular fractionation studies indicate that P2 fraction has the largest number of binding sites. Competition experiments fail to reveal significant binding interactions with putative neurotransmitters or with other drugs except the hydantoins and anticonvulsant barbiturates. Although it is premature to speculate on the clinical significance of these findings, it is encouraging to note that the low affinity site has a Kd very similar to the therapeutic levels of phenytoin found in cerebrospinal fluid and that there seems to be some relationship between binding potency and anticonvulsant potency within the hydantoin series.