Glycobiology vol. 10 no. 2 pp. 149–157, 2000
Mucous granule exocytosis and CFTR expression in gallbladder epithelium
Rahul Kuver1,2, Johanne Henriette Klinkspoor4, William
R.A.Osborne3 and Sum P.Lee2,4
Departments of 2Medicine and 3Pediatrics, University of Washington School
of Medicine and 4Department of Medicine, Veterans Affairs Puget Sound
Health Care System, Seattle, WA 98195, USA
Received on May 8, 1999; revised on July 26, 1999; accepted on July 27,
1999
A mechanistic model of mucous granule exocytosis by
columnar epithelial cells must take into account the unique
physical-chemical properties of mucin glycoproteins and
the resultant mucus gel. In particular, any model must
explain the intracellular packaging and the kinetics of
release of these large, heavily charged species. We studied
mucous granule exocytosis in gallbladder epithelium, a
model system for mucus secretion by columnar epithelial
cells. Mucous granules released mucus by merocrine exocytosis in mouse gallbladder epithelium when examined by
transmission electron microscopy. Spherules of secreted
mucus larger than intracellular granules were noted on
scanning electron microscopy. Electron probe microanalysis demonstrated increased calcium concentrations within
mucous granules. Immunofluorescence microscopic studies
revealed intracellular colocalization of mucins and the
cystic fibrosis transmembrane conductance regulator
(CFTR). Confocal laser immunofluorescence microscopy
confirmed colocalization. These observations suggest that
calcium in mucous secretory granules provides cationic
shielding to keep mucus tightly packed. The data also
suggests CFTR chloride channels are present in granule
membranes. These observations support a model in which
influx of chloride ions into the granule disrupts cationic
shielding, leading to rapid swelling, exocytosis and hydration of mucus. Such a model explains the physical-chemical
mechanisms involved in mucous granule exocytosis.
Key words: calcium/chloride channel/cystic fibrosis/mucin
Introduction
All columnar epithelial cells are lined by a blanket of mucus.
(The term mucus is used to indicate the native, physiological
gel-like secretion. Mucins are the main component of mucus
and its use implies (biochemically) a species of highly pure
glycoprotein and is synonymous with mucous glycoprotein.
1To
whom correspondence should be addressed at: Division of
Gastroenterology, Box 356424, University of Washington, Seattle, WA
98195
© 2000 Oxford University Press
Mucous is used as an adjective to indicate relatedness to mucus
or mucin, such as in mucous secretory granules, mucous
(secretory) cells, or mucous glycoprotein. The term mucinous
is not used, and its use is discouraged.)
This thin mucus coat separates the host mucosal cells from
the exterior milieu. The major constituents of mucus are
mucins, glycoproteins with oligosaccharides attached via Oglycosidic bonds to serine or threonine residues on peptide
backbones. Mucus is also a highly hydrated gel (95–98%
water) and contains a net negative charge due to sialic acid and
sulfate residues. In physical-chemical terms, it is a negatively
charged hydrogel with the matrix tangled in a randomly woven
polyionic network.
There is evidence that mucus is not an inert substance. Physiologically, it may mediate “cytoprotection,” as exemplified in
the stomach mucosa; regulate the diffusion and direction of ion
fluxes across the mucosa; and determine the pathogenicity of
microbial organisms. Little is known about the control of
mucus secretion. In the intestines, neural and humoral factors,
as well as toxins, induce mucus secretion (Forstner et al., 1981;
Specian and Neutra, 1982). In the gallbladder, where mucus
hypersecretion is an early and consistent accompaniment in
gallstone formation, and where mucus is believed to contribute
to stone formation, mucus secretion is hypothesized to be triggered by chemical mediators (LaMorte et al., 1986). These
studies have focused on mucus secretion at an organ or tissue
level. The intracellular events involved in mucus secretion,
including the signal transduction pathways involved, have
begun to be explored (Forstner, 1995).
Regardless of the trigger signal, any model of the molecular
mechanism of mucus secretion must account for the observation that within the cell, the mucous granule is extremely
tightly packed and condensed, usually contains an excess negative charge, and is relatively underhydrated. Verdugo et al.,
studying the mucous granule of the giant slug Ariolimax
columbianus, observed a marked elevation in calcium concentration within the granule and postulated that calcium acts as a
neutralizing cation (Verdugo et al., 1987). We hypothesize that
the net negative charge in the gallbladder epithelial cell
mucous granule is also shielded or neutralized by binding to
cations; we further hypothesize that, similar to the giant slug
mucous granule, a sudden unshielding of intragranular calcium
leads to the expansion of the mucus network. Unshielding may
be mediated by calcium channels found on the granule
membrane, as originally proposed (Verdugo et al., 1987).
Alternatively, chloride channels may provide the unshielding
mechanism. Whatever the mechanism, the unshielding of
cations would drive exocytosis and hydration of the mucous
granule to form the mucus gel.
Since secretion of fluid, electrolytes and mucin from biliary
epithelial cells can be mediated by a cAMP-dependent
149
R.Kuver et al.
pathway (Igimi et al., 1992; Lenzen et al., 1992; Kuver et al.,
1994), and since the cystic fibrosis transmembrane conductance regulator (CFTR) is a cAMP-dependent chloride channel,
we sought to localize CFTR in gallbladder epithelial cells.
Indeed, in human gallbladder epithelium, CFTR has been
demonstrated in the apical membrane (Dray-Charier et al.,
1995). CFTR may play a role in the mucin secretory pathway
in epithelial cells. Gallbladder epithelial cells that overexpress
CFTR via retroviral transduction demonstrated increased
mucin secretion (Kuver et al., 1994). In rat submandibular
acini, CFTR antibodies attenuated β-adrenergic-stimulated
mucin secretion (Mills et al., 1992). In human tracheal epithelial cells, protein kinase A–stimulated glycoconjugate release
was defective in CF cells, and was corrected with adenoviralmediated CFTR gene transfer (Mergey et al., 1995). ATPdependent mucin secretion by cystic fibrosis pancreatic epithelial cells is defective (Montserrat et al., 1996). Taken together,
these studies suggest a role for CFTR in mucin secretion,
although none of these studies addressed whether the
exocytotic process itself is the site of CFTR involvement.
The present study was designed to examine intracellular
mucous granules in gallbladder epithelium, their cationic
content and spatial relationship to CFTR. The objective was to
provide evidence for a mechanism of mucous granule
exocytosis in gallbladder epithelial cells that took into account
the unique physical chemical properties of mucus.
Results
Transmission electron microscopy (TEM) studies showed the
columnar epithelial cells of the mouse gallbladder contained
mucous secretory granules (Figure 1) and these were actively
secreted into the lumen of the gallbladder. The intracellular
secretory granules were 0.16 ± 0.04 µm in diameter (mean ±
SD, n = 20, range 0.1–0.25 µm), and were crowded in the
subapical membrane just before secretion. The membrane of
the secretory granule then fused with the apical membrane. By
way of a breach of the fused membrane, the contents of the
mucous granules were released by exocytosis to the exterior of
the cell. The phenomenon could be best described as a “merocrine” mechanism of secretion (Lee, 1980). Other detailed fine
structures of the mouse gallbladder epithelial cell have been
described elsewhere (Yamada, 1968; Lee and Scott, 1982). A
continuous mucus blanket was not seen by TEM, most probably because of artifactual disruption of the native mucus layer
during fixation and sectioning. However, occasionally cells
which had their intracellular mucous granules released extracellularly could be seen with the contents expanded to acquire
the form of a spherical mass. We have observed here that not
only specialized mucus secreting apparatus such as goblet cells
and glands, but also simple columnar epithelial cells of the
mouse gallbladder epithelium, can actively secrete mucus.
Scanning electron microscopy (SEM) studies with antibody
fixation resulted in consistent preservation of the mucus
blanket which covered and obscured mucosal details such as
the microvillous membrane (Figure 2b). However, the undulations of folds and valleys could still be discerned. If the mucus
layer was carefully dissected and lifted under a dissecting
microscope before SEM studies, spherical granules were easily
150
Fig. 1. Transmission electron microscopy showing active mucus secretion. (a)
Intracellular mucous granules, ranging from 0.1–0.25 µm in diameter are
present in the subapical region of the epithelial cell. Some of these granules are
immediately adjacent to the apical plasma membrane (arrows). (b) By a breach
of the continuity of the fused membrane (secretory granule-apical membrane),
the contents of the granule (G) are discharged into the gallbladder lumen. (c)
Some of the extracellular discharged contents (M) assume a spherical structure
with dimensions many times that of the intracellular granules. All
sections×13,300.
observed tangled amongst the microvilli (Figure 2c). These
spherules ranged from 0.25 to 0.95 µm (0.48 ± 0.22 µm; mean
± SD; n = 20), and coalesced to form a plaque which annealed
with the mucus blanket. In contrast, in mouse gallbladders not
fixed with anti-mucus antibodies, the mucus layer and spherules of mucus were not well seen (Figure 2a).
Electron probe microanalysis of freeze-dried cryosections of
the same mouse tissue revealed a striking accumulation of
calcium in the intracellular secretory granules (21 ± 13 mmol/
kg dry mass; n = 9) compared with the calcium contents in the
intracellular cytoplasm (6 ± 2 mmol/kg dry mass; n = 6, p <
0.02, Student’s unpaired t test). Larger perinuclear electron
dense mitochondria also had much lower calcium levels (3 ± 3
mmol/kg dry mass, n = 5, p < 0.001 compared with mucous
granules, Student’s unpaired t test). Nuclear calcium was 1.0 ±
2.3 mmol/kg dry mass; n = 5, p < 0.01 compared with mucous
granules, Student’s unpaired t test.
With the availability of a well-differentiated canine gallbladder epithelial cell culture system (Oda et al., 1991), studies
were undertaken to determine whether mucin granules colocalized with an anion channel. CFTR was chosen as a representative anion channel that may play a role in mucous granule
exocytosis as suggested by several studies (Mills et al., 1992;
Gallbladder mucus secretion
Fig. 2. Scanning electron microscopy showing the appearance of secreted
mucus in the mouse gallbladder. (a) Gallbladder sample prepared in
conventional way, showing apical cell membrane, with microvillous
projections; 2375×. (b) Gallbladder preparation preincubated with anti-mucus
antibody showing preservation of the mucus blanket. This provides a
continuous layer covering the apical membrane. The topographic features of
an unfilled gallbladder with its “folds and valleys” can still be discerned;
2375×. (c) By lifting the layer of mucus blanket, spheres of mucus can be seen
tangled within the microvillous filaments. They range from 0.25–0.95 µm in
diameter; 6175×.
Kuver et al., 1994; Mergey et al., 1995; Montserrat et al.,
1996).
We performed fluorescence immunocytochemistry using a
well-characterized CFTR C-terminal murine monoclonal antibody and a dog gallbladder mucin rabbit polyclonal antibody.
Negative controls were performed with combinations of
primary and secondary antibodies in order to ensure no crossreactivity existed between primary and/or secondary antibodies (not shown). In addition, negative controls with the
filters used (red for TRITC, green for FITC) showed no crossreactivity.
Additional controls were performed to ensure lack of crossreactivity between the mucin and CFTR antibodies. Mucins,
purified from dog gallbladder bile and culture media of the dog
gallbladder epithelial cells, were subjected to SDS–PAGE and
Western blotting. Whereas the mucin polyclonal antibody
showed a specific staining reaction with the purified mucins,
no staining was observed with the CFTR monoclonal antibody
(data not shown). Specifically, staining of high molecular
weight proteins, which migrated in the stacking and upper part
of the separating gel, was observed, consistent with immunostaining for mucin. However, no staining was noted in the
molecular weight range expected for CFTR (140–180 kDa).
No staining of mucin was seen with the preimmune serum.
Also, western blotting of purified dog gallbladder mucin using
the anti-mucin antibody did not show low-molecular weight
bands, indicating specificity of this antibody for high molecular weight mucins. Finally, the mucin antibody showed reactivity for mucins after periodate oxidation using an ELISA
assay, indicating that reactivity was not lost following cleavage
of carbohydrate groups (data not shown).
Dog gallbladder epithelial cells stained extensively for intracellular mucin using the specific dog gallbladder mucin antibody (Figure 3, red). Intracellular CFTR staining was seen
with the murine monoclonal CFTR antibody (Fig 3, green).
Computer-assisted digitization of the immunofluorescence
images was performed in order to demonstrate colocalization
with fusion of pseudocolors. The merged images on the same
fields (with the areas of colocalization shown in yellow) indicated intracellular granule colocalization of mucin and CFTR
in these cells.
We also performed fluorescence immunocytochemistry on
dog gallbladder epithelial cells transduced with the retroviral
vector LCFSN. These cells overexpress CFTR, and demonstrate increased constitutive mucin secretion (Kuver et al.,
1994). In addition, cells transduced with the control retroviral
vector LNL6 which contains the neomycin phosphotransferase
gene (Kuver et al., 1994) did not show increased CFTR signal
on immunofluorescence studies (not shown). As shown in
Figure 4, cells overexpressing CFTR showed a corresponding
increase in the CFTR signal (compare Figure 3 and 4, green
pseudocolor images). Furthermore, colocalization of mucin
and CFTR in LCFSN-transduced cells is clearly demonstrated
using digitized images and fusion of pseudocolors (Figure 4,
yellow). In these images, colocalization appears nearly
complete for CFTR staining granules, although the mucin
signal was more pronounced in the cells.
In order to increase the resolution of these images and
thereby more convincingly demonstrate colocalization, CFTRoverexpressing dog gallbladder epithelial cells were stained
with CFTR and mucin antibodies, then scanned using fluores151
R.Kuver et al.
cence filters with laser confocal microscopy at 0.2–0.5 µm
intervals. As shown in Figure 5, the fusion of the red (mucin)
and green (CFTR) signals to give the yellow color is clearly
demonstrated in intracellular granular structures. A strong
mucin signal is noted on the plasma membrane and intracellular regions. Similar images were obtained with untransduced dog gallbladder epithelial cells double-labeled with
mucin and CFTR antibodies. Within cells overexpressing
CFTR, more granules stained with the mucin antibody than
with the CFTR antibody.
Discussion
The molecular mechanisms that allow mucus to be densely
packaged, stored and released from cells are not well known.
How, once outside the cell, the granular contents rapidly
become a hydrated gel with many times its intracellular water
content and volume is equally unclear. Mucous granules are
discharged into the mouse gallbladder lumen (Lee, 1980;
Axelsson et al., 1979). We have extended this observation by
showing the granules were discharged as a “merocrine” secretion, existing as spherules many times larger than their intracellular dimensions. Nevertheless, such morphologic
observations were made at only one point in time of a complex
dynamic process. Verdugo used video enhanced microscopy to
examine exocytosis of rabbit tracheal mucous cells and
recorded the dimension of the mucous spherules at millisecond
intervals (Verdugo, 1984). A dramatic rapid expansion in the
radius of these spherules was observed. The swelling behavior
followed first order kinetics conforming to a Donnan equilibrium process. That is, the rate of swelling was dependent on the
pH and ionic concentration in the extracellular gel medium
(Verdugo, 1984).
Such a rapid explosive swelling cannot be due to simple
diffusion of water into the condensed mass of mucin molecules, and dilution to form a gel. The huge molecular weight of
mucins (2–10 × 106) would mean hours before any passive
diffusion would reach equilibrium. The sudden expansion of
the intracellular mucous granule after its release from the cell
must mean that there is intrinsic energy stored within the
granule, with exocytosis releasing this endogenously stored
force (Verdugo et al., 1987). This concept, and the subsequent
swelling characteristics, can be explained by the observation
that mucins contain net negative charges which repulse and
cause the gel to swell as predicted by the Donnan equilibrium
(Verdugo, 1984).
Two other findings in the present study are relevant in
considering mucin exocytosis. One is the selective entrapment
Fig. 3. Digitized images of immunofluorescence cytochemical studies on
subconfluent dog gallbladder epithelial cells grown on glass coverslips
showing intracellular colocalization of CFTR and mucin. Staining with the
rabbit polyclonal anti-mucin antibody and the mouse monoclonal anti-CFTR
antibody followed by incubation with the two secondary antibodies performed
sequentially on the same slide showed specific staining for mucin granules in
the majority of cells, as well as staining with the CFTR antibody. Shown are
representative slides. CFTR signal (green pseudocolor) corresponding to FITC
filter. Mucin signal (red pseudocolor) corresponding to TRITC filter.
Composite image analysis (yellow pseudocolor) showing areas of
colocalization (87×).
152
Gallbladder mucus secretion
Fig. 5. Confocal laser scanning immunofluorescence microscopy of
subconfluent dog gallbladder epithelial cells overexpressing CFTR. Shown is
a representative sample with CFTR signal in green, mucin signal in red and
composite signal in yellow (1920×).
of calcium within the secretory granules; the other is the intracellular colocalization of CFTR chloride channels with
mucous granules. To maintain electrical neutrality, the negative charges of mucins inside the granule must be neutralized
or shielded. Our observation of elevated calcium in the mucous
secretory granules is consistent with the idea that calcium
provides the cationic shielding mechanism (Nicaise et al.,
1992). The mucous granule in the gallbladder epithelium is not
the only intracellular product that uses cationic shielding. The
shielding species can be a protein, such as the spermine in
DNA which neutralizes negative charges on phosphate groups
(Marquet et al., 1985). Elevated calcium has been recorded in
secretory granules of cells in the salivary duct, labial glands
Fig. 4. Digitized images of immunofluorescence cytochemical studies on
subconfluent dog gallbladder epithelial cells overexpressing CFTR grown on
glass coverslips. Staining and digitized pseudocolor analysis were performed
as described for Figure 3. CFTR signal (green pseudocolor). Mucin signal (red
pseudocolor). Composite signal (yellow pseudocolor; 870×).
153
R.Kuver et al.
and the pancreas (Roomans et al., 1982; Sasake et al., 1983;
Izutsu et al., 1985). In mucous secretory granules of the terrestrial slug Ariolimax columbianus, an even more marked
concentration of calcium exists (Verdugo et al., 1987). There is
a sudden release of calcium from the mucous secretory granules which occurs immediately preceding the release of
granule contents, a step which is often associated with fusion
of the membrane of the secretory granule with the apical
plasma membrane (Zimmemberg and Whitaker, 1985). In
some tissues such as the chicken trachea, calcium is cosecreted with mucus (Kent and Mian, 1987).
A malfunction of CFTR (cystic fibrosis) will result in a relative deficiency of fluid and electrolytes and thereby not allow
proper hydration of the exocytosed mucus. This effect may be
enhanced by the sodium hyperabsorption characteristic of CF
cells resulting from abnormal regulation of an epithelial
sodium channel by CFTR (Stutts et al., 1995). Alternatively, if
CFTR functions on the mucous granule membrane as the
trigger for mucus release, then absent or defective CFTR may
lead to unregulated mucus release, contributing to the mucus
hypersecretion characteristic of CF. In CF, the secreted mucus
remains as a highly condensed tangled polyionic network and
manifests itself as highly viscous “inspissated” mucus. This
speculation can be examined by concomitantly testing CFTR
function with respect to Cl–, fluid, and mucus secretion.
For immunofluorescence studies, dog gallbladder epithelial
cells were grown on glass coverslips. No significant plasma
membrane CFTR staining was noted. This finding may have
several explanations. As culture conditions did not allow
differentiation into apical and basolateral membrane domains,
membrane localization of CFTR may have been affected.
Targeting of CFTR to the apical plasma membrane is linked to
polarization of cultured human pancreatic duct cells, and a
similar phenomenon may occur with cultured gallbladder
epithelial cells (Hollande et al., 1998). Alternatively, in these
cells under these culture conditions, CFTR may be present in
low copy number in unstimulated cells on the plasma
membrane. Stimulation with agonists that elevate intracellular
cAMP may be required to allow trafficking of CFTR from
intracellular vesicles. The combination of lack of polarization
and lack of CFTR stimulation may therefore account for the
absence of plasma membrane CFTR staining, as reported for
other cultured epithelial cells (Morris et al., 1994).
The question of cross-reactivity between the CFTR and
mucin antibodies was rigorously addressed in light of the colocalization seen with immunofluorescence studies. Several
controls were used to ensure lack of cross-reactivity. First, no
cross-reactivity between the primary and secondary antibodies
was noted, nor was there signal crossover between the two
filters used. Second, Western blotting of purified dog gallbladder mucin with the mucin antibody showed the expected
high molecular weight bands, which were clearly distinguished
from the molecular weight regions expected for CFTR.
Finally, the CFTR-overexpressing cells showed an increase in
staining for CFTR, but no increase in intracellular mucin
staining. While the latter finding supports the lack of crossreactivity between the two antibodies, this seems to contradict
our earlier observation that CFTR-overexpressing cells
showed increased mucin secretion (Kuver et al., 1994). One
explanation for this discrepancy is that while the rate of mucin
secretion was increased in CFTR-overexpressing cells, the
154
sum of intracellular mucin stores was not significantly
different. These findings, however, deserve further investigation.
Chloride channels are functionally involved in the regulation
of exocytosis (Gasser et al., 1988) and chloride channels and G
proteins have been found in the membrane of secretory granules (Gasser et al., 1988; Watson et al., 1992). In intestinal,
salivary and biliary epithelium, both Cl– (encompassing Na+
and fluid) secretion as well as mucus secretion can be concomitantly driven by a cAMP-dependent pathway (Igimi et al.,
1992; Fitz et al., 1993; Kuver et al., 1994). This simultaneous
response to a common biochemical intracellular signal transduction pathway is appropriate physical-chemically since
abundant fluid must accompany mucus exocytosis so that
extracellular mucus can be hydrated. Cholera toxin, for
example, stimulates copious intestinal secretion of fluid and
chloride ions, as well as mucin (Forstner et al., 1981).
One hypothesis to explain these observations is that in
mucus exocytosis, the initial event is Cl– moving into the
granule from the cytoplasmic compartment. This event may
coincide with changes mediated by the family of proteins
which controls vesicular docking and membrane fusion (Naren
et al., 1997; Low et al., 1998; Schimmoller et al., 1998).
Swelling due to this ionic movement leads to the expulsion of
the granule. Cationic shielding by calcium explains how
mucus can be condensed and stored in an intracellular granule.
This hypothesis also describes a common intracellular pathway
in the control of mucus secretion, whether this is initiated by
chemical mediators (LaMorte et al., 1986), by adrenergic,
peptidergic (Fleming et al., 1987), or cholinergic (Forstner et
al., 1981; Specian and Neutra, 1982) mechanisms. That the
mucous granule, in the form of a tightly packed and condensed
network of polyions, once released of its cationic shielding,
will rapidly expand according to a Donnan equilibrium process
(Verdugo et al., 1987; Nanavati and Fernandez, 1993; DrayCharier et al., 1995) and can be immediately hydrated with the
optimum amount of extracellular fluid is also predicted by this
hypothesis.
Functional data for chloride channel involvement in mucous
granule exocytosis is beyond the scope of this report. Others
have shown the presence of functional ionic channels on
zymogen granules from pancreas and salivary gland (Gasser et
al., 1988; Gasser and Hopfer, 1990). Characterization of these
channels on granule membranes has been carried out
(Thevenod et al., 1990). In order to perform similar studies on
mucous granules, methods for isolating intact and stable
mucous granules will need to be developed. Ongoing studies
address the feasibility of isolating such intact mucous granules
from columnar epithelial cells. The ultimate proof of a role for
CFTR or another anion channel in mucous granule exocytosis
will come from such functional studies.
Materials and methods
All procedures involving animals were approved by the institutional animal use committee, and received humane care
according to the criteria outlined in the “Guide for the Care and
Use of Laboratory Animals” published by the National Institutes of Health.
Gallbladder mucus secretion
Transmission electron microscopy (TEM)
Male inbred pathogen-free albino mice weighing 20–30 g were
used. They were housed four to a stainless steel cage with wire
mesh bottom to prevent coprophagia. The cages were kept in a
laboratory equipped with an automatic diurnal lighting system
and at a temperature of 21 ± 2.0°C. Animals were fed a regular
Purina Chow diet and were euthanized at 9:00–10:00 a.m.
without prior fasting by cervical dislocation. Mouse gallbladders (n = 20) were immediately removed, opened and carefully
washed with 0.15 N saline, cut into 2 mm sections and rapidly
transferred to 2.5% glutaraldehyde at 4°C for fixation. These
were post-fixed in osmic acid and dehydrated in graded
ethanol. They were oriented and embedded in Epon 812,
sectioned on a Reichert ultratome, stained with 50% methanolic uranyl acetate and Reynolds lead acetate, and studied
with a Philips EM 300 electron microscope as previously
described (Lee and Scott, 1982).
Scanning electron microscopy (SEM)
The relationship between columnar epithelial cells and the
adjacent mucus coat has not been studied adequately by SEM
because routine preparative techniques including fixation and
dehydration leave the mucus as condensed and disrupted
strands. In the rat colon, we have found that pretreatment of
specimens with anti-mucus antibodies reliably preserved the
mucus in relation to the epithelium (Bollard et al., 1986).
Cholecystectomies were performed on 35 mice and gallbladder mucus was collected by scraping away the native
mucus gel with a glass slide. Pooled mucus was homogenized in nine volumes of phosphate-buffered saline and
stored at –70°C until used in the immunization schedule in
rabbits. Phosphate-buffered saline (0.5 ml) containing 100 µg
of protein was emulsified with 1.5 ml of Freund’s complete
adjuvant using two syringes and a double hub connector. This
emulsion was injected subcutaneously at eight sites on both
sides of the rabbit’s spine.
Four weeks thereafter rabbits were given a booster injection
(0.5 ml) by the same technique. The animals were bled (10–
20 ml) at 7 days after the booster injection and the serum separated by centrifugation. Antibody activity was confirmed using
the tube precipitin test, double diffusion in agarose, and
binding to immobilized mucus on cellulose nitrate membranes.
The serum of the highest titer (1:100) was used throughout.
Complement activity in the sera was removed either by adding
solid Na2EDTA to a final concentration of 10 mM or by incubation at 56°C for 30 min.
Laparotomies were performed on mice (n = 20) and cholecystectomies performed after in vivo perfusion–fixation using
4% glutaraldehyde (Bollard et al., 1986). Mouse gallbladders
were opened, washed and oriented on a wax impregnated mesh
that was then pinned onto a wax petri dish. The mucus antibody was added to cover the specimen and incubated at 4°C for
30 min to allow antibody–mucus complexing. Samples were
immersed in 4% glutaraldehyde for 16 h and subjected to critical point drying. Control samples (n = 20) without antibody
pretreatment were also used. The dried tissue specimens,
mounted on aluminum stubs with copper conducting paint,
were coated with carbon, then gold by vacuum evaporation
before examination in an ISI DS-130 research scanning elec-
tron microscope fitted with a Robinson back scattered electron
detector (Bollard et al., 1986).
X-Ray elemental microanalysis
Immediately after removal, gallbladders (n = 4) were opened
and cut into 2 mm squares, then plunged rapidly into freon
slush at –160 to –180°C. We obtained cryosections at –110°C
by cutting the tissue, glued with toluene to metal chucks, using
a Sorvall MT2B ultramicrotome with FS 1000 cryokit. Thin
cryosections were then transferred dry to carbon-coated
Formvar support films on folding grids and freeze-dried in an
oil-free vacuum system.
Freeze-dried cryosections were viewed and analyzed in
STEM mode in a JEOL 1200 EX electron microscope. X-Ray
spectra were collected and processed using a Link AN 10,000
and 30 mm2 detector. Quantitation procedures which have
been described elsewhere (Izutsu et al., 1985) make use of
protein and salt standards (Shuman et al., 1976). Square rasters
were placed over regions of granules (200 × 200 nm), cytosol
(600 × 600 nm) and nuclei (1 µm × 1 µm). Electron dense
structures assumed to be secretory granules were identified on
the basis of their size and location (apical to nucleus) in the
cells. Larger perinuclear electron dense mitochondria were
also studied. Spectra were collected from samples at room
temperature and corrections were made for beam-induced
mass loss (Cantino et al., 1986).
Fluorescence immunocytochemistry
Mucin purification. Gallbladder bile from a mongrel dog was
obtained during surgery by aspiration of the gallbladder by
needle puncture and frozen at –20°C. Purification of mucin
from the gallbladder bile was performed as described previously (Smith, 1987), with slight modifications. Briefly, 5 ml of
bile was saturated with ammonium sulfate. After incubation at
4°C for 16 h, a floating layer of lipids and mucin was removed
and dissolved in PBS, containing 20 mM sodium cholate and
0.02% sodium azide, pH 7.4. Mucin was purified by repeated
preparative cesium chloride (CsCl) equilibrium-densitygradient ultracentrifugation. After Amicon filtration and
lyophilization, the purified dog gallbladder mucin was
dissolved in 0.5 ml of PBS. The purified mucin was subjected
to SDS–polyacrylamide gel electrophoresis. Staining with
Coomassie brilliant blue for total protein and periodic acid–
Schiff for sugar residues demonstrated that the preparation was
free from low-molecular weight contaminants and contained
only high-molecular weight glycoproteins.
Preparation of polyclonal antibody to dog gallbladder mucin.
A New Zealand White rabbit was injected subcutaneously with
100 µg of mucin purified from dog gallbladder bile in
complete Freund’s adjuvant. At 15 and 36 days, this was
repeated with injection of 50 µg of mucin in Freund’s incomplete adjuvant. On day 68 a final injection of 50 µg of mucin in
incomplete adjuvant was given and after 10 days 20 ml of
serum was collected and stored at –20°C. Preimmune serum
was obtained before the first immunization. The reactivity of
the polyclonal antibody with the purified dog gallbladder
mucin was tested after SDS–PAGE and Western blotting. The
antibody was shown to give a specific staining reaction with
155
R.Kuver et al.
the purified mucin, whereas no staining was seen with the
preimmune serum.
Immunofluorescence studies. Normal dog gallbladder epithelial cells and dog gallbladder epithelial cells overexpressing
CFTR were cultured on glass coverslips coated with 2.5 µg/
cm2 rat tail collagen type I to approximately 50% confluence.
CFTR-overexpression was achieved via retroviral transduction
with the vector LCFSN encoding the human CFTR cDNA
driven by the retroviral LTR promoter, as described previously
(Kuver et al., 1994). Cells were washed once with PBS, and
fixed in cold methanol (-20°C) for 5 min, followed by cold
acetone for 5 min (-20°C). The cells were washed with 70%
ETOH in PBS/Tween-20 (0.05%, v/v) then incubated with
mouse CFTR C-terminal monoclonal antibody (Genzyme) at
1:50 dilution in PBS/Tween-20/10% normal goat serum (NGS)
for 60 min. The rabbit mucin antibody was applied at 1:50 dilution in PBS/Tween-20/NGS for 60 min. Secondary antibodies
were goat-anti-rabbit-tetramethylrhodamine B isothiocyanate
(TRITC) and goat anti-mouse-fluorescein isothiocyanate
(FITC) (Molecular Probes, Eugene, OR), each applied at 1:50
dilution for 60 min. The slides were washed three times with
PBS/Tween-20 and sealed in mounting media containing pphenylenediamine.
The cells were examined using epifluorescence microscopy,
utilizing standard filter sets to visualize rhodamine and fluorescein fluorochromes. Photographs were taken on Kodak Ektachrome PRD200X film, using a Zeiss Axiovert 135
fluorescence microscope (Oberkochen, Germany), and a
Contax 167 MT camera (Tokyo, Japan). The cells were also
examined using a Zeiss Axiophot fluorescence microscope.
Images were captured using a Hamamatsu C4880 fast-cooled
integrating CCD camera and MCID M2 software (Imaging
Research, St. Catherines, Ontario, Canada). Images from the
same group of cells, but using different excitation filters were
digitized into separate channels, and assigned either a red
(mucin) or a green (CFTR) pseudocolor. The channels were
subsequently merged into a single image, using the software’s
image fusion function, resulting in the simultaneous visualization of two signals in the same group of cells. The same fields
were fused resulting in visualization of areas of colocalization.
Confocal laser scanning immunofluorescence microscopy
Samples were prepared as described above. Confocal images
were acquired on a Bio-Rad MRC-1024 fitted to a Nikon
Diaphot 200 fluorescence microscope. Samples were illuminated with all lines of a krypton/argon laser and scanned at
intervals of 0.2–0.5 µm using the 60× or 100× oil immersion
objective.
Acknowledgments
We thank Su Wan Chen, Manuel Villalon, Marie Cantino,
Marilyn Skelly and Andrew Shirk for valuable technical assistance; Dr. Andy McShea for confocal laser scanning fluorescence microscopy assistance; John Breiniger for fluorescence
immunocytochemistry assistance; and Drs. Kenneth Izutsu and
Pedro Verdugo for helpful discussions. NIH RO1 DK 50246,
Medical Research Service of the Department of Veterans
156
Affairs, NIH CF Core Grant DK 47754, and Cystic Fibrosis
Foundation Post-doctoral Fellowship.
Abbreviations
CFTR, cystic fibrosis transmembrane conductance regulator;
FITC, fluorescein isothiocyanate; SEM, scanning electron
microscopy; TEM, transmission electron microscopy; TRITC,
tetramethylrhodamine B isothiocyanate.
References
Axelsson,H., Danielsson,R., Henriksson,T. and Wahlin,T. (1979) Secretory
behavior and ultrastructural changes in mouse gallbladder principal cells
after stimulation with cholinergic drugs. Gastroenterology, 76, 335–340.
Bollard,J., Vanderwee,M., Smith,G., Tasman-Jones,C., Gavin,J. and Lee,S.
(1986) Preservation of mucus in situ in rat colon. Dig. Dis. Sci., 31, 1338–
1344.
Cantino,M., Wilkinson,L., Goddard,M. and Johnson,D. (1986) Beam induced
mass loss in high resolution biological microanalysis. J. Microsc., 144,
317–327.
Dray-Charier,N., Paul,A., Veissiere,D., Mergey,M., Scoazec,J.-Y., Capeau,J.,
Brahimi-Horn,C. and Housset,C. (1995) Expression of cystic fibrosis
transmembrane conductance regulator in human gallbladder epithelial
cells. Lab. Invest., 73, 828–836.
Fitz,J., Basavappa,S., McGill,J., Melhus,O. and Cohn,J. (1993) Regulation of
membrane chloride currents in rat bile duct epithelial cells. J. Clin. Invest.,
91, 319–328.
Fleming,N., Bilan,P., Slivinski-Lis,E. and Carvalho,V. (1987) Muscarinic, 1α adrenergic and peptidergic agonists stimulate phosphoinositide hydrolysis and regulate mucin secretion in rat submandibular gland cells. Eur. J.
Physiol., 409, 416–421.
Forstner,G. (1995) Signal transduction, packaging and secretion of mucins.
Annu. Rev. Physiol., 57, 585–605.
Forstner,J., Roomi,N., Fahim,R. and Forstner,G. (1981) Cholera toxin stimulates secretion of immunoreactive intestinal mucin. Am. J. Physiol., 240,
G10–G16.
Gasser,K. and Hopfer,U. (1990) Chloride transport across the membrane of
parotid secretory granules. Am. J. Physiol., 259, C413–C420
Gasser,K., DiDomenico,J. and Hopfer,U. (1988) Secretogogues activate chloride transport pathways in pancreatic zymogen granules. Am. J. Physiol.,
254, G93–G99.
Hollande,E., Fanjul,M., Chemin-Thomas,C., Devaux,C., Demolombe,S., Van
Rietschoten,J., Guy-Crotte,O. and Figarella,C. (1998) Targeting of CFTR
protein is linked to the polarization of human pancreatic duct cells in culture. Eur. J. Cell Biol., 76, 220–227.
Igimi,H., Yamamoto,F. and Lee,S. (1992) Gallbladder mucosal function: studies in absorption and secretion in humans and in dog gallbladder epithelium. Am. J. Physiol., 263, G69–G74.
Izutsu,K., Johnson,D., Schubert,M., Wang,E., Ramsey,B., Tamarin,A., Truelove,G., Ensign,W. and Young,M. (1985) Electron microprobe analysis of
human labial gland secretory granules in cystic fibrosis. J. Clin. Invest., 75,
1951–1956.
Kent,P. and Mian,N. (1987) Transmural calcium fluxes and role of mucins as
cellular calcium transport vehicles in chicken trachea in vitro. J. Physiol.
(Lond.), 388, 121–140.
Kuver,R., Savard,C., Oda,D. and Lee,S. (1994) PGE generates intracellular
cAMP and accelerates mucin secretion by cultured dog gallbladder epithelial cells. Am. J. Physiol., 267, G998–G1003.
Kuver,R., Ramesh,N., Lau,S., Savard,C., Lee,S. and Osborne,W. (1994) Constitutive mucin secretion linked to CFTR expression. Biochem. Biophys.
Res. Commun., 203, 1457–1462.
LaMorte,W., LaMont,J., Hale,W., Booker,M., Scott,T. and Turner,B. (1986)
Gallbladder prostaglandins and lysophospholipids as mediators of mucin
secretion during cholelithiasis. Am. J. Physiol., 251, G701–G709.
Lee,S. (1980) The mechanism of mucus secretion by the gallbladder epithelium. Br. J. Exp. Pathol., 131, 117–119.
Gallbladder mucus secretion
Lee,S. and Scott,A. (1982) The evolution of morphologic changes in the gallbladder before stone formation in mice fed a cholesterol cholic acid diet.
Am. J. Pathol., 108, 1–8.
Lenzen,R., Alpini,G. and Tavaloni,N. (1992) Secretin stimulates bile ductular
secretory activity through the cAMP system. Am. J. Physiol., 263, G527–
G532.
Low,S., Chapin,S., Wimmer,C., Whiteheart,S., Komuves,L., Mostov,K. and
Weimbs,T. (1998) The SNARE machinery is involved in apical plasma
membrane trafficking in MDCK cells. J. Cell Biol., 141, 1503–1513.
Marquet,R., Houssier,C. and Fredericz,E. (1985) An electro-optical study of
the mechanism of DNA condensation induced by spermine. Biochim. Biophys. Acta, 825, 365–374.
Mergey,M., Lemnaouar,M., Veissiere,D., Perricaudet,M., Gruenert,D.,
Picard,J., Capeau,J., Brahimi-Horn,M.-C. and Paul,A. (1995) CFTR gene
transfer corrects defective glycoconjugate secretion in human CF tracheal
epithelial cells. Am. J. Physiol., 269, L855–L864.
Mills,C., Pereira,M., Dormer,R. and McPherson,M. (1992) An antibody
against a CFTR-derived synthetic peptide, incorporated into living submandibular cells, inhibits beta-adrenergic stimulation of mucin secretion.
Biochem. Biophys. Res. Commun., 188, 1146–1152.
Montserrat,C., Merten,M. and Figarella,C. (1996) Defective ATP-dependent
mucin secretion by cystic fibrosis pancreatic epithelial cells. FEBS Lett.,
393, 264–268.
Morris,A., Cunningham,S., Tousson,A., Benos,D. and Frizzell,R. (1994)
Polarization-dependent apical membrane CFTR targeting underlies
cAMP-stimulated Cl- secretion in epithelial cells. Am. J. Physiol., 266,
C254–C268.
Nanavati,C. and Fernandez,J. (1993) The secretory granule matrix: a fast-acting smart polymer. Science, 259, 96–98.
Naren,A., Nelson,D., Xie,W., Jovov,B., Pevsner,J., Bennett,M., Benos,D.,
Quick,M. and Kirk,K. (1997) Regulation of CFTR chloride channels by
syntaxin and Munc18 isoforms. Nature, 390, 302–305.
Nicaise,G., Maggio,K., Thirion,S., Horoyan,M. and Keicher,E. (1992) The
calcium loading of secretory granules. A possible key event in stimulussecretion coupling. Biol. Cell., 75, 89–99.
Oda,D., Lee,S. and Hayashi,A. (1991) Long term culture and partial characterization of dog gallbladder epithelial cells. Lab. Invest., 64, 682–692.
Roomans,G., Wei,X., Ceder,O. and Kollberg,H. (1982) The reserpinized rat in
the study of cystic fibrosis–X-ray microanalysis of submandibular gland
and pancreas. Ultrastruct. Pathol., 3, 285–293.
Sasake,S., Nakagaki,I., Mori,H. and Imai,Y. (1983) Intracellular calcium
stores and transport of elements in acinar cells of the salivary gland determined by electron probe x-ray microanalysis. Jpn. J. Physiol., 33, 69–83.
Schimmoller,F., Simon,I. and Pfeffer,S. (1998) Rab GTPases, directors of vesicle docking. J. Biol. Chem., 273, 22161–22164.
Shuman,H., Somlyo,A. and Somlyo,A. (1976) Quantitative electron probe
microanalysis of biological thin sections: methods and validity. Ultramicroscopy, 1, 317–339.
Smith,B. (1987) Human gallbladder mucin binds biliary lipids and promotes
crystal nucleation in model bile. J. Lipid Res., 28, 1088–1097
Specian,R. and Neutra,M. (1982) Regulation of intestinal mucus secretion:
role of parasympathetic stimulation. Am. J. Physiol., 242, G370–G379.
Stutts,M., Canessa,C., Olsen,J., Hamrick,M., Cohn,J., Rossier,B. and
Boucher,R. (1995) CFTR as a cAMP-dependent regulator of sodium channels. Science, 269, 847–850.
Thevenod,F., Gasser,K. and Hopfer,U. (1990) Dual modulation of chloride
conductance by nucleotides in pancreatic and parotid zymogen granules.
Biochem. J., 272, 119–126.
Verdugo,P. (1984) Hydration kinetics of exocytosed mucins in cultured secretory cells of the rabbit trachea: a new model. Mucus and Mucosa. Ciba
Foundation Symposium 109. Pitman, London, pp. 212–234.
Verdugo,P., Deyrup-Olsen,I., Aitken,M., Villalon,M. and Johnson,D. (1987)
Molecular mechanism of mucin secretion. I. The role of intragranular
charge shielding. J. Dent. Res., 66, 506–508.
Watson,E., DiJulio,D., Kauffman,D., Iversen,J., Robinovitch,M. and Izutsu,K.
(1992) Evidence for G proteins in rat parotid plasma membranes and secretory granule membranes. Biochem. J., 285, 441–449.
Yamada,K. (1968) Some aspects of the fine structure of the gallbladder epithelium of the mouse. Folia Anat. Jpn., 45, 11–21.
Zimmemberg,J. and Whitaker,M. (1985) Irreversible swelling of secretory
granules during exocytosis caused by calcium. Nature, 315, 581–584.
157
R.Kuver et al.
158