Molecular Biology of the Cell
Vol. 18, 1056–1063, March 2007
VAMP8/Endobrevin as a General Vesicular SNARE
for Regulated Exocytosis of the Exocrine System
Cheng-Chun Wang,* Hong Shi,* Ke Guo,* Chee Peng Ng,* Jie Li,* Bin Qi Gan,*
Hwee Chien Liew,* Jukka Leinonen,†Hannu Rajaniemi,†Zhi Hong Zhou,*
Qi Zeng,* and Wanjin Hong*
*Institute of Molecular and Cell Biology, Singapore 138673, Singapore; and†Department of Anatomy and Cell
Biology, University of Oulu, 90014 Oulu, Finland
Submitted November 2, 2006; Revised December 21, 2006; Accepted January 2, 2007
Monitoring Editor: Vivek Malhotra
The molecular mechanism governing the regulated secretion of most exocrine tissues remains elusive, although VAMP8/
endobrevin has recently been shown to be the major vesicular SNARE (v-SNARE) of zymogen granules of pancreatic
exocrine acinar cells. In this article, we have characterized the role of VAMP8 in the entire exocrine system. Immuno-
histochemical studies showed that VAMP8 is expressed in all examined exocrine tissues such as salivary glands, lacrimal
(tear) glands, sweat glands, sebaceous glands, mammary glands, and the prostate. Severe anomalies were observed in the
salivary and lacrimal glands of VAMP8-null mice. Mutant salivary glands accumulated amylase and carbonic anhydrase
VI. Electron microscopy revealed an accumulation of secretory granules in the acinar cells of mutant parotid and lacrimal
glands. Pilocarpine-stimulated secretion of saliva proteins was compromised in the absence of VAMP8. Protein aggregates
were observed in mutant lacrimal glands. VAMP8 may interact with syntaxin 4 and SNAP-23. These results suggest that
VAMP8 may act as a v-SNARE for regulated secretion of the entire exocrine system.
Protein and lipid transport in the secretory and endocytic
pathways is primarily mediated by shuttling intermedi-
ates in the form of small vesicles (50–100 nm in diameter)
and/or larger containers (100–2000 nm). Studies over the
last three decades have identified molecular machineries
and have defined fundamental mechanisms responsible
for vesicle-mediated trafficking. Four key events have
been described for general vesicle-mediated transport be-
tween a donor and a target compartment. Various coat
protein complexes function in the process of vesicle for-
mation by causing membrane deformation and selecting
cargo proteins into the budding vesicle at a donor com-
partment. The resulting vesicles/containers are delivered
to the target compartment, a process facilitated by the
cytoskeleton network. The tethering event acts to position
the vesicles/containers in the precise vicinity of the target
compartment and is mediated by various tethering pro-
teins. The fusion of vesicles/containers with the target
compartment is catalyzed by SNARE (N-ethylmaleimide–
sensitive factor [NSF]-attachment protein [SNAP] recep-
tor) complexes. The coat proteins, tethering proteins and
SNAREs are converging points for the cell to regulate
vesicle trafficking. The pairing of vesicular SNARE (v-
SNARE) on the vesicle/container with t-SNARE on the
target compartment is the core event in membrane fusion
(Barr, 2000; Bonifacino and Glick 2004; Hong, 2005; Jahn
and Scheller, 2006; McNiven and Thompson, 2006).
SNAREs mediate secretion of small molecules and
polypeptides in many physiological processes. Synaptic
vesicles of neurons store neurotransmitters and are trig-
gered by the action potential to fuse with nerve terminals.
The core SNARE complex mediating the release of neu-
rotransmitters is well defined and is known to be regu-
lated by many accessory proteins (So ¨llner et al., 1993; Jahn
and Sudhof, 1999; Schoch et al., 2001; Rettig and Neher,
2002; Chandra et al., 2005; Rizo et al., 2006). VAMP2/
synaptobrevin2 acts as the v-SNARE of synaptic vesicles,
whereas syntaxin 1 and SNAP-25 interact to form the
t-SNARE complex on the plasma membrane of nerve
terminals. The interaction of VAMP2 with the syntaxin
1-SNAP-25 complex catalyzes the release of neurotrans-
mitters. In contrast to the small synaptic vesicles used by
neurons, endocrine and exocrine cells use larger contain-
ers called secretory granules to store proteins and bioac-
tive components, and their fusion with the plasma mem-
brane is regulated by various stimuli.
Despite the identification of ?38 SNAREs and the de-
tailed studies of the SNARE complex mediating the syn-
aptic transmission, the molecular mechanism and the
SNARE complexes responsible for regulated exocytosis of
the exocrine tissues remain elusive. Using the VAMP8
knockout mice generated recently (Wang et al., 2004), we
have examined the hypothesis that VAMP8 may partici-
pate in regulated exocytosis of the entire exocrine system.
Our study suggests that VAMP8 plays an important role
in parotid and lacrimal acinar cells and that it functions in
other exocrine tissues as well. This discovery will facili-
tate further analysis of the molecular mechanism respon-
sible for regulated exocytosis of the exocrine system.
This article was published online ahead of print in MBC in Press
on January 10, 2007.
Address correspondence to: Wanjin Hong (firstname.lastname@example.org.
1056© 2007 by The American Society for Cell Biology
MATERIALS AND METHODS
The VAMP8 polyclonal antibody was raised in rabbits with the cytosolic
N-terminal region of human VAMP8 as antigen (Wong et al., 1998). Other
antibodies were purchased from commercial suppliers: rabbit anti-amylase
from Calbiochem (La Jolla, CA); rabbit anti-KLK9 from Fitzgerald Industries
(Concord, MA); rabbit anti-syntaxin4, syntaxin13, and rabbit anti-SNAP23
from Synaptic Systems (Go ¨ttingen, Germany); mAb to syntaxin6 from AbCam
(Cambridge, United Kingdom).
The VAMP8 knockout mice have been described previously (Wang et al., 2004).
In this study, we crossed the original VAMP8 knockout mouse line with the
cre-transgenic strain (Schwenk et al., 1995) to remove the neomycin resistance
cassette. The neo-deleted VAMP8 knockout line was then bred onto a 129/SvJ
background for five generations. Sex- and age-matched 3–5-mo-old adult mice
were used for experiments. Male mice were used unless otherwise indicated.
Histological Analysis and Immunofluorescence Microscopy
Organs were fixed with 4% paraformaldehyde (PFA) in PBS and embedded in
paraffin. Sections were stained with hematoxylin and eosin (H&E) or PAS. For
immunofluorescence studies, PFA-fixed organs were embedded in OCT com-
pound (Sakura Finetek USA, Torrance, CA). Cryosections were mounted on
polylysine-coated slides and stained with primary antibodies followed by
appropriate FITC-conjugated secondary antibodies (Jackson ImmunoResearch
Laboratories, West Grove, PA). Nuclei were stained with DAPI or TO-PRO-3
(Molecular Probes, Eugene, OR).
For morphological electron microscopy (EM) study, fresh tissues were fixed
with 4% PFA and 0.5% glutaraldehyde in PBS and embedded in Spurrs’ resin.
Ultrathin sections were stained with uranyl acetate and lead citrate before
being observed with a transmission electronic microscope.
Fresh tissues were homogenized in a modified RIPA buffer (10 mM Tris-HCl,
pH 7.5, 150 mM NaCl, 5 mM EDTA, 2 mM DTT, 1% deoxycholic acid, 1%
Triton X-100, 0.1% SDS, 1 mM PMSF, and Complete proteinase inhibitors
from Roche Diagnostics, Mannheim, Germany). Proteins were transferred
onto Hybond-C extra membranes (Amersham, Piscataway, NJ) after being
separated by SDS-polyacrylamide gels. Protein identities were determined
with specific primary antibodies and appropriate peroxidase-conjugated sec-
ondary antibodies (Jackson ImmunoResearch Laboratories). Peroxidase activ-
ity was detected with the SuperSignal substrate and enhancer (Pierce, Rock-
ford, IL) and visualized on x-ray films. Quantitative analysis was carried out
with a GS-800 calibrated densitometer from Bio-Rad (Hercules, CA).
Saliva Collection under Resting and
Mice were anesthetized with Avertin (tribromoethanol from Morre Technol-
ogy, Union, New Jersey), and saliva was then collected from the oral cavity
with microcapillaries. Otherwise, mice were first anesthetized with Avertin
and then injected with pilocarpine at 1 mg/kg i.p. (intraperitoneally) or 10 ?l
of 1 mg/ml pilocarpine at a location close to salivary glands. Saliva secreted
into the oral cavity was collected with microcapillaries at 5-min intervals until
30-min after injection. Fifty to 100 ?l of saliva was collected from each mouse.
The saliva volume of mutant mice was not overtly different from that of
GST Pulldown Analysis
The GST pulldown experiment was performed as previously described
(Wang et al., 2004). Briefly, salivary glands were homogenized in homogeni-
zation buffer (20 mM HEPES, pH 7.4, 10 mM sucrose, 10 mM KCl, 2 mM
EDTA, 2 mM EGTA, 6 mM MgCl2, 1 mM DTT, 1 mM PMSF, Complete
proteinase inhibitors, and 1 mg/ml GST or GST-VAMP8 fusion protein).
Nuclei were removed by centrifugation at 500 ? g for 5 min, and total
membranes were pelleted from the postnucleus supernatant by a spin at
100,000 ? g for 1 h. Membranes were washed in washing buffer (500 mM KCl,
20 mM HEPES, 1 mM DTT, 1 mM EDTA, 1 mM PMSF, Complete proteinase
inhibitors, 1 mg/ml GST or GST-VAMP8, pH 7.4) and then resuspended in 2
ml binding buffer (20 mM HEPES, 100 mM KCl, 1 mM DTT, 4 mM EGTA, 4
mM MgCl2, 2 mM ATP, 1 mM PMSF, Complete proteinase inhibitors, 1%
BSA, 1 mg/ml GST or GST-VAMP8, pH 7.4) and incubated at 37°C for 5 min.
After the incubation, the volume of the mixture was topped up to 12 ml with
protein-free binding buffer before a spin at 100,000 ? g for 1 h. Membranes
were resuspended in extraction buffer (20 mM HEPES, 100 mM KCl, 1 mM
DTT, 10 mM EDTA, 0.2 mM ATP, 2% Triton X-100, pH 7.4) and incubated at
4°C for 1 h with rotation. Triton-insoluble materials were removed by cen-
trifugation at 200,000 ? g for 30 min. Membrane extracts were incubated
overnight with glutathione Sepharose 4B beads (Amersham). Beads were
washed three times with extraction buffer containing 0.5% Triton followed by
three times with Triton-free buffer. All the procedures were carried out at 4°C
except binding. Proteins were eluted by boiling the beads for 5 min in SDS gel
loading buffer and then were subjected to Western blotting analysis.
Isolation of Protein Aggregates from Lacrimal Glands
Lacrimal glands were homogenized in 280 mM sucrose supplemented with 10
mM HEPES, pH 7.4, 1 mM PMSF, and the Complete proteinase inhibitor
(Roche Diagnostics) with a motor homogenizer (model T8.01; IKA Labortech-
nik, Staufen, Germany). The tissue suspension was then laid on top of a
discontinuous sucrose gradient that consisted of 2, 1.5, and 1.0 M sucrose.
Samples were centrifuged at 100,000 ? g for 1 h. The black band at the
interface between 2 and 1.5 M was retrieved and diluted with 2 volumes of 1%
Triton X-100. Protein aggregates were pelleted after a spin at 10,000 ? g for 5
min. The whole procedure was carried out at 4°C.
VAMP8 Is Required for Regulated Secretion in Salivary
The requirement of VAMP8 in regulated exocytosis of the
pancreatic acinar cells (Wang et al., 2004) prompted us to test
the hypothesis that VAMP8 may have a more general role in
the exocrine system.
Salivary glands, which consist primarily of parotid, subman-
dibular, and sublingual glands and secrete saliva into the
mouth to moisten the mouth, initiate food digestion, and help
protect the teeth from decay, were first examined. Immunoflu-
orescence studies with an antibody specifically recognizing
VAMP8 showed that VAMP8 was expressed in all acinar cells
orescence staining for VAMP8 in major salivary glands. VAMP8
was stained green with an FITC-conjugated secondary antibody.
Nuclei were either stained blue with DAPI or red with TO-PRO-3.
D, duct; A, acinus. Scale bar, 50 ?m. (B) Western blotting analysis of
VAMP8 and SNAP-23 expression in parotid, submandibular, and
sublingual glands. ?/?, wild-type mice; ?/?, VAMP8-null mice.
VAMP8 is expressed in salivary glands. (A) Immunoflu-
VAMP8/Endobrevin, a General v-SNARE
Vol. 18, March 20071057
and lumen-lining duct epithelial cells of the three major sali-
vary glands (Figure 1A). VAMP8 appeared to be enriched in
the apical region of acinar cells and duct epithelial cells. The
labeling is specific because no signal was detected in VAMP8-
null sections. VAMP8 expression in salivary glands was con-
firmed by immunoblot analysis (Figure 1B).
Visual examination of salivary glands revealed that the
parotid glands of VAMP8-null mice exhibited an abnor-
mally milky appearance in contrast to the slightly tan ap-
pearance of control mice (Figure 2A), a phenotype similar to
that previously observed in the pancreas (Wang et al., 2004).
H&E staining of parotid gland sections showed that the cyto-
plasm of normal acinar cells was stained purple, whereas that
of VAMP8-null cells appeared pink (Figure 2A), indicating
verify this possibility, we examined the distribution of secre-
tory granules in parotid glands by immunolabeling their con-
stituents such as amylase and carbonic anhydrase VI. As
shown in Figure 2A, both amylase and carbonic anhydrase
VI were restricted to the apical region in normal cells. In
contrast, they were evenly distributed throughout the entire
cytoplasm in VAMP8-null cells. These results are similar to
those observed earlier in the pancreas and can be explained
by the accumulation of secretory granules in the acinar cells
due to defective exocytosis. The overall accumulation of
secretory granules by VAMP8-null acinar cells was con-
firmed by EM analysis (Figure 2B). However, we also no-
ticed that the dense granules were reduced in mutant cells.
The cause for this reduction is unknown. We speculate that
it might be due to premature activation of enzymes, which
we have previously found in the pancreas (Wang et al.,
2004). Consistent with a possible exocytosis defect, amylase
and carbonic anhydrase VI were substantially increased in
mutant parotid glands (Figure 2C). These enzymes were also
increased in VAMP8-null submandibular glands, but the
difference was not as obvious as in parotid glands (Figure
2C). Because VAMP8 was detected in all the three primary
salivary glands, we speculated that VAMP8 might be re-
quired not only for secretion of enzymes by serous acinar
cells but also for secretion of mucin by mucous cells. PAS
staining for carbohydrates showed that VAMP8-null mu-
cous cells in sublingual glands and submandibular glands
contained more mucin than their wild-type counterpart (Fig-
ure 2D). Taken together, these results suggest that the ab-
sence of VAMP8 caused a defect in exocytosis, leading to the
accumulation of secretory granules in acinar cells.
If exocytosis was affected, we would expect some alter-
ations in the protein profile of saliva. Therefore, we com-
pared the saliva proteins of control and VAMP-null mice.
After resolving saliva proteins on a SDS-polyacrylamide gel,
we observed a dramatic difference between normal and
VAMP8-null male mice (Figure 3A). The difference for fe-
male mice was, however, subtle. There was also a prominent
sexual difference: the protein profiles of normal male and
female saliva were distinct. It was clear that several proteins
were underrepresented in the saliva of VAMP8-null male
mice. Mass spectrometry analysis of these protein bands
identified A15 as carbonic anhydrase VI (also called carbon-
ate dehydratase VI) and A17 as kallikrein 9 (KLK9 or KLK-L3).
Carbonic anhydrase VI is a zinc metalloenzyme that cata-
lyzes the reversible hydration of CO2to form HCO3?and H?
and therefore plays a role in regulating the pH of saliva. The
accumulation of carbonic anhydrase VI in the acinar cells of
VAMP8-null parotid glands (Figure 2) suggested that the
reduced level of this protein in the saliva was due, at least in
part, to its defective exocytosis. KLK9 is a recently identified
member of the kallikrein family of serine proteases. This
protein was abundant in normal male saliva but could
hardly be detected in VAMP8-null male mice (Figure 3A).
Saliva collected over a period of 4 h under resting condition
from control and VAMP8-null male and female mice was
also analyzed by immunoblot (Figure 3B). Consistent with
the results of mass spectroscopic analysis, KLK9 was abun-
dant in normal male saliva but hardly detected in the saliva
of VAMP8-null male mice. It was absent from female saliva
regardless of genotypes (Figure 3B). These results suggest
that KLK9 is likely a male-specific salivary protein whose
secretion depends on VAMP8. Unlike the strict VAMP8
dependence of KLK9 secretion, the levels of amylase and
carbonic anhydrase IV in saliva were only partially reduced
in mutant mice (Figure 3B), and the differences in these two
enzymes between female mice were not statistically signifi-
cant, suggesting that VAMP8 is not the primary v-SNARE
mediating the constitutive exocytosis of these two proteins.
Pilocarpine (PLP) is an alkaloid obtained from plants of
the genus Pilocarpus (family Rutaceae). It stimulates secre-
tion by the salivary and lacrimal glands by mimicking the
effects of acetylcholine. It is a cholinergic drug used to treat
xerostomia (dry mouth) and dry eyes caused by Sjo ¨gren’s
syndrome and radiation therapy for cancers of the head and
neck. To provide more direct evidence for a role of VAMP8
in regulated exocytosis of salivary glands, we examined the
saliva elicited by pilocarpine. Mice were administrated with
1 mg/kg pilocarpine. Saliva was collected for a period of 30
min so that regulated secretion of secretory proteins trig-
gered by pilocarpine could be analyzed. As observed for
saliva collected over a period of 4 h under resting conditions,
acute secretion of KLK9 stimulated by pilocarpine was de-
tected only in normal male mice. Under this condition, sal-
ivary levels of both amylase and carbonic anhydrase IV were
significantly reduced in VAMP8-null mice compared with
the sex-matched normal mice. The results are quantified and
shown in Figure 3C. The results clearly demonstrate that
pilocarpine-stimulated exocytosis of amylase and carbonic
anhydrase VI was severely compromised in the absence of
VAMP8 in both male and female mice, suggesting that
VAMP8 plays a key role in regulated exocytosis of secretory
proteins of salivary glands.
Previous studies of pancreatic acinar cells indicates that
VAMP8 may act as a vesicular SNARE (v-SNARE) of zymo-
gen granules and mediate exocytosis by interacting with the
target SNAREs (t-SNARE) syntaxin 4 and SNAP-23 on the
plasma membrane. To assess whether a similar mechanism
exists in salivary glands, we used a GST-VAMP8 fusion
protein in affinity coisolation analysis to test if VAMP8 could
interact with syntaxin 4 and SNAP-23. As shown in Figure
3D, syntaxin 4 and SNAP-23 were efficiently corecovered
with GST-VAMP8, whereas syntaxin6 and syntaxin13 were
Ablation of VAMP8 Leads to Accumulation of Secretory
Granules and Protein Aggregates in Lacrimal Glands
Lacrimal (tear) glands are responsible for secretion of lubri-
cation material for the eyes. VAMP8 expression was also
detected in the acinar cells of this gland (Figure 4A). VAMP8
was primarily distributed in the apical pole of acinar cells.
EM revealed accumulation of secretory granules in VAMP8-
null acinar cells (Figure 4B). Unlike normal acinar cells
whose secretory granules were restricted to the apical re-
gion, mutant cells had secretory granules distributed all over
the cytoplasm. The overall number of secretary granules was
obviously increased in VAMP8-null cells. This phenotype
suggests that VAMP8 might be required for exocytosis in
C.-C. Wang et al.
Molecular Biology of the Cell1058
Amylase and CA VI were stained green with a FITC-conjugated secondary antibody, and the nuclei were stained red. Arrows mark apical poles
cells of parotid glands. The dense granules were, however, substantially reduced in mutant cells, probably due to premature activation of enzymes
as described in the pancreas. Scale bar, 5 ?m. (C) Western blotting showing that both parotid and submandibular glands of VAMP8-null mice
accumulate amylase and CA VI. Quantitative studies on five pairs of mice are shown at the bottom. The average enzyme levels in wild-type mice
are arbitrarily defined as one unit. Error bars, SD. p ? 0.01 for all the four groups. (D) PAS staining showing that sublingual and submandibular glands
of VAMP8-null mice accumulate mucin (stained red). m, mucin acinar cells. Scale bar, 50 ?m. ?/?, wild-type mice; ?/?, VAMP8-null mice.
VAMP8-null salivary glands retain secretory granules and accumulate mucin and enzymes. (A) Morphological defects in VAMP8-null
VAMP8/Endobrevin, a General v-SNARE
Vol. 18, March 20071059
The gross appearance of VAMP-null lacrimal glands was
distinguishable from that of normal lacrimal glands: whereas
normal glands were translucent and tan to slightly gray, mu-
tant glands were gray and opaque, with dark spots of varying
sizes present throughout the entire glands (Figure 5A). For an
unknown reason, the relative density of mutant glands was
reduced. As shown in Figure 5E, mutant glands floated on the
glands. (A) Protein profiles of saliva resolved by SDS-PAGE. Lanes
1 and 6, protein markers 25, 33, and 47 are 25, 33, and 47 kDa,
respectively. ?/?, wild-type mice; ?/?, VAMP8-null mice. Bands
A15 and A17 were identified as CA VI and KLK9, respectively. (B)
Western blotting analysis showing amylase, carbonic anhydrase VI
(CAVI), and kallikrein 9 (KLK9) in basal and PLP-stimulated saliva.
?/?, wild-type mice; ?/?, VAMP8-null mice. An equal volume of
saliva was loaded on each lane. (C) Quantitative analysis of salivary
VAMP8 plays a role in protein secretion by salivary
levels of amylase and carbonic anhydrase IV (CAVI). The average
enzyme levels for wild-type samples are arbitrarily defined as one
unit. Error bars, SD. n ? 5. The values for basal saliva are as follows:
M?/? (amylase) ? 1.00 ? 0.123, M?/? (amylase) ? 0.844 ? 0.111,
p ? 0.069 ? 0.10; M?/? (CA VI) ? 1.00 ? 0.088, M?/? (CA VI) ?
0.813 ? 0.111, p ? 0.040 ? 0.05; F?/? (amylase) ? 1.00 ? 0.222,
F?/? (amylase) ? 0.814 ? 0.088, p ? 0.140 ? 0.10; F?/? (CA VI) ?
1.00 ? 0.100, F?/? (CA VI) ? 0.886 ? 0.166, p ? 0.226 ? 0.10. The
values for PLP-stimulated saliva are: M?/? (amylase) ? 1.00 ?
0.049, M?/? (amylase) ? 0.445 ? 0.108, p ? 0.01; M?/? (CA
VI) ? 1.00 ? 0. 172, M?/? (CA VI) ? 0.514 ? 0.143, p ? 0.01; F?/?
(amylase) ? 1.00 ? 0.182, F?/? (amylase) ? 0.545 ? 0.145, p ? 0.01;
F?/? (CA VI) ? 1.00 ? 0.192, F?/? (CA VI) ? 0.504 ? 0.165, p ?
0.01. ?/?, wild-type; ?/?, VAMP8-null. M, male; F, female. PLP,
PLP-stimulated saliva. (D) GST pulldown experiment showing the
interaction SANP23 and syntaxin4 with VAMP8.
granules in lacrimal gland acinar cells. (A) Immunofluorescence
staining showing VAMP8 expression in the apical region of lacrimal
gland acinar cells. VAMP8 was stained green, whereas the nuclei
are red. Arrows indicate apical poles of acinar cells. Scale bar, 50
?m. (B) Electron microscopy showing accumulation of secretory
granules in VAMP8-null acinar cells of lacrimal glands. Scale bar,
2 ?m. ?/?, wild-type; ?/?, VAMP8-null.
Ablation of VAMP8 led to accumulation of secretory
C.-C. Wang et al.
Molecular Biology of the Cell1060
surface in PBS but normal glands sank to the bottom. H&E
staining of lacrimal gland sections revealed accumulation of
dark-brown deposits in the lumen of acini and ducts (Figure 5,
C and D) of VAMP8-null glands, whereas there were no such
deposits in normal lacrimal glands (Figure 5B). We then at-
tempted to determine the identity of the deposits. After frac-
tionation of gland homogenate by a sucrose density gradient, a
interface between 2 and 1.5 M sucrose (Figure 5F). The corre-
sponding band in normal samples was only 10–20% in size
relative to the mutant band and could hardly be seen (Figure
completely dissolved in 1% SDS, suggesting that the major
constituents of the precipitates might be protein. SDS-PAGE
showed that the precipitates contained multiple proteins with
two major small peptides (Figure 5G). Mass spectrometry anal-
ysis identified peptide “a” as histone 1 and H2b and peptide
“b” as hemoglobin beta-1 subunit. The yellow to brown pig-
ment that ran just ahead of bromphenol blue was probably
heme. It seems that hemoglobin combined with heme is the
major constituent that produced the dark-brown to black color.
Although much less than in mutant glands, this kind of protein
aggregates was also present in normal glands, and these ag-
gregates were similar to those found in mutant glands, regard-
ing their constituents (Figure 5G). Histone is a well-known
component of tear, but hemoglobin has not been reported in
tear although some plasma proteins, such as albumin and
immunoglobulin, are abundant in the tear (Zhou et al., 2006).
Hemoglobin probably has escaped from detection because of
its low concentration and/or propensity to form aggregates
with histone. The source of tear plasma proteins is unclear. We
suspect that plasma proteins and hemoglobin are probably
derived from the blood. These proteins might somehow be
selectively picked up by acinar cells and/or duct cells and then
be transported into the lumen of acini and ducts. Whatever the
source of hemoglobin is, VAMP8 did not seem to be involved
in the trafficking of this protein because the protein deposits
were found outside cells. Under normal condition, most hemo-
globin might be flushed out of the lacrimal gland, whereas a
small portion forms fine deposits and provides the gland with
a light gray color. These tiny aggregates can hardly be ob-
served on H&E-stained sections. When secretion from acinar
cells is reduced, hemoglobin may accumulate and form large
aggregates in the lumen of acini and ducts. The defects de-
scribed above suggest that exocytosis of lacrimal acinar cells is
compromised in the absence of VAMP8, indicating a role for
VAMP8 in tear secretion.
VAMP8 Expression in Other Exocrine Tissues
VAMP8 was also detected in the prostate (Figure 6A) and
the exocrine glands of the skin (Figure 6B). In the prostate,
VAMP8 is primarily present in the apical region of the
lumen-lining epithelial gland cells (Figure 6A).The seba-
ceous glands, which secrete an oily substance called sebum
into the follicular canal and eventually onto the surface of the
skin, expressed VAMP8 (Figure 6B, indicated as Se). The ec-
lacrimal glands. (A) Gross appearance of VAMP8-null
(top) and wild-type (bottom) lacrimal glands together
with eyeballs. (B) H&E staining of normal lacrimal
gland section. (C) H&E staining of VAMP8-null lacrimal
gland section showing accumulation of dark-brown de-
posits. (D) High magnification of an H&E-stained sec-
tion of VAMP8-null lacrimal glands showing protein
deposits in the lumen of ducts and acini. Arrows indi-
cate lumen of the ducts. Scale bar, 50 ?m. (E) Photog-
raphy of lacrimal glands in PBS showing gravity density
difference between normal (?/?) and mutant (?/?)
lacrimal glands. (F) Sucrose density fractionation of lac-
rimal gland homogenate showing a dark-gray band at
the interface between 2 and 1.5 M sucrose (indicated by
an arrow). (G) SDS-PAGE showing protein components
in the deposits. From bottom to top, the molecular
weights of the protein markers (M) were 6.5, 16.5, 25,
32.5, 47.5, 63, 83, and 175 kDa, respectively. Bands were
identified by mass spectrometry as histone 1 and H2b
(a) and hemoglobin beta-1 subunit (b).
Protein deposits (aggregates) in VAMP8-null
VAMP8/Endobrevin, a General v-SNARE
Vol. 18, March 2007 1061
crine sweat glands, which secrete sweat to cool the body, also
expressed VAMP8 (Figure 6B, indicated as Sw). The detection
of VAMP8 in the sebaceous and sweat glands was specific
because no labeling was detected in VAMP8-null mice.
In the mammary gland, VAMP8 expression was regulated
during development and pregnancy (Figure 6C). Low levels
of expression were detected in 2-wk-old mice, and the level
of expression increased substantially in 10-wk-old mice. The
highest level of expression of VAMP8 was detected in mice
at the 18th day of pregnancy, and VAMP8 was seen to be
concentrated in the apical pole of the alveoli cells (Figure
6C). High and more diffuse expression of VAMP8 was de-
tected in lactating alveoli. Again, the detection is specific as
no labeling was observed in VAMP8-null mice (data not
Our findings that VAMP8 is crucially involved in the
regulated exocytosis of the parotid and tear glands and it is
widely expressed in all the exocrine tissues examined to-
gether with its established function in the pancreatic exo-
crine cells (Wang et al., 2004), suggest that VAMP8 may play
a role in regulated exocytosis of the entire exocrine system.
The SNARE proteins were initially discovered as an essen-
tial component of the secretory pathway. They were later
discovered to mediate release of neurotransmitters. It has
been well established that SNARE proteins function as en-
gines for membrane fusion (Jahn and Scheller, 2006). The
focus of the SNARE field has recently shifted to the regula-
tion of SNARE functions (Giraudo et al., 2006). Another
challenge in the field is to define the physiological role of
individual SNAREs. This kind of studies will advance our
understanding of human diseases. Actually, mutations in
trafficking proteins have already been shown to be the bases
of many human diseases. However, of the some 38 SNAREs
identified, only a few have been studied in whole animals. It
has been shown by gene-targeting that synaptobrevin/
VAMP2 is important for neurotransmitter release (Schoch et
al., 2001), whereas syntaxin4 is required for surface deploy-
ment of GLUT4 (Yang et al., 2001a). Epimorphin/syntaxin2
negatively regulates epithelium growth (Wang et al., 2006).
Despite its wide expression, VAMP3 seems to be nonessen-
tial (Yang et al., 2001b), and Vti1b plays only a minor role in
the endocytic pathway (Atlashkin et al., 2003). VAMP8 has
recently been shown to play an important role in regulated
secretion of pancreas and platelets (Wang et al., 2004; Ren et
al., 2006). In this study, we extend our investigation into the
role of VAMP8 in the whole exocrine system.
Saliva has lubricative, protective, and digestive functions
and is produced by various salivary glands associated with
the oral cavity. The parotid, submandibular, and sublingual
glands represent the major salivary glands with indepen-
dent excretory ducts, whereas minor salivary glands are
located throughout the oral mucosa and the tongue. Based
on secreted materials, salivary glands are described as se-
rous or mucous glands. Serous cells secret a thin watery
fluid rich in enzymes, whereas mucous cells produce a thick
viscid secretion containing a large amount of glycoproteins.
The parotid gland is predominantly composed of serous
acini, whereas the sublingual gland is mainly mucous and
the submandibular gland is a mixture of serous and mucous
cells. VAMP8 was expressed in all three major salivary
glands. Two lines of evidence support a role for VAMP8 in
exocytosis of salivary glands. First, mucin and enzymes such
as amylase and carbonate anhydrase VI accumulated in
VAMP8-null glands due to accumulation of secretory gran-
ules. Second, mutant saliva protein levels were substantially
reduced, especially in pilocarpine (PLP)-stimulated saliva.
In our study, PLP-induced secretion was measured within a
period of 30 min compared with a period of 4 h for basal
secretion. As such, we were measuring acute regulated exo-
cytosis by the major salivary glands. Under this condition,
basal secretion is less significant than PLP-stimulated secre-
tion. The result is the key evidence suggesting that VAMP8
is more important for regulated secretion than for constitu-
tive secretion. This interpretation is supported by other mor-
phological (including EM) analysis. In line with earlier anal-
ysis of pancreatic acinar cells (Wang et al., 2004), VAMP8 is
likely to act as the v-SNARE of the secretory granule in
salivary glands as well. Because the difference between nor-
mal and VAMP8-null saliva amylase and CA VI are more
fluorescent labeling showing VAMP8 expression in (A) prostate, (B)
sebaceous glands (Se) and sweat glands (Sw), and (C) mammary
glands at different development stages and under different physiolog-
ical conditions. VAMP8 was stained green. Nuclei were stained red or
unstained. ?/?, wild-type; ?/?, VAMP8-null. Scale bar, 50 ?m.
VAMP8 expression in other exocrine tissues. Immuno-
C.-C. Wang et al.
Molecular Biology of the Cell1062
significant in PLP-stimulated secretion than in constitutive Download full-text
secretion, we conclude that VAMP8 is more important for
regulated exocytosis than for basal secretion.
KLK9 was molecularly identified in 2000 (Diamandis et al.,
2000; Yousef and Diamandis, 2000), but its function remains
to be determined. Our results here suggest that KLK9 is
likely a male-specific salivary protein whose secretion is
strictly dependent on VAMP8. In contrast to the partially
reduced secretion of well-characterized salivary proteins
such as amylase and carbonic anhydrase VI, KLK9 in saliva
was essentially abolished in the absence of VAMP8. This
disparity between KLK9 and amylase (or carbonic anhy-
drase VI) suggests that KLK9 and amylase (or carbonic
anhydrase VI) might be present in different secretary gran-
ules that express distinct sets of v-SNAREs. In fact, it is well
known that there are different types of secretory granules
that response to different stimuli (Castle et al., 2002). An
interesting question to be answered in future studies is
whether the secretion of other KLKs in salivary glands and
other tissues is also dependent on VAMP8.
The cornea of the eye is covered by a tear film which keeps
the eye moist and washes away dirty and small particles.
The tear film consists of three layers, with the middle aque-
ous layer of tear fluid being composed of water, salt, pro-
teins and other substances that help to lubricate the eye
(Zhou et al., 2006). Tear fluid is primarily formed in the
lacrimal tear glands; it is secreted by the tubuloalveolar
acinar cells and transported via a number of ducts to the
eye’s surface. In the absence of VAMP8, tear secretion might
be compromised, although no overt injury has been ob-
served. Consistent with this is the cellular observation that
acinar cells accumulated secretory granules, indicating that
exocytosis was defective. Furthermore, the lacrimal glands of
VAMP8 knockout mice accumulate dark brown protein aggre-
gates in the lumen of the acinus and duct, which may partially
affect the flow of tear. These defects ultimately result from the
lack of VAMP8 that is important for fusion of secretory gran-
ules with the plasma membrane. VAMP2, VAMP7, and
VAMP8 have been proposed to be candidate v-SNAREs in
lacrimal glands (Wu et al., 2006). Our study shows that VAMP8
is a major if not the only v-SNARE of lacrimal glands.
In human, there is a disease called Sjo ¨gren’s syndrome.
Patients of this syndrome suffer from dry mouth and dry
eyes. Because the PLP-stimulated saliva volume was not
significantly different between normal mice and VAMP8
knockout mice, despite the difference in protein composi-
tion, we do not believe the VAMP8 knockout mouse line is
a perfect model for human Sjo ¨gren’s syndrome. But human
with defective VAMP8 might suffer from a mild symptom of
dry mouth and dry eyes.
Our current study suggests that VAMP8 is important for
regulated exocytosis of salivary glands and lacrimal glands,
in addition to its role in regulated exocytosis of pancreatic
acinar cells (Wang et al., 2004). In the context of its wide
expression in all other exocrine tissues examined, we believe
that VAMP8 may also participate in regulated secretion of
additional exocrine tissues. It is reasonable to speculate that
VAMP8 is likely a common v-SNARE responsible for regu-
lated exocytosis in the entire exocrine system.
We thank staff of the Biological Resource Center (BRC) for maintaining the
VAMP8 knockout mice. We thank Singh Paramjeet for reading the manu-
script. This work is funded by Agency for Science, Technology, and Research
Atlashkin, V., Kreykenbohm, V., Eskelinen, E. L., Wenzel, D., Fayyazi, A., and
Fischer von Mollard, G. (2003). Deletion of the SNARE vti1b in mice results in
the loss of a single SNARE partner, syntaxin 8. Mol. Cell. Biol. 23, 5198–5207.
Barr, F. (2000). Vesicular transport. Essays Biochem. 36, 37–46.
Bonifacino, J. S., and Glick, B. S. (2004). The mechanisms of vesicle budding
and fusion. Cell 116, 153–166.
Castle, A. M., Huang, A. Y., and Castle, J. D. (2002). 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.
Chandra, S., Gallardo, G., Fernandez-Chacon, R., Schluter, O. M., and Sudhof,
T. C. (2005). Alpha-synuclein cooperates with CSPalpha in preventing neu-
rodegeneration. Cell 123, 383–396.
Diamandis, E. P., et al. (2000). New nomenclature for the human tissue
kallikrein gene family. Clin. Chem. 46, 1855–1858.
Giraudo, C. G., Eng, W. S., Melia, T. J., and Rothman, J. E. (2006). A clamping
mechanism involved in SNARE-dependent exocytosis. Science 313, 676–680.
Hong, W. (2005). SNAREs and traffic. Biochim. Biophys. Acta 1744, 493–517.
Jahn, R., and Scheller, R. H. (2006). SNAREs—engines for membrane fusion.
Nat. Rev. Mol. Cell Biol. 7, 631–643.
Jahn, R., and Sudhof, T. C. (1999). Membrane fusion and exocytosis. Annu.
Rev. Biochem. 68, 863–911.
McNiven, M. A., and Thompson, H. M. (2006). Vesicle formation at the
plasma membrane and trans-Golgi network: the same but different. Science
Ren, Q., Barber, H. K., Crawford, G. L., Karim, Z. A., Zhao, C., Choi, W.,
Wang, C.-C., Hong, W., and Whiteheart, S. W. (2007). Endobrevin/VAMP-8 Is
the Primary v-SNARE for the Platelet Release Reaction. Mol. Biol. Cell 18,
Rettig, J., and Neher, E. (2002). Emerging roles of presynaptic proteins in
Ca??-triggered exocytosis. Science 298, 781–785.
Rizo, J., Chen, X., and Arac, C. (2006). Unraveling the mechanisms of synap-
totagmin and SNARE function in neurotransmitter release. Trends Cell Biol.
Schwenk, F., Baron, U., and Rajewsky, K. (1995). A cre-transgenic mouse
strain for the ubiquitous deletion of loxP-flanked gene segments including
deletion in germ cells. Nucleic Acids Res. 23, 5080–5081.
Schoch, S., Deak, F., Konigstorfer, A., Mozhayeva, M., Sera, Y., Sudhof, T. C.,
and Kavalali, E. T. (2001). SNARE function analyzed in synaptobrevin/VAMP
knockout mice. Science 294, 1117–1122.
So ¨llner, T., Whiteheart, S. W., Brunner, M., Erdjument-Bromage, H., Geroma-
nos, S., Tempst, P., and Rothman, J. E. (1993). SNAP receptors implicated in
vesicle targeting and fusion. Nature 362, 318–324.
Wang, C.-C., Ng, C. P., Lu, L., Atlashkin, V., Zhang, W., Seet, L.-F., and Hong,
W. (2004). A role of VAMP8/endobrevin in regulated exocytosis of pancreatic
acinar cells. Dev. Cell 7, 359–371.
Wang, Y., Wang, L., Iordanov, H., Swietlicki, E. A., Zheng, Q., Jiang, S., Tang,
Y., Levin, M. S., and Rubin, D. C. (2006). Epimorphin(?/?) mice have
increased intestinal growth, decreased susceptibility to dextran sodium sul-
fate colitis, and impaired spermatogenesis. J. Clin. Invest. 116, 1535–1546.
Wong, S. H., Zhang, T., Xu, Y., Subramaniam, V. N., Griffiths, G., and Hong,
W. (1998). Endobrevin, a novel synaptobrevin/VAMP-like protein preferen-
tially associated with the early endosome. Mol. Biol. Cell 9,1549–1563.
Wu, K., Jerdeva, G. V., da Costa, S. R., Sou, E., Schechter, J. E., and Hamm-
Alvarez, S. F. (2006). Molecular mechanisms of lacrimal acinar secretory
vesicle exocytosis. Exp. Eye Res. 83, 84–96.
Yang, C., Coker, K. J., Kim, J. K., Mora, S., Thurmond, D. C., Davis, A. C.,
Yang, B., Williamson, R. A., Shulman, G. I., and Pessin, J. E. (2001a). Syntaxin
4 heterozygous knockout mice develop muscle insulin resistance. J. Clin
Invest. 107, 1311–1318.
Yang, C., Mora, S., Ryder, J. W., Coker, K. J., Hansen, P., Allen, L. A., and
Pessin, J. E. (2001b). VAMP3 null mice display normal constitutive, insulin-
and exercise-regulated vesicle trafficking. Mol. Cell. Biol. 21, 1573–1580.
Yousef, G. M., and Diamandis, E. P. (2000). The expanded human kallikrein
gene family: locus characterization and molecular cloning of a new member,
KLK-L3 (KLK9). Genomics 65, 184–194.
Zhou, L., Beueman, R. W., Foo, Y., Liu, S., Ang, L. P., and Tan, D. T. (2006).
Characterisation of human tear proteins using high-resolution mass spec-
trometry. Ann. Acad. Med. Singapore 35, 400–407.
VAMP8/Endobrevin, a General v-SNARE
Vol. 18, March 2007 1063