Tom Wang
Afferent Transmission Discussion
Overview
Afferent transmission at the inner hair cell (IHC) synapse results from depolarization of
the intracellular potential. This can result from either acoustic stimulation or random
fluctuations in IHC potential. In the case of acoustic stimulation, K+ ion influx through
mechanically gated transduction channels of the stereocilia depolarizes the IHC, opening
voltage-gated Ca2+ channels (VGCC). The subsequent Ca2+ influx induces a sequence
of events resulting finally in exocytosis. Synaptic vesicles located near the ribbon
synapse fuse with the cell membrane and release neurotransmitters onto boutons of
postsynaptic afferent nerve fibers. 10~30 afferent fibers form boutons on a single hair
cell and are each excited mostly by a single ribbon. Transmitter release induces
excitatory postsynaptic potentials (EPSP) on the afferent fibers, which generate action
potentials on the auditory nerve. The overall process translates acoustic stimuli into
electrical signals to be processed by the central auditory system. The discussion papers
address current knowledge of the mechanisms and kinetics of the transmission process.
[1] Moser, T. and Beutner, D. (2000)."Kinetics of exocytosis and endocytosis at the
cochlear inner hair cell afferent synapse of the mouse," Proc. Nat. Acad. Sci. 97,
883-888.
During exocytosis, vesicle fusion with the plasma membrane increases the total surface
area of the cell. Since cell membrane capacitance (Cm) is proportional to the area across
which it is measured, exocytosis can be expected to increase Cm. Conversely,
endocytosis would reduce cell surface area and likely decrease Cm. In this paper, the
authors utilizes capacitance measurements of the IHC membrane to substantiate the claim
that transmitter release is due to Ca2+-induced exocytosis; in addition, temporal aspects
of transmitter release and replenishment are described.
IHC membrane capacitance and intracellular [Ca2+] was measured while simulating
mechanoelectric depolarization using Ca2+-ion inward currents. Initial measurements
showed that increases in Cm were dependent on the influx of Ca2+ ions. When
extracellular [Ca2+] was removed, depolarization did not result in changes in Cm. The L-
type Ca2+ channel blocker nifedipine also inhibited changes in Cm and blocked Ca2+
influx as well; furthermore the Ca2+ ion channel blocker cobalt completely eliminated
Ca2+ influx and Cm changes upon depolarization. Finally, Cm measurements after
current injections slowly decreased. The authors argue based on these results that the
likely mode of transmitter release is exocytosis and that changes in Cm are indicative of
both exocytosis and endocytosis. Based on these arguments, kinetic aspects of IHC
afferent transmission were studied.
The time course for the change in Cm (dCm) was found to vary throughout the
presentation of the current stimulus and was fitted with two time constants (biphasic).
Near the onset of the stimulus, the time constant describing dCm was much faster than
that throughout the duration. In relating change in capacitance to the number of vesicles
released (~37 aF per vesicle), the rate of vesicle release was estimated as 28,000/s for the
initial fast component. During the sustained component of vesicle release, the rate was
estimated as 8,700/s. The initially large release rate is suggested to result from
exocytosis of a readily releasable pool (RRP) of vesicles at the active zone. The
replenishment of this RRP was studied by estimating accretion vesicle rates after short
stimuli; the recovery was found to be rapid and biphasic, with a maximum rate of 1200/s.
The time course of this recovery is significantly faster than that of endocytosis,
suggesting another mechanism of rapid replenishment. However, the authors do not rule
out the role of endocytosis for longer duration stimuli. Likewise, a significant question
posed by this data is how the sustained component of exocytosis maintains such a rate
high rate of vesicle release (8,700/s). The maximum rate of recovery for the RRP (1200/s)
is insufficient to replenish depletion of the IHC. The authors suggest two hypotheses: 1)
a rapid mechanism of vesicle mobilization near the synaptic ribbon allows the RRP to
maintain the rate at long durations or 2) vesicle diffusion occurs not only in the active
region but at longer distances away from it, thereby allowing sustained release.
It is noted that the biphasic behavior of vesicle release parallels auditory nerve fiber
behavior during long-duration sounds. In particular, action potential firing rates are
initially high and quickly decrease to a sustained rate; this is known as adaptation. This
behavior is similar to the initial high release rate of vesicles followed by a slower rate
when the RRP is assumed to be depleted. The authors therefore suggest that RRP
depletion may be one of many factors that contribute to adaptation.
[2] Griesinger Richards and Ashmore “Fast vesicle replenishment allows
indefatigable signaling at the first auditory synapse”. Nature 2005, 435:212-215
Synaptic vesicles are endocytosed from the apical end of the IHC and transported to the
basolateral regions for transmission. In this paper, the authors bathe the apical surface of
IHCs with a fluorescent dye (FM1-43) to label vesicles formed by this process. The
IHCs were then stimulated transepithelially, and fluorescence was used to track the
temporal behavior of vesicles as well as estimate vesicle count before, during, and after
stimulation. Initial trials showed destaining of fluorescence during stimulation,
suggesting exocytosis of labeled vesicles. A control experiment used nimodipine to
block L-type calcium channels responsible for this process and showed that destaining
was not altered during stimulation; therefore, vesicle release corresponded to destaining.
Vesicle numbers estimated throughout stimulation suggest a maximal vesicle release rate
of 3/ms initially, while later slowing to a sustained rate of 1.4/ms. Post-stimulus
fluorescence staining was also observed, suggesting vesicle replenishment. The rate of
recovery is estimated to be 1.1/ms. Though these exact rates are not consistent with those
estimated in [1], they do show biphasic behavior, thereby suggesting two stages of
vesicle release. Furthermore, the authors observe, from fluorescence, a small pool of
~100 vesicles near the synaptic ribbon prior to stimulation. This is consistent with the
hypothesis of a readily-releasable pool accounting for the initial large rate of release. A
smaller sustained release rate is observed during prolonged stimulation and is coupled
with a similar post-stimulus replenishment rate. The authors argue that the replenishment
mechanism is unlikely endocytosis, since it has a recovery rate of 10/s (much slower than
those observed). Furthermore, observed recovery rates are independent of stimulus
amplitude, suggesting that replenishment mechanisms are not coupled with initial release
rates. Replenishment of vesicles is suggested to result from a rapid mechanism whereby
the ribbon synapse mobilizes preformed vesicles in the cytosol toward the synapse during
sustained release.
A major caveat of this paper is that estimated recovery rates using the available technique
were made after the stimulus was off; therefore, it is not possible to confirm that recovery
rates during sustained stimulation are indeed those observed. Furthermore, like the
previous paper, the authors argue that endocytosis cannot be the rapid mechanism used to
replenish vesicles during sustained release; however, it cannot be ruled out as a
contributing factor to recovery at longer post-stimulus times.
[3] Glowatzki, E. and Fuchs, P. A. (2002)."Transmitter release at the hair cell
ribbon synapse," Nat Neurosci 5, 147-154.
Vesicle release at the IHC ribbon synapse induces excitatory postsynaptic currents on it
respective afferent nerve fiber. In this paper, EPSCs were recorded after inducing
afferent transmission to infer characteristics of vesicle release. In particular, the authors
sought to address the hypothesis that vesicle release can be either singular and
uncoordinated or multivesicular and coordinated. Cochlear sections were isolated, and
whole-cell patch-clamp methods were used to stimulate IHCs and measure afferent fiber
currents.
Initial results showed spontaneous EPSCs in an extracellular solution containing 5.8 mM
[K+] with roughly 1.5/s. Elevated [K+] of 40 mM resulted in 27.2/s. The shape of the
current in both cases fell into two groups: 1) 70% of EPSCs had a wide range of
amplitudes with fast rise times and slow decay times and were termed “monophasic” 2)
“Multiphasic” EPSCs also had different amplitudes but had varying waveforms (no
specific rise/fall time behavior); these represented 30% of all EPSCs. Based on their
hypothesis, the authors suggest that large monophasic responses could be due to
coordinate multi-vesicular release, while smaller monophasic EPSCs are due to singular
release. In contrast, multiphasic responses may be the result of uncoordinated singular
release.
Figure 1. Monophasic (left) vs. multiphasic EPSCs.
The observed waveforms were also assessed quantitatively. Figure 2 shows EPSC
amplitude plotted against the time constant of delay along with the rise time. The authors
argue that the data is consistent with their hypothesis based on several suggestions.
Large EPSCs from coordinated multivesicular release would have small rise and decay
times; however, as the release becomes more dispersed, the amplitude would be reduced
and both times would increase. In addition, smaller EPSCs arising from singular vesicle
release would also have short rise and decay times. Small rise and fall times should
therefore show a large range of amplitudes, but at larger rise and fall times, the
amplitudes should approach smaller values.
Figure 2. EPSC amplitudes with respect to decay (left) and rise (right) times.
Two further assessments were also done to address the hypothesis. Histograms of EPSC
amplitudes and interevent intervals were constructed for both 40 mM [K+] (depolarized)
and 5.8 mM (rest) (Figures 3, 4). The amplitude histograms both show a bias towards
smaller amplitudes, though in each case, the spread of amplitudes towards higher levels is
significant. The authors argue that this non-Gaussian distribution would not occur if only
single vesicles were released (in which case, a Gaussian distribution would likely result).
Furthermore, interevent histograms have peaks at smaller times, despite having a large
spread towards longer times. The interevent histograms are fitted with two different time
constants, which the authors suggest as arising from two different stochastic processes of
vesicle release: the smaller intervals represent the uncoordinated single vesicles while
larger intervals correspond to coordinate multivesicular release.
Figure 3. EPSC amplitude histograms for 40 mM (left) and 5.8 mM (right) [K+].
Figure 1. Monophasic (left) vs. multiphasic EPSCs.
Figure 4. EPSC interevent histograms for 40 mM (left) and 5.8 mM (right) [K+].
A final experiment sought to identify the receptor by which the EPSCs were mediated.
CNQX was administered which blocks both AMPA and kainate receptors; in this case,
EPSCs were completely blocked at positive and negative holding potentials. At positive
holding potentials amidst CNQX, NMDA receptors would likely become unblocked also.
Cyclothiazide was then administered to reduce desensitization of AMPA receptors but
not kainate. In this case, EPSCs were observed with a slower decay time. The authors
argue from the CNQX results that NMBA was unlikely the mediator, since it was
unblocked but no EPSCs were observed. Furthermore, they argue that since EPSCs
existed presumably without kainate (desensitized), the major receptor mediating EPSCs
was AMPA.