GLIA 46:437– 445 (2004)
Properties of Ca2ⴙ-Dependent
Exocytosis in Cultured Astrocytes
MARKO KREFT,1,2 MATJAŽ STENOVEC,2 MARJAN RUPNIK,1 SONJA GRILC,1
MOJCA KRŽAN,1 MAJA POTOKAR,1,2 TINA PANGRŠIČ,1,2 PHILIP G. HAYDON,3
1,2
AND ROBERT ZOREC
*
1
Medical Faculty, University of Ljubljana, Ljubljana, Slovenia
2
Celica Biomedical Sciences Center, Ljubljana, Slovenia
3
Department of Neuroscience, University of Pennsylvania, School of Medicine,
Philadelphia, Pennsylvania
KEY WORDS
calcium; regulated exocytosis; membrane capacitance; vesicular glutamate transporter; atrial natriuretic peptide
ABSTRACT
Astrocytes, a subtype of glial cells, have numerous characteristics that
were previously considered exclusive for neurons. One of these characteristics is a
cytosolic [Ca2⫹] oscillation that controls the release of the chemical transmitter glutamate and atrial natriuretic peptide. These chemical messengers appear to be released
from astrocytes via Ca2⫹-dependent exocytosis. In the present study, patch-clamp membrane capacitance measurements were used to monitor changes in the membrane area
of a single astrocyte, while the photolysis of caged calcium compounds by a UV flash was
used to elicit steps in [Ca2⫹]i to determine the exocytotic properties of astrocytes.
Experiments show that astrocytes exhibit Ca2⫹-dependent increases in membrane capacitance, with an apparent Kd value of ⬃20 ␮M [Ca2⫹]i. The delay between the flash
delivery and the peak rate in membrane capacitance increase is in the range of tens to
hundreds of milliseconds. The pretreatment of astrocytes by the tetanus neurotoxin,
which specifically cleaves the neuronal/neuroendocrine type of SNARE protein synaptobrevin, abolished flash-induced membrane capacitance increases, suggesting that
Ca2⫹-dependent membrane capacitance changes involve tetanus neurotoxin-sensitive
SNARE-mediated vesicular exocytosis. Immunocytochemical experiments show distinct
populations of vesicles containing glutamate and atrial natriuretic peptide in astrocytes.
We conclude that the recorded Ca2⫹-dependent changes in membrane capacitance
represent regulated exocytosis from multiple types of vesicles, about 100 times slower
© 2004 Wiley-Liss, Inc.
than the exocytotic response in neurons.
INTRODUCTION
Recent advances in astrocyte physiology showing
that these glial cells may not only support neuronal
networks, but may be key in modulating synaptic
transmission by releasing chemical messengers such as
ATP and glutamate (Jeftinija et al., 1996; Haydon,
2001), raised the question as to whether the mechanism of release of these chemical messengers involves
regulated exocytosis, which requires an increase in cytosolic calcium as a trigger. Key experiments that led to
the formulation of the “calcium hypothesis” were carried out about 50 years ago (Harvey and MacIntosh,
1940; Katz and Miledi, 1965; Dodge and Rahamimoff,
1967). This hypothesis holds that Ca2⫹ binds to a com©
2004 Wiley-Liss, Inc.
ponent of the presynaptic structure to trigger the release
of neurotransmitter. The introduction of electrophysiological membrane capacitance (Cm) measurements, a pa-
Grant sponsor: Ministry of Education, Sciences and Sports of the Republic of
Slovenia; Grant number: P3-521-0381; Grant sponsor: EC; Grant number:
DECG QLG3-CT-2001-02004; Grant sponsor: FIRCA; Grant number: R03TW01293; Grant sponsor: NIH; Grant number: R37-NS37585; Grant number:
RO1-NS43142.
*Correspondence to: Robert Zorec, Laboratory of Neuroendocrinology, Molecular Cell Physiology, Institute of Pathophysiology, Medical Faculty, University
of Ljubljana, Zaloska 4, 1000 Ljubljana, Slovenia.
E-mail: [email protected]
Received 7 October 2003; Accepted 23 December 2003
DOI 10.1002/glia.20018
Published online 15 March 2004 in Wiley InterScience (www.interscience.
wiley.com).
438
KREFT ET AL.
rameter linear to the membrane area (Neher and Marty,
1982), greatly advanced our understanding of regulated
exocytosis. Initial studies on neuroendocrine cells (Neher
and Marty, 1982; Sikdar et al., 1989; Thomas et al., 1990;
Zorec et al., 1991) and on neuronal cells (Heidelberger et
al., 1994; Schneggenburger and Neher, 2000; Bollmann
et al., 2000) revealed in detail the biophysical and biochemical properties of regulated exocytosis.
It has been shown that intracellular Ca2⫹ levels regulate the release of classical neurotransmitters such as
glutamate from astrocytes (Parpura et al., 1994; Bezzi
et al., 1998; Araque et al., 2000; Pasti et al., 2001), and
that glutamate release correlates to surface area
changes (Zhang et al., 2004). Astrocytes also possess
several components of SNARE molecules (Parpura et
al., 1995; Zhang et al., 2004); key elements for vesicular
exocytosis (Geppert and Südhof, 1998) and subcellular
organelles that may be involved in regulated exocytosis
have been described (Calegari et al., 1999; Coco et al.,
2003; Kržan et al., 2003), but direct evidence for Ca2⫹dependent membrane area increase due to SNARE-dependent vesicle exocytosis in astrocytes is not available.
Using membrane capacitance measurements and
flash photolysis of caged calcium compounds (Neher
and Zucker, 1993; Rupnik et al., 2000), we show that
astrocytes exhibit Ca2⫹-dependent membrane capacitance increases that are blocked by the pretreatment of
cells by the tetanus toxin, suggesting that Ca2⫹-dependent membrane capacitance increases represent neuron-like SNARE-associated vesicular fusion. Furthermore, immunocytochemical studies revealed distinct
vesicular compartments containing glutamate and
atrial natriuretic peptides, both thought to be released
from astrocytes via regulated exocytosis (Haydon,
2001; Kržan et al., 2003). We conclude that Ca2⫹-dependent increases in membrane capacitance represent
an exocytotic response of astrocytes consisting of multiple vesicular exocytotic pathways. In comparison
with the rapid kinetics of exocytotic response in neuronal cells assayed by similar measurements (Heidelberger et al., 1994; Parsons et al., 1994; Mennerick and
Matthews, 1996; Gomis et al., 1999; Neves and Lagnado, 1999; Beutner et al., 2001; Kreft et al., 2003),
the delay to the maximal rate of membrane capacitance
increase in astrocytes suggests an ⬃2 orders of magnitude slower exocytotic response, consistent with the
view that astrocytes play a role as integrators in gliato-neuron and glia-to-endothelia signaling.
MATERIALS AND METHODS
Materials
Unless mentioned otherwise, all materials were obtained from Sigma Chemical Company (St. Louis, MO).
Cell Culture and Electrophysiology
Cultures were prepared from the cortex of 3-day-old
rats as described (Schwartz and Wilson, 1992). Cells
were grown in high-glucose Dulbecco’s modified Eagle’s
medium (DMEM), containing 10% fetal bovine serum
(FBS), 1 mM pyruvate, 2 mM glutamine, and 25 ␮g/ml
penicillin/streptomycin in 95% air/5% CO2. Confluent
cultures were shaken at 225 rpm overnight, and the
medium was changed the next morning; this process
was repeated a total of three times. After the third
shaking, the cells were trypsinized and cultured for
24 h in 10 ␮M cytosine arabinoside. After reaching
confluence again, the cells were subcultured onto 22mm-diameter poly-L-lysine-coated coverslips. Compensated Cm measurements were used (Zorec et al., 1991)
employing a SWAM IIB amplifier (Celica, Ljubljana,
Slovenia), operating at 800-Hz lock-in frequency (sine
wave of 111 mV rms). Upon establishment of the
whole-cell configuration compensation with Cslow and
Ga controls provided readings of 18.6 ⫾ 1.6 pF and of
140.2 ⫾ 12.6 nS (n ⫽ 23, mean ⫾SE), respectively,
indicating a simple electrical geometry of nonspherical
astrocytes. The phase angle setting was determined by
a 1-pF calibration pulse and by monitoring the projection of this pulse from the out-of-phase signal C (proportional to Cm) to the in-phase signal G of the lock-in
amplifier. Signals were stored unfiltered (C-DAT4 recorder, Cygnus, Fort Atkinson, WI) for off-line analysis. Simultaneously, we recorded filtered (300-Hz, 4
pole Bessel) C and G signals, and the fluorescence
intensity from a C660 photon counter (Thorn EMI,
Byfleet, UK). PhoCal program (LSR, Cambridge, UK)
was used to acquire signals every 5 ms. For high temporal resolution measurements of Cm, the records on
DAT were played back and a 10-s epoch of the signal
enveloping each flash was digitized at 50 kHz, using a
CDR program (J. Dempster, Glasgow, UK). Signals
were digitally filtered at 1 kHz (2-way 150th order FIR
filter, Math Works MATLAB, Natick, MA) and resampled at 10 kHz. The pipette solution contained (in
mM): KCl 110, TEACl 10, KOH/HEPES (N-2-hydroxyethylpiperazine-N⬘-2-ethanesulfonic acid) 40, Na2ATP
2, MgCl2 2, K4 -NP-EGTA [o-nitrophenyl ethyleneglycol-bis(␤-aminoethyl ether)-N,N,N⬘,N⬘-tetrapotassium
salt] 4, CaCl2 3.6, furaptra 0.5, pH ⫽ 7.2. The bath
contained (in mM): NaCl 131.8, KCl 5, MgCl2 2,
NaH2PO4 0.5, NaHCO3 5, NaOH/HEPES 10, D-glucose
10, CaCl2 1.8, pH ⫽ 7.2. All recordings were made at
room temperature. Cells were voltage-clamped at a
holding potential of ⫺50 mV. Recordings were made
with pipette resistances between 1 and 4 M⍀ (measured in KCl-rich solution). The pipette and bath solutions were of similar osmolarity (within 5%) measured
by freezing point depression (Camlab, Cambridge, UK).
Toxin Microinjection
Transjector 4657 with a micromanipulator (Eppendorf, Hamburg, Germany) was used. Pulses (8 –12 hPa,
0.3–1s) were applied with a compensation pressure (1
hPa). Microinjection solution consisted of (in mM): Kgluconate 150, MgCl2 2, HEPES 10, RITC 6 mg/ml,
439
CA2⫹-DEPENDENT EXOCYTOSIS IN ASTROCYTES
tetanus toxin (20 ng/␮l, a gift from Dr. Das Gupta,
Madison, WI). Pipettes were prepared with a horizontal puller (P-87, Sutter Instruments, Novato, CA). Microinjected cells were identified by co-injected rhodamine-labeled dextran (25 kDa). Before continuing with
patch-clamp experiments, the cells were allowed to rest
for 1–2 h at 37°C.
Calcium Measurements and Flash Photolysis
The Ca2⫹ cage NP-EGTA (Molecular Probes, Eugene,
OR) was used to elevate [Ca2⫹]i by flash photolysis
(Ellis-Davies and Kaplan, 1994). A 1-ms ultraviolet
(UV) flash from a Xenon arc flash lamp (Hi-Tech, Salisbury, UK) was delivered to cells through a 40⫻ fluor oil
immersion objective (NA ⫽ 1.3) of a Nikon Diaphot
microscope (Rupnik et al., 2000). The same optical
pathway as in flash photolysis was used to illuminate
the [Ca2⫹] indicator furaptra (Molecular Probes, Eugene, OR). A combination of two dichroic mirrors was
used. The first was a 390-nm dichroic positioned at 45°,
which passed through the 420-nm light from a Xe arc
lamp for furaptra excitation and reflected the light
below 390 nm for NP-EGTA flash photolysis from a Xe
arc flash lamp. The second mirror, a 430-nm dichroic,
reflected both lights through the objective to the cell
studied and allowed only furaptra fluorescent light to
pass back to the photomultiplier through a 510-nm
barrier filter. [Ca2⫹]i measurements were performed as
described (Rupnik et al., 2000), using the equation:
关Ca 2 ⫹ 兴 i ⫽ K d
F ⫺ F min
F max ⫺ F
(1)
where Fmax is the autofluorescence in the cell-attached
configuration, Fmin is fluorescence in a resting wholecell recording, and F is fluorescence during the flash.
Kd for furaptra is 25 ␮M.
Lipofection, Immunocytochemistry, and
Confocal Microscopy
The plasmid encoding atrial natriuretic factor,
tagged with emerald green fluorescent protein, ANP.emd (Han et al., 1999), was introduced into the astrocytes by lipofection, using the Invitrogen (Carlsbad,
CA) protocol. DNA was mixed with 6 ␮l Plus Reagent,
diluted in 100 ␮l serum-free DMEM, and was incubated for 15 min at room temperature (RT). In this
study, 4 ␮l lipofectamine™ was diluted in 100 ␮l serum-free DMEM. After incubation, both solutions were
mixed and incubated further for 15 min at RT. Meanwhile, astrocytes were washed once with serum-free
DMEM and supplemented with 800 ␮l DMEM; 200 ␮l
of lipofection mixture was pipetted onto the cells, which
were further incubated for 3 h at 95% air/5% CO2 at
37°C. Then 30 ␮l of Ultroser G (Gibco, Grand Island,
NY) was added, and DMEM was exchanged the next
day. Astrocytes expressing ANP.emd were washed once
with 1 ml of phosphate-buffered saline (PBS) and fixed
in paraformaldehyde (4% in PBS) for 15 min. Cells
were permeabilized with 4% paraformaldehyde, 0.1%
Triton X-100 for 10 min and washed 4⫻ with PBS.
Subsequently, a rabbit serum containing either antiVGLUT1 (1:1,000) or anti-VGLUT2 (1:1,000) antibody
was applied to the astrocytes, which were further incubated overnight at 4°C. The cells were then washed
4⫻ with PBS and exposed to the respective secondary
antibody solution, an Alexa Fluor 546-conjugated antirabbit IgG (1:500). Cells were washed 4⫻ with PBS and
preparation finally supplied with Slow Fade Light Antifade Kit (Molecular Probes, Leiden, Netherlands).
Coverslips with double-labeled astrocytes were examined with a Zeiss confocal microscope (LSM 510, Jena,
Germany). Fluorescent images were acquired by a
plan-apochromatic oil immersion objective (63⫻ magnification, 1.4 NA) using 488-nm Ar-Ion and 543-nm
He-Ne laser excitation. ANP.emd fluorescence and fluorescence of labeled VGLUT1 and VGLUT2 were separated using BP 505–530-nm and LP 560-nm emission
filters, respectively. Images were stored on IBM-PC
compatible computer (Siemens Nixdorf, Germany). To
quantify images, these were exported into TIFF files
and analyzed with a custom-made MATLAB program
(Kreft et al., 2004). Briefly, the program counted all
red, all green, and all colocalized pixels in the image.
The degree of colocalization between green ANP.emd
and fluorescently labeled antibodies against ANP, VGLUT1 and V-GLUT2 in red, was expressed as percentage of colocalized pixels in comparison to all green
pixels. Colocalization data were statistically analyzed
using unpaired, two-tailed Student’s t-test.
RESULTS
Exocytosis involves the addition of membrane to the
plasma membrane, while endocytosis involves subtraction of membrane; thus, both processes contribute to
membrane area fluctuations (Neher and Marty, 1982).
If regulated Ca2⫹-triggered exocytosis exists in astrocytes, an increase in [Ca2⫹]i should elicit an increase in
Cm, a parameter that is linearly proportional to the
membrane area (Neher and Marty, 1982). We monitored secretory activity using compensated Cm measurements (Rupnik et al., 2000) combined with flash
photolysis to deliver rapid and spatially homogeneous
steps in cytosolic [Ca2⫹]i (Neher and Zucker, 1993). The
calcium cage NP-EGTA, preloaded with Ca2⫹, was introduced into the cytosol of single astrocytes using
patch pipette dialysis. Single astrocytes, devoid of contacts with neighboring cells, were used to avoid the
potential problems of electric non-homogeneity due to
gap junction coupling between cells. UV flash application (arrows) transiently elevated [Ca2⫹]i to a peak
value, which subsequently returned exponentially to
the baseline with a time constant of 2–5 s (Fig. 1, top
440
KREFT ET AL.
Fig. 1. Two types of Ca2⫹-induced membrane capacitance (Cm) increase in astrocytes. Time-dependent changes in Cm (bottom), elicited
by flash photolysis (arrow) of Ca2⫹ caged compound NP-EGTA to
increase cytosolic [Ca2⫹] (top) in a cultured astrocyte. A: Following the
UV flash induced peak in Cm, the amplitude persisted at an elevated
level. Resting capacitance of this cell was 29 pF. B: In cells exposed to
higher cytosolic [Ca2⫹], the peak in Cm was followed by a delayed
decrease in Cm, dominated by endocytosis. The extent of endocytosis
was determined by measuring the decrease in Cm 1 min after the flash
(endo). The resting capacitance of this cell was 17 pF.
traces). Figure 1 (bottom traces) shows two types of
time-dependent changes in Cm in astrocytes. First, increased [Ca2⫹]i elicited a large rise in Cm, which persisted at elevated level after the [Ca2⫹]i returned to
basal level (Fig. 1A, bottom trace). Second, the peak in
Cm was followed by a delayed decrease in Cm, caused by
endocytotic retrieval of membrane (Fig. 1B). Endocytosis was evaluated by measuring the decrease in Cm 1
min after the flash. With the peak [Ca2⫹]i between 10
and 20 ␮M, endocytosis was detected in 5 of 15 responses with an average decrease of 0.2 ⫾ 0.1 pF (mean
⫾standard error, N ⫽ 5). With a peak post-flash [Ca2⫹]i
of 20 – 60 ␮M, endocytosis was present in 12 of 14
responses with an average decrease in Cm of 2.8 ⫾ 0.9
pF (mean ⫾standard error, N ⫽ 12). From this one can
estimate the average rate of endocytosis to be around
50 fF/s. These experiments also indicate that the occurrence of endocytosis is dependent on the level of
stimulation (Artalejo et al., 1995; Engisch and Nowycky, 1998). It is unlikely that these responses are due
to an artifact, since when the NP-EGTA in the pipette
lacked Ca2⫹, a UV flash failed to elevate [Ca2⫹]i and no
change in Cm was observed. Moreover, repeated flash
applications resulted in repeated increases in Cm (not
shown), suggesting that flash applications did not result in phototoxicity.
At a peak [Ca2⫹]i of 15 ␮M, the amplitude of the
flash-induced Cm response was 700 fF (Fig. 1A),
whereas it increased to 1,750 fF at 28 ␮M [Ca2⫹]i (Fig.
1B), indicating a Ca2⫹-dependent addition of membrane due to exocytosis. At higher time resolution,
three responses in Cm are shown in Figure 2A, each
recorded at a different [Ca2⫹]i obtained by photolysis
with different UV flash intensities. Peak [Ca2⫹]i (in
␮M) is indicated near each of the traces. By increasing
Fig. 2. Calcium sensitivity of responses in membrane capacitance
(Cm) in astrocytes. A: Typical Cm responses in three single astrocytes
elicited by flash-induced (arrow) elevations in [Ca2⫹]i (numbers adjacent to traces indicate measured peak values of [Ca2⫹]i). B: Time
derivative of responses corresponding to traces in A, indicated by the
arrow. Dashed vertical lines denote the delays between the flash
delivery and the maximal rate of Cm increase. C: Ca2⫹-dependence of
the maximal rate of capacitance increase. The curve was obtained by
fitting the2⫹ data to the function with the form: (fF/s) ⫽ (3008 ⫾
792)*([Ca ](5.3 ⫾ 2.3))/((21.9 ⫾ 3.8 ␮M)(5.3 ⫾ 2.3) ⫹ [Ca2⫹](5.3 ⫾ 2.3)). Each
filled circle represents the average ⫾standard error of 3–5 measurements. D: Ca2⫹ dependence of the delay to the maximal rate of
capacitance increase. The curve was obtained by fitting the data with
the exponential regression algorithm: delay(s) ⫽ (0.55 ⫾
0.15 s)*(exp(⫺(0.057 ⫾ 0.015 ␮M⫺1)*[Ca2⫹])), where the parameters
are in the format mean ⫾standard error.
the [Ca2⫹]i beyond 10 ␮M we were able to reliably elicit
a detectable increase in Cm (Figs. 1 and 2A). Not only
the amplitude of responses, but also the rate of Cm
increase, a measure of the rate of exocytosis, depended
in a sigmoidal fashion on [Ca2⫹]i (Fig. 2C). The fitting
of a curve (Fig. 2C) suggests that ⬃20 ␮M [Ca2⫹]i is
required for the half-maximal response in Cm.
The delay between UV flash delivery and the peak
rate of Cm increase was reduced as [Ca2⫹]i increased
(Fig. 2B,D), consistent with the Ca2⫹ requirement for
the triggering of exocytosis in astrocytes. It is unlikely
that the very slow, stimulation-dependent endocytosis
CA2⫹-DEPENDENT EXOCYTOSIS IN ASTROCYTES
(Fig. 1B; ⬃50 fF/s) greatly affects the initial maximal
rate of Cm increase of 3,000 fF/s. Thus, the flash-induced increase in Cm is determined mainly by Ca2⫹regulated exocytosis.
To establish further that Ca2⫹-induced changes in
Cm are due to regulated vesicular exocytosis, we tested
whether Ca2⫹-induced changes in Cm are sensitive to
neurotoxins that specifically cleave the SNARE proteins, essential elements of neuronal/neuroendocrine
vesicle exocytosis (Montecucco and Schiavo, 1994), but
not those associated with the endosomal and lysosomal
subcellular compartments (Martinez-Arca et al., 2000).
Figure 3A shows a time-dependent change in [Ca2⫹]i
and in Cm, both in a control and in a neurotoxin-treated
cell. The maximal rate of Cm increase was strongly
inhibited by tetanus toxin pretreatment (Fig. 3B), supporting the view that Ca2⫹-dependent changes in Cm
consist mainly of tetanus toxin-sensitive vesicular exocytosis. The small residual tetanus toxin-insensitive
Ca2⫹-dependent changes in Cm (Fig. 3B) suggest that
tetanus toxin-insensitive vesicle exocytosis (MartinezArca et al., 2000) appears to be very small or absent in
cultured astrocytes.
The release of both glutamate (Parpura et al., 1994)
and atrial natriuretic peptide (ANP; Kržan et al., 2003)
requires Ca2⫹. While it is generally believed that small
molecular weight neurotransmitters and peptides segregate into distinct vesicles, some experiments suggest
that glutamate and peptide may coexist in a single
vesicle (Maechler and Wollheim, 1999). We used an
immunocytochemical approach to study the subcellular
distribution of ANP and markers for glutamate-containing vesicles, such as the vesicle glutamate transporters VGLUT1 and VGLUT2 (Bellocchio et al., 2000;
Takamori et al., 2000). Figure 4 shows a punctuate
appearance for the fluorescence signals of both markers, consistent with the notion that both markers are
localized to storage organelles such as vesicles. Furthermore, there appears to be no overlap between the
fluorescence signals, indicating the coexistence of distinct subcellular storage organelles for peptides and for
glutamate in astrocytes. To quantify further whether
the two markers are colocalized, we measured the level
of colocalization between ANP.emd fluorescence and
the fluorescence from the anti-ANP, anti-VGLUT1 and
anti-VGLUT2 antibodies, respectively. In controls, the
colocalization between ANP.emd and immunolabeled
anti-ANP antibody was 81 ⫾ 10% (mean ⫾SE), significantly higher with respect to the colocalization between ANP.emd and either anti-VGLUT1 (13 ⫾ 2%) or
anti-VGLUT2 (13 ⫾ 3%) antibodies, respectively (Fig.
5), indicating that ANP and vesicular glutamate transporters reside in distinct subcellular organelles in cultured astrocytes.
These data indicate that distinct types of vesicles
may contribute to the recorded flash-induced responses
in Cm, which is consistent with the presence of multiple
exocytotic pathways in astrocytes, as has been found in
a number of other cell types (Kasai, 1999; Rupnik et al.,
2000; Voets et al., 2001; Poberaj et al., 2002).
441
Fig. 3. Calcium-dependent Cm increase due to exocytosis is blocked
by neurotoxins. A: Typical time-dependent changes in [Ca2⫹]i and
membrane capacitance (Cm) in a control (top) and a tetanus toxintreated cell (bottom). B: Average maximal rate of Cm increase (top) as
elicited by an average peak [Ca2⫹] elevation (bottom) in control and
tetanus toxin-treated cells (⫹Tetx). Numbers adjacent to columns
indicate numbers of cells tested, asterisk indicates the level of significance (P ⫽ 0.03).
DISCUSSION
Although a large number of substances are reported
to be secreted from astrocytes (Martin, 1992; Parpura
et al., 1994), the underlying mechanisms are still debated. In the case of the transmitter glutamate secretion, for example, reversal of glutamate transporters,
anion transporter-dependent mechanisms, channel
permeation, as well as Ca2⫹- dependent exocytosis
have all been proposed (Attwell, 1994; Haydon, 2001;
Duan et al., 2003). In support of the latter hypothesis,
442
KREFT ET AL.
Fig. 4. Distinct subcellular compartments contain peptide and
chemical transmitter glutamate in cultured astrocytes. A: Confocal
image of an astrocyte transfected with a construct to express a fusion
protein between the atrial natriuretic factor and the emerald green
fluorescent protein (ANP.emd, green) (Han et al., 1999) and immunocytochemical detection of the vesicular glutamate transporter,
VGLUT1, using a specific VGLUT1 antibody and a fluorescent secondary antibody (red). B: Confocal image of an astrocyte transfected
with an ANP.emd (green) construct (Han et al., 1999) and immunocytochemically labeled for vesicular glutamate transporter, VGLUT2,
using a specific VGLUT2 antibody and a fluorescent secondary antibody (red). Scale bars ⫽ 10 ␮m.
it was shown recently that membrane surface area
changes and optically detected glutamate release are
associated, supporting the notion that glutamate is
released by exocytosis (Zhang et al., ).
Fig. 5. Quantitative colocalization analysis of fluorescent signals
from atrial natriuretic factor, tagged with emerald green fluorescent
protein, ANP.emd, containing vesicles and fluorescently labeled antiVGLUT1 and anti-VGLUT2 antibodies. A: Confocal image of an astrocyte showing high colocalization (yellow) between ANP.emd fluorescence (green) and fluorescence signals from an anti-ANP antibody
labeled by fluorescent secondary antibody (red). B: Confocal image of
an astrocyte showing distinct subcellular localization of ANP.emd
containing vesicles and compartments labeled by anti-VGLUT1 antibody and fluorescent secondary antibody (red). C: Level of colocalization between ANP.emd fluorescence and fluorescence from anti-ANP,
anti-VGLUT1 and anti-VGLUT2 antibodies, respectively. The colocalization between ANP.emd and immunolabeled anti-ANP antibody
was 81 ⫾ 10% (mean ⫾SE), significantly higher with respect to the
colocalization between ANP.emd and either anti-VGLUT1 antibody
(13 ⫾ 2%) or anti-VGLUT2 antibody (13 ⫾ 3%), respectively. Asterisks indicate statistically significant difference (P ⬍ 0.001) in colocalization amount between ANP.emd and either anti-VGLUT1 or antiVGLUT2 with respect to the colocalization between ANP.emd and
anti-ANP, n indicates the number of images analyzed. Scale bar ⫽ 5
␮m in A.
The aim of this study was to determine the properties of Ca2⫹-dependent exocytosis in astrocytes by using electrophysiological measurements of membrane
capacitance. Exocytosis involves addition of membrane
to the plasma membrane, while endocytosis involves
subtraction of membrane; thus, both processes actively
contribute to membrane area fluctuations (Neher and
Marty, 1982). If regulated Ca2⫹-triggered exocytosis
exists in astrocytes, an increase in [Ca2⫹]i should elicit
an increase in Cm, a parameter linearly proportional to
the membrane area (Neher and Marty, 1982). Consistent with the key involvement of Ca2⫹ in regulated
443
CA2⫹-DEPENDENT EXOCYTOSIS IN ASTROCYTES
exocytosis pathway flash photolysis of NP-EGTA, a
calcium cage compound, elicited an increase in [Ca2⫹]i
and a consequent rapid rise in Cm. The half-maximal
response was obtained at ⬃20 ␮M [Ca2⫹]i (Fig. 2C).
This concentration may be exceeded under physiological conditions (i.e., close to the mouth of open IP3
receptors that mediate calcium flux from the internal
stores to the cytosol). The relatively high Hill’s coefficient of 5.3 (see the equation in Fig. 2C) suggests that
at least five calcium ions must bind to a calcium sensor
to activate vesicle fusion or alternatively that there are
distinct mechanisms for calcium sensing by vesicles
that respond to calcium. Therefore, cultured astrocytes
appear to posses the machinery for regulated exocytosis.
It was shown previously that glutamate is released
from astrocytes at submicromolar cytosolic [Ca2⫹] (Parpura and Haydon, 2000). There are three possible reasons for the apparent discrepancy between the apparent Kd of 20 ␮M reported in this study and the previous
work. First, the previous study only demonstrated that
submicromolar elevations of [Ca2⫹] were sufficient to
evoke transmitter release. Because a full dose-response
relationship between [Ca2⫹] and glutamate release was
not determined, the Kd for [Ca2⫹] in the previous study
remains unknown. For example, it is possible that the
previous study only operated at the foot of the sigmoid
relationship between [Ca2⫹] and release, which would
explain why low [Ca2⫹] could stimulate release and
why a Hill coefficient of only 2 was measured, compared with 5 in the current study. Second, regulated
exocytosis may be of different modes. The kiss-and-run
fusion may result in a relatively small net increase in
membrane capacitance increase in comparison to complete fusion where the vesicle membrane is incorporated into the plasma membrane. In the kiss-and-run
mode of exocytosis, small-molecular-weight molecules
(e.g., glutamate) will escape the vesicle lumen during
relatively short (ms) transient openings of the fusion
pore. If glutamate release were mediated predominantly by kiss-and-run fusion, this contribution to the
whole-cell capacitance response measured in this case
would be relatively small compared with the total
membrane insertion measured. A third possibility is
that because capacitance measurements do not discriminate between different vesicle pools with different
Kd values the abundance of glutamate containing vesicles might be small in comparison with all vesicles
that contribute to the regulated exocytosis in astrocytes.
Other secretory vesicles, such as lysosomes, may contribute to changes in Cm as well. Secretory lysosomes
exocytose in a Ca2⫹-dependent manner (Martinez et
al., 2000), and it is possible that the Ca2⫹-dependent
changes in membrane capacitance records in astrocytes and fibroblasts (Ninomiya et al., 1996; Coorssen
et al., 1996) represent, at least in part, a contribution of
Ca2⫹-dependent lysosomal exocytosis required for
membrane wound repair (Reddy et al., 2001). Consistent with this, it was found that synaptotagmin VII, a
2⫹
protein required for Ca -dependent lysosome exocytosis (Martinez et al., 2000) was reported to be expressed
in astrocytes (Zhang et al. 2003). However, lysosomal
membranes contain VAMP 7 (Advani et al., 1999), a
SNARE protein that is insensitive to tetanus toxin
(Martinez-Arca et al., 2000). Therefore, it is likely that
fusion of lysosome membrane with other membranes is
insensitive to this neurotoxin. The small residual, tetanus toxin-insensitive flash induced Cm increase (Fig.
3B) suggests that, in astrocytes, the tetanus toxininsensitive vesicle exocytosis is small in comparison
with the tetanus toxin-sensitive vesicle exocytosis. This
component also shares the Ca2⫹-sensitivity of regulated exocytosis in various neuronal preparations (Heidelberger et al., 1994; Schneggenburger and Neher,
2000; Bollmann et al., 2000; Kreft et al., 2003). Therefore, the recorded Ca2⫹-dependent increases in Cm represent mainly the fusion of vesicles characteristic of
neurons and neuroendocrine cells. The nature and the
size of these vesicles are unknown and remain to be
investigated in the future.
Although stimulation-dependent endocytosis is
present in astrocytes (Fig. 1B), it is unlikely to have a
significant influence on the flash-induced rate of exocytosis. The average rate of decrease in Cm of 50 fF/s
(see Fig. 1B) indicates that the rate of endocytosis in
these cells is rather slow in comparison to the maximal
rate of Cm increase of 3,000 fF/s (Fig. 2C). To convert
the rate of Cm increase (Fig. 2B) into rate constants, we
divided the maximal rates of Cm increase of 3,000 fF/s
by 1,000 fF, the average amplitude of the Cm response
when [Ca2⫹]i was 20␮M. At high [Ca2⫹]i, the rate constant reached a limiting value of 3/s (time-constant of
333 ms). In contrast, the measurements of the delay
between the flash delivery and the maximal rate of Cm
increase revealed a limiting minimal delay of 50 –100
ms (Fig. 2C, D). Both measurements show that, in
cultured astrocytes, the kinetics of the Ca2⫹-regulated
exocytotic apparatus is similar to the exocytotic response in neuroendocrine cells (Rupnik et al., 2000;
chromaffin cell), but that it is ⬃2 orders of magnitude
slower in comparison with the rapid kinetic component
of exocytosis in neuronal cells studied by similar techniques (Heidelberger et al., 1994; Parsons et al., 1994;
Mennerick and Matthews, 1996; Gomis et al., 1999;
Neves and Lagnado, 1999; Beutner et al., 2001, Kreft et
al., 2003), and the delay in synaptic transmission (Sabatini and Regehr, 1999; Taschenberger and von Gersdorff, 2000). The relatively slow rate of exocytosis in
astrocytes is probably related to structural differences
between astrocytes and neurons. Rapid initial rates of
exocytosis can be achieved in neurons because many
vesicles are docked with the presynaptic plasma membrane in preparation for vesicle fusion and the release
of neurotransmitter. Additionally, it was recently suggested that adjacent vesicles can fuse with one another
through compound exocytosis to facilitate high rates of
release (Parsons and Sterling, 2003). Because astrocytes do not have vesicles arranged in distinct clusters
that are pre-docked with the plasma membrane, it is
444
KREFT ET AL.
perhaps not surprising that exocytosis is so slow compared with neurons. This property is consistent with
the integrating function of astrocytes in neuron-to-glia
and glia-to-endothelia signaling (Haydon, 2001; Kržan
et al., 2003).
Glial cells exhibit a special form of excitability consisting of [Ca2⫹]i oscillation, which may propagate between neighboring cells (Verkhratsky et al., 2002). Experimental evidence that neurotoxin sensitive Ca2⫹dependent exocytosis is present in astrocytes, and that
this process has a similar Ca2⫹-sensitivity to that
found in neurons, suggests a new perspective on the
role of astrocytes in the central nervous system. As in
neurons, exocytosis plays a role in astrocytes in the
release of chemical messengers such as ATP and glutamate, which lead to a modulation of synaptic transmission (Haydon, 2001). Moreover, exocytosis is involved in the release of peptides (Kržan et al., 2003),
playing a role in the regulation of blood perfusion of the
brain tissue. The kinetics of glial Ca2⫹-regulated exocytosis is slow in comparison with synaptic transmission (Sabatini and Regehr, 1999) (Figs. 1 and 2) and
provides Ca2⫹-excitable astrocytes with the cellular
machinery necessary for the integration and coordination of neurons as well as endothelia during CNS function.
ACKNOWLEDGMENTS
The authors thank Dr. E. Levitan for the ANP.emd
construct, and Dr. R. Jahn and Dr. S. Takamori for the
donation of VGLUT1 and VGLUT2 constructs and antibodies. This work was supported by grant P3-5210381 from the Ministry of Education, Sciences and
Sports of the Republic of Slovenia; by European Union
(EU) contract DECG QLG3-CT-2001-02004; by Fogarty
International Research Collaboration Award (FIRCA)
grant R03-TW01293 (to R.Z.); and by grants R37
NS37585, RO1 NS43142 from the National Institutes
of Health (to P.G.H.).
REFERENCES
Advani RJ, Yang B, Prekeris R, Lee KC, Klumperman J, Scheller RH.
1999. VAMP-7 mediates vesicular transport from endosomes to
lysosomes. J Cell Biol 146:765–775.
Araque A, Li N, Doyle RT, Haydon, PG. 2000. SNARE protein-dependent glutamate release from astrocytes. J Neurosci 20:666 – 673.
Artalejo CR, Henley JR, McNiven MA, Palfrey HC. 1995. Rapid endocytosis coupled to exocytosis in adrenal chromaffin cells involves
Ca2⫹, GTP, and dynamin but not clathrin. Proc Natl Acad Sci USA
92:8328 – 8332.
Attwell D. 1994. Glia and neurons in dialogue. Nature 369:707–708.
Bellocchio EF, Reimer RJ, Fremeau RT, Edwards JR. 2000. Uptake of
glutamate into synaptic vesicles by an inorganic phosphate transporter. Science 289:957–960.
Beutner D, Voets T, Neher E, Moser T. 2001. Calcium dependence of
exocytosis and endocytosis at the cochlear inner hair cell afferent
synapse. Neuron 29:681– 690.
Bezzi P, Carmignoto G, Past L, Vesce S, Rossi D, Lodi Rizzini B,
Pozzan T, Volterra A. 1998. Prostaglandins stimulate calcium-dependent glutamate release in astrocytes. Nature 391:281–285.
Bollmann JH, Sakmann B, Borst JG. 2000. Calcium sensitivity of
glutamate release in a calyx-type terminal. Science 289:953–957.
Calegari F, Coco S, Taverna E, Bassetti, Verderio C, Corradi N,
Matteoli M, Rosa P. 1999. A regulated secretory pathway in cultured hippocampal astrocytes. J Biol Chem 274:22539 –22547.
Coco S, Calegari F, Pravettoni E, Pozzi D, Taverna E, Rosa P, Matteoli
M, Verderio C. 2003. Storage and release of ATP from astrocytes in
culture. J Biol Chem 278:1354 –1362.
Coorsen J R, Schmitt H, Almers W. 1996. Ca2⫹ triggers massive
exocytosis in Chinese hamster ovary cells. EMBO J 15:3787–3791.
Dodge FA, Rahamimoff R. 1967. On the relationship between calcium
concentration and the amplitude of the end-plate potential.
J Physiol 193:419 – 432.
Duan S, Anderson CM, Keung EC, Chen Y, Chen Y, Swanson, RA.
2003. P2X7 receptor-mediated release of excitatory amino acids
from astrocytes. J Neurosci 23:1320 –1328.
Ellis-Davies GC, Kaplan JH. 1994. Nitrophenyl-EGTA, a photolabile
chelator that selectively binds Ca2⫹ with high affinity and releases
it rapidly upon photolysis. Proc Natl Acad Sci USA 91:187–191.
Engisch KL, Nowycky MC. 1998. Compensatory and excess retrieval:
two types of endocytosis following single step depolarisations in
bovine adrenal chromaffin cells. J Physiol 506.3:591– 608.
Geppert M, Südhof TC. 1998. RAB3 and synaptotagmin: the yin and
yang of synaptic membrane fusion. Annu Rev Neurosci 21:75–95.
Gomis A, Burrone J, Lagnado L. 1999. Two actions of calcium regulate
the supply of releasable vesicles at the ribbon synapse of retinal
bipolar cells. J Neurosci 19:6309 – 6317.
Han W, Ng YK, Axelrod D, Levitan, ES. 1999. Neuropeptide release
by efficient recruitment of diffusing cytoplasmic secretory vesicles.
Proc Natl Acad Sci USA 96:14577–14582.
Harvey AM, MacIntosh FC. 1940. Calcium and synaptic transmission
in a sympathetic ganglion. J Physiol 143:114 –137.
Haydon PG. 2001. GLIA: listening and talking to the synapse. Nat
Rev Neurosci 2:185–193.
Heidelberger R, Heinemann C, Neher E, Matthews G. 1994. Calcium
dependence of the rate of exocytosis in a synaptic terminal. Nature
371:513–515.
Jeftinija SD, Jeftinija KV, Stefanovic G, Liu F. 1996. Neuroligandevoked calcium-dependent release of excitatory amino acids from
cultured astrocytes. J Neurochem 66:676 – 684.
Kasai H. 1999. Comparative biology of Ca2⫹-dependent exocytosis:
implications of kinetic diversity for secretory function. Trends Neurosci 22:88 –93.
Katz B, Miledi R. 1965. The effect of calcium on acetylcholine release
from motor nerve terminals. Proc R Soc Biol 161:496 –503.
Kreft M, Križaj D, Grilc S, Zorec R. 2003. Properties of exocytotic
response in vertebrate photoreceptors. J Neurophysiol 90:218 –225.
Kreft M, Milisav I, Potokar M, Zorec R. 2004. Automated high
through-put colocalization analysis of multichannel confocal images. Comput Methods Prog Biol 74:63– 67.
Kržan M, Stenovec M, Kreft M, Pangršič T, Grilc S, Haydon PG, Zorec
R. 2003. Calcium-dependent exocytosis of atrial natriuretic peptide
from astrocytes. J Neurosci 23:1580 –1583.
Maechler P, Wollheim CB. 1999. Mitochondrial glutamate acts as a
messenger in glucose-induced insulin exocytosis. Nature 402:685–
689.
Martin DL. 1992. Synthesis and release of neuroactive substances by
glia. Glia 5:81–94.
Martinez I, Chakrabarti S, Hellevik T, Morehead J, Fowler K, Andrews NW. 2000. Synaptotagmin VII regulates Ca2⫹-dependent
exocytosis of lysosomes in fibroblasts. J Cell Biol 137:1– 4.
Martinez-Arca S, Alberts P, Galli T. 2000. Clostridial neurotoxininsensitive vesicular SNAREs in exocytosis and endocytosis. Biol
Cell 92:449 – 453.
Mennerick S, Matthews G. 1996. Ultrafast exocytosis elicited by calcium current in synaptic terminals of retinal bipolar neurons. Neuron 20:1241–1249.
Montecucco C, Schiavo G. 1994. Mechanism of action of tetanus and
botulinum neurotoxins. Mol Microbiol 13:1– 8.
Neher E, Marty A. 1982. Discrete changes of cell membrane capacitance observed under conditions of enhanced secretion in bovine
adrenal chromaffin cells. Proc Natl Acad Sci USA 79:6712– 6716.
Neher E, Zucker RS. 1993. Multiple calcium-dependent processes
related to secretion in bovine chromaffin cells. Neuron 10:21–30.
Neves G, Lagnado L. 1999. The kinetics of exocytosis and endocytosis
in the synaptic terminal of goldfish retinal bipolar cells. J Physiol
515:181–202.
Ninomiya Y, Kishimoto T, Miyashita Y, Kasai H. 1996. Ca2⫹-dependent exocytotic pathways in Chinese hamster ovary fibroblasts revealed by caged calcium compound. Kinetic diversity in the fusion of
exocytic vesicles. J Biol Chem 271:17751–17754.
CA2⫹-DEPENDENT EXOCYTOSIS IN ASTROCYTES
Parpura V, Haydon PG. 2000. Physiological astrocytic calcium levels
stimulate glutamate release to modulate adjacent neurons. Proc
Natl Acad Sci USA 97:8629 – 8634.
Parpura V, Basarsky T, Liu N, Jeftinija S, Haydon PG. 1994. Glutamate-mediated astrocyte-neuronal signalling. Nature 369:774 –777.
Parpura V, Fang H, Basarsky T, Jahn R, Haydon PG. 1995. Expression of synaptobrevin. II. Cellubrevin and syntaxin but not
SNAP-25 in cultured astrocytes. FEBS Lett 377:489 – 492.
Parsons TD, Sterling P. 2003. Synaptic ribbon. Conveyor belt or safety
belt? Neuron 37:379 –382.
Pasti L, Zonta M, Pozzan T, Vicini S, Carmignoto G. 2001. Cytosolic
calcium oscillations in astrocytes may regulate exocytotic release of
glutamate. J Neurosci 21:477– 484.
Parsons TD, Lenzi D, Almers W, Roberts WM. 1994. Calcium-triggered
exocytosis and endocytosis in an isolated presynaptic cell: capacitance
measurements in saccular hair cells. Neuron 13:875–883.
Poberaj I, Rupnik M, Kreft M, Sikdar SK, Zorec R. 2002. Modeling
excess retrieval in rat melanotroph membrane capacitance records.
Biophys J 82:226 –232.
Reddy A, Caler EV, Andrews NW. 2001. Plasma membrane repair is mediated by Ca2⫹-regulated exocytosis of lysosomes. Cell 106:157–169.
Rupnik M, Kreft M, Sikdar SK, Grilc S, Romih R, Zupančič G, Martin
TF, Zorec R. 2000. Rapid regulated dense-core vesicle exocytosis
requires the CAPS protein. Proc Natl Acad Sci USA 97:5627–5632.
Sabatini BL, Regehr WG. 1999. Timing of synaptic transmission.
Annu Rev Physiol. 61:521–542.
Schneggenburger R, Neher E. 2000. Intracellular calcium dependence
of transmitter release rates at a fast central synapse. Nature 406:
889 – 893.
Schwartz JP, Wilson DJ. 1992. Preparation and characterization of
type 1 astrocytes cultured from adult rat cortex, cerebellum and
striatum. Glia 5:75– 80.
445
Sikdar SK, Zorec R, Brown D, Mason WT. 1989. Dual effects of
G-protein activation on Ca-dependent exocytosis in bovine lactotrophs. FEBS Lett 253:88 –92.
Takamori S, Rhee JS, Rosenmund C, Jahn R. 2000. Identification of a
vesicular glutamate transporter that defines a glutamatergic phenotype in neurons. Nature 407:189 –194.
Taschenberger H, von Gersdorff H. 2000. Fine-tuning an auditory
synapse for speed and fidelity: developmental changes in presynaptic waveform, EPSC kinetics, and synaptic plasticity. J Neurosci
20:9162–9173.
Thomas P, Surprenant A, Almers W. 1990. Cytosolic Ca2⫹, exocytosis,
and endocytosis in single melanotrophs of the rat pituitary. Neuron
5:723–733.
Verkhratsky A, Solovyeva N, Toescu EC. 2002. Calcium excitability
of glial cells. In: Volterra A, Magistretti PJ, Haydon PG, editors.
The tripartite synapse. Oxford: Oxford University Press. p 99 –
109
Voets T, Moser T, Lund PE, Chow RH, Geppert M, Südhof TC, Neher
E. 2001. Intracellular calcium dependence of large dense-core vesicle exocytosis in the absence of synaptotagmin I. Proc Natl Acad Sci
USA 98:11680 –11685.
Zhang Q, Pangršič T, Kreft M, Kržan M, Li N, Sul J-Y, Halassa M,
Van Bockstaele E, Zorec R, Haydon PG. 2004. Fusion related release of glutamate from astrocytes J Biol Chem (in press).
Zhang Q, Pascual O, Perea G, Araque A, Fukuda M, Haydon PG.
2003. Expression of synaptotagmin IV, V and XI in CA1 area
hippocampal astrocytes Prog No. 765.3. Abstract Viewer/Itinerary
Planner. Washington, DC: Society for Neuroscience. Online.
Zorec R, Sikdar SK, Mason WT. 1991. Increased cytosolic calcium
stimulates exocytosis in bovine lactotrophs. Direct evidence from
changes in membrane capacitance. J Gen Physiol 97:473– 497.