| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 87 No. 5 May 2002, pp. 2297-2306
Copyright ©2002 by the American Physiological Society
The Vollum Institute, Oregon Health and Science University, Portland, Oregon 97201-3098
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Leão, Ricardo M. and
Henrique Von
Gersdorff.
Noradrenaline Increases High-Frequency Firing at the Calyx of
Held Synapse During Development by Inhibiting Glutamate Release.
J. Neurophysiol. 87: 2297-2306, 2002.
The
mammalian auditory brain stem receives profuse adrenergic innervation,
whose function is poorly understood. Here we investigate, during
postnatal development, the effect of noradrenaline (NA) at the calyx of
Held synapse in the rat medial nucleus of the trapezoid body (MNTB). We
observed that NA inhibits the large glutamatergic EPSC, evoked by
afferent fiber stimulation, in a dose-dependent manner. The inhibition
was maximal (approximately 48%) at the concentration of 2 µM. It was
antagonized by yohimbine and mimicked by the 
2-adrenergic specific
agonist UK14304. Both AMPA and NMDA receptor-mediated EPSCs were
inhibited in parallel by NA, suggesting a presynaptic effect.
Presynaptic recordings showed that NA inhibits the action potential
(AP) generated Ca current by about 20%; however, NA did not
significantly affect the presynaptic AP waveform. We thus conclude that
the calyx of Held presynaptic terminal expresses 2-adrenergic
receptors that inhibit its Ca current and thus glutamate release.
Noradrenaline was effective in all cells tested from postnatal days 6 to 7 (P6-P7), and thereafter the number of responsive cells
diminished, although half of the P14 cells tested still had EPSCs that
were inhibited by NA. By contrast, activation by
L-2-amino-5-phosphonovaleric acid-sensitive metabotropic
glutamate receptors strongly inhibited the EPSCs of all cells tested
from P6 to P14. The effect of NA on postsynaptic action potential
firing was dependent on the stimulus frequency. At 10 Hz, NA had no
effect on firing probability; however, NA helped MNTB cells fire more
action potentials during a 100-Hz train of stimuli, even though it did
not increase the steady-state depressed EPSC, because it produced a
smaller N-methyl-D-aspartate (NMDA)
receptor-activated depolarizing plateau. We therefore suggest that the
reduction by NA of the first few EPSCs in a train leads to a smaller
NMDA depolarizing plateau and thus to increased firing probability at
100 Hz in young synapses. Surprisingly, the inhibition of glutamate
release by NA can thus actually increase the excitability of MNTB
neurons during early postnatal development.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Many neurotransmitters and
modulators in the central and peripheral nervous system act on
presynaptic terminals inhibiting the release of neurotransmitters via
activation of metabotropic receptors (Wu and Saggau
1997
). Three mechanisms proposed to account for the presynaptic
inhibition of transmitter release are: inhibition of presynaptic Ca
channels, activation of presynaptic ion channels (e.g., K channels),
and regulation of the synaptic vesicle fusion complex by modulation of
the proteins involved in exocytosis (Jones and Elmslie
1997; Miller 1998). One main problem in
addressing the question of which mechanism a given transmitter uses to
inhibit transmitter release is the inaccessibility of most presynaptic bouton-type terminals to electrophysiological approaches.
Few preparations allow direct recordings of the presynaptic terminal.
The calyx of Held, a giant synaptic terminal in the mammalian auditory
brain stem (Rowland et al. 2000), is involved in
binaural sound localization (Guinan and Li 1990;
Oertel 1999; Spirou et al. 1990), and its
large size allows direct electrophysiological recordings (Borst
et al. 1995; Forsythe 1994; Sakaba and
Neher 2001). Glutamate release from the rat calyx of Held is
mediated mainly by P/Q-type calcium channels after postnatal day 10 (Forsythe et al. 1998; Iwasaki and Takahashi
1998; Wu et al. 1999). Direct recordings of the
calyx demonstrated that activation of presynaptic metabotropic
GABAB (Isaacson 1998;
Takahashi et al. 1998) and glutamate (Takahashi
et al. 1996) receptors inhibits glutamate release via
inhibition of presynaptic calcium channels.
Noradrenaline alters neuronal excitability and transmitter release by
G-protein-coupled receptors in central and peripheral synapses
(Boehm 1999; Dunlap and Fischbach 1981;
Kamisaki et al. 1992; Kondo and Marty
1998; Lipscombe et al. 1989; Madison and Nicoll 1986; Scanziani et al. 1993). The
mammalian auditory brain stem, and in particular the medial nucleus of
the trapezoid body (MNTB) (Wynne and Robertson 1996),
receives extensive adrenergic input (Jones and Friedman
1983; Klepper and Herbert 1991;
Kössl et al. 1988; Vincent
1988). However, the physiological function of this adrenergic
innervation is not well understood. Noradrenaline (NA) can profoundly
affect the response to auditory stimuli of neurons in the adult bat
cochlear nucleus because it reduced spontaneous neuronal activity and
it caused a twofold decrease in the latency jitter of the first
tone-evoked spikes (Kössl and Vater 1989). In
rats, NA applied iontophoretically increases the discharge of the
neurons in the cochlear nucleus (Ebert 1996). In
addition, NA also alters the excitability of neurons in the ventral
nucleus of the trapezoid body by decreasing K+
conductances (Wang and Robertson 1997) and in the MNTB
by modulating the size of a postsynaptic hyperpolarization-activated
current Ih (Banks et al.
1993).
Here we investigated the effects of NA on the calyx of Held synapse. We
found that NA inhibits glutamate release by inhibiting presynaptic Ca
channels via activation of 2-adrenoreceptors and that this effect is
developmentally regulated, being strongly present in immature synapses.
Surprisingly, presynaptic inhibition of glutamate release by NA allows
the postsynaptic MNTB neuron to fire more action potentials (APs)
during a 100-Hz stimulus train due to a reduction in the NMDA
receptor-activated depolarizing plateau.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Slice preparation
Brain stem slices were obtained from postnatal day 6 (P6) to P15 Sprague-Dawley rats. After rapid decapitation, the brain stem was immersed in ice-cold low-calcium artificial cerebral spinal fluid (ACSF) containing (in mM) 125 NaCl, 2.5 KCl, 3.0 MgCl2, 0.1 CaCl2, 25 glucose, 25 NaHCO3, 1.25 NaH2PO4, 0.4 ascorbic acid, 3 myo-inositol, and 2 Na-pyruvate, pH = 7.3 when bubbled with carbogen (95% O2-5% CO2). Transverse slices (170- to 200-µm thick) were cut proceeding from a caudal to rostral direction. Slices were rapidly transferred to an incubation chamber containing normal ACSF bubbled with carbogen and maintained at 37°C for 30-50 min and thereafter at room temperature. The normal ACSF was the same as the low-calcium ACSF except that 1.0 mM MgCl2 and 2.0 mM CaCl2 were used.
Electrophysiology
Whole cell patch-clamp recordings were performed in normal ACSF
at room temperature (21-23°C). The standard patch pipette solution
consisted of (mM) 130 K-gluconate, 20 KCl, 5 Na2-phosphocreatine, 10 HEPES, 5 EGTA, 4ATP-Mg,
and 0.5 GTP, pH = 7.3 with KOH. To record
N-methyl-D-aspartate receptor (NMDAR)-mediated
currents at depolarized resting membrane potentials, CsCl was
substituted for K-gluconate and KCl and 10 mM TEA-Cl was added. During
experiments, the slices were continuously perfused with normal
ACSF solution via a gravity-fed system. Neurons were visualized
by infrared-differential interference contrast (IR-DIC)
microscopy (Leica LDMLFS, Leica, Weltzar, Germany) through a
×40 water-immersion objective (Leica APO LU-V-I, Leica) and a CCD
camera (C79, Hammamatsu, Japan). A bipolar stimulation electrode was
placed on the brain stem midline. Connected cells were preselected by
the presence of evoked extracellular APs with a patch pipette
containing normal ACSF (Borst et al. 1995; Guinan
and Li 1990).
Presynaptic calcium currents were recorded in calyces identified
visually by fluorescence labeling with Lucifer yellow (0.25 mg/ml)
included in the internal solution. The extracellular solution was ACSF
with 20 mM TEA-Cl, substituted equimolarly for NaCl, plus TTX (1 µM).
The internal solution for calcium currents recordings consisted of (mM)
90 Cs-methanesulfonate, 20 CsCl, 1 MgCl2, 5 Na2-phosphocreatine, 40 HEPES, 10 TEA-Cl, 0.5 EGTA, 4 ATP-Mg, and 0.2 GTP, pH = 7.3 with CsOH. The osmolarity of
the internal solution was adjusted to approximately 310 mOsm with CsCl.
The calyx terminal was held at 
70 mV. The Ca2+
currents displayed in Fig. 5A were evoked by a pair of 15-ms square pulses to 10 mV, delivered at 0.1 Hz, with a 100-ms
depolarizing prepulse of 100 mV in between to induce relief of
G-protein inhibition of the calcium current (Elmsie et al.
1990). The relationship between first (I1) and second Ca
current (I2; elicited after the depolarizing prepulse to 100 mV) was
used to confirm the voltage-dependent inhibition and to discard effects
due to the slow and spontaneous run-down of the current that occurred
in some terminals. Leak subtraction was done using a P/n protocol. In
some cells, we recorded calcium currents evoked by a presynaptic AP
waveform, which was recorded previously from a calyx terminal of the
same age in fast current-clamp mode (EPC-9 amplifier).
Patch pipettes were pulled from soft thin-walled glass (WPI, Sarasota,
FL) using a Narishige puller (PP-830, Japan). Patch pipettes had an
open tip resistance of 1.5-3.0 M
for postsynaptic recordings and
2.5-7.0 M for presynaptic recordings. Postsynaptic access
resistance Rs were around 2-5 M
and Rs compensation was set to
75-90% (10 µs lag). Presynaptic terminals had a
Rs around 10-15 M and were also
electronically compensated (about 60-70%). Principal cells were
voltage-clamped at a holding potential of 70 mV if not stated
otherwise. No corrections were made for liquid junction potentials.
Pre- and postsynaptic APs were recorded in the fast current-clamp mode
of the EPC-9 after adjusting for the fast-capacitance cancellation
while in cell-attached mode. After break-in, the Rs value was determined in the
voltage-clamped cell at 70 or 80 mV. Current-clamp recordings were
continued only if the initial uncompensated
Rs was <10 M. For current-clamp AP
recordings, presynaptic terminals were identified by choosing connected
cells that, after whole cell, presented "action currents" instead
of excitatory postsynaptic currents (EPSCs) when electrically
stimulated (Forsythe 1994; Taschenberger and von
Gersdorff 2000). Presynaptic recordings were unequivocally
confirmed afterwards via Lucifer yellow fluorescence.
Afferent fiber stimulation was applied through a Master-8 stimulator (AMPI, Jerusalem, Israel) and had a duration of 100 µs and amplitudes of 2 to 25 V. Stimulation pulses were controlled using Pulse software (HEKA, Germany), and signals were recorded via a EPC-9 (HEKA) patch-clamp amplifier. Sampling intervals were 20 or 50 µs for AMPA or NMDA-EPSC recordings, and 10 µs for calcium current recordings. Data were low-pass filtered at 2.9 kHz (Bessel).
Drugs and off-line analysis
Yohimbine hydrochloride, UK 14304, L-2-amino-5-phosphonovaleric acid (L-AP4), and
(RS)--cyclopropyl-4-phosphonophenylglycine (CPPG) were
obtained from Tocris-Cookson (Bristol, UK). TTX was from Alomone Labs
(Jerusalem, Israel). All other salts and chemicals were from Sigma (St.
Louis, MO). All drugs were kept as 1,000-fold concentrated stock
solutions and added to the barrels with oxygenated ACSF during the
experiment. A fresh stock solution of NA was prepared every 10 days and
kept protected from light. NA was added to the ACSF immediately prior
to its perfusion in the bath to avoid oxidation and light degradation.
Off-line analysis was done with PulseFit (HEKA, Germany) or IgorPro software (Wavemetrics, Lake Oswego, OR). Statistical analysis and curve fitting were also performed with Microsoft Excel and Prism 3.0a (GraphPad, San Diego, CA). Paired and unpaired t-tests were performed to access statistical significance of the data, and means with two-tail P values less than 0.05 were considered significantly different. Data are reported as mean ± SE values.
For quantifying the mean postsynaptic NMDA receptor plateau depolarization in Fig. 10, C and D, during a 100-Hz train, we choose to take the midpoint between the initial depolarization caused by the first EPSP and the maximum value attained by the plateau depolarization during the train. This was important because in many occasions the initial depolarization was substantially more affected by NA application than the maximum plateau depolarization.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
NA inhibits the glutamatergic EPSC
Afferent fiber stimulation at 0.1 Hz elicited AMPA receptor-mediated EPSCs in the principal cells of the MNTB with a mean peak amplitude of 4.8 ± 0.6 nA (n = 64). The perfusion of noradrenaline (NA) in the recording chamber decreased the EPSC amplitude in 77.2% of the 101 cells tested (Figs. 1A and 2A). The effect was reversible, and re-application of the same concentration of NA produced the same magnitude of effect (Fig. 1A). The effect was dose dependent, and was observed in concentrations as low as 50 nM, and reached a plateau around 2 µM (Fig. 1C). The mean amount of inhibition produced by 2 µM of NA (46.7 ± 4%; n = 33) or 20 µM of NA (48.8 ± 2%; n = 37) was not significantly different (P = 0.56). These saturating concentrations inhibited the AMPA EPSCs by 47.8 ± 2% (n = 70). The amount of the inhibition ranged from 7.0 to 90.4%. The half-effective concentration (IC50), estimated by fitting a logistic function to the dose-response data, was 0.19 ± 0.05 µM.
|
|
The NA inhibited EPSC had the same average kinetics as the control EPSC (rise time control: 0.27 ± 0.02 ms, NA: 0.29 ± 0.04 ms; half-width control 1.6 ± 0.3 ms, NA 1.5 ± 0.2 ms; P > 0.05; n = 5 cells from slices of rats from P7 to P14). This lack of effect is clearly demonstrated when the EPSCs were normalized (Fig. 1B).
Effect of NA is dependent on development
During the range of ages used in this work, the rat calyx of Held
synapse undergoes a series of morphological (Kandler and Friauf
1993) and physiological (Chuhma and Omori 1998;
Iwasaki and Takahashi 1998, 2001; Taschenberger
and von Gersdorff 2000) changes. The expression of mGluR
subtypes is also dependent on age (Elezgarai et al.
1999). Because not all cells were affected by NA, we asked
whether this is due to a developmental process.
The effect of NA is clearly dependent on the developmental stage of the synapse as can be seen in Fig. 2A. We observed that the number of NA-responsive cells decreased with increasing age. All cells from P6 and P7 rats responded to NA (n = 26), but the percentage of responsive cells started to decrease progressively from P8 (93.3% of responsive cells, n = 15) to P15 (25% of responsive cells; n = 4).
The average effectiveness of NA did not seem to change considerably with development as can be seen in Fig. 2B. NA was still very effective in inhibiting the EPSC of responsive cells from P14 rats (61 ± 3% inhibition; n = 4). However, the mean inhibition at P7 was significantly higher than at other ages that had a similar number of tested cells (61.7 ± 4.4% at P7, n = 16; 46.4 ± 5% at P8, n = 13; 39.0 ± 3.5% at P9, n = 17; Fig. 2B). The amount of NA inhibition was not correlated with the EPSC peak amplitude (data not shown), and the potency (IC50) of NA did not seem to change significantly during development.
Effect of NA is mediated by 2-adrenoreceptors
The type of receptor involved in the effect of NA was investigated
using the antagonists of 
, 1, and
2 receptors, propranolol, prazosin, and
yohimbine, respectively (Fig.
3A). After application of 2 µM NA, subsequent addition of propranolol (1 µM) or prazosin (1 µM) did not antagonize the effect of NA (Fig. 3B), ruling
out the participation of adrenoreceptors of and 1 types in this effect. In contrast, after the addition of yohimbine (20 µM), the
EPSC amplitude returned to control (Fig. 3, A and
B). Application of yohimbine without previous application of
other antagonists completely antagonized the NA inhibition of the EPSC
(n = 6; not shown), confirming that the antagonism was
not due to a delayed effect of propranolol or prazosin. Further
evidence for a role of 2 receptors in this effect was obtained with
the specific 2 agonist UK 14304 (1 µM), which could mimic the
effect of NA, albeit less efficiently (Fig. 3, C and
D). Subsequent application of 2, 5, or 20 µM UK 14304 did
not increase the degree of inhibition in two cells tested. The effect
of UK 14304 was fully antagonized after application of 10 µM
yohimbine (Fig. 3, C and D). In cells that did
not respond to NA, subsequent application of UK 14304 also had no
effect (data not shown), showing that the lack of effect of NA in
slices from older rats is not due to an increase in the uptake or
oxidation of NA. In this regard, note that the recording solution
routinely contained 0.4 mM ascorbic acid, which should prevent NA
oxidation. These results thus indicate that 2 adrenergic receptors
mediate the effect of NA on the EPSC.
|
Presynaptic effect of NA
To know if NA is acting postsynaptically in the AMPA receptors or
presynaptically, we studied the effect of NA on the AMPA EPSC and NMDA
EPSC simultaneously (von Gersdorff et al. 1997). The
cells were kept at a depolarized holding potential (+30 mV) to relieve
the block of NMDA receptors by magnesium. Application of NA produced a
simultaneous and similar decrease in the amplitudes of both the AMPA
and the NMDA-EPSCs (30 ± 6% in the AMPA EPSC and 38 ± 6%
in the NMDA EPSC; n = 8; Fig.
4, A and B), and
the effect on both EPSCs was fully antagonized by application of
yohimbine. Although NA was slightly (8%) more efficient in blocking
the NMDA-EPSC, this decrease correlated very well with the decrease in
the AMPA-EPSC. This suggests that NA is acting presynaptically by
inhibiting glutamate release.
|
NA inhibits presynaptic Ca channels
Because inhibition of Ca channels on neuronal somas is a common
mechanism of action of the adrenoreceptors (Boehm and Huck 1996; Dunlap and Fischbach 1981;
Lipscombe et al. 1989), we asked whether NA is also
acting on presynaptic Ca channels. To directly investigate the action
of NA on presynaptic Ca channels, we recorded presynaptic Ca currents
in calyxes morphologically identified with Lucifer yellow fluorescence
(P6-P8 rats; Fig. 5).
|
We used a double pulse protocol to study the voltage-dependent
inhibition of Ca channels by NA (Bean 1989;
Dolphin 1998; Elmsie 1990; see
METHODS and Fig. 5A). When we perfused NA (20 µM), we observed a depression of the peak calcium current (I1) of
10.9 ± 1.5% (Fig. 5; from 0.94 ± 0.06 to 0.84 ± 0.06 nA; n = 11; P < 0.01). Characteristic
of a G protein inhibition, NA inhibition of the Ca current was almost
completely relieved by the 100-mV depolarizing prepulse (Fig. 5;
current at I2 from 0.88 ± 0.07 nA to 0.86 ± 0.07 nA;
n = 11; P > 0.05). Accordingly, when
the ratio I2/I1 was compared, we observed that it significantly
increased from 0.93 ± 0.04 to 1.02 ± 0.03 (P < 0.01).
In cells from older animals (P9-P12), NA had a similar effect inhibiting Ca channels by 10.2% (P < 0.01; n = 5). In three of eight P9-P12 terminals tested, the Ca current was insensitive to NA application (I2/I1 control = 1.0 ± 0.05; I2/I1 NA = 1.0 ± 0.06; n = 3) as expected by the developmental decrease of responsive cells observed in Fig. 2A. Terminals from animals older than P12 were not tried due to the heavy myelinization of fibers present in the brain stem slices after P12 that severely impairs visualization of the terminals.
Because the effect of NA and other G-protein-linked transmitter
receptors affects mainly the activation phase of the current, we
supposed that in a calcium current evoked with a AP waveform NA should
have a more pronounced effect (Brody et al. 1997;
Park and Dunlap 1998). In fact, when we evoked the
calcium current with a presynaptic AP waveform recorded from a P7 calyx
(Fig. 6A), NA was almost twice
as potent in inhibiting the peak of the AP-evoked Ca current than the
peak of the square-pulse-evoked Ca current in the same P7/8 terminals
(8.6 ± 1.9%, square pulse vs. 16.6 ± 1.7%, AP evoked;
P < 0.01; n = 3; Fig. 6B).
|
In the rat calyx of Held synapse, neurotransmitter release is triggered
mainly by activation of P/Q-type voltage-dependent Ca channels with a
significant contribution of N- and R-type voltage-dependent Ca channels
in young animals (Iwasaki and Takahashi 1998; Wu
et al. 1998, 1999). Because we observed a substantial effect of
NA in animals older than P10 when the calyx Ca-channel is exclusively of the P/Q type (Iwasaki and Takahashi 1998), this
suggests that the decrease of NA responsive cells with age is not due
to the observed decline of N- or R-type Ca channel with increasing age.
NA does not affect the presynaptic AP waveform
Manipulations of the presynaptic K currents can alter the shape of
the presynaptic AP (Ishikawa and Takahashi 2000;
Wang and Kaczmareck 1998), while the presynaptic Ca
current does not take part in the shaping of the presynaptic AP
waveform (Borst et al. 1995). So it is possible that the
NA effect could also reduce release by shortening the duration of the
presynaptic AP by activation of presynaptic K+
channels. We therefore recorded presynaptic APs evoked by afferent fiber stimulation. Figure 5B shows an example of two
presynaptic APs before and after the application of NA. As can be seen,
NA did not alter the presynaptic AP waveform. In four calyces tested from P6 to P9 rats, NA application did not change the amplitude (control, 97 ± 9 mV; NA, 93 ± 12 mV; P > 0.05), the 20-80% rise time (control, 0.21 ± 0.03 ms; NA,
0.21 ± 0.03 ms; P > 0.05), or the half-width
(control, 0.8 ± 0.14 ms; NA, 0.9 ± 0.15 ms; P > 0.05) of the AP. In addition, the presynaptic
resting potential was also not changed by NA (control 67 ± 3 mV; NA, 66 ± 3 mV; P > 0.05, n = 4). We conclude that NA has no significantly effect on presynaptic
APs or in the presynaptic resting membrane potential.
NA does not inhibit all G-protein-sensitive Ca channels in the calyx
Activation of metabotropic glutamate receptors by the agonist
L-AP4 also leads to inhibition of glutamate release
(Barnes-Davies and Forsythe 1995) by
inhibition of presynaptic calcium channels (Takahashi et al.
1996). We observed that application of L-AP4 (50 µM) inhibited the EPSC amplitude more potently (70.2 ± 3.5%; n = 16) than NA. In accordance with this more-potent
inhibitory effect, the sequential application of L-AP4
after NA application produced a further inhibition in the EPSC
amplitude (NA 46.1, ± 6.9%; NA + L-AP4, 76.4 ± 3.2, P < 0.01; n = 5; Fig.
7A). Application of yohimbine
(20 µM) did not relieve the effect (NA + L-AP4 + yohimbine, 73.5 ± 6.8%; n = 4; P > 0.05 when compared with NA and L-AP4 or
L-AP4 alone) showing that L-AP4 and NA have
nonadditive effects. Inversely, application of the mGluR group II/III
antagonist CPPG, reverts the effect of L-AP4 on the EPSC
after co-application with NA (Fig. 7B), and the remaining
inhibition is reverted completely by yohimbine (Fig. 7B). In
fact after inhibition of presynaptic Ca currents by NA,
L-AP4 is able to further inhibit the Ca current (Fig.
7C), and L-AP4 occludes the NA effect in
inhibiting the EPSC when applied first (Fig. 7B). We
conclude that L-AP4 acts in the same pool of Ca channels
accessible to NA but can inhibit another additional pool of Ca channels
that are not affected by NA in the calyx of Held.
|
We also observed that the effect of L-AP4 is not dependent on development. Cells where the EPSC or the presynaptic Ca current did not respond to NA still invariably responded to L-AP4 (P6-P14 cells; data not shown). This demonstrates that the developmental decline of the effect of NA is a phenomenon specific to its receptor and not due to some general change in the G-protein machinery.
Effects of NA during 10- and 100-Hz trains of stimuli
So far we have studied the effect of NA in EPSCs evoked at the
frequency of 0.1 Hz, but the physiological frequencies of discharge of
this synapse may be much higher even in immature animals (Spirou et al. 1990; Wu and Kelly 1993). For example, in
adult cats the spontaneous rate of firing of the calyciferous axon
varied from 10 to 110 Hz (Spirou et al. 1990). However,
at frequencies as low as 10 Hz, the EPSCs may already present some
depression that is especially severe (more than 90%) in young rat pups
(Borst et al. 1995; Taschenberger and von
Gersdorff 2000; von Gersdorff et al. 1997). We
tested the effect of NA on the EPSCs elicited during 10- and 100-Hz
frequencies trains of stimuli. We observed that NA is less effective in
inhibiting the amplitude of the depressed steady-state EPSCs (ssEPSCs)
at 10 Hz (16.3 ± 4.7% inhibition; n = 15;
P < 0.05; Fig.
8A) and ineffective at
depressed ssEPSCs generated by a 100-Hz train (4.9 ± 6.4%
inhibition; n = 13; P > 0.05; Fig.
9). A similar result was also observed
when L-AP4 (50 µM) was applied to the slice (data not
shown). Thus in contrast to the avian nucleus magnocelularis, an
auditory region that also receives a calyceal input, we did not observe
a general enhancement of the depressed ssEPSCs after presynaptic
inhibition (Brenowitz et al. 1998).
|
|
Despite the depression of the EPSC, the postsynaptic cell can reliably
fire APs when subjected to a 10-Hz train of stimuli. But we observed
that EPSC inhibition by NA has no effect on the firing of APs elicited
by a 10-Hz train of stimuli (Fig. 8B). Because the
noradrenergic AMPA-EPSC amplitude inhibition per se is not sufficient
to prevent the postsynaptic cell from firing, could NA affect the
firing of the postsynaptic cell by other means? Although the
postsynaptic cell is able to reliably fire APs during a high-frequency
train (100 Hz or more) at synapses older than P14, at immature synapses
(P5-P8), high-frequency trains of stimuli produce a plateau
depolarization due to the activation of NMDA receptors that impairs the
firing of APs (Futai et al. 2001; Taschenberger and von Gersdorff 2000). When postsynaptic cells from P6 to P9 rats in current-clamp were subjected to a 100-Hz train of 20 stimuli, we observed that the number of APs fired in the train was inversely correlated to the mean amplitude of the NMDA-plateau depolarization as
can be seen in Fig. 10C. We
thus suggest that the size of the NMDA-plateau depolarization regulates
the number of APs fired in a 100-Hz train.
|
Because the amplitude of the NMDA-EPSC decays in a developmental time
frame similar to that observed for the NA effect, we asked whether the
presynaptic inhibition of glutamate release by NA can increase the
firing of P6-P8 postsynaptic cells by diminishing the amplitude of the
plateau depolarization. From 11 cells tested, 8 fired less than the
maximum response (Fig. 10A) and 3 fired maximally (20 APs or
slightly more due to some aberrant firing; Fig. 10B) (see
Futai et al. 2001). In the cells that fired less than
the maximum rate, application of NA reduced significantly the size of
the mean plateau depolarization (from 39 ± 4 to 45 ± 3 mV; P < 0.01; Fig. 10, A and D),
and the number of APs fired was concomitantly increased (ratio
APs/EPSPs from 0.38 ± 0.1 to 0.5 ± 0.1; P < 0.01; Fig. 10, A and D). In cells that already
fired at the maximum rate (Fig. 10B), NA had no effect in
increasing the firing (ratio APs/EPSPs from 1 ± 0.02 to 1 ± 0.04), although it decreased the NMDA-plateau depolarization (from
54 ± 0.1 to 58 ± 2 mV; P > 0.05; Fig.
9B), and it significantly increased the AP amplitude of
these cells. These 100-Hz effects and the lack of effect at 10 Hz (Fig.
8), a frequency that does not produce a large NMDA plateau
depolarization, suggest that the increase in firing observed at 100 Hz
was not due to an increase of the general excitability of the
postsynaptic cell. Also corroborating this, we observed that the
relationship between the number of APs fired and the size of the
plateau depolarization was the same either in the absence or in the
presence of NA (Fig. 10C). Because we did not observe a
potentiation of the depressed EPSCs after presynaptic inhibition of
glutamate release, as observed by Brenowitz et al.
(1998), it is not likely that an increase in the depressed
EPSCs amplitude is responsible for the increased firing. We suggest
instead that the large depolarizing plateau produced in the absence of
NA probably prevents the recovery from inactivation of the postsynaptic
sodium channels (as suggested also by the decreased AP amplitude, Fig.
10B), and this then prevents the firing of APs during rapid stimulation.
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
We have characterized the effect of NA in the calyx of Held
synapse in the auditory brain stem during early development. NA inhibits glutamate release at the calyx by activation of
2-adrenoreceptors that inhibit presynaptic Ca channels. This effect
is clearly associated with immature synapses because the number of
responsive cells decreases from 100% in immature P6-P7 calyxes to
25% in the more mature P15 calyx. NA was relatively weak at inhibiting
presynaptic Ca channels because the activation of mGluRs by the agonist
L-AP4 is able to inhibit a larger fraction of Ca channels,
including those inhibited by NA. Finally, we observed that presynaptic
inhibition by NA in immature calyces makes some postsynaptic neurons
fire significantly more APs during a 100-Hz train of stimuli. Immature calyces have large NMDA receptor EPSCs that produce a plateau depolarization that can inactivate postsynaptic Na channels when the
synapse is firing at a high-frequency. So, presynaptic inhibition of
glutamate release in immature calyces would lead to a smaller NMDA
receptor plateau depolarization and thus to less Na channel inactivation and thereby to more postsynaptic firing.
A series of evidences was presented that suggest this is the mechanism
involved in the increase of firing of the postsynaptic neuron and not
some other change in the postsynaptic membrane excitability that NA may
produce. First, NA did not change the general relationship between the
plateau depolarization amplitude and the number of APs fired. Second,
at the frequency of 10 Hz, which does not produce a large sustained
plateau depolarization, NA has no effect on the number of APs fired.
Third, NA has no effect in the amplitude of the depressed ssEPSCs
during the 100-Hz train, showing that the potentiation observed by
Brenowitz et al. (1998) cannot contribute to the
increased firing.
We have also demonstrated pharmacologically that NA acts via
2-adrenoreceptors. Calcium channel inhibition by these receptors and
others is due to the direct interaction of the 
subunits of the G
protein with the presynaptic Ca channels (Dolphin 1998). Subtypes of the 2-adrenoceptors (2A/D,
2B, and 2C) are known from biochemical and genetic studies (MacKinnon et al.
1994), but a pharmacological distinction is difficult, due to a
lack of highly selective agonists and antagonists suitable for
functional studies. Genetically engineered mice lacking, or
overexpressing, the different subtypes of 2-adrenoceptors
(Kable et al. 2000) have shown that in most cases the
2A/D subtype is responsible for most of the
classical central and peripheral actions of 2-adrenoreceptors.
In the calyx of Held, the presynaptic inhibition of glutamate release
by mGluR and GABAB receptors has been attributed
to the inhibition of presynaptic Ca channels, without any participation of potassium channels (Isaacson 1998; Takahashi
et al. 1996, 1998), although one cannot exclude completely a
direct action on the release machinery, like that recently demonstrated
by Blackmer et al. (2001). We observed an approximately
10% inhibition of the peak of depolarizing-step evoked presynaptic Ca
current by NA. In Ca currents evoked by an AP waveform, the peak
inhibition was almost twice as potent (1.9 times). If we suppose a
power-relation between intracellular Ca and transmitter release of
around 3-4, as has been demonstrated for the calyx of Held
(Bollmann et al. 2000; Borst and Sakmann,
1999; Schneggenburger and Neher 2000; Wu
et al. 1999), a 17-19% inhibition of the peak AP-evoked
calcium current will produce a 43-53% inhibition of the transmitter
release in the range of what we observed (average of 48%).
We observed that the agonist for group III mGluR receptors,
L-AP4 is more effective than NA in blocking Ca current and
glutamate release. Their effects were not additive showing that they
are sharing the same mechanism. We suggest that NA acts on a smaller pool of Ca channels than L-AP4 and that L-AP4
acts in a pool of channels that engulfs the NA-sensitive pool of Ca
channels. Possibly the number of mGlu receptors is bigger than the
number of 2-adrenoreceptors, producing more free subunits
that can inhibit more Ca channels, or they are specifically targeted to
more Ca channels than the adrenergic receptors. In agreement with this
last hypothesis, NA did not inhibit the R-type current, on which
L-AP4 is effective (not shown) (Wu et al.
1998). Interestingly, at another calyx-type synapse in the
chick ciliary ganglion, NA has a very different effect on synaptic
transmission. In contrast to our present results in the calyx of Held,
NA potentiated the size of the EPSC and this was due to a
cGMP-dependent mechanism that increased the Ca2+
sensitivity of the exocytotic process (Yawo 1999).
In contrast to L-AP4, the effect of NA decays with
development. This suggests that the EPSC inhibition by NA may have some role in the development of this auditory synapse. At present, we do not
know if this effect disappears completely with age or if it becomes
restricted to some subset of cells. However, we emphasize that in some
P14 calyxes (an age when calyxes are morphologically mature)
(Kandler and Friauf 1993) NA was still able to produce a
strong inhibition of release. Interestingly, using transgenic animals
it has been shown that NA has a role in mouse brain development (Thomas et al. 1995) and that the
2D-adrenoreceptor subtype is particularly
important (Kable et al. 2000).
What may be the role of NA during development? The onset of hearing in
rats occurs at P12 (Blatchey et al. 1987), and an
important developmental change in this synapse is the marked reduction
of the size of the NMDA-EPSC with increasing age (Futai et al.
2001; Taschenberger and von Gersdorff 2000).
Here we report that there also occurs an almost parallel reduction of
the NA effect. In addition, we demonstrated that during high-frequency
stimulation the presynaptic inhibition of glutamate release makes the
postsynaptic cell fire more impulses. Presynaptic inhibition might thus
be of physiological relevance in the MNTB pathway during this critical period when the auditory brain stem adapts itself to high-frequency transmission. In addition, the presynaptic inhibition could decrease the amount of Ca entering the postsynaptic cell via NMDA receptors; this accounts for approximately 30% of the total Ca that enters the
principal cell during an EPSC in P8-P10 rats (Bollmann et al.
1998). Interestingly, we also observed that NA inhibits the somatic Ca currents of the MNTB principal cell (data not shown); this
accounts for approximately 70% of the total Ca that enters the cell
during an EPSC (Bollmann et al. 1998). Because calcium influx can affect gene expression in neurons (Gallin and
Greenberg 1995), the conjunction of these effects could
strongly regulate developmental changes.
Noradrenaline may thus be playing an important developmental role in the maturation of this synapse. These results are somewhat surprising given the canonical view of this synapse as operating solely as a fail-safe relay that simply follows its massive calyceal input. Contrary to this view, our results suggest that the output of this synapse can be modified by hormones and neuromodulators during development.
| |
ACKNOWLEDGMENTS |
|---|
We thank Drs. John Willians, Hitoshi Morikawa, Mary Palmer, Holger Taschenberger, and Larry Trussell for comments and helpful suggestions. R. M. Leão thanks J. Negrão for support and incentive.
This research was funded by National Institute on Deafness and Other Communication Disorders Grant RO1 DC-04274, an Alfred P. Sloan Research Scholar Award, and a Pew Biomedical Research Scholar Award.
| |
FOOTNOTES |
|---|
Address for reprint requests: H. von Gersdorff, The Vollum Institute, L-474, Oregon Health Sciences University, 3181 SW Sam Jackson Park Rd., Portland, OR 97201 (E-mail: vongersd{at}ohsu.edu).
Received 7 September 2001; accepted in final form 14 January 2002.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
subunit-mediated presynaptic inhibition: regulation of exocytotic fusion downstream of Ca2+ entry.
Science
292:
293-297, 20012-adrenoceptors control excitatory, but not inhibitory, transmission at rat hippocampal synapses.
J Physiol (Lond)
519:
439-449, 19992-autoreceptors control sympathetic transmitter release.
Eur J Neurosci
8:
1924-1931, 1996[ISI][Medline].2-adrenergic receptors.
J Pharmacol Exp Ther
293:
1-7, 20002-Adrenoceptors: more subtypes but fewer functional differences.
Trends Pharmacol Sci
15:
119-123, 1994[Medline].
, the bath perfusion of artificial cerebrospinal fluid
(ACSF) containing NA (2 µM). Experiment done in a slice from a P7
rat. B: comparison of the EPSC waveforms of the
experiment shown in A. Left: an average
of 4-5 EPSCs immediately before (
) and after
(
) the perfusion of NA. Right: both EPSCs
normalized to the peak. C: dose-response curve for the
inhibition of the EPSC peak amplitude by NA. The number of cells used
for each point is shown in parentheses. The points were fitted with a
logistic equation, which gave an IC50 value of 0.19 µM, a
Hill coefficient of 1.2, and a maximum inhibition value of 56.4%. Data
collected from slices of P6 (2), P7 (6), P8 (4), P9 (3), P10 (5), and
P14 (1) rats.
, the average of 3-5 traces
in control conditions; - - -, the average of the same number traces
obtained after bath perfusion of ACSF containing 20 µM NA. Recordings
done in a slice from a P6 rat. B: NA does not
affect presynaptic AP. Presynaptic AP evoked by afferent stimulation at
0.5 Hz before (
