Previous Article | Next Article 
The Journal of Neuroscience, February 1, 2002, 22(3):886-900
Activity-Dependent Presynaptic Effect of Serotonin 1B Receptors
on the Somatosensory Thalamocortical Transmission in Neonatal
Mice
Alban
Laurent1,
Jean-Marc
Goaillard1,
Olivier
Cases2,
Cécile
Lebrand2,
Patricia
Gaspar2, and
Nicole
Ropert1
1 Laboratoire de Neurophysiologie et Nouvelles
Microscopies, Institut National de la Santé et de la Recherche
Médicale (INSERM) EPI-0002, Ecole Supérieure de Physique et
de Chimie Industrielle, 75231 Paris cedex 5, France, and
2 INSERM U106, Bât. Pédiatrie, Hôpital
Pitié Salpétrière, 75651 Paris cedex 13, France
 |
ABSTRACT |
The disruptive effect of excessive serotonin (5-HT) levels on the
development of cortical sensory maps is mediated by 5-HT1B receptors,
as shown in barrelless monoamine oxidase A knock-out mice, in which the
additional inactivation of 5-HT1B receptors restores the barrels.
However, it is unclear whether 5-HT1B receptors mediate their effect on
barrel formation by a trophic action or an activity-dependent effect.
To test for a possible effect of 5-HT1B receptors on activity, we
studied the influence of 5-HT on the thalamocortical (TC) synaptic
transmission in layer IV cortical neurons. In TC slices of postnatal
day 5 (P5)-P9 neonate mice, we show that 5-HT reduces monosynaptic TC
EPSCs evoked by low-frequency internal capsule stimulation and relieves the short-term depression of the EPSC evoked
by high-frequency stimulation. We provide evidence that 5-HT decreases
the presynaptic release of glutamate: 5-HT reduces similarly the
AMPA-kainate and NMDA components and the paired pulse depression of TC
EPSCs. We show also that 5-HT1B receptors mediate exclusively the
effect of 5-HT: first, the effect of 5-HT on the TC EPSC is correlated
with the transient expression of 5-HT1B receptor mRNAs in the
ventrobasal thalamic nucleus during postnatal development; second, it
is mimicked by a 5-HT1B agonist; third, 5-HT has no effect in 5-HT1B
receptor knock-out mice. Our results show that in the developing barrel
field of the neonatal mice, 5-HT1B receptors mediate an
activity-dependent regulation of the TC EPSC that could favor the
propagation of high-frequency TC activity.
Key words:
primary somatosensory cortex; barrel field; 5- hydroxytryptamine; serotonin; CP93129; 5-HT1B receptor; 5-HT1D
receptor
| |
INTRODUCTION |
In the primary somatosensory (S1)
cortex of the rodent, the barrel field is a characteristic topographic
projection of the peripheral sensory receptors. Sensory inputs from
each mystacial vibrissa converge to a restricted zone in the cortical
layer IV called a barrel, and the barrels form an ordered map that is a point to point representation of the contralateral mystacial pad (Woolsey and Van der Loos, 1970
; Welker, 1971; Armstrong-James and Fox,
1987). The development of the barrel field requires a signal from the
whiskers during a critical period (Van der Loos and Woolsey, 1973;
Woolsey and Wann, 1976), but the nature of this signal is unknown. In
particular, the relative contributions of trophic factors and neuronal
activity is still disputed. The genetic inactivation of NMDA receptors
(Iwasato et al., 1997) prevents the barrel field formation, indicating
that glutamatergic transmission is a key signal, but trophic actions
seem also critical: normal barrels can form after blockade of activity
in the infraorbital nerve (Henderson et al., 1992) and in S1 cortex
(Chiaia et al., 1992; Schlaggar et al., 1993); the remodeling of the
barrel field induced by a peripheral lesion is maintained after
blockade of activity in S1 cortex (Chiaia et al., 1994), and the
thalamocortical (TC) axons segregate in S1 cortex after
selective cortical NMDA receptor inactivation (Iwasato et al.,
2000).
The importance of 5-HT for the barrel field development was first
suggested by the observation of a transient serotonergic hyperinnervation of the barrels in neonate rodents (Fujimiya et al.,
1986; D'Amato et al., 1987) and confirmed in monoamine oxidase A
(MAOA) knock-out mice in which excessive 5-HT levels during the
neonatal period prevent the formation of barrels (Cases et al., 1996;
Vitalis et al., 1998). A rescue of the barrel phenotype is observed in
the MAOA knock-out mice by the inactivation of the 5-HT1B receptor gene
(Salichon et al., 2001). These results show the importance of 5-HT1B
receptors that are transiently expressed by the TC terminals (Leslie et
al., 1992; Bennett-Clarke et al., 1993), but it is unclear how 5-HT and
the 5-HT1B receptors influence the formation of barrels. A trophic
action has been suggested because the depletion of 5-HT in neonates
causes an activity-independent reduction of barrel size (Rhoades et
al., 1998), and 5-HT1B receptor activation increases thalamic axonal
growth in vitro (Lieske et al., 1999; Lotto et al., 1999).
Neuronal activity may also play an important role: in neonates, the
5-HT1B receptor activation inhibits polysynaptic TC responses in S1
cortex (Rhoades et al., 1994). Finally thalamic neurons express the
5-HT plasmic transporter 5-HTT (Bennett-Clarke et al., 1996), and the
type 2 vesicular monoamine transporter VMAT2 (Lebrand et al., 1996;
1998), therefore TC axons could use 5-HT as a cotransmitter of glutamate.
To test the hypothesis of an activity-dependent effect of 5-HT on the
barrel development, we have studied the effect of 5-HT on the
monosynaptic TC EPSC in TC slice preparation of the neonate mice.
Preliminary results have been presented in abstract form (Lebrand et
al., 1998a; Laurent et al., 1999).
| |
MATERIALS AND METHODS |
Thalamocortical slice preparation. Slices were
prepared from wild-type C3H and 129SV mice and from 5-HT1B receptor
knock-out mice with a 129SV genetic background (Saudou et al., 1994),
and a C3H background was back-crossed for 10 generations. In early experiments, the animals were raised locally. Later gravid females were
obtained mostly from Iffa Credo (L'Arbresle, France) and sometimes
from Janvier (Le Genest Saint Isle, France), and the mice were born in
the local animal house. The day of birth was counted as postnatal day 0 (P0). Most experiments were done with neonate mice (P7 ± 1 d; n = 124; range, P5-P9). Some controls were done
with older mice (P23 ± 4 d; n = 22; range,
P18-P29). The animals were anesthetized by intraperitoneal injection
of pentobarbital (15 mg/kg) and decapitated. The brain was rapidly removed and placed in oxygenated (5% CO2, 95%
O2) cold artificial CSF (ACSF). The TC
slices were cut (300 µm thickness) with a vibratome (DSK DTK-1000 or
VT1000S; Leica, Nussloch, Germany) as previously described
(Agmon and Connors, 1991). The slices were maintained 1 hr at 34°C
and later at room temperature (25°C) in oxygenated standard ACSF with
5-10 mM lactate (Izumi et al., 1994).
Electrophysiological recordings and data analysis. The
slices were placed in a small recording chamber (1 ml), maintained by a
platinum grid, and submerged in a constant (2 ml/min) ACSF perfusion at
25°C. Renewal of the medium was achieved within 1 min as shown by the
bath application of GABAA, AMPA-kainate (KA), or
NMDA antagonists that induced a maximal effect within 1 min. The S1
cortex and the layer IV neurons were recognized in the recording
conditions by visualization of the large dorsal barrels (Fig.
1) using an upright fixed stage
microscope (Axioskop FS; Zeiss, Oberkochen, Germany) with Nomarski
optics and an infrared video camera (Newvicon C2400; Hamamatsu,
Shizouka, Japan). Stainless steel bipolar semimicroelectrodes (100 µm
tip diameter, 250 µm intertip distance; Rhodes Medical Instruments)
were placed in the internal capsule (IC) near the thalamic border to
activate the TC axons. The frequency of the IC stimulation to obtain
stable responses was 0.07 Hz. The intensity of the stimulation was
adjusted to evoke a unitary TC EPSC (Fig.
2). Whole-cell voltage- and current-clamp recordings were obtained with Axopatch 1D or 200A amplifiers (Axon Instruments, Foster City, CA) from the soma of S1 cortical neurons (Stuart et al., 1993). Recording pipettes were pulled from cleaned borosilicate glass tubes and coated with beeswax. The tip resistance was 4-6 M
. During recording, the series resistance
(Rs) was not compensated and was
monitored continuously by applying short (20 msec) negative voltage
steps (
1 to 3 mV) before each IC stimulation (Rs = 23 ± 7 M;
n = 94). Recordings were discarded if the
Rs increased >20%. The voltage and
current signals were filtered (5 kHz), digitized (10 kHz), and stored
directly on the computer by a Labmaster TL-1 DMA acquisition board
(Axon Instruments). The data were analyzed using programmable software
(Acquis1; Biologic, Claix, France). The latency of the response was
obtained by measuring the time between the beginning of the stimulation
artifact and a threshold value of the first derivative of the response.
The amplitude of the TC EPSC was measured between the baseline and the
peak of the response; the baseline was obtained by averaging the signal
for 1 or 2 msec between the stimulation artifact and the onset of the
response; the peak of the response was measured by detection of the
maximal absolute value and averaging 5 or 7 data points around this
value. The time constant of decay of the EPSC was obtained by fitting
the decay of the response with a double exponential function. In some
cases, a single time constant 
was found. When the decay was
biexponential, an estimation of the duration of the responses was
obtained by calculating a weighted time constant = 1 × [A1 
(A1 + A2)] + 2 × [A2 (A1 + A2)], 1 and
2 being the fast and the slow time constants
respectively, and A1 and A2 the amplitude of the
fast and slow components, respectively. To quantify the responses to
5-HT and to the 5-HT1B agonist CP93129, the relative amplitude of the
TC EPSCs was calculated by normalization of the individual values to
the mean value obtained before drug application. The maximal effect of
the drug was estimated by calculating the running average
(n = 10) of the relative amplitude of the TC EPSC and
taking the minimum reached by the running averaged at the end of the
drug application. Summary data were obtained by averaging the relative
amplitude of the individual TC EPSCs obtained from several cells in the
same experimental conditions. Paired pulse and trains of high-frequency
IC stimulation were used to assess the short-term changes of the TC
EPSC; the interval between the stimuli was 20 msec, and five stimuli
were applied in a train; we measured the amplitude of the AMPA
component of the EPSC recorded at 80 mV. The short-term change
induced by the paired pulse IC stimulation was quantified by the paired
pulse ratio (PPR) equal the amplitude ratio of the second response to the first response
(a2/a1).
The short-term changes induced by the high-frequency trains of IC
stimulation were quantified by estimation of the relative EPSC
amplitude (R2 to
R5) calculated as the amplitude ratio
of the second to the fifth EPSCs to the first EPSC
(R2 = a2/a1;
R3 = a3/a1;
R4 = a4/a1;
R5 = a5/a1).
Results are given as mean ± SD for statistical analysis,
and mean ± SEM for graphical representations of the
summary data. The variability of the response latency was estimated by
the coefficient of variation CV = SD mean. The
significance levels were calculated using paired or unpaired Student's
two-tailed t test.
View larger version (68K):
[in this window]
[in a new window]
|
Figure 1.
TC slice preparation and identification by
infrared videomicroscopy and spike discharge of layer IV neurons in the
primary somatosensory (S1) cortex of the neonate mouse.
A, The TC slice preparation of a P5 mouse is visualized
at a low magnification by infrared videomicroscopy in recording
conditions (1) and schematized for structure
identification (2). The S1 cortex with six layers
(I-VI) is recognized by the presence of barrels in layer IV. The
larger barrels of the PMBSF are located dorsally to hippocampal regions
[Ammon's horn: CA1, and CA3 and fascia dentata (FD)]. A
bipolar stimulating electrode (Stim) is placed in the
internal capsule (IC) to activate the TC axons
originating in the ventrobasal (VB) thalamic nucleus and
projecting through the nucleus reticularis (Re) to layer
IV in S1 cortex. The recording electrode (Rec) is
located over a PMBSF barrel. B, View of S1 cortex at an
intermediate magnification with the tip of the recording pipette over a
large PMBSF barrel in layer IV. C, View at a higher
magnification of a layer IV neuron with a nonpyramidal soma and no
apical dendrite. Pial orientation is identical in B and
C. D, E, Current-clamp recordings of the spike discharge
evoked by current injection in a fast-spiking (D)
and a regular-spiking (E) layer IV neuron. The
bottom traces illustrate the current steps that were
injected for 1 sec. The fast-spiking neuron exhibits a maintained spike
frequency. The so-called regular-spiking neuron shows a strong spike
adaptation.
|
|
View larger version (32K):
[in this window]
[in a new window]
|
Figure 2.
Identification of the
AMPA- KA and NMDA components of the monosynaptic TC EPSC
recorded in layer IV S1 cortex of P6-P8 wild-type mice. The IC is
stimulated at 0.07 Hz. A, A minimal IC stimulation
evoked a unitary TC EPSC at +30 mV holding potential. The IC
stimulation intensity was changed from 10 to 20 V at intervals of 1 V
to find the intensity that evokes unitary responses. 1,
2, Five individual (top traces) and averaged
(bottom traces) TC EPSCs evoked by a subthreshold
(1, 12V) and a minimal
(2, 16 V) intensity.
3, In the same cell, the individual
(squares), mean (circles,
n = 5), and SD of the peak amplitude of the TC
EPSCs are plotted against stimulation intensity. From 10 to 12 V,
responses fluctuate around baseline (1.65 ± 5.38 pA;
n = 15). At 13 V, responses were seen, and
the failure rate was high (four of five). From 14 to 20 V, the failure
rate was reduced (3 of 35), and the amplitude of the TC EPSCs fluctuate
around a constant plateau value (23.4 ± 4.9 pA;
n = 32), as expected for a unitary EPSC.
B, The IC stimulation evoked a monosynaptic EPSC.
1, Superimposition of 10 individual TC EPSCs at 80 and
+30 mV. The latency of the response was identical at both potentials
and showed little variability. 2, In the same cell, the
response latency measured at 80 mV was 6.28 ± 0.18 msec
(n = 30) and followed a Gaussian distribution as
expected for a monosynaptic response. C,
I-V curves of the early and late components of the TC
EPSC. 1, Averaged (n = 10) TC EPSCs
at several holding potentials (V) between
80 and +40 mV with a reversal potential near 0 mV. The kinetics of
the TC EPSCs was changed with voltage, being faster at 80 mV
than at more positive potentials. The amplitude of the TC EPSCs
was measured at the peak of the early response at 80 mV
(a) and at the peak of the late response at +40
mV (n). 2, The I-V
curve of the early component (a) was almost
linear with a reversal at 0.43 mV. The I-V curve of
the late component (n) reversed at 4 mV.
D, Effect of CNQX and AP-5 on the TC EPSCs. Averaged
(n = 5) TC EPSCs at +30, 0, and 80 mV in control,
in 10 µM CNQX, and in 10 µM CNQX plus 100 µM D,L-AP-5. The fast and the slow inward
currents seen at 80 mV in standard ACSF were blocked by the AMPA-KA
antagonist CNQX. The peak amplitude of the outward response at +30 mV
was unchanged in CNQX and blocked by the NMDA antagonist AP-5. A small
triphasic residual response is seen in CNQX plus AP-5: it precedes the
TC EPSC, and it is insensitive to the holding potential and corresponds
most likely to the presynaptic TC fiber activity.
|
|
In the neonate mice, we confirmed that a minimal IC stimulation evoked
a TC EPSC without disynaptic IPSCs, as shown by the absence of a
delayed outward postsynaptic response at 0 mV. In older mice, a
disynaptic GABAA mediated IPSC was evoked by IC stimulation. The IPSC was identified by a longer and more variable latency than the EPSC and by a reversal potential near the chloride equilibrium (40 mV in our recording conditions). To block the disynaptic IPSC, a GABAA antagonist (10 µM BIC or 100 µM PTX) was added to the
standard ACSF. In this condition, it was necessary to add AMPA-KA or
NMDA antagonist (in µM: 10 CNQX, 50 D-AP-5, or 100 D,L-AP-5) to reduce the spontaneous synchronized
discharges (Gutnick et al., 1982), and to record in isolation either
the NMDA or the AMPA component of the TC EPSC.
Solutions and pharmacological compounds. The external
standard ACSF contained (in mM): NaCl 126, KCl
1.5, KH2PO4 1.25, MgSO4 1.5, CaCl2 2, NaHCO3 26, and glucose 10; osmotic pressure,
298 ± 3 mOsm. The following drugs were applied in the perfusion:
5-hydroxytryptamine (5-HT) creatinine sulfate (0.5-20
µM; Sigma, St. Louis, MO);
D,L-2-amino-5-phosphonopentanoic acid
(D,L-AP-5) (100 µM;
Tocris Cookson, Ballwin, MO); D-AP-5 (50 µM; Tocris);
6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) (10 µM; Tocris); picrotoxin (PTX) (100 µM; Research Biochemicals, Natick, MA);
bicuculline metochloride (BIC) (10 µM; Sigma);
1,4-dihydro-3-(1,2,3,6-tetrahydro-4-pyridinyl)-5H-pyrolo[3,2-b]pyridin-5-one (CP93129 dihydrochloride) (100 nM; a gift from
Pfizer, Groton, CT). The pipette solution contained (in
mM): (1) for whole-cell voltage-clamp recordings,
Cs gluconate 120, CsCl 10, K-ATP 4, MgCl2 2, HEPES 10, Na-GTP 0.4, EGTA(CsOH) 0.2, pH 7.35, adjusted with CsOH;
osmotic pressure, 273 ± 2 mOsm. (2) For whole-cell current-clamp
recordings, K gluconate 144, MgCl2 3, HEPES 10, EGTA(KOH) 0.2, Na2-ATP 2, Na-GTP 0.2, pH 7.35, adjusted with KOH. The potentials were corrected for the liquid
junction potential (10 mV). For extracellular field recordings, we
used large (1-2 M) patch pipettes that were filled with ACSF.
In situ hybridization. Swiss OF1 and C3H mice from Iffa
Credo were analyzed at P0, P7, P10, P15, and as adults. Fresh frozen sections (15 µm) were cut serially in the coronal plane. Alternate sections were used for Nissl staining and for in situ
hybridization of the 5-HT1B and 5-HT1D receptors. cDNA fragments of the
5-HT1B and 5-HT1D receptor genes (kind gift of Luc Marotaux CNRS,
Strasbourg, France), containing the full-length coding sequence,
were subcloned in pBluescript KS vectors (Stratagene, La Jolla, CA).
5-HT1B and 5-HT1D receptor plasmids were linearized with
EcoRI (Roche Diagnostics) for antisense RNA synthesis and
with XbaI (Roche Diagnostics) for sense synthesis. The
in vitro transcription was performed using the Promega
(Madison, WI) kit, and probes were labeled with 35S-UTP (>1000 Ci/mmol; Amersham,
Arlington Heights, IL). Tissue sections were post-fixed for 15 min in
4% paraformaldehyde, washed in PBS, acetylated, dehydrated, air-dried,
and hybridized with 106 cpm overnight in a
humid chamber at 50°C. Sections were sequentially treated 30 min at
42°C in 5× SSC, 0.1% dithiothreitol (DTT); 20 min at 60°C
in 2× SSC, 50% formamide, 0.5% DTT; 30 min at 37°C in RNase buffer
(23% NaCl, 10 mM Tris, 5 mM EDTA) containing 0.02% RNase A (Roche
Diagnostics); 15 min at 37°C in RNase buffer; 15 min at 37°C in 2×
SSC; and 15 min at 37°C in 0.1× SSC. The slides were dehydrated,
air-dried, dipped in photographic emulsion (NTB2; Eastman Kodak,
Rochester, NY), and exposed for ~10 d. After development of the
emulsion, the sections were counterstained with cresyl violet.
| |
RESULTS |
Characterization of the monosynaptic TC EPSC
Experiments were done in neonate mice (P5-P9) with two different
genetic backgrounds (C3H and 129SV). The results were similar in both
backgrounds and were pooled. The recordings were made in S1 cortex
identified by the presence of barrels visualized by infrared
videomicroscopy in the recording conditions (Agmon and Connors, 1991;
Fleidervish et al., 1998; Feldmeyer et al.,1999). The recordings were
limited to the soma of the layer IV neurons in the large barrels of the
posteromedial barrel subfield (PMBSF) that correspond to the vibrissae
of the posterior mystacial pad, and they were restricted to neurons
with a nonpyramidal soma (Fig. 1). In preliminary experiments, we
characterized the neurons in the layer IV using current-clamp
recordings and potassium as the main cation of the pipette solution
(see Materials and Methods). The neurons (n = 34) were
identified by their spike discharge in response to current injection
(McCormick et al., 1985; McCormick and Prince, 1987; Chagnac-Amitai and
Connors, 1989; Feldmeyer et al., 1999; Lubke et al., 2000). Most
neurons (n = 29) displayed long-duration spike
(2.18 ± 0.46 msec), small afterhyperpolarization (AHP) (12 ± 5 mV), and strong spike adaptation, typical of the regular spiking
neurons (Fig. 1E), indicating that they are probably spiny glutamatergic stellate neurons. The remaining neurons
(n = 5) displayed significant (p < 0.01) shorter duration spikes (1.54 ± 0.16 msec), larger AHP
(19 ± 2 mV), and weak spike adaptation typical of the fast
spiking GABAergic neurons (Fig. 1D). In the subsequent voltage-clamp recordings it was necessary to use cesium in
the pipette solution to record synaptic responses at potentials more
positive than 70 mV. These conditions do not allow the identification of neurons from their spike discharge. However, based on our
preliminary experiments, voltage-clamp recordings were probably
obtained from a heterogeneous population of layer IV neurons with a
majority of spiny stellate neurons and a minority of GABAergic neurons. In all the neurons that we recorded in the layer IV, we found that the
IC-evoked EPSCs were sensitive to 5-HT, therefore the data from all
neurons were pooled.
The low-frequency stimulation (0.07 Hz) of the IC evoked stable
postsynaptic current responses in layer IV neurons (Fig. 2). When the
intensity of the IC stimulation was increased above the threshold, the
amplitude of the responses fluctuated around a constant value (Fig.
2A) that corresponds to a unitary response caused by
the release of glutamate by a single presynaptic axon (Stern et al.,
1992). As an attempt to detect a possible input specific effect of
5-HT, the intensity of the IC stimulation was chosen to evoke a unitary
response. The latency of the response was relatively long (6.09 ± 1.14 msec; n = 106), as expected for neonate mice with
partially myelinated axons (Jacobson, 1963). The latency distribution
calculated for individual cells could be fitted with a single Gaussian
with weak variability (CV = 0.06 ± 0.04; n = 50), indicating that the IC-evoked responses were monosynaptic (Fig.
2B). In adult rats, the layer VI pyramidal neurons of
S1 cortex project through the IC to the thalamus (Bourassa et al.,
1995) and send axon collaterals to layer IV (Zhang and Deschenes,
1997). Therefore, the IC stimulation may activate the corticothalamic
(CT) pathway, giving rise to antidromic spikes in the layer VI
pyramidal neurons, and evoking a monosynaptic response in layer IV
(Agmon and Connors, 1991). To identify the pathway that gives rise to
the monosynaptic responses in layer IV in our TC slice preparation, we
tested the ability of the IC stimulation to evoke an antidromic action
potential in the layer VI pyramidal neurons. In seven TC slices, the
threshold of monosynaptic unitary responses in layer IV neurons was
first measured in response to IC stimulation (Fig.
2A). In the same slices, layer VI pyramidal neurons
of the same cortical column were recorded in the current-clamp mode and
identified by their morphological aspect in infrared videomicroscopy
and by their regular spike discharge. After the IC stimulation, a
monosynaptic EPSP was recorded in several cells (n = 7), but no antidromic spike discharge was seen in layer VI pyramidal
neurons (n = 16) even when the IC stimulation intensity was increased up to 10 times that necessary to evoke a unitary EPSC in
layer IV. We also used extracellular field recordings to further
characterize the layer IV cortical responses evoked by IC stimulation.
As reported before (Agmon and Connors, 1991), the layer IV
intracellular monosynaptic EPSCs were correlated with an extracellular
slow negativity preceded by an extracellular fast negativity caused by
the TC fiber spike discharge (data not shown). In the infragranular
layers V/VI of the same slices, we did not record an extracellular fast
negativity caused by antidromic spike discharge of pyramidal neurons
and preceding the layer IV EPSC. These results indicate that the
probability to evoke an antidromic spike discharge in the layers V/VI
pyramidal neurons is probably very weak in our TC slice preparation and
that the layer IV monosynaptic EPSCs evoked by IC stimulation are more likely attributable to the activation of the ascending TC axons.
The monosynaptic TC responses to IC stimulation were further identified
by their physiological and pharmacological properties (Fig. 2). The
kinetics of the TC responses was sensitive to the membrane potential
being faster at 80 mV and slower at potentials more positive than
60 mV (Table 1). The decay of the TC
responses at 80 mV could be fitted with a two exponential function
(1 = 4 ± 1 msec, 90 ± 5%;
2 109 ± 98 msec) with a weighted time constant of 14 ± 13 msec (n = 42) (see Materials
and Methods). Similar fast and slow values were obtained for the AMPA,
the kainate (KA) (Kidd and Isaac, 1999), and the NMDA components of the
TC EPSC (Barth and Malenka, 2001). As reported before (Barth and Malenka, 2001), the decay of the TC responses at +30 mV was fitted either with a single or with a two exponential function with a time
constant of 175 ± 33 msec (n = 13) significantly
different from the decay at 80 mV (p < 0.001). The reversal potential of the TC responses was near 0 mV and
corresponds to the reversal potential of the NMDA and AMPA-kainate
responses in our recording conditions. The early component of the TC
EPSC was blocked by CNQX, an AMPA-kainate receptor antagonist
(n = 3) (Fig. 2D) and its current-voltage
(I-V) curve (measured at the peak of the response at
80 mV) was linear (Fig. 2C2a), as reported for AMPA
receptors that do not express polyamine block (Ozawa et al., 1998). At
80 mV both the NMDA and the KA receptors appear to contribute
to the slow component of the TC EPSC, as shown by the antagonistic effects of AP-5, an NMDA receptor blocker, in a majority of cells (n = 7 of 11) and of CNQX in the remaining cells
(n = 4 of 11). As expected from the strong
rectification of KA receptors at +30 mV (Kidd and Isaac, 1999), the
NMDA receptors contribute exclusively to the slow component of the TC
EPSC that was completely blocked by AP-5 at +30 mV (n = 13). The I-V curve of the slow component of the TC EPSC
(measured at the peak of the response at +30 mV) expressed an inward
rectification with a maximum near 30 mV (Fig. 2C2n)
reminiscent of the NMDA receptor I-V curve. These results confirm that the activation of the TC axons by the IC stimulation activates the AMPA-KA and the NMDA glutamate receptors (Agmon and
O'Dowd, 1992; Kidd and Isaac, 1999). The relative contribution of the
NMDA component was estimated by calculating the ratio of the peak
amplitude of the responses at +30 and at 80 mV. Confirming previous
results (Crair and Malenka, 1995), we found a major contribution of the
NMDA component in neonate mice (NMDA-AMPA ratio = 1.64 ± 1.23; n = 82). In CNQX and AP-5, the TC EPSC was
completely blocked, indicating that, in our conditions, there was no
contribution of a serotonergic postsynaptic response to the EPSC
(Roerig et al., 1997). In several experiments (n = 21),
we recorded a small triphasic response that preceded the EPSC (Fig.
2D). This response was probably caused by the TC
afferent fiber discharge: it had a lower threshold than the TC EPSC, it
was synchronized with the early fast negative wave of the layer IV
extracellular field, and it was voltage-insensitive and followed 50 Hz
stimulation.
View this table:
[in this window]
[in a new window]
|
Table 1.
Physiological properties of the TC EPSC in neonate and
juvenile wild-type and in neonate 5-HT1B receptor knock-out mice
|
|
In the neonate mice (Fig. 2), we confirmed that the minimal IC
stimulation evoked a unitary TC EPSC without disynaptic
GABAA-mediated IPSCs (Agmon and O'Dowd, 1992;
Barth and Malenka, 2001). In older mice, disynaptic IPSCs were evoked
by IC stimulation (see Materials and Methods). Therefore, in the
neonatal mice experiments were done in standard ACSF, whereas
GABAA antagonists were needed in older mice to
record the EPSC in isolation.
Presynaptic inhibitory effect of 5-HT on the monosynaptic TC EPSC
in neonatal mice
At the beginning of each recording session, the responses at 0 mV
were first examined to control the absence of an evoked IPSC (Fig. 2).
The effect of 5-HT was then tested on the larger NMDA component of the
unitary TC EPSC recorded at +30 mV. After a control period of at least
10 min to establish the stability of the TC EPSC, 5-HT (0.5-20
µM) was applied in the bath for 5 min before returning to
standard ACSF. Increasing the duration of the application to 10 min did
not increase the effect of 10 µM 5-HT. Therefore, 5 min
applications were used as a standard.
In the neonate mice, the application of 1-20 µM 5-HT had
no effect on the holding current of the layer IV neurons but induced a
significant (p < 0.0001) reduction of the peak
amplitude of the NMDA component of the TC EPSC (Fig.
3). The inhibitory effect of 5-HT on the
TC EPSC was reversible and was enhanced with increased concentrations
of 5-HT in the ACSF. We found no significant effect of 0.5 µM 5-HT. At higher concentrations, the effect
of 5-HT was increased with the concentration (Fig. 3C). We
did not attempt to obtain the IC50 of the effect
of 5-HT because the widespread expression of the 5-HTT in the neonate
and in the older S1 cortex (Lebrand et al., 1998b) limits the diffusion
of 5-HT, and one cannot estimate the actual concentration of 5-HT at
the receptors in the slice preparation. At relatively high
concentrations of 5-HT (
10 µM), we saw an
increase of the frequency of spontaneous unitary EPSCs and IPSCs,
indicating that 5-HT has several actions in S1 cortex. In the present
study, we focused on the effect of 5-HT on TC transmission.
View larger version (27K):
[in this window]
[in a new window]
|
Figure 3.
Inhibitory effect of 5-HT on the TC EPSC
of neonate wild-type mice. The effect of bath applications of 5-HT at
several concentrations has been tested on the NMDA component of the TC
EPSC evoked by IC stimulation at 0.07 Hz in two neurons
(A, B) maintained at +30 mV.
A, Inhibitory effect of 1 and 5 µM 5-HT
applied to the same neuron of a P7 mouse. 1,
2, TC EPSCs were averaged (n = 5) in
control (thick line) and during the response to the
successive applications to 5-HT (dotted line) at 1 (1) and 5 µM
(2). The applications of 5-HT lasted 5 min, and
the second application was made 20 min after the first.
3, The amplitude of each TC EPSC was normalized to the
control value (107 ± 14 pA; n = 40) before
the first 5-HT application and plotted against time as individual
(dots) and running average (n = 10;
continuous line). After a 10 min control period, 1 µM 5-HT applied for 5 min induced a small reduction of
the TC EPSC. The minimal relative EPSC at the end of the application
(0.78 ± 0.22; n = 10) was significantly
(p < 0.0001) smaller than the control
(1.00 ± 0.13; n = 40). After recovery, 5 µM 5-HT applied for 5 min induced a larger reduction of
the TC EPSC with a minimal relative value (0.54 ± 0.24;
n = 10) significantly different from the control
(p < 0.0001) and from the response at 1 µM (p < 0.05).
B, Inhibitory effect of 20 µM 5-HT on the
TC EPSC of another neuron of a P5 mouse. 1, The TC EPSCs
were averaged (n = 45) in control
(thick line) and during the response to 5-HT
(dotted line). 5-HT was applied for 5 min.
2, The same traces in control and in 5-HT were
normalized at their peak for a comparison of their kinetics. The
inhibitory effect of 5-HT did not affect the shape of the EPSC.
3, In the same cell, the time course of the inhibitory
effect of 5-HT is plotted against time. The peak amplitude of
individual TC EPSCs was normalized to the mean value in control and
plotted against time as individual (dots) and running
average (n = 10; continuous line).
The relative amplitude of the TC EPSC reached a minimal value within 1 min of 5-HT application that was maintained during the rest of the
application. The amplitude of the TC EPSC was significantly
(p < 0.0001) smaller in 5-HT (39 ± 24 pA; n = 45) than in control (138 ± 31 pA;
n = 45). Complete recovery was obtained within 30 min of wash. C, Summary data of the dose-response curve
of the inhibitory effect of 5-HT on the TC EPSC. The maximal inhibition
of the TC EPSC was 0.33 ± 0.09 (n = 3) at
1-2 µM; 0.55 ± 0.25 (n = 13)
at 5 µM; 0.62 ± 0.13 (n = 5) at
10 µM; and 0.74 ± 0.23 (n = 6)
at 20 µM.
|
|
The inhibitory effect of 5-HT on the NMDA component of the TC EPSC
could be attributable either to a postsynaptic action of 5-HT on the
glutamate receptor responsiveness or to a presynaptic action of 5-HT on
the release of glutamate. To evaluate these two possibilities, we
compared the effect of 5-HT on the NMDA and the AMPA components of the
TC EPSC, and we studied the effect of 5-HT on the paired pulse ratio.
A postsynaptic action of 5-HT is expected to change differentially the
NMDA and the AMPA-KA components of the TC EPSC. To demonstrate a
possible differential effect of 5-HT on the AMPA-KA and the NMDA
components of the TC EPSC we used two strategies: first, we compared
the kinetics of the TC EPSC in control and in 5-HT; second we compared
the effect of 5-HT on the AMPA-KA and on the NMDA components of the TC EPSC.
The kinetics of the TC EPSC was first estimated visually by
normalization and superimposition of averaged traces in control and in
5-HT (Fig. 3B2; see Fig. 5A2,B2); we found a
complete overlap of the traces in control and in 5-HT in 10 cells
recorded at 80 mV and in 10 different cells recorded at +30 mV. The
kinetics of the TC EPSC was also quantified by the measurement in 10 cells of the time constant of decay of the TC EPSC recorded at 80 mV in control and in 5-HT. At 80 mV the AMPA-KA and the NMDA receptors contribute to the decay of the EPSC. When the relative amplitude of the
EPSC was reduced by 5-HT to a minimum (0.25 ± 0.17;
n = 10), the weighted time constant of decay of the
EPSC was not significantly different (4.96 ± 0.77 and 5.02 ± 0.82 msec; n = 10, in control and in 5-HT, respectively).
To compare the effect of 5-HT on the NMDA and AMPA-KA components of
the TC EPSC, we also quantified the relative inhibitory effect of 5-HT
on the TC EPSC recorded at +30 and at 80 mV. As discussed earlier
(Fig. 2), the NMDA receptors and the AMPA-KA receptors contribute
exclusively to the peak of the EPSC recorded at +30 and at 80 mV,
respectively. In one single neuron, the effect of 5 µM
5-HT was tested at both potentials (Fig.
4). We found that the minimal relative
amplitude of the TC EPSC reached at the end of the 5-HT application was
similar at +30 mV (0.16 ± 0.28) and at 80 mV (0.28 ± 0.19), indicating that the NMDA and the AMPA-KA components of the TC
EPSCs were similarly reduced by 5-HT. However comparing the effect of
5-HT at +30 and 80 mV in the same neuron requires excellent recording
conditions for at least 90 min that are very difficult to obtain in the
neonate. Therefore, we compared the effect of 10 µM 5-HT
in two separate samples of layer IV neurons: six cells were recorded at
80 mV, and four cells at were recorded at +30 mV (Fig.
5). We found that the minimal relative
amplitude of the EPSC obtained in 10 µM 5-HT was similar
at 80 and at +30 mV (0.25 ± 0.30, n = 6;
0.22 ± 0.15, n = 4, respectively) confirming that
5-HT reduces similarly the AMPA-KA and the NMDA components of the TC
EPSC. These results support the idea that the inhibitory effect of 5-HT
on the TC EPSC is attributable to a presynaptic reduction of the
release of glutamate by the TC axons.
View larger version (43K):
[in this window]
[in a new window]
|
Figure 4.
Comparison of the inhibitory effect of 5 µM 5-HT on the NMDA and the AMPA components of the TC
EPSC. The NMDA (1) and the AMPA
(2) components of the TC responses were recorded
in the same neuron at +30 and at 80 mV, respectively. At each
potential, the TC EPSCs were recorded in control, in 5 µM
5-HT applied for 5 min, and during recovery. Sequential individual TC
EPSCs (n = 5) at +30 mV (1)
and at 80 mV (2). 3, The
amplitude of the NMDA and AMPA components were normalized to the
responses in control (1.00 ± 0.15, n = 50;
1.00 ± 0.17, n = 30; at +30 and 80 mV,
respectively), and the relative amplitude of the EPSC was plotted
against time as individual values (circles) and running
average (n = 9; continuous line) at
+30 and 80 mV. 5-HT induced a significant reduction
(p < 0.0001) of the TC EPSC at +30 and 80
mV. The minimal relative amplitude of the TC EPSC reached in 5-HT was
similar at +30 and 80 mV (0.16 ± 0.28 and 0.28 ± 0.19;
n = 10, respectively).
|
|
View larger version (41K):
[in this window]
[in a new window]
|
Figure 5.
Inhibitory effect of 10 µM 5-HT on
the AMPA and NMDA components of the TC EPSC. The effect of 10 µM 5-HT applied for 5 min was compared in different layer
IV neurons recorded at 80 or at +30 mV. A, In one
neuron, the effect of 10 µM 5-HT was tested on the AMPA
component of TC EPSC recorded at 80 mV. B, In another
neuron, the effect of 10 µM 5-HT was tested on the NMDA
component of the TC EPSC recorded at +30 mV. A,
B, Traces in control (thick line) and
during the responses to 5-HT (dotted line) were averaged
(n = 20) superimposed (1),
and normalized (2). The inhibitory effect of 5-HT
(1) did not change the kinetics of the TC EPSC at
80 or at +30 mV (2). 3, The
amplitude of the EPSC was normalized to the response in control, and
the relative amplitude was plotted against time as individual values
(dots) and running average (n = 10;
continuous line). In each neuron, the relative amplitude
of the TC EPSC was significantly (p < 0.0001) reduced at both potentials at the end of the 5-HT application
(from 1.00 ± 0.21, n = 50 to 0.52 ± 0.25, n = 10 at 80 mV; and from 1.00 ± 0.25, n = 50 to 0.28 ± 0.30, n = 10 at + 30 mV). C, Summary data
(mean ± SEM) of the inhibitory effect of 10 µM 5-HT
on the AMPA (1; n = 6 cells) and
NMDA (2; n = 4 cells) components of
the TC EPSC. The minimal relative amplitude of the TC EPSC at the end
of the 5-HT application was not significantly different at 80 mV
(0.25 ± 0.30; n = 6) and at +30 mV (0.22 ± 0.15; n = 4).
|
|
When a synapse is activated twice at short intervals, a reduction
[paired pulse depression (PPD)] or a facilitation [paired pulse
facilitation (PPF)] of the second EPSC is often seen. PPD and PPF are
sensitive to presynaptic changes of transmitter release and insensitive
to changes of the postsynaptic receptor sensitivity (Manabe et al.,
1993) when AMPA receptors do not express polyamine block (Rozov and
Burnashev, 1999). We have seen before that the early component of the
TC EPSC displays a linear I-V curve typical of AMPA
receptors insensitive to polyamines (Fig. 2). The absence of inward
rectification is probably not attributable to the absence of spermine
in the pipette solution because the dissipation of the polyamine is
slow in the whole-cell configuration, and inward rectification is
recorded in such condition (Kidd and Isaac, 1999; Métin et al.,
2000). Therefore, the short-term changes of the EPSC depend mainly on a
presynaptic modification of the amount of glutamate release. PPD and
PPF can be quantified by the paired pulse ratio (PPR) that is the
amplitude ratio of the second response to the first response
(a2/a1).
To estimate the PPR in control and in 10 µM
5-HT, two IC stimuli were applied at 20 msec intervals in six cells
(Fig. 6). We confirmed that in the
neonate TC slice preparation, it was possible to record paired pulse
monosynaptic TC EPSCs in isolation without disynaptic IPSCs by
examining the responses at 0 mV. The AMPA component of the TC EPSC was
recorded at 80 mV. In control, we saw a PPD of the TC EPSC (PPR = 0.75 ± 0.04; n = 6) similar to that reported in
mature animals (Gil et al., 1997, 1999). The inhibitory effect of 10 µM 5-HT on the first TC EPSC (39 ± 18 pA
in control and 12 ± 10 pA in 5-HT; n = 6) was
associated with a PPF of the TC EPSC (PPR = 2.23 ± 1.05; n = 6). After removal of 5-HT, full recovery of the
first EPSC amplitude and of the PPD were obtained (PPR = 0.88 ± 0.15; n = 6). These results are consistent with a
presynaptic inhibitory action of 5-HT on the release of glutamate by
the TC relay neurons.
View larger version (37K):
[in this window]
[in a new window]
|
Figure 6.
Activity-dependent effect of 5-HT on the
TC EPSC. A, A 10 µM concentration of 5-HT
converts the paired pulse depression of the control into a paired pulse
facilitation. 1, TC EPSCs evoked by two IC stimuli
applied at 20 msec interval every 15 sec and recorded at 80 mV. The
averaged (n = 20) TC EPSCs are shown in control
(thick trace), during the response to 10 µM 5-HT (dotted trace), and after recovery
(thin trace). 2, 3, In the
same neuron, the amplitude of the first
(a1) and the second
(a2) TC EPSCs are plotted against
time; individual (dots) and running average
(n = 10; continuous line) are shown.
The application of 10 µM 5-HT for 5 min induced a
significant (p < 0.0001) reduction of the
first EPSC without changing the second EPSC
(a2), and the PPD of the control was
converted into a PPF in 5-HT. 4, The PPR calculated from
the running average (n = 10) is plotted against
time. The maximal PPR in 5-HT (2.33 ± 0.34; n = 10) was significantly (p < 0.0001)
different from the PPR in control (0.82 ± 0.11;
n = 45). 5, Summary data of the
effect of 10 µM 5-HT applied for 5 min on the PPR in
six cells. Individual (continuous lines) and mean
(discontinuous line) PPRs are plotted in control during
the inhibitory effect of 5-HT and after recovery. The inhibitory effect
of 5-HT reduced the EPSC to a minimal normalized value of 0.28 ± 0.12 (n = 6) and changed the PPD of the control
into a PPF in 5-HT. The PPR of the control (0.75 ± 0.04;
n = 6) was significantly
(p < 0.01) different from the PPR in 5-HT
(2.23 ± 1.05; n = 6). After recovery the PPR
(0.82 ± 0.15; n = 6) was not different from
the control. B, 5-HT changes the short-term depression
of the control into a short-term facilitation. 1, Averaged (n = 40) TC EPSC in response to a
train of five high-frequency (50 Hz) IC stimuli applied every 15 sec in
control (thick line) and during the inhibitory effect of
10 µM 5-HT (thin line). We
applied 10 µM BIC and 50 µM D-AP-5 during the whole session.
The successive TC EPSCs of the train were normalized to the first
response (R2 5 = a2 5/a1). In 5-HT, the amplitude
of the first EPSC was reduced to 13% of its control value. In control,
the repetitive stimulation induced a short-term depression of the TC
EPSCs (R2 = 0.74; R3 = 0.44; R4 = 0.23; R5 = 0.15) that was
changed into a short-term facilitation in 5-HT
(R2 = 1.20; R3 = 2.00;
R4 = 2.00; R5 = 1.40).
2, Summary data (mean ± SEM) of the effect of 5-HT
on the short-term changes of the TC EPSC in four layer IV cells. In
abscissa is shown the stimulus number in the train, in the ordinate the
EPSC amplitude normalized to the first response
(Rn). 5-HT induced a significant
(p < 0.01) inhibition of the first TC EPSC
(39 ± 16 pA in control; 7 ± 5 pA; n = 4 in 5-HT), associated with a change of the short-term depression seen in
control (continuous line) into a short-term facilitation
in 5-HT (dashed line).
|
|
Spontaneous bursts of activity have been recorded in the TC pathways
during development, and it has been proposed that they play an
important role in the activity-dependent refinement of the axonal
projections in sensory systems (Weliky and Katz, 1999). To investigate
the possibility that 5-HT regulates the propagation of the spontaneous
burst of activity in the TC pathways of the neonate mice, we decided to
test the effect of 5-HT on the responses to high-frequency (50 Hz)
trains of five IC stimuli (Fig. 6B) reminiscent of
the bursts seen in vivo (Weliky and Katz, 1999). The
experiments (n = 4) were done in the presence of
GABAA and NMDA receptor antagonists to isolate
the AMPA-KA component of the TC EPSC. We compared the effect of 50 Hz
trains of IC stimulation in control and in 10 µM 5-HT. In control the IC trains induced a
progressive decline of the TC EPSC amplitude or short-term depression similar to that seen in older rodents (Gil et al., 1997, 1999). In 10 µM 5-HT, we found a reduction of the first TC
EPSC amplitude (39 ± 16 and 7 ± 5 pA; n = 4; in control and in 5-HT, respectively; p < 0.01). At
the same time, the short-term depression seen in control
(R2 = 0.79 ± 0.19;
R3 = 0.58 ± 0.29;
R4 = 0.27 ± 0.09; R5 = 0.17 ± 0.10;
n = 4) was changed into a short-term facilitation in 10 µM 5-HT (R2 = 1.83 ± 1.02; R3 = 2.03 ± 0.30; R4 = 1.48 ± 0.71;
R5 = 1.28 ± 0.59;
n = 4); R4 and
R5 were significantly different in
control and in 5-HT (p < 0.01). These results
show that the effect of 5-HT on the TC EPSC depends on the recent
history of the TC synapse and that 5-HT may change the propagation of
high-frequency bursts of TC activity in the primary somatosensory TC
pathways of the neonatal mice.
5-HT1B receptors mediated the presynaptic inhibitory action
of 5-HT
Previous studies using a genetic approach have shown that the
excessive activation of the 5-HT1B receptors plays a critical role in
the appearance of the barrelless phenotype in the MAOA knock-out mice
(Salichon et al., 2001). It has also been shown that 5-HT1B receptor
binding sites disappear after thalamic lesion and are present in the
layer IV of S1 cortex in P8 rats and mice (Leslie et al., 1992; Cases
et al., 1996), suggesting that 5-HT1B receptors are expressed by the
axon terminals of the VB relay neurons (Bennett-Clarke et al., 1993).
Substantial coexpression of the 5-HT1B and 5-HT1D receptors has also
been reported in the adult rat brain (Bruinvels et al., 1993). Both
receptor types are addressed to axon terminals and are negatively
coupled to adenylyl cyclase (Maroteaux et al., 1992). In embryos, the
expression of the 5-HT1D receptors appears to precede the expression of
the 5-HT1B receptors (Bolanos-Jimenez et al., 1997), raising the
possibility that both the 5-HT1B and the 5-HT1D receptors may
contribute to the effect of 5-HT on the TC EPSC of the neonatal mice.
To investigate this possibility we studied the distribution of the
5-HT1B and 5-HT1D receptors in the somatosensory TC pathways during
development by in situ hybridization (Fig.
7). Full-length mRNA probes were used to
compare the distribution of the 5-HT1B and 5-HT1D receptor mRNAs in
neonate and adult mice. We observed 5-HT1B receptor gene expression in
the dorsal lateral geniculate nucleus (dLGN) and in the VB thalamic
nucleus from P0 to P7. This expression decreased by P10 and had
completely disappeared by P14 (Fig. 7F). In S1
cortex, 5-HT1B receptor mRNAs were detected only in the infragranular
cortical layers, essentially in layer V. The expression of the 5-HT1D
receptor mRNAs was very low at P4, P7, and in adults: a weak signal was
detected in the striatum, and no signal was found in the somatosensory
thalamic and cortical structures (data not shown).
View larger version (189K):
[in this window]
[in a new window]
|
Figure 7.
Transient expression of 5-HT1B
receptor mRNAs in the mouse somatosensory system. Expression
pattern is illustrated at P7 (A, C) and P15 (D,
F) in the cerebral cortex (A, D) and the
thalamus (C, F). Nissl-counterstained sections
are shown for the cortex to illustrate the layer IV in S1 cortex where
the recordings were done (B, E). In S1 cortex, the
specific hybridization signal is limited to layer V at P7 and P15,
although the relative extent of the expression appears to be broader at
P7 than at P15 (the position of the cortical layers is indicated by
roman numbers). In the thalamus, a strong hybridization signal is
visible in the dLGN and the VB at P7 but is absent at P15. Scale bar,
450 µm.
|
|
According to our in situ hybridization studies (Fig. 7), the
expression of 5-HT1B receptor mRNA by the TC relay neurons is transient
and disappears after the second week of life. To further investigate
the possible implication of the 5-HT1B receptor in the regulation of
the TC activity in the neonate and to test whether the effect of 5-HT
is developmentally regulated, we studied the effect of 5-HT on the TC
EPSC in older mice (23 ± 4 d; n = 22). The
AMPA component of the monosynaptic TC EPSC was recorded at 80 mV in
the presence GABAA antagonists to block the
disynaptic TC IPSC evoked by the minimal IC stimulation in the older
mice and in the presence of AP-5 to prevent seizure activity (see
Materials and Methods). We found that in mice older than 21 d, the
latency of the TC EPSC and the variability of the latency were smaller than in the neonate, as expected from the axonal myelin formation (Table 1). The effect of 10 µM 5-HT applied for
5 min was tested in six cells (Fig. 8).
We found that the amplitude of the AMPA component of the TC EPSC was
very slightly reduced by 5-HT. Within individual neurons, the effect of
5-HT was weak and not consistent. When the cells were averaged, we
found a small reduction the EPSC amplitude with a minimal relative
value (0.89 ± 0.06; n = 6); significantly smaller
than the control (p < 0.0001), but the
inhibitory effect of 5-HT was significantly smaller
(p < 0.001) in older mice than in the neonate
(compare Figs. 5C1 and 8B). These results show that during barrel development the downregulation of the 5-HT1B
receptor mRNA expression in the VB relay neurons is correlated with a
significant reduction of the effect of 5-HT on the TC EPSC. They
also indicate that in older mice 5-HT receptors other than 5-HT1B
receptors are probably not involved in the control of the TC EPSC.
View larger version (25K):
[in this window]
[in a new window]
|
Figure 8.
Effect of 5-HT on the TC EPSC in the older
wild-type mice. A, The AMPA component of the TC EPSC
evoked by the IC stimulation at 0.07 Hz was recorded at 80 mV in
presence of 100 µM PTX and 100 µM
D,L-AP-5. 1, Superimposition of sequential
individual TC EPSCs (n = 20) in control, during the
application of 10 µM 5-HT, and at the end of the
recording session. 2, In the same neuron, the TC EPSC
amplitudes were normalized to the mean value in control (176 ± 25 pA; n = 50). The individual values
(squares) and the running average (continuous
line) of the relative amplitude of TC EPSC are plotted against
time. The application of 10 µM 5-HT for 5 min induced a
very small reduction of the TC EPSC amplitude with a minimal value
(0.87 ± 0.06; n = 10) significantly
(p < 0.01) different from the control
(1.00 ± 0.14; n = 40). B,
Summary data of the effect of 10 µM 5-HT on the AMPA
component of TC EPSC recorded in six cells from wild-type mice between
P18 and P29. The mean (continuous line) and the SEM
(vertical lines) are plotted against time. At the end of
the application of 10 µM 5-HT, the amplitude of the TC
EPSC reached a minimal value (0.89 ± 0.06; n = 10) significantly (p < 0.01) different
from the control (1.00 ± 0.08; n = 40).
|
|
To further investigate the possible implication of the 5-HT1B receptors
in the effect of 5-HT on the TC EPSC, we studied the effect of CP93129
(100 nM), a specific 5-HT1B agonist, on the TC EPSC evoked
by a single IC stimulus and recorded at +30 mV (Fig.
9). In all tested cells
(n = 4), we found that CP93129 induced a significant
(p < 0.0001) and reversible reduction of the TC EPSC. These results indicate that the activation of the 5-HT1B receptors mimics the effect of 5-HT on the TC EPSC in the somatosensory system of the neonatal mice.
View larger version (26K):
[in this window]
[in a new window]
|
Figure 9.
CP93129, a 5-HT1B receptor agonist, mimics the
inhibitory effect of 5-HT on the TC EPSC in the neonate wild-type mice.
A, In a neuron maintained at +30 mV, the NMDA component
of the TC EPSC was evoked by IC stimulation at 0.07 Hz in control and
during the application of 100 nM CP93129 for 5 min.
1, Superimposition of sequential individual TC EPSCs
(n = 5) in control, during the response to 100 nM CP93129, and after recovery. CP93129 induced a
reversible reduction of the TC EPSC amplitude, and failures occurred.
2, The amplitude of the TC EPSC was normalized to the
mean amplitude in control (84 ± 14 pA; n = 40). The individual (dots) and running average
(continuous line) of the relative amplitude of TC EPSCs
are plotted against time. The relative amplitude of the TC EPSC reached
minimal value (0.16 ± 0.09; n = 10) in 100 nM CP93129 that was significantly
(p < 0.0001) different from the control
(1.00 ± 0.17; n = 40). B,
Summary data (mean ± SEM) of the effect of 100 nM
CP93129 applied for 5 min in four cells. The recording conditions are
identical to those described in A. The minimal relative
amplitude of the TC EPSC reached in 100 nM CP93129
(0.35 ± 0.33; n = 4) was significantly
different from the control.
|
|
Finally, to examine the specificity of the effect of 5-HT1B receptors
on the TC EPSC of the neonate, we tested the effect of 5-HT on the TC
EPSC in the layer IV barrels of the neonatal 5-HT1B receptor knock-out
mice (Fig. 10). In these mice, we found that the minimal stimulation of the IC evoked monosynaptic unitary EPSCs with the same characteristics as in the wild-type mice (Table 1),
indicating that 5-HT1B receptors are not critically involved in the
development of the TC synapses. Both the NMDA and the AMPA components
of the TC EPSC appear normal, and their relative contribution was not
significantly different from that observed in the wild-type neonate
mice. We also observed a PPD of the TC EPSC with 50 Hz IC stimulation
with a PPR similar to that seen in the neonate wild-type mice (Table
1). We also found that the bath application of 10 µM 5-HT
in the 5-HT1B receptor knock-out neonate mice could still induce an
increase of the frequency of the spontaneous unitary EPSCs similar to
that seen in wild-type neonate mice, supporting the idea that this
effect of 5-HT on spontaneous EPSC is not attributable to 5-HT1B
receptors. In contrast, 10 µM 5-HT applied for 5 min on
five neurons had no effect on the amplitude of TC EPSC recorded at 80 mV in the 5-HT1B receptor knock-out mice (Fig. 10). These results support our pharmacological and in situ
hybridization data showing that 5-HT1B receptors mediate the regulation
of the TC EPSC in the neonate wild-type mice. They also indicate that the 5-HT1D receptors do not compensate for the absence of 5-HT1B receptors in the knock-out neonate mice.
View larger version (30K):
[in this window]
[in a new window]
|
Figure 10.
The TC EPSC of the 5-HT1B receptor knock-out
neonatal mice is insensitive to 5-HT. A, Physiological
characterization of the TC EPSC in a neonatal 5-HT1B knock-out mouse
evoked by IC stimulation. 1, The responses displayed a
faster kinetics at 80 mV than at +30 mV and reversed near 0 mV.
2, The latency distribution of the response at 80 mV
was stable (7.48 ± 0.14 msec; n = 80).
3, The 50 Hz paired pulse IC stimulation evoked a
depression of the second response (PPR = 0.91). B,
10 µM 5-HT applied for 5 min had no inhibitory effect on
the TC EPSC in a 5-HT1B knock-out mouse at P5. The averaged
(n = 10) TC EPSC recorded at 80 mV was evoked by
a paired pulse IC stimulation in control, during the application of
5-HT and after washing 5-HT. The application of 5-HT had no significant
effect on the amplitude of the first TC EPSC (37.73 ± 18.51 pA,
n = 39 in control; 37.87 ± 17.27 pA,
n = 20 in 5-HT) and on the PPR (0.95 and 1.03 in
control and in 5-HT, respectively). C, Summary data of
the results (mean ± SEM) obtained in five cells from 5-HT1B
receptor knock-out mice. 10 µM 5-HT applied for 5 min had
no effect on the amplitude of the first TC EPSC recorded at 80
mV.
|
|
| |
DISCUSSION |
In the barrel field of the neonate mice, we demonstrate that 5-HT
reduces the monosynaptic TC EPSC evoked by a low-frequency IC
stimulation. We also provide the first evidence that the effect of 5-HT
depends on TC activity because 5-HT relieves the short-term depression induced by high-frequency IC stimulation. As first suggested
by Rhoades et al. (1994), we show that the effect of 5-HT is caused by
a presynaptic action on the release of glutamate by the TC fibers, and
we demonstrate that the effect of 5-HT appears to rely exclusively on
5-HT1B receptors. Our results support the hypothesis that the excessive
activation of 5-HT1B receptors that takes place in MAOA knock-out mice
during the critical period (Cases et al., 1996; Salichon et al., 2001)
may prevent the barrel formation by changing TC activity.
Presynaptic effect of the 5-HT1B receptor activation on TC
excitatory transmission
The stimulation of the VB thalamic nucleus evokes a monosynaptic
EPSC in layer IV neurons that results probably from the release of
glutamate by ascending TC axons. As shown before in older mice (Agmon
and Connors, 1991), IC stimulation that could activate neurons in
layers V/VI and cause the release of glutamate by corticocortical collaterals of layer VI pyramidal neurons to layer IV (Zhang and Deschenes, 1997). To investigate this possibility we used whole-cell and extracellular field recordings in the layers V/VI of TC slices with
a functional monosynaptic connection between the IC and the layer IV S1
cortex. We recorded no antidromic responses in the infragranular layers
of S1 cortex. Therefore, although we cannot completely exclude a minor
contribution of descending CT axons, the monosynaptic EPSCs evoked in
layer IV S1 cortex by IC stimulation result probably mainly from the
release of glutamate by ascending TC axons. This advantageous feature
of the TC slice preparation may be attributable to the slightly
different trajectories of ascending and descending fibers (Bicknese et
al., 1994).
It has been shown previously that 5-HT reduces the EPSPs evoked by VB
stimulation in S1 cortex (Rhoades et al., 1994), but most neurons were
recorded in layer V and the EPSPs were polysynaptic, raising the
question of the site of action of 5-HT. Our recordings were obtained
from neurons in layer IV, as demonstrated by the localization of their
soma in the barrels; they receive a direct input from the VB thalamic
nucleus (White, 1978). We showed that the latency of the TC EPSC
exhibits little variability and follows a Gaussian distribution, as
expected for a monosynaptic response. Therefore the effect of 5-HT that
we analyzed in layer IV neurons was probably restricted to TC synapses.
Our in situ hybridization and physiological data show that
5-HT acts presynaptically on the glutamate release by TC terminals via
the activation of 5-HT1B receptors. The possible transient expression
of the 5-HT1B receptor by the TC terminals has been previously assessed
by combining lesions and binding of a 5-HT1B agonist cyanopindolol
(Bennett-Clarke et al., 1993). We used full-length mRNA riboprobes to
provide direct evidence for a transient expression of the 5-HT1B
receptor gene in the VB thalamic neurons during postnatal development.
We also showed that within S1 cortex, the expression of the 5-HT1B
receptor gene remains confined to neurons in layer V in the neonatal
and adult mice. This is in contrast to previous descriptions of a
5-HT1B receptor mRNA expression in layer IV S1 cortex of the adult
rodents (Voigt et al., 1991; Boschert et al., 1994). Our physiological
results using CP93129, a specific 5-HT1B agonist, show also that the
activation of 5-HT1B receptors reduces the monosynaptic TC EPSC.
Finally we show that 5-HT1B receptors appear to contribute exclusively
to the inhibition of the TC EPSC: first, we were unable to detect
5-HT1D receptor gene expression in the thalamus and cerebral cortex;
second, the downregulation of the expression of the 5-HT1B receptor
gene in the VB thalamic nucleus was correlated to a decrease of the
sensitivity of the TC EPSC to 5-HT in mice older than 3 weeks; third,
we showed that in the neonate 5-HT1B receptor knock-out mice, the TC
synaptic contacts were functional, but the TC EPSCs were insensitive to 5-HT. Our results demonstrate also that the inhibitory effect of 5-HT
on the TC EPSC is caused by a presynaptic reduction of the release of
glutamate by the VB relay neurons: first, we found a similar inhibitory
effect of 5-HT on the AMPA and the NMDA components of the TC EPSC;
second, the effect of 5-HT is associated with a change of the paired
pulse ratio and a relief of the short-term depression of the TC EPSC. A
presynaptic effect of 5-HT through 5-HT1B heteroreceptors has also been
reported on retinal inputs to the superior colliculus (Mooney et al.,
1994) and to the suprachiasmatic nucleus (Pickard et al., 1999), where
the effect was maintained in adults.
Postsynaptic 5-HT1A and 5-HT2 receptors are expressed by cortical
neurons (Hoyer et al., 1994) and control potassium channel activity
(Andrade and Nicoll, 1987; McCormick, 1992). Their contribution to the
inhibition of the TC EPSC by 5-HT is unlikely because the recording
pipettes contained cesium, a potassium channel blocker. A possible
contribution of 5-HT3 receptors found in developing sensory cortex
(Roerig and Katz, 1997; Roerig et al., 1997), is also unlikely because
5-HT did not induce a postsynaptic current. The possibility that
5-HT1A, 5-HT2, and 5-HT3 receptors contribute to the increase of
spontaneous unitary IPSCs and EPSCs in the neonate S1 cortex remains to
be studied.
Possible effects of the 5-HT1B receptor activation on the
barrel formation
TOP
ABSTRACT
INTRODUCTION
