In vivo Demonstration of a Late Depolarizing Postsynaptic...

In vivo Demonstration of a Late Depolarizing Postsynaptic Potentialin CA1 Pyramidal Neurons

Yuan Fan* 1, Bende Zou*, Yiwen Ruan, Zhi-Ping Pang2 and Zao C. Xu

Department of Anatomy & Cell Biology,Indiana University School of Medicine, Indianapolis, Indiana, U.S.A

(FINAL ACCEPTED VERSION)

Running title: In vivo L-PSPs in CA1 pyramidal neurons

Correspondence to: Zao C. Xu, M.D., Ph.D.MS 507, Department of Anatomy & Cell Biology,Indiana University School of Medicine,635 Barnhill Drive, Indianapolis, IN 46202E-mail: [email protected]: (317) 274- 0547 (voice)Fax: (317) 278- 2040

* These authors contribute equally to this manuscript.

Current address: 1. Department of Neuroscience, Baylor College of Medicine; 2. Center forBasic Neuroscience, University of Texas Southwestern Medical Center

Articles in PresS. J Neurophysiol (October 13, 2004). doi:10.1152/jn.00734.2004

Copyright © 2004 by the American Physiological Society.

In vivo L-PSP in CA1 pyramidal neuronsJN-00734-2004.R1

2

ABSTRACT

Previous studies have shown that GABA can have a depolarizing and excitatory action

through GABAA receptors in mature CNS neurons in vitro. However, it remains unknown

whether this occurs under physiological conditions. In the present study, using intracellular

recording and staining in vivo technique, we demonstrate a late depolarizing post-synaptic

potential (L-PSP) in CA1 pyramidal neurons of adult Wistar rats under halothane anesthesia.

This L-PSP was elicited in approximately 70% of the recorded neurons upon stimulation of the

Schaffer collaterals or the contralateral commissural path. The size of L-PSP was linearly

correlated to the decay time constant but not the rising slope of the initial EPSP. Intravenous

administration of the NMDA receptor blocker MK-801 and the GABAA receptor blocker

picrotoxin significantly reduced the size of the L-PSP. The spine density and apical dendritic

branching length of the neurons that displayed L-PSPs was significantly greater than those that

do not. These results indicate that NMDA receptor and GABAA receptor-mediated depolarizing

post-synaptic potentials can be revealed in CA1 pyramidal neurons of adult rats in vivo,

supporting the physiological relevance of GABAA-mediated depolarization in normal neuronal

information processing. The difference in electrophysiological properties and morphological

features between neurons that display the L-PSP and the other neurons suggest that they might

represent two different subtypes of CA1 pyramidal neurons.

Keywords: Hippocampus, synaptic transmission, intracellular recording, NMDA, GABA.


3

INTRODUCTION

The hippocampus plays a fundamental role in certain forms of learning and memory, and

exhibits extraordinary vulnerability to pathological insults such as epilepsy and cerebral

ischemia. Each region of the hippocampal formation is linked by an excitatory tri-synaptic

pathway (Andersen et al. 1969). The major excitatory neurotransmitter in the hippocampus is

glutamate (Roberts et al. 1981). The action of glutamate is mediated through ionotropic and

metabotropic receptors (Hicks et al. 1987). The ionotropic glutamate receptors consist primarily

of a-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA), kainate and N-methyl-D-

aspartate (NMDA) receptors. AMPA and kainate receptors mediate fast excitatory post-synaptic

potentials (EPSP) whereas NMDA receptors mediate slower-rising and slower-decaying EPSPs.

The excitability of hippocampal pyramidal neurons is also influenced by the feedback

recurrent (Andersen et al. 1964, 1963) and the feedforward (Alger and Nicoll 1982a) g-

aminobutyric acid (GABA) receptor-mediated synaptic inhibition. GABA mediates its action

through both ionotropic GABAA and metabotropic GABAB receptors, which underlie the fast and

slow phases of inhibitory post-synaptic potentials (IPSP) respectively (Sivilotti and Nistri 1991).

It is traditionally believed that, in the adult mammalian CNS, GABA assumes an inhibitory

role, keeping neuronal excitability under control (Macdonald and Olsen 1994). However, this

over-simplified notion has been challenged by many in vitro studies that use adult animals (Stein

and Nicoll 2003). In those studies, exogenous and synaptically released GABA was shown to

mediate a long-lasting depolarizing potential that enhanced neuronal excitability under certain

conditions (Alger and Nicoll 1982b; Andersen et al. 1980; Avoli 1992; Grover et al. 1993;

Gulledge and Stuart 2003; Perreault and Avoli 1988; Staley et al. 1995; Taira et al. 1997;

Thalmann et al. 1981; Wong and Watkins 1982).


4

The exact mechanisms of the depolarizing effect of GABA in adult animals have been under

active investigation. Some studies suggest that there are two types of GABAA receptors on

hippocampal pyramidal neurons. The hyperpolarizing responses reflect the activation of

synaptic receptors, which are highly concentrated on the pyramidal cell soma and initial

segment, whereas depolarizing responses reflect the activation of extra-synaptic or dendritic

receptors (Alger and Nicoll 1982a, b; Gulledge and Stuart 2003). Other studies show that a

collapse in the transmembrane Cl- gradient, as well as outflux of HCO3- through GABAA receptor

channels might also play a role (Kaila 1994; Staley et al. 1995).

Despite the wealth of literature describing the depolarizing action of GABA in the mature

mammalian central neurons, it is not yet clear whether this occurs in vivo. Thus the

physiological relevance of the GABAA-mediated depolarization in adult animal has been

challenged. We address this issue using sharp-electrode intracellular recording and staining in

vivo in CA1 pyramidal neurons of adult rats. This approach offers the advantage of preserving

neuronal circuitry as well as minimizing the perturbation to the cytoplasmic content. We found a

long-lasting depolarizing post-synaptic potential that resembles the effect of GABA described in

vitro but has an additional NMDA receptor-mediated component. These data support the view

that GABAergic interneuron networks can have an excitatory role in vivo in information

processing in mature central nervous system.


5

MATERIALS AND METHODS

All procedures were performed in accordance with the National Institute of Health Guide for

the Care and Use of Laboratory Animals (NIH Publication No. 80-23) and were approved by the

Indiana University Institutional Animal Care and Use Committee. Male adult Wistar rats (200-

300 g, Charles Rivers, MA) were used in the current study. The rats were anesthetized with 1-

2% halothane mixed with 33% oxygen and 66% nitrogen gas. In some animals, one femoral

vein was cannulated for subsequent intravenous (i.v.) administration of dizocilpine maleate

(MK-801, 2 mg/kg) or/and picrotoxin (PTX, 2 mg/kg). These drugs were purchased from Sigma

(St. Louis, MO) and dissolved in 0.9% saline in experiments.

Intracellular recording and staining in vivo.

Anesthetized rats were fixed in the stereotaxic apparatus. The skull was opened to expose the

recording site and to place the electrodes. Recording electrodes were pulled with a Kopf pipette

puller (model 750, David Kopf Instruments, Tujunga, CA) from glass capillaries with a filament.

Tip resistance ranged from 50 to 80 MΩ when filled with a solution of 5% neurobiotin (Vector

Laboratories, Burlingame, CA) in 2 M potassium acetate. Bipolar stimulating electrodes (1 mm

apart) were made from insulated stainless steel pins with 1 mm of exposed tip. The stimulating

electrodes were placed into the contralateral commissural pathway (AP: 3.7- 4.7 mm, ML: 1.5

mm, DV: 3.5 mm) and/or ipsilateral Schaffer collaterals pathway (AP: 4.7- 5.7 mm, ML: 5.6

mm, DV: 3.9 mm, Vertical angle 25o) (Fig. 1).

Cerebrospinal fluid was drained via a cisternal puncture to reduce brain pulsation. The

animal was then suspended with a clamp applied to the tail, in order to further reduce the

pulsation of the brain caused by respiratory movement. After placement of the recording


6

microelectrode in the cortex above the hippocampus, the exposed surface of the brain was

covered with soft paraffin wax. A stimulator (Master-8, A.M.P.I., Israel) with a stimulus

isolation unit (Isoflex, A.M.P.I., Israel) was used to deliver stimulus pulses. The microelectrode

was advanced slowly, with a motorized micro-motion controller (model ESP300, Newport

Corporation, Irvine, CA), into the hippocampus at 2 µm increment to impale the CA1 pyramidal

neurons. After impalement, the neurons with a stable membrane potential of at least – 60 mV,

and action potential overshoot were selected for further recording.

Signals were amplified by a high input-impedance amplifier (Axoclamp 2B, Axon

instrument, Foster city, CA). The bridge balance was monitored throughout the experiments and

adjusted appropriately. Data were digitized at 4- 5 KHz via a computer interface (ITC-16,

Instrutech Corporation, Long Island, NY) controlled by the data acquisition program (Axodata,

Axon Instruments) and stored on a Macintosh computer for off-line analysis. After each

successful recording, neurobiotin was iontophoresed into the cell by applying depolarizing

current pulses (2 Hz, 300 ms, 1.0- 1.5 nA) for more than 10 minutes to ensure sufficient loading.

At the end of the experiment, the animal was deeply anesthetized and transcardially perfused

with 0.01 M phosphate-buffered saline followed by 4% paraformaldehyde. The brain was

removed and stored in the same fixative overnight at 4 oC. Coronal sections were cut at 80 mm

thickness using a vibratome and incubated in 0.01 M potassium phosphate-buffered saline

(KPBS, pH 7.4) containing 0.1% horseradish peroxidase-conjugated avidin-D (Vector) and 0.5%

Triton X-100 overnight at room temperature. After detection of peroxidase activity with 3,3’-

diaminobenzidine (DAB), sections were examined in KPBS. Sections containing labeled cell

and stimulating electrodes track were mounted on gelatin- coated slides and counterstained with

Cresyl Violet for light microscopy.


7

3-D reconstruction of the labeled neurons and quantitative analysis of the spine density

Brain sections containing DAB-labeled neurons were incubated with 0.25% osmium

tetroxide (OsO4) for 1 hour to intensify the staining. The sections were then dehydrated and

embedded with Embed 812 (EMS, Hatfield, PA). Optimally labeled CA1 neurons in the serial

sections were reconstructed using a 60¥ water-immersion lens and a computer-based neuronal

reconstruction program (Neurolucida, Microbrightfield, Colchester, VT). Morphometric data for

the dendrites was calculated with the data analysis program NeuroExplore (Microbrightfield,

Colchester, VT). These data includes total dendritic branch end number (TDBEN), total

dendritic branch length (TDBL), and mean dendritic branch length (MDBL, by dividing TDBL

with the number of primary dendrites). The dendritic spines of CA1 neurons were examined

using a 100¥ oil-immersion objective lens. From each neuron, five branches of ~1 µm in

diameter were selected from the basal dendrites (30- 100 µm from soma), the proximal portion

(50- 100 µm from the soma) and the distal portion (200- 300 µm from the soma) of the apical

oblique dendrites respectively. The number of the spines on a 20 µm segment of each dendrite

was counted and spine density was calculated as the number of spines per unit length of dendrite.

Electrophysiological data analysis

Both evoked post-synaptic potentials and intrinsic membrane properties of CA1 pyramidal

neurons were analyzed using the data analysis program AxoGraph (Axon Instruments). The

amplitudes of the post-synaptic potentials were measured as the difference between resting

membrane potential and the peak of the response. The duration of the post-synaptic potential

was measured at half of the peak amplitudes. The slope of the initial EPSP was measured as the

time between 10% and 90% of the peak amplitudes. The decay time constant of the EPSP was


8

derived from mono-exponential fitting of the decay part of the potentials. The rheobase was

defined as the minimum magnitude required for a 200 ms depolarizing current pulse to evoke an

action potential at soma. Analysis of action potentials (spike) was performed from the action

potentials evoked by the rheobasic current. The spike height was measured as the difference

between the resting membrane potentials and the peak of action potentials. The spike duration

was measured at half of the peak amplitude of the action potentials. The spike threshold was

measured at the visually determined inflection point at the onset of an action potential. The

membrane time constant was calculated as the time required for the membrane potential to reach

63% of the steady-state amplitudes in response to a hyperpolarizing current pulse (-0.3 nA, 200

ms). Input resistance of the membrane was derived from the linear portion of the I-V

relationship as the slope of the linear regression fitting line. For calculating the reversal potential

of each post-synaptic potential, the amplitude of each component was measured and plotted as a

function of corresponding membrane potential. The reversal potential was then extrapolated by

the intersection of the linear regression line with the membrane potential axis at which the

amplitude of the component would be zero.

All values were presented as mean ± S.E. ANOVA, post-hoc Scheffé’s test, paired or

unpaired Student’s t- tests were used to detect statistical difference (StatView, Abacus

Concepts). P < 0.05 was considered to be significant.


9

RESULTS

Intracellular recording and staining in vivo were performed on 71 rats. A total of 91 neurons

with stable resting membrane potentials of – 60 mV or more hyperpolarized and action potential

overshoot were analyzed in the present study. Among these cells, 40 neurons were

intracellularly stained and identified as CA1 pyramidal neurons based on their location and

morphology. The remainder was considered CA1 pyramidal neurons based on the stereotaxic

parameters and their responses to the afferent stimulation.

Synaptic responses of the CA1 pyramidal neurons to afferent stimulation

Consistent with previous literature (Schwartzkroin 1986), low intensity stimuli (1.5 times of

the threshold intensity to evoke EPSP, 1.5T) to the contralateral commissural path elicited a post-

synaptic potential starting with an initial EPSP followed by an IPSP (Fig. 2A). However, with

the increase of stimulus intensity to about 3T, many neurons (62 out of 86) began to exhibit a

late depolarizing post-synaptic potential (L-PSP) interposed between the fast and slow IPSPs

(Fig. 2B). This late component typically saturated at 6 times of the threshold stimulus intensity

(6T). At 6T stimulation, the amplitude of L-PSPs was 7.3 ± 0.7 mV (n = 29) and the duration

was 186.2 ± 10.8 ms (n = 29). Occurrence of the L-PSP was associated with an increase in

membrane conductance as indicated by the decrease in the amplitude of membrane potential

deflections in response to hyperpolarizing current pulse (Fig. 2C). Although the induction of the

L-PSP required relatively strong afferent stimulation, it was not dependent on the appearance of

evoked action potentials arising from the EPSPs. Furthermore, L-PSP could not be evoked by

direct depolarizing current injection to the soma (Fig. 2D). In some cases, a single stimulus to


10

the afferent could generate spikes from the late depolarizing component while paired-pulse

stimulation consistently potentiated the late depolarizing components (Fig. 2E and F).

Since CA1 pyramidal neurons mainly receive their excitatory innervations from the

contralateral commissural path and Schaffer collaterals (Andersen et al. 1969), we tested the

pathway specificity of the L-PSP induction. L-PSP could be evoked from either the contralateral

commissural path stimulation or Schaffer collaterals stimulation (Fig. 3A). The input-output

property of the L-PSP elicited by stimulation of either pathway showed no distinguishable

difference (Fig. 3B). In contrast, some CA1 pyramidal neurons (24 out of 86) failed to exhibit L-

PSP response regardless of the stimulation intensities (Fig. 3C).

L-PSP responses became apparent at a stimulation intensity of 3T or higher, whereas at 1.5T

stimulation, the primary evoked potentials was the initial EPSP. We studied the kinetics of the

initial EPSPs evoked at 1.5T and examined the relationship between the initial EPSPs and the L-

PSPs evoked at 6T. We found that the duration and decay time constant of the initial EPSPs in

the neurons that displayed the L-PSP were significantly larger than those of the neurons that did

not display the L-PSP (p < 0.05), whereas the amplitude and rising slope of the initial EPSPs

showed no significant difference (Fig. 4 A and B). In addition, the decay time constant of the

initial EPSPs was linearly related to the amplitude (R2 = 0.73) and duration (R2 = 0.70) of the L-

PSPs (Fig.5A and B). In contrast, there was no linear relationship between the slopes of the

initial EPSPs and the amplitude and duration of the L-PSPs (Fig. 5C and D).

Electrophysiological and pharmacological properties of the L-PSP

To understand the underlying ionic mechanism of the L-PSP, we first measured the reversal

potentials of different components of the post-synaptic potentials. Currents were injected into


11

the recorded neurons to maintain the membrane potential at different levels and the amplitudes of

each post-synaptic potential were be measured accordingly (Fig. 6A). The initial EPSP reversal

potential was – 46.1 mV (n = 18); the fast IPSP reversal potential was – 65.5 mV (n = 18); the

late IPSP reversal potential was – 94.4 mV (n = 18); the L-PSP reversal potential was – 58.4 mV

(n = 18) (Fig. 6B).

To directly elucidate the underlying mechanisms of the L-PSP, we applied the NMDA

receptor antagonist MK-801 and the GABAA receptor antagonist PTX i.v., after eliciting a L-PSP

with 6T afferent stimulation. MK-801 (2 mg/kg) rapidly reduced the amplitude of L-PSPs from

8.6 ± 1.3 mV to 3.0 ± 1.3 mV (n = 8, p < 0.01). The reduction ratio was 68.6 ± 13.1% (Fig. 7A

and D). Upon application of PTX (2 mg/kg), the amplitude of L-PSPs was rapidly reduced from

7.0 ± 1.0 mV to 3.1 ± 0.6 mV (n = 8, p < 0.01). The reduction ratio was 62.5 ± 7.6% (Fig. 7B

and D). We performed additional experiments in which both MK-801 and PTX were applied

simultaneously. L-PSPs were suppressed from 8.0 ± 0.8 mV to 2.8 ± 1.0 mV (n = 5, p < 0.01).

The reduction ratio was 67.0 ± 9.1% (Fig.7C and D). These results indicate that that L-PSP is

mediated by both NMDA receptors and GABAA receptors in a non-additive fashion.

The difference in the post-synaptic response between the neurons that displayed the L-PSP

and those that did not prompted us to hypothesize that they might represent two different cell

subtypes in the CA1 region. Therefore we characterized their membrane properties and

examined potential differences. As shown in Table 1, there was no significant difference in

parameters such as resting membrane potential, spike height, spike duration, spike threshold and

rheobase. However, the input resistance and membrane time constant of the neurons that

displayed a L-PSP were significantly greater than those that did not. The stability and

comparability of the resting membrane potential and active membrane properties of these two


12

groups of neurons suggest that it is not likely that the relatively lower input resistance of neurons

that do not display the L-PSP is due to more severe injury caused by microelectrode impalement.

The morphology of the L-PSP neurons and non L-PSP neurons

To test if the difference in electrophysiological properties of recorded neurons had any

relationship to the their difference in anatomical locations within hippocampus, we plotted the

recording sites with microscopy after staining. We found that the recording sites were randomly

distributed within the CA1 region, and the distribution pattern of the morphologically identified

L-PSP neurons and non L-PSP neurons had no detectable difference. Both types of the cells

scattered fairly homogeneously within the CA1 region (Fig. 8A). Since the L-PSP had a MK-

801-sensitive component, we hypothesized that neurons that displayed the L-PSP might receive a

relatively higher excitatory drive. Since the excitatory inputs of the CA1 pyramidal neurons

mainly impinge on the dendritic spines, we compared spine density between the neurons that

displayed L-PSP and those neurons that did not. On the basal dendrites, the spine density in the

neurons that displayed the L-PSP (34.5 ± 2.8, n = 12) was significantly higher than that of the

neurons that did not display the L-PSP (27.1 ± 2.9, n = 5, p < 0.05) (Fig. 8B and C). The spine

density on the proximal portion of the apical dendrites in the neurons that displayed the L-PSP

(37.3 ± 3.9) was significantly higher than that of the neurons that did not (31.6 ± 2.3, p < 0.05)

(Fig. 8C). The spine density on the distal portion of the apical dendrites in the neurons that

displayed the L-PSP (30.6 ± 3.6) was also higher than that of the neurons that did not (28.9 ±

3.4), but no statistical significance was detected (Fig. 8C).

Since higher density of spines did not neccessarily translate to a higher overall excitatory

input, we also examined potential differences in the general dendritic branching morphology


13

between neurons that displayed the L-PSP and neurons that did not (Fig. 9A). Sholl’s analysis

revealed that the number of branches in the proximal apical dendrites (170 µm – 300 µm) was

significantly higher in the neurons that displayed the L-PSP (Fig. 9B). In addition, the total

branching ends and length of the apical dendrites were significantly greater in the neurons that

displayed the L-PSP than the neurons that did not (Fig. 9C). Taken together, these results

suggest that the neurons that display the L-PSP receive a higher overall level of excitatiry input

than the neurons that do not.


14

DISCUSSION

The main finding of the present study is that a partly GABAA-receptor mediated long-lasting

depolarization, which is termed “L-PSP”, can be revealed in mature CA1 pyramidal neurons in

vivo by stimulating either Schaffer collaterals or the contralateral commissural pathways. Such a

response can be elicited from approximately 70% of the recorded neurons. Both NMDA

receptors and GABAA receptors contribute to the generation of the L-PSP. Approximately 30%

of the recorded neurons fail to exhibit the L-PSP response under the same stimulation conditions.

These neurons have a significantly lower input resistance and display differences in dendritic

spine density and branching complexity.

Stimulation of afferent, but not direct suprathreshold depolarizing current injection to the

soma, can evoke the L-PSP, suggesting that the L-PSP is a synaptically-driven event instead of a

somatic regenerative event. Consistent with this observation, the generation of the L-PSP is not

dependent on the initiation of the evoked action potentials arising from the initial EPSPs. Thus it

is unlikely that back-propagating action potentials play a role in L-PSP induction. The induction

of L-PSPs requires relatively strong stimulation intensities (> 3T), suggesting that a threshold

amount of pre-synaptic fibers has to be activated (“co-operativity”). The input-output property

of L-PSPs reaches an asymptotic level at higher stimulation intensities (> 5T), which might

reflect saturation of receptor activation. L-PSP response does not exhibit input-specificity, since

both SC and CC stimulation can elicit indistinguishable response. The lack of input-specificity

suggests that the L-PSP might not be specifically mediated through the CA3–CA1 connections.

The L-PSP can evoke bursts of spikes, supporting the excitatory nature of this component. It

has been shown that strong excitatory synaptic activation can lead to “action potentials”


15

generated within the dendrites, independent of the soma or axon initial segment (Regehr et al.

1993; Stuart et al. 1997). The burst of spikes on the L-PSP could also have a dendritic origin.

Since glutamate is the main excitatory neurotransmitter in the hippocampus and the NMDA

component of the synaptic current rises slowly to a peak (~20 ms) and then decays bi-

exponentially with time constants of 40 and 200 ms respectively (Lester et al. 1990), the

prolonged duration of the L-PSP prompted us to initially hypothesize that L-PSPs were mediated

by NMDA receptors. Regression analysis indicated that the amplitude and duration of the L-PSP

was positively correlated to the decay time constant, but not to the rising slope of initial EPSPs.

Since in synapses containing both AMPA and NMDA receptors, the rising slope of EPSP is

mainly contributed by the AMPA component whereas the decay of EPSP is mainly contributed

by the NMDA component, the most parsimonious interpretation of this result is that the L-PSP

bears certain relationship with the NMDA receptor strength. More direct evidence supporting

the role of NMDA receptors in mediating the L-PSP comes from the result that MK-801, the

NMDA receptor- gated channel blocker, significantly suppresses this response.

On the other hand, a possible role for GABAA receptor involvement needs to be considered.

In our study, the GABAA receptor blocker, PTX, significantly reduced the amplitude of the L-

PSPs. This indicates that GABAA receptor activation can indeed mediate a depolarizing effect in

vivo.

In support of its mixed nature, L-PSPs reverse at – 58.4 mV, which lies between the reversal

potentials of the glutamatergic receptor-mediated EPSP and the GABAergic IPSPs. It should be

noted that, as a caveat, due to the lack of pharmacological isolation in our in vivo preparation,

each post-synaptic potential (i.e. excitatory and inhibitory), temporally overlaps with each other

to a certain degree. Thus, the actual reversal potential for the EPSP should be more depolarized


16

and the fast IPSP reversal potential should be more hyperpolarized than our measured values.

Similarly, the true reversal potential of L-PSPs might be more depolarized if measured in the

absence of the hyperpolarizing effect of its neighboring fast and slow IPSPs. Additionally, the

reversal potential of more distal dendritic synapses is probably not accurately assessed with

somatic current injection.

Although the L-PSP has both a MK-801-sensitive component and a PTX-sensitive

component, due to the lack of an additive effect of MK-801 and PTX in blocking the L-PSP, it is

unlikely that it is simply an arithmetic sum of independent NMDA and GABAA receptor-

mediated conduction. According to the feed-forward inhibition theory postulated by Alger and

Nicoll (Alger and Nicoll 1982a), a population of interneurons receive afferent excitation and

make synapses on pyramidal neuronal dendrites to form the feed-forward inhibition. Indeed,

there is a strong excitatory drive to the GABAergic interneurons in CA1 region which is

mediated by ionotropic glutamatergic receptors (Davies et al. 1990). Furthermore, the effect of

GABA on dendrites has been shown to be depolarizing (Alger and Nicoll 1982a, 1979, 1982b;

Andersen et al. 1980; Gulledge and Stuart 2003). In line with this theory, it is plausible that

MK-801 mainly exerts its effect pre-synaptically, which leads to a reduction in the NMDA

receptor-mediated excitatory drive to these interneurons and subsequent reduction in GABA

release. The reduction in GABA release in turn causes less GABAA receptor- mediated

depolarization on CA1 pyramidal neurons.

Nevertheless, one should not rule out the contribution of the post-synaptic NMDA receptors

on CA1 pyramidal neuron to L-PSP expression. The post-synaptic NMDA receptor activation

could occur secondary to the depolarization initially caused by GABAA receptor activation,

greatly boosting the amplitude of the L-PSP. The findings that MK-801 preferentially affects the


17

later phase of the L-PSP whereas PTX suppresses the L-PSP fairly uniformly fit such an

explanation.

A small proportion of the CA1 neurons failed to exhibit the L-PSPs regardless of the

stimulation intensities and the stimulation pathways (contralateral commissural path or ipsilateral

Schaffer collateral path). This group of neurons shows indistinguishable resting membrane

potentials and action potential properties compared with the group of neuron exhibiting L-PSP.

However, the input resistance and membrane time constant of these neurons are significantly

lower. In theory, these passive membrane properties determine the temporal and spatial

summation of the synaptic potential (Koester and Siegelbaum 2000), thus one can argue that the

neurons with smaller input resistance and time constant are not able to amplify and propagate the

remotely generated synaptic potentials and therefore fail to display a L-PSP. However, this

possibility is not supported by the fact that the failure to elicit a L-PSP is not dependent on the

stimulation intensities in these neurons. Rather, the failure to exhibit a L-PSP in these neurons is

more likely due to the difference in their synaptic inputs. Since L-PSP generation is dependent

on NMDA receptors and GABAA receptors, the failure to generate a L-PSP response could result

from the lack of enough glutamatergic excitatory synaptic input to these neurons, as well as the

lack of innervations from feed-forward interneurons. Such an explanation is supported by the

discernable difference in spine density and the proximal apical dendrite branch pattern between

the L-PSP neurons and non L-PSP neurons. Since spines are the site of glutamatergic synaptic

contact on the hippocampal pyramidal neurons (Harris and Kater 1994), the difference in spine

expression levels between these two groups of neurons implies that they receive different levels

of excitatory input. In addition, previous studies have shown that the proximal dendrites of

pyramidal neurons can initiate regenerative local spikes, including sodium, calcium and NMDA


18

receptor-mediated spikes (Larkum et al. 2001; Schiller et al. 2000), it is likely that these

morphological features might correlate to the generation of the L-PSP.

L-PSPs may have important functional implications. Since the hippocampus is implicated in

learning and memory formation, and hippocampal long-term potentiation (LTP) is postulated to

be an underlying cellular substrate (Bliss and Collingridge 1993; Eichenbaum 1994), we

speculate that the L-PSP could play a potentially important role in synaptic integration and

activity-dependent synaptic plasticity. First, prolonged synaptic depolarization would lead to

more Ca2+ influx through NMDA receptors and voltage-gated calcium channels, activating a

variety of Ca2+ -dependent signal transduction cascades that are critically involved in information

processing and storage. Second, prolonged synaptic depolarization could promote dendritic

spike initiation and back-propagation of somato-axonal action potentials, both of which are

critical for spike-timing dependent long-term synaptic plasticity (Hoffman et al. 1997; Magee

and Johnston 1997; Watanabe et al. 2002). Indeed, the spiking activities elicited from the L-PSP

resemble previously well-documented complex spike bursting that is typical of hippocampal

pyramidal cells (Ranck 1973). This bursting may represent an important form of information

coding in the hippocampus (Lisman 1997). In agreement with this notion, it has been shown that

strong stimulation to the Schaffer collateral at theta frequencies can trigger the complex spike

bursting, which enables the induction of LTP (Thomas et al. 1998). It is important to note that

the GABAA receptor-mediated depolarization will relieve the Mg2+ block of the NMDA receptor,

thus facilitating NMDA receptor-dependent plasticity (Ben-Ari et al. 1997). Furthermore, the

dendritic spike initiation may be facilitated by the co-operative activation of both NMDA and

depolarizing GABAA receptors (Larkum and Zhu 2002; Schiller et al. 2000). In addition to the

implication in physiological situation, the depolarizing effect of GABA has also been implicated


19

in the pathogenesis of epilepsy. There is evidence that interictal epileptic activity in the human

temporal lobe is critically dependent on the excitatory action of GABA (Cohen et al. 2002).

The present results provide the first in vivo evidence for the existence of GABAA-mediated

excitation under normal physiological conditions in mature neurons. These findings suggest that

GABA may play a far more sophisticated role than its postulated inhibitory role in controlling

neuronal excitability.


20

ACKNOWLEDGEMENTS

We thank Drs. Roger A. Nicoll, Amiel Rosenkranz and Andreas Frick for their critical

comment on this manuscript. This work was supported by grants AHA 0070048 to Z.C.Xu.

AHA 0110275Z to Y.Fan, 0320081Z to B.D.Zou and 0120566Z to Z.P.Pang.


21

REFERENCES

Alger BE and Nicoll RA. Feed-forward dendritic inhibition in rat hippocampal pyramidal cells

studied in vitro. J Physiol 328: 105-123, 1982a.

Alger BE and Nicoll RA. GABA-mediated biphasic inhibitory responses in hippocampus.

Nature 281: 315-317, 1979.

Alger BE and Nicoll RA. Pharmacological evidence for two kinds of GABA receptor on rat

hippocampal pyramidal cells studied in vitro. J Physiol 328: 125-141, 1982b.

Andersen P, Bliss TV, Lomo T, Olsen LI, and Skrede KK. Lamellar organization of

hippocampal excitatory pathways. Acta Physiol Scand 76: 4A-5A, 1969.

Andersen P, Dingledine R, Gjerstad L, Langmoen IA, and Laursen AM. Two different responses

of hippocampal pyramidal cells to application of gamma-amino butyric acid. J Physiol 305: 279-

296, 1980.

Andersen P, Eccles JC, and Loyning Y. Location of Postsynaptic Inhibitory Synapses on

Hippocampal Pyramids. J Neurophysiol 27: 592-607, 1964.

Andersen P, Eccles JC, and Loyning Y. Recurrent inhibition in the hippocampus with

identification of the inhibitory cell and its synapses. Nature 198: 540-542, 1963.

Avoli M. Synaptic Activation of GABAA Receptors Causes a Depolarizing Potential Under

Physiological Conditions in Rat Hippocampal Pyramidal Cells. Eur J Neurosci 4: 16-26, 1992.

Ben-Ari Y, Khazipov R, Leinekugel X, Caillard O, and Gaiarsa JL. GABAA, NMDA and

AMPA receptors: a developmentally regulated 'menage a trois'. Trends Neurosci 20: 523-529,

1997.

Bliss TV and Collingridge GL. A synaptic model of memory: long-term potentiation in the

hippocampus. Nature 361: 31-39, 1993.


22

Cohen I, Navarro V, Clemenceau S, Baulac M, and Miles R. On the origin of interictal activity in

human temporal lobe epilepsy in vitro. Science 298: 1418-1421, 2002.

Davies CH, Davies SN, and Collingridge GL. Paired-pulse depression of monosynaptic GABA-

mediated inhibitory postsynaptic responses in rat hippocampus. J Physiol 424: 513-531, 1990.

Eichenbaum H. The Hippocampal System and the Declarative Memory in Humans and Animals:

Experimental analysis and Historical Origins. Cambridge: MIT Press, 1994.

Grover LM, Lambert NA, Schwartzkroin PA, and Teyler TJ. Role of HCO3- ions in depolarizing

GABAA receptor-mediated responses in pyramidal cells of rat hippocampus. J Neurophysiol 69:

1541-1555, 1993.

Gulledge AT and Stuart GJ. Excitatory actions of GABA in the cortex. Neuron 37: 299-309,

2003.

Harris KM and Kater SB. Dendritic spines: cellular specializations imparting both stability and

flexibility to synaptic function. Annu Rev Neurosci 17: 341-371, 1994.

Hicks T, Lodge D, and H McLennan. Excitatory Amino Acid Transmission. New York: Alan R.

Liss, 1987.

Hoffman DA, Magee JC, Colbert CM, and Johnston D. K+ channel regulation of signal

propagation in dendrites of hippocampal pyramidal neurons. Nature 387: 869-875, 1997.

Kaila K. Ionic basis of GABAA receptor channel function in the nervous system. Prog

Neurobiol 42: 489-537, 1994.

Koester J and Siegelbaum S. Local signaling: passive electrical properties of the neuron. In:

Principles of Neural Science, edited by Kandel E, Schwartz J and Jessell T. New York:

McGraw-Hill, 2000.


23

Larkum ME and Zhu JJ. Signaling of layer 1 and whisker-evoked Ca2+ and Na+ action

potentials in distal and terminal dendrites of rat neocortical pyramidal neurons in vitro and in

vivo. J Neurosci 22: 6991-7005, 2002.

Larkum ME, Zhu JJ, and Sakmann B. Dendritic mechanisms underlying the coupling of the

dendritic with the axonal action potential initiation zone of adult rat layer 5 pyramidal neurons. J

Physiol 533: 447-466, 2001.

Lester RA, Clements JD, Westbrook GL, and Jahr CE. Channel kinetics determine the time

course of NMDA receptor-mediated synaptic currents. Nature 346: 565-567, 1990.

Lisman JE. Bursts as a unit of neural information: making unreliable synapses reliable. Trends

Neurosci 20: 38-43, 1997.

Macdonald RL and Olsen RW. GABAA receptor channels. Annu Rev Neurosci 17: 569-602,

1994.

Magee JC and Johnston D. A synaptically controlled, associative signal for Hebbian plasticity in

hippocampal neurons. Science 275: 209-213, 1997.

Perreault P and Avoli M. A depolarizing inhibitory postsynaptic potential activated by

synaptically released gamma-aminobutyric acid under physiological conditions in rat

hippocampal pyramidal cells. Can J Physiol Pharmacol 66: 1100-1102, 1988.

Ranck JB, Jr. Studies on single neurons in dorsal hippocampal formation and septum in

unrestrained rats. I. Behavioral correlates and firing repertoires. Exp Neurol 41: 461-531, 1973.

Regehr W, Kehoe JS, Ascher P, and Armstrong C. Synaptically triggered action potentials in

dendrites. Neuron 11: 145-151, 1993.

Roberts P, Storm-Mathisen J, and Johnston G. Gultamate Transmission in the Central Nervous

System. Chichester: John Wiley & Sons, 1981.


24

Schiller J, Major G, Koester HJ, and Schiller Y. NMDA spikes in basal dendrites of cortical

pyramidal neurons. Nature 404: 285-289, 2000.

Schwartzkroin PA. Regulation of Excitability in Hippocampal Neurons. New York: Plenum

Press, 1986.

Sivilotti L and Nistri A. GABA receptor mechanisms in the central nervous system. Prog

Neurobiol 36: 35-92, 1991.

Staley KJ, Soldo BL, and Proctor WR. Ionic mechanisms of neuronal excitation by inhibitory

GABAA receptors. Science 269: 977-981, 1995.

Stein V and Nicoll RA. GABA generates excitement. Neuron 37: 375-378, 2003.

Stuart G, Schiller J, and Sakmann B. Action potential initiation and propagation in rat

neocortical pyramidal neurons. J Physiol 505 ( Pt 3): 617-632, 1997.

Taira T, Lamsa K, and Kaila K. Posttetanic excitation mediated by GABA(A) receptors in rat

CA1 pyramidal neurons. J Neurophysiol 77: 2213-2218, 1997.

Thalmann RH, Peck EJ, and Ayala GF. Biphasic response of hippocampal pyramidal neurons to

GABA. Neurosci Lett 21: 319-324, 1981.

Thomas MJ, Watabe AM, Moody TD, Makhinson M, and O'Dell TJ. Postsynaptic complex spike

bursting enables the induction of LTP by theta frequency synaptic stimulation. J Neurosci 18:

7118-7126, 1998.

Watanabe S, Hoffman DA, Migliore M, and Johnston D. Dendritic K+ channels contribute to

spike-timing dependent long-term potentiation in hippocampal pyramidal neurons. Proc Natl

Acad Sci U S A 99: 8366-8371, 2002.

Wong RK and Watkins DJ. Cellular factors influencing GABA response in hippocampal

pyramidal cells. J Neurophysiol 48: 938-951, 1982.


25

FIGURE LEGENDS

Figure 1. Schematic diagram of the stimulating and recording configuration. CC:

contralateral commissural pathway; SC: Schaffer collaterals pathway.

Figure 2. Synaptic response in CA1 neurons to contralateral commisural pathway

stimulation. A. Representative trace showing that a weak synaptic stimuli (1.5T) elicited an

EPSP followed by an IPSP. B. Stronger synaptic stimuli elicited L-PSPs in addition to the initial

EPSPs. The amplitude of L-PSPs increased with increasing stimulus intensity and saturated at

about 5T stimulus intensity (1 - 6 traces are responses to 1 - 6 times of threshold stimulus

intensities respectively). C. L-PSP (arrow) was associated with an increase in membrane

conductance as indicated by the decrease in the membrane potential deflections in response to a

series of 0.5 nA hyperpolarizing current steps. D. L-PSP could not be induced by direct

suprathreshold current injection to the soma. E. Sometimes a single shock to the afferent could

induce a burst of spikes from L-PSP. F. Paired- pulse stimulation (interpulse interval: 50 ms)

potentiated L-PSP responses and the burst of spikes could consistently be generated from the L-

PSP. Except panel D, all traces are averages of 4 consecutive recordings and the action

potentials are truncated. Scale bars in panel E applies to panel F.

Figure 3. Different responses evoked from CA1 neurons. A. Representative traces showing

the L-PSPs evoked from the same neuron by stimulation the CC and the SC respectively. B.

Input-output curves obtained from the CC and SC stimulation showed no significant difference.

C. In some neurons, stimulation of CC or SC could not induce L-PSP. Traces in panel A and C


26

are the average of four consecutive recordings and the action potentials are truncated. The scale

bar in panel A applies to panel C.

Figure 4. Comparison of the initial EPSP between the neurons that display the L-PSP and

other neurons. A. Representative traces of the initial EPSPs from the L-PSP and non L-PSP

neurons at 1.5T stimulus intensity. Scale bar in a also applies to b and c. B. The amplitude and

rising slope of the initial EPSP showed no significant difference between the two groups,

whereas the duration and decay time constant of the initial EPSP in the L-PSP neurons are

significantly greater than those of the non-L-PSP neurons (*, p <0.05). Numbers of cells are

denoted in parenthesis.

Figure 5. Linear regression analysis between the L-PSPs elicited at 6T stimulus intensity

and the initial EPSPs elicited at 1.5T stimulus intensity. A and B. A linear relationship was

detected between the initial EPSP decay time constant and L-PSP amplitude (R2 = 0.73, p < 0.01,

n = 40) and the L-PSP duration (R2 = 0.70, p < 0.01) respectively. C and D. There was no linear

relationship between the initial EPSP rising slope and the L-PSP amplitude (R2 = 0.004, p = 0.68)

and the L-PSP duration (R2 = 0.02, p = 0.34, n = 42) respectively.

Figure 6. Reversal potentials of different components of the post-synaptic potentials

evoked at 6T stimulus intensity. A. Representative traces showing the post-synaptic potentials

at different membrane potentials. The dashed lines indicate the positions for measurement of the

initial EPSP, fast IPSP, L-PSP and slow IPSP amplitudes respectively. B1. Linear regression of

the initial EPSP amplitude as a function of the membrane potentials. The reversal potential of


27

the initial EPSP was – 46.1 mV (n = 18). B2. Linear regression of the fast IPSP amplitude as a

function of the membrane potentials. The reversal potential of the fast IPSP was – 65.5 mV (n =

18). B3. Linear regression of the L-PSP amplitude as a function of the membrane potentials.

The reversal potential of the L-PSP was – 58.4 mV (n = 18). B4. Linear regression of the slow

IPSP amplitude as a function of the membrane potentials. The reversal potential of the slow

IPSP was – 94.4 mV (n = 18).

Figure 7. The effect of systemic administration of parmacological blockers on the L-PSPs.

L-PSP was elicited with 6T afferent stimulation. A. Representative traces showing that the

amplitude of L-PSPs was significantly reduced by the NMDA channel blocker MK-801. B.

Application of GABAA channel blocker PTX significantly reduced the amplitude of the L-PSP.

C. Simultaneous application of both MK-801 and PTX significantly reduced the amplitude of the

L-PSP. D. Group data summarizing the reduction ratio of the L-PSP amplitude after application

of MK-801 or/and PTX (**, p < 0.01).

Figure 8. The anatomical distribution and spine density comparison between the L-PSP

neurons and the non-L-PSP neurons. A. The non L-PSP neurons (left panel, open circles) and

L-PSP neurons (right panel, filled squares) showed no detectable difference in the patterns of

their anatomical distribution. B. Example light photomicrographs showing the dendritic spines

from non L-PSP neurons (left column) and L-PSP neurons (right column). Panel 1 and 2 are the

spines on the basal dendrites. Panel 3 and 4 are the spines on the proximal portion of the apical

dendrites. Panel 5 and 6 are the spines on the distal portion of the apical dendrites. C.

Quantitative comparison of spine density between the L-PSP and non L-PSP neurons. The spine


28

density of the basal dendrites and the proximal portion of the apical dendrites in the L-PSP

neurons was significantly higher than those of the non-L-PSP neurons (*, p < 0.05).

Figure 9. Morphometric comparisons of the dendritic arborization between the neurons

that display the L-PSP and other neurons. A. Representative 3-dimensional neuro-lucida

reconstructions of the L-PSP and non L-PSP neurons in the CA1 region. B. Sholl plots showing

the number of intersections made by basal and apical dendrites as a function of distance from the

soma in the L-PSP and non L-PSP neurons. The branches of the proximal apical dendrites (170

– 300 µm) in the L-PSP neurons are significantly more than those of the non L-PSP neurons (**,

p < 0.01). C. Summary of numeric analysis results. In basal dendrites, no significant difference

was detected between the L-PSP and non L-PSP neurons in the total dendritic branch ends

(TDBE), the total dendritic branch length (TDBL) and the average dendritic branch length

(ADBL). In apical dendrites, the values of TDBE, TDBL and ADBL in L-PSP neurons were

significantly greater than those of non L-PSP neurons (*, p < 0.05). ADBL values were obtained

by dividing the corresponding TDBL by the number of primary branches directly from the soma.

SC stimulation

CC stimulation Intracellular recording

Fan et al. Fig. 1

100 ms

10 mVEPSPIPSP

-70 mV

1 23

4 5 6&

L-PSP

-72 mV

10 mV

100 ms

10 mV100 ms

-70 mV

0.5 nA

10 mV100 ms

-72 mV

0.5 nA

-70 mV

10 mV

100 ms

-70 mV

A B

C D

E F

Fan et al. Fig. 2

10 mV

100 ms

SC stimulation

-70 mV

CC stimulation

-70 mV

L-PSP neuronA

SC stimulation

-70 mV

Non L-PSP neuron

-72 mV

CC stimulation

B

1 2 64 53-4

-2

0

2

4

6

8CC stimulationSC stimulation

Stimulus intensity (T)

L-P

SPam

plitu

de(m

V)

C

Fan et al. Fig. 3

L-PSP neuron

Non L-PSP neuron

0

2

4

6

8

10

12

14

16

(21)

(8)

(5)

(21)

(21)

(5)

(19)*

*

Amplitude (mV) Rising slope (ms) Duration (ms) Decay time (ms)

Mea

nva

lue

(mV

)

B

(8)

Non L-PSP neuron

10 mV20 ms-70 mV

a

L-PSP neuron-70 mV

b

Superimposed traces

cA

Fan et al. Fig. 4

2

4

6

8

10

12

14

16

18

20

22

0 5 10 15 20 25 30

L-P

SPam

plitu

de(m

V)

35

A

50

100

150

200

250

300

350

0 5 10 15 20 25 30 35L

-PSP

dura

tion

(ms)

Initial EPSP decay (ms)

B

50

100

150

200

250

300

350

L-P

SPdu

ratio

n(m

s)

0 1 2 3 4 5Initial EPSP rising slope (ms)

D

02468

10

1214161820

0 1 2 3 4 5

22

L-P

SPam

plitu

de(m

V)

Initial EPSP rising slope (ms)

CInitial EPSP decay (ms)

Fan et al. Fig. 5

-120 -100 -80 -60 -40

60

50

40

30

20

10

0

-10

Initial EPSP amplitude1 2B Fast IPSP amplitude

Membrane Potential (mV)

Am

plitu

de(m

V)

Reversal potential -46.10 mV

-100 -80 -60 -40 -20


Am

plitu

de(m

V)

50

40

30

20

10

0

-10

-20


L-PSP amplitude3 Slow IPSP amplitude4


Am

plitu

de(m

V)

Reversal potential -94.43 mV30

20

10

0

-10

-20

-30

-40

-120 -100 -80 -60 -40

-120 -100 -80 -60 -40Am

plitu

de(m

V)



50

40

30

20

10

0

-10

-20

A

-110 mV-100 mV

-90 mV-80 mV-70 mV-60 mV-50 mV

10 mV50 ms

Fan et al. Fig. 6

-72 mV

control

PTX

100 ms

5 mV

B

-69 mV

MK-801

control

A

0

0.2

0.4

0.6

0.8

1.0

1.2

Before After Before AfterBeforeAfter

L-P

SPam

plitu

de(%

)

MK-801 PTX MK-801+PTX

D

MK-801+PTX

control

-71 mV

C

** ** **

Fan et al. Fig. 7

DG CA3

CA1

DG CA3

CA1

A

1

3

5

2

4

6

2 µm

B

Proximal Distal

L-PSP 12 34.5±2.8 37.3±3.9 30.6±3.6

Non L-PSP 5 26.7±2.9* 31.6±2.3* 28.9±3.4

n

The spine density of L-PSP and Non L-PSP neurons (per 20 µm)C

Non L-PSP neuron L-PSP neuron

Fan et al. Fig. 8

X-plan Y-plan Z-plan

L-PSP

Non L-PSP

100 µm

Basal dendrite

Apical dendrite

Morphometric analysis of dendrites in L-PSP and non L-PSP neurons

Cell type n TDBE TDBL ADBL

L-PSP (6) 41.17±3.22 7228.97±940.96 2304.95±291.06

Non L-PSP (4) 44.75±2.32 7105.78±293.91 2000.40±287.12

L-PSP (6) 58.83±6.79* 9887.30±902.95 * 9887.30±902.95 *

Non L-PSP (4) 42.50±5.24 7327.53±382.01 7327.53±382.01

0

10

20

30

40

50

10 60 110 160 210 260 310

Basal dendrite

Distance from soma (µm)

Inte

rsec

tion

0

5

10

15

20

25

30

10 110 210 310 410 510 610

Inte

rsec

tion

Apical dendrite

Distance from soma (µm)

**

L-PSP

Non L-PSP

A

B

C

Fan et al. Fig. 9

Table 1. Membrane properties of the L-PSP neurons and the non L-PSP neurons

Group RMP

(mV)

-67.0 + 0.80 75.5 + 1.01 1.07 + 0.05 -57.6 + 1.00 0.16 + 0.01 30.1 + 1.53 11.0 + 0.76

(22) (19) (19) (18) (13) (29) (29)

SpkH

(mV)

SpkD

(ms)

SpkT

(mV)

Rheob

(nA)

Rin

(M½ ) (ms)

L-PSP

Non L-PSP -67.6 + 0.79 73.5 + 2.53 1.04 + 0.08 -58.5 + 0.87 0.18 + 0.02 24.3 + 1.48 8.5 + 0.76

(11) (11) (11) (11) (9) (13) (13)

**

Value are Mean + S.E., with number of neurons in parentheses. Spike height (SpkH) was measured from the resting membranepotential. Spike duration (SpkD) was measured at the half of the peak amplitude of the action potential. Input resistance (Rin)was derived from the linear portion of the current-voltage curve (-0.5 to 0 nA). Time constant ( ) was derived from the tran-sient of hyperpolarizing pulse (-0.3 nA, 200 ms). RMP, resting membrane potential. SpkT, spike threshold. Rheob, rheobase.* p < 0.05. Un-paired Student’s t test.

τm

τm

Fan et al. Table. 1

In vivo Demonstration of a Late Depolarizing Postsynaptic...

Documents

Transcript of In vivo Demonstration of a Late Depolarizing Postsynaptic...