Detecting synaptic connections in the medial nucleus of the trapezoid body using calcium imaging

Abstract

The study of synaptic transmission in brain slices generally entails the patch-clamping of postsynaptic neurones and stimulation of identified presynaptic axons using a remote electrical stimulating electrode. Although patch recording from postsynaptic neurones is routine, many presynaptic axons take tortuous turns and are severed in the slicing procedure, blocking propagation of the action potential to the synaptic terminal and preventing synaptic stimulation. Here we demonstrate a method of using calcium imaging to select postsynaptic cells with functional synaptic inputs prior to patch-clamp recording. We have used this method for exploring transmission in the auditory brainstem at the medial nucleus of the trapezoid body neurones, which are innervated by axons from the contralateral cochlear nucleus. Brainstem slices were briefly loaded with the calcium indicator fura-2 AM and stimulated with an electrode placed on the midline. Electrical stimulation caused a rise in intracellular calcium concentration in those postsynaptic neurones with active synaptic connections. Since <10% of the medial nucleus of the trapezoid body neurones retain viable synaptic inputs following the slicing procedure, preselecting those cells with active synapses dramatically increased our recording success. This detection method will greatly ease the study of synaptic responses in brain areas where suprathreshold synaptic inputs occur but connectivity is sparse.

Full text

Pflugers Arch - Eur J Physiol (2002) 444:663–669
DOI 10.1007/s00424-002-0861-6
ORIGINAL ARTICLE
B. Billups · A. Y. C. Wong · I. D. Forsythe
Detecting synaptic connections in the medial nucleus
of the trapezoid body using calcium imaging
Received: 11 January 2002 / Revised: 4 March 2002 / Accepted: 16 April 2002 / Published online: 29 June 2002
 Springer-Verlag 2002
Abstract The study of synaptic transmission in brain
slices generally entails the patch-clamping of postsynaptic
neurones and stimulation of identified presynaptic axons
using a remote electrical stimulating electrode. Although
patch recording from postsynaptic neurones is routine,
many presynaptic axons take tortuous turns and are
severed in the slicing procedure, blocking propagation of
the action potential to the synaptic terminal and prevent-
ing synaptic stimulation. Here we demonstrate a method
of using calcium imaging to select postsynaptic cells with
functional synaptic inputs prior to patch-clamp recording.
We have used this method for exploring transmission in
the auditory brainstem at the medial nucleus of the
trapezoid body neurones, which are innervated by axons
from the contralateral cochlear nucleus. Brainstem slices
were briefly loaded with the calcium indicator fura-2 AM
and stimulated with an electrode placed on the midline.
Electrical stimulation caused a rise in intracellular
calcium concentration in those postsynaptic neurones
with active synaptic connections. Since <10% of the
medial nucleus of the trapezoid body neurones retain
viable synaptic inputs following the slicing procedure,
preselecting those cells with active synapses dramatically
increased our recording success. This detection method
will greatly ease the study of synaptic responses in brain
areas where suprathreshold synaptic inputs occur but
connectivity is sparse.
Keywords Brain slice · Calyx of Held · Electrical
stimulation · Fura-2 · Kynurenate · Medial nucleus of the
trapezoid body · Patch-clamp · Presynaptic terminal
Introduction
The study of synaptic transmission was revolutionized by
the development of whole-cell patch-clamp recording
from brain slices [6]. The ability to record from neurones
and electrically stimulate their synaptic inputs in a
preparation that has undergone normal development has
greatly increased our understanding of neuronal commu-
nication in local circuits. However, the process of cutting
brain slices unavoidably severs many longer axons,
making it very difficult to identify functional pathways
in certain brain areas. To enhance our ability to study one
such pathway, we have developed a method to select
neurones with intact synaptic inputs by loading the brain
slice with a fluorescent calcium-sensitive probe and
observing the postsynaptic calcium concentration rise in
response to distal electrical stimulation.
We have used this method to detect functional synaptic
connections at the calyx of Held synapse. This synapse
mediates auditory transmission from the anteroventral
cochlear nucleus to the contralateral medial nucleus of the
trapezoid body (MNTB) via a long axon crossing the
midline in the trapezoid body (Fig. 1) [5, 7, 20]. It is a fast
relay synapse that faithfully transmits the presynaptic
action potential pattern at frequencies over 600 Hz [21,
24, 27] and forms part of the auditory pathway respon-
sible for sound source localization [8, 18]. Its large size
allows direct whole-cell patch-clamp recording from the
presynaptic terminal [7]. The abilities to image presyn-
aptic calcium [11], to release caged-calcium within the
terminal [2, 19] and to measure vesicle cycling with FM
1–43 [13] or capacitance change [23] make it one of the
most comprehensively studied synapses in the brain.
However, the ability to examine the postsynaptic response
at the calyx of Held synapse is limited by the difficulty in
recording from MNTB neurones that retain active
presynaptic terminals following the intrusive slicing
procedure. An excitatory postsynaptic current (EPSC)
can be stimulated in fewer than 10% of the MNTB cells
[1], even though anatomical studies indicate all the
principal cells receive calyceal inputs [21]. This makes a
B. Billups ()) · A.Y.C. Wong · I.D. Forsythe
Department of Cell Physiology and Pharmacology,
University of Leicester, PO Box 138, Leicester LE1 9HN, UK
e-mail: bjb10@le.ac.uk
Tel.: +44-116-2523303
Fax: +44-116-2525045

method for preselecting cells with functional synaptic
inputs highly advantageous.
Materials and methods
Brain slice preparation
Ten- to 14-day-old Lister hooded rats were killed by decapitation
and transverse brainstem slices (@200 m thick) were cut as
previously described [1]. The slicing medium was maintained at
around 0C and contained (mM): 250 sucrose; 2.5 KCl; 10 glucose;
1.25 NaH
2
PO
4
; 26 NaHCO
3
; 4 MgCl
2
; 0.1 CaCl
2
;3myo-inositol;
0.5 ascorbic acid and 2 Na-pyruvate (pH 7.4 when gassed with 95%
O
2
/5% CO
2
). Slices were incubated for 1 h at 37C in aCSF, and
then stored at room temperature before being transferred to the
experimental chamber for recording. MNTB neurones and calyces
of Held were visualized with infrared differential interference
contrast (DIC) optics on a Nikon E600FN microscope with a 60,
N.A. 1.0, water-immersion, fluor lens. The aCSF contained (mM):
125 NaCl; 2.5 KCl; 10 glucose; 1.25 NaH
2
PO
4
; 26 NaHCO
3
;
1 MgCl
2
; 2 CaCl
2
;3myo-inositol; 0.5 ascorbic acid and 2 Na-
pyruvate (pH 7.4 when gassed with 95% O
2
/5% CO
2
).
Calcium imaging
Slices were loaded with fura-2 by incubation in the recording
chamber with 7 M fura-2 acetoxymethyl (AM) ester for 5 min at
37C. Fluorescence loaded cells were imaged with light at 350 nm
and 380 nm from a Polychrome II monochromator (T.I.L.L.
Photonics). The incident light was heavily attenuated with neutral-
density filters and the exposure time for each frame was 100 ms per
wavelength. Emitted light was separated by a 400-nm dichroic
mirror and filtered with a 420-nm long-pass filter. Fluorescence
image pairs were acquired every 600 ms with a PentaMAX cooled
CCD camera via a Gen IV image intensifier (Princeton Instru-
ments) and analysed with MetaFluor software (Universal Imaging).
Similar experiments have also been performed using an Optoscan
monochromator (Cairn Research) and CoolSNAP fx camera (Roper
Scientific). Intracellular calcium concentrations were calculated
using the following equation [9]:
Ca2þ

i¼Kd
RRmin
Rmax R

F380
max
F380
min
 ð1Þ
where Ris the ratio of fluorescence emitted following 350 nm and
380 nm excitation. A K
d
value of 135 nM was used and R
min
and
R
max
were established as 0.38€0.02 and 2.7€0.6 (n=10) using 2 M
ionomycin and external solutions containing 0 mM (1 mM EGTA
added) and 10 mM calcium. The 350-nm excitation light was
significantly absorbed by kynurenate, so in experiments using this
drug intracellular calcium concentrations were calculated using
only the fluorescence at 380 nm excitation and the following
equation [11, 12]:
Ca2þ

i¼Ca2þ

restþKdDF=FðÞ=DF=FðÞ
max
1DF=FðÞ=DF=FðÞ
max

ð2Þ
Electrophysiological recording
Whole-cell patch-clamp recordings were made from both the
synaptic terminal and postsynaptic cell using thick-walled glass
pipettes (GC150F-7.5, Clark Electromedical) with an Axopatch
200B amplifier, filtered at 10 kHz (8-pole Bessel filter) and
sampled at 20–50 kHz. Currents were recorded with pCLAMP 8
software (Axon Instruments). Whole-cell access resistances were
<10 MWfor the postsynaptic cell and <20 MWfor the presynaptic
terminal, and were compensated >70% with a 10 s lag time. The
intracellular solution contained (mM) 110 CsCl; 40 HEPES; 10
TEA-Cl; 12 Na
2
-phosphocreatine and 1 EGTA (pH adjusted to 7.3
with CsOH); 2 mM QX314 was routinely added to the intracellular
solution to block postsynaptic sodium currents and improve the
quality of the voltage-clamp. Excitatory postsynaptic currents were
stimulated with a bipolar platinum electrode consisting of two
platinum wires 300 m apart placed above and below the slice at
the midline (Fig. 1). Electrical pulses of 2–8 V and 0.2 ms duration
were provided by a Digitimer DS2 A isolated stimulator triggered
by the pCLAMP software.
Fura-2 AM was obtained from Molecular Probes and was used
from a 1 mM stock solution in dry DMSO containing 5% pluronic
acid. All other chemicals were obtained from Sigma. Data are
expressed as the mean €SEM and statistical significance (P<0.05)
was tested with 2-tailed t-tests. All experiments were performed at
physiological temperature (35–37C).
Results
Calcium responses in synaptically connected neurones
Following fura-2 loading, fluorescence was detected in
many of the superficial MNTB principal neurones
(Fig. 2A). Brief trains of electrical stimulation at the
midline were used to activate axons in the pathway from
the contralateral cochlear nucleus to the MNTB. Principle
cells in the MNTB with active synaptic inputs showed an
obvious increase in intracellular calcium concentration
Fig. 1 Schematic diagram of the auditory brainstem. The output
from the anteroventral cochlear nucleus projects across the midline
to the contralateral medial nucleus of the trapezoid body (MNTB),
giving rise to the calyx of Held presynaptic terminals. Principle
neurones of the MNTB send inhibitory projections to the ipsilateral
medial superior olive (MSO) and lateral superior olive (LSO) in the
plane of the slice. MNTB neurones also project to the ipsilateral
ventral nucleus of the lateral lemniscus, and dorsomedial and
ventromedial periolivary nuclei (not shown). The forked stimulat-
ing electrode is placed on the midline with one prong above and
one below the slice
664

([Ca
2+
]
i
) following stimulation (Fig. 2B). However, the
majority of cells (>90%) did not respond to remote
electrical stimulation (Fig. 2C) suggesting that their axons
no longer extend as far back as the midline stimulating
electrode or that the synaptic terminal has not survived
the slicing procedure. Calcium responses in MNTB cells
resulting from antidromic stimulation were also occa-
sionally observed, giving a “false-positive” response in
the calcium imaging data. The proportion of false-positive
responses was less than half of the total number of
responding cells and could be minimized by positioning
the stimulating electrode slightly contralateral to the
MNTB under observation, since the output from the
MNTB neurones innervates the ipsilateral medial superior
olive and lateral superior olive (MSO and LSO; Fig. 1)
but does not cross to the contralateral nuclei [21, 22].
False-positive electrical stimulation was distinguished
from orthodromic stimulation by its insensitivity to the
neurotransmitter receptor antagonist. Postsynaptic [Ca
2+
]
i
responses resulting from synaptic activation were sensi-
tive to the glutamate receptor antagonist kynurenate
(Fig. 2D), whereas responses to antidromic stimulation
were unaffected. Application of 1 M TTX abolished
action potential propagation and all responses to electrical
stimulation (Fig. 2D), confirming the involvement of
regenerative sodium currents in the electrical activation
and excluding the possibility that calcium responses are a
result of a direct electrical effect on the postsynaptic cell.
Synaptic currents in identified cells
Whole-cell voltage-clamp recordings were made to
confirm that the neurone identified by calcium imaging
does receive a synaptic input. During the formation of the
electrode seal, the extracellular current recording in
response to synaptic stimulation can be examined in the
“loose-patch” configuration [10]. The postsynaptic action
potential can be detected as a small current approximately
1–2 ms after the stimulus artefact (Fig. 3A). The presence
of this current confirmed that the MNTB neurone has an
active synaptic input. When whole-cell current recording
was accomplished, the large glutamatergic EPSC was
clearly visible following stimulation (Fig. 3B). This
current is composed entirely of current through gluta-
mate-gated ion channels and is not contaminated by
regenerative sodium currents, as judged by its reversal
around 0 mV and insensitivity to the QX314 included in
the internal solution. Occasionally the recording electrode
picks up the extracellular presynaptic action potential
from the presynaptic terminal, as can be seen immediately
preceding the postsynaptic EPSC in Fig. 3C. If non-
synaptic activation of the MNTB cell was present, it
would be seen as an regenerative sodium current imme-
diately following the stimulus artefact, before the QX314
has had time to dialyse into the postsynaptic cell, as
shown in Fig. 3D.
Fig. 2A–D Detection of active
synapses with calcium imaging.
AA fluorescence image
(380 nm excitation) over the
MNTB nucleus from a slice
loaded with fura-2 AM. Scale
bar is 50 m. BThe change of
[Ca
2+
]
i
following electrical
stimulation (200 Hz for 200 ms)
of the same field of cells as A.
Scale bar is 50 m. CGraph of
the calcium concentration of the
cells in Aduring the electrical
stimulation (at arrow). Only 1
of the 17 cells (indicated by the
arrow head in Aand B) re-
sponds to the electrical stimu-
lation. D[Ca
2+
]
i
recorded from
two further cells during repeat-
ed stimulation (at arrows), in
the presence and absence of
4 mM kynurenate and 1 M
tetrodotoxin (TTX). The re-
sponse in the top cell represents
antidromic stimulation, whereas
the lower cell demonstrates or-
thodromic, synaptic activation
665

Measurement of [fura-2] in pre-
and postsynaptic elements
Calcium indicators such as fura-2 also act as calcium
buffers inside cells, possibly altering their physiological
responses. The 5-min loading time was the minimum
needed to provide enough dye in the majority of
superficial cells for an adequate signal to noise ratio
without introducing an unacceptable amount of exoge-
nous calcium buffer into the slice. To quantify fura-2
loading into the MNTB neurones, a concentration stan-
dard curve was produced by whole-cell patch recording
MNTB neurones with varying concentrations of fura-2
salt in the patch pipette. After full equilibration of the dye
with the intracellular solution (approximately 20 min)
somatic images at the calcium-insensitive (isosbestic)
wavelength were used to produce a calibration curve
(Fig. 4A). This was compared to the fluorescence levels in
the MNTB neurones recorded at the isosbestic wave-
length following loading with fura-2 AM. The average
concentration of fura-2 loaded into MNTB cells was
calculated to be 31€3 M (n=39), and the maximum
concentration observed was 87 M.
Exogenous calcium buffering can influence neuro-
transmitter release [3, 26], so it was essential that fura-2
loading of the presynaptic terminal is minimal. Although
the small amount of fura-2 loading produced a detectable
level of fluorescence in the postsynaptic cells, presynaptic
terminals did not appear to be loaded with dye. It was
possible that the presynaptic fluorescence was obscured
by the signal from the postsynaptic cell. To exclude this
possibility we examined fluorescent images following
quenching of the postsynaptic fluorescence by whole-cell
patching the MNTB neurone with 2 mM MnCl
2
added to
the intracellular solution [16]. Figure 4B and C show the
DIC and fluorescent images from a field of cells loaded
with fura-2. Whole-cell patch-clamping the central neu-
rone in the field (Fig. 4D) confirmed the presence of an
active synaptic connection upon stimulation (Fig. 4D,
inset). Following removal of the postsynaptic fluores-
cence by manganese quenching, fluorescence in the
presynaptic terminal was not observed (Fig. 4E). Even
without MnCl
2
in the internal solution, the postsynaptic
fluorescence could also be removed by dialysis of the
fura-2 dye into the patch pipette, but this occurred more
slowly (data not shown). No fluorescence from the
presynaptic terminal region could be observed under
these conditions, indicating that the [fura-2] in the
terminal is below the optical limits of the imaging
system. To confirm that fura-2 could be detected were it
present in the presynaptic terminal, whole-cell recording
was performed from a presynaptic terminal [7] with 50 M
fura-2 salt in the patch pipette (Fig. 4F; see also [11]).
When this amount of dye was present, the presynaptic
terminal was clearly visible around the postsynaptic cell
(Fig. 4G). This was repeated with 10 M fura-2 salt in the
patch pipette (Fig. 4H, I). Although significantly dimmer,
the fluorescence from the terminal is clearly visible above
the background fluorescence of the slice at a higher
camera gain (Fig. 4I). Taken together, these results
indicate that there was no detectable fura-2 present in the
presynaptic terminal when the slice was briefly loaded
with fura-2 AM.
Fura-2 loading does not alter synaptic properties
Although no fura-2 can be visually detected in the
presynaptic terminals following AM loading of the slice,
there remains the possibility that a small amount of fura-2
may enter the terminal. Wang and Kaczmarek [26]
showed that the rate of recovery of the MNTB synaptic
currents from synaptic depression was slowed by loading
a brainstem slice with EGTA-AM (200 M for 30 min).
To ensure that fura-2 loaded into the slice in our
Fig. 3A–D Electrophysiological recordings from MNTB neurones.
ACell attached extracellular recording of the postsynaptic action
potential (t) following a single electrical stimulus at the midline
(arrow). The stimulus artefact is clearly visible immediately
following the stimulation. BWhole-cell voltage-clamp recording
of the excitatory postsynaptic current (EPSC; t) following
electrical stimulation (arrow) of the same cell as in A.CWhole-
cell voltage-clamp recording of a different cell showing the
postsynaptic EPSC (t) following electrical stimulation. Also
evident in this recording is the extracellular recording of the
presynaptic action potential (asterisk). The inset shows the current
on an expanded scale. The scale bar is 500 s and 100 pA. D
Whole-cell voltage-clamp recording of a cell without an active
synaptic input demonstrating an unclamped action potential
immediately following the electrical stimulation
666

experiments did not have a similar effect on the recovery
from synaptic depression we examined the effects of fura-
2 loading on trains of synaptic responses. Postsynaptic
recordings were obtained without previous dye loading,
and trains of stimuli were elicited during and for 30 min
after dye loading. Trains of stimuli were delivered at
300 Hz for 100 ms, resulting in a depressing train of
postsynaptic EPSCs. A test pulse, 1 s after the train, was
Fig. 4A–I Presynaptic terminals do not load with fura-2 AM. A
Concentration standard curve for the fluorescence of fura-2 salt in
MNTB neurones, excited at its isosbestic wavelength (358 nm).
Data are mean €SEM of three cells at each concentration. The
dashed line indicates the average fluorescence of neurones loaded
with fura-2 AM (100€10 fluorescence units ”31€3 M fura-2). B
Differential interference contrast (DIC) image of a field of MNTB
neurones loaded with fura-2 AM. CFluorescence image (358 nm
excitation) of the same field of cells as in B.DWhole-cell
recording from the middle MNTB neurone with 2 mM MnCl
2
in the
patch pipette. This cell has an active presynaptic terminal as
evident by the EPSC upon stimulation shown in the inset (scale bar
is 1 nA and 5 ms). EFluorescence image (358 nm excitation) of the
same field of cells following manganese quenching of the
postsynaptic fluorescence. Despite there being an active terminal
around this cell, no synaptic terminal fluorescence is observed. F
DIC image of an MNTB neurone with a presynaptic terminal
(surrounding the postsynaptic cell, in between the two arrow
heads), whole-cell voltage-clamped with a pipette containing 50 M
fura-2 salt. GFluorescence image (358 nm excitation) of the same
cell as Fdemonstrating fluorescence clearly visible in the
presynaptic terminal region. HDIC image of an MNTB neurone
with a presynaptic terminal, whole-cell voltage-clamped with
10 M fura-2 salt in the internal solution. IFluorescence image
(358 nm excitation) of the same cell as Hdemonstrating
fluorescence in the presynaptic terminal region, visible above the
background autofluorescence of the slice. All images are to the
same scale. The bars on Band Care 20 m
Fig. 5A–D Fura-2 loading does not affect synaptic transmission. A
Whole-cell voltage-clamp recordings from an MNTB neurone
before and after loading of the slice with 7 M fura-2 AM for 6 min.
The synapse was stimulated at 300 Hz for 100 ms, followed 1 s
later by a test pulse to determine the degree of recovery from
synaptic depression. EPSCs are shown with the stimulus artefact
erased for clarity. BTest EPSCs from Aoverlaid for comparison.
The control trace is shown in black and the fura-2 trace in grey.C
The paired-pulse ratio (2nd EPSC/1st EPSC amplitude) and the test
EPSC ratio (test EPSC/1st EPSC amplitude) plotted as a function of
time during fura-2 AM loading of the slice and continuing for the
30-min de-esterification period. DComparison of the amplitude of
the 1st EPSC in the train, the paired-pulse ratio and the test EPSC
ratio before and after fura-2 AM loading. Data are mean €SEM of
three cells. None of these parameters were significantly altered by
fura-2 loading (P=0.28, 0.45 and 0.15 respectively; paired t-tests)
667

used to assess the degree of recovery from the synaptic
depression (Fig. 5A). Following fura-2 loading, the
magnitude and time-course of the test EPSC were
unchanged (two superimposed EPSCs in Fig. 5B). Fig-
ure 5C (open symbols) shows the ratio of the test pulse to
the amplitude of the first EPSC in the train, plotted during
and after fura-2 loading. There was clearly no effect of
fura-2 loading on the test EPSC, whereas EGTA loading
resulted in an approximately 15% decrease in the EPSC
magnitude 1 s after a 300-Hz, 100-ms train [26]. The ratio
of the second EPSC to the first EPSC (the paired-pulse
ratio) was similarly unaffected by fura-2 loading (Fig. 5C,
solid symbols). Averaged data from three cells (Fig. 5D)
showed that the magnitude of the EPSCs, paired-pulse
ratio and recovery from synaptic depression were all
unchanged by fura-2 loading.
To gauge the level of fura-2 loading needed to
significantly influence neurotransmitter release we ex-
posed to slice to a very high concentration of fura-2 AM.
Loading the slice for 30 min with 50 M fura-2 AM (a
sevenfold higher concentration for six times as long as the
usual loading conditions) resulted in substantial loading
of the postsynaptic neurones, but again no loading of fura-
2 into the presynaptic terminal was observed. To deter-
mine whether sufficient fura-2 had entered the presynap-
tic terminal to influence neurotransmitter release we
assessed the rate of recovery of the synapse from short-
term synaptic depression following trains of stimuli.
Following 100-ms trains of stimuli at 300 Hz, test pulses
were delivered with varying time delays, and plotted as a
ratio of the initial EPSC in the stimulating train (Fig. 6).
Compared to control (solid line), high fura-2 loading
(grey line) produced no significant change in the time
constants of recovery from depression. The fast time
constant was 52€9 ms (36% of total amplitude) in control
and 68€25 ms (43% of total amplitude) following high
fura-2 loading (n=3, P=0.58), and the slow time constant
was 2.6€0.6 s in control and 3.0€1.7 following high fura-
2 loading (n=3, P=0.83). This result indicates that even
this substantial fura-2 load does not adversely alter the
presynaptic calcium buffering.
Discussion
We have demonstrated that calcium imaging can be used
in a brain slice to preselect cells with functional synaptic
inputs prior to attempting patch-clamp recording. Loading
the brain slice with a calcium indicator as described above
has minimal effects on the physiology of the pre- and
postsynaptic cells for the following reasons. First, the
concentrations of calcium-sensitive dye needed to detect
activity are very low, and will be rapidly removed (5–
10 min) from the postsynaptic cell by dialysis with the
whole-cell pipette solution during recording. Second,
fura-2 loading of the presynaptic terminals was below the
optical resolution limit of the imaging system (Fig. 4E),
less than 10 M (Fig. 4I). And third, the concentration of
mobile calcium buffers in the presynaptic terminal is
estimated to be equivalent to approximately 200 M
EGTA [3], a much larger amount than would be loaded
with 7 M fura-2 AM for 5 min. Low cytoplasmic
esterase activity probably accounts for the lack of fura-2
in the presynaptic terminals. Fura-2 AM loading had no
effect on the magnitude of the postsynaptic EPSCs,
paired-pulse ratio and recovery from synaptic depression
(Fig. 5). Excessive fura-2 AM loading (a high concentra-
tion for a long time) also has no significant effect on the
rate of recovery from synaptic depression (Fig. 6), in
contrast to EGTA AM loading which eliminated the fast
component of the recovery curve [26]. This result
indicates that there is a broad safety margin between the
amount of dye needed to observe postsynaptic calcium
rises and the amount that would influence presynaptic
calcium buffering. Our results clearly demonstrate that
the calyx physiology was unchanged by our loading
protocols.
Detection of the extracellular action potentials with a
loose cell-attached electrode (Fig. 3A) has also been used
to screen for synaptic inputs [4]. However, the calcium
imaging method presented here allows a larger number of
cells to be screened in a relatively short time without
compromising the physiology. As a result, our rate of data
acquisition has increased. Peterlin et al. [17] have used a
similar “optical probing” technique to detect the postsyn-
aptic targets of individual cells in the cortex. Whereas
their method has been utilized to map-out synaptic
circuits (in the neocortex; [15]), we have used optical
probing to detect surviving synaptic connections by
grossly stimulating an entire input pathway. In the present
situation, the identity of the postsynaptic cells is known
but the nature of the slicing procedure and length of the
axon pathway reduce the connectivity from 100% to
<10%. This method will be of general application for
studies of sparse connections in other areas of the brain,
Fig. 6 Exposure to excessive fura-2 AM does not influence
recovery from synaptic depression. Following a depressing stim-
ulation train (300 Hz for 100 ms) the recovery of the EPSC was
measured as a function of time (expressed as ratio of the test EPSC
to the 1st EPSC amplitude at each time point). The control recovery
curve is shown in black. The recovery curve from cells in slices that
had been loaded with 50 M fura-2 AM for 30 min is shown in
grey. Data from three slices for each curve were fit with a double
exponential function
668

for example the climbing fibre to Purkinje cell synapse in
the cerebellum [14]. Suprathreshold postsynaptic respons-
es are an absolute requirement for detection of functional
synapses by this optical probing method. The calyx of
Held-MNTB synapse is ideal in this respect since each
presynaptic action potential produces a postsynaptic
action potential over an extensive frequency range [21,
24, 27]. The importance of this calyx of Held as a model
for studying synaptic physiology [7, 25] makes this
simple optical method a reliable means of detecting
functional synaptic connections, without adversely affect-
ing their physiological responses.
Acknowledgements We are grateful to Martine Hamann for
encouragement, suggestions and comments on the manuscript.
This work was supported by the Wellcome trust.
References
1. Barnes-Davies M, Forsythe ID (1995) Pre- and postsynaptic
glutamate receptors at a giant excitatory synapse in rat auditory
brainstem slices. J Physiol (Lond) 488:387–406
2. Bollmann JH, Sakmann B, Borst JG (2000) Calcium sensitivity
of glutamate release in a calyx-type terminal. Science 289:953–
957
3. Borst JG, Sakmann B (1996) Calcium influx and transmitter
release in a fast CNS synapse. Nature 383:431–444
4. Borst JG, Helmchen F, Sakmann B (1995) Pre- and postsyn-
aptic whole-cell recordings in the medial nucleus of the
trapezoid body of the rat. J Physiol (Lond) 489:825–840
5. Cowie RJ, Carpenter MB (1985) Transneuronal transport in the
vestibular and auditory systems of the squirrel monkey and the
arctic ground squirrel. II. Auditory system. Brain Res 358:264–
275
6. Edwards FA, Konnerth A, Sakmann B, Takahashi T (1989) A
thin slice preparation for patch clamp recordings from neurones
of the mammalian central nervous system. Pflugers Arch
414:600–612
7. Forsythe ID (1994) Direct patch recording from identified
presynaptic terminals mediating glutamatergic EPSCs in the rat
CNS, in vitro. J Physiol (Lond) 479:381–387
8. Grothe B, Park TJ (2000) Structure and function of the bat
superior olivary complex. Microsc Res Tech 51:382–402
9. Grynkiewicz G, Poenie M, Tsien RY (1985) A new generation
of Ca
2+
indicators with greatly improved fluorescence proper-
ties. J Biol Chem 260:3440–3450
10. Guinan JJ Jr. Li RY (1990) Signal processing in brainstem
auditory neurons which receive giant endings (calyces of Held)
in the medial nucleus of the trapezoid body of the cat. Hear Res
49:321–334
11. Helmchen F, Borst JG, Sakmann B (1997) Calcium dynamics
associated with a single action potential in a CNS presynaptic
terminal. Biophys J 72:1458–1471
12. Jaffe DB, Johnston D, Lasser-Ross N, Lisman JE, Miyakawa H,
Ross WN (1992) The spread of Na
+
spikes determines the
pattern of dendritic Ca
2+
entry into hippocampal neurons.
Nature 357:244–246
13. Kay AR, Alfonso A, Alford S, Cline HT, Holgado AM,
Sakmann B, Snitsarev VA, Stricker TP, Takahashi M, Wu LG
(1999) Imaging synaptic activity in intact brain and slices with
FM1–43 in C. elegans, lamprey, and rat. Neuron 24:809–817
14. Konnerth A, Llano I, Armstrong CM (1990) Synaptic currents
in cerebellar Purkinje cells. Proc Natl Acad Sci USA 87:2662–
2665
15. Kozloski J, Hamzei-Sichani F, Yuste R (2001) Stereotyped
position of local synaptic targets in neocortex. Science
293:868–872
16. Kwan CY, Putney JW Jr. (1990) Uptake and intracellular
sequestration of divalent cations in resting and methacholine-
stimulated mouse lacrimal acinar cells. Dissociation by Sr
2+
and Ba
2+
of agonist-stimulated divalent cation entry from the
refilling of the agonist-sensitive intracellular pool. J Biol Chem
265:678–684
17. Peterlin ZA, Kozloski J, Mao BQ, Tsiola A, Yuste R (2000)
Optical probing of neuronal circuits with calcium indicators.
Proc Natl Acad Sci USA 97:3619–3624
18. Sanes DH (1990) An in vitro analysis of sound localization
mechanisms in the gerbil lateral superior olive. J Neurosci
10:3494–3506
19. Schneggenburger R, Neher E (2000) Intracellular calcium
dependence of transmitter release rates at a fast central synapse.
Nature 406:889–893
20. Smith PH, Joris PX, Carney LH, Yin TC (1991) Projections of
physiologically characterized globular bushy cell axons from
the cochlear nucleus of the cat. J Comp Neurol 304:387–407
21. Smith PH, Joris PX, Yin TC (1998) Anatomy and physiology of
principal cells of the medial nucleus of the trapezoid body
(MNTB) of the cat. J Neurophysiol 79:3127–3142
22. Sommer I, Lingenhohl K, Friauf E (1993) Principal cells of the
rat medial nucleus of the trapezoid body: an intracellular in
vivo study of their physiology and morphology. Exp Brain Res
95:223–239
23. Sun JY, Wu LG (2001) Fast kinetics of exocytosis revealed by
simultaneous measurements of presynaptic capacitance and
postsynaptic currents at a central synapse. Neuron 30:171–182
24. Taschenberger H, Gersdorff H von (2000) Fine-tuning an
auditory synapse for speed and fidelity: developmental changes
in presynaptic waveform, EPSC kinetics, and synaptic plastic-
ity. J Neurosci 20:9162–9173
25. Von Gersdorff H, Borst JG (2002) Short-term plasticity at the
calyx of held. Nat Rev Neurosci 3:53–64
26. Wang LY, Kaczmarek LK (1998) High-frequency firing helps
replenish the readily releasable pool of synaptic vesicles.
Nature 394:384–388
27. Wu SH, Kelly JB (1993) Response of neurons in the lateral
superior olive and medial nucleus of the trapezoid body to
repetitive stimulation: intracellular and extracellular recordings
from mouse brain slice. Hear Res 68:189–201
669