The C2B Domain of Rabphilin Directly Interacts with SNAP-25 and Regulates the Docking Step of Dense Core Vesicle Exocytosis in PC12 Cells*

Rabphilin is a membrane trafficking protein on secretory vesicles that consists of an N-terminal Rab-binding domain and C-terminal tandem C2 domains. The N-terminal part of rabphilin has recently been shown to function as an effector domain for both Rab27A and Rab3A in PC12 cells (Fukuda, M., Kanno, E., and Yamamoto, A. (2004) J. Biol. Chem. 279, 13065–13075), but the function of the C2 domains of rabphilin during secretory vesicle exocytosis is largely unknown. In this study we investigated the interaction between rabphilin and SNAREs (soluble N-ethylmaleimide-sensitive factor attachment protein receptors, VAMP-2/synaptobrevin-2, syntaxin IA, and SNAP-25) and SNARE-associated proteins (Munc18-1 and Munc13-1) and found that the C2B domain of rabphilin, but not of other Rab27A-binding proteins with tandem C2 domains (i.e. Slp1-5), directly interacts with a plasma membrane protein, SNAP-25. The interaction between rabphilin and SNAP-25 occurs even in the absence of Ca2+ (EC50 = 0.817 μm SNAP-25), but 0.5 mm Ca2+ increases the affinity for SNAP-25 2-fold (EC50 = 0.405 μm SNAP-25) without changing the Bmax value (1.06 mol of SNAP-25/mol of rabphilin). Furthermore, vesicle dynamics were imaged by total internal reflection fluorescence microscopy in a single PC12 cell expressing a lumen-targeted pH-insensitive yellow fluorescent protein (Venus), neuropeptide Y-Venus. Expression of the wild-type rabphilin in PC12 cells significantly increased the number of docked vesicles to the plasma membrane without altering the kinetics of individual secretory events, whereas expression of the mutant rabphilin lacking the C2B domain, rabphilin-ΔC2B, decreased the number of docked vesicle or fusing at the plasma membrane. These findings suggest that rabphilin is involved in the docking step of regulated exocytosis in PC12 cells, possibly through interaction between the C2B domain and SNAP-25.

The functional relationship between Rab27A⅐effector complex and SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) proteins, fundamental fusion machinery of vesicle exocytosis (20), have recently been reported. Slp4-a/granuphilin-a, another Rab27A effector that functions in certain endocrine cells (21)(22)(23) and parotid acinar cells (24), was reported to directly interact with syntaxin IA and/or Munc18-1 (23,25,26). In addition, the results of a genetic analysis of Caenorhabditis elegans rabphilin and SNARE (syntaxin, SNAP-25, or VAMP/synaptobrevin) double mutants have suggested that rabphilin modulates SNARE function (8). Thus, it is highly possible that rabphilin physically associates with SNARE proteins itself and/or SNARE-associated proteins but that possibility has never been investigated. In this study we systematically investigated the interaction between rabphilin and SNAREs and SNARE-associated proteins by coimmunoprecipitation assays and found that the C2B domain of rabphilin directly interacts with isolated SNAP-25, a target SNARE localized on the plasma membrane. Because there has never been a detailed description of the function of the C2B domain of rabphilin in the recruitment, docking, priming, and fusion of secretory vesicles in live cells, we also investigated the function of the C2B domain of rabphilin on the motion of a single dense core vesicle during exocytosis in PC12 cells by total internal reflection fluorescence microscopy (TIRF, also called evanescent wave or evanescence microscopy) (27) using vesicletargeted fluorescent proteins (28 -31). Expression of the rabphilin-⌬C2B mutant lacking SNAP-25 binding activity inhibited vesicle docking and markedly decreased the number of exocytotic events without affecting the kinetics of vesicle fusion. Based on our findings, we discuss * This work was supported in part by the Ministry of Education, Culture, Sports, and Technology of Japan (Grants 15689006, 16044248, 17024065, and 17657067 (to M. F.)), by the Kato Memorial Bioscience Foundation (to M. F.), by The Sumitomo Foundation (to M. F.), and by the FY2005 DRI Research Grant (to T. T.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Fig. 1  the possible function of the C2B domain of rabphilin in docking of dense core vesicles to the plasma membrane.
Co-immunoprecipitation Assay in COS-7 Cells-COS-7 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin G, and 100 g/ml streptomycin, at 37°C under 5% CO 2 . pEF-T7, pEF-FLAG, and/or pEF-HA vectors (total of 4 g of plasmids) were transfected into COS-7 cells (7.5 ϫ 10 5 cells, the day before transfection/10-cm dish) by using Lipofectamine Plus reagent (Invitrogen) according to the manufacturer's instructions. Three days after transfection, cells were harvested and homogenized, and total cell lysates were prepared as described previously (37). The total cell lysates (400 l) were incubated with anti-T7 tag antibody-conjugated agarose beads (wet volume 20 l) with gentle agitation at 4°C for 1 h, and the proteins bound to the beads were analyzed by 10% (or 12.5%) SDS-PAGE followed by immunoblotting with HRPconjugated anti-T7 tag antibody (1/10,000 dilution), HRP-conjugated anti-FLAG M2 antibody (1/10,000 dilution), and/or HRP-conjugated anti-HA tag antibody (1/10,000 dilution) as described previously (33,38). The blots shown in this paper are representative of at least two independent experiments.
Direct Interaction between T7-rabphilin and GST-SNAP-25-The T7-rabphilin beads (wet volume, 10 l) were prepared as described above. GST-SNAP-25 was expressed in bacteria and purified with glutathione-Sepharose beads by standard protocols (18,39). The concentration of GST-SNAP-25 was determined with a Bio-Rad protein assay kit (Bio-Rad Laboratories), using bovine serum albumin as a reference. The T7-rabphilin beads were incubated for 1 h at 4°C with 400 l of various concentrations of GST-SNAP-25 (0.02-5 M) in 50 mM HEPES-KOH, pH 7.2, 150 mM NaCl, 1 mM MgCl 2 , 0.2% Triton X-100, and 2 mM EGTA (or 0.5 mM CaCl 2 ), and proteins bound to the beads were analyzed by 10% SDS-PAGE followed by staining with Coomassie Brilliant Blue R-250. The intensity of the bands on the gel was captured and quantified with Lane Analyzer software (version 3.0) (ATTO Corp., Tokyo, Japan). All statistical analyses and curve fitting were performed with a GraphPad Prism computer program (version 4.0, GraphPad Software, San Diego, CA).
We monitored exocytosis of NPY-Venus at the single vesicle level by using a TIRF microscope similar to that described previously (28,31). In brief, a high numerical aperture objective lens (Plan Apochromatic, 100ϫ, numerical aperture ϭ 1.45, infinity-corrected, OLYMPUS, Tokyo, Japan) was mounted on an inverted microscope (IX71, OLYM-PUS), and incident light for total internal reflection illumination was introduced from the high numerical aperture objective lens through a single mode optical fiber and two illumination lenses (IX2-RFAEVA-2, OLYMPUS). To observe the NPY-Venus fluorescence images, we used a diode-pumped solid state 488-nm laser (HPU50100, 20 milliwatt, Furukawa Electronic, Chiba, Japan) for total internal fluorescence illumination and a band pass filter (HQ535/30m, Chroma, Rockingham, VT) as an emission filter. The laser beam was passed through an electromagnetically driven shutter (VMM-D3J, Unibritz, Rochester, NY), and the shutter was opened synchronously with electron multiplier charge-coupled device camera (C9100-02, Hamamatsu Photonics, Hamamatsu, Japan) exposure controlled by MetaMorph software (version 6.3, Universal Imaging Corporation, Downingtown, PA). Images were acquired every 200 ms or otherwise as indicated. To analyze the TIRF imaging data, single exocytotic events were manually selected, and the average fluorescence intensity of individual vesicles in a 0.7 ϫ 0.7-m square positioned over the center of the vesicle was calculated. The number of fusion events was counted manually during a 5-min period. Data are reported as means Ϯ S.E. of at least five individual experiments. Means were compared by one-way analysis of variance with GraphPad Prism software.
Interaction between rabphilin and SNAP-25 was further investigated by co-immunoprecipitation assay using COS-7 cells, which do not endogenously express either rabphilin or neuronal SNAREs (or SNARE-associated proteins) (35). In brief, agarose beads coupled with T7-rabphilin were incubated with COS-7 cell lysates containing FLAGsyntaxin IA, FLAG-SNAP-25, FLAG-VAMP-2, or FLAG-Munc18-1, in the presence and absence of HA-Rab27A, and FLAG-tagged proteins and HA-Rab27A that had bound to the beads were detected by immu-  NOVEMBER 25, 2005 • VOLUME 280 • NUMBER 47

JOURNAL OF BIOLOGICAL CHEMISTRY 39255
noblotting with the specific antibodies indicated (Fig. 2). As expected, rabphilin specifically interacted with SNAP-25, but not with other SNAREs or Munc18-1, irrespective of the presence of Rab27A (Fig. 2,  third panel, lanes 2 and 6). It should be noted that SNAP-25 did not interact with Slp1-5 (11,14,36), another Rab27A-binding protein with tandem C2 domains at the C terminus (supplemental Fig. 1), suggesting that rabphilin is the only Rab27A-binding protein that physically associates with SNAP-25.
The C2B Domain of Rabphilin Functions as a SNAP-25-binding Site-Next, we attempted to map the minimal SNAP-25-binding site in rabphilin. To do so, we prepared three truncated mutants of rabphilin, i.e. the RBD, C2A domain, and C2B domain, and tested their SNAP-25 binding activity by co-immunoprecipitation assay using COS-7 cells. As shown in Fig. 3A, the C2B domain interacted with SNAP-25, the same as the full-length protein (third panel, lanes 1 and 4), but the C2A domain and the RBD did not. Similarly, the rabphilin-⌬C2B mutant did not interact with SNAP-25, but the rabphilin (E50A/I54A) mutant, which lacks Rab27A binding activity (13), did (Fig. 3B, third panel, lanes 2-4). The direct interaction between rabphilin and SNAP-25 was further investigated by using purified components, i.e. T7-rabphilin and GST-SNAP-25. Interaction between rabphilin and SNAP-25 was also observed when the purified proteins were used, even in the presence of 2 mM EGTA (Fig. 4A, closed arrowhead), and 0.5 mM Ca 2ϩ slightly increased the SNAP-25 binding activity (Fig. 4B, closed arrowhead). A calculation of the EC 50 values by analysis of the dose dependence curve of GST-SANP-25 (Fig. 4C) yielded 0.817 and 0.405 M for the interaction between rabphilin and SNAP-25 in the absence and presence, respectively, of Ca 2ϩ , whereas the B max values were unchanged (1.06 mol of GST-SNAP-25/mol of T7-rabphilin). These findings indicate

. The C2B domain of rabphilin is required for interaction with SNAP-25.
Mapping of the site in rabphilin responsible for SNAP-25 binding is shown. Agarose beads coupled with T7-rabphilin (Rph) mutants (Rph, Rph-RBD, Rph-C2A, and Rph-C2B in A and Rph(E50A/I54A) and Rph-⌬C2B in B) were incubated with COS-7 cell lysates containing FLAG-SNAP-25 in the presence and absence of pEF-HA-Rab27A. Proteins that bound to the beads were analyzed by 10% SDS-PAGE followed by immunoblotting with HRP-conjugated anti-FLAG M2 antibody (Blot, anti-FLAG; IP, anti-T7; third panels), HRPconjugated anti-HA tag antibody (Blot, anti-HA; IP, anti-T7; fourth panels), and HRP-conjugated anti-T7 tag antibody (Blot, anti-T7; IP, anti-T7; bottom panels). Input means 1/80 volume of the reaction mixture used for immunoprecipitation (top two panels). Note that the C2B domain of rabphilin alone is necessary and sufficient for SNAP-25 binding and that the binding is independent of Rab27A binding to the N-terminal RBD. The positions of the molecular mass markers (ϫ10 Ϫ3 ) are shown on the left. that 1 molecule of rabphilin binds 1 molecule of SNAP-25 via the C2B domain and that Ca 2ϩ promotes SNAP-25 binding to rabphilin because of the 2-fold increase in affinity. Because SNAP-25 and Rab27A interact with the distinct domains of rabphilin, a tripartite protein complex of Rab27A, rabphilin, and SNAP-25 can be formed (Fig. 4D, lane 2).
Rabphilin Promotes Dense Core Vesicle Exocytosis by Increasing the Number of Docked Vesicles to the Plasma Membrane in PC12 Cells-Although expression of rabphilin in certain endocrine cells has been shown to promote dense core vesicle exocytosis (3,4,13), the step or steps in vesicle exocytosis, i.e. recruitment, docking, and fusion, that are regulated by rabphilin have never been identified. To resolve this matter, we co-expressed rabphilin and NPY-Venus, which is specifically targeted to dense core vesicles (30,34), in PC12 cells and monitored the fate of NPY-Venus during and after individual exocytotic events near the plasma membrane by TIRF microscopy. First, we investigated colocalization between mRFP-rabphilin and NPY-Venus in resting PC12 cells by confocal microscopy, and as expected, most NPY-Venus-positive vesicles were found to be colocalized with mRFP-rabphilin-positive vesicles (Fig. 5A, top panels), consistent with  the previous findings regarding dense core-vesicle localization of rabphilin in PC12 cells (4,13).
To determine whether mRFP-rabphilin regulates vesicle transport, docking to the plasma membrane, or fusion in live PC12 cells, we counted the number of plasma membrane-associated vesicles before and after high KCl stimulation (70 mM) by TIRF microscopy (Fig. 5B and supplemental movies 1-3). As shown in Fig. 5C, the numbers of plasma membrane-docked vesicles both before and after stimulation were significantly higher in the mRFP-rabphilin-expressing cells than in the control cells that expressed mRFP alone. The total number of NPY-Venus release events from cells expressing mRFP-rabphilin was also markedly higher than in the control cells (Fig. 5D, open bar and closed bar, respectively), but the kinetics of the individual fusion events were identical in both type of cells (Fig. 6A, top and middle panels, respectively). The NPY-Venus-containing spots grew brighter and expanded suddenly during the release of the fluorescent peptide in response to high KCl stimulation (31), with an identical time course, in both the control cells and mRFP-rabphilin-expressing cells (Fig. 6B). These results strongly indicate that rabphilin promotes docking of dense core vesicles to the plasma membrane in PC12 cells rather than modulates vesicle fusion (or the kinetics of vesicle fusion).
The C2B Domain of Rabphilin Is Required for the Dense Core Vesicle Docking Step in PC12 Cells-To further determine whether the interaction between the C2B domain of rabphilin and SNAP-25 is required for the docking step of dense core vesicle exocytosis in PC12 cells, a truncated mRFP-rabphilin-⌬C2B mutant that completely lacked SNAP-25 binding activity in vitro (see Fig. 3B) was coexpressed with NPY-Venus in PC12 cells, and the dynamics of single vesicle docking, movement, and exocytosis were analyzed by TIRF microscopy. Confocal microscopic analysis showed that the mRFP-rabphilin-⌬C2B mutant still localized on NPY-Venus-containing dense core vesicles in PC12 cells (Fig. 5A, lower panels), consistent with the previous finding that the dense core vesicle localization of rabphilin is specifically mediated by the RBD (13). In contrast to the wild-type mRFP-rabphilin, however, no significant difference in the number of plasma membranedocked vesicles before stimulation was detected between control and mRFP-rabphilin-⌬C2B-expressing cells (Fig. 5C), and mRFP-rabphilin-⌬C2B reduced the number of plasma membrane-docked vesicles after stimulation (Fig. 5C). Consistent with this, there were significantly fewer NPY-Venus release events in cells expressing mRFP-rabphilin-⌬C2B than control cells (Fig. 5D, gray bar). Exocytotic events were also detected much less frequently in mRFP-rabphilin-⌬C2B-expressing cells than in mRFP-rabphilin-expressing cells, or even in the control cells (Fig. 5D), although the kinetics of individual fusion events were identical in each case (Fig. 6). Taken together, these findings suggest that the C2B domain of rabphilin regulates the docking step of dense core vesicles to the plasma membrane rather than the dense core vesicle fusion step.

DISCUSSION
Although a previous genetic analysis of C. elegans rabphilin-SNARE double mutants suggested that rabphilin modulates SNARE function (8), rabphilin-binding proteins identified thus far (e.g. ␣-actinin, ␤-adducin, GTP cyclohydrolase I, Rabaptin5, and annexin A4) (43)(44)(45)(46)(47)(48) have no relation to the function of neuronal SNAREs or SNARE-related proteins, and a physical association between rabphilin and SNAREs had never been elucidated even in vitro. In the present study we obtained for the first time biochemical evidence that the C2B domain of rabphilin is necessary and sufficient for direct Ca 2ϩ -independent interaction with SNAP-25 but not with syntaxin IA or VAMP-2 ( Figs. 1-4). We also demonstrated by TIRF microscopy that expression of rabphilin significantly increases the number of docked dense core vesicles to the plasma membrane in PC12 cells (Fig. 5) without altering the kinetics of individual exocytotic events (Fig. 6), whereas expression of a rabphilin-⌬C2B mutant lacking SNAP-25 binding activity significantly decreased the number of high KCl-induced NPY-Venus release events and the number of plasma membrane-docked vesicles after high KCl stimulation (Fig. 5). The most straightforward explanation for this result is that rabphilin promotes docking of dense core vesicles to the plasma membrane by linking the Rab27A on the vesicle via the RBD and SNAP-25 at the plasma membrane via the C2B domain. Very recently, another Rab27A effector, Slp2-a in melanocytes, has been shown to be involved in the anchoring of melanosomes to the plasma membrane through simultaneous interaction with Rab27A on the melanosome and phosphatidylserine in the plasma membrane via the C2A domain (49), and Slp4-a/granuphilin-a has been shown to be involved in the docking of insulin-containing vesicles to the plasma membrane through interaction with Rab27A on the vesicle and syntaxin IA/Munc18-1 at the plasma membrane in pancreatic ␤-cell lines (26,50).
It is interesting that neither melanocytes nor pancreatic ␤-cells endogenously express rabphilin, and thus it is highly possible that the general function of Rab27A effectors with tandem C2 domains (i.e. Slp1-5 and rabphilin) may be the docking of Rab27A-bound organelles to the plasma membrane with "two hands." These Rab27A effectors seize specific organelles with Rab27A by the N-terminal RBD (one hand) and directly or indirectly tether them to the plasma membrane by the C-terminal tandem C2 domains (the other hand). Because the biochemical properties of the tandem C2 domains of Slp1-5 and rabphilin differ in terms of phospholipid and protein interactions (14,23,25,36,(51)(52)(53) and these Rab27A effectors are differentially expressed in mouse tissues (1,36,42,51), the docking mechanisms of these Rab27A effectors may be different, because Slp1-5 did not interact with SNAP-25 at all (sup- plemental Fig. 1), and rabphilin may be the only Rab27A effector that utilizes SNAP-25 for docking process in certain neuroendocrine cells (e.g. chromaffin cells and PC12 cells).
Because the rabphilin-⌬C2B mutant used in this study completely lacked the C2B domain, we cannot rule out the possibility that unidentified C2B ligands other than SNAP-25 regulate the docking process of dense core vesicle exocytosis in PC12 cells. One of the candidate ligands is phosphatidylinositol 4,5-bisphosphate in the plasma membrane (19). However, because phosphatidylinositol 4,5-bisphosphate interacts with both C2 domains of rabphilin, mostly in a Ca 2ϩ -dependent manner (19), the observed rabphilin-dependent increase in number of docked vesicles in resting cells (Fig. 5C) cannot be explained by the Ca 2ϩ -dependent phosphatidylinositol 4,5-bisphosphate binding activity of rabphilin alone. Production and analysis of a rabphilin mutant that specifically lacks SNAP-25 binding activity, but not phosphatidylinositol 4,5bisphosphate binding activity, will clarify which C2B ligands are essential for the docking step of regulated exocytosis.
In conclusion, we have demonstrated that rabphilin directly interacts with SNAP-25 via the C2B domain and that the C2B domain is required for promotion of dense core vesicle exocytosis in PC12 cells, specifically the docking step to the plasma membrane. Based on our findings, we hypothesize that the C2B domain of rabphilin plays an essential role in the docking step of regulated exocytosis in PC12 cells, possibly through interaction with Rab27A on the vesicle via the RBD and with t-SNARE SNAP-25 at the plasma membrane via the C2B domain.