| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Neurophysiol (January 1, 2003). 10.1152/jn.00691.2002
Submitted on Submitted 6 June 2002; accepted in final form 11 September 2002
Swammerdam Institute for Life Sciences, University of Amsterdam, 1098 AM Amsterdam, The Netherlands
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Karst, Henk and
Marian Joëls.
Effect of Chronic Stress on Synaptic Currents in Rat Hippocampal
Dentate Gyrus Neurons.
J. Neurophysiol. 89: 625-633, 2003.
We investigated the effect of chronic
stress on synaptic responses of rat dentate granule cells to perforant
path stimulation. Rats were subjected for 3 wk to unpredictable
stressors twice daily or to control handling. One day after the last
stressor, hippocampal slices were prepared and synaptic responses were
determined with whole-cell recording. At that time, adrenal weight was
found to be increased and thymus weight as well as gain in body weight were decreased in the stressed versus control animals, indicative of
corticosterone hypersecretion during the stress period. In slices from
rats with basal corticosteroid levels (at the circadian trough, under
rest), no effect of prior stress exposure was observed on synaptic
responses. However, synaptic responses of dentate granule cells from
chronically stressed and control rats were differently affected by in
vitro activation of glucocorticoid receptors, i.e., 1-4 h after
administration of 100 nM corticosterone for 20 min. Thus the maximal
response to synaptic activation of dentate cells at holding potential
of 
70 mV [when N-methyl-D-aspartate (NMDA)
receptors are blocked by magnesium] was significantly enhanced after
corticosterone administration in chronically stressed but not in
control animals. In accordance, the amplitude of
-amino-3-hydroxy-5-methylisolazole-4-propionic acid (AMPA) but
not of NMDA receptor-mediated currents was increased by corticosterone
in stressed rats, over the entire voltage range. Corticosterone
treatment also decreased the time to peak of AMPA currents, but this
effect did not depend on prior stress exposure. The data indicate that
following chronic stress exposure synaptic excitation of dentate
granule cells may be enhanced when corticosterone levels rise. This
enhanced synaptic flow could contribute to enhanced excitation of
projection areas of the dentate gyrus, most notably the CA3 hippocampal region.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The rat adrenal hormone
corticosterone is secreted in high amounts after acute stress (review
by Dallman et al. 1994
). Corticosterone can enter the
brain and bind to the high-affinity mineralocorticoid receptor (MR) and
to the glucocorticoid receptor (GR) with 10-fold lower affinity
(Reul and de Kloet 1985). Principal neurons in the CA1
area and dentate gyrus (DG) of the hippocampus express high amounts of
both receptor types; CA3 pyramidal neurons contain high amounts of MR
but much lower levels of GR (review by McEwen et al.
1986). Previous studies have shown that differential activation of hippocampal MR and GR affects functional characteristics of CA1 and
DG neurons. It has been proposed that signal transfer is maintained at
a stable level under conditions of predominant MR activation, such as
occurs under rest at the circadian trough; perturbations of hippocampal
activity following stress are normalized by activation of the GR
(review by Joëls 1997). These effects of the
hormone add to its adaptational role following stress exposure.
Chronic exposure of animals to stress though is often associated with
maladaptation. It is well documented that exposure of animals to
elevated corticosteroid levels for several weeks results in atrophy of
CA3 pyramidal cells (Magarinos and McEwen 1995a,b; Sapolsky et al. 1985; Watanabe et al.
1992b). Based on pharmacological studies, it was concluded that
excitatory amino acid mediated transmission, in particular through
N-methyl-D-aspartate (NMDA) receptors, plays a
role in the stress-induced atrophy. Thus pretreatment of animals with
the amino acid release compound phenytoin or the competitive NMDA
receptor blocker CPG 43487 both prevented atrophy (Magarinos and
McEwen 1995a; Watanabe et al. 1992a).
Presynaptic as well as postsynaptic elements of excitatory amino acid
transmission could be involved. With respect to the latter, short-term
manipulation of corticosterone level rather than chronic stress
appeared to increase NMDA receptor binding and subunit expression
(Watanabe et al. 1995; Weiland et al.
1997). More recently, however, RT-PCR in hippocampal tissue
revealed that chronic stress increases expression of the GluR1
(-amino-3-hydroxy-5-methylisolazole-4-propionic acid; AMPA) subunit
while the NMDA-R1 subunit expression was unaffected (Schwendt
and Jezova 2000). With respect to the presynaptic
glutamate-mediated input, bilateral lesion of the entorhinal cortex,
which projects to the CA3 neurons directly but also indirectly via the
dentate gyrus (Steward 1976), fully prevented dendritic
atrophy in the CA3 region after chronic stress (Sunanda et al.
1997). Also, mossy fiber terminals from DG cells exhibited
changes after chronic stress, showing more densely packed vesicle
clusters localized in the vicinity of active zones (Magarinos et
al. 1997). The latter suggests that DG cell function and output
to CA3 pyramidal neurons may alter during chronic stress exposure.
We here investigated the effect of chronic stress on glutamate-mediated
entorhinal input to DG cells, which in turn provide a major
glutamatergic input to the CA3 area via mossy fibers (Crawford and Connor 1973). Rats were subjected to chronic
unpredictable stress or control handling for 3 wk. This paradigm was
earlier shown to be associated with apparent corticosterone
hypersecretion during the stress period (Herman et al.,
1995; Paskitti et al., 2000) and to result in
CA3 dendritic atrophy (Magarinos and McEwen 1995b).
Hippocampal slices were prepared 1 day after the last stressor or
control treatment. At that time, gain in body weight during the stress
period, basal corticosterone level, and weight of the adrenals and
thymus were determined. Apart from basal cell characteristics (input
resistance and membrane capacitance), the responsiveness of granule
cells to perforant path stimulation was examined with whole-cell
patch-clamp recording at a holding potential (70mV) close to the
resting membrane potential. In addition, the conductance, voltage
dependency, and kinetic properties of both the NMDA and AMPA
receptor-mediated components of synaptic responses were investigated.
Since chronic stress was reported to either attenuate or sensitize
central responses to subsequent GR activation, depending on the
stressor used or parameter tested (Lanfumey et al. 1999;
Nisenbaum et al., 1991), we here also tested the
efficacy of GR activation in dentate neurons, in chronically stressed
and control animals. To this end, all animals entered the experiment in
the morning when they display basal plasma corticosterone levels (i.e.,
with predominant MR activation) (Reul and de Kloet 1985); responses under these conditions were compared with
those recorded 1-4 h after GRs activation, induced in vitro by a
20-min perfusion with 100 nM corticosterone.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Stress paradigm
Young adult male Wistar rats (n = 23, Harlan) of ±150 g were housed two (or, in one case, 3) in a cage, with a light/dark cycle of 12 h (lights on at 0800). Food and water were given ad libitum. At the start of the stress exposure, rats were randomly assigned to a group receiving chronic unpredictable stress (n = 11) or to the control group (n = 12). The experiments were carried out with permission of the local Animal Experiment Committee (DED protocols 79 and 80). Seven different stressors were given randomly during a period of 21 days, twice daily, i.e., in the early morning and in the afternoon: 1) The rats were forced to swim in cold water (10-15°C) for 5 min 2) or in water of 25-30°C for 30 min; 3) they were immobilized at room temperature in a restrainer tube for 1 h or 4) immobilized at 4°C for 1 h; 5) shaken for 1 h on a horizontal shaker 30 cpm; 6) crowded for 24 h (4-6 per cage); or 7) isolated for 24 h. Control rats were handled daily. During this period of 21 days, the stressed and control rats were weighed every morning.
Rats entered the electrophysiological experiment 1 day after the last stressor and were decapitated around 0930. Trunk blood was collected to determine plasma corticosterone levels with a RIA. The adrenals and thymus were removed and weighed. The brain was removed from the skull and stored for 3 min in ice-cold artificial cerebrospinal fluid (ACSF) containing low calcium and high magnesium of the following composition (in mM): 124 NaCl, 3.5 KCl, 1.25 NaH2PO4, 5.0 MgSO4, 0.2 CaCl, 25 NaHCO3, and 10 glucose; pH 7.4 gassed with 95% O2-5% CO2. The osmolarity (300 mOsm) of this ACSF was adjusted with a Wescor 5100C vapor pressure osmometer. Horizontal slices of the brain were made with a vibratome (Campden Instruments, Sileby, UK). The slices containing the hippocampus were stored in continuously gassed (95% O2-5% CO2) ACSF containing (in mM): 124 NaCl, 3.5 KCl, 1.25 NaH2PO4, 1.5 MgSO4, 2 CaCl, 25 NaHCO3, and 10 glucose; pH 7.4, at room temperature. After a delay of 1 h, some of the slices were treated with 100 nM corticosterone (Sigma) for 20 min in ACSF of 32°C. After this treatment the slices were moved to a storage bath with normal ACSF at room temperature. The same procedure was carried out for the vehicle-treated controls. Before moving the slice to the recording chamber, an incision was made between the DG and the CA3 area.
Electrophysiology
One slice at a time was placed in a recording chamber mounted on an upright microscope (Nikon Optiphot-2). Slices were continuously perfused with ACSF (32°C, 2-3 ml/s) and kept fully submerged. Bicuculline methiodide (20 µM, Sigma) was added to the buffer to prevent GABA-mediated inhibition that could be activated by pathway stimulation.
The surface of the suprapyramidal blade of the DG was cleaned with a
cleaning pipette. Patch-clamp electrodes for recording (1.5 mm OD,
borosilicate glass; impedance approximately 3-4 M
) were pulled on a
Sutter micropipette puller and placed above the slice. The
intracellular pipette solution contained (in mM) 120 Cs methane
sulfonate, 17.5 CsCl, 10 Hepes, 2 MgATP, 0.1 NaGTP, 5 BAPTA, and 10 QX-314; pH 7.4, adjusted with CsOH. The osmolarity of the pipette fluid
was 295 mOsm. BAPTA was obtained from Molecular Probes (Leiden, The
Netherlands); the sodium channel blocker QX-314 was from
Alomone (Jerusalem, Israel). Under visual control (40× objective and
10× ocular magnification) the electrode was directed toward a granule
neuron using positive pressure. Once a patch electrode was sealed on
the cell (approximately 1 G) the membrane patch under the electrode
was ruptured and the cell was held at a holding potential of 70 mV.
Signals were fed into an Axopatch 200B amplifier (Axon Instruments).
Data acquisition and analysis was performed with an Atari computer with
in-house-developed software (courtesy T. Juta and W. Wadman). Series
resistance was compensated to ±70%.
A bipolar stainless steel stimulus electrode (60 µm diam, insulated
except for the tip) was placed in the perforant path (Stienstra et al. 1998). Biphasic stimuli (250 µs) were applied through
a Neurolog stimulus isolator (NL 800) driven by a homemade software program. Input-output curves of excitatory postsynaptic currents (EPSCs) evoked in DG neurons were made at holding potential by increasing stimulus intensities from 7 to 600 µA (see example in Fig.
2), given once every 10 s. EPSCs were recorded with a sampling
frequency of 1 or 10 kHz, depending on whether the focus was on slow or
fast (e.g., risetime) components, respectively; signals were stored and
off-line corrected for leak. Fast components of the synaptic currents
(e.g., rise time) were studied separately with a 10-kHz sampling
frequency. Input-output curves were fit with a Boltzmann equation
R(i) = Rmax/[1 + exp{(i iH)/IC}], in which Rmax is the maximal evoked
current, iH the half-maximal stimulus
intensity, and IC proportional to the
slope. Based on this curve the half-maximal stimulus intensity was
determined. This intensity was used to evoke EPSCs at holding
potentials between 90 and +40 mV, increasing voltage in subsequent
steps by 10 mV, using intervals of 10 s. During approximately half
of the recordings D-()-2-amino-5-phosphonopentanoic acid
(APV, Sigma; 50µM) was perfused to block the NMDA receptor; in a
limited number of neurons, CNQX (10 µM, Sigma) was applied to block
AMPA receptors.
Statistics
Statiscal analysis of the input-output curves, the current-voltage (I-V) relationship, and the gain in weight were tested with ANOVA for repeated measurements (MANOVA). Other data sets were tested with ANOVA and a nonparametric Mann-Whitney U test (P < 0.05), making the less powerful assumption that the distribution of the data was not Gaussian.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Neuroendocrine parameters
Several standard procedures were carried out to examine the effectiveness of the chronic stress procedure. One of the most indicative parameters is the gain in weight of the adrenals. As shown in Fig. 1A, the increase in absolute weight of the adrenals and the weight corrected for body weight (per 100 g) was significantly increased in chronically stressed compared with control animals. Another index of chronic stress is the weight of the thymus. Following chronic stress, the absolute thymus weight (549 ± 22 mg) was significantly reduced compared with the control group (658 ± 38 mg), although this difference did not attain statistical significance when corrected for body weight (Fig. 1). Gain in body weight (relative to the body weight when animals started to be exposed to stress or handling) was attenuated in stressed compared with control rats (Fig. 1). The plasma corticosterone level of the stressed group was somewhat elevated [control: 2.2 ± 1.3 µg/100 ml; chronically stressed: 4.7 ± 2.5 µg/100 ml, based on observations in part of the control (n = 4) and stressed (n = 7) animals]. These data support that rats subjected to unpredictable stress indeed experienced prolonged corticosterone hypersecretion prior to the electrophysiological experiments.
|
Effect of chronic stress and in vitro corticosterone on basal properties and input-output curves
In total, 72 DG granule cells were recorded with whole-cell recording. Input resistance (Rin) was comparable for DG cells of chronically stressed and control rats, both before and after corticosterone treatment. Interestingly, corticosterone treatment increased Rin by approximately 30%. The corticosterone-induced increase of Rin was significant in control rats though not in chronically stressed rats (Table 1). Capacitance of the cells was unaffected by chronic stress and/or corticosterone treatment (Table 1).
|
Excitatory synaptic currents in DG granule cells (recorded in the
presence of bicuculline) involve activation of NMDA and AMPA receptors,
while the role of kainate receptors is negligible (Behr et al.
2001). At a holding potential of 70 mV NMDA receptors are
blocked by magnesium (see following text), so that synaptically evoked
currents are supposedly mediated by AMPA receptors. Figure 2A shows an example of AMPA
receptor-mediated currents evoked in DG granule cells at a holding
potential of 70 mV by perforant path stimulation, using increasing
stimulus intensities. Maximal amplitude was reached with stimulus
intensities higher than 400 µA. With the Boltzmann equation
R(i) = Rmax/[1 + exp{(i iH)/IC}], the maximal evoked current (Rmax), the
half-maximal stimulus intensity (iH),
and a factor IC proportional to the
slope factor were calculated. In Fig. 2A, an example of a
Boltzmann plot is shown.
|
Averaged input-output curves for all the experimental groups are shown in Fig. 2B. No effect of chronic stress was seen on the input-output relationship under basal conditions when corticosterone levels are low. In vitro treatment of the slices of control rats with 100 nM corticosterone had a marginal but nonsignificant [F(1,36) = 2.56; P = 0.119] effect on the input-output curves. However, a major and significant change in the input-output curve [MANOVA: F(3,50) = 4.588; P = 0.0065] was seen when slices of chronically stressed rats had been treated with 100 nM corticosterone for 20 min, 1-4 h prior to recording. The main change was a significant increase in the maximal amplitude of synaptically evoked EPSCs (Fig. 2C). No change was observed with respect to the iH (Fig. 2D) or IC (Fig. 2E). Subsequent testing of the cells was performed at half-maximal stimulus intensity.
NMDA receptor- and AMPA receptor-mediated components of the EPSCs
Depending on the holding potential, EPSCs evoked in granule cells
of the DG by stimulation of the perforant path contained one or two
components: a fast component was seen at all holding potentials while a
slow component appeared at holding potentials more depolarized than
40 mV (Fig. 3A). The slow
currents disappeared in the presence of the APV, indicating that they
were mediated by the NMDA receptor (Fig. 3B). The remaining
fast current was fully blocked when CNQX was added in addition to APV.
The fast current therefore represents the AMPA receptor-mediated
component. Since no currents were evoked in the presence of the two
blocking agents, we conclude that only NMDA and AMPA receptor-mediated currents take part in the synaptically evoked currents.
|
In a first series of experiments, we analyzed fast peak currents and
slow synaptic currents 50 ms after stimulation (Fig 3A, a
and b, respectively). The slow current displayed a nonlinear I-V relation (Fig. 3F, b), with no apparent
conductance up to 40 mV, increasing conductance between 40 and 20
mV and linear characteristics at levels more depolarized than 20 mV.
Two observations indicate that the slow currents measured at b are
mediated by NMDA receptors only. First, the slow currents were
abolished in the presence of the NMDA receptor blocker APV (Fig
3B, e). Second, by subtracting AMPA currents (recorded in
the presence of APV) from the total synaptic currents, presumably NMDA
receptor-mediated currents could be studied in isolation (Fig.
3C, d). The I-V relation of signal d was
indistinguishable from that of b (Fig. 3F). We conclude that
currents measured 50 s after stimulation are a reliable indicator
of NMDA receptor-mediated synaptic activity.
Interpretation of the peak total synaptic current was less
straightforward. Thus the I-V relationship of the evoked
currents (Fig. 3E, a) was nonlinear: A clear deviation was
seen, starting from a holding potential of 40 mV. The nonlinearity
was clearly caused by activation of NMDA receptors since a linear
I-V relationship was obtained for the peak amplitudes in the
presence of the NMDA receptor antagonist APV (Fig. 3E, c).
In a second series of experiments we therefore studied AMPA
receptor-mediated currents in isolation, during administration of APV.
Effect of chronic stress and in vitro corticosterone on I-V relation
EPSCs were evoked with half-maximal stimulus intensities at
holding potentials varying from 90 to +40 mV. In the first series of
experiments, currents were determined in the absence of any blockers
(Fig. 4A). The averaged
I-V plot of the peak current is shown in Fig.
5A, the I-V plot of
the slow, NMDA receptor-mediated currents is shown in Fig.
5B. Under basal corticosterone conditions, no differences
were observed between the I-V relations of synaptic responses evoked in the DG neurons of chronically stressed animals and
control rats (typical examples in Fig. 4A). However, 1-4 h after a brief (20 min) in vitro application of 100 nM corticosterone, peak currents over a voltage range of 90 to 0 mV were significantly enhanced in chronically stressed but not control animals (Figs. 4A and 5A). No differences between experimental
groups were observed for the NMDA receptor-mediated currents [Fig.
5B; MANOVA for currents between 90 and 0 mV:
F(3,31) = 0.59; P = 0.63]. This
suggested that AMPA rather than NMDA receptor components were affected
by corticosterone treatment in tissue from chronically stressed rats.
|
|
To address this specifically, synaptic currents were studied in a
second series of experiments while perfusing APV, yielding AMPA
receptor-mediated currents in isolation (typical traces in Fig.
4B). As shown in Fig. 5C, the AMPA
receptor-mediated currents showed very similar sensitivity to
corticosterone and stress as the peak current in the absence of APV.
Thus no changes by chronic stress were seen under basal conditions.
While corticosterone did not significantly alter the I-V
relation of AMPA receptor-mediated responses in DG cells of control
rats, marked enhancement of the current amplitude was seen after
corticosterone treatment in chronically stressed rats, in the voltage
range of 90 to 0 mV. We therefore conclude that corticosterone
specifically enhances AMPA receptor-mediated synaptic responses of DG
cells in chronically stressed but not in control rats.
Effect of chronic stress and in vitro corticosterone treatment on rise time and decay of synaptic currents
We looked at two aspects of the kinetic properties of AMPA
receptor-mediated currents. First, the effect of corticosterone treatment on the rise time (interval between 10-90% of the peak amplitude) was studied (Fig.
6A). Second, we examined the
effect on the decay time constant (
) of the AMPA currents (Fig.
6B).
|
Typically, granule cells in untreated slices from control rats
displayed a rise time of 4.5 ms (Fig. 6C). This is
comparable to the values published by others in various
hippocampus cell types (Obokata et al. 1997;
Rammes et al. 1999; Xie et al.
1997). The rise time is rather slow compared with the rise
times recorded from dendrites (Benke et al. 1998), which
can be explained by dendritic filtering of the currents (Major
et al. 1994). Corticosterone treatment consistently shortened
the rise time of the AMPA current. The histogram of Fig. 6C
shows that the mean rise time was reduced in corticosterone-treated
slices of both the control and the chronically stressed rats. In
stressed as well as nonstressed groups, corticosterone treatment
reduced the rise time by approximately 20%.
Decay time constants were obtained by fitting the current traces after the peak with a single exponential. In nearly all cells this yielded a good fit (r > 0.90). In contrast to the rise time, no group differences were observed with respect to the decay time (Fig. 6D). We conclude that corticosterone decreases the rise time of the AMPA receptor-mediated synaptic response, without affecting the decay time.
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
In this study we investigated the effect of chronic stress on DG
granule cell responses to perforant path stimulation. The model of
chronic unpredictable stress was selected since this paradigm induces
dendritic atrophy in CA3 pyramidal neurons and shows less habituation
to the stressor than, e.g., found after 3 wk of daily restraint stress
(Magarinos et al. 1995b). In agreement with the presumed
chronic hypercorticism of this model, animals included in the present
study displayed increased adrenal weight, somewhat decreased thymus
weight (only significant if not corrected for body weight), and
attenuated body weight gain, compared with the control animals, which
were only handled. The changes were relatively small, as was also
reported earlier for this model (Cullinan and Wolfe
2000), indicating that the animals probably did not experience
severe stress.
In thus stressed rats, the conductance, voltage dependency, and kinetic properties of NMDA receptor and AMPA receptor-mediated synaptic currents induced by perforant path stimulation in DG granule cells were not altered under basal conditions, i.e., when corticosterone levels are quite low. Only when GRs were substantially activated by in vitro administration of a high dose of corticosterone did DG granule cells respond in a different way to synaptic input. Thus the amplitude of AMPA but not NMDA receptor-mediated synaptic currents of granule cells was markedly enhanced after GR activation in stressed but not in control rats. It should be noted that we can presently not fully exclude that we missed an effect on the NMDA receptor-mediated currents: NMDA receptor-mediated currents were measured at 50 ms after stimulation, i.e., during the period when the channels are closing. No apparent shifts in voltage dependency could be discerned. A consistently faster rise time of the AMPA receptor-mediated synaptic current was observed after GR activation, but this effect was seen in stressed as well as in control rats and therefore apparently is not linked to chronic stress.
Although clear changes were thus observed in association with chronic
stress and corticosterone treatment, these effects need to be
interpreted with some caution since voltage control over DG granule
cells in situ is incomplete, although less so than in CA1 or CA3
pyramidal neurons (Jaffe and Carnevale 1999). This is
relevant since lateral perforant path input mostly impinges on distal
dendrites, while we here recorded synaptic currents in the soma. If the
cable properties of the dendrites are comparable between experimental
groups, one can assume that all groups are equally affected by the
incomplete voltage control and filtering of signals (due to the cable
properties), so that quantitative rather than qualitative aspects of
group differences may be influenced. However, we observed in dentate
cells that (contrary to CA1 neurons) the input resistance is increased
after corticosterone treatment in the nonstressed controls. It is
therefore necessary to consider whether the observed effects of chronic
stress and/or corticosterone treatment can be merely explained by
changed cable properties and incomplete voltage control. Several
observations argue against this possibility. First, although reversal
potentials of AMPA receptor and NMDA receptor-mediated synaptic
currents were rather depolarized-which is expected in case of
incomplete voltage control and was also seen in earlier studies using
comparable recording conditions (e.g., Otmakhova et al.
2002)-there was no clear difference between the four
experimental groups. Second, an increased input resistance such as seen
after corticosterone treatment in nonstressed animals is expected to
result in a decreased amplitude and slowing down of the rise time as
well as the decay of synaptic currents recorded in the soma. Contrary
to this prediction, synaptic currents were either not altered
(nonstressed group) or even increased in amplitude (stressed group)
after corticosterone treatment. Also, corticosterone caused a faster
rather than slower rise time in the stressed as well as the control
group, while it did not affect the decay of synaptic currents. We
tentatively conclude that the presently observed differences in
synaptic currents between the experimental groups cannot be explained
(and if anything are masked) by the changes in basic cell properties in
combination with incomplete voltage control.
Interestingly, very little effect of prior chronic stress exposure was observed when synaptic currents were recorded in slices obtained from animals that at the start of the experiment were at the circadian trough of corticosterone release and at rest. Plasma corticosterone level of the stressed group was somewhat elevated, but the influence of this rise was apparently limited, probably since, in both the stressed and the control animals, extensive GR activation can be ruled out. The lack of functional differences also argues against major changes in MR function and slow stress-induced adaptational changes in basal cellular and network properties, e.g., in morphology and innervation.
Prior stress exposure, however, markedly altered the response of
synaptic currents to in vitro perfused corticosterone. Thus, in control
animals, GR activation had little effect on the amplitude of the
synaptic currents. A lack of effect of GR (in addition to MR)
activation on synaptic responses in the DG of nonstressed rats was
earlier also observed in extracellular recording studies investigating
the field response evoked by perforant path stimulation (Stienstra et al. 2000). Yet, a clear enhancement of
synaptic currents by GR was seen in chronically stressed rats, pointing to increased GR activity after chronic stress. Enhanced functional responses to GR activation in chronically stressed compared with control animals was earlier also reported, most notably for the release
of hippocampal norepinephrine (Nisenbaum et al. 1991). Enhanced GR activity could be due to an increased capacity of the
receptor but also, for instance, to enhanced efficacy of nuclear translocation, DNA binding, or transcriptional activity. In an earlier
study using the same paradigm for chronic stress, no differences in MR
and GR mRNA or hnRNA levels were observed in the DG (Paskitti et
al. 2000). If this is translated to the protein level, it seems unlikely that increased GR capacity underlies the observed phenomena. Recently, it was reported that chronic stress prolongs GR interaction with the DNA (Kitchener et al. 2001), which could
enhance GR functional activity. Clearly, extensive studies addressing
transcriptional regulation after chronic stress, including the role of
all proteins involved (not only GR but also, e.g., coactivators and
repressors), will be necessary to resolve the molecular mechanism
underlying the here observed enhanced excitatory responses after GR activation.
In theory, the enhanced effect of GR on synaptic responses in
chronically stressed rats could involve pre- or postsynaptic targets
determining synaptic responsiveness. If GR-mediated actions change
entorhinal innervation of the DG and/or glutamate release this would be
expected to affect NMDA and AMPA receptor-mediated responses to an
equal extent. The present findings do not strongly point in that
direction. Postsynaptically, GR effects could involve altered
functionality of AMPA receptors. Corticosterone treatment was indeed
found to change AMPA receptor binding (Clark and Cotman 1992; Watanabe et al. 1995), while modulation of
NMDA receptor binding and subunit expression still is equivocal
(Watanabe et al. 1995; Weiland et al.
1997). Expression of the Glu-R1 AMPA receptor subunit was also
significantly increased in hippocampal tissue after chronic stress,
while the NMDA-R1 subunit was not changed, as recently demonstrated
with competitive RT-PCR (Schwendt and Jezova 2000),
although not with in situ hybridization (Watanabe et al.
1995). Functionality of the AMPA receptor, however, could also
be altered by posttranslational modification following GR activation,
as was earlier proposed for GR effects on other membrane properties
(Karten et al. 1999). A distinct possibility is that GR
activation leads to recruitment of existing AMPA receptors, similar to
that assumed in the case of long-term potentiation (Shi et al.
1999).
The functional consequence of increased granule cell responsiveness to
entorhinal input after chronic stress can be considerable: It is likely
to enhance the excitation of these cells at least during part of the
day, i.e., after activation of GR by an acute stressor or during the
circadian peak of corticosterone release. If not counteracted by local
inhibitory processes, such altered DG granule cell firing properties
could possibly contribute to the earlier described changes in mossy
fiber terminals (Magarinos et al. 1997) and to the
presumed increased excitation of CA3 pyramidal neurons during stress
(Watanabe et al. 1995b). It may also at least partly
explain why lesioning of the entorhinal cortex protects against
stress-induced dendritic atrophy in the CA3 region (Sunanda et
al. 1997). Importantly, the present study for the first time shows that chronic stress results in functional changes in hippocampal signal transfer, not only in the CA3 region but also in the DG, which
provides a major afferent pathway to the CA3 area.
| |
FOOTNOTES |
|---|
Address for reprint requests: H. Karst, Section Neurobiology, SILS; University of Amsterdam, Kruislaan 320, 1098 SM Amsterdam, The Netherlands.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
This article has been cited by other articles:
|
|
 |
|
  Z. Pu, H. J. Krugers, and M. Joels Corticosterone time-dependently modulates {beta}-adrenergic effects on long-term potentiation in the hippocampal dentate gyrus Learn. Mem., May 3, 2007; 14(5): 359 - 367. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Karst and M. Joels Corticosterone Slowly Enhances Miniature Excitatory Postsynaptic Current Amplitude in Mice CA1 Hippocampal Cells J Neurophysiol, November 1, 2005; 94(5): 3479 - 3486. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Quintanar-Stephano, R. Chavira-Ramirez, K. Kovacs, and I. Berczi Neurointermediate pituitary lobectomy decreases the incidence and severity of experimental autoimmune encephalomyelitis in Lewis rats J. Endocrinol., January 1, 2005; 184(1): 51 - 58. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
