Department of Pharmacology and Cancer Biology and Department of Neurobiology, Duke University Medical Center, Durham, North Carolina 27710
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Hardison, Jeremy L., Maxine M. Okazaki, and J. Victor Nadler. Modest Increase in Extracellular Potassium Unmasks Effect of Recurrent Mossy Fiber Growth. J. Neurophysiol. 84: 2380-2389, 2000. 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 GABAA 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 ~50% when [K+]o was raised to either 4.75 or 6 mM. Addition of bicuculline (3 or 30 µM) 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 µM), 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 GABAA receptor-mediated inhibition is less efficacious and frequency facilitation may not be a significant factor.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The dentate granule cells of the
hippocampus are believed to function as a buffer against seizure
propagation through the limbic circuit (Collins et al.
1983; Lothman et al. 1992
; Stringer et
al. 1989
). Studies of animal models of epilepsy suggest that seizures that originate from or invade the entorhinal cortex normally reach the hippocampus only if they are strong enough to activate the
dentate granule cells synchronously (Stringer and Lothman 1992
). These cells resist seizure propagation for several
reasons, including a highly negative resting membrane potential, strong tonic inhibition from GABA interneurons (Freund and
Buzsáki 1996
; Otis et al. 1991
), low
Ca2+ conductance (Fricke and Prince
1984
), and lack of a synaptic mechanism for synchronization of
cell discharge. With respect to the latter point, spontaneous cellular
bursts can be induced by increasing the K+
concentration or reducing the Ca2+ concentration
in the medium bathing a hippocampal slice, but synchronized granule
cell activity appears only under extreme conditions (Pan and
Stringer 1996
, 1997
; Patrylo et al. 1994
; Schweitzer and Williamson 1995
; Schweitzer et al.
1992
). Granule cell synchrony is difficult to achieve, in part,
because the few synaptic connections among these cells present in
normal brain (Molnár and Nadler 1999
;
Okazaki et al. 1999
) are insufficient to support
recurrent synchronous discharge. However, in many persons with temporal
lobe epilepsy the recurrent mossy fiber pathway of the dentate gyrus
expands dramatically (Babb et al. 1991
; Franck et
al. 1995
; Represa et al. 1989
; Sutula et
al. 1989
). These new connections among granule cells could
reduce their threshold for synchronization, thus enhancing the
participation of these cells in seizures.
Despite formation of new recurrent excitatory connections, granule
cells still remain difficult to recruit into epileptiform activity.
Stimulation of the perforant path in vivo does not usually evoke
reverberating excitation (Buckmaster and Dudek 1997).
Similarly, responses of the granule cell population usually remain
unaltered during low-frequency stimulation in hippocampal slices
superfused with standard artificial cerebrospinal fluid (ACSF). For
example, antidromic stimulation of the mossy fibers seldom evokes
extracellularly recorded granule cell activity beyond the antidromic
population spike (Cronin et al. 1992
; Patrylo and
Dudek 1998
; Tauck and Nadler 1985
). Under what
circumstances then will these novel connections mediate synchronous
activity? Previous research suggested that several factors present
during the period just before seizure initiation may combine to
strengthen recurrent excitatory circuitry and enhance its ability to
support epileptiform activity. Among the most obvious candidate factors
are increased [K+]o
(Patrylo and Dudek 1998
; Traub and Dingledine
1990
) and reduced GABA-mediated inhibition (Cronin et
al. 1992
; Miles et al. 1984
; Patrylo and
Dudek 1998
). With respect to the recurrent mossy fiber pathway,
use-dependent forms of synaptic plasticity may also play an important
role. Mossy fiber synapses on CA3 pyramidal cells exhibit marked
frequency facilitation. Increasing the stimulus frequency from 0.0125 to 0.33 Hz increases the size of the excitatory postsynaptic current
(EPSC) by more than fivefold (Salin et al. 1996
). The
range of presynaptic activity that provokes the greatest facilitation
(0.1-1 Hz) corresponds to the normal firing frequencies of granule
cells in vivo (Jung and McNaughton 1993
). If mossy fiber-granule cell synapses also exhibit this form of plasticity, then
the recurrent pathway might support synchronous granule cell discharge
preferentially when driven at particular frequencies or in particular
patterns. Thus we tested, alone and in combination, the ability of
modestly elevated [K+]o,
reduced synaptic inhibition, and frequency facilitation to unmask the
effect of recurrent mossy fiber circuitry.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Pilocarpine-induced status epilepticus
Adult male Sprague-Dawley rats (175-200 g; Zivic-Miller
Laboratories, Allison Park, PA) were treated with pilocarpine
hydrochloride (335-365 mg/kg ip) 30 min after receiving 2 mg/kg ip
each of scopolamine methyl bromide and terbutaline hemisulfate. Status
epilepticus, defined as a continuous limbic motor seizure of stage 2 or
higher (Racine 1972), usually began within 90 min and
was terminated 3.5 h after onset with a single injection of
phenobarbital sodium (50 mg/kg ip). No systematic behavioral monitoring
was performed after recovery from status epilepticus. However, all rats
that develop status epilepticus after administration of pilocarpine exhibit chronic spontaneous limbic motor seizures after a latent period
of 1-3 wk (Lemos and Cavalheiro 1996
; Mello et
al. 1993
).
Some pilocarpine-treated rats exhibited only a few brief seizures that
never progressed to status epilepticus. Previous work demonstrated the
absence of both neuronal degeneration and supragranular mossy fiber
growth in these animals (Okazaki et al. 1999). They were
therefore used as controls.
Preparation and incubation of hippocampal slices
Animals were studied 10 wk after status epilepticus because
previous studies suggested that the recurrent mossy fiber pathway required this period of time to become fully mature (Okazaki et al. 1995
). The rat was decapitated under ether anesthesia, the brain was removed, and 400-µm-thick transverse slices were cut from
the caudal third of the hippocampal formation with a vibratome. Slices
used for electrophysiological recording corresponded to horizontal
plates 98-100 of Paxinos and Watson (1986)
. These
slices were trimmed and transferred to a beaker of standard ACSF, which contained (in mM) 122 NaCl, 25 NaHCO3, 3.1 KCl,
1.8 CaCl2, 1.2 MgSO4, 0.4 KH2PO4, and 10 D-glucose, pH 7.4. They were continuously gassed at room
temperature with 95% O2-5%
CO2.
Beginning 1.5 h after preparation, a slice was transferred to a
small experimental chamber, placed on a nylon net and submerged in ACSF
equilibrated with 95% O2-5%
CO2 at 35°C. The superfusion rate was 1 ml/min.
A monopolar stimulating electrode fashioned from 25-µm-diam nichrome
wire and insulated to the tip with polymerized polyvinyl resin
(Formvar) was placed in stratum lucidum of area CA3b >100 µm from
the opening of the dentate hilus. Then an extracellular recording
electrode fashioned from borosilicate glass and filled with 0.15 M NaCl
(2-8 M) was placed in the granule cell body layer of the dentate
gyrus. The recording electrode was positioned at numerous locations in
this layer until the location was found at which the antidromic
population spike was of maximal amplitude. The stimulating electrode
was then moved perpendicular to the pyramidal cell body layer until
stimulation evoked an antidromic population spike of the greatest
possible amplitude. The final optimization of antidromic spike
amplitude was achieved by adjusting both electrodes along the
z axis.
Constant-current stimuli (rectangular pulses, 100-µs duration) were varied from 100 to 800 µA at 100-µA intervals to determine the smallest current that evoked a maximal antidromic response. This stimulus current was used for the rest of the experiment. After allowing 1 h for the response to stabilize, a train of 10 pulses was presented at a frequency of 0.2 Hz and then again at a frequency of 1 Hz. Then 10 trains of 10 pulses each were presented at a frequency of 10 Hz with a 20-s intertrain interval. Responses were filtered below 2 kHz, digitized at 10 kHz, and stored to disk with use of a Digidata board and Axoscope (Axon Instruments, Foster City, CA) running on a Micron P-100 microcomputer.
The superfusion medium was then modified, and 45 min later the stimulation protocol was carried out as before. In some experiments, a second change of medium and third round of stimulation were performed.
Data analysis
Antidromically evoked responses were inspected visually for the
presence of epileptiform activity. This judgment was based on the
presence or absence of any activity on the trace after the antidromic
population spike. The magnitude of antidromically evoked epileptiform
activity was quantitated by measurement of the coastline index
(Dingledine et al. 1986). For this purpose, the Axoscope
files were transferred to Microsoft Excel and the Pythagorean theorem
was used to determine the change in response between successive data
points. These changes were then summed for the 30-ms period that began
at the end of the antidromic population spike. Unless stated otherwise,
quantitative analyses were performed on recordings made at a stimulus
frequency of 0.2 Hz, because all 10 responses recorded during such
trains were practically identical. The coastline index was determined
for each response, and indices for the 10 responses in the train were
averaged. It should be noted that the presence or absence of
epileptiform activity was determined only from visual inspection of the
traces and not from the value of the coastline index. The coastline
index is not useful for this purpose because differences in baseline
noise could alter the coastline index by as much as brief, but clearly recognizable, epileptiform activity.
The percentage effect of (2S,2'R,3'R)-2-(2',3'-dicarboxycyclopropyl)glycine (DCG-IV) on antidromically evoked epileptiform activity was calculated by subtracting the coastline index after addition of DCG-IV to the superfusion medium from the coastline index before addition of DCG-IV and dividing this value by the difference of the coastline index before the addition of DCG-IV and the expected baseline. The expected baseline was the value of the coastline index in the absence of epileptiform activity. It was calculated by averaging the coastline indices obtained from all slices in the status epilepticus group that had no visible epileptiform response during recording in standard ACSF. This calculation yielded a value of 2.54.
Timm histochemistry
After the last set of recordings, the slice was immersed
in a solution of 0.1% (wt/vol) Na2S, 0.1 M
sodium phosphate buffer, pH 7.3, for 90 min. Then it was fixed in
phosphate-buffered 10% formalin/0.9% (wt/vol) NaCl at 4°C for 1-2
days. After being embedded in albumin-gelatin, the slice was cut into
30-µm-thick sections with a Vibratome. Slide mounted sections were
stained for the presence of heavy metals as described by
Danscher (1981) and lightly counterstained with cresyl
violet. The density of Timm staining in the supragranular zone of the
dentate gyrus was estimated on an ordinal scale of 0-3 as previously
described (Okazaki et al. 1995
; Tauck and Nadler
1985
). Sections were scored without knowledge of how the animal
had been treated or of the electrophysiological results. In slices from
control rats, scattered clusters of mossy fiber-like Timm stain were
usually present in the supragranular zone (Molnár and
Nadler 1999
; Okazaki et al. 1999
), qualifying them for a Timm score of 1. Some slices from the status epilepticus group were also assigned a score of 1. In these instances, there were
more clusters of supragranular stain than were present in controls, but
not enough to qualify for a score of 2.
Materials
DCG-IV was purchased from Tocris Cookson (Bristol, UK) and
bicuculline methiodide was from Research Biochemicals (Natick, MA).
Phenobarbital sodium, pilocarpine hydrochloride, ()scopolamine methyl
bromide, and terbutaline hemisulfate were obtained from Sigma Chemical
(St. Louis, MO).
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Control group
Antidromic stimulation of the mossy fibers did not evoke epileptiform activity in slices from control rats under any condition tested (15 slices from 12 animals; Table 1). Conditions tested included 4.75 mM [K+]o followed by 4.75 mM [K+]o + 3 µM bicuculline (n = 2), 6 mM [K+]o (n = 1), 3 µM bicuculline (n = 4), and 6 mM [K+]o + 30 µM bicuculline (n = 8). The coastline index measured under the most extreme condition tested, 6 mM [K+]o + 30 µM bicuculline, was not significantly different from the coastline index measured when the same slice was superfused with standard ACSF (98 ± 4%, mean ± SE). No epileptiform activity emerged when the stimulus frequency was increased from 0.2 to 1 or 10 Hz.
|
Status epilepticus group
EFFECTS OF INCREASING [K+]O. In slices from rats that had developed status epilepticus, antidromic stimulation of the mossy fibers during superfusion with standard ACSF (3.5 mM K+) usually evoked only a single population spike (Fig. 1, Table 1). We observed epileptiform activity in only 14% of these slices.
|
|
|
EFFECTS OF BICUCULLINE. To test the hypothesis that block of GABA inhibition alone could unmask the effect of recurrent excitation in the dentate gyrus, slices were superfused with the GABAA receptor antagonist bicuculline. Two concentrations were used, 3 and 30 µM. At a concentration of 3 µM, bicuculline reduces monosynaptic inhibition onto dentate granule cells by ~40%, whereas 30 µM bicuculline abolishes the response (P. Molnár and J. V. Nadler, unpublished observations). However, there was no consistent difference between additions of 3 and 30 µM bicuculline in the present study.
Addition of bicuculline to standard ACSF never caused antidromic stimulation to evoke epileptiform activity (n = 8; Table 2). Similar results were obtained when the slice was first exposed to 4.75 or 6 mM K+ provided that epileptiform activity had not been observed in the presence of elevated K+ alone (n = 3; Fig. 1A, Table 2). If epileptiform activity was already present, addition of bicuculline to the superfusion medium increased the amplitude and duration of the response without changing the waveform (Fig. 1, B and C). Similar effects were observed with [K+]o concentrations of 3.5 (n = 2), 4.75 (n = 3), and 6 (n = 3) mM. These effects were reflected in a substantial enhancement of the coastline index (Fig. 2; P < 0.02, Wilcoxon signed ranks test). In some experiments, bicuculline was added to the superfusion medium at the same time [K+]o was increased from 3.5 to either 4.75 or 6 mM. The proportion of experiments in which such treatment caused antidromic stimulation to evoke epileptiform activity was no greater than that obtained by increasing [K+]o alone (Table 2). However, the elevation of [K+]o with addition of bicuculline produced the greatest average increase in coastline index of any treatment studied (Fig. 2).OCCURRENCE OF EPILEPTIFORM ACTIVITY DID NOT CORRELATE WITH
ANTIDROMIC SPIKE AMPLITUDE.
Antidromic stimulation of the mossy fibers evoked epileptiform activity
in at most about half the slices tested, even under conditions expected
to maximize excitability (6 mM
[K+]o + 30 µM
bicuculline). In several instances, epileptiform activity was evoked in
one slice from an animal and not in another. One potential explanation
for these findings is that different numbers of mossy fibers were
activated by the stimulus. The more mossy fibers are activated, as
indicated by increased amplitude of the antidromic population spike
(Andersen et al. 1971), the more recurrent excitatory circuitry should be recruited. There was, however, no
significant difference between the size of the antidromic spike in
slices with (3.9 ± 0.3 mV; n = 30) and without
(4.1 ± 0.3 mV; n = 33) evoked epileptiform activity.
In fact, spike amplitude tended to decline when
[K+]o was increased. Yet
the smaller antidromic spike was frequently followed by epileptiform
activity that was not present before.
CORRELATION WITH TIMM SCORE.
The Timm score serves as a valid measure of recurrent mossy fiber
growth (Okazaki et al. 1995). Antidromic stimulation
evoked epileptiform activity in nine slices during superfusion with
standard ACSF. All but one of these slices was assigned a Timm score of 3 (Fig. 3). The likelihood of evoking
epileptiform discharge correlated significantly with the value of the
Timm score (P < 0.05,
2 test). For
each value of the Timm score, raising
[K+]o increased the
percentage of slices that responded with epileptiform activity. This
effect was particularly dramatic in slices with the lowest density of
supragranular mossy fibers, but there was still a trend toward higher
probability of epileptiform activity with more robust mossy fiber
growth. The magnitude of epileptiform activity, however, showed no
relationship to the Timm score.
|
DCG-IV REDUCES ANTIDROMICALLY EVOKED EPILEPTIFORM ACTIVITY.
Mossy fiber terminals express a type II metabotropic glutamate receptor
(Shigemoto et al. 1997). Activation of this receptor inhibits synaptic transmission (Kamiya et al. 1996
). We
tested the ability of DCG-IV, an agonist of type II metabotropic
glutamate receptors (Conn and Pin 1997
), to attenuate
antidromically evoked epileptiform activity. The results of one such
experiment are shown in Fig. 4,
A-C. In this example, raising
[K+]o to 4.75 mM with
addition of 30 µM bicuculline converted the antidromically evoked
response from a single population spike to multiple population spikes.
Addition of a maximal concentration of DCG-IV (1 µM) to the
superfusion medium strongly depressed this response. A similar result
was obtained in each of six experiments (Fig. 4D). DCG-IV
reduced the coastline index by 79 ± 8%.
|
EFFECTS OF REPETITIVE STIMULATION. When 10 antidromic stimulus pulses were presented at a frequency of 0.2 or 1 Hz, all responses were essentially superimposable on one another. The coastline index did not change significantly during the train (Fig. 5D). This result was obtained at all values of [K+]o tested and in either the presence or absence of bicuculline. There was no consistent evidence of frequency facilitation, even when we excluded results from experiments in which antidromic stimulation failed to evoke epileptiform activity. For example, in the 11 positive experiments with 6 mM [K+]o and 30 µM bicuculline, the ratio of coastline indices for response 10 compared with response 1 was 1.07 ± 0.04 for stimulation at 0.2 Hz and 1.08 ± 0.07 for stimulation at 1 Hz.
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Our results provide insights into the factors that regulate expression of recurrent excitation in the dentate gyrus of epileptic brain. They demonstrated that antidromic stimulation of the mossy fibers in standard ACSF evoked epileptiform granule cell discharge only in hippocampal slices from pilocarpine-treated rats with recurrent mossy fiber growth, increasing [K+]o from 3.5 to 4.75 or 6 mM increased the proportion of slices in which epileptiform discharge was evoked, increased [K+]o also reduced the extent of recurrent mossy fiber growth required to support epileptiform activity, depression or block of GABAA receptor-mediated inhibition only increased the magnitude of antidromically evoked epileptiform activity not the probability of evoking such activity, facilitation of epileptiform activity did not occur during antidromic stimulation at frequencies reported to enhance the EPSC at mossy fiber-CA3 pyramidal cell synapses, and repetitive stimulation at 10 Hz reduced, and sometimes abolished, the epileptiform response.
Recurrent mossy fiber growth as a prerequisite for antidromically evoked epileptiform activity
Pilocarpine-induced status epilepticus triggers in the dentate
gyrus several pathologic sequelae that could facilitate epileptiform activity. Besides expansion of the recurrent mossy fiber pathway, some
hilar GABA neurons degenerate (Obenaus et al. 1993) and
possible disturbances of inhibitory synaptic transmission from other
causes have also been described (Gibbs et al. 1997
;
Isokawa 1996
). Any of these changes could, in principle,
have contributed to antidromically evoked epileptiform activity.
However, two lines of evidence suggest that expansion of the recurrent
mossy fiber pathway played a key role. First, the occurrence of
antidromically evoked epileptiform activity was clearly related to the
presence and extent of recurrent mossy fiber growth. Antidromic
stimulation never evoked epileptiform activity in slices from control
rats. Furthermore in slices from the status epilepticus group, robust
recurrent mossy fiber growth was essentially required to evoke
epileptiform activity during superfusion with standard ACSF. These
observations closely resemble results obtained with hippocampal slices
from kainic acid-treated rats (Cronin et al. 1992
;
Patrylo and Dudek 1998
; Tauck and Nadler 1985
). Second, DCG-IV, an agonist of type II metabotropic
glutamate receptors, markedly reduced the magnitude of antidromically
evoked epileptiform activity. Within the rat hippocampal formation,
type II receptors have a restricted distribution. Immunoreactivity is
found only on the preterminal axons of the mossy fiber pathway and the
medial perforant path (Shigemoto et al. 1997
). The
concentration of DCG-IV used in the present study (1 µM) is expected
to depress mossy fiber transmission maximally
(IC50 = 44 nM) (Kamiya et al. 1996
) and has no detectable effect on type I or type III
metabotropic glutamate receptors. We have confirmed with whole cell
patch-clamp recordings that 1 µM DCG-IV inhibits transmission at
mossy fiber synapses on dentate granule cells (P. Molnár and
J. V. Nadler, unpublished observations), as it does at all other
mossy fiber synapses studied to date (Doherty and Dingledine
1998
; Kamiya et al. 1996
; Maccaferri et
al. 1998
). DCG-IV can also activate N-methyl-D-aspartate receptors, but this action
is significant only when it is applied at higher concentrations
(EC50 = 144 µM) (Wilsch et al.
1994
). Thus our data support the involvement of recurrent mossy
fiber synapses in antidromically evoked epileptiform activity.
The only other excitatory input to the dentate granule cells reported
to be activated by stimulation of the CA3 area is the disynaptic
dentate associational pathway (CA3 pyramidal cells-hilar mossy
cells-granule cells) (Scharfman 1994). Although the
associational pathway may have contributed in some way to
antidromically evoked epileptiform activity, several considerations
suggest that its contribution, if any, was much less than that of the
recurrent mossy fiber pathway. 1) Limbic status epilepticus
kills the majority of hilar mossy cells (Buckmaster and
Jongen-Rêlo 1999
; Sloviter 1987
).
Nevertheless, epileptiform activity was observed only in slices from
rats that had developed status epilepticus. 2) We could not
evoke an EPSC in granule cells by uncaging glutamate within the dentate
hilus in the presence of bicuculline (Molnár and Nadler
1999
). This result suggests that most mossy cells either do not
survive slice preparation under our conditions or are not connected to
granule cells within the plane of our slices. 3) Associational terminals are not immunoreactive for type II metabotropic glutamate receptors (Shigemoto et al. 1997
).
4) Associational terminals do express mGluR4, a type III
metabotropic glutamate receptor (Shigemoto et al. 1997
).
However, L-AP4, an agonist of this receptor, does not significantly
reduce the magnitude of antidromically evoked epileptiform activity
(Okazaki and Nadler 1999
).
Effect of modestly increasing [K+]o
Of the three variables examined in the present study,
[K+]o was clearly of
greatest significance with respect to unmasking the effect of recurrent
excitation. Even a modest elevation of [K+]o (by 1.25 mM) more
than tripled the probability of observing antidromically evoked
epileptiform activity of dentate granule cells in slices with recurrent
mossy fiber growth. Increases in [K+]o of this magnitude
have been observed in response to low-frequency electrical stimulation,
intense natural stimuli, and rhythmic neuronal activity in vivo
(Krnjevi et al. 1982
; Walz and Hertz 1983
). Thus the K+ concentrations we
tested are within the physiological range. They were also below the
concentrations achieved in the dentate gyrus during a seizure
(Lux et al. 1986
; Stringer and Lothman 1989
). It is not clear, however, which stimuli raise
[K+]o to 4.75 or 6 mM in
the dentate gyrus of epileptic brain. Studies of this issue are needed.
Elevated [K+]o can itself
precipitate epileptiform discharge (Korn et al. 1987;
Rutecki et al. 1985
; Traynelis and Dingledine 1988
). In addition, elevated
[K+]o facilitates
epileptiform discharge in hippocampal slices prepared from animal
models of epilepsy (Patrylo and Dudek 1998
;
Stringer and Lothman 1988
). Patrylo and Dudek
(1998)
used hippocampal slices from kainic acid-treated rats to
study the effect of
6 mM
[K+]o in the presence of
recurrent mossy fiber growth. They reported evidence of
hyperexcitability to "hilar" stimulation, ranging from multiple
granule cell population spikes to negative field potential shifts with
superimposed spikes, in 82% of slices from kainic acid-treated rats
that were superfused with 6 mM K+. These abnormal
discharges were shown to depend on the activation of glutamate
synapses. The incidence of hyperexcitability in their study was greater
than that observed in the present study with hippocampal slices from
pilocarpine-treated rats, despite comparable recurrent mossy fiber
growth. Although we cannot rule out the possibility that kainic
acid-induced status epilepticus and pilocarpine-induced status
epilepticus differentially alter granule cell excitability, a more
likely explanation lies in the different sites of stimulation. Stimulation of mossy fibers within the dentate hilus brings more granule cells to threshold than does stimulation in area CA3b, as
evidenced by much larger antidromic population spikes. The more granule
cells are activated, the more recurrent mossy fiber circuitry can
potentially be recruited and the easier it is to detect unmasking by
elevated [K+]o. However,
hilar stimulation is more likely to involve other elements of dentate
gyrus circuitry. Both excitatory mossy cells and inhibitory GABA
neurons may be activated, either synaptically or by DC application.
Hilar stimulation should also activate more effectively the sparse
mossy fiber-granule cell synapses that are normally present in the rat
dentate gyrus.
It should be noted that Timm histochemistry visualizes only mossy fiber boutons and provides no information about connectivity within the recurrent mossy fiber circuit. It is possible that antidromic stimulation failed to evoke much recurrent excitation in some of our experiments because key segments of the circuit were not present in that particular slice. This may account for a few instances in which epileptiform activity was evoked in one slice but not in another slice from the same rat tested under the same experimental conditions.
Kindling reduces the value of
[K+]o at which
stimulation of the Schaffer collateral-commissural projection evokes
epileptiform bursting of CA1 pyramidal cells (Stringer and
Lothman 1988). The concentration of extracellular
K+ required depends critically on
[Ca2+]o; the greatest
incidence of epileptiform activity is observed with a combination of
high K+ and low Ca2+
concentrations. Increased
[K+]o and reduced
[Ca2+]o also promote
spontaneous cellular bursts in the normal dentate gyrus (Pan and
Stringer 1997
). The concentrations of these ions required to
promote cellular bursting are more extreme than those used in the
present study, and there is still no evidence of epileptiform activity
in the field recording. Our results demonstrate that, similar to the
effect of kindling on excitability in area CA1, pilocarpine-induced
status epilepticus reduces the threshold ionic conditions for
epileptiform activity in the dentate gyrus. Antidromic stimulation
evoked epileptiform activity in the presence of 1.8 mM
Ca2+ and 4.75 mM K+. The
recurrent mossy fiber pathway provided the substrate for synchronization of the cellular bursts.
[K+]o regulates granule
cell excitability in several ways. The membrane potential of these
neurons varies linearly with
[K+]o up to ~6 mM
(Pan and Stringer 1997). Raising
[K+]o from 3.5 to 4.75 or
6 mM is expected to depolarize dentate granule cells by 8 and 16 mV,
respectively. In addition, increasing [K+]o can shift
ECl to a more positive value, thus reducing the
magnitude of GABAA receptor-mediated inhibition
(Chamberlin and Dingledine 1988
), depress
K+-dependent hyperpolarizations produced by
activation of postsynaptic receptors (e.g., GABAB
and adenosine A1 receptors), reduce K+-mediated
afterhyperpolarizations (Jensen et al. 1994
), increase the frequency of spontaneous EPSPs (Traub and Dingledine
1990
) and induce glial swelling, thus increasing extracellular
resistance (Lux et al. 1986
; Traynelis and
Dingledine 1989
; Walz 1987
). Further studies are
needed to assess the relative contributions of each factor in the
presence and absence of recurrent excitatory circuitry.
Effect of blocking GABAA receptor-mediated synaptic inhibition
Blocking GABAA receptor-mediated inhibition
never caused the emergence of antidromically-evoked epileptiform
activity, although it did enhance the magnitude of such activity. Thus
the effect of the recurrent mossy fiber pathway was not masked by such
inhibition. This result appears at odds with the study of
Patrylo and Dudek (1998), in which addition of 10 µM
bicuculline to the superfusion medium increased the proportion of
slices in which hilar stimulation evoked epileptiform activity from 27 to 84%. Again this discrepancy may be explained by the difference in
stimulation sites. Stimulation within the dentate hilus would be
expected to activate greater synaptic inhibition than stimulation in
area CA3b, due both to greater driving of inhibitory neurons through
mossy fiber connections and to direct activation of interneurons by the
stimulus current. The recurrent mossy fiber pathway may be more
effectively regulated by GABA inhibition under these conditions; thus
blockade of GABA inhibition would have a more dramatic effect. Hilar
stimulation may also activate more effectively the excitatory
associational-commissural pathway, especially when GABA inhibition is
reduced. Even in slices from previously untreated rats, hilar
stimulation can sometimes evoke multiple population spikes in the
presence of bicuculline (Fournier and Crepel 1984
;
Jackson and Scharfman 1996
; Patrylo and Dudek
1998
). No such response was observed in slices from the control
group when we stimulated the mossy fibers outside the dentate hilus.
Our results suggest that GABA inhibition regulates the recurrent mossy
fiber pathway less strongly than
[K+]o. They do not,
however, exclude the possibility that status epilepticus produced some
deficit in GABA inhibition that allowed epileptiform activity to emerge
when [K+]o was increased.
The existence of such a deficit is largely hypothetical at this time.
Studies of synaptic inhibition in the dentate gyrus of epileptic brain
have yielded mixed results as to the direction of any change
(Okazaki et al. 1999).
Effect of stimulus trains
If mossy fiber synapses on dentate granule cells operate similarly
to mossy fiber synapses on CA3 pyramidal cells (Salin et al.
1996), one would expect to observe marked frequency
facilitation during stimulation at rates of 0.1-1 Hz. This increase in
synaptic strength should then manifest itself as facilitation of
epileptiform activity evoked by similar stimulus trains. Our results
did not confirm this expectation. No significant change in response
magnitude occurred during stimulus trains delivered at a frequency of
0.2 or 1 Hz, and stimulation at a frequency of 10 Hz caused a
progressive and profound loss of the orthodromic component of the
response. Perhaps mossy fiber synapses on dentate granule cells differ
from those on CA3 pyramidal cells in having little frequency
facilitation. There is precedent in this pathway for regulation of
presynaptically mediated plasticity by the postsynaptic cell; long-term
potentiation, which is induced through a presynaptic change, occurs at
mossy fiber synapses on CA3 pyramidal cells but not at mossy fiber
synapses on CA3 interneurons (Maccaferri et al. 1998
).
With respect to both synaptic size (Okazaki et al. 1995
)
and unitary EPSC amplitude (Molnár and Nadler
1999
), mossy fiber synapses on dentate granule cells are
intermediate between mossy fiber synapses on CA3 pyramidal cells and
interneurons. If these synapses do exhibit frequency facilitation, this
process must have been masked by some inhibitory effect of repeated stimulation.
Further research is needed to explain the profound suppression of
epileptiform discharge observed during stimulation at 10 Hz. The
magnitude of this effect appeared to depend on the initial size of the
discharge. One possibility is that glutamate released from the
recurrent mossy fiber boutons inhibited subsequent glutamate release by
activating type II metabotropic receptors. This mechanism is
significant only at high rates of presynaptic activity; transport processes clear extracellular glutamate adequately during low frequency
activity (Scanziani et al. 1997). Other possible
contributory factors include buildup of extracellular adenosine,
activation of Ca2+- dependent
K+ conductance, and synaptic fatigue. Whatever
the cause, our results suggest that transmission in the recurrent mossy
fiber circuit fails at rates of activity similar to that occurring
during the tonic phase of seizures.
Implications for seizure propagation through the dentate gyrus
Recurrent mossy fiber growth appears not to be required for
pilocarpine-treated rats to become epileptic, because suppression of
mossy fiber growth with cycloheximide does not prevent the development
of spontaneous seizures (Longo and Mello 1997). This finding does not exclude a role for recurrent mossy fiber growth in
epileptogenesis, however. Pilocarpine-induced status epilepticus severely damages many forebrain regions (Clifford et al.
1987
; Okazaki et al. 1999
; Turski et al.
1983
), which may result in sprouting of residual excitatory
pathways, loss of synaptic inhibition, and altered expression of many
genes. Thus development of epilepsy in this model probably involves
multiple changes in several regions of the brain. Removal of any single
pro-epileptic consequence of the insult would not by itself be expected
to prevent spontaneous seizures.
Our results reinforce the suggestion of Patrylo and Dudek
(1998) that elevation of
[K+]o increases seizure
susceptibility in the dentate gyrus when a significant recurrent mossy
fiber pathway is present. The dramatic expansion of this pathway could
contribute to episodic spontaneous seizures in both pilocarpine-treated
rats and humans with temporal lobe epilepsy by the following mechanism.
During interictal periods, when
[K+]o is at resting
levels, the recurrent mossy fiber pathway is seldom activated and
dentate granule cells behave normally. That is, they discharge
asynchronously at low frequency. Seizures are presumably triggered by
excessive discharge in the entorhinal cortex that is propagated to the
dentate gyrus through the perforant path. Based on studies in other
regions of the brain,
[K+]o might rise to a
level that would unmask recurrent excitation if the granule cells were
driven by perforant path activity at a rate near the upper limit of
normal (~1 Hz) (Jung and McNaughton 1993
). However,
the entorhinal cortex may discharge abnormally in the epileptic brain.
In both human temporal lobe epilepsy and animal models of this
condition, the entorhinal cortex suffers damage (Du et al. 1993
,
1995
), which is presumably followed by synaptic reorganization
and perhaps other changes. Whatever the mechanism, entorhinal cortical
discharge assumes a prolonged burst-like morphology (Scharfman
et al. 1998
). This abnormally strong excitatory input may not
only serve as a more effective stimulus for granule cell activation but
might easily raise [K+]o
in the dentate gyrus by enough to facilitate reverberating granule cell
excitation. Importantly, our results support a synergistic interaction
between [K+]o and
recurrent mossy fiber growth; the combination of a relatively small
increase in the number of mossy fiber-granule cell synapses with a
relatively small increase in
[K+]o appears sufficient
to facilitate the recruitment of granule cells into synchronous
epileptiform activity. Once the granule cell population has been
engaged, prolonged field bursts appear. The recurrent mossy fiber
pathway becomes nonoperational during the high-frequency activity
associated with these bursts and synchronization then depends entirely
on non-synaptic mechanisms (Pan and Stringer 1996
;
Schweitzer and Williamson 1995
; Schweitzer et al.
1992
). The field bursts then propagate into and through the
hippocampus. This hypothesis does not exclude the involvement of
additional abnormalities that may be present in the dentate gyrus of
epileptic brain, including weakened synaptic inhibition, reduced
calcium buffering, or enhanced glutamate receptor function. To the
extent that recurrent mossy fiber growth reduces seizure threshold,
however, it seems rational to pursue an approach to pharmacotherapy of temporal lobe epilepsy based on depression of activity in that pathway.
![]() |
ACKNOWLEDGMENTS |
---|
We thank Dr. J. O. McNamara for timely assistance and K. Gorham for secretarial help.
This study was supported by National Institute of Neurological Disorders and Stroke Grant NS-17771 (J. V. Nadler) and a Four Schools Research Scholarship (J. L. Hardison).
![]() |
FOOTNOTES |
---|
Address for reprint requests: J. V. Nadler, Dept. of Pharmacology and Cancer Biology, Duke University Medical Center, Durham, NC 27710 (E-mail: nadle002{at}acpub.duke.edu).
Received 27 March 2000; accepted in final form 14 July 2000.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|