Cortical Filamentous Actin Disassembly and Scinderin
Redistribution during Chromatfin Cell Stimulation Precede
Exocytosis, A Phenomenon Not Exhibited by Gelsolin
M. L. Vitale, A. Rodriguez Del Castillo, L. Tchakarov, and J.-M. Trifar6
Secretory Process Research Program, Department of Pharmacology, Faculty of Medicine, University of Ottawa,
Ottawa, Ontario, Canada K1H 8M5
Abstract. Immunofluorescence and cytochemical
studies have demonstrated that filamentous actin is
mainly localized in the cortical surface of the chromaffin
cell. It has been suggested that these actin filament
networks act as a barrier to the secretory granules,
impeding their contact with the plasma membrane.
Stimulation of chromaffin cells produces a disassembly
of actin filament networks, implying the removal of the
barrier. The presence of gelsolin and scinderin, two
Ca2+-dependent actin filament severing proteins, in the
cortical surface of the chromatfin cells, suggests the
possibility that cell stimulation brings about activation
of one or more actin filament severing proteins with
the consequent disruption of actin networks. There-
fore, biochemical studies and fluorescence microscopy
experiments with scinderin and gelsolin antibodies and
rhodamine-phalloidin, a probe for filamentous actin,
were performed in cultured chromaffin cells to study
the distribution of scinderin, gelsolin, and filamentous
actin during cell stimulation and to correlate the pos-
sible changes with catecholamine secretion. Here we
report that during nicotinic stimulation or K+-evoked
depolarization, subcortical scinderin but not gelsolin is
redistributed and that this redistribution precedes cate-
cholamine secretion. The rearrangement of scinderin
in patches is mediated by nicotinic receptors. Cell
stimulation produces similar patterns of distribution of
scinderin and filamentous actin. However, after the
removal of the stimulus, the recovery of scinderin cor-
tical pattern of distribution is faster than F-actin reas-
sembly, suggesting that scinderin is bound in the corti-
cal region of the cell to a component other than F-actin.
We also demonstrate that peripheral actin filament dis-
assembly and subplasmalemmal scinderin redistribution
are calcium-dependent events. Moreover, experiments
with an antibody against dopamine-~-hydroxylase sug-
gest that exocytosis sites are preferentially localized to
areas of F-actin disassembly.
sponse to cholinergic stimulation and upon Ca 2+ entry, the
granules fuse with the plasma membrane and release their
soluble contents to the cell exterior by exocytosis (Trifar6,
1977; Viveros, 1974). Immunofluorescence and cytochemi-
cal studies have described the presence of a mesh of filamen-
tous actin (F-actin) underneath the chromattin cell plasma
membrane (Lee and Trifar6, 1981; Trifar6 et al., 1984;
Cheek and Burgoyne, 1986). It has been proposed that actin
networks act as a barrier to the secretory granules by block-
ing their movement towards the plasma membrane (Trifar6
et al., 1982, 1984, 1989; Cheek and Burgoyne, 1986, 1987;
Burgoyne and Cheek, 1987; Burgoyne et al., 1989). Evi-
HROMAFFIN cells of the adrenal medulla store their
secretory products in specialized organelles, the chro-
marlin granules (Smith, 1968; Trifar6, 1977). In re-
M. L. Vitale is on leave from the CONICET-Buenos Aires, Argentina.
A. Rodrfguez Del Castillo is on leave from the University of La Laguna,
La Laguna, Spain. L. Tchakarov is a postdoctoral fellow from the Medical
Research Council of Canada.
dence obtained from different experimental approaches has
demonstrated that stimulation of chromaflin cells brings
about a disassembly of cortical F-actin networks, suggesting
the removal of the physical barrier to granule movement
(Cheek and Burgoyne, 1986, 1987; Burgoyne and Cheek,
1987; Burgoyne et al., 1989; Trifar6 et al., 1989). The exis-
tence of actin-binding proteins that regulate the dynamics
of actin networks (Yin and Stossel, 1979; Craig and Pol-
lard, 1982; Stossel et al., 1985; Maekawa et al., 1989;
Rodriguez Del Castillo et al., 1990) strongly suggests a role
for these proteins in the disassembly of actin filaments trig-
gered by cell stimulation. Therefore, it was of interest to in-
vestigate the participation in this process of gelsolin (Yin and
Stossel, 1979) and scinderin (Rodriguez Del Castillo et al.,
1990) two Ca~+-dependent actin-binding proteins that con-
trol actin filament length.
Gelsolin is an actin filament capping and severing protein
found in many cells, including chromaffin cells, and in ex-
tracellular fluids (Yin and Stossel, 1989; Yin et al., 1981;
Stossel et al., 1985; Trifar6 et al., 1985; Bader et al., 1986).
© The Rockefeller University Press, 0021-9525/91/06/1057/II $2.00
The Journal of Cell Biology, Volume 113, Number 5, June 1991 1057-1067 I057
Previous work from our laboratory has described the pres-
ence in chromaflin cells of another actin binding protein that
can be eluted by an EGTA containing buffer from actin-
DNase I affinity columns along with gelsolin (Bader et al.,
1986). Recently we have isolated, characterized, and given
the name of "scinderin" to this new protein (Rodriguez Del
CastiUo et al., 1990). Scinderin is an 80-kD cytosolic pro-
tein that shortens actin filament length provided Ca 2+ is pres-
ent in the medium (Rodriguez Del Castillo et al., 1990).
The fact that, when chromaffin ceils are stimulated there
is an entry of Ca 2+ (Douglas, 1968) with a consequent in-
crease in its intracellular level (Cheek et al., 1989) prompted
us to investigate whether this condition would influence gel-
solin and/or scinderin distributions in a way that could be
correlated to actin filament disassembly. This paper describes
biochemical and immunocytochemical experiments performed
in cultured chromaflin ceils. We have studied the cellular lo-
calization of scinderin and gelsolin under different experi-
mental conditions and compared their su.bcellular redistribu-
tion with F-actin disassembly and catecholamine secretion.
The present experiments demonstrate that during cell stimu-
lation, subplasmalemmal scinderin, but not gelsolin, is re-
distributed in chromaflin cells, that this redistribution pre-
cedes exocytosis, and that exocytosis sites are preferentially
localized to areas of F-actin disassembly. The redistribution
of scinderin is mediated by nicotinic receptors. The results
also show that the redistribution of scinderin and F-actin dis-
assembly are Ca2+-dependent events and that similar patterns
of distribution for scinderin and F-actin are observed during
stimulation. A preliminary account of this work has been
presented elsewhere (Vitale, M. L., A. Rodrfguez del Cas-
tillo, L. Tchakarov, M. L. Novas, and J.-M. Trifar6. 1990.
J. Cell Biol. 111:424a).
Materials and Methods
(a) Chromaffin Cell Culture
Bovine adrenal glands were obtained from a local slaughterhouse and
chromaffin cells were isolated by collagenase digestion and further purified
using a Percoll gradient (Trifar6 and Lee, 1980). Cells were plated on
collagen-coated glass coverslips contained within plastic Petri dishes at a
density of 0.25 x 106 cells/35-mm dish for fluorescence microscopy
studies or in collagen-coated plastic Petri dishes at a density of 0.5 × 10 ~
cells/35-mm dish for catecholamine release studies. Cells were grown at
37°C in a humidified incubator under a CO2 + air atmosphere for 48 h as
previously described (Trifar6 and Lee, 1980).
(b) Source of Antibodies
Polyclonal antibodies were raised in rabbits against purified bovine scinde-
tin, gelsolin, and dopamine-~-hydroxylase (D/~H) 1 as previously described
(Bader et al., 1986; Rodriguez Del Castillo et al., 1990; Trifar6 et al.,
1976). Scinderin antiserum 6 thus obtained does not recognize gelsolin and
gelsolin antiserum does not cross-react with scinderin (Bader et al., 1986;
P-,odriguez Del Castillo et al., 1990; Tchakarov et al., 1990). Moreover,
scinderin was the only protein immunoprecipitated from an adrenal medul-
lary cytosolic preparation by antiserum 6 (Tchakarov et al., 1990). Anti-
D/3H IgG has been previously characterized (Trifiu-6 et al., 1976). A mouse
mAb against gelsolin (clone GS-2C4) was purchased from Sigma Chemical
Co. (St. Louis, MO). This antibody is specific for an epitope localized on
a 47-kD peptide derived from a chymotryptic cleavage of human gelsolin
(Chaponnier et al., 1986).
1. Abbreviations used in this paper: D/~H, dopamine-~-hydroxylase.
(c) Extraction of ChromaOin Cell
Chromaflin cell cytoskeleton was prepared essentially as described by
Bader et al. (1984). Briefly, chromaflin cells were cultured in a 100-mm-
diam plastic Petri dish at a density of 60 x 106 cells/dish. After 24 h in
culture, the cells were resuspended by washing several times with buffer A
(PBS: 100 mM sodium phosphate, 130 mM NaC1, pH 7.2 containing 5 mM
EGTA, and 2 mM PMSF). Cells were collected by centrifugation at 5,000 g
for 10 rain. The sediment thus obtained was resuspended in 1 ml buffer A
containing 600 mM KCI, 10 mM MgC12, 1% Triton X-100, and 0.5 nag
DNAse l/ml. The mixture was centrifuged at 27,000 g for 30 rain and the
sediment containing Triton-insoluble proteins (cytoskeleton) was washed
three times with buffer A, resuspended in 1 ml electrophoresis buffer and
boiled for 5 rain.
(d) Preparation of Total Proteins and Actin-binding
Proteins from Bovine Adrenal Medulla
Bovine adrenal medullae (60 g) were washed in ice-cold Locke's solution to
remove the blood and then homogenized in 300 mM sucrose, 20 raM Tris-
HC1, pH 7.5, 100 mM KC1, 5 mM DTT, 1 m.M PMSF, 5 ram N-ethyl-
maleimide, I mM EGTA and 1 mM Na-ATP (1 g of medulla in 2 mi of solu-
tion), using a Sorvail omnimixer (Sorvall Instruments Div., Newton, CT).
The homogenate was centrifuged at 1,000 g for 10 rain. The supernatant
thus obtained (total protein sample) was centrifuged at 100,000 g for 60 rain
and CaC12 was added to obtain a final concentration of 2 raM. The prepa-
ration was then applied to a DNase I-Sepharose 413 column prepared as de-
scribed by Bader et al. (1986). The column was pre-cquilibrated with buffer
A (20 mM Tris-HCl pH 7.5, 2 mM CaC12, 0.5 mM Na-ATP, 1 mM DTT,
and 1 mM PMSF) containing 100 mM KCI. The column was then washed
extensively (300 ml) with the same buffer A but this time containing 300
mM KCI. Finally the Ca2+-dependent actin binding-proteins were eluted
with buffer B (20 mM Tris-HC1 pH 7.5, 100 mM KC1, 1 mM DTT, 2 mM
EGTA, and 1 mM PMSF). Samples from either total proteins or EGTA elu-
ares were used in the experiments.
(e) Preparation of Scinderin
Scinderin was purified from bovine adrenal medullae following a chromato-
graphic procedure previously described (Rodriguez Del Castillo et al.,
1990) except that the last chromatographic step (HPLC) was omitted.
Therefore, the preparation obtained in this case also contains small amounts
of gelsolin and other cae+-dependent actin-binding proteins.
09 Electrophoresis and Immunoblotting
Monodimensional SDS-PAGE was performed according to Doucet and
Trifar6 (1988); gels were usually run at 60 V overnight in a Bio-Rad Protean
I apparatus (Bio-Rad Laboratories, Inc., Richmond, CA). The protocol for
immunoblotting was as described by Towbin et al., (1979). After electro-
phoresis, proteins were electrotransferred onto nitrocellulose membranes
(Hoefer Scientific Instruments, San Francisco, CA). Membranes were first
blocked with 3 % BSA in PBS and then incubated with scinderin antiserum
6 (1:500 dilution) or gelsolin antiserum (1:500 dilution) for 60 rain. Mem-
branes were next incubated with goat antirabbit immunoglobulin G-alkaline
phosphatase conjugate (1:3,000 dilution) for another 60 min. Color was de-
veloped by treatment with a mixture ofp-nitroblue tetrazolium chloride a,d
5-bromo-4-chloride-3-indolyl phosphate-toluidine salt.
(g) Fluorescence Microscopy
Single Staining. Experiments were started by rinsing the cultured cells
three times with regular Locke's solution (in millimolar: NaC1, 154; KCI,
2.6; K2HPO4, 2.15; KI-I2PO4, 0.85; MgCI2, 1.2; CaCI2, 2.2; and glucose,
10.0; pH 7.2). Cells were then incubated for different periods of time with
Locke's solution in the absence (control) or presence (stimulated) of differ-
ent compounds. Chromaffin cells were fixed in 3.7% formaldehyde in
Locke's solution for 20 rain at different times after initiated the stimulation
and processed for fluorescence microscopy as described previously (Lee
and Trifart, 1981). Briefly, cells were permeabilized by three successive ex-
posures of 5 rain each to 50, 100, and 50% acetone. Preparations were
washed several times with PBS and then incubated at 370C with either
scinderin antiserum 6 (1:20 dilution), gelsolin antiserum (1:20 dilution) or
mouse monoclonal antibody against gelsolin (1:40 dilution) for 60 rain.
The Journal of Cell Biology, Volume 113, 1991 1058
Coverslips were thoroughly washed with PBS and were next incubated at
37°C with goat antirahbit immunoglobulin G-fluorescein isothiocyanate
conjugate (FITC-IgG) (1:80 dilution) for another 60 min. When a mouse
mAb against gelsolin was used, goat antimouse immunoglobulin G-tetra-
methylrhodamine isothiocyanate conjugate (TRITC-IgG; 1:256 dilution)
was used as a second antibody. Coverslips were washed several times with
PBS and mounted in glycerol-PBS (1:1; vol/vol). Control experiments were
performed with (a) second antibody alone and (b) first antiserum after ad-
sorption with the correspondent antigen.
Double Staining. Chromaflin cells were incubated with Locke's solution
for different periods of time in the absence (control) or presence (stimu-
lated) of different compounds. For scinderin/F-actin or gelsolin/F-actin
staining, fixation and acetone-induced permeabilization were as above. Per-
meabilized cells were incubated at room temperature for 40 min with 0.3
#M rhodamine-labeled phalloidin, a probe for filamentous actin (Faultish
et al., 1988). Coverslips were then washed six times with PBS and in-
cubated with scinderin antiserum 6 or gelsolin antiserum followed by FITC-
IgG as described above.
For DflH/F-actin staining, cells incubated for 40 s with 10/~M nicotine
were fixed with 3.7% formaldehyde and then incubated with anti-DflH IgG
(1:100 dilution) for 60 min, washed as indicated above, and further in-
cubated with FITC-IgG as described previously. This was followed by incu-
bation for 40 min with 0.3/~M rhodamine-labeled phalloidin.
Preparations were observed with a Leitz Ortholux II fluorescent micro-
scope equipped with a 200-W high-pressure lamp and a Ploemopack II inci-
dent light illuminator equipped with an 1-filter block (KP 490 plus 1 mm
GG 455 exciting filter, TK dichroic beam splitting mirror, K 515 suppres-
sion filter) for fluorescein and an M-filter block (2 mm BG 36 plus S 546
exciting filter, TK 580 dichroic beam splitting mirror, K 580 suppression
filter) for rhodamine. Photographs were taken with Kodak Tri-X pan films
(400 ASA). 100 cells per coverslip were examined and classified as having
either "continuous staining" as shown in Fig. 3 A, a. or having "discontinu-
ous patched staining pattern" as can be seen in Fig. 3 A, b. This was done
without knowing whether cells were from control or stimulated prepara-
tions (single blind design).
(h) Catecholamine Release Studies
Catecho]amine OUlpUt was evaluated by ~
output essentially as described previously (Trifar6 and Lee, 1980; Kenigs-
berg and Trifar6, 1980). Previous experiments from our laboratory have
demonstrated that when chromattin cells are loaded with [3H]NA under
controlled conditions, there was a concomitant and parallel release of en-
dogenous catecholamines and [3H]NA upon stimulation (Trifar6 and Lee,
1980; Trifar6 and Bourne, 1981).
Chromaltin cells were plated on collagen-coated plastic dishes (0.5 x
106 cells/35-mm dish) and grown for 48 h. Dishes were washed three times
with special medium (in millimolar: NaCI, 110; NaI-ICO3, 40; KC1, 5;
MgSO4, 1; NaH2PO4, 1; NaPyruvate, 1; CaCI2, 2; Fe(NO3)3, 2.5 x 10-4;
ascorbic acid, 0.1; pH 7.2 adjusted with Hepes). Chromatlin cells were in-
cubated at room temperature for 5 rain with 1 ml of special medium contain-
ing 0.1 nmol [3H]NA (42.1 Ci/mmol sp act; New England Nuclear, Bos-
ton, MA). After this labeling step, each dish was incubated with six changes
of I ml regular Locke's solution over a 60-rain period before the experiments
were commenced. Cells were then incubated with 1 ml regular Locke's solu-
tion with or without 10 ~,M nicotine or 56 mM K + for different periods.
After these periods of time, the entire (1 ml) incubation medium was re-
moved and radioactivity determined. [3H]NA cell content was determined
by treating the dishes with 1 ml of 10% TCA for 10 rain followed by two
washes of 0.5 ml 6% TCA; the three aliquots were combined and the radio-
activity was measured in a liquid scintillation spectrometer (Beckman In-
strumants, Fullerton, CA). Total [3H]NA cell content was calculated by
adding the [3H]NA released into the medium to the [3H]NA extracted with
"I'CA. The assay allowed the determination in 1 ml sample of a level of
[3H]NA equivalent to 0.35% of the total cell catecholamine content.
(i) Protein Determination
Protein concentrations were determined by the method of Bradford (1976)
using BSA as standard.
Nicotine (hydrogen tartrate salt), muscarine (chloride), D-tubocuradne (chlor-
ide), monoclonal anti-gelsolin clone GS-2C4, TRrrC-IgG, and FITC-IgG
were purchased from Sigma Chemical Co. Materials for SDS-PAGE and
Figure 1. Specificity of scinderin and gelsolin antibodies. (A) Chro-
maffin cell cytoskeletal proteins (50 #g, lane 1) and a partially
purified scinderin preparation (5 t~g, lane 2) were subjected to
SDS-PAGE and stained with Coomassie brilliant blue. Lane St
shows the position of molecular weight standards. Lanes 1', 2', 1",
and 2" are the immunoblots of the same preparations after incuba-
tion with scinderin antiserum 6 (lanes/' and 2') or gelsolin antise-
rum (lanes 1" and 2"). The presence of myosin (M), tubulin (T),
and actin (A) in the chromaffin cell cytoskeleton and of scinderin
(Sc) and gelsolin (G) in a partially purified scinderin preparation
is indicated by arrowheads. (B) Cytoskeletal proteins (200 ~g, lane
1), adrenal medullary total proteins (200/zg, lane 2) and DNase
I-actin affinity-purified Ca2÷-dependent actin-binding proteins (30
/zg, lane 3) were subjected to SDS-PAGE and stained with Coo-
massie brilliant blue. Lane St shows the position of molecular
weight standards. Lanes 1'-3" are the immunoblots of the same
preparations after incubation with scinderin antiserum 6 (lanes 1',
2', and 3') or gelsolin antiserum (lanes F, 2", and Y). A faint reac-
tion is observed with scinderin antiserum 6 (lanes/' and 21 and with
gelsolin antiserum (lanes/" and 2°). The presence of gelsolin (G),
scinderin (Sc), tubulin (T), and actin (A) is indicated by arrow-
immunoblotting were from Bio-Rad Laboratories, Ltd. (Mississauga, On-
tario, Canada). Rhodamine-labeled phalloidin was from Molecular Probes,
Inc. (Eugene, OR).
(a) Specificity of Scinderin and Gelsolin Antisera
Scinderin and gelsolin distribution in bovine adrenal chro-
matiin cells in culture was studied by using polyclonal an-
Vitale et al. Actin Disassembly and Scinderin Redistribution in Exocytosis
Figure 2. Control experiments for immunocytochemical staining.
No staining was observed when chromaffin cells were incubated
with antisera preadsorbed with the correspondent antigens (a',
scinderin;/1, gelsolin), a and b show phase contrast images of cells
present in a' and V, respectively. Bars, 10 #M.
tisera against both proteins. To study the distribution of
scinderin and gelsolin in bovine adrenal chromaflin cells in
culture, it was important to demonstrate the specificity of the
polyclonal antisera to be used. Previous work from our labo-
ratory demonstrated that scinderin and gelsolin antisera are
specific for these proteins and that do not cross-react with
any other chromaffin cell cytosolic protein (Bader et al.,
1986; Rodriguez Del Castillo et al., 1990; Tchakarov et al.,
1990). Cross-reactivity of both antisera with chromaffin cell
cytoskeletal proteins was also investigated. Fig. 1 A shows
electrophoretic patterns of chromaflin cell cytoskeletal pro-
teins (lane/) and partially purified scinderin preparation also
containing a small amount of gelsolin (lane 2). Myosin (M),
tubulin (T), and actin (A) can be recognized as components
of the cytoskeleton. Preparations of chromaffin cell cyto-
skeleton (Fig. 1 A, lanes 1 ', I") and partially purified scinde-
rin (Fig. 1 A, lanes 2', 2") were transferred onto nitrocellulose
membranes and incubated with either scinderin (Fig. 1 A,
lanes 1' and 2') or gelsolin antisera (Fig. 1 A, lanes 1" and
2"). Scinderin antiserum 6 did not cross-react with either
gelsolin or any cytoskeletal protein. Furthermore, gelsolin
antiserum did not cross-react with any cytoskeletal protein
or scinderin. To determine whether small amounts of scinde-
rin or gelsolin could be associated with the cytoskeleton,
gels (Fig. 1 B) were overloaded with four times more cy-
toskeletal protein (200/~g of 1% Triton-X-100 insoluble pro-
teins). The immunoblots obtained from these gels indicated
the presence in the cytoskeletal preparation of small amounts
of both scinderin (Fig. 1 B, lane 1 ') and gelsolin (Fig. 1 B,
lane 1"). However, it should be pointed out that the amount
of scinderin associated with the cytoskeleton represents <1%
of the total cellular scinderin. In addition, to discard a possi-
ble cross-reactivity of scinderin or gelsolin antisera with any
chromattin cell protein, gels were also overloaded with 200
/~g of total protein. Here again, the immunoblots showed
bands corresponding to either scinderin (Fig. 1 B, lane 2')
or gelsolin (Fig. 1 B, lane 2'), and indicated no cross-reac-
tivity of antisera with any additional chromattin cell protein.
These results show the specificity of antisera against scinde-
rin and gelsolin. Therefore, changes in the distribution of ei-
ther of the two antibodies will reflect changes in scinderin
or gelsolin subcellular localization.
(b) Distribution of Scinderin, Gelsolin, and of F-Actin
in Resting and Nicotine-stimulated ChromaJ~in Cells:
A Fluorescence Microscopy Study
Chromaffin cells, cultured for 48 h, were incubated with
regular Locke's solution alone or in the presence of 10 #M
nicotine for 5, 20, or 40 s. At the end of these incubation
periods, cells were processed for immunoffuorescence using
scinderin or gelsolin antibodies. Control experiments demon-
strated that chromaflin cells were not stained when incu-
bated with antisera adsorbed with the correspondent anti-
gens (Fig. 2).
To investigate within the same cell the subcellular organi-
zation of scinderin or gelsolin together with that of F-actin
some preparations were also stained with rhodamine-labeled
phalloidin, a probe for filamentous actin. Scinderin distribu-
tion in control cells showed a bright and continuous conical
fluorescent ring and a less intense and diffuse cytoplasmic
fluorescence (Fig. 3, a and e'). Nicotine stimulation caused
a fragmentation of the bright fluorescent ring suggesting a
redistribution of conical scinderin. Patches of scinderin ap-
peared clearly as a fragmented fluorescent ring at the
equatorial plane of the cells (Fig. 3 A, b, c, and d). The effect
of nicotine on scinderin reorganization was seen as early as
5 s of stimulation (Fig. 3 A, b).
Distribution of F-actin in chromaflin cells under resting
conditions showed a continuous cortical fluorescent ring (Fig.
3 B, e and Fig. 4 B, a3. Stimulation of ceUs with 10/~M nico-
tine produced a disruption in the rhodamine-phaUoidin coni-
cal fluorescent pattern (Fig. 3 B,f, g, h; and Fig. 4 B, b', c'),
suggesting depolymerization of F-actin. Interestingly, in 88 +
3 % of the nicotine-stimulated cells showing scinderin reor-
ganization (916 of 1,200 total cells examined), there was a
concomitant distribution of F-actin and scinderin (compare
fandf', g and g', h and h' in Fig. 3 B).
Gelsolin distribution in chromattin cells was studied using
monoclonal and polyclonal antibodies (Fig. 4, A and B, re-
spectively). Cells incubated with regular Locke's solution
and treated with gelsolin polyclonal antiserum (1:20 dilution)
showed a cortical cytoplasmic fluorescent pattern (Fig. 4 B,
a'). The fluorescent ring underneath the plasma membrane
was weaker than that observed in anti-scinderin-stained cells.
Under resting conditions, only 14 + 3 % of antigelsolin-la-
beled cells (700 cells examined) showed a discontinuous cor-
tical fluorescent pattern for gelsolin. Although exposure of
chromaflin cells to nicotine for 40 s produced the disassem-
bly of conical F-actin network it did not cause any modifica-
tion in the gelsolin cortical fluorescence (compare b' and b",
c' and c" in Fig. 4 B). Additional experiments, using a mouse
rnAb (1:40 dilution) against the 47-kD chymoptryptic frag-
ment of gelsolin, showed a diffuse cytoplasmic staining (Fig.
4 A, a, b, and c). To be certain that there was not redistribu-
tion of gelsolin or that gelsolin redistribution in response to
cell stimulation was masked by a strong fluorescence due to
a relative high concentration of antibodies used in the stain-
ing, different higher dilutions of either polyclonal (1:40, 1: 80,
The Journal of Cell Biology, Volume 113, 1991 1060
Figure 3. Localization of scinderin
and actin by fluorescence micros-
copy. (,4) Single staining: chromaf-
fin cells were stained by incuba-
tion with scinderin antiserum 6
followed by FITC-anti-rabbit IgG.
In resting cells (a) scinderin anti-
body fluorescence consists in a
bright and continuous peripheral
ring (open arrowhead) and a less
intense and diffuse cytoplasmic
staining. Incubation with 10 #M
nicotine for 5 (b), 20 (c), or 40 s
(d) produces a disruption of the
cortical fluorescent pattern. Some
patches are marked by arrowheads.
Bar, 10 #M. (B) Double staining:
chromaflin cells were incubated
with Locke's solution for 40 s in
the absence (control) or presence
(stimulated) of 10 #M nicotine.
Cells were sequentially stained
with rhodamine-labeled phalloi-
din followed by scinderin anti-
body and FITC-anti-rabbit IgG.
A control cell in (e, e') shows
continuous and intense rings of
fluorescence for F-actin (e) and
scinderin (e') colocalized at the
subplasmalemmal region (open
arrows). Stimulated cells display
a disrupted cortical fluorescent pat-
tern either for F-actin (f, g, h) or
scinderin (f', g', h'). There is a cor-
respondence between the patched
distribution of both actin and scin-
derin in each cell (comparefand
f', g and 4, h and h'). Some patches
are indicated by arrows.
and 1:100) or monoclonal (1:80 and 1:160) gelsolin antibod-
ies were tested. The results obtained from these experiments
were similar to those described above with higher concentra-
tions of antibodies.
(c) Time Courses of F-Actin Disassembly
and Scinderin Redistribution in Chromaffin Cells
Stimulated with Nicotine
Cells displaying scinderin redistribution in response to 10
gM nicotine stimulation also showed a similar time course
for actin filament disassembly (Fig. 5). However, upon
removal of the stimulus, the rate of recovery of cortical
scinderin distribution was faster than that of cortical actin
filament reassembly (Fig. 5). 80 s after the removal of nico-
tine from the incubation medium, 24 5:3 % (n = 300) of
cells showed normal cortical distribution of scinderin with
a concomitant fragmented cortical rhodamine-phalloidin
fluorescence (Fig. 5).
(d) Time Courses of Scinderin and Gelsolin
Redistribution and Catecholamine Release
in Chroma2~n Cells Stimulated with Either Nicotine
or a Depolarizing Concentration of K ÷
In view of the earlier redistribution (5 s) of scinderin oh-
served upon cell stimulation, time courses of scinderin redis-
tribution and catecholamine output were performed and
compared. Chromatfin cells were incubated for O, 5, 10, 20,
or 40 s with either 10 #M nicotine or 56 mM K ÷ or 40 s
with either 10 #M nicotine or 56 mM K ÷ followed by 50 or
80 s with regular Locke's solution. After these incubation
periods, the cells were immediately fixed and processed for
scinderin (antiserum 6) and gelsolin (antiserum) immuno-
fluorescence microscopy. Cells were classified as having a
continuous or a discontinuous cortical fluorescent pattern
and the percentage of cells showing redistribution was plotted.
During nicotinic stimulation, there was a sharp increase in
the percentage of cells displaying a redistribution of the cor-
tical scinderin fluorescence (Fig. 6 A, v). The maximum
value (78 + 2 %) was reached 40 s after nicotinic stimulation
was started. Removal of the stimulus at that point produced
a decline in the number of cells showing a discontinuous
fluorescent pattern. 80 s after the removal of nicotine from
the incubation medium, scinderin redistribution reached
In contrast to these observations the number of cells dis-
playing a disrupted cortical fluorescent ring for gelsolin (9
+ 1%) was not modified by nicotinic stimulation (Fig. 6 A,
,). Moreover, incubation with 10 gM nicotine for longer
periods of time (up to 180 s) did not increase this percentage
Vitale et al. Actin Disassembly and Scinderin Redistribution in Exocytosis
(data not shown). Control values for gelsolin redistribution
were always lower than those for scinderin redistribution (9
± 1 vs. 22 ± 3%).
Catecholamine output also rose sharply during nicotinic
.-~ 80 t I t°''M ~ I
N 2 t
0 20 40 60 80 100 120
Figure 5. Time courses of F-actin disassembly and scinderin redis-
tribution in response to nicotine stimulation in cultured chromaffin
cells. Chromattin cells cultured for 48 h were incubated for 0, 5,
20, or 40 s with 10 #M nicotine or for 40 s with nicotine followed
by an additional 50 or 80 s with regular Locke's solution. After
these periods of incubation cells were immediately fixed, permea-
bilized, and processed for fluorescence microscopy using rho-
damine-phalloidin and scinderin antiserum 6 as described in Ma-
terials and Methods. 100 cells per coverslip were examined and
classified as having either a "continuous cortical fluorescent pat-
tern" (see Fig. 3, e-e) or a "discontinuous cortical fluorescent pat-
tern" (see Fig. 3, f-f'). This was done without knowing whether
cells were control or stimulated with nicotine. Each value shown
represents the mean + SEM of the percentage of discontinuous cor-
tical fluorescence patterns of 6-8 coverslips (600-800 cells for each
value) containing ceils from three different chromaffin cell cultures.
Upon removal of the stimulus the rate of recovery was different
for scinderin and F-actin (90 and 120 s).
Figure 4. Localization of gelsolin
and actin by fluorescence micros-
copy in cultured chromaflin ceils.
(A) Single staining: chromattin
ceils were stained after incuba-
tion with a monoclonal antibody
against gelsolin (1:40 dilution) fol-
lowed by TRITC-anti-mouse IgG.
Control cells ((3 s) or ceils incu-
bated for 40 s with 10 tiM nico-
tine (b and c) showed similar
staining for gelsolin. Bar, 10 #M.
(B) Double staining: chromafiin
ceils grown for 48 h were stained
for F-actin with rhodamine-labeled
phalloidin and then with gelsolin
antiserum followed by FITC-anti-
rabbit IgG. Control cells (0 s) dis-
play a bright peripheral fluores-
cent pattern for F-actin (a') and a
less intense cortical ring and a
diffuse cytoplasmic staining for
gelsolin (a"). Incubation of ceils
with 10 #M nicotine for 40 s pro-
duces a disruption of the F-actin
cortical fluorescent pattern (b' c',
arrows) without affecting gelso-
lin staining (b ~, c"). Continuous
peripheral fluorescent rings are
marked by open arrows.
stimulation (Fig. 6 A, o). However, the increase in catechol-
amine release lagged 15-20 s behind scinderin redistribu-
tion. [3H]NA output leveled off when nicotine was removed
from the medium (Fig. 6 A, o).
To test whether direct cell depolarization also induces
scinderin redistribution, chromaffin cells were exposed to
high K + concentrations. Depolarization of chromaflin cells
with 56 mM K ÷ also caused a reorganization of subplas-
malemmal scinderin (Fig. 6 B, v). An increase in the num-
ber of cells displaying a patched fluorescent pattern could be
detected 10 s after K + depolarization was initiated. The
maximum percentage of cells displaying a scinderin rear-
rangement (68 ± 7%) was observed after 20 s of depolariza-
tion; this value was lower than that obtained for nicotinic
stimulation (78 ± 2%). Replacement of high K ÷ medium
for regular Locke's solution produced a decline in the num-
ber of cells showing a disrupted scinderin cortical fluores-
cence. However, 80 s after lowering K + concentration the
values for scinderin redistribution were still higher than
those of controls. The distribution of gelsolin was not modi-
fied by K ÷ depolarization (Fig. 6 B, *).
K ÷ depolarization stimulated catecholamine output (Fig.
6 B, o); a sharp increase in release was detected 20 s after
initiation of the K + challenge. Removal of high K ÷ by re-
placing the incubation medium for regular Locke's solution
leveled off [3H]NA release (Fig. 6 B, o). As in the case of
nicotinic stimulation, the rise in the percentage of cells
showing discontinuous scinderin cortical fluorescent ring
preceded the increase in catecholamine output.
(e) Characterization of Nicotine-induced
Scinderin Redistribution: Evidence for a Nicotinic
We have previously demonstrated that stimulation of chro-
The Journal of Cell Biology, Volume 113, 1991 1062
1 O0 - -
8 0 .......
6 2b ,o go Bo ~6o,io
Figure 6. Time courses of
scinderin and gelsolin redis-
tribution and [3H]NA output
in response to stimulation in
cultured chromaffin cells. (A)
Nicotinic stimulation: chro-
marlin cells cultured for 48 h
~0 were incubated for 0, 5, 10,
i 20, or 40 s with 10 ~M nico-
tine or for 40 s with nicotine
) followed by an additional 50-
or 80-s period with regular
Locke's solution. After these
periods of incubation, cells
were immediately fixed, per-
meabilized, and processed for
copy using either scinderin
(v) or gelsolin (¢) antisera. 100 cells per coverslip were examined
and classified, as described in the legend to Fig. 5. This was done
without knowing whether cells were control or stimulated with nic-
otine. Each value plotted represents the mean + SEM of the per-
centage of discontinuous cortical fluorescent pattern of 6-8 cover-
slips (600-800 cells for each value) containing cells from three
different cell cultures. [3H]NA output (o): cultured chromaffin
cells with catecholamine stores labeled with [3H]NA were incu-
bated for 0, 5, 10, 20, 30, or 40 s with 10/xM nicotine or for 40 s
with 10/~M nicotine followed by an additional 50- or 80-s period
with regular Locke's solution. After each of those periods media
were removed and their radioactivity measured. Basal I~H]NA out-
put was determined by incubating the cells with nicotine-free Locke's
solution for the same periods of time as above. Basal values (0.7-
1.0%) were subtracted from the corresponding data obtained dur-
ing stimulation. Nicotine-induced [3H]NA secretion is expressed
as percentage of total [3H]NA cell content. Each point represents
the mean + SEM of values obtained from three different culture
dishes. (B) K ÷ depolarization: Chromaftin cells grown in cover-
slips for 48 h were incubated at room temperature for 0, 5, 10, 20,
or 40 s with a Locke's solution containing 56 mM K + (high K +) or
40 s in high K ÷ followed by an additional 50- or 80-s period with
regular Locke's solution (regular K+). [3H]NA output (o): chro-
marlin cells with their catecholamine stores labeled with [3H]NA
were incubated with a 56 mM K ÷ Locke's solution for 0, 5, 10, 20,
30, or 40 s or for 40 s with high K ÷ Locke's solution followed by
50 or 80 s with regular Locke's solution. The procedure followed
in these experiments was as described above for nicotine stim-
martin ceils by acetylcholine induces disassembly of cortical
actin networks (Trifar6 et al., 1989). Because acetylcholine
stimulates both, muscarinic and nicotinic receptors, experi-
ments were performed to determine whether scinderin redis-
tribution was exclusively a nicotinic mediated response or
was also affected by muscarinic receptor stimulation. Stimu-
lation of chromatfin cells with 10/~M nicotine for 40 s pro-
duced a fourfold increase in the percentage of cells showing
scinderin reorganization when compared with control levels
(Fig. 7). The redistribution of scinderin was blocked when
10/~M D-tubocurarine (dtc), a nicotinic antagonist, was pres-
ent in the incubation medium (Fig. 7). Moreover, incubation
of ceils with 10/~M muscarine (musc) for 40 s did not cause
any rearrangement of subplasmalemmal scinderin (Fig. 7) or
F-actin disassembly (control: 20 + 4%, n = 300 cells and
musc: 18 5: 3% disassembly, n = 300 cells). The results
clearly indicate that scinderin redistribution is a nicotinic
._c o 4o-
'~ 20 -
Figure 7. Effects of cholinergic and anticholinergic drugs on scin-
derin redistribution in chromaffin cells in culture. 48-h-old chromaf-
fin cell cultures grown on collagen coverslips were incubated for
40 s under the following experimental conditions: (a) regular Locke's
solution alone (control), Locke's solution containing (b) 10 t~M
nicotine (nic), (c) 10 #M nicotine plus 10 #M D-tubocurarine (dtc)
or (d) 10 t~M muscarine (muse). After these treatments, cells were
fixed, permeabflized, and processed for immunofluorescence using
scinderin antiserum 6. 100 cells per coverslip were examined and
scinderin distribution was classified as described in the legend to
Fig. 5. Each value represents the mean + SEM of the percentage
of discontinuous scinderin distribution of 6-8 coverslips (600"800
cells for each value) containing cells from three different cell
09 Calcium Dependence of Sdnderin Redistribution
and F-Actin Disassembly by Nicotine-stimulated
or K ÷~epolarized Chromaffin Cells
Catecholamine release from chromaflin cells in response to
1°° 1 Fl°°
60 60 ~ ©
lOpM nic - + +
56 mM K +
2.2 mid Co 2+
+ + --
Figure 8. Effect of extracellular Ca 2+ on nicotine-evoked or high
K+-induced scinderin redistribution and identical F-actin disas-
sembly in cultured chrornaflin cells. Chromatiin cells grown for
48 h on collagen coated coverslips were incubated with Locke's so-
lution (control) or with Locke's solution containing 10 #M nico-
tine (nic) or 56 mM K + either each case containing either 2.2 mM
Ca 2+ or 0.1 mM EGTA (Ca 2+ free). Following these incubations,
cells were fixed, permeabilized, and processed for double staining
fluorescence microscopy using rhodarnine-labeled phalloidin and
scinderin antiserum 6 as indicated in Materials and Methods. 100
cells per coverslip were examined for scinderin (fluorescein stain-
ing) and for F-actin (rhodamine fluorescence) peripheral distribu-
tion and were classified as having a continuous or a patched cortical
staining. Each value represents the mean + SEM of the percentage
of discontinuous scinderin and F-actin distribution of 4-5 cover-
slips (400-500 cells for each value) containing cells from two
different cell cultures.
Vitale et al. Actin Disassembly and Scinderin Redistribution in ExocyWsis
Figure 9. Presence of surface D~H in cortical areas devoid of
rhodamine-phalloidin staining. Chromaffin cells grown for 48 h
were incubated for 40 s with 10 #M nicotine, fixed and stained for
D/~H and F-actin as described in Materials and Methods. Fluores-
cent patterns of two stimulated cells (a, a' and b, b') are shown as
they appeared after incident light illumination for rhodamine (a and
b) and fluorescein (a' and b'). Arrows show the absence of rhoda-
mine-phalloidin staining (F-actin) and the presence of FITC-IgG
staining. Bar, 5 #M.
stimulation is triggered by the entry of extracellular Ca ~+
through so-called slow Ca 2+ channels. Consequently, it was
of interest to determine whether scinderin redistribution also
requires the presence of extracellular Ca 2+. In resting con-
ditions, 25 + 4% of chromaflin cells showed discontinuous
distribution of subplasmalemmal scinderin (Fig. 8). This
percentage rose when the cells were incubated for 40 s with
10/zM nicotine or were exposed for the same period of time
to 56 mM K + (77 + 3 and 63 + 2 %, respectively) (Fig. 8).
F-actin disassembly either during resting conditions or stim-
ulation also followed a similar pattern of distribution (Fig.
8). Disruption of scinderin and actin cortical fluorescent
rings in both experimental conditions was not observed in
the absence of extracellular Ca ~+ (Fig. 8). The results indi-
cate that stimulation evoked Ca 2÷ influx is a necessary re-
quirement for scinderin redistribution and F-actin depoly-
merization in response to either nicotine stimulation or K ÷
depolarization. Moreover, the data also show that Ca 2÷ en-
try affects F-actin disassembly and scinderin reorganization
in the same manner, suggesting again a close functional rela-
tionship between the two proteins.
(g) Exocytosis Is Observed in Areas
of F-Actin Disassembly
The disassembly of cortical F-actin produced by cell stimula-
tion would suggest that subplasmalemmai areas devoid of
F-actin are formed and that these might be zones of low cyto-
plasmic viscosity and probably high secretory granule mo-
bility. The absence of a cytoskeletal barrier in these areas,
would allow the interaction of secretory granules with plasma
membranes with the subsequent release of granule contents
to the cell exterior by exocytosis. To test the possibility that
exocytotic pits might be present in plasma membrane areas
devoid of F-actin, chromaffin cells were stimulated with 10
#M nicotine for 40 s, fixed, and incubated with anti-D/~H-
IgG to detect the presence of chromaflin granule membranes
on the cell surface. Fig. 9 (arrows) shows DflH cell surface
staining in areas devoid of F-actin as indicated by the absence
of rhodamine phalloidin fluorescence.
Work from our laboratory, as well as others, has demon-
strated that filamentous actin is mainly localized in the corti-
cal surface of the chromaffin cell (Lee and Trifar6, 1981;
Trifar6 et al., 1984, 1989; Cheek and Burgoyne, 1986). We
have also suggested that cortical F-actin acts as a barrier to
the secretory granules, impeding their contact with the plasma
membrane. Chromaflin granules contain ot-actinin (Aunis et
al., 1980; Trifar6 et al., 1982) and fodrin (Perrin and Aunis,
1985), anchorage proteins which mediate filamentous actin
association with these vesicles. Stimulation of chromaflin
cell produces disassembly of actin networks and removal of
the barrier (Cheek and Burgoyne, 1986, 1987; Burgoyne et
al., 1989; Trifar6 et al., 1982, 1984, 1989). This interpreta-
tion is based on the following evidence. Cytochemical experi-
ments with rhodamine-labeled phalloidin and actin antibod-
ies indicated that in resting chromaflin cells, a filamentous
actin network is visualized as a strong cortical fluorescent
ring (Lee and Trifar6, 1981; Cheek and Burgoyne, 1986,
1987; Trifar6 et al., 1989). Nicotinic receptor stimulation
produces a fragmentation of this fluorescent ring leaving cell
cortical areas devoid of fluorescence (Cheek and Burgoyne,
1986, 1987; Trifar6 et al., 1989). These changes are accom-
panied by a decrease in F-actin associated with a concomi-
tant increase in G-actin as evaluated by the DNase I inhibi-
tion assay (Cheek and Burgoyne, 1986; Trifar6 et al., 1989).
These changes are also accompanied by a decrease in the
amount of F-actin recovered with the Triton-X-100 insolu-
ble (cytoskeleton) protein (Burgoyne et al., 1989; Trifar6,
1990). F-actin network disassembly has also been observed
in mast cells upon stimulation (Koefer et al., 1990) and in
depolarized (high K*) synaptosomes (Bernstein and Barn-
The present experiments clearly demonstrate that stimula-
tion of chromaffin cells with either nicotine or a depolarizing
concentration of K ÷ causes disassembly of cortical F-actin
networks and redistribution of subplasmalemmal scinderin.
Gelsolin on the other hand does not show such a rearrange-
ment. To be certain that this was the case, polyclonal and
monoclonal antibodies against gelsolin were used and these
were tested at different dilutions on resting and stimulated
cells. Under these conditions, no rearrangement or changes
in the fluorescence pattern of gelsolin were observed. These
observations ruled out the possibility that gelsolin redistribu-
tion in response to cell stimulation was masked by a strong
antibody fluorescence. Thus, the effect of cell stimulation
seems to be quite specific for scinderin. Previous studies
The Journal of Cell Biology, Volume 113, 1991 1064
from our laboratory have demonstrated that scinderin is a
structurally different protein from gelsolin (Rodriguez Del
CastiUo et al., 1990). Scinderin and gelsolin have different
molecular weights, isoelectric points, amino acid composi-
tion and yield different peptide maps after limited proteolytic
digestion (Rodriguez Del Castillo et al., 1990). Both pro-
teins have an actin filament severing activity which is Ca ~+
dependent; in the case of gelsolin, severing activity is in-
hibited by phosphatidylinositol 4,5 biphosphate (Yin et al.,
1988). Further work from our laboratory has demonstrated
different tissue expressions for scinderin and gelsolin. Scinde-
rin seems to be expressed in neuronal, endocrine, and exo-
crine tissues (Tchakarov et al., 1990) systems in which secre-
tion is a main function. Immunocytochemical studies showed
that in chromaflin cells scinderin has a diffuse cytoplasmic
and a more dense subplasmalemmal distribution (Rodriguez
Del Castillo et al., 1990). Instead, gelsolin only showed a
diffuse cytoplasmic distribution (Rodriguez Del Castillo et
al., 1990). Therefore, experimental data suggest that gelso-
lin and scinderin are two distinct Ca:÷-dependent F-actin
severing proteins that may also differ in their fine regulation
by intracellular messengers.
The present studies also show that scinderin redistribution
and actin filament disassembly, induced by either nicotine or
high K +, precedes catecholamine release. The lag period
observed between scinderin redistribution/F-actin disassem-
bly and catecholamine release was not due to the lack of sen-
sitivity of the catecholamine release assay used. We have
previously demonstrated that the [3H]NA taken up by cul-
tured chromaffin cells is stored and released together with
endogenous catecholamines and that the measurement of
[3H]NA in the incubation medium gives a precise indica-
tion of total catecholamine release (Trifar6 and Lee, 1980;
Kenigsberg and Trifar6, 1980; Trifar6 and Bourne, 1981).
The catecholamine assay used in the present experiments de-
tect catecholamine concentrations equal to 0.35% of total
content. The present results show that after 5 s of initiated
cell stimulation, 65 % of the cells showed scinderin redistri-
bution and F-actin disassembly. The catecholamine assay
could easily have detected 65 % (2.5 % of total catechol-
amine content) of the total release (4.0% of total catechol-
amine content) observed at the end of the stimulation period
if catecholamines were released concomitantly with scinde-
rin redistribution and F-actin disassembly. Moreover, simi-
lar time courses and amounts of catecholamine release dur-
ing the first minute of stimulation have also been previously
observed by us (Ctt6 et al., 1986) and other laboratories
(Baker et al., 1985; TerBush et al., 1988; Bittner and Holz,
1990). It can also be argued that the lag period observed was
due to a slow diffusion of catecholamines into the incubation
medium. This seems unlikely, since release experiments on
cultured cells eliminates physiological barriers such as capil-
lary endothelial walls, et cetera.
The rates of F-actin disassembly and scinderin redistribu-
tion during stimulation were found to be similar and subplas-
malemmal areas showing filamentous actin also showed cor-
tical scinderin. Immunocytochemical studies have also
shown that caldesmon (Burgoyne et al., 1986) and fodrin
(Perrin and Aunis, 1985) are preferentially localized in the
cortical region of the chromaffin cell. Moreover, in one of
these studies (Perrin and Aunis, 1985) a redistribution of
cortical fodrin antibody fluorescence was observed upon
nicotinic or high K ÷ stimulation. However, in this case the
time course of fodrin redistribution was much slower than
that described for scinderin in the present experiments. In
view of the observations described in this paper, it can also
be argued that scinderin shows a subplasmalemmal distribu-
tion because it is bound to filamentous actin. However, this
notion should be discarded since after removal of the stimu-
lus the rate of recovery of scinderin cortical fluorescence was
faster than that of rhodamine-phalloidin fluorescence. In other
words, during the post stimulation period, a significant num-
ber of cells displayed cortical continuance of scinderin fluo-
rescence in the presence of a fragmented ring of rhodamine-
phalloidin fluorescence. This would suggest that scinderin is
retained in the cortical region of the resting cell through its
binding to a site other than filamentous actin. Maekawa and
Sakai (1990) have shown the presence in chromattin cells of
a 74-kD actin filament-severing protein which binds to phos-
phatidylinositol and phosphatidylserine in a Ca2÷-dependent
manner. Although the possibility exists that scinderin (79.6
kD) and the 74 kD protein described above are the same pro-
tein, there is no evidence of these at the present time. More-
over, scinderin seems to be bound or retained in the cortical
region of the cell under resting conditions and during recov-
ery from secretion, conditions which are characterized by
low intraceUular Ca :÷ levels. Furthermore, muscarinic stim-
ulation does not release catecholamines (Wilson and Kirsh-
ner, 1977; Fisher et al., 1981), redistribute scinderin or pro-
duce actin filament disassembly. It is known that, in adrenal
chromaflin cells, nicotine and high K ÷ induce the entry and
a rise in cytosolic Ca 2+, which is necessary for catechol-
amine secretion (Douglas and Rubin, 1961; Douglas, 1968;
Cheek et al., 1989; Kim and Westhead, 1989; O'Sullivan et
al., 1989). Muscarine produces mobilization of Ca 2+ from
intracellular stores (Wilson and Kirshner, 1977; Kim and
Westhead, 1989) an effect which is independent of extracel-
lular Ca 2+ (Kao and Schneider, 1985) and is mediated by
inositol 1,4,5 triphosphate (Hughes and Putney, 1990). The
reduced and localized release of Ca 2+ induced by muscarine
is not enough to trigger catecholamine release and, as shown
in these studies, scinderin redistribution. Therefore, only
secretagogues that induce Ca 2+ entry are able to redistrib-
ute subplasmalemmal scinderin and produce the disassem-
bly of F-actin networks leaving cytoplasmic areas devoid of
these two proteins. We have previously demonstrated by low
shear viscometry that the concentrations of Ca 2÷ required
by scinderin to induce a fall in the viscosity of actin gels are
in the range of Ca 2÷ concentrations expected to be found in
the chromaftin cell cytoplasm as a result of cell stimulation
(Rodriguez Del Castillo et al., 1990). These observations
suggest that Ca 2÷ entry might regulate the actin filament-
severing activity of scinderin. One is tempted to speculate
that cell stimulation and Ca 2÷ entry bring about activation
of proteins such as scinderin with a consequent severing of
cortical actin filament networks. This should produce sub-
plasmalemmal areas of decreased viscosity and high secre-
tory granule mobility, allowing subsequent interaction of
granules with the plasma membrane. The experiments with
anti-D~H described here seem to indicate that exocytotic
pits are preferentially present in plasma membrane areas de-
void of F-actin. Bound D/$H is a chromaffin granule compo-
nent with a specific membrane topology (Joh and Hwang,
1982). No granule surface D/3H can be found and the en-
Vitale et at. Actin Disassembly and Scinderin Redistribution in Exocytosis
zyme has an intragranular domain recognized by the anti-
body. Therefore, when secretory vesicle membranes are in-
serted into plasma membranes during exocytosis, antigenic
D#H sites are exposed on the cell surface, allowing visual-
ization of plasma membrane exocytosis sites (Phillips et al.,
Activation of scinderin and actin filament disassembly
seem to precede exocytosis. However, to consider this as the
only important phenomenon in secretion would be an over-
simplified notion of what might be the fine regulation of exo-
cytosis in which intervention of other messengers and modu-
lators such as calmodulin (Kenigsberg and Trifar6, 1985),
cAMP (Cheek and Burgoyne, 1987), G-proteins (Matter et
al., 1989; Bader et al., 1989), polyphosphatidylinositol
breakdown (Janney and Stossel, 1987; Forsher, 1989), et
cetera, may occur.
We are grateful to Mrs. R. Tang for technical assistance and preparation
of chromaflin cell cultures and to Mrs. S. J. Dunn for typing the manu-
This work was supported by grants from the Medical Research Council
of Canada to J.-M. Trifar6.
Received for publication 26 October 1990 and in revised form 19 February
Aunis, D., B. Guerold, M.-F. Bader, and J. Cieselki-Treska. 1980. Im-
munochemical and biochemical demonstration of contractile proteins in
chromaffin cell culture. Neuroscience. 5:2261-2277.
Bader, M.-F., E. Georges, W. E. Mushynski, and J.-M. Trifar6. 1984.
Nearofilament proteins in cultured chromaflin cells. J. Neurochem. 43:
Bader, M.-F., J.-M. Trifar6, O. K. Langley, D. Thierst, and D. Aunis. 1986.
Secretory cell actin-binding proteins: identification ofa gelsolin-like protein
in chromafi~n cells. J. Cell Biol. 102:636-646.
Bader, M.-F., J.-M. Sontag, D. Thierst, and D. Aunis. 1989. A reassessment
of guanine nucleotide effects of catecholamine secretion from permeabilized
adrenal chromaffin cells. J. Biol. Chem. 264:16426-16434.
Baker, E. M., T. R. Sbeek, and R. D. Burgoyne. 1985. cyclic AMP inhibits
secretion from bovine adrenal chromaffin cells evoked by carbamylcholine
but not by high K ÷. Biochim. Biophys. Acta. 846:388-393.
Bernstein, B. W., and J. R. Bamburg. 1985. Reorganization of actin in depola-
rized synaptosomes. J. Neuroscience. 5:2565-2569.
Bittner, M. A., and R. W. Holz. 1990. Phorbol esters enhance exocytosis from
chromafl~n cells by two mechanisms. J. Neurochem. 54:205-210.
Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of
microgram quantities of protein utilizing the principle of protein-dye bind-
ing. Anal. Biochem. 72:248-254.
Burgoyne, R. D. and T. R. Cheek. 1987. Reorganization of peripheral actin
filaments as a prelude to exocytosis. Biosci. Rep. 7:281-288.
Burgoyne, R. D., T. R. Cheek, and K. M. Norman. 1986. Identification of a
secretory granule-binding protein as caldesmon. Nature (Lond.). 319:68-70.
Burgoyne, R. D., A. Morgan, and A. J. O'Sullivan. 1989. The control ofcy-
toskeletal actin and exocytosis in intact and permeabilized adrenal chro-
maflin cells: role of calcium and protein kinase C. Cell. Signalling. 1:
Chaponnier, C., P. A. Janmey, and H. L. Yin. 1986. The actin filament-
severing domain of plasma gelsolin. J. Cell Biol. 103:1473-1481.
Cheek, T. R., and R. D. Burgoyne. 1986. Nicotine-evoked disassembly of cor-
tical actin filaments in adrenal chromaitin cells. FEBS (Fed. Fur. Biochem.
Soc.) Lett. 207:110-114.
Cheek, T. R., and R. D. Burgoyne. 1987. cAMP inhibits both nicotine-induced
actin disassembly and catecholamine secretion from bovine adrenal chro-
maffin cells. J. Biol. Chem. 262:11663-11666.
Cheek, T. R., T. R. Jackson, A. J. O'Sullivan, R. B. Moreto, M. J. Berridge,
and R. D. Burgoyne. 1989. Simultaneous measurements of cytosolic Ca 2÷
and secretion in single bovine adrenal chromaflin cells by fluorescent imag-
ing of fura-2 in cocultured cell. J. Cell Biol. 109:1219-1227.
Ctt~, A., J.-P. Doucet and J.-M. Trifart. 1986. Phosphorylation and dephos-
phorylation of chromaffin cell proteins in response to stimulation. Neurosci-
Craig, S. W., and T. D. Pollard. 1982. Actin-binding proteins. Trends Bio-
chem. Sci. 7:88-92.
Doucet, J.-P., and J.-M. Trifar6. 1988. A discontinuous and highly porous so-
dium dodecyl sulfate-polyacrylamide slab gel system of high resolution.
Anal. Biochem. 168:265-271.
Douglas, W. W. 1968. Stimulus-secretion coupling: the concept and clues from
chromaflin and other cells. Br. J. Pharmacol. 34:451-474.
Douglas, W. W., and R. P. Rubin. 1961. The role of calcium in the secretory
response of adrenal medulla to acetylcholine. J. Physiol. (Lond.). 159:40-
Faultish, H., S. Zobeley, G. Rinnerthaler, and J. V. Small. 1988. Fluorescent
phallotoxins as probes for filamentous actin. J. Muscle Res. Cell Motil.
Fisher, S. K., R. W. Holz, and B. W. Agranoff. 1981. Muscarine receptors
in chromaliin cell cultures mediate enhanced phospholipid labelling but not
catecbolamine secretion. J. Neurochem. 37:491-497.
Forscber, P. 1989. Calcium and phosphoinositide control of cytoskeletal dy-
namics. Trends Neurol. Sci. 12:468-494.
Hughes, A. R., and J. W. Putney, Jr. 1990. Inositol phosphate formation and
its relationship to calcium signals. Environ. Health Persp. 84:141-147.
Janney, P. A., and T. P. Stossel. 1987. Modulation of gelsolin function by
phosphatidylinositol 4,5 biphosphate. Nature (Lond.). 352:362-364.
Joh, T. H., and O. Hwang. 1987. Dopamine-B-hydroxylase: biochemistry and
molecular biology. R. G. Johnson, Jr., editor. Ann. NY Acad. Sci. 493:
Kao, L.-S., and A. S. Schneider. 1985. Muscarine receptors on bovine
chromaflin cells mediate a rise in cytosolic calcium that is independent of ex-
traeellular calcium. J. Biol. Chem. 260:2019-2022.
Kenigsberg, R. L., and J.-M. Trifar6. 1980. Presence of high alt~nity uptake
system for catecholamines in cultured bovine adrenal chromaffin cells. Neu-
Kenigsberg, R. L., and J.-M. Trifart. 1985. Microinjections ofcalmodulin an-
tibodies into cultured chromaffin cells blocks catecholamine release in re-
sponse to stimulation. Neuroscience. 14:335-337.
Kim, K.-T., and E. W. Westhead. 1989. Cellular responses to Ca 2÷ from ex-
tracellular and intracellular sources are different as shown by simultaneous
measurements of cytosolic Ca :+ and secretion from bovine chromaflin cells.
Proc. Natl. Acad. Sci. USA. 86:9881-9885.
Koefer, A., P. E. R. Tatham, and B. D. Gompers. 1990. Changes in the state
of actin during the exocytotic reaction in mast cells. J. Cell Biol. 111:919-
Lee, R. W. H., and J.-M. Trifart. 1981. Characterization of anti-actin antibod-
ies and their use in immunocytocbemical studies on the localization of actin
in adrenal chromaflin cells. Neuroscience. 6:2087-2108.
Maekawa, S., and S. Sakai. 1990. Inhibition of actin regulatory activity of the
74 kDa protein from bovine adrenal medulla (Adseverin) by some phospho-
lipids. J. Biol. Chem. 265:10940-10942.
Maekawa, S., M. Toriyama, S.-I. Hisanaga, N. Yonezawa, S. Endo, N.
Hirokawa, and H. Sakai. 1989. Purification and characterization of a Ca :+-
dependent actin filament severing protein from bovine adrenal medulla. J.
Biol. Chem. 264:7458-7465.
Matter, K., F. Dreyer, and K. Aktories. 1989. Actin involvement in exocytosis
from PC 12 cells. Studies on the influence of botulinun C2 toxin on stimulated
noradrenaline release. J. Neurochem. 52:370-376.
O'Sullivan, A. J., T. R. Cheek, R. B. Moreton, M. J. Berridge, and R. D. Bur-
goyne. 1989. Localization and heterogeneity of agonist-induced changes in
cytosolic calcium concentration in single bovine adrenal chromal~n cell from
video imaging of fura-2. EMBO (Eur. Mol. Biol. Organ.) J. 8:401-411.
Pert'in, D., and D. Aunis. 1985. Reorganization of fodrin induced by stimula-
tion in secretory cells. Nature (Lond.). 315:589-591.
Phillips, J. M., K. Burridge, S. O. Wilson, and N. Kirshner. 1983. Visualiza-
tion of the exocytosis/endocytosis secretory cycle in cultured adrenal
chromaffin cells. J. Cell Biol. 97:1906-1917.
Rodrfguez Del Castillo, A., S. Lemaire, L. Tchakarov, M. Jeyapragasan, J.-P.
Doucet, M. L. Vitale, and J.-M. Trifart. 1990. Chromaflin cell scinderin:
a novel calcium-dependent actin-filament severing protein. EMBO (Eur.
Mol. Biol. Organ.)J. 9:43-52.
Smith, A. D. 1968. The storage of hormones. Biochem. J. 109:17-19.
Stossel, T. P., C. Chaponnier, R. Ezzell, J. H. Hartwig, P. A. Janney, D. J.
Kwiatowski, S. E. Lind, D. B. Smith, F. S. Southwick, and H. L. Yin. 1985.
Non-muscle actin binding proteins. Annu. Rev. Cell Biol. 1:353--402.
Tchakarov, L., M. L. Vitale, M. Jeyapragasan, A. Rodrfguez Del Castilto, and
J.-M. Trifart. 1990. Expression ofscinderin, an actin filament-severing pro-
tein, in different tissues. FEBS (Fed. Eur. Biochem. Soc.) Len. 268:209-
TerBush, D. R., M. A. Bittner, and R. W. Holz. 1988. Ca 2+ influx causes
rapid translocation of PKC to membranes. J. Biol. Chem. 263:18873-18879.
Towbin, H., T. Staebelin, and J. Gordon. 1979. Electrophoretic transfer of pro-
teins from acrylamide gels to nitrocellulose sheets: procedure and some ap-
plications. Proc. Natl. Acad. Sci. USA. 76:4350-4354.
Trifart, J.-M. 1977. Common mechanisms of hormone secretion. Annu. Rev.
Pharmacol. Toxicol. 17:27-47.
Trifart, J.-M. 1990. The dynamics of chromaflin cell (CC) cytoskeleton during
neurohormone secretion. The 18th Aharon Katzir-Katchalsky Conference.
From Receptors to Cell Response. Physiology and Pharmacology. Chieti,
Italy, September, 1990. 40.
Trifart, J.-M., and R. W. H. Lee. 1980. Morphological characteristics and
stimulation-secretion coupling in bovine adrenal chromaffin cell cultures.
Trifart, J.-M., and G. W. Bourne. 1981. Differential effects of concanavalin
A on acetylcholine and potassium-evoked release of catecholamines from
The Journal of Cell Biology, Volume 113, 1991 1066
cultured chromatfin ceils. Neuroscience. 6:1823-1833.
Trifar6, J.-M., A. C. Duerr, and J. E. B. Pinto. 1976. Membranes of the adre-
nal medulla: a comparison between the membranes of the Golgi apparatus
and chromaffin granules. Mol. PharmacoL 12:536-545.
Trifar6, J.-M., R. W. H. Lee, R. L. Kanigsberg, and A. C6t6. 1982. Contrac-
tile proteins and chromaflin cell function. Adv. Biosci. V:151-158.
Trifar6, J.-M., R. L. Kenigsberg, A. C6t6, R. W. H. lee, and T. Hikita. 1984.
Adrenal paraneuron contractile proteins and stimulus-secretion coupling.
Can. J. Physiol. Pharmacol. 62:493-501.
Trifar6, J.-M., M.-F. Bader, and J.-P. Doucet. 1985. Chromaffin cell
cytoskeleton: its possible role in secretion. Can. J. Biochem. Cell Biol.
Trifar6, J.-M., M. L. Novas, S. Fournier, and A. Rodrfguez Del Castillo.
1989. Cellular and molecular mechanisms in hormone and neurotransmitter
secretion. In Recent Advances in Pharmacology and Therapeutics. M.
Velasco, A. Israel, E. Romero, and H. Silva editors. Elsevier Science Pub-
Viveros, O. H. 1974. Mechanisms of secretion of catecholamines from the
adrenal medulla. In Handbook of Endocrinology. Part 1. H. Blasko, G.
Sayers, and A. D. Smith, editors. American Physiology Society. Waverly
Press, Inc., Baltimore, MD. 389-426.
Wilson, S. P., and N. Kirshner. 1977. The acetylcholine receptor of the adrenal
medulla. J. Neurochem. 28:687-695.
Yin, H. L., and T. P. Stossel. 1979. Control of cytoplasmic actin gel-sol trans-
formation by gelsolin, a calcium dependent regulatory protein. Nature
Yin, H. L., J. H. Albrecht, and A. Fattoum. 1981. Identification of gelsolin,
a calcium dependent regulatory protein of actin-gel transformation and its
intracellular distribution in a variety of cells and tissues. J. Cell Biol.
Yin, H. L., K. Iida, and P. A. Janney. 1988. Identification of a poly-
phosphoinositide-modulated domain in gelsolin which binds to the sides of
actin filaments. J. Cell Biol. 106:805-812.
Vitale et al. Actin Disassembly and Scinderin Redistribution in Exocytosis