Brain, Vol. 122, No. 10, 1807-1822,
October 1999
© 1999 Oxford University Press
Review Article |
Chandelier cells and epilepsy
Instituto Cajal (CSIC), Madrid, Spain
Correspondence to:
Dr Javier DeFelipe, Instituto Cajal (CSIC), Avenida Dr Arce, 37, 28002-Madrid, Spain E-mail: defelipe{at}cajal.csic.es
 |
Abstract |
|---|
 Top Abstract IntroductionTypes of neocortical neuronsSurgical treatment of temporal...A chandelier cell hypothesis:...References |
|---|
The main goal of this article is to review certain aspects of the circuitry of the human cerebral cortex that may be particularly relevant for the development, maintenance or spread of seizures. There are a number of different structural abnormalities that are commonly found in the cortex of epileptic patients, but these abnormalities do not appear to be intrinsically epileptogenic, since some patients displaying them are epileptic (after variable delays) whereas others are not. Therefore, cortical circuits in an affected brain may undergo a series of changes that finally cause epilepsy. In this article, it is proposed that the chandelier cell, which is considered to be the most powerful cortical GABAergic inhibitory interneuron, is probably a key component of cortical circuits in the establishment of human intractable temporal lobe epilepsy. These cells (among other types) have been found to be lost or reduced at epileptic foci in both experimental animals and epileptic patients. A hypothesis is presented by which the normal variability in the number of interneurons might explain the predisposition of some individuals to develop epilepsy more than others as a result of a lesion or other precipitating factors that lead to loss of neurons. The sources of GABAergic input on dendrites and somata of cortical pyramidal cells originate from many and diverse types of interneurons but, at the level of the axon initial segment of these cells, all synapses come from a few chandelier cells (five or less). Loss of one class of interneurons ending on soma and dendrites might have relatively little impact on the inhibitory control of the pyramidal cell. However, if chandelier cells were affected, it would have serious consequences for the inhibitory control of the pyramidal cells. Evidence suggests that the loss of chandelier cells may be non-specific and that when this occurs epilepsy may develop. Therefore, these cells might represent a key component in the aetiology of human temporal lobe epilepsy.
human; cortical circuitry; interneurons; parvalbumin; neuronal loss
Ch terminal = terminal portion of the axon of chandelier cells
|
Introduction |
|---|
|
TopAbstractIntroductionTypes of neocortical neuronsSurgical treatment of temporal...A chandelier cell hypothesis:...References |
|---|
Epilepsy frequently is associated with a variety of brain lesions, tumours or other structural abnormalities which mostly affect the cerebral cortex. However, these alterations are not intrinsically epileptogenic, since some patients are epileptic, whereas others with similar abnormalities are not, or they develop epilepsy after a variable delay. One of the best examples is that of brain tumours, which may or may not cause epilepsy, even when they are of the same type and located in the same brain area. Furthermore, seizures associated with brain tumours or head injuries may appear relatively soon or after a long `latent' period that may extend for several years (Ketz, 1974;
Spencer et al., 1984; Morrell, 1989; Dinner, 1993; Morris and Estes, 1993). Thus, it seems that cortical tissue in the altered brain suffers a series of changes that eventually lead to epilepsy.
Temporal lobe epilepsy is one of the most frequent types of human epilepsy, and a variety of surgical procedures have been developed for the treatment of intractable temporal lobe epilepsy (e.g. Falconer, 1971a; Wieser, 1996). Resected tissue can be examined successfully using a variety of sophisticated biochemical and anatomical techniques for investigating the mechanisms underlying epilepsy. However, the basic mechanism by which seizure activity is related to abnormal cortical circuits is still under debate (Ben-Ari and Represa, 1990; Sloviter, 1991; McNamara, 1994; Schwartzkroin, 1994a). Most hypotheses are based on alterations of glutamatergic (excitatory) and GABAergic (inhibitory) cortical neuronal systems in either the hippocampus, the neocortex or both (Lloyd et al., 1986; Prince and Connors, 1986; Sherwin and van Gelder, 1986; Houser, 1991; Ribak, 1991; Avanzini et al., 1992; Meldrum, 1995; DeFelipe et al., 1996). However, some researchers (including my group) maintain that, in spite of the variety of brain abnormalities associated with epilepsy, it is possible that a common basic underlying mechanism exists. In this work, I shall propose that a particular type of interneuron, the chandelier cell, is probably a key component of cortical circuits in the establishment of human epilepsy. The following presentation will be focused mainly on the human neocortex in temporal lobe epilepsy, unless otherwise specified.
|
Types of neocortical neurons |
|---|
|
TopAbstractIntroductionTypes of neocortical neuronsSurgical treatment of temporal...A chandelier cell hypothesis:...References |
|---|
The neocortex contains two major classes of neurons: spiny neurons and smooth non-pyramidal neurons. Spiny neurons are represented by pyramidal cells and spiny non-pyramidal cells. Pyramidal cells are the projection neurons, they are present in all layers except layer I, and they are the most abundant (~70–85% of the total neuron population). Spiny non-pyramidal cells are interneurons confined to the middle layers of the cortex (principally layer IV), the most typical being the spiny stellate cell (e.g. Fairén et al., 1984; Lund, 1984). Smooth non-pyramidal neurons are also interneurons, but with smooth or sparsely spiny dendrites. Pyramidal and spiny stellate cells are excitatory and are likely to use glutamate and/or aspartate, whereas the majority of smooth interneurons are likely to use GABA and, thus, are inhibitory (for review, see Gilbert, 1983; Houser et al., 1984; Ottersen and Storm-Mathisen, 1984; Streit, 1984; White, 1989; Conti, 1991; DeFelipe and Fariñas, 1992; DeFelipe, 1993; Jones, 1993; Somogyi et al., 1998). In this presentation, I will refer to smooth non-pyramidal neurons as aspiny interneurons. These interneurons are found in all cortical areas and layers and make up ~15–30% of the total neuron population. These interneurons form a morphologically heterogeneous group, and different morphological types are recognized mainly on the basis of their patterns of axonal arborization (e.g. Jones, 1975; Fairén et al., 1984).
What is a chandelier cell?
The chandelier cell is a type of aspiny interneuron with smooth dendrites and is distinguished by the terminal portions of its axon that form short vertical rows of boutons (Ch terminals) resembling candlesticks (Szentágothai and Arbib, 1974) (Fig. 1). These cells have been found not only in the neocortex, but also in several areas of the paleocortex, archicortex and mesocortex, which include the piriform cortex, entorhinal cortex, subiculum, hippocampus, fascia dentata and cingulate cortex (Szentágothai and Arbib, 1974; Jones, 1975; Fairén and Valverde, 1980; Somogyi et al., 1982, 1983b, 1985; Soriano and Frotscher, 1989; Soriano et al., 1990, 1993; Schmidt et al., 1993; Buhl et al., 1994; Gabbott and Bacon, 1996; Martínez et al., 1996; Mikkonen et al., 1997). In general, the chemical and morphological characteristics of these cells are similar in all divisions of the cerebral cortex, but in the present work I shall deal only with chandelier cells of the neocortex, unless otherwise specified.
|
These cells, or the Ch terminals, have been visualized using a variety of methods: the Golgi method (e.g. Jones, 1975; Szentágothai, 1975; Somogyi, 1977; Fairén and Valverde, 1980; Peters et al., 1982; Somogyi et al., 1982; DeFelipe et al., 1985; Kisvárday et al., 1986; Marin-Padilla, 1987; Lewis and Lund, 1990); immunocytochemical methods using antibodies directed against parvalbumin (DeFelipe et al., 1989
; Lewis and Lund, 1990; Williams et al., 1992; Condé et al., 1994; del Río and DeFelipe, 1994; Anderson et al., 1995), against calbindin (del Río and DeFelipe, 1997), against corticotropin-releasing factor (Lewis et al., 1989; Lewis and Lund, 1990) or against GABA transporter GAT-1 (DeFelipe and González-Albo, 1998; Woo et al., 1998); by intracellular injections of horseradish peroxidase (Freund et al., 1983) or Lucifer Yellow (Naegele and Katz, 1990); and by intracellular staining with biocytin (Kawaguchi, 1995; Kawaguchi and Kubota, 1998).
Studies using immunocytochemistry for GABA or its synthesizing enzyme glutamic acid decarboxylase showed that chandelier cells are GABAergic and, thus, inhibitory (Peters et al., 1982, Freund et al., 1983; Hendry et al., 1983, DeFelipe et al., 1985; Somogyi et al., 1985; Buhl et al., 1994). On the basis of immunocytochemistry, chandelier cells may be chemically defined as GABAergic cells which contain the calcium-binding proteins parvalbumin and calbindin, but not calretinin, and which contain the neuropeptide corticotropin-releasing factor, but not other neuropeptides, namely cholecystokinin, somatostatin, neuropeptide Y, vasoactive intestinal polypeptide and tachykinins (for a review, see DeFelipe and Fariñas, 1992) (Fig. 2). However, these cells are chemically heterogeneous, since, depending on the species, cortical areas and layers, chandelier cells are either immunoreactive or not for selected combinations of these substances (see Lewis and Lund, 1990; del Río and DeFelipe, 1994, 1997; and references therein). For example, using immunocytochemistry for corticotropin-releasing factor to examine the prefrontal and occipital cortex of Macaca mulatta, Macaca fascicularis and Saimiri sciureus, Lewis and Lund found that Ch terminals were labelled only in the squirrel monkey (S. sciureus), mainly in layer IV of the prefrontal cortex (Lewis and Lund, 1990). Calbindin immunocytochemistry labels a small subpopulation of Ch terminals located mainly in layers V and VI of the human neocortex, but not in other species (del Río and DeFelipe, 1997). Nevertheless, the most general, prominent and complete staining of Ch terminals is obtained with immunocytochemistry for parvalbumin (Figs 2 and 3) and GAT-1 (DeFelipe and González-Albo, 1998).
|
|
The axon terminals that make up Ch terminals have been shown to innervate only the axon initial segments of pyramidal cells on which they form symmetrical synapses (e.g. Somogyi, 1977; Fairén and Valverde, 1980; Peters et al., 1982; Somogyi et al., 1982; Freund et al., 1983; DeFelipe et al., 1985, 1993; del Río and DeFelipe, 1994). Chandelier cells are the major or sole source of synapses on pyramidal cell axon initial segments, and each single chandelier cell axon can give off several hundred Ch terminals (e.g. Freund et al., 1983; DeFelipe et al., 1985). Each Ch terminal innervates a single axon initial segment. However, a single axon initial segment may be innervated by one or a few Ch terminals (five or less) which, in turn, may originate from the same or different chandelier cells (Fairén and Valverde, 1980
; Peters et al., 1982; Freund et al., 1983; Somogyi et al., 1983c; reviewed in Somogyi et al., 1982; DeFelipe and Fariñas, 1992).
Finally, because these cells are inhibitory and ubiquitously distributed and because the axon initial segments of pyramidal cells appear to be strategically important regions for the control of pyramidal cell excitability (Douglas and Martin, 1990; Stuart and Sakmann, 1994), chandelier cells are considered to be the most powerful cortical inhibitory interneuron. Because pyramidal cells are the main source of cortical efferent axons and of intracortical collaterals, they are responsible for the spread of epileptiform activity; this makes chandelier cells particularly relevant in epilepsy, as originally proposed by Somogyi et al. and then by Ribak and by Kisvárday et al. (Somogyi et al., 1983b; Ribak, 1985; Kisvárday et al., 1986; for a review, see Ribak, 1991).
|
Surgical treatment of temporal lobe epilepsy and pathological findings |
|---|
|
TopAbstractIntroductionTypes of neocortical neuronsSurgical treatment of temporal...A chandelier cell hypothesis:...References |
|---|
Neuronal loss
Neuronal loss (and gliosis) is characteristic of epileptogenic tissue; however, a comprehensive review of the current voluminous literature about the relationship between neuronal loss and epilepsy is beyond the scope of this presentation.
In patients with symptomatic epilepsy, neuronal loss occurs irrespective of the primary cause of the lesion (e.g. underlying tumour, infection, trauma, vascular accidents) (Meldrum and Bruton, 1992). In cases not associated with brain tumours and without obvious pathological defects, standard histopathological examination of resected brain tissue reveals a range of findings that vary from severe neuronal loss to no loss at all (Meldrum and Bruton, 1992). The most typical alteration observed in patients with temporal lobe epilepsy is neuronal loss and gliosis in the hippocampus (hippocampal or Ammon's horn sclerosis) (Margerison and Corsellis, 1966; Falconer, 1974; Babb and Brown, 1987; Meldrum and Bruton, 1992; Babb and Pretorius, 1993), although the neocortex of the anterolateral superior, middle and inferior temporal gyri (to which I shall refer collectively as temporal neocortex) frequently is found to be normal (Babb et al., 1984). In these patients, MRI may be normal or may show hippocampal atrophy or abnormal signals (Cascino et al., 1991; Jackson et al., 1994). This is in line with the well-known fundamental role attributed to the hippocampus in human temporal lobe epilepsy (Schwartzkroin, 1994b), which is often considered (together with the amygdala) as `the epileptogenic zone'. However, significant interictal activity arising from the temporal neocortex and extrahippocampal mesial temporal structures frequently is observed (including cases with hippocampal sclerosis). Furthermore, due to the strong reciprocal connections between mesial temporal structures and the temporal neocortex (Van Hoesen, 1982; Suzuki and Amaral, 1994), it is often difficult to ascertain whether spiking activity recorded from a given electrode is primarily a projected or an intrinsic activity (e.g. Walczak, 1995; O'Brien et al., 1996). The experience of many surgeons indicates that, in general, a better result is obtained from a large resection compared with small and restricted removals. Thus, the most common surgical procedure for temporal lobe epilepsy is anterior temporal lobectomy, which consists of the ipsilateral removal of the anterolateral temporal cortex, the amygdala and the anterior portion of the hippocampus and adjacent cortex (Olivier, 1992).
Selective amygdalohippocampectomy (removal of the amygdala, the anterior hippocampus and part of the parahippocampal gyrus) is an alternative surgical procedure (Wieser and Yasargil, 1982) which, in principle, would be the most appropriate for surgical treatment when sclerosis of the hippocampus is the most probable lesion. Nevertheless, this surgical procedure is applied only to a relatively few patients in whom seizure onset is clearly restricted to mesial temporal structures; but even in these cases, although the amygdalohippocampectomy is highly effective in some patients, there is no guarantee of a good surgical outcome in all patients (see Wieser, 1991, 1996; Mayanagi et al., 1996; Alarcon et al., 1997). Furthermore, as pointed out by Babb and Brown (Babb and Brown, 1987), because of the occurrence of extrahippocampal pathologies in some cases with hippocampal sclerosis (dual pathology), amygdalohippocampectomy should be a less successful surgical technique than anterior temporal lobectomy, which would remove both hippocampal and extrahippocampal lesions. For example, in a series of 178 epileptic patients who underwent en bloc temporal lobectomies (reviewed by Lévesque et al., 1991), dual pathology was present in 30% of the cases (see also Margerison and Corsellis, 1966). Additionally, although there are a number of clinical and electrophysiological differences between temporal lobe epileptic patients presenting mesial temporal sclerosis and those with discrete temporal neocortical lesions (Saygi et al., 1994; O'Brien et al., 1996; Gil-Nagel and Risinger, 1997), in practice these differences are not sufficient to establish such a distinction in an individual patient (see O'Brien et al., 1996). Finally, Jones-Gotman et al. have shown that selective, minimal removals of temporal lobe structures do not preserve certain intellectual capabilities better than larger resections, since neocorticectomies, selective amygdalohippocampectomies and anterior temporal lobectomies led to similar neuropsychological deficits (regarding learning and memory for verbal and visuospatial material) (Jones-Gotman et al., 1997). Thus, many surgeons are inclined toward anterior temporal lobectomy as the surgical procedure to be used.
In a recent study made by Alarcon et al. on the relationship between patterns of discharge propagation during acute electrocorticography and surgical treatment of temporal lobe epilepsy, a very significant association between a poor outcome and the retaining of regions where discharges showed earliest peaks (`leading regions') was found (Alarcon et al., 1997). Therefore, these authors concluded that leading regions were potentially epileptogenic regions that should be removed in order to improve seizure control. Within a given patient, multiple leading regions, which were not necessarily associated with the anatomically pathological region, were found. Interestingly, in the 23 cases with mesial temporal sclerosis studied by Alarcon and colleagues, the main leading regions were localized in the hippocampus in only 43% of the patients, whereas in the remaining 57% these leading regions were localized either in the temporal neocortex or in the extrahippocampal basal temporal cortex. These results support previous studies which emphasized the contribution of the neocortex to seizure development and propagation. However, the temporal neocortex was found by these authors to be normal on the basis of standard histopathological methods, which is the most common finding in the resected temporal neocortex (see above). Therefore, it is generally thought that the temporal neocortex without any apparent structural abnormality may be epileptogenic due to unknown mechanisms and that this may occur irrespective of the primary pathology found in the resected tissue.
Loss of interneurons
Three long-continuing questions still prevail in the field of epilepsy pathology: first, how and when neuronal loss occurs; secondly, whether all types of neurons are equally affected; and thirdly, whether this loss is secondary to epileptic seizures or represents a primary process which initiates the development of seizure activity.
Standard histopathological techniques do not usually detect subtle changes in neuronal circuitry, especially if particular types of interneurons are selectively affected, because they represent a minor population of neurons whose loss may pass unnoticed. Along these lines, in certain experimental animal models of focal epilepsy, it has been shown, using immunocytochemical and electron microscope techniques, that in the epileptogenic neocortex there is a preferential loss of GABAergic neurons and axon terminals; two of the GABAergic cell types that are lost at epileptic foci are the basket cell and chandelier cell (Ribak et al., 1979, 1982; Ribak, 1985; Houser et al., 1986; for reviews, see Houser, 1991; Ribak 1991). Loss of inhibition at the epileptic foci would cause epileptic activity. Finally, Williams et al. found, in a patient showing generalized slow and paroxysmal activity recorded by EEG and diagnosed with neuronal ceroid-lipofuscinosis (Batten disease), but with no evidence of neuronal degeneration or loss, that the perikarya, axon hillock and axon initial segment of neocortical pyramidal cells were surrounded by reactive processes of astrocytes and possessed few or no inhibitory synapses at these sites, whereas presumptive excitatory synapses were abundant in the neuropil (Williams et al., 1977). These authors suggested that the loss of the inhibitory synapses might be the cause of the paroxysmal activity.
Parvalbumin immunostaining as a tool for identifying abnormal circuits
Since parvalbumin immunocytochemistry labels a subpopulation of GABAergic interneurons (which include chandelier cells and basket cells) (Celio, 1986; DeFelipe et al., 1989; Hendry et al., 1989; Lewis and Lund, 1990; Andressen et al., 1993), we used this staining to label these cells in cortical tissue removed from epileptic patients (Figs 4–6). The aim of these studies (DeFelipe et al., 1993; Marco et al., 1996, 1997) was mainly to investigate whether chandelier cells and basket cells were lost from sections of neocortex from patients with temporal lobe epilepsy, while adjacent sections showed normal appearances in standard histopathological preparations. The examination was independent of whether or not the condition was associated with brain tumours or presented neuronal cell loss in the hippocampus. In these studies, multiple small regions or patches (often 200 µm to 1 mm wide) showing variable degrees of decreased parvalbumin immunostaining commonly were observed (see also Ferrer et al., 1994; Spreafico et al., 1998), and were independent of the primary pathological findings (Figs 4–6). Since no patches of decreased parvalbumin immunostaining were observed in sections from neurologically normal human neocortical tissue (Marco et al., 1996; Marco and DeFelipe, 1997), we assume that these changes are probably related to the occurrence of seizures in the brains of these patients (see below).
|
|
|
For simplicity, I shall refer to all abnormal patterns of parvalbumin immunostaining collectively as `decreased parvalbumin immunostaining', unless otherwise specified. The decrease apparently always affected chandelier cell axon terminals and, to a greater or lesser extent, other types of parvalbumin-positive neurons, including basket cell axon terminals. For example, there were regions showing a pattern of parvalbumin immunostaining that, at first glance, was indistinguishable from the normal pattern of immunostaining, but closer inspection showed a clear decrease or a virtual lack of chandelier cell axon terminals (Fig. 6B
), whereas in other regions all types of parvalbumin-positive cells including chandelier cells appeared to be affected to various degrees (Marco et al., 1996, 1997) (Fig. 5A and B). Nevertheless, there are still always relatively numerous terminals in the neuropil and around the somata of pyramidal cells in the regions with decreased parvalbumin immunostaining, whereas the decrease in chandelier cell axon terminals appears to be complete, because vertical rows of boutons, which are typical of chandelier cell axon terminals, are not stained (Fig. 7).
|
Examination of Nissl-stained sections adjacent to those showing decreased parvalbumin immunostaining suggested that the cortical tissue was apparently normal in most cases, but no quantitative analyses were performed. However, regions showing the so-called pattern C (which consists of a virtual lack of parvalbumin-immunoreactive neurons and a reduction in parvalbumin-immunoreactive processes, affecting mainly chandelier cell axon terminals) were associated with focal cell loss (Marco et al., 1996
). This focal cell loss consisted of a moderate or severe decrease in neurons that affected (at least in the severe cases) both pyramidal cells and GABAergic interneurons. Frequently, these regions were so small that they were very difficult to identify in Nissl-stained sections. However, once a region with pattern C of parvalbumin immunostaining was detected, then a decrease in cells was readily identified in the adjacent Nissl-stained sections. This correlation was demonstrated more directly after examination of 2 µm semi-thin plastic sections from patches of decreased immunostaining in material prepared for electron microscopy (Marco and DeFelipe, 1997).
It should be mentioned that patches of decreased parvalbumin immunostaining were not found in all patients. As pointed out previously (Marco et al., 1996), only a relatively small amount of neocortex was analysed using parvalbumin immunocytochemistry in each patient and, therefore, there may be small regions with altered patterns that might not be included in the analysis. In this regard, it is important to note that after post-mortem pathological examination of the brains of epileptic patients, patchy neuronal loss was found in the frontal cortex in some cases with hippocampal sclerosis (Margerison and Corsellis, 1966). Thus, neocortical alterations may be more widespread than previously thought.
Decreased parvalbumin immunostaining and electrocorticography
Patches of decreased parvalbumin immunostaining were found, surprisingly, in both spiking and non-spiking cortex, as revealed by intraoperative electrocorticography, making it difficult to correlate unambiguously abnormal parvalbumin immunostaining with patterns of spiking cortex (see Marco et al., 1996). Similarly, Ferrer et al. found multiple patches of decreased parvalbumin (and calbindin) immunoreactivity in the neocortex of patients with temporal lobe epilepsy and mesial sclerosis, and in patients with cryptogenic frontal lobe epilepsy (Ferrer et al., 1994). These authors concluded that, because in these patients no morphological abnormalities were found in standard histopathological preparations and because of the lack of correlation between abnormal areas of staining and electrically abnormal foci recorded with electrocorticography, the existence of these patches might be explained simply as non-specific fluctuations in the immunocytochemical staining due to the release of calcium from intracellular calcium-binding proteins during depolarization of the cells. Later studies showed that the patches of decreased immunostaining for parvalbumin were coincident in adjacent sections with patches of decreased immunostaining for the GABA-synthesizing enzyme glutamic acid decarboxylase (Marco et al., 1996). Furthermore, using correlative light and electron microscopic methods to examine quantitatively the synaptic density in the neuropil of some of the regions showing focal decreases, these zones were also found to correspond to fine disorganizations in synaptic circuits which consist of increases and decreases of asymmetrical (presumptive excitatory) and symmetrical (presumptive inhibitory) synapses, compared with the adjacent normal cortex (DeFelipe et al., 1993; Marco and DeFelipe, 1997). We suggest that these changes might be due to a focal proliferation of excitatory axon terminals (probably originating from the local axonal arborizations of neighbouring pyramidal cells) and loss of inhibitory terminals, which may lead to hyperexcitatory synaptic circuits.
The pathophysiological significance of intraoperative electrocorticography recordings is unclear because of the lack of a good correlation between seizure outcome after surgery and the localization of regions with maximal spike incidence or amplitude (Engel et al., 1975; Cascino et al., 1995; Schwartz et al., 1997). As pointed out by Alarcon et al. (Alarcon et al., 1997), among others, this is probably due to the complexity of electrocorticography and the difficulty in distinguishing between primary and propagated spikes simply by visual inspection. Primary spikes would be associated with structural lesions (i.e. epileptogenic regions), whereas regions with propagated spikes would be normal (see Alarcon et al., 1997, and references therein; see also Marco and DeFelipe, 1997). Further studies are necessary to investigate the relationship between patches of decreased immunostaining for parvalbumin and electrocorticography, using more sophisticated methods (such as computerized analysis of spike propagation proposed, for example, by Alarcon and colleagues) to identify epileptogenic zones. Nevertheless, as occurs with many pathologies associated with brain disease, a particular alteration is often difficult to ascertain as the cause or the consequence of the disease. For example, hippocampal sclerosis, which is often considered to be a primary lesion associated with epilepsy, has been shown to be induced after repeated kindled seizures in the rat (Cavazos et al., 1994). Also, seizures may not be present in patients with severe damage of the hippocampal formation and amygdala (Corsellis, 1957; Haymaker et al., 1958; Gloor, 1991). Thus, it is also possible that in some patients epileptic activity originating primarily in the neocortex or in the cortex of the parahippocampal gyrus induces neuronal loss in the hippocampus.
In conclusion, regions of neocortex showing decreases in parvalbumin immunoreactivity and altered synaptic circuits may be particularly relevant to epileptogenesis, but we do not know whether these changes are consequences of epileptic seizures or represent key components of a primary process that initiates the development of seizures. I propose a hypothesis below which supports the idea that the non-specific loss of chandelier cells may lead to seizures.
|
A chandelier cell hypothesis: variability in numbers of interneurons and non-specific loss of chandelier cells as a risk for the development of epilepsy |
|---|
|
TopAbstractIntroductionTypes of neocortical neuronsSurgical treatment of temporal...A chandelier cell hypothesis:...References |
|---|
Since a primary pathology found in an epileptic patient may also be present in patients who are not epileptic (e.g. Meldrum and Bruton, 1992), what alterations are necessary in the neuronal circuitry to induce epilepsy, and why does this happen in some patients but in others it does not? It has been argued that a genetic component appears to be an important factor in the development of both idiopathic and symptomatic epilepsy (Falconer, 1971b
; Treiman and Treiman, 1993; Cendes et al., 1998). It is possible that there is a normal, genetically determined variability in the number of interneurons which might predispose some individuals to develop epilepsy more than others in the face of a lesion or other precipitating factor that leads to neuronal loss. Evidence for variability comes from several quarters. First, the well-known differences in brain size suggest a considerable variability in the constitutive elements of the brain and their connections (e.g. see Purves et al., 1996; Peters et al., 1998). Secondly, it is the experience of many researchers working with human brain tissue (including ourselves) that there are great interindividual differences in the number of cortical interneurons labelled for GABA or calcium-binding proteins. In the monkey cerebral cortex, evidence for these differences is seen in counts of GABA neurons, which may vary considerably between one monkey and another (see Hendry et al., 1987). At the electron microscope level, significant interindividual differences have been demonstrated in the number of inhibitory synapses in the cat visual cortex (Beaulieu and Colonnier, 1985). We have observed variations of up to 30% in the density of symmetrical synapses in the human temporal neocortex (unpublished work). Although no data for variability in the number of chandelier cells is available from normal cortical areas, I propose (Fig. 8) that there may be individuals whose brains have relatively low numbers of GABA neurons and axon terminals (`GABAL type' brains), while the brains of other individuals have high numbers (`GABAH type' brains), and that the same factor inducing a loss of interneurons may produce epilepsy in one individual but not in the other.
|
This hypothesis is based on the fact that the GABAergic inputs on the dendrites and somata of pyramidal cells originate from many and diverse types of interneurons (for recent publications, see Kawaguchi and Kubota, 1997, 1998; Tamás et al., 1997; Thomson and Deuchars, 1997; Somogyi et al., 1998), whereas at the level of the axon initial segment, virtually all synapses arise from one or a few chandelier cells (Fairén and Valverde, 1980
; Peters et al., 1982; Freund et al., 1983; Somogyi et al., 1983c; reviewed in Somogyi et al., 1982; DeFelipe and Fariñas, 1992). The number of GABAergic synapses which a pyramidal cell may receive has been estimated to be in the order of several hundreds (e.g. see DeFelipe, 1997). However, cells with axons ending on dendrites or somata form relatively few synapses with a given pyramidal cell (in the order of tens or less; e.g. Peters and Fairén, 1978; Somogyi et al., 1983a; Kisvárday et al., 1985; Tamás et al., 1997, 1998). Consequently, the loss of a few such cells might have little impact on the inhibitory control of the pyramidal cell, whereas if chandelier cells were affected, it would be crucial. The loss of chandelier cells could be selective or non-selective as a response to a given precipitating factor. However, as far as I know, there is no lesion or precipitating factor which is epileptogenic per se since, as pointed out above, the primary pathology or precipitating factor may also be present in patients who are not epileptic. Thus, if a given precipitating factor did produce a selective loss of chandelier cells, then many more epileptic patients should be expected.
Finally, since chandelier cells are also components of neuronal circuits in the entorhinal cortex, subiculum, hippocampus and fascia dentata (see section `What is a chandelier cell?'), this hypothesis may also be valid in cases which present alterations only in these cortical regions. For example, Freund and Buzsáki found a selective loss of chandelier cell axons in rat hippocampal transplants which showed epileptiform activity (Freund and Buzsáki, 1988). Work in progress in our laboratory suggests that there is a decrease in chandelier cells in the apparently histologically normal subiculum and/or entorhinal cortex in epileptic patients who show a severe neuronal loss in the CA1 field. This may be particularly relevant in epileptic patients in whom seizure onset is restricted to mesial temporal structures and who present hippocampal sclerosis. As mentioned above, the incomplete removal of the parahippocampal gyrus, and particularly of the subiculum, in these patients resulted in a less satisfactory outcome (Siegel et al., 1990). Therefore, it has been hypothesized that the histologically normal parahippocampal gyrus may serve as a hyperexcitable fringe area which would amplify the relatively few excitatory outputs from the sclerotic Ammon's horn (Babb and Brown, 1986; Wieser, 1986; Siegel et al., 1990; Wieser et al., 1993). Our findings support this amplifier hypothesis, because the loss of these interneurons would contribute to the hyperexcitability of subicular pyramidal cells. In conclusion, it seems that the loss of chandelier cells might be non-specific and that when this occurs epilepsy may develop. Therefore, these cells might represent a key component in the aetiology of human epilepsy.
|
Acknowledgments |
|---|
The author wishes to thank Edward G. Jones and Antonio Gil-Nagel for critically reading the manuscript. This work was supported by FIS grant 98/0933 and Comunidad de Madrid grant 08.5/014/1997.
|
References |
|---|
|
TopAbstractIntroductionTypes of neocortical neuronsSurgical treatment of temporal...A chandelier cell hypothesis:...References |
|---|
Alarcon G, Garcia Seoane JJ, Binnie CD, Martin Miguel MC, Juler J, Polkey CE, et al. Origin and propagation of interictal discharges in the acute electrocorticogram. Implications for pathophysiology and surgical treatment of temporal lobe epilepsy. Brain 1997; 120: 2259–82.
Anderson SA, Classey JD, Condé F, Lund JS, Lewis DA. Synchronous development of pyramidal neuron dendritic spines and parvalbumin-immunoreactive chandelier neuron axon terminals in layer III of monkey prefrontal cortex. Neuroscience 1995; 67: 7–22.
Andressen C, Blümcke I, Celio MR. Calcium-binding proteins: selective markers of nerve cells. [Review]. Cell Tissue Res 1993; 271: 181–208.
Avanzini G, Engel J Jr, Fariello R, Heinemann U, editors. Neurotransmitters in epilepsy. Amsterdam: Elsevier; 1992.
Babb TL, Brown WJ. Neuronal, dendritic, and vascular profiles of human temporal lobe epilepsy correlated with cellular physiology in vivo. Adv Neurol 1986; 44: 949–66.
Babb TL, Brown WJ. Pathological findings in epilepsy. In: Engel J Jr, editor. Surgical treatment of the epilepsies. New York: Raven Press; 1987. p. 511–40.
Babb TT, Pretorius JK. Pathologic substrates of epilepsy. In: Wyllie E, editor. The treatment of epilepsy: principles and practice. Philadelphia: Lea and Febiger; 1993. p. 55–70.
Babb TL, Brown WJ, Pretorious J, Davenport C, Lieb JP, Crandall PH. Temporal lobe volumetric cell densities in temporal lobe epilepsy. Epilepsia 1984; 25: 729–40.
Beaulieu C, Colonnier M. A laminar analysis of the number of round-asymmetrical and flat-symmetrical synapses on spines, dendritic trunks, and cell bodies in area 17 of the cat. J Comp Neurol 1985; 231: 180–9.
Ben-Ari Y, Represa A. Brief seizure episodes induce long-term potentiation and mossy fibre sprouting in the hippocampus. [Review]. Trends Neurosci 1990; 13: 312–8.
Buhl EH, Han Z-S, Lörinczi Z, Stezhka VV, Karnup SV, Somogyi P. Physiological properties of anatomically identified axo-axonic cells in the rat hippocampus. J Neurophysiol 1994; 71: 1289–307.
Cascino GD, Jack CR Jr, Parisi JE, Sharbrough FW, Hirschorn KA, Meyer FB, et al. Magnetic resonance imaging-based volume studies in temporal lobe epilepsy: pathological correlations. Ann Neurol 1991; 30: 31–6.
Cascino GD, Trenerry MR, Jack CR Jr, Dodick D, Sharbrough FW, So EL, et al. Electrocorticography and temporal lobe epilepsy: relationship to quantitative MRI and operative outcome. Epilepsia 1995; 36: 692–6.
Cavazos JE, Das I, Sutula TP. Neuronal loss induced in limbic pathways by kindling: evidence for induction of hippocampal sclerosis by repeated brief seizures. J Neurosci 1994; 14: 3106–21.
Celio MR. Parvalbumin in most 
-aminobutyric acid-containing neurons of the rat cerebral cortex. Science 1986; 231: 995–7.
Cendes F, Lopes-Cendes I, Andermann E, Andermann F. Familial temporal lobe epilepsy: a clinically heterogeneous syndrome. Neurology 1998; 50: 554–7.
Condé F, Lund JS, Jacobowitz DM, Baimbridge KG, Lewis DA. Local circuit neurons immunoreactive for calretinin, calbindin D-28k or parvalbumin in monkey prefrontal cortex: distribution and morphology. J Comp Neurol 1994; 341: 95–116.
Conti F. Toward the anatomical identification of glutamatergic neurons and synapses in the cerebral cortex. In: Meldrum BS, Moroni F, Simon RP, Woods JH, editors. Excitatory amino acids. Fidia Research Foundation Series, Vol. 5. New York: Raven Press; 1991. p. 45–53.
Corsellis JA. The incidence of Ammon's horn sclerosis. Brain 1957; 80: 193–208.
DeFelipe J. Neocortical neuronal diversity: chemical heterogeneity revealed by colocalization studies of classic neurotransmitters, neuropeptides, calcium binding proteins and cell surface molecules. [Review]. Cereb Cortex 1993; 3: 273–89.
DeFelipe J. Microcircuits in the brain. In: Mira J, Moreno-Díaz R, Cabestany J, editors. Biological and artificial computation: from neuroscience to technology. Lecture Notes in Computer Science, No. 1240. Berlin: Springer; 1997. p. 1–14.
DeFelipe J, Fariñas I. The pyramidal neuron of the cerebral cortex: morphological and chemical characteristics of the synaptic inputs. Prog Neurobiol 1992; 39: 563–607.
DeFelipe J, González-Albo MC. Chandelier cell axons are immunoreactive for GAT-1 in the human neocortex. Neuroreport 1998; 9: 467–70.
DeFelipe J, Hendry SH, Jones EG, Schmechel D. Variability in the terminations of GABAergic chandelier cell axons on initial segments of pyramidal cell axons in the monkey sensory-motor cortex. J Comp Neurol 1985; 231: 364–84.
DeFelipe J, Hendry SH, Jones EG. Visualization of chandelier cell axons by parvalbumin immunoreactivity in monkey cerebral cortex. Proc Natl Acad Sci USA 1989; 86: 2093–7.
DeFelipe J, Garcia Sola R, Marco P, del Río MR, Pulido P, Ramón y Cajal S. Selective changes in the microorganization of the human epileptogenic neocortex revealed by parvalbumin immunoreactivity. Cereb Cortex 1993; 3: 39–48.
DeFelipe J, Huntley GW, del Río MR, Sola RG, Morrison JH. Microzonal decreases in the immunostaining for non-NMDA ionotropic excitatory amino acid receptor subunits GluR 2/3 and GluR 5/6/7 in the human epileptogenic neocortex. Brain Res 1994; 657: 150–8.
DeFelipe J, Sola RG, Marco P. Changes in excitatory and inhibitory synaptic circuits in the human epileptogenic neocortex. In: Conti F, Hicks TP, editors. Excitatory amino acids and the cerebral cortex. Cambridge (MA): MIT Press; 1996. p. 299–312.
del Río MR, DeFelipe J. A study of SMI 32-stained pyramidal cells, parvalbumin-immunoreactive chandelier cells, and presumptive thalamocortical axons in the human temporal neocortex. J Comp Neurol 1994; 342: 389–408.
del Río MR, DeFelipe J. Colocalization of parvalbumin and calbindin D-28k in neurons including chandelier cells of the human temporal neocortex. J Chem Neuroanat 1997; 12: 165–73.
Dinner DS. Posttraumatic epilepsy. In: Wyllie E, editor. The treatment of epilepsy: principles and practice. Philadelphia: Lea and Febiger; 1993. p. 654–8.
Douglas RJ, Martin, KA. Control of neural output by inhibition at the axon initial segment. Neural Comp 1990; 2: 283–92.
Engel J Jr, Driver MV, Falconer MA. Electrophysiological correlates of pathology and surgical results in temporal lobe epilepsy. Brain 1975; 98: 129–56.
Fairén A, Valverde F. A specialized type of neuron in the visual cortex of cat: a Golgi and electron microscope study of chandelier cells. J Comp Neurol 1980; 194: 761–79.
Fairén A, DeFelipe J, Regidor J. Nonpyramidal neurons. General account. In: Peters A, Jones EG, editors. Cerebral cortex, Vol. 1. New York: Plenum Press; 1984. p. 201–53.
Falconer M. Anterior temporal lobectomy for epilepsy. In: Logue V, editor. Operative surgery, Vol. 14. Neurosurgery. 2nd edn. London: Butterworths; 1971a. p. 142–9.
Falconer MA. Genetic and related aetiological factors in temporal lobe epilepsy. A review. Epilepsia 1971b; 12: 13–31.
Falconer MA. Mesial temporal (Ammon's horn) sclerosis as a common cause of epilepsy: aetiology, treatment, and prevention. Lancet 1974; 2: 767–70.
Ferrer I, Oliver B, Russi A, Casas R, Rivera R. Parvalbumin and calbindin-D28k immunocytochemistry in human neocortical epileptic foci. J Neurol Sci 1994; 123: 18–25.
Freund TF, Buzsáki G. Alterations in excitatory and GABAergic inhibitory connections in hippocampal transplants. Neuroscience 1988; 27: 373–85.
Freund TF, Martin KA, Smith AD, Somogyi P. Glutamate decarboxylase-immunoreactive terminals of Golgi-impregnated axoaxonic cells and of presumed basket cells in synaptic contact with pyramidal neurons of the cat's visual cortex. J Comp Neurol 1983; 221: 263–78.
Gabbott PL, Bacon SJ. Local circuit neurons in the medial prefrontal cortex (areas 24a,b,c, 25 and 32) in the monkey: I. Cell morphology and morphometrics. J Comp Neurol 1996; 364: 567–608.
Gil-Nagel A, Risinger MW. Ictal semiology in hippocampal versus extrahippocampal temporal lobe epilepsy. Brain 1997; 120: 183–92.
Gilbert CD. Microcircuitry of the visual cortex. Annu Rev Neurosci 1983; 6: 217–47.
Gloor P. Mesial temporal sclerosis: historical background and an overview from a modern perspective. In: Lüders H, editor. Epilepsy surgery. New York: Raven Press; 1991. p. 689–703.
Haymaker W, Pentschew A, Margoles C, Bingham WG. Occurrence of lesions in the temporal lobe in the absence of convulsive seizures. In: Baldwin M, Bailey P, editors. Temporal lobe epilepsy. Springfield (IL): Charles C. Thomas; 1958. p. 166–202.
Hendry SH, Houser CR, Jones EG, Vaughn JE. Synaptic organization of immunocytochemically identified GABA neurons in the monkey sensory-motor cortex. J Neurocytol 1983; 12: 639–60.
Hendry SH, Schwark HD, Jones EG, Yan J. Numbers and proportions of GABA-immunoreactive neurons in different areas of monkey cerebral cortex. J Neurosci 1987; 7: 1503–19.
Hendry SH, Jones EG, Emson PC, Lawson DE, Heizmann CW, Streit P. Two classes of cortical GABA neurons defined by differential calcium binding protein immunoreactivities. Exp Brain Res 1989; 76: 467–72.
Houser CR. GABA neurons in seizure disorders: a review of immunocytochemical studies. [Review]. Neurochem Res 1991; 16: 295–308.
Houser CR, Vaughn JE, Hendry SH, Jones EG, Peters A. GABA neurons in the cerebral cortex. In: Jones EG, Peters A, editors. Cerebral cortex, Vol. 2. New York: Plenum Press; 1984. p. 63–89.
Houser CR, Harris AB, Vaughn JE. Time course of the reduction of GABA terminals in a model of focal epilepsy: a glutamic acid decarboxylase immunocytochemical study. Brain Res 1986; 383: 129–45.
Jackson GD, Kuzniecky RI, Cascino GD. Hippocampal sclerosis without detectable hippocampal atrophy. Neurology 1994; 44: 42–6.
Jones EG. Varieties and distribution of non-pyramidal cells in the somatic sensory cortex of the squirrel monkey. J Comp Neurol 1975; 160: 205–67.
Jones EG. GABAergic neurons and their role in cortical plasticity in primates. [Review]. Cereb Cortex 1993; 3: 361–72.
Jones-Gotman M, Zatorre RJ, Olivier A, Andermann F, Cendes F, Staunton H, et al. Learning and retention of words and designs following excision from medial or lateral temporal-lobe structures. Neuropsychologia 1997; 35: 963–73.
Kawaguchi Y. Physiological subgroups of nonpyramidal cells with specific morphological characteristics in layer II/III of rat frontal cortex. J Neurosci 1995; 15: 2638–55.
Kawaguchi Y, Kubota Y. GABAergic cell subtypes and their synaptic connections in rat frontal cortex. Cereb Cortex 1997; 7: 476–86.
Kawaguchi Y, Kubota Y. Neurochemical features and synaptic connections of large physiologically-identified GABAergic cells in the rat frontal cortex. Neuroscience 1998; 85: 677–701.
Ketz E. Brain tumours and epilepsy. In: Vinken PJ, Bruyn GW, editors. Handbook of clinical neurology, Vol. 16. Amsterdam: North-Holland; 1974: p. 254–69.
Kisvárday ZF, Martin KA, Whitteridge D, Somogyi P. Synaptic connections of intracellularly filled clutch cells: a type of small basket cell in the visual cortex of the cat. J Comp Neurol 1985; 241: 111–37.
Kisvárday ZF, Adams CB, Smith AD. Synaptic connections of axo-axonic (chandelier) cells in human epileptic temporal cortex. Neuroscience 1986; 19: 1179–86.
Lévesque MF, Nakasato N, Vinters HV, Babb TL. Surgical treatment of limbic epilepsy associated with extrahippocampal lesions: the problem of dual pathology. J Neurosurg 1991; 75: 364–70.
Lewis DA, Lund JS. Heterogeneity of chandelier neurons in monkey neocortex: corticotropin-releasing factor- and parvalbumin-immunoreactive populations. J Comp Neurol 1990; 293: 599–615.
Lewis DA, Foote SL, Cha CI. Corticotropin-releasing factor immunoreactivity in monkey neocortex: an immunohistochemical analysis. J Comp Neurol 1989; 290: 599–613.
Lloyd KG, Bossi L, Morselli PL, Munari C, Rougier M, Loiseau H. Alterations of GABA-mediated synaptic transmission in human epilepsy. [Review]. Adv Neurol 1986; 44: 1033–44.
Lund JS. Spiny stellate neurons. In: Peters A, Jones EG, editors. Cerebral cortex, Vol. 1. New York: Plenum Press; 1984. p. 255–308
Marco P, DeFelipe J. Altered synaptic circuitry in the human temporal neocortex removed from epileptic patients. Exp Brain Res 1997; 114: 1–10.
Marco P, Sola RG, Pulido P, Alijarde MT, Sánchez A, Ramón y Cajal S, et al. Inhibitory neurons in the human epileptogenic temporal neocortex: an immunocytochemical study. Brain 1996; 119: 1327–47.
Marco P, Sola RG, Ramón y Cajal S, DeFelipe J. Loss of inhibitory synapses on the soma and axon initial segment of pyramidal cells in human epileptic peritumoural neocortex: implications for epilepsy. Brain Res Bull 1997; 44: 47–66.
Margerison JH, Corsellis JA. Epilepsy and the temporal lobes. A clinical, electroencephalographic and neuropathological study of the brain in epilepsy, with particular reference to the temporal lobes. Brain 1966; 89 : 499–530.
Marin-Padilla M. The chandelier cell of the human visual cortex: a Golgi study. J Comp Neurol 1987; 256: 61–70.
Martínez A, Lübke J, del Río JA, Soriano E, Frotscher M. Regional variability and postsynaptic targets of chandelier cells in the hippocampal formation of the rat. J Comp Neurol 1996; 376: 28–44.
Mayanagi Y, Watanabe E, Kaneko Y. Mesial temporal lobe epilepsy: clinical features and seizure mechanism. Epilepsia 1996; 37 Suppl 3: 57–60.
McNamara JO. Cellular and molecular basis of epilepsy. J Neurosci 1994; 14: 3413–25.
Meldrum BS. Neurotransmission in epilepsy. Epilepsia 1995; 36 Suppl 1: S30–5.
Meldrum BS, Bruton CJ. Epilepsy. In: Adams JH, Duchen LW, editors. Greenfield's neuropathology. 5th edn. London: Edward Arnold; 1992. p. 1246–83.
Mikkonen M, Soininen H, Pitkänen A. Distribution of parvalbumin-, calretinin-, and calbindin-D28k-immunoreactive neurons and fibers in the human entorhinal cortex. J Comp Neurol 1997; 388: 64–88.
Morrell F. Varieties of human secondary epileptogenesis. [Review]. J Clin Neurophysiol 1989; 6: 227–75.
Morris HH, Estes ML. Brain tumors and chronic epilepsy. In: Wyllie E, editor. The treatment of epilepsy: principles and practice. Philadelphia: Lea and Febiger; 1993. p. 659–66.
Naegele JR, Katz LC. Cell surface molecules containing N-acetylgalactosamine are associated with basket cells and neurogliaform cells in cat visual cortex. J Neurosci 1990; 10: 540–57.
O'Brien TJ, Kilpatrick C, Murrie V, Vogrin V, Morris K, Cook MJ. Temporal lobe epilepsy caused by mesial temporal sclerosis and temporal neocortical lesions. A clinical and electroencephalographic study of 46 pathologically proven cases. Brain 1996; 119: 2133–41.
Olivier A. Temporal resections in the surgical treatment of epilepsy. In: Theodore WH, editor. Surgical treatment of epilepsy. Epilepsy Research, Suppl 5. Amsterdam: Elsevier; 1992. p. 175–88.
Ottersen OP, Storm-Mathisen J. Neurons containing or accumulating transmitter amino acids. In: Björklund A, Hökfelt T, Kuhar MJ, editors. Handbook of chemical neuroanatomy, Vol. 3. Amsterdam: Elsevier; 1984. p. 141–246.
Peters A, Fairén A. Smooth and sparsely-spined stellate cells in the visual cortex of the rat: a study using a combined Golgi–electron microscopic technique. J Comp Neurol 1978; 181: 129–71.
Peters A, Proskauer CC, Ribak CE. Chandelier cells in rat visual cortex. J Comp Neurol 1982; 206: 397–416.
Peters A, Morrison JH, Rosene DL, Hyman BT. Are neurons lost from the primate cerebral cortex during normal aging? [Review]. Cereb Cortex 1998; 8: 295–300.
Prince DA, Connors BW. Mechanisms of interictal epileptogenesis. [Review]. Adv Neurol 1986; 44: 275–99.
Purves D, White LE, Riddle DR. Is neural development darwinian? [see comments]. Trends Neurosci 1996; 19: 460–4. Comment in: Trends Neurosci 1997; 20: 291–3.
Ribak CE. Axon terminals of GABAergic chandelier cells are lost at epileptic foci. Brain Res 1985; 326: 251–60.
Ribak CE. Epilepsy and the cortex: anatomy. In: Peters A, Jones EG, editors. Cerebral cortex, Vol. 9. New York: Plenum Press; 1991. p. 427–83.
Ribak CE, Harris AB, Vaughn JE, Roberts E. Inhibitory, GABAergic nerve terminals decrease at sites of focal epilepsy. Science 1979; 205: 211–4.
Ribak CE, Bradburne RM, Harris AB. A preferential loss of GABAergic, symmetric synapses in epileptic foci: a quantitative ultrastructural analysis of monkey neocortex. J Neurosci 1982; 2: 1725–35.
Saygi S, Spencer SS, Scheyer R, Katz A, Mattson R, Spencer DD. Differentiation of temporal lobe ictal behavior associated with hippocampal sclerosis and tumors of temporal lobe. Epilepsia 1994; 35: 737–42.
Schmidt S, Braak E, Braak H. Parvalbumin-immunoreactive structures of the adult human entorhinal and transentorhinal region. Hippocampus 1993; 3: 459–70.
Schwartz TH, Bazil CW, Walczak TS, Chan S, Pedley TA, Goodman RR. The predictive value of intraoperative electrocorticography in resections for limbic epilepsy associated with mesial temporal sclerosis. Neurosurgery 1997; 40: 302–11.
Schwartzkroin PA. Cellular electrophysiology of human epilepsy. [Review]. Epilepsy Res 1994a; 17: 185–92.