Synaptic Depression: Many Different Possible Mechanisms

Synaptic Depression
Many different possible

mechanisms
Von Gersdorff and Borst (2002)

Nat Rev Neurosci 3, 53-64
Average PSP = npq
n = number of ready-releasable vesicles at the synapse (RRP: ready-
releasable pool);
n coincides with the number of primed vesicles at synapses where multivesicular
release can occur or with the number of active zones at synapses where only one
vesicle per active zone can be released
p= probability of release of a ready-releasable vesicle
p depends on:
- the local Ca2+ transient (which depends on the number and kinetics of CaV at the
active zone and their distance from the primed vesicle, the AP shape, the Ca2+ buffer
and the Ca2+ extrusion mechanisms at the active zone)
- the affinity of the Ca2+ sensor
-q= quantal amplitude (PSP produced by release of one vesicle)

q depends on the number and affinity of postsynaptic receptors and the content of
neurotransmitter in a vesicle
n and p depend on (and may change during AP sequences as a consequence of )

presynaptic mechanisms
q depends on (and may change during AP sequences as a consequence of)

postsynaptic mechanisms (assuming constant content of neurotrasmitter in vesicle)
At most synapses:
1) If one decreases [Ca]ext, the depression of the EPSC during

repetitive activity is reduced
 Short-term depression (STD) depends on the initial probability of
release at a given synapse: the higher is the probability of
release, the larger is the depression of the EPSC during
repetitive activity
2) STD depends on the frequency of APs during repetitive activity: :

the larger the frequency, the larger is the depression
The time constant of recovery from depression varies from

hundreds of ms to tens of seconds at different synapses
 The STD is due (at least in part) to the depletion of the ready-
releasable pool (RRP) of vesicles; the velocity of recovery from
depression mainly reflects the velocity of repopulation of the RRP.
Calyx of Held
Simple model in which it is assumed that:
i) the probability of release of a vesicle does not depend on the dimension of the RRP;
ii) the velocity of repopulation of the pool is constant
EPSC = npq
EPSC1 = n1p1 q
n2 = n1-n1p1 = n1(1-p1)
EPSC2 = n1(1-p1)p2 q
EPSC2/EPSC1 = (1-p1) (p2/p1)
if p2=p1 EPSC2/EPSC1 = 1-p1
 The larger is the probability of release p1, the smaller is the EPSC2/EPSC1 ratio,
i.e.larger is the depression
Simple model in which it is assumed that:
i) the probability of release of a vesicle does not depend on the dimension of the RRP;
ii) the velocity of repopulation of the pool is constant
EPSC = npq
EPSC1 = n1p1 q
n2 = n1-n1p1 = n1(1-p1)
EPSC2 = n1(1-p1)p2 q
EPSC2/EPSC1 = (1-p1) (p2/p1)
if p2=p1 EPSC2/EPSC1 = 1-p1
 The larger is the probability of release p1, the smaller is the EPSC2/EPSC1 ratio,
i.e.larger is the depression
But the model is too simple: in the Calyx of Held the STD during a 10-100 Hz train of
APs is smaller than expected on the basis of this model, indicating that either
-p is not constant but decreases with decreasing RRP or
-the velocity of repopulation is not constant but increases during the repetitive stimulation
.
Experiments at the Calyx of Held (performed by E Neher and collaborators)

have shown that:
There are two distinct pools of RRP vesicles with different p release and
different kinetics of repopulation:
-fast releasing (high p, synchronous release during AP, slow recruitment-
repopulation velocity: hundreds ms-sec),
-slowly releasing (low p, asynchronous release during AP, high repopulation
velocity: < 100 ms).
Velocity of recruitment of fast-releasing vesicles

-depends on [Ca]in (and also on cAMP); it depends linearly on [Ca]in
-it increases during a train of APs at high frequency
Molecular priming: involves SNARE, Munc13, complexin etc metastable state for vesicle fusion
Positional priming: primed vesicles are near Ca channels (cf RIM, RIMBP linking Ca channels to vesicular
proteins)
Fast releasing vesicles are both molecularly and positionally primed

Slowly releasing vesicles are molecularly, but not positionally primed
Molecular priming (step 4) is fast (<100 ms); positional priming (step4a) is rate limiting for
recruitment of fast releasing vesicles and is Ca-dependent
Since there is a limited number of binding sites mediating Ca channel-vesicle interaction, site
clearing may be the rate limiting step for positional priming (cf recent evidence that endocytosis limits the
extent of depression by accelerating this step)
Complexin
CaV2.1 and CaV2.2 channels bind to the active zone proteins RIM and RIM-BPs.
RIM binds to the Rab3/27 small GTPase proteins of the synaptic vesicles
Sudhof (2013) Neuron

Synaptic Depression
mechanisms

Calyx of Held
STD is not due to change in

shape and amplitude of the AP
during repetitive activity
Action potential clamp experiments

have shown that, if anything, the
changes in shape and amplitude of
the AP would produce a small
increase of the EPSC during the train
(because the effect of prolonging the
duration of the AP and hence of the
Ca current prevails over the the
effect of decreasing the amplitude of
the AP and of the Ca current)
Calyx of Held
STD is not due to change in

shape and amplitude of the AP
during repetitive activity.
In contrast, at the inhibitory

synapses between Purkinje cells
and deep cerebellar nuclei
neurons (PC-DCN synapses:
patch-clamp on 3 um terminals
of PC!!), the effect of reducing
the amplitude of the AP during
the train prevails over the effect
of prolonging its duration (due to
the different NaV and KV
channels at the terminals).
AT PC-DCN synapses this
mechanism accounts for the
majority of STD (which indeed
does not depend much on the
initial p release at this synapse)
(Kawaguchi and Sakaba, 2015,
Neuron)
Calyx of Held
STD during high-frequency
train of APs (100 Hz) is not
due to inactivation of
presynaptic CaV channels.
However, at lower AP frequency

(<30 Hz) Ca-dependent
inactivation of CaV channels does
contribute to STD. Moreover the
contribution of this mechanism
depends also on the age: larger
contribution in young animals.
 The mechanism of STD at a

given synapse can vary with
the frequency of repetitive
activity and with age
Calyx of Held
STD during high-frequency
train of APs (100 Hz) is not
due to inactivation of
presynaptic CaV channels.
However, at lower AP frequency

(<30 Hz) Ca-dependent
inactivation of CaV channels does
contribute to STD. Moreover the
contribution of this mechanism
depends also on the age: larger
contribution in young animals.
The contribution of Ca-

dependent inactivation of CaV
channels to STD is different at
different synapses:
e.g. in transfected superior cervical
ganglion neurons deletion of CBD
from the C terminus of CaV2.1
channels reduces STD during AP
trains up to 40 Hz
Synaptic Depression
mechanisms

Calyx of Held:
desensitization of postsynaptic glutamate receptors contributes to STD
In the presence of cyclothiazide (a drug that inhibits desensitization)

the STD is reduced
In general (but not always) synapses with high probability of release (e.g. cerebellar CF-PC
synapses) depress during repetitive activity, while synapses with low probability of release
(e.g. cerebellar PF-PC synapses) facilitate during repetitive activity.
Synapses with
intermediate probability of
release (e.g. hippocampal
Ca3-CA1 synapses) may
first facilitate and then
depress during repetitive
activity.
Atwood and Karunanithi (2002) Nature

Rev. Neurosci. (2002) 3 497
Even synapses formed by the same presynaptic neuron on different
postsynaptic neurons can have different STP properties
(e.g. cf synapses between cortical L2/3 pyramidal neurons and FS interneurons or
SOM interneurons)
- Target cell-dependent STP properties

Due to feedback modulation by postsynaptic neuron
The synapse between layer 2/3 pyramidal cells and FS interneurons show short-term
depression during 10 Hz AP trains in wild-type (WT) mice.
The STD at this synapse is enhanced in knockin mouse models of migraine (FHM1 KI),
which carry a gain-of-fnction mutation in the CaV2.1 channel that increases the probability
of glutamate release (cf two times larger EPSP at the 1° pulse in the train)
PYR FS
WT
FHM1 KI
WT (n=16)
FHM1 KI (n=17)
WT FHM1
Tottene et al, 2009 Neuron

The synapse between layer 2/3 pyramidal cells and SOM interneurons show short-term
facilitation during 25 Hz AP trains in wild-type (WT) mice.
The short-term facilitation is unaltered at this synapse in knockin mouse models of

migraine (FHM1 KI), which carry a gain-of-fnction mutation in the CaV2.1 channel that
increases the probability of glutamate release (cf about two times larger EPSP at the 1°
pulse in the train and also subsequent pulses)
25 10 EPSP5/EPSP1
25 Hz
20 8
EPSPi / EPSP1
EPSPi / EPSP1
15 6
PYR SOM+
EPSP2/EPSP1
10 4
5 2
0.2 mV 50 ms 15 18 15 18
0 0
WT FHM1 WT FHM1
Pilati, Forli et al.

As a consequence of synaptic plasticity, the effect of an AP in
a presynaptic neuron on the activity of the postsynaptic
neuron can greatly vary depending on the previous sequence
of presynaptic APs.
Irregular train of presynaptic APs (average frequency 20 Hz)
to simulate physiological activity
EPSC10/EPSC1 as a function of the
AP frequency in a regular
Low-pass filter
EPSC
High
Prelease
High-pass filter
Low
Prelease
Band-pass filter
Intermediate
Prelease
NB Neuromodulators modulate the synaptic plasticity!

Synaptic Depression: Many Different Possible Mechanisms

Uploaded by

Copyright:

Available Formats

Synaptic Depression: Many Different Possible Mechanisms

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Synaptic Depression: Many Different Possible Mechanisms

Uploaded by

Copyright:

Available Formats

Synaptic Depression

Many different possible

Von Gersdorff and Borst (2002)

p= probability of release of a ready-releasable vesicle

-q= quantal amplitude (PSP produced by release of one vesicle)

n and p depend on (and may change during AP sequences as a consequence of )

q depends on (and may change during AP sequences as a consequence of)

1) If one decreases [Ca]ext, the depression of the EPSC during

2) STD depends on the frequency of APs during repetitive activity: :

The time constant of recovery from depression varies from

EPSC2/EPSC1 = (1-p1) (p2/p1)

if p2=p1 EPSC2/EPSC1 = 1-p1

EPSC2/EPSC1 = (1-p1) (p2/p1)

if p2=p1 EPSC2/EPSC1 = 1-p1

Experiments at the Calyx of Held (performed by E Neher and collaborators)

Velocity of recruitment of fast-releasing vesicles

Fast releasing vesicles are both molecularly and positionally primed

Sudhof (2013) Neuron

Von Gersdorff and Borst (2002)

STD is not due to change in

Action potential clamp experiments

STD is not due to change in

In contrast, at the inhibitory

However, at lower AP frequency

 The mechanism of STD at a

However, at lower AP frequency

The contribution of Ca-

Von Gersdorff and Borst (2002)

In the presence of cyclothiazide (a drug that inhibits desensitization)

Atwood and Karunanithi (2002) Nature

- Target cell-dependent STP properties

Tottene et al, 2009 Neuron

The short-term facilitation is unaltered at this synapse in knockin mouse models of

Pilati, Forli et al.

NB Neuromodulators modulate the synaptic plasticity!

You might also like