THE JOURNAL

OF

BIOLOGICAL CHEMISTRY

Vol. 277, No. 18, Issue of May 3, pp. 1595715961, 2002

Printed in U.S.A.

Sweeping Model of Dynamin Activity

VISUALIZATION OF COUPLING BETWEEN EXOCYTOSIS AND ENDOCYTOSIS UNDER AN EVANESCENT
WAVE MICROSCOPE WITH GREEN FLUORESCENT PROTEINS*
Received for publication, January 25, 2002, and in revised form, February 20, 2002
Published, JBC Papers in Press, February 20, 2002, DOI 10.1074/jbc.C200051200

Takashi Tsuboi, Susumu Terakawa**, Bethe A. Scalettar, Claire Fantus, John Roder,
and Andreas Jeromin
From the Laboratory of Cell Imaging, Photon Medical Research Center, Hamamatsu University School of Medicine,
1-20-1 Handayama, Hamamatsu 431-3192, Japan, Department of Physics, Lewis & Clark College, Portland, Oregon
97219, and Samuel Lunenfeld Research Institute, Mt. Sinai Hospital, SLRI-860, Toronto, Ontario M5G 1X5, Canada

Exocytosis and endocytosis are linked and regulated coordinately by a cascade of protein-protein interactions (1) to ensure
the highly complex spatial and temporal patterns of membrane
recycling. Previous studies focused mainly on the last step of
exocytosis and inferred the kinetics of endocytosis only indirectly (25). In the present study, using the green fluorescent
protein (GFP)1 technique, we have focused on the coupling of
exocytosis and endocytosis. We observed the vesicle-associated
* This study was supported by Grants-in-aid for Scientific Research
10557003 and 11794015 (to S. T.) and 8701 (to T. T.) from the Ministry
of Education, Science, Sport, and Culture of Japan, grants from the
Medical Research Council of Canada (to C. F., J. R., and A. J.) and
National Institutes of Health Grant GM61539 (to B. A. S.). The costs of
publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Recipient of Research Fellow of the Japan Society for the Promotion
of Science (JSPS).
** To whom correspondence should be addressed. Tel. and Fax: 8153-435-2092; E-mail: terakawa@hama-med.ac.jp.
1
The abbreviations used are: GFP, green fluorescent protein; EGFP,
enhanced green fluorescent protein; Syb, synaptobrevin.
This paper is available on line at http://www.jbc.org

membrane protein, which is also referred to as synaptobrevin

(Syb), and the vesicle-producing protein, dynamin, simultaneously under an evanescent field microscope (6 8). The fluorescence imaging restricted to the plasma membrane allowed
us to capture the exact moment and the site of exocytosis and
compare them with the foci of endocytotic activity. In many
cases, dynamin clusters appeared in the void space between the
sites of exocytotic responses, and then they moved around
continuously beneath the membrane, as if they were searching
or scanning for the proper site of membrane retrieval. Here,
based on these observations, we will propose a novel hypothesis, sweeping model of dynamin, for an efficient retrieval of
superfluous membranes by endocytosis.
EXPERIMENTAL PROCEDURES

Expression Vector and TransfectionA construct of Syb fused with

enhanced GFP (EGFP) was produced by subcloning the rat Syb cDNA
as a HindIII/BamHI fragment into pEGFP C1 (CLONTECH, Tokyo,
Japan). A construct of Syb fused with DsRed was made by shuffling the
Syb cDNA from EGFP fusion construct into DsRed C1 (CLONTECH).
Constructs of dynamin I-EGFP and dynamin K44A-EGFP (9, 10) were
provided by Dr. Patrizia Okamoto and Dr. Richard Vallee (University of
Massachusetts). To transfect pheochromocytoma cells (PC12), an expression vector containing (2 g/l) LipofectAMINE 2000 (Invitrogen,
Tokyo, Japan) solution was mixed with a culture medium, and the cells
were cultured in the mixture for 4 h. Then, the cells were washed with
the culture medium and stored in a fresh medium. Under this condition,
many cells (70%) were successfully transfected. The cells were used
for experiments in 24 48 h after transfection.
The culture medium was Dulbeccos modified Eagles medium supplemented with 10% fetal bovine serum (both from Invitrogen). For the
fluorescence imaging, a modified Krebs solution containing (in mM)
NaCl, 135; KCl, 5.0; CaCl2, 2.0; MgCl2, 1.0; Na-HEPES, 10; glucose, 10
(buffered at pH 7.3 by titration with NaOH) was used.
Imaging SystemsFor most fluorescence observations, we employed
a total internal reflection fluorescence microscope (TIRFM or evanescent field microscope) described previously by Tsuboi et al. (7). The
incident light for evanescent illumination was introduced from the
objective lens (NA 1.45, 60 magnification) installed on an inverted
microscope (IX70, Olympus, Tokyo, Japan). To observe the EGFP fluorescence image, we used a 473-nm laser (diode-pumped solid state, 30
milliwatts; Shimadzu, Tokyo, Japan) for the evanescent wave excitation
and a long pass filter (515 nm) for barrier. The laser beam was passed
through an electromagnetically driven shutter (Uniblitz; Vincent Associates, Rochester, NY). The shutter was opened synchronously with
camera exposure under control with a personal computer (Endeavor
Pro-400, Epson, Tokyo, Japan) running a MetaMorph software package
(Universal Imaging Co., West Chester, PA). Fluorescence images were
captured with a monochromatic ICCD camera (DAS-512, Imagista,
Tokyo, Japan) combined with an image intensifier (VS4 1845, VideoScope, Sterling, VA). The video images were contrast-enhanced with a
digital image processor (ARGUS-20, Hamamatsu Photonics,
Hamamatsu, Japan), then recorded on digital videotape (GV-D900,
Sony, Tokyo, Japan) continuously, and stored finally on a computer
hard disk.

15957

Downloaded from www.jbc.org at UNIVERSIDAD AUSTRAL DE CHILE, on March 18, 2010

Vesicle recycling through exocytosis and endocytosis

is mediated by a coordinated cascade of protein-protein
interactions. Previously, exocytosis and endocytosis
were studied separately so that the coupling between
them was understood only indirectly. We focused on the
coupling of these processes by observing the secretory
vesicle marker synaptobrevin and the endocytotic vesicle marker dynamin I tagged with green and red fluorescent proteins under an evanescent wave microscope
in pheochromocytoma cells. In control cells, many synaptobrevin-expressing vesicles were found as fluorescent spots near the plasma membrane. Upon electrical
stimulation, many of these vesicles showed an exocytotic response as a transient increase in fluorescence
intensity followed by their disappearance. In contrast,
fluorescent dynamin appeared as clusters increasing
slowly in number upon stimulation. The clusters of fluorescent dynamin moved around beneath the plasma
membrane for a significant distance. Simultaneous observations of green fluorescent dynamin and red fluorescent synaptobrevin indicated that more than 70% of
the exocytotic responses of synaptobrevin had no immediate dynamin counterpart at the same site. From these
findings it was concluded that dynamin-mediated recycling is not directly coupled to exocytosis but rather
completed by a scanning movement of dynamin for the
sites of invaginating membrane destined to endocytosis.

15958

Sweeping Model of Dynamin Activity

To observe fluorescence images of EGFP and DsRed simultaneously

under the evanescent wave illumination, we used a single laser line
(473 nm) for excitation and the W-view optical system (A-4313,
Hamamatsu Photonics) for emission, as described previously (6). All
images were reproduced from the hard disk or from the digital videotape using a personal computer (Power Macintosh G4/450, Apple Computer, Inc., Cupertino, CA).
Immunofluorescence StainingThe Syb-DsRed expressing PC-12
cells were fixed with 99.5% ethanol for 10 min at 4 C. After fixation,
the cells were incubated with monoclonal anti-synaptophysin antibody
(Roche Molecular Biochemicals, Tokyo, Japan) for about 1 h at room
temperature. Then the cells were incubated with fluorescein isothiocyanate-conjugated goat anti-mouse IgG (Southern Biotechnology Associates, Inc., Birmingham, AL) for 12 h at 4 C. After a final washing, the
cells were observed with a confocal microscope equipped with a microlens-attached Nipkow-disc scanner (CSU-10, Yokogawa Electric Co.,
Tokyo, Japan).
RESULTS AND DISCUSSION

Imaging of Vesicles Expressing Syb-EGFPTo assess the

exocytotic response of secretory vesicles, we observed the fluorescence of Syb-EGFP expressed in the vesicles under the evanescent field microscope. Many fluorescent spots of a quite
uniform size were distributed all over the bottom area of the
cell (Fig. 1a). These fluorescent spots showed short linear
movements (average velocity: 3.21 0.82 m/s, n 10 cells) in
various directions in control cells. The average velocity was
significantly decreased to 0.12 0.02 m/s (n 10 cells, p
0.05) when colchicine (10 M, 30 min) was applied to the cells.
However, no change was observed when ML-9, a myosin light
chain kinase inhibitor (20 M, 30 min), was applied (average
velocity: 2.94 0.64 m/s, n 10 cells).
When electrical stimulation (1-A current pulse, 1-ms duration) was applied with a micropipette attached to the cell (11),
many of the Syb-EGFP vesicles (76 4%, n 10 cells) abruptly
disappeared after their transient increase in fluorescence intensity (flash) with individual delays. During the flash response for about 30 ms, the image of the vesicle became diffusively larger (Fig. 1b, middle panel). After their abrupt

disappearance, the fluorescence could no longer be detected on

the plasma membrane (Fig. 1, d and e).
The abrupt disappearance of fluorescent vesicles indicated
their exocytotic responses. The flash could be ascribed to a
lateral diffusion of the Syb-EGFP from a vesicle membrane to
the plasma membrane in the exponentially rising evanescent
field (4 6). We did not detect a significant increase in fluorescence of the plasma membrane after stimulation. This could be
accounted for by lateral diffusion of the Syb-EGFP-carrying
vesicles also, because EGFP would be diluted at least 10-fold by
the surrounding plasma membranes after the exocytotic fusion.
A lack of increase in background fluorescence of the plasma
membrane even after many exocytotic responses suggested
that Syb-EGFP on the plasma membrane was efficiently internalized after exocytotic membrane fusion. Alternatively, SybEGFP transferred to the plasma membrane might be quenched
by some unknown mechanism or by enzymatic degradation for
recycling.
Imaging of Dynamin I-EGFPWe observed the dynamic
activity of dynamin I-EGFP near the plasma membrane of
PC12 cells by using the same evanescent wave microscope.
Against a faint background, dynamin I-EGFP in control cells
was visualized as punctuate fluorescent spots of variable sizes
(0.4 m). Hereafter, we will refer to the large fluorescent
spots as clusters. Electrical stimulation induced a marked increase in number of these clusters near the plasma membrane
(224 5% on average, n 10 cells; Fig. 2a). Stimulation
induced a significant increase in background fluorescence of
the plasma membrane as well (Fig. 2, a and e). The number of
clusters formed was larger in the high background areas. It is
possible that dynamin molecules first translocated from the
cytoplasm to the plasma membrane and then they assembled
into clusters in an initial stage of the membrane-related activity. Once the clusters appeared, they did not grow later. In
contrast to the rapid response of the vesicles expressing SybEGFP, the fluorescent clusters of dynamin I-EGFP appeared

Downloaded from www.jbc.org at UNIVERSIDAD AUSTRAL DE CHILE, on March 18, 2010

FIG. 1. Effect of electrical stimulation on Syb-EGFP-expressing vesicles in a single PC12 cell. a, image observed by evanescent wave
excitation before (0 s) and after (1 and 5 s) stimulation. b, sequential images of a single vesicle observed after electrical stimulation (figures indicate
time after stimulation in milliseconds). The second image shows a diffuse cloud of the Syb-EGFP and the third image a disappearance of the
fluorescent spot. c, evanescent wave (total internal reflection fluorescence) microscopic images of an EGFP fluorescence of a cell observed with a
color CCD camera without additional staining (left panel) and with a 5-min staining with 3 M acridine orange (middle panel). The yellow color
indicates co-localization of Syb-EGFP with acridine orange. A magnified view of the portion in a box in the middle panel is shown (right panel).
The arrowhead indicates a vesicle that was free from acridine orange (middle and right panels). The scale bar indicates 5 m (a and c in center
panel) and 1 m (b and c in right panel). d, fluorescence intensity measured in the center of three different vesicles. An arrow indicates the time
of electrical stimulation (single current pulse of 1-ms duration and 1 A). e, time course of the fluorescence intensity change of Syb-EGFP measured
in the whole part of a single cell. Stimulation was given at the time indicated by an arrow.

Sweeping Model of Dynamin Activity

15959

with a clear delay of 10 60 s after stimulation (Fig. 2d). The

time course of this endocytotic response estimated with dynamin I-EGFP was consistent with recent observations (12
14). All clusters and the background fluorescence soon faded to
restore the initial fluorescence level in the entire field of cell
observed. In individual clusters, the large increase in fluorescence of dynamin I-EGFP was followed by a sharp decrease in
fluorescence in 10 20 s (Fig. 2d) without a detectable change in
shape. The decrease in fluorescence reflected a movement of
clusters away from the evanescent field, possibly showing dynamin activity of pinching endocytotic vesicles from the plasma
membrane. There were clusters linked loosely to others, forming a super cluster in the shape of a ribbon (Fig. 2c) or sometimes a quite large ring (Fig. 3c).
Many clusters showed a continuous zig-zag movement with a
maximum step length of about 1 m (per s) without altering
their fluorescence intensity, suggesting their lateral movement
under the plasma membrane (Fig. 3a). Some drifted laterally
for final distances longer than 3 m before they faded (Fig. 3b).
Almost the entire area of the cell under observation was covered by traces of some clusters. Some ribbons of dynamin
moved sideways for a few micrometers. The ring structures
grew in diameter to scan a large area of the plasma membrane
(Fig. 3c), enlarging like a wave made by a stone thrown into a
pond. More precisely, rings fragmented into several arcs during
enlargement, keeping their shape and intensity for a certain
traveling distance. These various modes of gliding suggested
that dynamin clusters scan the membrane to meet the invaginating pits destined to endocytosis. We observed such responses reproducibly in 5 of 7 cells tested in a series.
We also observed the dynamics of mutant type dynamin
I-EGFP (Fig. 2b). A marked difference between the wild type
and mutant dynamin was noticed in the rates of disappearance
from the plasma membrane after their stimulation-induced
appearance. The mutant dynamin did not disappear for more
than 10 min of the observation period (Fig. 2e, bottom panel).
When compared before stimulation, mutant clusters were always larger in number than wild type clusters. The ribbons or
the rings of clusters did not appear after stimulation in mutant
cells, although individual clusters of a small size were observed
similarly at a high density moving randomly in lateral directions. All these properties of mutant dynamin are probably
pertinent to the impeded endocytosis.

In normal cells, the appearance of fluorescent clusters of

dynamin I-EGFP after electrical stimulation was blocked significantly (to 124 5%, n 10 cells) by application of an
anti-mitotic agent, colchicine (10 M, 30 min). However, no
blockade was observed when an inhibitor of myosin light chain
kinase, ML-9 (20 M, 30 min), was applied (211 11%, n 10
cells).
Simultaneous Imaging of Syb-DsRed and Dynamin
I-EGFPA fate of Syb-DsRed-expressing vesicles after the exocytotic event was examined in the cell co-expressing dynamin
I-EGFP, using a dual window (W-view) evanescent wave microscope. Syb-DsRed was monitored in the red-filtered window
and dynamin I-EGFP in the green-filtered window. In the
resting cell, most of vesicular images appeared only in the red
window (Fig. 4a, top panels). When a depolarizing pulse was
applied to the cell, many of Syb-DsRed-expressing vesicles in
this window disappeared in a period of several seconds. Instead, many of the fluorescent spots (clusters) of dynamin IEGFP appeared in the green window (Fig. 4a, bottom panels).
The decrease in fluorescence intensity of Syb-DsRed ended
earlier than the onset of the increase in intensity of dynamin
I-EGFP fluorescence. We obtained such reproducible results of
simultaneous imaging in 7 of 12 cells we tested. Failure was
mainly due to a low frequency of responses.
We next examined the coincidence of the dynamin I-EGFP
response with the Syb-DsRed response by overlaying the two
images of pseudocolor representing both responses separately
(Fig. 4b). Among all disappearing responses of Syb-DsRed fluorescence, 27 1% (n 8 cells) had the appearance of dynamin
I-EGFP in exactly the same site (Fig. 4b, arrows). The SybDsRed vesicles that showed such successive responses were
mostly large in diameter (0.43 0.1 m, n 8 cells). Many of
such large vesicles were also stained with acridine orange (data
not shown). These results suggested that large dense core granules are more likely to attract dynamin I at active sites of
endocytosis. The dual window evanescent wave microscopy
described here will allow one to investigate the modes of different types of secretory vesicles in further detail.
Syb-DsRed was present exclusively on secretory vesicles in
PC-12 cells as judged by comparison with the distribution of
another vesicle protein, synaptophysin (Fig. 4c). Co-localization
averaged 81 5% (n 3 cells), minimizing the possibility that
the DsRed-conjugated protein was mistargeted. We calculated

Downloaded from www.jbc.org at UNIVERSIDAD AUSTRAL DE CHILE, on March 18, 2010

FIG. 2. Effect of electrical stimulation on dynamin I-EGFP in a single PC12 cell. a, image obtained by evanescent wave excitation in a
single PC12 cell expressing wild-type dynamin I-EGFP observed before (0 s) and after stimulation (10 and 60 s). b, single PC12 cell expressing
mutant type dynamin I-EGFP observed before (0 s) and after stimulation (10 and 60 s). c, sequential images of wild-type dynamin I-EGFP
fluorescence observed after electrical stimulation. Scale bars represent 5 m (in a and b) and 1 m (in c). d, time courses of the fluorescence change
measured in the center of dynamin I-EGFP spots, which were relatively stable in position. The arrow indicates the time of electrical stimulation.
e, time courses of the fluorescence intensity of wild type dynamin I-EGFP (closed circle) and mutant type dynamin I-EGFP (open circle) measured
in the whole part of a cell shown in a and b. The arrow indicates the time of electrical stimulation.

15960

Sweeping Model of Dynamin Activity

FIG. 4. Dual imaging of dynamin IEGFP- and Syb-DsRed-expressing

vesicles. a, sequential images of SybDsRed (top panels) and dynamin I-EGFP
(bottom panels) fluorescence observed simultaneously in a single cell. b, left panel,
the yellow color indicates co-localization
of dynamin I-EGFP (green) and SybDsRed (red). Time points of Syb-DsRed
and dynamin I-EGFP images were 0 and
10 s after electrical stimulation, respectively. b, right panel, high magnification
view of the boxed portion in the left panel.
Arrows indicate co-localization (yellow) of
dynamin I-EGFP (green) and Syb-DsRed
(red). Arrowheads indicate vesicles free
from Syb-DsRed. c, distribution of synaptophysin marked by green fluorescence
(left panel) and of Syb-DsRed (center
panel) in a cell, showing their co-localization in secretory vesicles. An overlap of
the two colors is shown in yellow (right
panel). The scale bar indicates 5 (a and b,
left panel, and c) and 2 m (b, right
panel).

the diffusion coefficient of Syb-EGFP- and Syb-DsRed-expressing vesicles to be 2.5 0.5 1010 cm2/s (n 8 cells) and 2.6
0.2 1010 cm2/s (n 8 cells), respectively. No significant
difference was observed indicating that Syb-EGFP and SybDsRed were targeted similarly in a non-aggregated form and
that fusion of Syb to DsRed did not alter its targeting.

Appearance and disappearance of some dynamin clusters

was assessed by measuring the fluorescence intensity at the
exact site where an exocytotic response of a Syb-DsRed vesicle
occurred (Fig. 5a). When the cell was stimulated, the fluorescent cluster of dynamin I appeared with an approximately 30-s
delay after an exocytotic response (Fig. 5c). After a certain

Downloaded from www.jbc.org at UNIVERSIDAD AUSTRAL DE CHILE, on March 18, 2010

FIG. 3. Movement of dynamin clusters during the membrane-associated

activity. a, sequence of video frames
showing random movements of several
clusters appearing after electrical stimulation. The stimulation was applied immediately after 0 s. The brightest cluster
was associated with the membrane from
the beginning (before the stimulation).
This cluster and others, which appeared
after electrical stimulation disappeared
from the plasma membrane 18 s later. b,
traces of the movement of 8 of 150 200
clusters transiently appearing beneath
the membrane in a cell. The shape of the
cell is indicated by solid curves (outline)
and dashed lines (frame). c, sweeping activity of a dynamin ring. A sequence of
video frames was taken after electrical
stimulation at time 0. The figure in each
frame indicates the time in seconds (a and
c). The scale bar indicates 1 m in a and c
and 5 m in b.

Sweeping Model of Dynamin Activity

15961

FIG. 5. Dynamics of dynamin

I-EGFP- and Syb-DsRed-expressing
single vesicles. a, dual evanescent wave
image of a dynamin I-EGFP fluorescence
(left) and Syb-DsRed fluorescence (right)
observed simultaneously before stimulation. b, sequential images of a single dynamin I-EGFP and Syb-DsRed vesicle
(circle 1 in a) observed after electrical
stimulation. c, time courses of the fluorescence intensity of dynamin I-EGFP
(closed circle) and Syb-DsRed (open circle)
measured in the center of vesicles in the
circles shown in a. The arrows indicate
the time of electrical stimulation. The
scale bars represent 5 (a) and 1 m (b).

coated pits are formed away from sites of exocytotic response

marked by Syb-DsRed, and the pits are swept up by dynamin
clusters. Some secretory vesicles expressing phogrin (phosphatase on the granule of insulinoma cells) remain in a vesicular
shape even after a complete exocytotic response (15), indicating
the kiss and glide of vesicles (6). Such a gliding of empty
vesicles together with the sweeping activity of dynamin would
bring about a frequent collision, thus facilitating the membrane retrieval process. The present study provides evidence
that membrane recycling is mediated by dual activities of dynamin, i.e. sweeping-up and pinching-out the pits under the
plasma membrane.
AcknowledgementsWe thank Drs. P. Okamoto and R. B. Vallee for
providing the dynamin I-EGFP construct.
REFERENCES
1. Bottomlet, M. J., Surdo, P. L., and Driscoill, P. C. (1999) Curr. Biol. 9,
R301R304
2. Takei, K., Slepnev, V. I., Haucke, V., and De Camilli, P. (1999) Nat. Cell Biol.
1, 3339
3. Steyer, J. A., Horstmann, H., and Almers, W. (1997) Nature 388, 474 478
4. Schmoranzer, J., Goulian, M., Axelrod, D., and Simon, S. M. (2000) J. Cell Biol.
149, 2332
5. Toomre, D., Steyer, J. A., Keller, P., Almers, W., and Simons, K. (2000) Cell
Biol. 49, 33 40
6. Tsuboi, T., Zhao, C., Terakawa, S., and Rutter, G. A. (2000) Curr. Biol. 10,
13071310
7. Tsuboi, T., Kikuta, T., Warashina, A., and Terakawa, S. (2001) Biochem.
Biophys. Res. Commun. 282, 621 628
8. Tokunaga, M., Kitamura, K., Saito, K., Iwane, A. H., and Yanagida, T. (1997)
Biochem. Biophys. Res. Commun. 235, 4753
9. Obar, R. A., Collins, C. A., Hammarback, J. A., Shpetner, H. S., and Vallee,
R. B. (1990) Nature 347, 256 261
10. Vasquez, R. J., Okamoto, P., Wheatley, S., Wang, Y. L., Maxfield, F., and
Vallee, R. B. (1998) Mol. Biol. Cell 9, 196 (abstr.)
11. Manivannan, S., and Terakawa, S. (1994) J. Neurosci. 14, 59175928
12. Sankaranarayanan, S., and Ryan, T. A. (2001) Nat. Neurosci. 4, 129 136
13. Sankaranarayanan, S., and Ryan, T. A. (2000) Nat. Cell Biol. 2, 197204
14. Li, Z., and Murthy, V. N. (2001) Neuron 31, 593 605
15. Pouli, A. E., Emmanouilidou, E., Zhao, C., Wasmeier, C., Hutton, J. C., and
Rutter, G. A. (1998) Biochem. J. 333, 193199

Downloaded from www.jbc.org at UNIVERSIDAD AUSTRAL DE CHILE, on March 18, 2010

period, the dynamin cluster disappeared in 30 s. Because the

stable period and disappearing period of dynamin fluorescence
were clearly distinguishable, the disappearance was judged not
to be a result of photobleaching. An enhanced photobleaching of
the dynamin fluorescence with a stronger illumination led to
an exponential fading and not to a stepwise fading. The fluorescence intensity of a cluster was more than a dozenfold
higher than the background intensity of dynamin fluorescence.
These findings suggested that many dynamin molecules were
involved in a single cluster.
In conclusion, using an evanescent wave microscope, we
demonstrated here dynamics of exocytosis and endocytosis of
secretory vesicles in PC12 cells. The fluorescent probes, SybDsRed and dynamin I-EGFP, made it possible to distinguish
the exocytotic process from the endocytotic process and to observe the sequence of two responses in the same cell. Thus the
present study provided insight into the recycling process of
exocytosis and endocytosis in a neuronal-like cell.
The increased fluorescence of dynamin I-EGFP in stimulated
cells (Fig. 3) suggests a recruitment of dynamin I molecules
from the cytoplasm to the plasma membrane for promotion of
the membrane recycling. The increased number in clusters of
dynamin I near the plasma membrane in stimulated cells (Fig.
4) suggests a Ca2-dependent enhancement of the endocytotic
activity. The formation of clusters of dynamin may be a microtubule-related response, as it was affected by colchicine at low
concentration.
Only a portion (27%) of the exocytotic responses showed a
corresponding formation of the dynamin I cluster coupled or
targeted directly to the membrane-fusion site. The rest of the
exocytotic responses probably are coupled to a membrane retrieval process at some distance. Our important observation is
that dynamin I clusters sweep the plasma membrane. Taken
together, it is highly likely that the major fraction of clathrin-