Am J Physiol Cell Physiol 289: C1209 –C1219, 2005.
First published July 6, 2005; doi:10.1152/ajpcell.00159.2005.
Spatiotemporal analysis of exocytosis in mouse parotid acinar cells
Ying Chen,1 Jennifer D. Warner,1 David I. Yule,2 and David R. Giovannucci1
1
Department of Neurosciences, Medical College of Ohio, Toledo, Ohio; and 2Department
of Pharmacology and Physiology, University of Rochester Medical Center, Rochester, New York
Submitted 5 April 2005; accepted in final form 29 June 2005
THE COMPLEX AND HIGHLY CONTROLLED process governing the
exocytotic release of cargo molecules from the secretory granules of exocrine, endocrine, or neuronal cells involves the
interaction, fusion, and dynamic reorganization of internal and
surface membranes (12, 24). Although there is substantial
insight into the molecular mechanisms directing the exocytosis
of neurotransmitter or hormone (13, 25), there is currently an
insufficient understanding of the controlled secretion of protein
from the exocrine cells of the digestive system. A notable
exception has been the use of electrophysiological and optical
methods to probe Ca2⫹ handling and the exocytosis of digestive enzyme from isolated or small clusters of pancreatic acinar
cells (2, 6, 8, 16 –18, 30, 32, 33). In a recent study, however,
we demonstrated (7) that although apparently morphologically
and functionally similar, mouse pancreatic and parotid acinar
cells exhibit important functional divergence in their Ca2⫹
signaling properties, and thus pancreatic cells alone may not be
universally applicable for understanding other exocrine cells.
For example, in parotid and pancreatic acinar cells, fluid and
protein secretion is regulated by both parasympathetic and
sympathetic input. In the parotid, cholinergic or ␤-adrenergic
receptor activation evokes rises in cytosolic Ca2⫹ concentration ([Ca2⫹]i) and CAMP concentration, respectively, and both
messengers can act as intracellular messengers to induce amylase release (34). A primary difference between parotid and
pancreatic acinar cells is that robust exocytosis in the parotid
can be stimulated by increased cytosolic levels of cAMP.
According to the current rubric of excitation-secretion coupling, this secretory activity is believed to occur without a
change in [Ca2⫹]i. Moreover, biosynthetic labeling analysis
has shown that the Ca2⫹- or cAMP-mobilizing agonists carbachol (CCh) and isoproterenol (Iso) induced trafficking of
␣-amylase-containing vesicles in parotid acini via a process
that involves both major and minor regulated secretory pathways (3, 5). Indeed, a number of candidate proteins for controlling protein secretion have been identified in parotid acinar
cells by molecular, protein biochemical, and immunocytochemical methods (3, 9, 10). However, the cellular and molecular mechanisms by which Ca2⫹ and cAMP mediate exocytosis in the parotid have not been clarified.
In the present study we sought to address this gap in our
understanding by using time-differential imaging analysis in
combination with highly time-resolved membrane capacitance
(Cm) measurements to image the secretory activity of single
mouse parotid acinar cells in real time after agonist-induced or
direct photolytic elevation of Ca2⫹. Moreover, optical methods
were successfully applied to a parotid gland organotypic slice
preparation. Our observations revealed that there are two
apparent Ca2⫹-dependent vesicle pools. After moderate elevation of Ca2⫹, one pool was rapidly induced to fuse with the
plasma membrane. Although imaged with time-resolved Cm
methods, single exocytotic events could not be resolved either
electrically or optically. The other vesicle pool underwent
exocytosis after a delay, was resolvable by both electrophysiological and optical methods, and most likely corresponded to
the individual fusion events of dense core zymogen-containing
secretory granules (ZSGs). ZSG fusion events produced persistent postfusion structures, were often found to coincide at
similar subcellular sites or “hot spots,” and produced a spatial
pattern of exocytotic activity that progressed outward from
apical or lateral aspects of the cell toward the cell interior when
robustly stimulated. These observations suggest that there are
Address for reprint requests and other correspondence: D. R. Giovannucci,
Dept. of Neurosciences, Medical College of Ohio, Toledo, OH 43614 (e-mail:
[email protected]).
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
membrane capacitance; differential imaging; zymogen; gland slice;
exocrine cells
http://www.ajpcell.org
0363-6143/05 $8.00 Copyright © 2005 the American Physiological Society
C1209
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
Chen, Ying, Jennifer D. Warner, David I. Yule, and David R.
Giovannucci. Spatiotemporal analysis of exocytosis in mouse parotid
acinar cells. Am J Physiol Cell Physiol 289: C1209 –C1219, 2005.
First published July 6, 2005; doi:10.1152/ajpcell.00159.2005.—Exocrine cells of the digestive system are specialized to secrete protein
and fluid in response to neuronal and/or hormonal input. Although
morphologically similar, parotid and pancreatic acinar cells exhibit
important functional divergence in Ca2⫹ signaling properties. To
address whether there are fundamental differences in exocytotic
release of digestive enzyme from exocrine cells of salivary gland
versus pancreas, we applied electrophysiological and optical methods
to investigate spatial and temporal characteristics of zymogen-containing secretory granule fusion at the single-acinar cell level by direct
or agonist-induced Ca2⫹ and cAMP elevation. Temporally resolved
membrane capacitance measurements revealed that two apparent
phases of exocytosis were induced by Ca2⫹ elevation: a rapidly
activated initial phase that could not be resolved as individual fusion
events and a second phase that was activated after a delay, increased
in a staircaselike fashion, was augmented by cAMP elevation, and
likely reflected both sequential compound and multivesicular fusion
of zymogen-containing granules. Optical measurements of exocytosis
with time-differential imaging analysis revealed that zymogen granule
fusion was induced after a minimum delay of ⬃200 ms, occurred
initially at apical and basolateral borders of acinar cells, and under
strong stimulation proceeded from apical pole to deeper regions of the
cell interior. Zymogen granule fusions appeared to coordinate subsequent fusions and produced persistent structures that generally lasted
several minutes. In addition, parotid gland slices were used to assess
secretory dynamics in a more physiological context. Parotid acinar
cells were shown to exhibit both similar and divergent properties
compared with the better-studied pancreatic acinar cell regarding
spatial organization and kinetics of exocytotic fusion of zymogen
granules.
C1210
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
multiple classes of secretory vesicles in parotid acinar cells that
differ in their size, Ca2⫹ sensitivity, kinetics of exocytosis, and
membrane retrieval mechanisms. Thus exocytotic activity in
parotid acinar cells exhibits characteristics both similar to and
divergent from those of the better-studied pancreatic acini.
MATERIALS AND METHODS
AJP-Cell Physiol • VOL
RESULTS
Exocytosis induced by cAMP and Ca2⫹ photorelease. In
pancreatic acinar cells, Cm measurements have been used to
demonstrate that application of a cholinergic agonist or artificial Ca2⫹ elevation can induce changes in cell surface area (8,
15, 16, 22). In the current study we determined whether this
technique could be used to monitor membrane surface dynamics of parotid acinar cells. To achieve this, we patch-clamped
acutely isolated parotid cells that largely retained polar distribution of ZSGs and monitored changes in Cm induced by
cAMP elevation. In contrast to the rapid, smooth rises in Cm
evoked by Ca2⫹ elevation in pancreatic acinar cells that we and
others have previously reported, cAMP-raising agonists produced staircaselike changes in Cm that persisted over several
minutes of treatment. As shown in Fig. 1A, local bath application of 10 ␮M Iso, a ␤-adrenergic receptor agonist previously shown to effectively induce amylase release in the
parotid, induced stepwise increases in Cm that we interpreted to
reflect the exocytotic fusion of individual or multiple ZSGs.
Similar results were obtained whether 10 ␮M forskolin or 1
mM dibutyryl cAMP was used to elevate cAMP levels. As
shown in Fig. 1B, individual events in cells treated with
cAMP-raising agonists ranged in magnitude from 5 to 270 fF.
The majority of events were clustered around 20, 40, or 60 fF
(20 fF was estimated to be the amount of membrane equivalent
to the addition of 1 ZSG, given that the average diameter of
these organelles is ⬃0.8 ␮m). Correspondingly, transmitted
light images of the patch-clamped acinar cells before and after
Iso treatment for several minutes revealed an apparent reduction in the number of phase-dark ZSGs from the cells’ apical
regions. The average cumulative rise in Cm over an ⬃3-min
interval after cAMP elevation was 351 ⫾ 190 fF (n ⫽ 9),
corresponding to the exocytosis of ⬃18 ZSGs. As Cm measurements are a net measure of cell surface dynamics, the
staircaselike profile of the Cm change indicated that little
apparent endocytosis was activated over the time that the cells
were monitored. In contrast to the Cm responses evoked by 10
289 • NOVEMBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
Preparation of mouse parotid gland acinar cells and salivary gland
slices. Animal study protocols were approved by the Medical College
of Ohio Institutional Animal Care and Use Committee. Single acinar
cells or small groups of acinar cells were isolated from parotid glands
that were freshly dissected from 18- to 25-g NIH Swiss Webster mice
(Charles River). The glands were minced with fine scissors and
subjected to stepwise digestion in minimum essential medium (Eagle’s Spinner modification) containing either 1% BSA, 0.01% trypsin,
and 0.5 mM EDTA (step 1) or 1% BSA, collagenase P (0.04 mg/ml)
(Boehringer Mannheim), and protease (bacterial) type XIV (1 mg/ml;
Sigma, St. Louis, MO) (step 2). Digestion was performed in a water
bath shaking at 1 Hz for 10 min (step 1) and twice for 20 min each
(step 2) at 37°C under continuous 95% O2-5% CO2. After gentle
trituration, single acinar cells were collected by centrifugation at 70 g,
resuspended in BSA-free basal Eagle medium supplemented with 2
mM glutamine and penicillin-streptomycin, filtered through 80-␮m
nylon mesh, and attached onto poly-L-lysine-coated glass coverslips.
Cells were maintained in DMEM in a tissue culture incubator and
used within 1– 4 h after plating.
For slice preparation, 2 ml of a 3% (wt/vol) low-temperature
gelling point agarose solution (Sigma type VIIA) was prepared in
physiological salt solution (PSS) that contained (in mM) 140 NaCl, 10
HEPES-NaOH, 4.7 KCl, 1.13 MgCl2, 1 CaCl2, and 10 D-glucose, pH
7.3, and warmed to 90°C. The agarose solution was then maintained
at ⬃35°C, and isolated lobes, or in some cases the entire parotid
glands, were embedded. The tissue-agarose mixture was then gelled at
4 – 8°C, and the block was trimmed so that minimal agarose remained
around the tissue. The block was then glued to the tissue stand of a
vibratome, and the chamber was filled with ice-cold nominal-Ca2⫹
PSS. Parotid gland tissue prepared by this method was cut into 40- to
50-␮m-thick sections. The slices were collected and kept at 25°C until
recording. Experiments monitoring Ca2⫹ or secretory dynamics in
isolated cells or slices were performed with PSS at ⬃25°C 1– 4 h after
slice preparation.
Time-resolved Cm measurements. Cm measurements were performed with an Axopatch 200B patch-clamp amplifier (Axon Instruments, Foster, CA), an ITC-16 digital interface (Instrutech, Port
Washington, NY), and an Apple G4 computer running IGOR Pro
(Wavemetrics, Lake Oswego, OR) and a software-based phase-sensitive detector (Pulse Control software, Dr. Richard Bookman, University of Miami Medical School, Coral Gables, FL) as previously
described (8). The standard intracellular recording/caged-Ca2⫹ solution contained (in mM) 110 CsMeSO4, 20 CsCl, 10 HEPES-Tris, 10
NP-EGTA, 5 CaCl2, 3.2 MgCl2, 2 Tris-ATP, 0.2 GTP, and 0.075
Oregon Green BAPTA-5N, pH 7.2.
Flash photolysis, digital imaging, and time-differential imaging
analysis. Caged compounds were photoconverted with a pulsed xenon
arc flash lamp system (TILL Photonics, Eugene, OR) ported to a
Nikon TE2000 microscope equipped with differential interference
contrast (DIC) optics through a fiber-optic guide and epifluorescence
condenser. High-intensity UV light flashes (0.1–1 ms) were focused
onto the image plane with a DM400 dichroic mirror and a Nikon
SuperFluor ⫻40 oil-immersion objective. Fluorescence or transmitted
light images were obtained with a TILL Photonics monochrometerbased high-speed digital imaging system. For measurement of [Ca2⫹]i
cells were illuminated with 340-, 380-, or 488-nm light and fluorescence was collected through a 525 ⫾ 25-nm band-pass filter (Chroma
Technologies, Brattleboro, VT). The numerical apertures of the objective and condenser for time-differential imaging analysis were 1.3
and 0.52, respectively. Transmitted light optical resolution was estimated to be ⬃350 nm. For experiments in which alternating brightfield and fluorescent images were collected, a high-speed Uniblitz
VS35 optical shutter (Vincent Associates, Rochester, NY) was placed
in the tungsten lamp illumination path. Pairs of brightfield and
fluorescence images (30- and 50-ms exposure, respectively) were
obtained at 5 Hz. Time-differential images were generated by subtraction of each brightfield frame by its preceding frame. Timedifferential imaging has been used and validated previously in pancreatic acinar cells and other cell types as a method of visualizing
individual secretory granule fusions (2). Parotid acinar cells were
amenable for this approach, likely because of the large size and
phase-dark core of their ZSGs. Exocytotic fusion created a change in
optical density that was detected as a spot in the difference image.
Average spot diameter after calibration with 1-␮m beads (Molecular
Probes) as standards was estimated to be 890 nm. Only those spots
that appeared and disappeared within one or two frames were considered to be exocytotic events.
Data analysis. Fluorescence and time-differentiated images were
generated, analyzed, and represented with TillVisION (TILL Photonics) and ImageJ (W. S. Rasband, National Institutes of Health,
Bethesda, MD) software packages. For statistical analysis, all numerical values of the data are expressed as means ⫾ SE, where n is the
number of cells examined. Statistical significance was determined
where appropriate by Student’s t-test with Prism3 software.
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
C1211
␮M Iso, treatment with 1–3 ␮M Iso had no effect on resting Cm
or induced a slow, smooth increase or decrease in Cm in
patch-clamped single cells (n ⫽ 5). Occasionally, one or two
stepwise events were seen superimposed on the smooth change
in Cm. The absence of Cm steps was surprising because 1 ␮M
Iso has been shown to efficiently induce amylase release from
parotid acini (34) and may indicate that adrenergic signaling is
compromised in single cells or that another vesicle population
whose fusions cannot be resolved as individual events can be
induced by lower concentrations of agonist to release amylase.
Because the stepwise changes in Cm occurred in cells in
which [Ca2⫹]i was buffered at ⬃170 nM by the patch pipette
dialysis solution, we tested whether Ca2⫹ played a role in the
Iso-induced response. In cells in which Ca2⫹ was clamped
below 30 nM by addition of BAPTA to the pipette, no events
were observed after Iso treatment, suggesting that despite the
inability to detect a volume-averaged change in cytosolic
levels, Ca2⫹ was required for cAMP-induced exocytotic activity (n ⫽ 4). Therefore, we next determined the exocytotic
response evoked by elevation of Ca2⫹ in the absence or
presence of cAMP. To achieve this, we used xenon flash
lamp-based UV photolysis of caged Ca2⫹ to apply controlled,
spatially uniform Ca2⫹ challenges. As shown in Fig. 1Ca,
flash-evoked photorelease of Ca2⫹ induced a rapid, smooth
change in Cm similar to the exocytotic burst evoked in pancreatic acinar cells (8, 15, 16). This type of Cm change contrasted
AJP-Cell Physiol • VOL
with the stepwise changes induced by Iso treatment. The
evoked rise in Cm had a maximum amplitude that averaged
594 ⫾ 63 fF (n ⫽ 10) and was best fit by an exponential line
with multiple time constants. Because the exocytotic burst
response exhibited a smooth Cm profile that lacked the stepwise increases observed after application of Iso, we hypothesized that the rapid, artificial elevation of Ca2⫹ with flash
photorelease evoked multiple, overlapping fusion events that
were not individually resolvable. To test this we applied a
continuous, low-level intensity photolytic stimulus to induce a
more gradual “ramplike” increase in Ca2⫹ (7). Surprisingly, in
contrast to the staircaselike Cm rises that we had expected to
observe, Ca2⫹ ramps induced two apparent phases of exocytosis, as shown in Fig. 1, Cb and Cc. The initial change in Cm
was induced rapidly during photorelease and increased in a
continuous, graded manner until it either reached a new steady
state or decreased back to or below prestimulus levels. After
the initial Cm increase, a second phase of exocytotic activity
was observed after a variable delay (83 ⫾ 15 s; n ⫽ 8) that
resembled the staircaselike rises in Cm that were induced by Iso
treatment. Next, we pretreated cells with forskolin or dibutyryl
cAMP to determine the effects of elevated cAMP on these
phases of exocytotic activity. As shown in Fig. 1, Cd and Ce,
elevation of cAMP had no significant effect on the initial rise
in Cm induced by Ca2⫹ ramps but enhanced the magnitude of
the response of the slower phase of activity by ⬃30% com-
289 • NOVEMBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
Fig. 1. Time-resolved membrane capacitance (Cm)
measurements of membrane dynamics in isolated
patch-clamped parotid acinar cells. A: application of
10 ␮M isoproterenol (Iso, horizontal bars) induced
stepwise increases in Cm (a), whereas 3 ␮M Iso was
much less effective at inducing stepwise changes (b).
Resting intracellular Ca2⫹ concentration ([Ca2⫹]rest)
was buffered at ⬃170 nM (traces a and b) or ⬍10 nM
by addition of BAPTA to the pipette solution (trace c).
B: time dependence and amplitudes of exocytotic
events induced by treatment with low-level photolysis
of caged Ca2⫹ (black symbols) or 10 ␮M Iso (red
symbols). C: high-intensity flash photorelease of Ca2⫹
(a) evoked a rapid increase in Cm, whereas lowintensity flash (b) or continuous low-level photorelease (c, dotted line) induced 2 apparent phases of
exocytosis. Symbols denote high (F)- or low (E)intensity UV flash application. Note longer time base
in trace c. Comparison between low-level photolysisevoked increases in Cm in a control cell (d) and
another cell pretreated for 3 min with 3 ␮M Iso (blue
bar, e) is also shown. D: maximal Cm change and
corresponding estimate of zymogen-containing secretory granule (ZSG) fusions induced within 5 min after
photorelease of Ca2⫹ or cAMP elevation under different experimental conditions.
C1212
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
pared with control (614 ⫾ 80 fF vs. 951 ⫾ 207 fF; n ⫽ 9).
Figure 1D shows the average maximal cumulative amplitude
induced by Iso, Ca2⫹ release, or Ca2⫹ release in the presence
of elevated cAMP.
Optical measurements of secretory dynamics. The observations with time-resolved Cm measurements suggested that
distinct vesicle populations of differing sizes might underlie
the two apparent phases of activity induced by Ca2⫹ elevation.
To test whether ZSG fusion was linked to one or both of these
phases of exocytosis, we applied time-differential imaging
methods to DIC images of patch-clamped acinar cells to
visualize exocytotic fusion dynamics at the single-cell level.
These events are detected as changes in optical density as
phase-dark core material is extruded from the ZSG on exocytotic fusion. This method has been previously applied to image
single exocytotic events in pancreatic acinar cells (1, 2, 11) and
submandibular salivary gland cells (26). Initially, to determine
applicability of time-differential imaging in the parotid, we
used 1–3 ␮M Iso treatment for 3 min before high-intensity
flash photorelease of Ca2⫹ to activate robust exocytotic activity
in acinar cells. As shown in Fig. 2, application of a highintensity flash to an acinar cell loaded via the patch pipette with
caged Ca2⫹ evoked a burst of ZSG fusions after a delay of
⬃400 ms. On average the latency between the flash and the
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
Fig. 2. Sequential montage of exocytotic events evoked by high-intensity flash photorelease of caged Ca2⫹ in a single patch-clamped parotid acinar cell. The
first 100 time-differential image frames (10 s total duration) after flash (caught in frame 1) demonstrate the detection of single or multiple ZSG fusions (red
arrowheads). The cell was treated for 3 min with 3 ␮M Iso before the photorelease of Ca2⫹ to induce a robust exocytotic response. A map of the evoked fusions
is shown in cell 1 in Fig 3.
AJP-Cell Physiol • VOL
289 • NOVEMBER 2005 •
www.ajpcell.org
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
AJP-Cell Physiol • VOL
close to the end of the recording interval, these lifetime values
should be considered a lower-end estimate of the stability of
these structures. An example of the postfusion structures observed in parotid acinar cells is shown in Fig. 4, A and B. To
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
first exocytotic event was 600 ⫾ 144 ms (n ⫽ 6). This delay
was significantly longer than the rapid activation of increases
in membrane surface area measured by the Cm method. The
highest rates of exocytosis occurred within the first 10 –20 s
after flash but persisted, generally at a lower rate for much
longer periods (see Fig. 4C). An advantage of the timedifferential imaging analysis method over the Cm method is the
ability to create a spatial map of exocytotic activity. For
example, as shown in Fig. 3B, ZSG fusions induced by flash
photolysis of caged Ca2⫹ began at the apical border of the cell,
and subsequent events occurred in progressively deeper regions of the cell. Thus the burst of ZSG fusions generally
initiated in the apical region and proceeded sequentially into
deeper regions of the cell interior. These events appeared and
disappeared within 1–2 frames and had an average spot diameter of 1.2 ⫾ 0.03 ␮m (estimated from pixel numbers and
before calibration with bead standards; n ⫽ 160). In addition,
larger events with a more complex profile were occasionally
observed. These events appeared to reflect the multivesicular
fusion of two to four ZSGs. As shown in Fig. 3C, a Gaussian
fit to a histogram describing the distributions of diameter sizes
of events that could be resolved as discrete spots indicated that
events were distributed around 1.14 ⫾ 0.04 ␮m. Spot diameter
after correction by calibration with bead standards was estimated to be 890 nm, comparable to the average size of a single
parotid ZSG determined from electron micrographs. Moreover,
ZSG fusions often occurred in clusters or exocytotic hot spots
that were largely confined to the apical or lateral granulecontaining regions of the cells (1). In many cases where
exocytotic activity was robust, ZSG fusions at these hot spots
occurred over several successive image frames. This sequential
activity suggested that the fusion of one ZSG can recruit or
coordinate the subsequent fusion of other granules. An example of this coordinated behavior is shown in Fig. 4B.
In addition to the direct visualization of ZSG exocytosis, a
remarkable feature regarding the fate of fused ZSGs was
observed. In contrast to neurons or endocrine cells, where
secretory vesicles either transiently fuse or fully collapse into
the surface membrane so that they may be rapidly retrieved or
recycled, the release of the phase-dark cores of the ZSGs after
fusion produced stable, phase-bright structures that remained
for tens to hundreds of seconds. Similar structures have been
observed in pancreatic acinar cells and alveolar septal cells
with multiphoton or confocal microscopy (14, 18, 28). Relative
intensities from regions of interest (ROIs) encompassing sites
of individual exocytotic events selected from corresponding
time-differential images from cell 1 in Fig. 3A were quantified,
and some examples are shown in Fig. 3D. (Line traces of the
time derivative of these events are shown in Fig. 3E.) It should
be noted that in some cases, multiple fusions were detected
within the same ROI, consistent with the observation of hot
spots of exocytotic activity. The average lifetime of phasebright structures obtained from single fusion events from
several cells that initiated within the first 20 s after flash were
grouped as either short lived (⬍100 s) or long lived (⬎100 s).
Short-lived event intensity returned to baseline with an average
duration of 41 ⫾ 11 s (n ⫽ 7). Long-lived events had an
average duration of at least 228 ⫾ 18 s (n ⫽ 11). However,
considering that some movement of the patched cell during
long periods of imaging sometimes occurred and that many of
these events either remained stable for the duration or occurred
C1213
Fig. 3. Spatial and temporal properties of ZSG exocytosis as measured by
optical method. A: maps of ZSG exocytotic activity of 3 cells induced by flash
photorelease of caged Ca2⫹ made by placing masks over ZSG fusion events on
time-differentiated image frames (outlines of cells traced from transmitted
light images). Images were collected at 100-ms intervals, and maps represent
the events induced in the first 10 s after flash photolysis. B: map of events
evoked in cell 2 replotted in ⬃1.25-s intervals demonstrating apical-to-basal
exocytotic activity. Last 2 maps (horizontal bar) are cumulative responses in
⬃10-s bins each. C: distribution of spot diameter determined from pixel
number before correction with bead standards. Gaussian fit to the histogram
(black line) indicated that the diameters of fusion events were distributed
around 1.14 ␮m. D: representative change in optical density relative to optical
density before stimulation (⌬T/T0) induced by flash photolysis and recorded
from regions of interest (ROIs) placed at fusion sites as indicated on cell 1 in
A. Corresponding line traces from time-differentiated images are shown in E.
289 • NOVEMBER 2005 •
www.ajpcell.org
C1214
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
follow the lifetime of these structures, transmitted light images
were expressed as ⌬T/T0, where T0 is the average optical
density of 10 frames before stimulation and changes in optical
density (⌬T) are relative to the preflash image.
Estimates of exocytotic rates with time-differential imaging
analysis. Differential imaging analysis was used to investigate
the Ca2⫹ dependence of ZSG exocytosis. Events evoked after
activation of exocytosis were counted and expressed as the rate
of ZSG fusion (events/min). As shown in the histogram in Fig.
4C, the delay following flash photorelease of Ca2⫹ for the first
10 fusion events/cell was generally ⬃14 s. These data demonstrated that Ca2⫹-activated ZSG fusion follows a biphasic
relationship in which the majority of events are evoked in the
first 20 s after Ca2⫹ elevation. Average rates of ZSG fusions
under different experimental conditions are shown in Fig. 4D.
The rates of exocytotic activity in single patch-clamped cells
after attainment of whole cell configuration with pipettes containing no added Ca2⫹ and BAPTA, 60 nM Ca2⫹, or 600 nM
Ca2⫹ were 0.25 ⫾ 0.25, 2.1 ⫾ 1.5 and 4 events/min (1 ⱕ n ⱕ
4), respectively. In contrast, the instantaneous elevation of
Ca2⫹ to micromolar levels by UV flash photorelease evoked an
average response of 63 events/min (n ⫽ 2). To assess the
magnitude of the Ca2⫹-induced exocytotic response in the
presence of elevated cAMP, acinar cells were treated with 3
AJP-Cell Physiol • VOL
␮M Iso for ⬃3 min before flash photolysis. The average
response to the photolytic release of Ca2⫹ after Iso treatment
was 165 ⫾ 13 events/min (n ⫽ 6). The rate of exocytotic
events was augmented nearly threefold compared with control
cells. Furthermore, in contrast to the enhancement estimated by
Cm measurements, the total number of ZSG fusions determined
by imaging over a 30-s interval after flash photolysis was
nearly fivefold that of control. These data indicate that elevated
cAMP levels greatly enhanced the number of ZSGs available
for Ca2⫹-evoked exocytotic release. The mismatch between
estimates of the cAMP-mediated enhancement obtained with
Cm measurements and optical methods suggests that Cm measurements may not properly track changes in surface area due
to sequential compound exocytosis, perhaps because of space
clamp errors.
Exocytotic rates in parotid slices. Enzymatically dispersed
single acinar cells or small clusters of three to eight acinar cells
(small acini) have been used extensively as models to study
exocrine secretion and calcium dynamics, and their study has
resulted in a wealth of information regarding these processes.
However, some reports have debated whether dissociated acinar cells can exhibit altered morphology and function (1, 20).
To address this possibility, we developed an organotypic slice
preparation of the mouse parotid gland, using a vibratome that
289 • NOVEMBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
Fig. 4. Optical assessment of ZSG exocytosis. A:
sequence of ⌬T/T0 images of an acinar cell doublet
(patch-clamped cell is on right) taken at times
indicated after flash photolysis of caged Ca2⫹.
Arrowheads denote appearance and disappearance
of stable phase-bright structures, indicating addition and retrieval of ZSGs. B: sequential timedifferentiated images from the boxed region of cell
in A showing an example of coordinated exocytotic activity. C: histogram of the delay between
flash and first 10 ZSG fusions for 5 different cells.
Most events occur within the first 20 s after flash
and peak with a 140-s delay estimated by the
Gaussian fit (black line). D: rates of ZSG fusion
induced by inclusion of 0, 0.06, and 0.6 ␮M Ca2⫹
in pipette of photorelease of caged Ca2⫹. Rates for
flash-evoked release were estimated from events
counted within the first 10 s after the flash.
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
C1215
circumvented the need to use enzymes. As shown in Fig. 5A,
this preparation largely preserved the normal three-dimensional arrangements of cell types in this tissue as indicated by
F-actin localization. Moreover, as shown in Fig. 5, B–E,
parotid slices were readily imaged with conventional digital
brightfield and fluorescence microscopy. For example, an intact acinus from a slice (Fig. 5B) was treated with 1 ␮M Iso
and imaged with transmitted light and time-differential analysis. A projection of nearly 1,000 frames of the exocytotic
events (image inverted for clarity) induced over 8 min of
continuous Iso application is shown in Fig. 5C. In addition, as
shown in Fig. 5, D and E, spatial and temporal aspects of Ca2⫹
dynamics could also be visualized in slices. Slices were loaded
for 1 h at ⬃25°C with fura-2 AM, placed onto glass-bottomed
dishes, and held in place with a small metal washer. The
majority of the cells in slices responded to CCh treatment with
AJP-Cell Physiol • VOL
increases in cytosolic Ca2⫹. For example, the Ca2⫹ response of
a small three- to four-cell acinus of a slice to application of 300
nM CCh is shown in the montage in Fig. 5D. Line traces
obtained from ROIs placed in apical and basal regions of one
acinar cell from this cluster demonstrate that CCh application
could induce Ca2⫹ oscillations that were qualitatively similar
to those characterized in isolated acini.
Using the slice preparation, we quantified the rates of exocytosis induced by adrenergic or cholinergic activation in the
semi-intact gland with time-differential imaging analysis methods. Spontaneous ZSG fusions in slices had a relatively low
frequency of occurrence. Basal rates of exocytosis were similar
between slices and single cells (0.35 ⫾ 0.17 and 0.25 ⫾ 0.25
events/min, respectively; n ⫽ 4). Application of 0.3, 3, or 10
␮M CCh evoked concentration-dependent increases in ZSG
exocytosis (1.0 ⱕ 0.5, 1.4 ⫾ 0.5, and 4.4 ⫾ 1.7 events/min,
289 • NOVEMBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
Fig. 5. Morphological and functional characteristics of a parotid gland slice preparation. A: immunofluorescent image of a
50-␮m-thick parotid slice fixed in 4% paraformaldehyde and stained with Alexa-543
phallacidin to visualize filamentous actin
structure. B: transmitted light image of live
acinus of 12–15 cells from a parotid slice.
For clarity, the rest of the slice was masked
from the image. The slice was subsequently
treated with 1 ␮M Iso, and exocytotic response was monitored by time-differential
imaging analysis. Events detected over an
8-min interval are shown in C. Image map
was constructed by projection of time-differentiated images, and resulting image was
inverted. D: montage of Ca2⫹ dynamics induced by 300 nM carbachol (CCh) treatment
in a small acinus (⬃3 cells) of a parotid
gland slice. Slice was loaded with 5 ␮M
fura-2 AM for 1 h and imaged at 2 Hz. E:
fluorescence ratio changes expressed as ratio
units (ru) from ROIs placed on 1 cell of the
acinus in D as indicated. Red line is recorded
from the apical region, and black line is from
ROI placed in the basal region.
C1216
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
respectively; 4 ⱕ n ⱕ 7). As shown in Fig. 6, A and B, 10 ␮M
CCh activated ZSG fusions that largely mapped to apical or
lateral regions of cells within an acinus, mirroring the distribution of ZSGs, similar to that demonstrated in single cells.
Similar results were also obtained after 1 ␮M Iso treatment.
The total number of events per acinus over 6 min in the
presence of CCh or Iso was 137 and 121, respectively. To
determine the kinetics of exocytosis, events of individual cells
of acini were counted in 10-s intervals, binned, and plotted as
shown in Fig. 6, C and F. These experiments revealed that CCh
application induced a biphasic response in which the rate of
exocytosis was initially ⬃7 fusions/min but decreased rapidly
thereafter. In contrast, Iso treatment produced a lower initial
rate but a persistent level of exocytotic fusions. The overall
rates of exocytosis over a 6-min interval for these CCh- and
Iso-treated acini were 1.7 and 4.4 events/min for Iso and CCh,
respectively. On average, the overall rates of exocytosis were
similar between 1 ␮M Iso- and 1 ␮M CCh-treated acini. The
average rate of exocytosis induced under different experimental conditions is shown in Fig. 6G.
Interestingly, in contrast to isolated acinar cells, in which
1–3 ␮M Iso was largely unable to induce obvious ZSG fusions,
AJP-Cell Physiol • VOL
1 ␮M Iso was an effective secretagogue in semi-intact gland
slices. To test whether Ca2⫹ was necessary for the Iso-induced
exocytosis, slices were treated with 100 ␮M dimethyl
BAPTA-AM (DMB), a high-affinity cell-permeant Ca2⫹
buffer (Kd ⫽ 40 nM). Incubation for 1 h at room temperature
with DMB decreased the average fura-2 resting ratio value
from 0.25 to 0.18 ratio units (n ⫽ 4). As shown in Fig. 6G,
treatment with DMB significantly reduced the Iso-evoked exocytotic response (2.8 ⫾ 0.8 events/min to 1.2 ⫾ 0.6 events/
min; n ⫽ 5 and 8, respectively). However, DMB treatment
was unable to completely abolish the response as introduction
of a strong exogenous Ca2⫹ buffer did in patch-clamped
acinar cells.
DISCUSSION
In this study, we applied electrophysiological and optical
methods to investigate ZSG fusion activated by direct or
agonist-induced elevation of Ca2⫹ and cAMP in mouse parotid
acinar cells and compared our findings with recent observations regarding exocytosis in pancreatic acinar cells. The findings indicate that parotid and pancreatic acinar cells exhibit
289 • NOVEMBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
Fig. 6. Transmitted light images, exocytotic maps, and histograms of ZSG fusion events in acini from parotid slice preparations evoked during continuous
application of 1 ␮M Iso (A–C) or 10 ␮M CCh (D–F) for 6 min. Insets in C and F show fusion events for each individual cell of an acinus. Histograms represent
the time course of fusion events measured at 2 Hz and analyzed in 10-s bins. E: exocytotic map for first 100 ms after stimulation only. G: average rates of
exocytosis determined from parotid slices in response to 1 ␮M Iso treatment alone, Iso after 1-h treatment with 100 ␮M dimethyl-BAPTA-AM (IS-DMB), or
CCh (0.3–10 ␮M). Rates were determined from total number of recorded events over the time interval of agonist treatment. Scale bars ⫽ 5 ␮m.
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
AJP-Cell Physiol • VOL
present study reflects recruitment of the minor pathway. Indeed, ultrastructural studies of parotid and submandibular salivary glands have demonstrated the presence of small-diameter
vesicles (⬍150 nm) at the subluminal face of apical membrane
or on or around ZSGs (23, 27). However, it is also possible that
transient fusion of ZSGs, delayed fusion pore expansion, or
retention of the phase-dense material underlies the initial phase
reported by Cm measurements. Accordingly, as differential
imaging detects the loss of phase-dark material, we would not
be able to visualize ZSG fusions faster than the time required
to extrude core material. Indeed, we did occasionally detect a
fusion event that projected over two image frames (100 ms/
frame), indicating that this process may require 0.1– 0.3 s to
complete. Therefore, additional study specifically addressing
this will be needed for a complete understanding of the relationship between these two phases of exocytosis.
Comparison of ZSG exocytosis in parotid and pancreatic
acinar cells. In part because Cm measurements distinguish
neither the specific site nor the identity of a vesicle that fuses
with the cell surface, we used time-differential imaging methods pioneered by others to optically assess parotid ZSG fusion
dynamics in real time. We found that parotid cells had a
relatively low rate of basal exocytosis but that high rates of
ZSG fusion could be evoked after the controlled photorelease
of caged Ca2⫹ with a minimum delay of ⬃200 ms. In fact, the
average delay between the flash application and the first ZSG
fusion (600 ms) was more than an order of magnitude shorter
in parotid cells than the delay reported for photolysis-evoked
fusion in pancreatic cells (10 s). In general the kinetics of the
responses were similar to those reported with the time-differential imaging analysis of pancreatic acinar cells. We found
that increases in cytosolic Ca2⫹ by direct elevation or CCh
application produced a biphasic response characterized by an
initial burst of exocytosis with rates that decreased over the
subsequent 25–30 s and then persisted at a lower rate over
minutes. (Interestingly, Iso application produced a largely
monotonic, but persistent, level of fusion events that lasted
throughout the stimulus application.) However, whereas the
overall rates of exocytosis and total numbers of fusion events
induced by agonist treatment were comparable between parotid
cells and those reported for pancreatic cells, the rate and
number of events evoked by flash photolysis were much
greater. Furthermore, under strong stimulation, ZSG fusion
proceeded from apical or sometimes lateral borders toward the
cell interior. We interpreted this activity as the sequential
compound fusion of ZSGs that proceeded from the primary
sites of fusion at the cell surface into successively deeper
regions of the cell interior (see below). Consistent with this
idea, ZSG fusions were shown coordinate subsequent fusions
at so-called hot spots. Indeed, our observations and interpretations match well with those of Campos-Toimil et al. (2), with
the exception that sequential fusion at hot spots was much
more rapid in parotid than in pancreatic acinar cells. For
example in pancreatic cells, the delay between events at defined hot spots was ⬃50 s. In the parotid, frame-to-frame
analysis showed that multiple events at hot spots could occur
over several successive frames (⬃100 ms of each other).
Moreover, unlike the pancreatic cells, a number of events
appeared to be comprised of multivesicular fusions.
Although we concede the possibility that under strong stimulus conditions primary fusions might be induced at distal sites
289 • NOVEMBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
both similar and divergent properties regarding the spatiotemporal control of exocytotic fusion of zymogen granules.
Comparison of time-resolved Cm measurements and timedifferential imaging analysis of membrane dynamics. The rate
of exocytosis induced by continuous low-level photolysis and
determined by counting stepwise events from Cm records
roughly matched that of measured by differential imaging
analysis (⬃4 events/min). However, when absolute rises in Cm
rather than steps were counted, the estimate of exocytotic
activity was two- or threefold greater. This apparent discrepancy could largely be accounted for if a significant number of
the stepwise increases observed reported compound exocytotic
events. Consistent with this idea, many of the Cm steps were
indeed multiples of 20 fF (the amount of membrane estimated
for 1 ZSG fusion). Similarly, with the understanding that the
time resolution was significantly slower than Cm measurements
(100 ms vs. 10 ms), examination of time-differentiated images
often indicated what appeared to be multivesicular fusion
events. On the other hand, high-intensity flash photorelease of
Ca2⫹ produced an exponential rise in Cm that rapidly reached
a new steady state and suggested the nearly simultaneous,
unresolved fusion of ⬃30 ZSGs. These estimates indicated a
greater number of individual fusion events than that observed
by direct visualization of fusion events by optical methods.
Moreover, the increase in Cm initiated within a few milliseconds after cessation of the flash, in contrast to the delay (600
ms on average) between the flash and the initial fusion event
detected by differential imaging. Indeed, the shortest delay
observed with optical methods was at least 200 ms. This
observation suggested that, in addition to ZSG fusion, Cm
measurements also detected the rapid fusion of another vesicle
type. The idea that there are multiple vesicle types that differ in
size, Ca2⫹ sensitivity, and kinetics of exocytosis is not without
precedence. Moreover, multiple pathways of amylase release
have been identified in parotid acinar cells (4). Constitutive and
minor regulated pathways account for the majority of release of
␣-amylase under basal conditions. The majority of evoked
release is via regulated fusion of ZSGs in response to adrenoceptor and muscarinic cholinoceptor receptor activation. However, recent studies have demonstrated that the minor regulated
pathway is functionally linked to granule exocytosis. Exocytotic fusion of the minor regulated pathway is apically directed,
precedes granule exocytosis, and can be selectively discharged
under low-level stimulus conditions (i.e., 1 ␮M Iso) (3, 5). It
has been proposed that the minor pathway can recruit syntaxin
3 to the apical membrane (or remove via endocytosis) and
modulate the number of docking sites for exocytotic fusion of
ZSGs (3). The initial phase of Cm increase we observed
exhibits features that are qualitatively similar to the minor
regulated pathway in the parotid. The rise in Cm was continuous, suggesting fusion of a population of vesicles that were too
small to be resolved by either Cm or optical methods. The
initial, sometimes biphasic Cm change was induced by lowlevel stimulation and was abolished by strong intracellular
Ca2⫹ buffering with BAPTA. In addition, BAPTA treatment
and block of the initial phase of exocytosis abolished the
secondary stepwise increases in exocytosis. Conversely, pretreatment with low concentrations of Iso enhanced the secondary phase of ZSG fusion evoked by low-level photolysis of
caged Ca2⫹. Thus it is tempting to speculate that the initial
phase of exocytosis induced by Ca2⫹ ramps identified in the
C1217
C1218
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
AJP-Cell Physiol • VOL
subject to more dynamic regulation than pancreatic cells.
These differences may reflect the multifunctional role for and
the need to rapidly and dynamically adjust saliva output in
processes such as mastication, digestion, oral health, and
speech.
GRANTS
This work was supported by National Institute of Dental and Craniofacial
Research Grant DE-014756-03 (to D. I. Yule and D. R. Giovannucci).
REFERENCES
1. Campos-Toimil M, Edwardson JM, and Thomas P. Acetylcholineinduced zymogen granule exocytosis: comparison between acini and
single pancreatic acinar cells. Pancreas 24: 179 –183, 2002.
2. Campos-Toimil M, Edwardson JM, and Thomas P. Real-time studies
of zymogen granule exocytosis in intact rat pancreatic acinar cells.
J Physiol 528: 317–326, 2000.
3. Castle AM, Huang AY, and Castle JD. The minor regulated pathway, a
rapid component of salivary secretion, may provide docking/fusion sites
for granule exocytosis at the apical surface of acinar cells. J Cell Sci 115:
2963–2973, 2002.
4. Castle JD. Protein secretion by rat parotid acinar cells. Pathways and
regulation. Ann NY Acad Sci 842: 115–124, 1998.
5. Castle JD and Castle AM. Two regulated secretory pathways for newly
synthesized parotid salivary proteins are distinguished by doses of secretagogues. J Cell Sci 109: 2591–2599, 1996.
6. Dolman NJ, Tepikin AV, and Petersen OH. Generation and modulation
of cytosolic Ca2⫹ signals in pancreatic acinar cells: techniques and
mechanisms. Biochem Soc Trans 31: 947–949, 2003.
7. Giovannucci DR, Bruce JI, Straub SV, Arreola J, Sneyd J, Shuttleworth TJ, and Yule DI. Cytosolic Ca2⫹ and Ca2⫹-activated Cl⫺ current
dynamics: insights from two functionally distinct mouse exocrine cells.
J Physiol 540: 469 – 484, 2002.
8. Giovannucci DR, Yule DI, and Stuenkel EL. Optical measurement of
stimulus-evoked membrane dynamics in single pancreatic acinar cells.
Am J Physiol Cell Physiol 275: C732–C739, 1998.
9. Imai A, Nashida T, and Shimomura H. mRNA expression of membrane-fusion-related proteins in rat parotid gland. Arch Oral Biol 46:
955–962, 2001.
10. Imai A, Nashida T, Yoshie S, and Shimomura H. Intracellular localisation of SNARE proteins in rat parotid acinar cells: SNARE complexes
on the apical plasma membrane. Arch Oral Biol 48: 597– 604, 2003.
11. Ishihara Y, Sakurai T, Habara Y, Busik JV, Kanno T, and Terakawa
S. Exocytosis in the dissociated pancreatic acinar cells of the guinea pig
directly visualized by VEC-DIC microscopy. Biochem Biophys Res Commun 277: 134 –137, 2000.
12. Jahn R. Principles of exocytosis and membrane fusion. Ann NY Acad Sci
1014: 170 –178, 2004.
13. Lang J. Molecular mechanisms and regulation of insulin exocytosis as a
paradigm of endocrine secretion. Eur J Biochem 259: 3–17, 1999.
14. Mair N, Haller T, and Dietl P. Exocytosis in alveolar type II cells
revealed by cell capacitance and fluorescence measurements. Am J Physiol
Lung Cell Mol Physiol 276: L376 –L382, 1999.
15. Maruyama Y. Agonist-induced changes in cell membrane capacitance
and conductance in dialysed pancreatic acinar cells of rats. J Physiol 406:
299 –313, 1988.
16. Maruyama Y, Inooka G, Li YX, Miyashita Y, and Kasai H. Agonistinduced localized Ca2⫹ spikes directly triggering exocytotic secretion in
exocrine pancreas. EMBO J 12: 3017–3022, 1993.
17. Muallem S. Calcium transport pathways of pancreatic acinar cells. Annu
Rev Physiol 51: 83–105, 1989.
18. Nemoto T, Kimura R, Ito K, Tachikawa A, Miyashita Y, Iino M, and
Kasai H. Sequential-replenishment mechanism of exocytosis in pancreatic
acini. Nat Cell Biol 3: 253–258, 2001.
19. Nemoto T, Kojima T, Oshima A, Bito H, and Kasai H. Stabilization of
exocytosis by dynamic F-actin coating of zymogen granules in pancreatic
acini. J Biol Chem 279: 37544 –37550, 2004.
20. Park MK, Lee M, and Petersen OH. Morphological and functional
changes of dissociated single pancreatic acinar cells: testing the suitability
of the single cell as a model for exocytosis and calcium signaling. Cell
Calcium 35: 367–379, 2004.
289 • NOVEMBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
on the cell surface, our interpretations are largely consistent
with recent observations on the exocytotic process in pancreatic acinar cells. Recently, several groups have applied multiphoton and confocal methods to directly visualize exocytotic
activity in pancreatic acinar cells. An emerging consensus
regarding the process of ZSG fusion in pancreatic acinar cells
is that after activation by Ca2⫹ levels ⬎1 ␮M, a subpopulation
of ZSGs fuse initially at the apical membrane, and additional
granule-granule fusions proceed in a sequential fashion as they
“daisy-chain” on each other (18). It is believed that sequential
compound exocytosis is an adaptation to increase the efficiency
of protein secretion where access to the subluminal apical
domain is limited. Importantly, after primary fusion events,
additional ZSG fusions appear to proceed independently of
receptor activity or Ca2⫹ gradients (19). Therefore, it is hypothesized that sequential compound exocytosis requires the
redistribution of a surface membrane protein to sites on the
newly fused granules that in turn act as nuclei for subsequent
granule-granule fusions (18, 21). Thus the recruitment of
secondary or tertiary granules is believed to be limited by the
diffusion of protein factor(s) from the plasma membrane to the
granular membranes without the mixing of plasma membrane
and granule membrane lipids (18, 28). Consistent with this
idea, ZSG fusions in parotid acinar cells were shown to
produce persistent, phase-clear structures that generally lasted
for 220 s. The lifetimes of these events were similar to the
F-actin-stabilized postfusion structures, or secretory granule
“ghosts,” produced in pancreatic acinar cells shown by multiphoton microscopy (19, 29, 31).
Monitoring exocytosis in parotid gland slice. In contrast to
those reported for pancreatic acinar cells, many of the exocytotic events in single parotid cells appeared to comprise multivesicular ZSG fusions. To ascertain whether this was a relevant
characteristic of parotid cells or a consequence of patchclamping isolated cells and artificial stimulus conditions, we
treated parotid organotypic slices with agonist and optically
monitored ZSG fusions. Numerous multivesicular fusions were
also observed in response to Iso or CCh treatment in slices, and
thus this mode of secretory activity likely represents a real
physiological property of these cells. In contrast, we found that
isolated cells did not respond to low levels of Iso or CCh with
ZSG fusions, whereas cells in acini in semi-intact slices responded to 1 ␮M Iso with robust secretory activity. Therefore,
although isolated cells were a good model for studying mechanistic aspects of Ca2⫹ and secretory activity, enzymatic treatment may alter the amount or integrity of surface receptors and
reduce the exocytotic response. Thus this indicates that some
caution in using isolated cells is warranted and that the parotid
slice preparation may provide a useful system to address
questions regarding these processes in a less invasive and more
physiologically relevant context. It is also important to note
that the overall rate of secretion, as measured by time-differential image analysis, in parotid slices is lower than that
reported for pancreatic acini. However, the total mass of
zymogen release may be similar because the percentage of
multivesicular fusions appears greater in parotid. Whether the
lower rate of exocytosis reflects a true difference in the magnitude of secretory activity remains to be determined.
In conclusion, despite remarkable similarity in morphology
and general function between these two cell types, parotid cells
demonstrate secretory kinetics that are more rapid and perhaps
SINGLE-CELL ASSAYS OF EXOCYTOSIS IN PAROTID GLAND
AJP-Cell Physiol • VOL
28. Thorn P, Fogarty KE, and Parker I. Zymogen granule exocytosis is
characterized by long fusion pore openings and preservation of vesicle
lipid identity. Proc Natl Acad Sci USA 101: 6774 – 6779, 2004.
29. Thorn P and Parker I. Two phases of zymogen granule lifetime in mouse
pancreas: ghost granules linger after exocytosis of contents. J Physiol 563:
433– 442, 2005.
30. Tomita Y, Inooka G, Shimada H, and Maruyama Y. Ca2⫹-dependent
unidirectional vesicular release detected with a carbon-fibre electrode in
rat pancreatic acinar cell triplets. Pflügers Arch 428: 69 –75, 1994.
31. Turvey MR and Thorn P. Lysine-fixable dye tracing of exocytosis shows
F-actin coating is a step that follows granule fusion in pancreatic acinar
cells. Pflügers Arch 448: 552–555, 2004.
32. Wasle B and Edwardson JM. The regulation of exocytosis in the
pancreatic acinar cell. Cell Signal 14: 191–197, 2002.
33. Williams JA, Groblewski GE, Ohnishi H, and Yule DI. Stimulussecretion coupling of pancreatic digestive enzyme secretion. Digestion 58,
Suppl 1: 42– 45, 1997.
34. Yoshimura K, Fujita-Yoshigaki J, Murakami M, and Segawa A.
Cyclic AMP has distinct effects from Ca2⫹ in evoking priming and
fusion/exocytosis in parotid amylase secretion. Pflügers Arch 444: 586 –
596, 2002.
289 • NOVEMBER 2005 •
www.ajpcell.org
Downloaded from http://ajpcell.physiology.org/ by 10.220.33.6 on June 18, 2017
21. Pickett JA, Thorn P, and Edwardson JM. The plasma membrane
Q-SNARE syntaxin 2 enters the zymogen granule membrane during exocytosis in the pancreatic acinar cell. J Biol Chem 280: 1506 –1511, 2005.
22. Rohlicek V and Schmid A. Dual-frequency method for synchronous
measurement of cell capacitance, membrane conductance and access
resistance on single cells. Pflügers Arch 428: 30 –38, 1994.
23. Segawa A, Loffredo F, Puxeddu R, Yamashina S, Testa Riva F, and
Riva A. Exocytosis in human salivary glands visualized by high-resolution scanning electron microscopy. Cell Tissue Res 291: 325–336, 1998.
24. Sollner TH. Regulated exocytosis and SNARE function. Mol Membr Biol
20: 209 –220, 2003.
25. Sudhof TC. The synaptic vesicle cycle. Annu Rev Neurosci 27: 509 –547,
2004.
26. Takagi K, Yamaguchi K, Sakurai T, Asari T, Hashimoto K, and
Terakawa S. Secretion of saliva in X-irradiated rat submandibular glands.
Radiat Res 159: 351–360, 2003.
27. Thomopoulos GN, Garrett JR, and Proctor GB. Ultrastructural histochemical studies of secretory processes in rat submandibular granular
tubules during intermittent sympathetic nerve stimulation. Eur J Morphol
38: 69 –79, 2000.
C1219