RESEARCH ARTICLE
1757
Rabphilin dissociated from Rab3 promotes
endocytosis through interaction with Rabaptin-5
Thierry Coppola1, Harald Hirling1, Véronique Perret-Menoud1, Sonia Gattesco1, Stefan Catsicas1,*,
Gérard Joberty2, Ian G. Macara2 and Romano Regazzi1,‡
1Institut de Biologie Cellulaire et de Morphologie, University
2Markey Center for Cell Signalling, Health Sciences Center,
of Lausanne, Switzerland
University of Virginia, Charlottesville, VA 22908, USA
*Present address: Swiss Federal Institute of Technology, EPFL Lausanne, Switzerland
‡Author for correspondence (e-mail: [email protected])
Accepted 21 February 2001
Journal of Cell Science 114, 1757-1764 © The Company of Biologists Ltd
SUMMARY
Rabphilin is a secretory vesicle protein that interacts with
the GTP-bound form of the small GTPase Rab3. We
investigated the involvement of Rabphilin in endocytosis
using different point mutants of the protein.
Overexpression of wild-type Rabphilin in the insulinsecreting cell line HIT-T15 did not affect receptor-mediated
transferrin endocytosis. By contrast, Rabphilin V61A, a
mutant that is unable to interact with Rab3, enhanced the
rate of transferrin internalization. The effect of Rabphilin
V61A was not mimicked by Rabphilin L83A, another
mutant with impaired Rab3 binding. Careful analysis of
the properties of the two mutants revealed that Rabphilin
V61A and Rabphilin L83A are both targeted to secretory
vesicles, have stimulatory activity on exocytosis, and bind
equally well to α-actinin. However, Rabphilin L83A fails to
interact with Rabaptin-5, an important component of
the endocytotic machinery. These results indicate that
Rabphilin promotes receptor-mediated endocytosis and
that its action is negatively modulated by Rab3. We propose
that the hydrolysis of GTP that is coupled to the exocytotic
event disrupts the Rabphilin-Rab3 complex and permits
the recruitment of Rabaptin-5 at the fusion site. Our data
show that immediately after internalization the transferrin
receptor and VAMP-2 colocalize on the same vesicular
structures, suggesting that Rabphilin favors the rapid
recycling of the components of the secretory vesicle.
INTRODUCTION
Rab3 forms a cytosolic complex with the regulatory protein
Rab GDP-dissociation inhibitor (RabGDI) (Regazzi et al.,
1992, Ullrich et al., 1993). RabGDI prevents the binding of
Rab3 to inappropriate cellular compartments and delivers the
protein to the secretory vesicle membrane (Pfeffer et al., 1995).
There, through interaction with the guanine nucleotide
exchange factor Rab3-GEP (Wada et al., 1997), Rab3
substitutes GDP for GTP. This causes a conformational change
that allows interaction with specific effector proteins. During
or immediately after fusion of the secretory vesicle with the
plasma membrane, a regulatory protein called Rab3-GAP
enhances the intrinsic enzymatic activity of Rab3 and promotes
the hydrolysis of GTP (Fukui et al., 1997). Once this has
occurred, GDP-bound Rab3 reforms the complex with RabGDI
and dissociates from the membrane.
Rabphilin is a 78 kDa protein purified by chemical
crosslinking of Rab3 with bovine brain extracts (Shirataki et
al., 1992). Rabphilin is associated with the membrane of
secretory vesicles and interacts selectively with the GTP-bound
form of Rab3 (Shirataki et al., 1993; Stahl et al., 1996).
Structural and biochemical analyses indicate that Rabphilin
can be subdivided into two functionally distinct domains
(Yamaguchi et al., 1993; Ostermeier and Brunger, 1999). The
N-terminal portion contains an amphipathic α-helix required
for Rab3 binding, whereas the C-terminal region possesses two
C2-like domains that bind Ca2+ and phospholipids. Several
observations indicate that Rabphilin performs major regulatory
Tight coupling between exocytosis and endocytosis is required
to prevent excessive enlargement of the plasma membrane and
to permit rapid recycling of the components of the secretory
vesicle. This is most dramatically demonstrated in neurons,
where impairment of endocytosis leads to rapid depletion of
the secretory vesicle pool (Koenig and Ikeda, 1989). The
mechanism ensuring the spatial and temporal link between
exo- and endocytosis is not well understood. The observation
that synaptotagmin binds with high affinity to the clathrin
adaptor complex AP-2 led to the proposal that the coupling
between the two processes could be provided by proteins
associated with the membrane of secretory vesicles (Zhang et
al., 1994). According to this model, the insertion of these
proteins in the plasma membrane would provide the signal that
triggers endocytosis. Multidomain adaptor molecules such as
intersectin that bind concomitantly to components of the
exocytotic and endocytotic machinery (Okamoto et al., 1999)
could contribute an additional connection between the two
pathways.
The four isoforms of the small GTPase Rab3 modulate
exocytosis in most secretory systems, including neurons and
exocrine and endocrine cells (Lledo et al., 1993; Holz et al.,
1994; Johannes et al., 1994; Geppert et al., 1997; Ohnishi et
al., 1997; Iezzi et al., 1999). Rab3 undergoes a functional cycle
that is coupled to GTP hydrolysis. In the GDP-bound form,
Key words: GTP, Recycling, Exocytosis
1758
JOURNAL OF CELL SCIENCE 114 (10)
functions in exocytosis. Overexpression of the full-length
protein stimulates exocytosis in chromaffin and insulinsecreting cells (Chung et al., 1995; Arribas et al., 1997; Joberty
et al., 1999). By contrast, fragments of Rabphilin lacking one
or both C2 domains act as potent dominant inhibitors of
secretion (Chung et al., 1995). In addition, microinjection of
the isolated N- or C-terminal parts of Rabphilin inhibits Ca2+triggered cortical granule exocytosis in mouse eggs (Masumoto
et al., 1998). The precise mechanism of action of Rabphilin in
exocytosis is still unclear. The protein was initially postulated
to mediate Rab3 action. However, more recent data clearly
indicate that Rab3 and Rabphilin can modulate secretion
independently. Thus, mutants of Rabphilin that do not interact
efficiently with Rab3 can still potentiate stimulus-induced
secretion (Joberty et al., 1999). Conversely, mutants of Rab3
that cannot bind to Rabphilin inhibit exocytosis (Chung et al.,
1999; Coppola et al., 1999). These observations are further
substantiated by the fact that the synaptic properties that are
impaired in Rab3A-deficient mice are not altered in Rabphilin
knock-out mice (Schlüter et al., 1999). A yeast two-hybrid
screen revealed that the N-terminal portion of Rabphilin can
bind to α-actinin (Kato et al., 1996). The binding of Rabphilin
enhances the actin filament bundling activity of α-actinin
(Kato et al., 1996). Interestingly, the GTP-bound form of Rab3
competes for the binding of α-actinin (Kato et al., 1996). These
findings suggest that at least part of the function of Rabphilin
is mediated by the reorganization of actin filaments.
In addition to its effect on exocytosis, Rabphilin has been
suggested to participate in the control of endocytosis. Thus, the
membrane area of endosomes is increased following injection
of full-length Rabphilin in squid giant synapse, whereas
injection of the N-terminal fragment prevents membrane
retrieval from the plasma membrane (Burns et al., 1998). Yeast
two-hybrid screening showed that Rabphilin can bind to
Rabaptin-5, a protein that associates with the GTP-bound form
of Rab5 (Ohya et al., 1998). The interaction with Rabaptin-5
occurs through the N-terminal fragment of Rabphilin and is
prevented by the presence of the GTP-bound form of Rab3
(Ohya et al., 1998).
In this study, we took advantage of different point mutants
of Rabphilin to investigate the role of this protein in
endocytosis. We found that a Rabphilin mutant that is unable
to bind Rab3 can promote transferrin endocytosis by a
mechanism that requires interaction with Rabaptin-5. We
propose that Rabphilin plays a dual role in vesicular trafficking
at the plasma membrane by participating in the regulation of
distinct stages of the exocytic and endocytic pathways.
MATERIALS AND METHODS
The generation of Rabphilin mutants has been described previously
(Joberty et al., 1999). Human transferrin receptor cDNA obtained
from L. Kuhn (Swiss Cancer Research Institute, Epalinges) was
subcloned in pcDNA3 (Invitrogen). Epitope-tagged full-length human
Rabaptin-5 in pcDNA3 was kindly provided by G. Grenningloh
(University of Lausanne). The 385-753 fragment of Rabaptin-5
amplified by PCR from the full-length cDNA was subcloned in
pcDNA3. The VAMP-2-Tag plasmid and the monoclonal antibody
(KT3) against the T antigen tag were obtained from R. B. Kelly
(University of California, San Francisco). Human transferrin (hTfr)
coupled to FITC was from Molecular Probes; hTfr coupled to
horseradish peroxidase (HRP) was purchased from Pierce. The Cy3conjugated anti-mouse antibody was from Jackson Immuno Research
Laboratories. Purified chicken gizzard α-actinin and the antibody
against α-actinin were obtained from Sigma.
Cell culture and transfection
HIT-T15 cells were cultured in RPMI 1640 supplemented with 5%
fetal calf serum as described previously (Regazzi et al., 1990).
Transient cotransfection was performed by electroporating 3×106
cells in the presence of 30 µg of each plasmid (Coppola et al., 1999).
Immediately after electroporation, the cells were resuspended in
culture medium and distributed in 24-multiwell plates at a
concentration of approximately 3×105 cells/well. The expression level
of each Rabphilin construct was assessed by western blotting using a
mouse monoclonal antibody directed against the HA epitope tag
(12CA5).
Subcellular localization of hTfr and VAMP-2
HIT-T15 cells were transiently transfected with hTfr or, for the
experiments shown in Fig. 9, with both hTfr and VAMP-2-Tag (Grote
and Kelly, 1996). For the experiments in Fig. 1, the cells cultured for
3 days on glass coverslips were pre-incubated in culture medium for
1 hour at for 4°C with hTfr coupled to FITC. For the experiments
shown in Fig. 9, the cells were pre-incubated at 4°C for 45 minutes
with hTfr coupled to FITC, and with a monoclonal antibody directed
against the T antigen tag (KT3). After washing, the cells were exposed
for 30 minutes at 4°C to a Cy3-labeled anti-mouse antibody in the
constant presence of hTfr-FITC.
At the end of the different pre-incubation procedures the coverslips
were washed and shifted at 37°C for the indicated times. The cells
were fixed in 4% paraformaldehyde and analyzed by confocal
microscopy (model TCS NT, Leica, Lasertechnik, Heidelberg,
Germany). Excitation was obtained with an Argon-Krypton laser, with
line set at 488 nm for fluorescein excitation and 588 nm for rhodamine
excitation. The emitted light was filtered through appropriate filters
(BF 530/30 for fluorescein, LP 590 for rhodamine). Single section
images of 512×512 pixels were taken with a 40× objective, numerical
aperture 1.32.
Measurement of transferrin endocytosis
HIT-T15 cells transiently cotransfected with a plasmid encoding the
hTfr receptor and with the Rabphilin constructs were seeded in 24multiwell plates at a concentration of about 2×105 cell/well. Three
days after transfection, the cells were incubated in culture medium for
60 minutes at 4°C in the presence of 10 µg/ml hTfr coupled to HRP.
After this period, the cells were washed once in culture medium and
four times in PBS to eliminate unbound hTfr. They were then either
kept at 4°C for another 5 minutes or incubated for 2-5 minutes at 37°C
in culture medium. After the incubation, the cells were washed three
times at 4°C with PBS or with PBS supplemented with 30 mM glycine
at pH 2.4. At the end of this procedure, the cells were lysed at room
temperature in 100 µl PBS supplemented with 1% Triton-X 100. The
amount of hTfr bound to the surface of the cells (washes with PBS)
or internalized (washes with PBS supplemented with glycine) was
assessed using a colorimetric assay. For this purpose, an aliquot of the
lysate (5-10 µl) was diluted in 50 µl 3,3′,5,5′-tetramethylbenzidine
and incubated for 15 minutes at room temperature. The reaction was
stopped by adding 50 µl stop solution containing 1.5 M H2SO4 and 1
M HCl. HRP activity was estimated by measuring the absorbance at
450 nm.
Assessment of the interaction of Rabphilin mutants with
Rab3A
The ability of Rabphilin mutants to bind to Rab3A was tested using
a mammalian two-hybrid system (CheckMateTM, Promega). For this
purpose, the full-length cDNAs of the Rabphilin mutants were
subcloned in frame with VP16 in the expression vector pACT,
Involvement of Rabphilin in endocytosis
1759
whereas the constitutively active Rab3A Q81L mutant was subcloned
in frame with GAL4 in the expression vector pBIND. The GAL4 and
VP16 fusion proteins were cotransfected in HIT-T15 cells along with
a third plasmid (pG5luc) encoding five GAL4 binding sites upstream
of the firefly luciferase gene. Three days after transfection, the cells
were lysed and the amount of firefly luciferase was quantitated using
the Dual-LuciferaseTM Reporter Assay System (Promega).
Interaction of Rabphilin mutants with α-actinin
Approximately 20 pmol of bacterially expressed fusion proteins
between Glutathion-S-transferase (GST) and the N-terminal fragment
(amino acids 1-206) of wild-type or mutated Rabphilin were
immobilized on glutathion-agarose beads (Sigma). The affinity
columns were incubated for 90 minutes at 4°C with 100 pmol of
purified α-actinin in 20 mM Tris-HCl, pH 7.5, 1 mM dithiothreitol, 1
mM EDTA, 0.26% CHAPS and 150 mM NaCl. After five washes in
the same buffer the proteins remaining attached to the beads were
analyzed by SDS-PAGE. The α-actinin associated with the affinity
columns was detected by immunoblotting using a polyclonal antibody
directed against the purified protein.
Interaction of Rabphilin mutants with Rabaptin-5
For analysis of the ability of different mutants of Rabphilin to bind to
Rabaptin-5, the GST-affinity columns described above were incubated
in the presence of [35S]-labeled Rabaptin-5 (amino acids 385-753)
generated by in vitro translation (Promega).
RESULTS
The involvement of Rabphilin in the regulation of endocytosis
was investigated using cells of the hamster insulin-secreting
line HIT-T15 that were transiently cotransfected with plasmids
encoding the human transferrin receptor and different
Rabphilin mutants. As shown in Fig. 1, if 3 days after
transfection the cells were incubated at 4°C in the presence of
human transferrin (hTfr) coupled to FITC, the plasma
membrane of the cells expressing the receptor was
fluorescently labeled (Fig. 1A). The binding of hTfr to
untransfected hamster cells was undetectable (not shown). As
expected, the labeling observed on transfected cells was
abolished by acid washes, indicating that, under these
conditions, hTfr was retained on the surface of the cells (Fig.
1B). If after the initial labeling at 4°C the cells were incubated
for 5 minutes at 37°C, the fluorescence localized on punctate
structures inside the cells. In this case the signal was resistant
to acid washes confirming that hTfr was internalized (Fig. 1C).
To measure the amount of hTfr bound on the cell surface
at 4°C or internalized during the incubation at 37°C we
performed experiments using hTfr coupled to horseradish
peroxidase. As shown in Fig. 2, the amount of hTfr bound at
4°C in cells transfected with hTfr receptor was four to five
times higher than that in cells transfected with the vector alone.
As expected, the peroxidase activity was reduced to
background levels after acid washes. By contrast, if the cells
were incubated for 5 minutes at 37°C prior to the acid
treatment, a fraction of the peroxidase activity was internalized
and recovered in the cell extract (Fig. 2).
This approach was used to measure the effect on endocytosis
of wild-type Rabphilin and of two point mutants, Rabphilin
R60A and Rabphilin V61A. The R60A mutant displays wildtype affinity for Rab3 and is targeted to secretory vesicles but
has no stimulatory effect on exocytosis (Joberty et al., 1999).
Fig. 1. Subcellular localization of hTfr coupled to FITC in HIT-T15
cells expressing the hTfr receptor. HIT-T15 cells transiently
transfected with the hTfr receptor were incubated for 60 minutes at
4°C with FITC-coupled hTfr. They were then washed and either kept
for another 5 minutes at 4°C (A,B) or transferred for 5 minutes at
37°C (C). At the end of the incubation the cells were washed with
PBS (A) or PBS supplemented with 30 mM glycine pH 2.4 (B,C)
and fixed with paraformaldehyde. The localization of fluorescently
labeled hTfr was analyzed by confocal analysis. The figure shows
one representative cell for each condition.
Fig. 2. Measurement of transferrin endocytosis in HIT-T15 cells
expressing the hTfr receptor. HIT-T15 cells were transfected with
vector alone (open bars) or with the hTfr receptor (filled bars). Three
days later the cells were pre-incubated for 60 minutes at 4°C with
hTfr coupled to HRP. They were then kept for another 5 minutes at
4°C or transferred for 5 minutes at 37°C. At the end of the incubation
the cells were washed with PBS or PBS supplemented with 30 mM
glycine pH 2.4 (+G). The amount of HRP-coupled hTfr bound at the
surface of the cells or internalized was determined using a
colorimetric assay as described in Materials and Methods.
1760
JOURNAL OF CELL SCIENCE 114 (10)
Fig. 3. Effect of different Rabphilin mutants on transferrin
endocytosis. HIT-T15 cells were transiently cotransfected with a
plasmid encoding the hTfr receptor and with an empty vector
(control) or plasmids encoding wild-type Rabphilin (wt), Rabphilin
R60A or Rabphilin V61A. Three days later the cells were preincubated for 60 minutes at 4°C in the presence of hTfr coupled to
HRP, washed with PBS and transferred for 2 or 5 minutes at 37°C.
At the end of the incubation the cells were washed with PBS
supplemented with 30 mM glycine at pH 2.4 and lysed with 1%
Triton-X 100. Internalized hTfr is expressed as a percentage of the
amount of hTfr bound to the surface of the cells during the 60
minutes pre-incubation at 4°C.
The V61A mutant cannot bind to Rab3 and is targeted to
secretory vesicles less efficiently than wild-type Rabphilin but
retains the ability to enhance K+-induced exocytosis (Joberty
et al., 1999). As shown in Fig. 3, hTfr endocytosis was not
affected by the overexpression of wild-type Rabphilin or the
R60A mutant. By contrast, in cells transfected with the V61A
mutant, the amount of hTfr internalized after 2 or 5 minutes at
37°C was approximately doubled. This indicates that
Rabphilin is indeed involved in endocytosis and suggests that
this function is negatively modulated by interaction with Rab3.
We attempted to clarify the role of Rabphilin in the
endocytotic process by taking advantage of another Rabphilin
mutant that cannot interact with Rab3. This mutant was
selected using a so-called mammalian two-hybrid system. For
this purpose, we generated fusion proteins between the
transcriptional activator VP16 and Rabphilin, and between
Gal4 and Rab3A. These constructs were transiently
cotransfected in HIT-T15 cells together with a plasmid
encoding the Gal4 binding sequence upstream of the luciferase
gene. As expected, transfection of the VP16/Rabphilin or
Gal4/Rab3A proteins alone led to a poor expression of the
luciferase gene (Fig. 4). However, co-expression of the fusion
protein including wild-type Rabphilin with Gal4/Rab3A
resulted in a strong induction of the reporter gene (Fig. 4). By
contrast, the fusion protein between VP16 and the V61A
mutant caused no induction of the luciferase gene, confirming
that this mutant cannot bind efficiently to Rab3 (Joberty et al.,
1999). A VP16 fusion protein including a Rabphilin mutant in
which leucine 83 was replaced by alanine (L83A) was also
inactive, indicating that this mutant cannot interact with Rab3
(Fig. 4).
We then tested the effect of the L83A mutant on endocytosis
(Fig. 5). In this series of experiments, the V61A mutant
enhanced hTfr internalization by about two-fold. This effect
was comparable with that obtained after overexpression of
wild-type Rabaptin-5, a known component of the endocytotic
Fig. 4. Interaction of Rabphilin mutants with Rab3A. HIT-T15 cells
were cotransfected with a plasmid containing five GAL4 binding
sites upstream of the firefly luciferase gene, a fusion protein of
GAL4 with Rab3A Q81L (Rab3A), and a fusion protein of VP16
with the indicated mutants of Rabphilin. The interaction between
Rabphilin mutants and Rab3A was quantitated by measuring the
firefly luciferase activity. The luciferase activity produced by the
association of the wild-type Rabphilin construct with Rab3A Q81L
was set at 100%. The figure shows the mean±s.d. of a representative
experiment.
machinery (Fig. 5). Surprisingly, however, the L83A mutant
had no significant effect on endocytosis.
In an attempt to elucidate the cause(s) of the different
behaviors of Rabphilin V61A and Rabphilin L83A on
endocytosis, the properties of the two mutants were analyzed
in detail. The lack of effect of Rabphilin L83A was not due to
an instability of the protein, because all the Rabphilin
constructs used in the study were expressed to similar levels
(Fig. 6). We have shown previously that both Rabphilin V61A
and Rabphilin L83A have stimulatory effects on K+-triggered
Fig. 5. Effect of Rabphilin L83A on transferrin endocytosis. HIT-T15
cells were transiently cotransfected with a plasmid encoding the hTfr
receptor and with an empty vector (control), or a plasmid encoding
Rabphilin V61A, Rabphilin L83A or Rabaptin-5. Three days later
the cells were pre-incubated for 60 minutes at 4°C in the presence of
hTfr coupled to HRP, washed with PBS and transferred for 2 minutes
at 37°C. At the end of the incubation the cells were washed with PBS
supplemented with 30 mM glycine at pH 2.4 and lysed with 1%
Triton-X 100. Internalized hTfr is expressed as percentage of the
amount of hTfr bound to the surface of the cells during the 60 minute
pre-incubation at 4°C. The results are the mean±s.d. of one
representative experiment out of four.
Involvement of Rabphilin in endocytosis
Fig. 6. Expression level of Rabphilin constructs in HIT-T15 cells.
HIT-T15 cells were transiently transfected with an empty vector
(control), wild-type Rabphilin, Rabphilin R60A, Rabphilin V61A
and Rabphilin L83A. Three days later the cells were homogenized
and the proteins analyzed by western blotting using a monoclonal
antibody against the HA epitope tag.
exocytosis (Joberty et al., 1999). In addition, we have reported
that in PC12 cells both proteins associate with secretory
vesicles but their targeting efficiency is reduced (Joberty et al.,
1999). To verify that these observations were not restricted to
PC12 cells we analyzed by confocal microscopy the
subcellular localization of the two mutants in HIT-T15 cells.
We found that both the V61A and the L83A mutants colocalize
with the same efficiency with secretory granules (not shown).
The N-terminal domain of Rabphilin has been shown to
interact with α-actinin in a Rab3-dependent manner (Kato et
al., 1996). Because the reorganization of the actin cytoskeleton
could potentially underlie some of the effects of Rabphilin on
the endocytotic process, we investigated the impact of the
mutations in valine 61 and in leucine 83 on α-actinin binding.
To this end, we prepared GST-affinity columns containing
either GST alone or GST fusion proteins including the Nterminal fragment of Rabphilin. As shown in Fig. 7, α-actinin
did not interact with the affinity column containing GST alone.
The columns containing the N-terminal fragment of wild-type
and mutated Rabphilin retained the same amount of α-actinin,
indicating that valine 61 and leucine 83 are not directly
involved in the binding of this regulatory component of the
actin cytoskeleton.
We next tested whether the V61A and the L83A mutants
could interact with Rabaptin-5. In this case, GST affinity
columns analogous to those used for the experiments in Fig. 7
were incubated in the presence of radioactively labeled
Rabaptin-5. Wild-type Rabphilin and the V61A mutant were
found to interact with similar efficiency with Rabaptin-5 (Fig.
1761
Fig. 8. Binding of Rabphilin mutants to Rabaptin-5. GST alone or
GST-fusion proteins containing the Rab3-binding domain of
Rabphilin (amino acids 1-206) were immobilized on glutathioneagarose beads. The beads were then incubated with [35S]-labeled
Rabaptin-5. The proteins remaining associated with the GST affinity
columns were resolved by SDS-PAGE and visualized by
autoradiography. The first lane on the left contains an aliquot of the
in vitro translated product.
8). By contrast, Rabaptin-5 was not retained on affinity
columns containing GST alone or including the L83A mutant
(Fig. 8). This indicates that leucine 83 is essential for
interaction with both Rab3 and Rabaptin-5.
The recycling of the granule components after insulin
exocytosis is thought to occur by a process similar to receptormediated endocytosis. This process has, however, not been
studied in detail. For this reason, we investigated whether the
vesicle membrane protein VAMP-2 is co-internalized with the
transferrin receptor. To this end, we co-expressed in HIT-T15
cells the human transferrin receptor and a VAMP-2 construct
containing a C-terminal T antigen tag. When the secretory
vesicles fuse with the plasma membrane this tag is exposed to
the exterior of the cell (Grote and Kelly, 1996). Thanks to this
property, it is possible to follow the fate of VAMP-2 after
exocytosis with an antibody directed against the tag that is
added in the extracellular medium. As shown in Fig. 9, after a
1 minute incubation at 37°C, human transferrin and VAMP-2
were internalized in small punctate structures. Comparison of
the two confocal images revealed that part of the endosomal
structures positive for transferrin were labeled by the antibody
against the tag of VAMP-2. The colocalization of the two
proteins indicates that, at least in its initial steps, the recycling
of VAMP-2 occurs by a mechanism analogous to that of
receptor-mediated endocytosis.
DISCUSSION
Fig. 7. Binding of Rabphilin mutants to α-actinin. GST alone or
GST-fusion proteins containing the Rab3-binding domain of
Rabphilin (amino acids 1-206) were immobilized on glutathioneagarose beads. The beads were then incubated with purified αactinin. The fraction of α-actinin remaining associated with the GST
affinity columns was assessed by western blotting. The first lane on
the left contains 100 ng of purified α-actinin.
Exocytosis and endocytosis are two inter-related processes that
need to be coordinated to preserve the efficiency of the
secretory apparatus. This is most dramatically demonstrated in
nerve terminals, where genetic defects that affect endocytosis
cause a rapid depletion of the ready releasable pool of synaptic
vesicles (Koenig and Ikeda, 1989). However, tight coupling
between exo- and endocytosis is likely to be required in all
secretory systems. Indeed, recent evidence obtained in
chromaffin cells suggests that recycling of the membrane
components of dense core secretory granules is required for the
biogenesis of secretory vesicles (Slembrouk et al., 1999).
During the past few years, several components of the molecular
machinery involved in exocytosis and endocytosis have been
1762
JOURNAL OF CELL SCIENCE 114 (10)
Fig. 9. Subcellular localization of hTfr
and VAMP-2 after internalization. HITT15 cells were transiently transfected
with a plasmid encoding the hTfr
receptor and with a VAMP-2 construct
containing a T antigen tag that is
exposed to the luminal side of secretory
vesicles (Grote and Kelly, 1996). Three
days after transfection the cells were
pre-incubated at 4°C with FITCcoupled hTfr and with a monoclonal
antibody against the T antigen tag. The
monoclonal antibody was then labeled
by exposing the cells to a Cy3-coupled
anti-mouse antibody. At the end of the
pre-incubation the cells were transferred for 1 minute at 37°C, washed with PBS supplemented with 30 mM glycine at pH 2.4 and fixed for
analysis at the confocal microscope. The figure shows a single section image.
identified, but the mechanism insuring the coupling between
the two pathways is still unclear. In this study, we provide
evidence that Rabphilin, a putative player in the exocytotic
process (Chung et al., 1995; Arribas et al., 1997; Joberty et al.,
1999), is also involved in endocytosis.
Our findings are consistent with the results obtained in squid
giant synapses and PC12 cells. In squid giant synapses,
microinjection of full-length Rabphilin increases the
membrane area of endosomes (Burns et al., 1998). In PC12
cells, the N-terminal fragment of Rabphilin was reported to
inhibit transferrin endocytosis (Ohya et al., 1998). Using a
transient transfection system, we found that a mutant of
Rabphilin that is unable to form a complex with Rab3 increases
transferrin internalization. By contrast, wild-type Rabphilin
was inactive. The most likely explanation for this finding is
that the function of wild-type Rabphilin is negatively
modulated by endogenous Rab3. The loss of this inhibitory
action would enable the mutants that cannot bind to the
GTPase to stimulate transferrin internalization. The pool of
Rab3 associated with secretory vesicles is known to be
predominantly in the GTP-bound conformation (Stahl et al.,
1994; Burstein et al., 1993). Therefore, under resting
conditions, the majority of Rabphilin attached to secretory
vesicles is expected to be complexed to Rab3 (Fig. 10). When
exocytosis is triggered, the GTPase activity of Rab3 is
enhanced leading to the hydrolysis of GTP and the release of
inorganic phosphate (Pi) (Stahl et al., 1994). Thus, during or
immediately after the fusion of secretory vesicles with the
plasma membrane, Rab3 switches from the GTP-bound to the
GDP-bound state. The conformational change associated with
the hydrolysis of GTP will cause a drop in the affinity for
Rabphilin (Shirataki et al., 1993), favoring the formation of a
soluble Rab3-RabGDI complex (Matsui et al., 1990). Thus,
immediately after the insertion of the secretory vesicle
membrane in the plasma membrane, Rabphilin becomes free
to interact with other binding partners and to trigger
endocytosis (Fig. 10). This scenario provides a possible
explanation for the spatio-temporal link between exo- and
endocytosis. In this study, the role of Rabphilin in endocytosis
was investigated by measuring transferrin internalization.
However, because immediately after internalization we
observe a partial colocalization of VAMP-2 and transferrin on
the same endosomal compartment, Rabphilin will most likely
accelerate the recycling of secretory vesicle proteins.
Rabphilin has been shown to bind to Rabaptin-5 (Ohya et
al., 1998), a known component of the endocytotic machinery
(Stenmark et al., 1995). The association between the two
proteins occurs through the N-terminal portion of Rabphilin
and is prevented by the binding of Rab3 (Ohya et al., 1998).
Our results confirm these findings and narrow down the domain
of interaction to a portion of the amphipathic α-helix at the Nterminus of Rabphilin. Taking advantage of the properties of
different Rabphilin mutants, we show that the interaction with
Rabaptin-5 is required for the stimulation of endocytosis.
Rabaptin-5 and the guanine nucleotide exchange factor Rabex5 are known to form a complex that activates the GTPase Rab5
(Horiuchi et al., 1997). This leads to the assembly of a
multiprotein complex including the early endosome-associated
protein EEA1 and phosphatidylinositol 3-kinase that promotes
the fusion of clathrin-coated vesicles with early endosomes,
and the homotypic fusion of early endosomes (McBride et al.,
1999; Christoforidis et al., 1999). We propose that immediately
after the fusion of secretory vesicles with the plasma
membrane, Rabphilin dissociates from Rab3 and binds to
Rabaptin-5 (Fig. 9). This would cause the recruitment of an
important component of the machinery involved in the budding
Fig. 10. Model for the role of Rabphilin in the coupling of exo- and
endocytosis. Under resting conditions, Rabphilin is localized on the
membrane of secretory vesicles and is tightly associated to the GTPbound form of Rab3. Upon vesicle fusion, Rab3 hydrolyses GTP and
forms a complex with RabGDI that dissociates from the membrane.
This permits Rabphilin to recruit Rabaptin-5 near the exocytotic site,
thus contributing to the rapid recycling of the components of the
secretory vesicle.
Involvement of Rabphilin in endocytosis
of endocytic vesicles close to the release site. Rabaptin-5
together with Rabex-5 could then activate Rab5 and, in turn,
promote the assembly of the multiprotein complex controlling
the fusion of clathrin-coated vesicles with early endosomes.
This hypothesis is in agreement with the fact that
overexpression of Rabphilin V61A, Rabaptin-5, or a
constitutively active mutant of Rab5 (Stenmark et al., 1994) all
have similar effects on transferrin endocytosis.
Rabphilin is not the first example of a protein associated with
the membrane of secretory vesicles that is proposed to generate
a signal that initiates membrane retrieval. A similar role was
postulated for synaptotagmin, the putative Ca2+ sensor of
exocytosis. Synaptotagmin was found to be a high-affinity
binding site for the clathrin adaptor complex AP-2 (Zhang et
al., 1994). Genetic evidence obtained in Caenorhabditis
elegans confirmed the involvement of synaptotagmin in
synaptic vesicle recycling (Jorgensen et al., 1995). Newly
assembled clathrin-coated pits are unable to proceed to ligand
sequestration in the absence of Rab5 (McLauchlan et al.,
1997). Thus, it is possible that the recycling of secretory
vesicles requires the concerted action of both synaptotagmin
and Rabphilin. Synaptotagmin would initiate the assembly of
the clathrin coat by binding to AP-2, whereas Rabphilin would
recruit Rabaptin-5 and its partners.
In summary, we have identified Rabphilin as a possible
candidate for coordination of the exo- and endocytic pathways.
Future studies will have to elucidate the precise sequence of
events that are elicited by Rabphilin in order to increase the
internalization rate of clathrin-coated vesicles. This important
task will certainly be facilitated now we know more about the
characteristics of different point mutants of Rabphilin.
We thank L. Kühn, G. Grenningloh and R. B. Kelly for reagents,
and P. Clarke for critical reading of the manuscript. This work was
supported by the Swiss National Science Foundation grants 3150640.97 and 32-61400.00 (R. R.), and 31-52587.97 (H. H. and S.
C.).
REFERENCES
Arribas, M., Regazzi, R., Garcia, E., Wollheim, C. B. and De Camilli, P.
(1997). The stimulatory effect of Rabphilin 3a on regulated exocytosis from
insulin-secreting cells does not require an association-dissociation cycle
with membranes mediated by Rab3. Eur. J. Cell Biol. 74, 209-216.
Burns, M. E., Sasaki, T., Takai, Y. and Augustine, G. J. (1998). Rabphilin3A: a multifunctional regulator of synaptic vesicle traffic. J. Gen. Physiol.
111, 243-255.
Burstein, E. S., Brondyk, W. H., Macara, I. G, Kaibuchi, K. and Takai, Y.
(1993). Regulation of the GTPase cycle of the neuronally expressed Raslike GTP-binding protein Rab3A. J. Biol. Chem. 268, 22247-22250.
Chung, S. H., Takai, Y. and Holz, R. W. (1995). Evidence that the Rab3abinding protein, Rabphilin3a, enhances regulated secretion. J. Biol. Chem.
270, 16714-16717.
Chung, S.-H., Joberty, G., Gelino, E. A., Macara, I. G. and Holz, R. W.
(1999). Comparison of the effects on secretion in chromaffin and PC12 cells
of Rab3 family members and mutants. J. Biol. Chem. 274, 18113-18120.
Christophoridis, S., McBride, H. M., Burgoyne, R. D. and Zerial, M.
(1999). The Rab5 effector EEA1 is a core component of endosome docking
and fusion. Nature 397, 621-625.
Coppola, T., Perret-Menoud, V., Lüthi, S., Farnsworth, C. C., Glomset, J.
A. and Regazzi, R. (1999). Disruption of Rab3-calmodulin interaction, but
not other effector interactions, prevents Rab3 inhibition of exocytosis.
EMBO J. 18, 5885-5891.
Fukui, K., Sasaki, T., Imazumi, K., Matsuura, Y., Nakanishi, H. and Takai,
Y. (1997). Isolation and characterization of a GTPase activating protein
1763
specific for the Rab3 subfamily of small G proteins. J. Biol. Chem. 272,
4655-4658.
Geppert, M., Goda, Y., Stevens, C. F. and Südhof, T. C. (1997). The small
GTP-binding protein Rab3A regulates a late step in synaptic vesicle fusion.
Nature, 387, 810-814.
Grote, E. and Kelly, R. B. (1996). Endocytosis of VAMP is facilitated by a
synaptic vesicle targeting signal. J. Cell Biol. 132, 537-547.
Holz, R. W., Brondyk, W. H., Senter, R. A., Kuizon, L. and Macara, I. G.
(1994). Evidence for the involvement of Rab3A in Ca(2+)-dependent
exocytosis from adrenal chromaffin cells. J. Biol. Chem. 269, 10229-10234.
Horiuchi, H., Lippé, R., McBride, H. M., Rubino, M., Woodman, P.,
Stenmark, H., Rybin, V., Wilm, M., Ashman, K., Mann, M. et al. (1997).
A novel Rab5 GDP/GTP exchange factor complexed to Rabaptin-5 links
nucleotide exchange to effector recruitment and function. Cell 90, 11491159.
Iezzi, M., Escher, G., Meda, M., Charollais, A., Baldini, G., Darchen, F.,
Wollheim, C. B. and Regazzi, R. (1999). Subcellular distribution and
function of Rab3A, B, C, and D isoforms in insulin-secreting cells. Mol.
Endocrinol. 13, 202-212.
Joberty, G., Stabila, P. F., Coppola, T., Macara, I. G. and Regazzi, R.
(1999). High affinity Rab3 binding is dispensable for Rabphilin-dependent
potentiation of stimulated secretion. J. Cell Sci. 112, 3579-3587.
Johannes, L., Lledo, P. M., Roa, M., Vincent, J. D., Henry, J.-P. and
Darchen, F. (1994). The GTPase Rab3a negatively controls calciumdependent exocytosis in neuroendocrine cells. EMBO J. 13, 2029-2037.
Jorgensen, E. M., Hartwieg, E., Schuske, K., Nonet, M. L., Jin, Y.,
Horwitz, H. R. (1995). Defective recycling of synaptic vesicles in
synaptotagmin mutants of Caenorhabditis elegans. Nature 378, 196-199.
Kato, M., Sasaki, T., Ohya, T., Nakanishi, H., Nishioka, H., Imamura, M.
and Takai, Y. (1996). Physical and functional interaction of Rabphilin-3A
with α-actinin. J. Biol. Chem. 271, 31775-31778.
Koenig, J. H. and Ikeda, K. (1989). Disappearance and reformation of
synaptic vesicle membrane upon transmitter release observed under
reversible blockade of membrane retrieval. J. Neurosci. 9, 3844-3860.
Lledo, P. M., Vernier, P., Vincent, J. D., Mason, W. T. and Zorec, R. (1993).
Inhibition of Rab3B expression attenuates Ca2+-dependent exocytosis in rat
anterior pituitary cells. Nature 364, 540-544.
Masumoto, N., Sasaki, T., Tahara, M., Mammoto, A., Ikebuchi, Y., Tasaka,
K., Tokunaga, M., Takai, Y. and Miyake, A. (1998). Involvement of
Rabphilin-3A in cortical granule exocytosis in mouse eggs. J. Cell Biol. 135,
1741-1747.
Matsui, Y., Kikuchi, A., Araki, J., Hata, Y., Kondo, J., Teranishi, Y. and
Takai, Y. (1990). Molecular cloning and characterization of a novel type of
regulatory protein (GDI) for smg p25A, a Ras p21-like GTP-binding
protein. Mol. Cell. Biol. 10, 4116-4122.
McBride, H. M., Rybin, V., Murphy, C., Giner, A., Teasdale R., Zerial, M.
(1999). Oligomeric complexes link Rab5 effectors with NSF and drive
membrane fusion via interactions between EEA1 and syntaxin 13. Cell 98,
377-386.
McLauchlan, H., Newell, J., Morrice, N., Osborne, A., West, M. and
Smythe, E. (1997). A novel role for Rab5-GDI in ligand sequestration into
clathrin-coated pits. Curr. Biol. 8, 34-45.
Ohnishi, H., Samuelson, L. C., Yule, D. I., Ernst, S. A. and Williams, J. A.
(1997). Overexpression of Rab3D enhances regulated amylase secretion
from pancreatic acini of transgenic mice. J. Clin. Invest. 100, 3044-3052.
Ohya, T., Sasaki, T., Kato, M. and Takai, Y. (1998). Involvement of
Rabphilin3 in endocytosis through interaction with Rabaptin5. J. Biol.
Chem. 273, 613-617.
Okamoto, M., Schoch, S. and Südhof, T. C. (1999). EHSH1/Intersectin, a
protein that contains EH and SH3 domains and binds to dynamin and SNAP25. J. Biol. Chem. 274, 18446-18454.
Ostermeier, C. and Brunger, A. T. (1999). Structural basis of Rab effector
specificity: crystal structure of the small G protein Rab3A complexed with
the effector domain of Rabphilin3A. Cell 96, 363-374.
Pfeffer, S. R., Dirac-Svejstrup, B. A. and Soldati, T. (1995). Rab GDP
dissociation inhibitor: putting Rab GTPases in the right place. J. Biol. Chem.
270, 17057-17059.
Regazzi, R. Li, G., Deshusses, J. and Wollheim, C.B. (1990). Stimulusresponse coupling in insulin-secreting HIT cells. Effects of secretagogues
on cytosolic Ca2+, diacylglycerol, and protein kinase C activity. J. Biol.
Chem. 265, 15003-15009.
Regazzi, R., Kikuchi, A., Takai, Y. and Wollheim, C. B. (1992). The small
GTP-binding proteins in the cytosol of insulin-secreting cells are complexed
to GDP dissociation inhibitor proteins. J. Biol. Chem. 267, 17512-17519.
1764
JOURNAL OF CELL SCIENCE 114 (10)
Schlüter, O. M., Schnell, E., Verhage, M., Tzonopoulos, T., Nicoll, R. A.,
Janz, R., Malenka, R. C., Geppert, M. and Südhof, T. C. (1999).
Rabphilin knock-out mice reveal that Rabphilin is not required for Rab3
function in regulating neurotransmitter release. J. Neurosci. 19, 58345846.
Shirataki, H., Kaibuchi, K., Yamaguchi, T., Wada, K., Horiuchi, H. and
Takai, Y. (1992). A possible target protein for smg-25A/rab3A small GTPbinding protein. J. Biol. Chem. 267, 10946-10949.
Shiritaki, H., Kaibuchi, K., Sakoda, T., Kishida, S., Yamaguchi, T., Wada,
K., Miyazaki, M. and Takai, Y. (1993). Rabphilin-3A, a putative target
protein for smg p25A/rab3A p25 small GTP-binding protein related to
synaptotagmin. Mol. Cell. Biol. 13, 2061-2068.
Slembrouck, D., Annaert, W. G., Wang, J. M. and De Potter, W. P. (1999).
Rab3 is present on endosomes from bovine chromaffin cells in primary
culture. J. Cell Sci. 112, 641-649.
Stahl, B., Fischer von Mollard, G., Walch-Solimena, C. and Jahn, R.
(1994). GTP cleavage by the small GTP-binding protein Rab3A is
associated with exocytosis of synaptic vesicles induced by α-latrotoxin. J.
Biol. Chem. 269, 24770-24776.
Stahl, B., Chou, J. H., Li, C., Südhof, T. C. and Jahn, R. (1996). Rab3
reversibly recruits Rabphilin to synaptic vesicles by a mechanism analogous
to Raf recruitment by Ras. EMBO J. 15, 1799-1809.
Stenmark, H., Parton, R. G., Steele-Mortimer, O., Lütcke, A., Gruenberg,
J. and Zerial, M. (1994). Inhibition of rab5 GTPase activity stimulates
membrane fusion in endocytosis. EMBO J. 13, 1287-1296.
Stenmark, H., Vitale, G., Ullrich, O. and Zerial, M. (1995). Rabaptin-5 is
a direct effector of the small GTPase Rab5 in endocytic membrane fusion.
Cell 83, 423-432.
Ullrich, O., Stenmark, H., Alexandrov, K., Huber, L. A., Kaibuchi, K.,
Sasaki, T., Takai, Y. and Zerial, M. (1993). Rab GDP dissociation inhibitor
as a general regulator for the membrane association of Rab proteins. J. Biol.
Chem. 268, 18143-18150.
Wada, M., Nakanishi, H., Satoh, A., Hirano, H., Obaishi, H., Matsuura,
Y. and Takai, Y. (1997). Isolation and characterization of a GDP/GTP
exchange protein specific for the Rab3 subfamily small G proteins. J. Biol.
Chem. 272, 3875-3878.
Yamaguchi, T., Shirataki, H. Kishida, S. Miyazaki, M., Nishikawa, J.,
Wada, K., Numata, S., Kaibuchi, K. and Takai, Y. (1993). Two
functionally different domains of Rabphilin-3A, Rab3A p25/smg p25Abinding and phospholipid- and Ca(2+)-binding domains. J. Biol. Chem. 268,
27164-27170.
Zhang, J. Z., Davletov, B. A., Südhof, T. C. and Anderson, R. G. (1994).
Synaptotagmin I is a high affinity receptor for clathrin AP-2: implications
for membrane recycling. Cell 78, 751-760.