| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 87 No. 1 January 2002, pp. 157-165
Copyright ©2002 by the American Physiological Society
Ethel Percy Andrus Gerontology Center, USC Program in Neuroscience, University of Southern California, Los Angeles, California 90089-0191
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Akopian, Garnik and John P. Walsh. Corticostriatal Paired-Pulse Potentiation Produced by Voltage-Dependent Activation of NMDA Receptors and L-Type Ca2+ Channels. J. Neurophysiol. 87: 157-165, 2002. AMPA and N-methyl-D-aspartate (NMDA) receptor-mediated synaptic responses expressed differential paired-pulse plasticity when examined in the same cell using intracellular or whole cell voltage-clamp recordings. Electrical stimulation of corticostriatal afferents in brain slices bathed in artificial cerebrospinal fluid containing bicuculline produces excitatory postsynaptic potentials and excitatory postsynaptic currents (EPSCs) mediated primarily by AMPA receptors. Cell-to-cell variation existed in AMPA receptor paired-pulse plasticity, but within-cell plasticity was stable over a range of stimulation intensities. Addition of 6-cyano-7-nitroquinoxalene-2,3-dione blocked most of the synaptic response leaving behind a small AP-5-sensitive component. Increasing the stimulation intensity produced large, long-lasting NMDA receptor-mediated responses. In contrast to AMPA receptor-mediated responses, NMDA receptor responses consistently showed an increase in paired-pulse potentiation with increasing stimulation intensity. This relationship was restricted to interstimulus intervals shorter than 100 ms. Paired-pulse potentiation of NMDA receptor responses was voltage-dependent and reduced by removal of extracellular Mg2+. Block of postsynaptic L-type Ca2+ channels with nifedipine produced a voltage-dependent reduction of NMDA receptor excitatory postsynaptic currents (EPSCs) and a voltage-dependent reduction of NMDA receptor paired-pulse potentiation. These data indicate depolarization during the first NMDA receptor response causes facilitation of the second by removing voltage-dependent block of NMDA receptors by Mg2+ and by activating voltage-dependent Ca2+ channels.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Paired-pulse
plasticity is traditionally viewed as an activity-dependent change in
the behavior of presynaptic terminals. The number of vesicles released
in response to the second of two closely timed action potentials is a
function of vesicle availability and the level of
Ca2+ reached within the terminal (Charlton
et al. 1982
; Katz and Miledi 1968;
Swandulla et al. 1991; Zucker 1999).
Residual Ca2+ from prior activity increases
neurotransmitter release by facilitating the fusion of vesicles at
active zones and influencing Ca2+-mediated
reactions associated with priming the terminal for release (Charlton et al. 1982; Katz and Miledi
1968; Llinás et al. 1991; Swandulla et al. 1991; Zucker 1999).
Simultaneously, synaptic vesicles can become temporarily depleted with
the arrival of closely timed action potentials (Applegate and
Landfield 1988; Charlton et al. 1982;
Swandulla et al. 1991; Zucker 1999). The
combined effect of Ca2+-mediated facilitation and
vesicle depletion determines the amount of neurotransmitter release.
Consistent with this mechanism, lowering extracellular
Ca2+ in striatal slices reduces release from
corticostriatal terminals and increases paired-pulse potentiation
(Choi and Lovinger 1997a; Mori et al.
1994; Ou et al. 1997).
The presynaptic expression of paired-pulse plasticity is often used as
an analytic tool for interpreting changes in behavior of presynaptic
terminals associated with long-term plasticity (Akopian et al.
2000; Kuhnt and Voronin 1994; Liao et al.
1992; Malinow and Tsien 1990; Wang and
Kelly 1997). The utility of this measure has come into
question, however (Kuhnt and Voronin 1994; Liao
et al. 1992). Postsynaptic factors like the addition or removal of AMPA receptors can influence measures of short- and long-term plasticity (Selig et al. 1995). Changes in the percent
representation of AMPA and 2-amino-5-phosphonovaleric acid (NMDA)
receptors can also influence paired-pulse plasticity due to differences
in their voltage dependence of activation. Previous work in the
hippocampus has shown release of glutamate from excitatory terminals
can produce differential paired-pulse plasticity at AMPA and NMDA
receptors (Clark et al. 1994). It has also been shown in
the hippocampus that separate excitatory pathways innervating the same
neuron can express differential plasticity depending on the degree of NMDA and AMPA receptor activation (Christie and Abraham
1994; Zalutsky and Nicoll 1990) The present
study examined differences between AMPA and NMDA receptor-mediated
plasticity produced by paired activation of corticostriatal afferents
and provides evidence for postsynaptic, voltage-dependent changes in
synaptic strength occurring when NMDA receptors are activated.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Preparation
Experiments were performed on 4- to 6-mo-old (150-300 g) male Fischer 344 rats bred by Harlan Laboratories for intracellular analysis and on 11- to 18-day-old Fischer 344 rat pups for whole cell analysis. Rats were anesthetized with halothane vapors and decapitated immediately in accordance with protocols approved by USC's Animal Use and Care Committee. Their brains were removed and placed in cooled (1-4°C), oxygenated artificial cerebrospinal fluid (ACSF; concentrations in mM were 124 NaCl, 1.3 MgSO4, 3.0 KCl, 1.5 NaH2PO4, 20 NaHCO3, 2.4 CaCl2, and 10.0 glucose, equilibrated with a 95% O2-5% CO2 mixture to obtain a pH value of 7.3-7.4).
Hemi-coronal striatal slices were cut at a thickness of 450 µm with a Camden vibroslicer (WPI). The slices were immediately placed in a oxygenated ACSF solution containing 30 µM bicuculline methiodide (BIC; Sigma), and they were slowly brought to room temperature (23°C). BIC was used to block gamma-amino butyric acid-A (GABAA) receptor-mediated inhibition in an attempt isolate excitatory synaptic events. Slices remained in this solution for at least 1 h prior to and throughout all recording sessions. Single slices were transferred to the recording chamber and bathed continuously with the oxygenated BIC-ACSF solution. Intracellular records were obtained using a gas:liquid interface slice and whole cell recordings were obtained from submerged slices. In some experiments brain slices were perfused with Mg2+-free ACSF replaced with equal molar Ca2+. AMPA receptor-mediated responses were blocked in some experiments with 10 µM CNQX (Sigma).
Recording and extracellular stimulation
Bipolar insulated tungsten wire (50 µm diam) stimulating electrodes were used for delivering paired extracellular stimuli during intracellular recording experiments and a monopolar stimulating electrode consisting of a glass micropippette filled with saline (tip diameter, 5 µM). Stimulating electrodes were placed at the border between the striatum and the overlying corpus callosum. Intracellular records were obtained at a 1-mm distance from the extracellular electrodes, and whole cell records were obtained at a distance of ~100 µM. Test synaptic responses were delivered as pairs separated by an interstimulus interval (ISI) of 50 ms with a constant current stimulus at a pulse duration <0.2 ms. The threshold intensity for evoking a synaptic response was determined for each cell under each pharmacological condition. Once the threshold was determined, an input/output relationship was determined by increasing the stimulation intensity in 10- or 20-µA steps until an action potential could be produced by activation of corticostriatal afferents.
Intracellular records were obtained with glass microelectrodes pulled
by a Flaming-Brown P-87 pipette puller (Sutter Instruments). Electrodes
filled with 2 M potassium acetate had resistance values ranging from
100 to 120 M
. Intracellular signals were amplified with an Axoclamp
2A (Axon Instruments) amplifier, digitized with a LABMASTER interface
and stored on disk using pCLAMP software (Axon Instruments). Cells had
to maintain certain electrophysiological criteria throughout the
experiment to be included in the study. These criteria included resting
membrane potential of at least 
75 mV, input resistance >20 M, and
the ability to generate action potentials overshooting 0 mV.
Whole cell records were obtained with 3-6 M electrodes pulled by a
Flaming-Brown P-87 pipette puller (Sutter Instruments). Patch
electrodes were filled with an internal solution consisting of (in mM)
125 Cs+-gluconate, 2.0 MgCl2, 0.5 EGTA, 10 HEPES, 10 TEA-Cl, 3.0 QX-314, and 3.0 ATP-Na, pH adjusted to 7.2 with CsOH. The extracellular recording solution was identical to that used for intracellular recording current-clamp experiments. Cells were visualized using a
Zeiss fixed-stage microscope using water-immersion lenses (×40). Patch
pipettes were positioned with an MX-530 hydraulic micromanipulator (Soma Scientific). The recordings reported in this study were amplified
with an Axopatch-1B amplifier (Axon Instruments). Membrane voltage and
current were digitized with a TL-1 Labmaster computer interface (Axon
Instruments). Experiments were controlled and analyzed using the pClamp
6.0 software package (Axon Instruments). Unless stated otherwise, cells
were maintained at a holding potential of 60 mV when examining EPSCs
and their paired-pulse plasticity. This holding potential was chosen
because it was a stable potential where the voltage clamp did not
impose additional holding current. We found, as have others, that
striatal neurons become unstable when held at hyperpolarized holding
potentials of 80 or more (Mori et al. 1994). In
addition, this holding potential is close to the negative slope
resistance region of the NMDA receptor-mediated response and thus
provided the best opportunity for having the first response of the pair
influence the second response during voltage-clamp analysis of
paired-pulse plasticity. To reduce biasing inherent to response
fluctuation, three 50-ms paired responses were evoked (20-s sampling
interval) and averaged for both AMPA/kainate and NMDA receptor-mediated
synaptic responses during the initial comparison of synaptic responses
mediated by these separate classes of excitatory amino acid receptors
(Fig. 2). Single paired-responses were evoked, however, at each holding
potential during the analysis of response voltage dependency (Figs. 3
and 4). Data analysis and plotting was performed using Microsoft Excel
and CorelDraw. All results are reported as means ± SE.
Significance was tested using paired, two-tailed Student's
t-tests because treatment comparisons were performed in the
same cell.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Input/output relationship for AMPA and NMDA receptor-mediated paired-pulse plasticity
AMPA/kainate receptor and then NMDA receptor synaptic responses
were tested sequentially in each cell. Intracellular records in adult
(6 mo) and whole cell recordings in juvenile (p11-p18) striatal
neurons confirmed corticostriatal afferents produce predominantly AMPA
receptor mediated excitatory postsynaptic potentials (EPSPs) or EPSCs
at the high resting membrane potentials characteristically found in
vitro (Cherubini et al. 1988; Walsh et al.
1989). Increasing the stimulus intensity to the threshold for
action potential generation did not change this property of the
synaptic response (Fig.
1). NMDA
receptor-mediated synaptic responses were observed after blocking the
AMPA/kainate component of the EPSP with 10 µM CNQX and increasing the
stimulation intensity. While most of the excitatory synaptic response
was blocked by the AMPA/kainate antagonist CNQX, it is possible that
some NMDA receptor contribution existed when using weaker stimuli.
Increasing the stimulus intensity in CNQX, however, produced large
AP-5-sensitive synaptic responses (Fig. 1). Cell-to-cell variation
existed in paired-pulse plasticity for AMPA receptor-mediated EPSPs and
EPSCs but within-cell plasticity was relatively stable. The plasticity
expressed by activation of AMPA/kainate receptors showed little change
when increasing the stimulation intensity to the threshold for action
potential generation (Fig. 1) (Akopian et al. 2000;
Ou et al. 1997). In some cases, slight changes in
paired-pulse plasticity were observed with changes in stimulation
intensity, but there was not a consistent trend. By contrast,
paired-pulse potentiation consistently increased for NMDA
receptor-mediated synaptic responses when the stimulation intensity was
increased (Fig. 1).
|
Paired-pulse plasticity depends on the interstimulus interval
Maximal, subthreshold AMPA/kainate receptor-mediated synaptic
responses showed variable sensitivity to changes in paired-pulse ISIs.
Some synapses showed greater facilitation at ISIs shorter than 100 ms,
while others continued to express paired-pulse depression even at short
ISIs (Fig. 2). By contrast, NMDA receptor
responses of equivalent amplitude consistently showed strong
facilitation at ISIs <100 ms. Using a test interval of 50 ms, we found
a clear difference between paired-pulse plasticity produced by AMPA
receptor (BIC) and NMDA receptor activation (BIC + CNQX) (Fig. 2;
EPSPs: P < 0.001; EPSCs: P < 0.02, paired Student's t-tests). Current-clamp cells had an
average resting membrane potential of 87 ± 6.5 (SE) mV
(n = 6). The 50-ms paired-pulse plasticity of AMPA
EPSPs was 88.9 ± 6.2% and NMDA EPSPs was 160.2 ± 11.8%
(n = 6, means ± SE). Stimulation intensity was
increased from 193 ± 15 (SE) µA for AMPA EPSPs to 727 ± 140 µA for NMDA EPSPs to elicit EPSPs of equivalent size
(n = 6). The 50-ms paired- pulse plasticity of AMPA
EPSCs was 83.3 ± 5.1% and NMDA EPSCs was 161.4 ± 20%
(n = 7). Stimulation intensity was increased from
80 ± 10.6 µA for AMPA EPSCs to 143 ± 17 µA for NMDA
EPSCs to elicit EPSCs of equivalent size (n = 7). The
50-ms paired-pulse plasticity was calculated from the average of three
50-ms pairings for each group (predominantly AMPA, NMDA). The
inter-pairing sampling interval was 20 s.
|
Voltage dependence of paired-pulse plasticity
The requirement of large-amplitude synaptic responses and short ISIs for NMDA receptor potentiation suggested a postsynaptic, voltage-dependent mechanism may be involved. We hypothesized the degree of depolarization present during the arrival of the second synaptic response influenced the amplitude of the second response and thus the paired-pulse plasticity. The influence of membrane potential on paired-pulse plasticity was examined by clamping the membrane potential at sequentially more depolarized membrane potentials. Smaller-amplitude synaptic responses were evoked because large EPSCs in combination with depolarization consistently evoked runaway Ca2+ currents. The average stimulation intensity used to evoke the AMPA receptor EPSC in these experiments was 45.7 ± 7.2 µA (n = 6). Following addition of CNQX, the remaining CNQX-insensitive EPSC (NMDA receptor mediated) was small. We therefore doubled the stimulation intensity used to study the AMPA EPSC to sample the NMDA EPSC (BIC + CNQX; 91.4 ± 14.4 µA). This intensity was held constant when the solution was switched to Mg2+-free ACSF containing BIC + CNQX.
The voltage dependency of unitary synaptic responses was initially
tested for each pharmacological condition. Depolarization caused a
linear decrease in the amplitude of the predominantly AMPA receptor
EPSC (BIC), consistent with a decrease in the driving force for cation
movement through the channel (Fig.
3B) (Hestrin et al.
1990). The deviation from the straight line between 40 to
20 mV of the current-voltage plot illustrated in Fig. 3B
can be attributed to NMDA receptor activation at more depolarized membrane potentials. By contrast, following isolation of the
NMDA-receptor EPSC with BIC + CNQX, the current-voltage relationship of
the synaptic response showed a clear voltage-dependent increase with depolarization (Fig. 3). Consistent with
Mg2+-block underlying the voltage dependency of
the NMDA receptor synaptic response, we found
Mg2+-free conditions reduced the voltage
dependency of the response (Hestrin et al. 1990;
Konnerth et al. 1990). The residual voltage dependency
in Mg2+-free ACSF has been attributed to
incomplete removal of Mg2+ from the brain slice
(Clark et al. 1994; Hestrin et al. 1990; Konnerth et al. 1990).
|
Paired-pulse plasticity of AMPA and NMDA receptor-mediated EPSCs showed
differential voltage dependency (Fig. 3). All cells included in this
experiment were sequentially tested under the three pharmacological
conditions of BIC, BIC + CNQX, and BIC + CNQX in
Mg2+-free ACSF. On average, paired-pulse
plasticity of AMPA receptor-mediated EPSCs was relatively insensitive
to changes in striatal neuron membrane potential and, much like the
larger responses shown in Figs. 1 and 2, the tendency was toward
paired-pulse depression (Fig. 3). NMDA receptor-mediated synaptic
responses were negligible at membrane potentials of 60 mV, and little
paired-pulse interaction was observed. Depolarization to 40 mV,
however, revealed clear NMDA receptor-mediated EPSCs and created a
shift in paired-pulse plasticity for NMDA receptor responses toward
potentiation (Fig. 3). The increase in potentiation seen at 40 mV was
interpreted to be due to a depolarization-dependent removal of the
Mg2+ block during the second synaptic response of
the pair. This interpretation assumes the voltage clamp did not control
dendritic depolarization created by the first synaptic response of the
pair. Consistent with this idea, we found removal of extracellular
Mg2+ increased the size of the first response,
and it eliminated the shift toward potentiation at 40 mV
(P < 0.04, paired Student's t-test,
n = 6; Fig. 3).
L-type Ca2+ channels contribute to the EPSC and paired-pulse plasticity
Depolarization can also activate voltage-dependent
Ca2+ channels that could contribute to the
amplitude of NMDA receptor synaptic responses. The L-type
Ca2+ channel antagonist nifedipine (10 µM)
produced a voltage-dependent reduction in the amplitude of the NMDA
receptor-mediated EPSCs tested in Mg2+-free ACSF
(Fig. 4). The stimulation intensity was
held constant as cells were tested sequentially in
Mg2+-free ACSF and
Mg2+-free ACSF containing nifedipine (10 µM).
Nifedipine did not effect the amplitude of the NMDA receptor-mediated
EPSC at membrane potentials more negative than 60 mV or more positive
than +20 mV, but it reduced NMDA receptor EPSCs at intermediate
depolarized membrane potentials. At a holding potential of 40 mV,
nifedipine reduced the amplitude of the NMDA receptor-mediated EPSC by
18.1 ± 4.6% (n = 7; Fig. 4). Nifedipine also
produced a voltage-dependent reduction of paired-pulse potentiation of
the NMDA receptor EPSC. A difference was observed in the paired-pulse
plasticity at 60 mV but not at more hyperpolarized holding potentials
(n = 7; P < 0.05, paired Student's
t-test). This membrane potential is hyperpolarized for
activation of L-type Ca2+ channels, but
activation of these channels can be explained by synaptic
current-mediated depolarization of the dendrites and poor space clamp
(DeFazio and Walsh 1996; Spruston et al.
1994). Cells were first tested after perfusing
Mg2+-free ACSF for 15 min and then again after
perfusing Mg2+-free ACSF containing nifedipine
for 15 min (10 µM; paired analysis). Increasing the stimulation
intensity demonstrated a greater voltage-dependent Ca2+ channel contribution in the form of run-away
plateau currents.
|
The voltage dependence of nifedipine's block of NMDA receptor-mediated synaptic responses and their paired-pulse plasticity suggested it was not changing release from corticostriatal terminals but rather was acting on postsynaptic L-type Ca2+ channels. We tested for presynaptic effects of nifedipine by examining the sensitivity of predominantly AMPA receptor-mediated synaptic responses (slices bathed in BIC) and their paired-pulse plasticity. Nifedipine had no effect on the amplitude of predominantly AMPA receptor-mediated synaptic responses or their paired-pulse plasticity (n = 5; Fig. 5). To examine the postsynaptic block of NMDA responses by nifedipine independent of glutamate release from synaptic terminals, we applied NMDA to striatal neurons with pressure (0.2 mM, 5 s). In three cells examined, NMDA application produced 1.65 ± 0.18 nA of current in brain slices bathed in Mg2+-free ACSF. Addition of 10 µM nifedipine reduced the NMDA evoked current on average by 13% to 1.44 ± 0.17 nA (n = 3; Fig. 5). The decrease in amplitude produced by nifedipine was different from the control prenifedipine amplitude (paired t-test, P < 0.05). The nifedipine effect did not wash out with removal of nifedipine, and postwash responses continued to be different from the prenifedipine control (P < 0.05). The NMDA response was reduced by 18% to 1.35 ± 0.19 nA of control 6 min after nifedipine washout.
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
These experiments demonstrate a postsynaptic
contribution to paired-pulse plasticity at juvenile (p11-p18) and adult
(6 mo) corticostriatal synapses. Paired-pulse plasticity at
hyperpolarized membrane potentials, where the synaptic response is
mediated primarily by AMPA receptors, can be explained by presynaptic
variation in the release of glutamate (Ou et al. 1997).
By contrast, NMDA receptor-mediated synaptic responses displayed
properties consistent with a postsynaptic, voltage-dependent threshold
for triggering facilitation. The voltage dependency of paired-pulse
plasticity appears to be determined by the extent of depolarization
created by the first synaptic response of the pair. These data indicate
depolarization reduces Mg2+ block of NMDA
receptors, which allows greater NMDA receptor-mediated depolarization
during the second of two closely timed synaptic responses
(Clarke et al. 1994; Herron et al. 1986).
Slow-NMDA receptor-mediated depolarizations also triggered the
activation of voltage-dependent L-type Ca2+
channels, which also contributed to paired-pulse potentiation.
EPSPs and EPSCs evoked in BIC-ACSF were almost entirely blocked by CNQX
(Fig. 1) (Cherubini et al. 1988; Walsh et al.
1989). Stronger, suprathreshold stimulation of
corticostriatal afferents or experimental depolarization of the
membrane potential reveals an NMDA receptor component to the EPSP
(Cherubini et al. 1988; Ou and Walsh
1997) (Figs. 1 and 2). These requirements are explained by the
fewer numbers of NMDA receptors found at corticostriatal synapses, as
compared with other telencephalic structures, and by the
characteristically "hyperpolarized" state of striatal neurons that
maximizes the Mg2+ block of NMDA receptors
(Castorina et al. 1994; Wilson and Kawaguchi 1996).
Cell-to-cell plasticity varied for AMPA receptor-mediated synaptic
responses, but within-cell paired-pulse plasticity was relatively
stable over repeated trials and changed little when the stimulation
intensity was increased to the synaptic threshold for generating action
potentials (Akopian et al. 2000; Mori et al.
1994; Ou et al. 1997; Walsh and Ou
1994). Previous work has demonstrated subthreshold
stimulation of corticostriatal axons produces a predominantly AMPA-
receptor mediated EPSP whose short-term plasticity is explained by
changes in presynaptic release properties (Mori et al.
1994; Ou et al. 1997). Considerable variation
exists in the literature concerning paired-pulse plasticity at
corticostriatal synapses, due in part to laboratory differences in the
Ca2+/Mg2+ ratio of
extracellular solutions (Akopian et al. 2000;
Cromwell et al. 1995; Kita et al. 1985;
Mori et al. 1994; Nisenbaum et al. 1993;
Ou et al. 1997). Another factor contributing to
differences in corticostriatal paired-pulse plasticity is age.
Corticostriatal synapses examined in animals younger than postnatal day
20 tend to show depression or no plasticity, as do animals 24 mo of
age (Choi and Lovinger 1997; Mori et al.
1994; Ou et al. 1997). The simplest
interpretation for the tendency of corticostriatal synapses to express
short-term depression is an apparent depletion of the readily
releasable pool of vesicles (Figs. 1A, 2, and 3)
(Applegate and Landfield 1988; Charlton et al.
1982; Fitzpatrick et al. 2001; Katz and
Miledi 1968; Llinás et al. 1991;
Mori et al. 1994; Ou et al. 1997;
Swandulla et al. 1991; Zucker 1999).
Traditional presynaptic theory for facilitation suggests new
Ca2+ entering during the second of two closely
timed action potentials sums with residual Ca2+
in the terminal to enhance neurotransmitter release (Ou et al. 1997; Zucker 1999). The time course of this
phenomena depends on mechanisms of Ca2+ buffering
and sequestration (Zucker 1999). This presynaptic
mechanism should influence the activation of postsynaptic AMPA and NMDA receptors equally. However, in the same cell we found paired-pulse depression of AMPA receptor responses and paired-pulse potentiation of
NMDA receptor responses at short ISIs (Figs. 1-3). The difference between AMPA and NMDA receptor-mediated paired-pulse plasticity suggests two independent processes were acting to produce the final
plasticity. We found it necessary to increase the stimulation intensity
to examine paired-pulse plasticity of NMDA receptor-mediated responses
(Fig. 1). Increasing the stimulus intensity is likely to recruit new
afferents which did not contribute to the initial, predominantly AMPA
receptor-mediated synaptic response. The new population of afferents
recruited by the stronger stimulation intensities may have had unique
release properties that could have contributed to the increase in
paired-pulse potentiation seen when we examined NMDA receptor responses
in isolation.
Alternatively, we hypothesized the paired-pulse potentiation produced
by large depolarizing NMDA receptor-mediated responses was due to the
removal of Mg2+-block during the arrival of the
second synaptic event of the pair. This hypothesis was tested by
examining the relationship between membrane potential and paired-pulse
plasticity using whole cell voltage clamp. Voltage clamping imposes a
somatic membrane potential that is attenuated in the dendrites
(DeFazio and Walsh 1996; Spruston et al.
1994). We observed a depolarized reversal potential
measured for EPSCs and found stronger stimulation to trigger
Ca2+ spikes, verifying the limits of our clamp
(Fig. 3B). Nonetheless, we were able to make qualitative
interpretations of the effect of postsynaptic membrane potential on
paired-pulse plasticity. We found paired-pulse potentiation increased
at membrane potentials expected to reduce Mg2+
block of NMDA receptors (Fig. 3) (Hestrin et al. 1990;
Konnerth et al. 1990). We also found
Mg2+-free ACSF eliminated the
depolarization-induced increase in potentiation (Figs. 3 and 4)
(Clarke et al. 1994; Herron et al. 1986).
Because the voltage-clamp opposes depolarizing synaptic current, it is safe to assume it minimized the effect the first synaptic event had on
the second. By contrast, unopposed depolarization in response to
synaptic current, as occurs naturally, would optimize conditions for
reducing Mg2+ block of NMDA receptors and
enhancement of the second response. This same mechanism has been used
to explain non linear increases in dendritic spine
Ca2+ following paired activation of NMDA receptor
EPSPs in the hippocampus (Mainen et al. 1999).
Our initial use of strong stimulation to study NMDA receptor-mediated
synaptic physiology is in contrast to hippocampal studies where NMDA
receptor contribution to EPSPs is significant under normal stimulation
conditions (Clarke et al. 1994; Herron et al. 1986). Two properties of striatal neurons may explain the
weaker influence of NMDA receptors at resting membrane potentials.
First, more NMDA receptors are found at hippocampal versus striatal
excitatory synapses and second, striatal neurons have higher resting
membrane potentials and thus greater Mg2+ block
to overcome (Castorina et al. 1994; Wilson and
Kawaguchi 1996). When we used weaker stimulation, we still
found a dissociation between AMPA and NMDA receptor-mediated synaptic
plasticity at depolarized versus hyperpolarized membrane potentials,
although it was not as pronounced as when we used strong stimuli (Fig. 3). We are thus left to speculate about the significance of these findings in behaving animals. In contrast to the hyperpolarized membrane potential encountered in striatal neurons in vitro, in vivo
analysis demonstrates cortically driven, long-duration synaptic oscillations that are mimicked by cortical stimulation (Salt et al. 1995). These large, long-duration synaptic events drive
striatal neurons into sustained depolarizations known as "up
states" that last hundreds of milliseconds to seconds (Wilson
and Kawaguchi 1996). The up-states maximize conditions for NMDA
receptor activation, suggesting a role for postsynaptic NMDA
paired-pulse potentiation in the initiation and maintenance of up-state depolarizations.
Our finding of L-type Ca2+ channel participation
in NMDA receptor responses and paired-pulse facilitation indicates
synaptic current also depolarizes the dendrites beyond the activation
threshold for voltage-dependent Ca2+ channels
(VDCCs; Fig. 4). Voltage-clamp experiments showed a significant
contribution of L-type channels to paired-pulse potentiation when the
soma was clamped at 60 mV. This membrane potential is below the
threshold for activation of L-type channels, but again it can be
explained by poor space clamp of the dendrites and it is likely
synapses drove the membrane potential of the dendrites well beyond the
activation threshold of L-type Ca2+ channels.
Alternatively, nifedipine could have acted presynaptically to block
Ca2+ channels and reduce synaptic transmission.
However, most terminals including corticostriatal are known to use N-
and P/Q-type Ca2+ channels for
excitation-secretion coupling (Krieger et al. 1999; Lovinger et al. 1994; Sheng et al. 1998).
L-type Ca2+ channels have been shown to
participate in the priming of presynaptic terminals during
high-frequency activity but not in the actual release process
(Jensen et al. 1999). Arguing against this idea, nifedipine showed a voltage-dependent reduction of NMDA
receptor-mediated EPSCs that indicates it is acting
postsynaptically (Fig. 4). We also found nifedipine reduced the
response to exogenous application of NMDA, but it did not effect the
amplitude or paired-pulse plasticity of AMPA/kainate receptor
mediated-synaptic responses (Fig. 5). Together, these data strongly
suggest nifedipine reduced NMDA receptor-mediated paired-pulse
plasticity by blocking postsynaptic L-type Ca2+ channels.
The nifedipine results demonstrate L-type Ca2+
channels are activated by NMDA receptor-mediated depolarizations.
Synapses are known to activate dendritic T-type channels and increase
dendritic Ca2+ through VDCCs (Markram and
Sakmann 1994; Miyakawa et al. 1992). Our data
provide further evidence for VDCCs participating in subthreshold synaptic integration (Markram and Sakmann 1994). The
voltage dependence of NMDA receptors and VDCCs may act as threshold
detectors for enhancing synaptic signals, a property increased when
inputs arrive in rapid succession (Bernander et al.
1994; Yuste et al. 1999). In addition,
subthreshold voltage-dependent enhancement of synaptic function
increases the likelihood of triggering active dendritic responses
(Magee et al. 1995; Schwindt and Crill
1997). VDCCs are known to increase spine
Ca2+ during responses to glutamate application as
well as during low-frequency subthreshold synaptic activity
(Magee et al. 1995; Schiller et al. 1998;
Yuste et al. 1999). It is likely, therefore that VDCCs also participate in the facilitation of spine
Ca2+ produced by paired NMDA receptor EPSPs
(Mainen et al. 1999). The combined entry of
Ca2+ through these two voltage-dependent
mechanisms impacts discharge probability and likely cooperates to
activate intracellular cascades important for modifying synaptic strength.
| |
ACKNOWLEDGMENTS |
|---|
We thank G. Allen for technical assistance.
This study was supported by National Institute on Aging Grants R29 AG-12679, AG-05142, and AG-00093.
| |
FOOTNOTES |
|---|
Address for reprint requests: J. P. Walsh, Andrus Gerontology Center, University of Southern California, Los Angeles, CA 90089-0191 (E-mail: jwalsh{at}usc.edu).
Received 12 February 2001; accepted in final form 23 August 2001.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Pawlak and J. N. D. Kerr Dopamine Receptor Activation Is Required for Corticostriatal Spike-Timing-Dependent Plasticity J. Neurosci., March 5, 2008; 28(10): 2435 - 2446. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. G. Carter, G. J. Soler-Llavina, and B. L. Sabatini Timing and Location of Synaptic Inputs Determine Modes of Subthreshold Integration in Striatal Medium Spiny Neurons J. Neurosci., August 15, 2007; 27(33): 8967 - 8977. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Akopian and J. P. Walsh Reliable long-lasting depression interacts with variable short-term facilitation to determine corticostriatal paired-pulse plasticity in young rats J. Physiol., April 1, 2007; 580(1): 225 - 240. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. S. Bamford, S. Robinson, R. D. Palmiter, J. A. Joyce, C. Moore, and C. K. Meshul Dopamine Modulates Release from Corticostriatal Terminals J. Neurosci., October 27, 2004; 24(43): 9541 - 9552. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Lape and J. A. Dani Complex Response to Afferent Excitatory Bursts by Nucleus Accumbens Medium Spiny Projection Neurons J Neurophysiol, September 1, 2004; 92(3): 1276 - 1284. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
) produced a slow inward current in
striatal neurons bathed in Mg2+-free ACSF
(left). Addition of 10 µM nifedipine reduced the
inward current produced by NMDA application. Bottom: the
net current blocked by nifedipine (subtraction of NMDA response
produced before and after nifedipine addition). D: plot
shows the change in amplitude of the NMDA response produced by
nifedipine addition. Nifedipine produced a significant reduction in the
size of the NMDA response. The bracketed time points were all
significantly different from the prenifedipine control (paired
t-test, P < 0.05). The holding
potential for cells in A and B was