Isoproterenol Stimulates Transient SNAP23–VAMP2 Interaction in Rat

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

FEBS Letters 587 (2013) 583–589

journal homepage: www.FEBSLetters.org

Isoproterenol stimulates transient SNAP23–VAMP2 interaction in rat parotid glands ⇑ Taishin Takuma a, , Akiko Shitara a, Toshiya Arakawa a, Miki Okayama b, Itaru Mizoguchi b, Yoshifumi Tajima c

a Department of Biochemistry, School of Dentistry, Health Sciences University of Hokkaido, Tobetsu, Hokkaido 061-0293, Japan b Department of Orthodontics, School of Dentistry, Health Sciences University of Hokkaido, Tobetsu, Hokkaido 061-0293, Japan c Department of Oral Pathology, School of Dentistry, Meikai University, Sakado, Saitama 350-0283, Japan

article info abstract

Article history: The exocytosis of salivary proteins is mainly regulated by cAMP, although soluble N-ethylmalei- Received 11 December 2012 mide-sensitive factor attachment protein receptors (SNAREs), which mediate cAMP-dependent exo- Revised 10 January 2013 cytic membrane fusion, have remained unidentiﬁed. Here we examined the effect of isoproterenol Accepted 20 January 2013 (ISO) and cytochalasin D (CyD) on the level of SNARE complexes in rat parotid glands. When SNARE Available online 1 February 2013 complexes were immunoprecipitated by anti-SNAP23, the coprecipitation of VAMP2 was signiﬁ- Edited by Gianni Cesareni cantly increased in response to ISO and/or CyD, although the coprecipitation of VAMP8 or syntaxin 4 was scarcely augmented. These results suggest that the SNAP23–VAMP2 interaction plays a key role in cAMP-mediated exocytosis from parotid glands. Keywords: Parotid gland Exocytosis Structured summary of protein interactions: Soluble N-ethylmaleimide-sensitive factor Snap23 physically interacts with Vamp8, Vamp3 and Syn-4 by anti bait coimmunoprecipitation (View attachment protein receptor (SNARE) Interaction: 1, 2, 3) VAMP2 Vamp2 physically interacts with Syn-4, Syn-3 and Snap23 by anti bait coimmunoprecipitation (View SNAP23 interaction) Cytochalasin D Syn-3 physically interacts with Snap23 by anti bait coimmunoprecipitation (View interaction) Vamp2 physically interacts with Snap23 and Syn-4 by anti bait coimmunoprecipitation (View interaction) Snap23 physically interacts with Vamp8, Syn-4 and Vamp2 by anti bait coimmunoprecipitation (View interaction: 1, 2)

1. Introduction However, the mechanism of cAMP-induced exocytic membrane fusion is totally unknown. The salivary glands are unique exocrine glands that secrete var- The regulated exocytosis is a spatio-temporal enhancement of ious salivary proteins mainly through a cyclic AMP-mediated reg- membrane fusion between the plasma membrane and the mem- ulatory pathway [1,2]. Exocytosis of salivary proteins, including brane of secretory granules that are produced and pooled in ad- amylase and mucins, is evoked by various agonists that elevate vance. Intracellular membrane fusion is widely recognized to be the intracellular cAMP level in intact acinar cells [1,2], and also mediated by soluble N-ethylmaleimide-sensitive factor attachment by exogenously added cAMP analogs [3,4] and by the catalytic sub- protein receptors (SNAREs) and a variety of their associated proteins unit of cAMP-dependent protein kinase (PKA) [5] in permeabilized [7,8]. In general a set of cognate SNARE proteins, which reside either acinar cells. Cyclic AMP in the salivary glands triggers 2 separate on the membrane of transporting vesicles (v-SNAREs) or on target signaling pathways, one involving PKA and the other Epac [6]; membranes (t-SNAREs), makes a four-helix-bundle (3Q + 1R) and the PKA pathway is so far recognized to be the main route. structure that is composed of 3 Q-SNARE motifs (Qa, Qb, Qc) on 2 or 3 t-SNAREs and 1 R-SNARE motif on a v-SNARE, when the two Abbreviations: SNARE, soluble N-ethylmaleimide-sensitive factor attachment membranes dock and fuse. Gene knockout studies have ﬁrmly protein receptor; ISO, isoproterenol; CyD, cytochalasin D; GFP, green ﬂuorescent established that several SNAREs are essential for certain regulated protein; VAMP, vesicle-associated membrane protein secretory pathways, especially in the case of synaptic exocytosis ⇑ Corresponding author. Fax: +81 1332 3 1391. for neurotransmitter release [9–13]. Salivary glands contain most E-mail address: [email protected] (T. Takuma).

0014-5793/$36.00 Ó 2013 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.febslet.2013.01.039 584 T. Takuma et al. / FEBS Letters 587 (2013) 583–589

SNARE proteins [14–16], although the quantity of each SNARE in the and/or ISO, the tissue suspensions were first divided into four glands is much less than that in the brain. Previous studies using parts, and then incubated in the presence or absence of 10 lM clostridial neurotoxins and specific antibodies against Rabs and CyD for 5 min followed by incubation with or without 10 lM ISO Rab-associated proteins, those of which were introduced into for 25 min, both at 37 °C. permeabilized parotid acinar cells, suggested that some proteins, including VAMP2 and Rab27B, are involved in the exocytic process 2.3. Immunoprecipitation and immunoblotting [14,17]. Furthermore, gene knockout studies clearly established that VAMP8 is one of major v-SNAREs in exocrine glands, including the After the incubation, the parotid minces were pelleted and pancreas, salivary, and lacrimal glands; although salivary protein homogenized in a motor-driven Teflon-glass homogenizer with secretion stimulated by pilocarpine is incompletely blocked in 1 ml lysis buffer containing 1% Nonidet P-40, 10 mM Tris–HCl (pH VAMP8-null mice [12,13]. These findings suggest that several 7.4), 150 mM NaCl, 10 mM NaF, 5 mM EDTA, and protease inhibitor combinations of SNARE proteins may produce SNARE complexes cocktail (Complete Mini, Rosche). The homogenates were centri- that mediate exocytic membrane fusion in salivary glands. fuged at 15000 rpm for 15 min at 4 °C, and the resulting superna- We previously isolated a binary t-SNARE complex composed of tants were used for subsequent immunoprecipitation. The SNAP23 and syntaxin 4 from the lysate of rat parotid acinar cells, supernatants were pre-cleared by rotation at 4 °C for 30 min with but failed to detect SNARE complexes that contained both v- and 30 ll of anti-mouse or anti-rabbit IgG-agarose beads (Sigma) to re- t-SNAREs [18]. Then plausible ternary SNARE complexes contain- move non-specific binding to the agarose beads, and centrifuged at ing syntaxin 4, SNAP23, and VAMP2 were isolated, when FLAG- 5000 rpm for 10 s; then the resulting supernatants were further ro- tagged and GFP-tagged SNARE proteins were overexpressed in tated at 4 °C for 3 h with 2 lg of monoclonal anti-VAMP2 antibody COS-7 cells [19]; although the interaction of SNARE proteins is (Synaptic Systems) or polyclonal anti-SNAP23 antibody (abcam) known to be highly promiscuous under artificial experimental con- and 30 ll of anti-mouse or anti-rabbit IgG-agarose beads (Sigma). ditions [7,8]. In a recent study we succeeded to immunoprecipitate The agarose beads were washed five times with PBS containing endogenous SNARE complexes containing both v- and t-SNAREs 1% Triton X-100 and two times with PBS alone. SNARE complexes from HeLa cells [20]. Thus in this study we tried again to detect were eluted from the beads with 50 ll of SDS sample cocktail with- SNARE complexes in rat parotid glands and successfully demon- out boiling, and then the complexes were boiled for 5 min accord- strated that isoproterenol (ISO) and/or cytochalasin D (CyD) treat- ing to the instructions of the Rabbit IgG TrueBlot kit (eBioscience). ments significantly increased the transient interaction between The supernatants (2.5–5 ll for input) and aliquots of isolated SNAP23 and VAMP2. SNARE complexes, which were equivalent to 125–250 ll supernatants, were applied to a 5–20% gradient SDS–PAGE gel (ATTO, To- kyo), resolved at 30 mA for 50 min, and then transferred to a PVDF 2. Materials and methods membrane (Immobilon-P, Millipore) at 100 mA per mini-gel (90 Â 73 Â 1 mm) for 60 min in a semi-dry blotter (ATTO) with 2.1. Materials 0.1 M Tris—0.192 M glycine buffer containing 5% methanol. The membrane was blocked with Block Ace (Snow Brand Milk Products, Rabbit polyclonal antibodies against SNAREs were purchased Co., Tokyo) or 5% skim milk in PBS at room temperature for 2 h or from Synaptic Systems (Goettingen, Germany). In addition, anti- overnight at 4 °C, and then incubated for 1 h at room temperature SNAP23 (rabbit polyclonal; ab3340) and anti-syntaxin 3 (rabbit with primary antibodies properly diluted in PBS containing 0.05% polyclonal; S5547) were obtained from Abcam and Sigma, respec- Tween-20 (PBST) and 20% Block Ace. Next the membrane was tively. Rabbit IgG TrueBlot was from eBioscience (San Diego, CA, washed three times with 50 ml PBST for 15 min each time, and USA). was further incubated for 1 h with secondary antibodies specific for mouse IgG or rabbit IgG (TruBlot, eBioscience) according to the 2.2. Preparation of parotid minces and treatment with NEM, CyD, and manufacturer’s protocol. It was again washed three times with PBST ISO as above, and HRP activity of the secondary antibodies was visual- ized with the Immobilon Western (Millipore). Chemiluminescent Male Wistar strain rats were anesthetized with ether and then images were captured and analyzed by an ATTO Cool Saver (ATTO). sacrificed by heart puncture. The experiments with rats had received the approval of the Animal Care and Use Committee of 2.4. Determination of amylase release the Health Sciences University of Hokkaido and were conducted in accordance with the Guideline for the Care and Use of Labora- Parotid tissue suspensions (1 ml) were preincubated for 5 min tory Animals of the Health Sciences University of Hokkaido. Rat with or without 10 lM CyD at 37 °C, and then further incubated with parotid glands were removed, carefully trimmed, and then minced or without 10 lM ISO for 20 min at 37 °C. After incubation the tissue thoroughly with fine scissors (500 times) in a CO2-independent suspensions were thoroughly mixed and centrifuged at 2000 rpm medium (#18045, Gibco) at room temperature. The parotid minces for 1 min. The supernatant was used for the amylase assay. For the were washed with the same medium (10 ml) two times by centri- determination of total amylase activity, the tissue suspension con- fugation at 1000 rpm for 10 s to remove the fatty tissues. The tissue taining 1% Triton X-100 was homogenized in a Polytron homoge- suspension was equally divided into two parts. One part, desig- nizer, and the homogenate was centrifuged at 15000 rpm for nated NEM(+), was treated with 0.1 mM N-ethylmaleimide 5 min. The supernatant was used for the amylase assay. Amylase (NEM) for 15 min and then treated with 1 mM dithiothreitol activity in the properly diluted samples was measured by the meth- (DTT) for 15 min, with both treatments at room temperature; be- od of Bernfeld as described previously [3], and the released activity cause the exposure of parotid acinar cells to cold itself induces was expressed as a percentage of the total activity. amylase exocytosis [21]. The other part, designated as NEM(À), was treated with pre-mixed NEM and DTT for 30 min at room tem- 2.5. Statistical analysis perature. The NEM-treated tissues were further divided into two parts, and one part was incubated in the presence, and the other The statistical significance of the differences between control in the absence, of 10 lM isoproterenol (ISO) at 37 °C for 30 min. and experimental results was estimated by the Welch two-sample When parotid minces were treated with cytochalasin D (CyD) t-test using R (version 2.6.1). T. Takuma et al. / FEBS Letters 587 (2013) 583–589 585

3. Results because we had previously detected only a t-SNARE complex of syntaxin 4 and SNAP23 in rat parotid cells [18]. As shown in 3.1. Effect of NEM treatment on immunoprecipitation of SNARE Fig. 1A, antibodies against syntaxin 4 and SNAP23 clearly coprecip- complexes itated each other and further precipitated v-SNAREs (i.e., VAMP3 and VAMP8). Anti-syntaxin 4 preferentially coprecipitated VAMP3; In this study we used finely minced rat parotid glands instead of and anti-SNAP23, VAMP8. Antibody against syntaxin 3 efficiently enzyme-digested parotid cells to maintain intact polarity of the immunoprecipitated syntaxin 3 itself and weakly coprecipitated parotid cells and to avoid abnormal SNARE complex formation be- SNAP23 but none of the v-SNAREs examined. In these experiments, tween SNAREs on the secretory granules and those on the basolat- however, VAMP2 was scarcely detectable, as reported previously eral plasma membrane. To stabilize SNARE complexes by [18]. inhibiting N-ethylmaleimide sensitive factor (NSF), which dissoci- On the contrary, anti-VAMP2 antibody obviously coprecipitated ates cis-SNARE complexes after membrane fusion for the recycling t-SNAREs, including SNAP23, syntaxin 3, and syntaxin 4 (Fig. 1B). of SNARE proteins for subsequent membrane fusion, we treated the Especially the immunoprecipitation of syntaxin 4 was markedly parotid tissues with NEM(À) or NEM(+) for 30 min at room tem- augmented by the NEM treatment, suggesting that the interaction perature and then incubated them for 30 min at 37 °C in the pres- between syntaxin 4 and VAMP2 was highly susceptible to NSF- ence or absence of 10 lM ISO. We performed the NEM treatment at mediated dissociation of the SNARE complexes. The interaction be- room temperature instead of 0 °C, because exposure to cold by it- tween VAMP2 and SNAP23 appeared to be less sensitive to the self induces amylase exocytosis from parotid cells [21]. NEM treatment, implying that the interaction was rather stable We first examined whether it was possible to immunoprecipi- even in the presence of intact NSF. The immunoprecipitation tate the SNARE complexes that contained both v- and t-SNAREs, of these SNARE complexes, however, was not increased by ISO

A Antibodies for IP 4% Input Con S3 S4 SN

Syn3

Syn4

SN23

VA-2

VA-3

VA-8

NEM +++++

BC 4% IP: anti-SNAP23 Input IP: anti-VAMP2

VA-2 SN23

SN23 Syn4

Syn3 VA-3

Syn4 VA-8

ISO -+- + ISO -+ NEM -++- NEM + +

Fig. 1. Immunoprecipitation of SNARE complexes from the lysates of rat parotid glands. Parotid minces were pre-treated with NEM(+) or NEM(À) at room temperature and further incubated with or without 10 lM isoproterenol (ISO) for 30 min at 37 °C. After incubation tissue lysates were prepared for the immunoprecipitation of SNARE complexes, as described in Section 2. (A) SNARE proteins (equivalent to 125 ll supernatant) were immunoprecipitated with various antibodies (aCon, control rabbit IgG; aS3, anti-syntaxin 3; aS4, anti-syntaxin 4; and aSN, anti-SNAP23) and detected with the same and additional antibodies by immunoblotting. Syn3, syntaxin 3; Syn4, syntaxin 4; SN23, SNAP23; VA-2, VAMP2; VA-3, VAMP3; VA-8, VAMP8; 4% Input, 5 ll supernatant. (B) Immunoprecipitation with monoclonal anti-VAMP2. (C) Immunoprecipitation with polyclonal anti-SNAP23. 586 T. Takuma et al. / FEBS Letters 587 (2013) 583–589

A B CyD + ISO 50

Amylase release (% of total) Amylase release 0 CON ISO CyD CyD + ISO

Fig. 2. Effects of cytochalasin D (CyD) and ISO. (A) Schematic presentation of morphological effects of CyD and ISO on parotid acinar cells. SG, secretory granule. (B) Effects of CyD and ISO on amylase release from rat parotid glands. Parotid tissue suspension (1 ml) were preincubated with or without 10 lM CyD for 5 min at 37 °C, and then further incubated with or without 10 lM ISO for 20 min at 37 °C. After incubation the tissue suspensions were centrifuged at 2000 rpm for 1 min, and amylase activity in the supernatant was then determined. CON, control. Data are means ± S.D. (n = 4).

treatment either in the presence or absence of NEM (Fig. 1B). Sim- with CyD plus ISO, although the amount of SNARE complex con- ilarly the SNARE complexes that were immunoprecipitated with taining syntaxin 4 was not increased under the same conditions. anti-SNAP23 antibody also failed to respond to ISO treatment Similarly, polyclonal anti-SNAP23 antibody highly concentrated (Fig. 1C). SNAP23 and weakly coprecipitated SNARE complexes containing syntaxin 4 and VAMP8, although it remained unknown whether 3.2. Effect of CyD on amylase release and SNARE complex formation these complexes were binary or ternary ones (Fig. 3C). SNAP23 induced by ISO interacted constantly with syntaxin 4 and VAMP8 at readily detectable levels, but these interactions were insensitive to either ISO or Previous studies showed that amylase release from parotid CyD treatment. To our great surprise, an extremely small amount glands is partially decreased by cytochalasin D (CyD), which depo- of VAMP2 was detected by a Light-Capture II (ATTO) when the lymerizes actin filaments; although the exocytic membrane fusion blotted membranes were exposed for a prolonged time (more than between the luminal plasma membrane and secretory granules is 15 min at the high-sensitivity setting). Only less than 0.01% of the not inhibited, at least morphologically [22–25]. Hence the lumen total VAMP2 in parotid lysates was coprecipitated with SNAP23 of parotid acini was enlarged with invagination of the luminal plas- under the unstimulated condition, where up to 20% of the total ma membrane that would have been derived from sequential fu- SNAP23 in the lysates was immunoprecipitated (Fig. 3C). Further- sions of secretory granule membranes, as revealed by high- more, the coprecipitation of VAMP2 was markedly and signifi- density substances, i.e., secretory cargos, inside the lumen cantly increased in response to ISO and/or CyD treatments, (Fig. 2A). From these findings we hypothesized that CyD might suggesting that the interaction between SNAP23 and VAMP2 was have inhibited the recycling process of SNARE proteins in addition very unstable and that the complex was transiently formed when to the retrieval pathway of secretory granule membranes. Thus we the exocytic membrane fusion was stimulated by ISO and/or CyD next examined the effect of CyD on the level of SNARE complexes (Fig. 3D). induced by ISO. Parotid gland minces were preincubated for 5 min As shown in Fig. 3E, polyclonal anti-syntaxin 4 antibody with or without 10 lM CyD at 37 °C and then further incubated for markedly concentrated syntaxin 4 and weakly coprecipitated 25 min in the presence or absence of 10 lM ISO at 37 °C. As re- SNAP23. In addition, small amounts of VAMP2 and VAMP8 were ported previously [22–25], CyD only partially inhibited the ISO-in- detected by prolonged exposure; although the amounts of these duced amylase release from the parotid minces, probably because SNARE complexes were not increased by ISO and/or CyD treat- some of the secreted amylase had remained in the enlarged lumen ments. On the other hand, polyclonal anti-VAMP8 antibody used of the parotid gland (Fig. 2B). in this study readily immunoprecipitated VAMP8 but did not Based on the quantitative analysis by immunoblotting, mono- coprecipitate any t-SNAREs examined in the parotid lysates (data clonal anti-VAMP2 antibody immunoprecipitated more than 50% not shown). of the total VAMP2 in parotid lysates, since the chemiluminescent intensity of the IP bands of VAMP2 was more than 50 times as 4. Discussion strong as that of the 1% input (Fig. 3A). However, the chemiluminescent intensity of coprecipitation bands of SNAP23 and syntaxin 4.1. Demonstration of increase in SNARE complex formation by 4 was weaker than that of their 1% inputs, indicating that only less secretory stimuli than 1% of the total SNAP23 or syntaxin 4 in the lysates was coprecipitated with VAMP2 (Fig. 3A). These results suggest that the In the present study we for the first time clearly demonstrated majority of the VAMP2 immunoprecipitated was not in SNARE that the interaction of t-SNARE (i.e., SNAP23) and v-SNARE (i.e., complexes with other SNAREs tested, although other SNAREs or VAMP2) was significantly increased in response to secretory stim- SNARE associated proteins untested might be bound to VAMP2. uli (ISO and/or CyD treatments) in rat parotid glands, although the As can be seen in Fig. 3B, the SNARE complex containing SNAP23 combinatorial syntaxin involved in this SNARE complex remained was slightly but significantly increased in amount by the treatment unidentified. Although an increase in SNARE complex formation T. Takuma et al. / FEBS Letters 587 (2013) 583–589 587

AB 2 1% Input IP : anti-VAMP2 *

VA-2 1.5

1 SN23

0.5 Syn4 SNAP23 in SNARE complex 0 ISO - + - + CON ISO CyD CyD+ISO CyD - - + +

C D

Fig. 3. Effects of CyD and ISO on the amount of SNARE complexes. Parotid minces were preincubated with 10 lM CyD for 5 min at 37 °C and further incubated with 10 lM ISO for 25 min at 37 °C. After incubation tissue lysates were prepared for the immunoprecipitation of SNARE complexes, as described in Section 2. (A) SNARE complexes (equivalent to 250 ll supernatant) were immunoprecipitated with monoclonal anti-VAMP2 and each SNARE protein was detected by immunoblotting. I% Input, 2.5 ll supernatant. (B) Quantitative determination of SNAP23 in SNARE complexes immunoprecipitated by anti-VAMP2. Data are means + S.D. (n = 3). ⁄P = 0.013 vs. control. (C) Immunoprecipitation of SNARE complexes with anti-SNAP23. (D) Quantitative determination of VAMP2 in SNARE complexes immunoprecipitated by anti-SNAP23. Data are means + S.D. (n = 3). ⁄P = 0.0017, ⁄⁄P = 0.0069, ⁄⁄⁄P = 0.0025 vs. control. (E) Immunoprecipitation of SNARE complexes with anti-syntaxin 4. 588 T. Takuma et al. / FEBS Letters 587 (2013) 583–589 by secretory stimuli is strong evidence to support the involvement stable pool of SNAP23–VAMP8 complex exists on the intracellular of SNARE proteins in the secretory process, the secretion-dependent membranes of rat parotid glands and thus a small transient in- increase in SNARE complex has not been readily detectable, proba- crease in the SNARE complex on the plasma membrane would be bly because the SNARE complexes that mediate exocytic membrane masked; and (2) the interaction between SNAP23 and VAMP8 is fusion might be highly unstable and rapidly recycled. The anti- not critical for ISO-stimulated exocytosis from the parotid gland. SNAP23 polyclonal antibody could immunoprecipitate various SNARE complexes containing syntaxin 4, VAMP2, and VAMP8 4.3. Role of SNARE complex of SNAP23 and VAMP2 (Fig. 3C and D); and among them the complex of SNAP23–VAMP2 was the only one increased by ISO and/or CyD. The steady-state level Since the epitope (aa193–210) of anti-SNAP23 polyclonal anti- of this complex without ISO or CyD was extremely low, suggesting body (Abcam, ab3340) overlapped partly with the distal Qc SNARE that the SNAP23–VAMP2 complex was very unstable and transiently motif (aa156–208) of SNAP23, the accessibility of the antibody to formed under physiological conditions when the recycling process its epitope would have been markedly attenuated by steric hin- of SNARE proteins was not prevented. drance when SNAP23 made complexes with other SNARE proteins; In this study, however, we obtained some contradictory results; although the t-SNARE complex containing syntaxin 4 was always i.e., when complexes were immunoprecipitated with monoclonal immunoprecipitated (Figs. 1A, C and 3C). The anti-SNAP23 co- anti-VAMP2, the interaction between VAMP2 and SNAP23 was immunoprecipitated v-SNAREs in order of VAMP8 > VAMP3 readily detectable, but this interaction responded very weakly to VAMP2 (Figs. 1A, C and 3C), suggesting that this order reflected (1) the ISO plus CyD treatment (Fig. 3A and B). In contrast, with the steady-state pool size of the SNARE complexes in parotid cells anti-SNAP23 polyclonal antibody used for the immunoprecipita- or (2) the accessibility of the antibody to the epitope in each binary tion, the interaction between SNAP23 and VAMP2 was hardly or ternary SNARE complex, in which the coiled-coil structure in the detectable under the steady-state condition but increased mark- vicinity of distal Qc SNARE motif might vary in response to the bind- edly by the treatment with ISO and/or CyD, as described above ing of different sets of combinatorial VAMPs and syntaxins. (Fig. 3C and D). The anti-VAMP2 readily immunoprecipitated As described above, the steady-state level of the VAMP2– SNAP23, suggesting that the steady-state level of the VAMP2– SNAP23 complex is not very low (Figs. 1B and 3A), but the major SNAP23 complex itself was not very low (Figs. 1B and 3A); but this portion of this population of SNARE complexes did not respond VAMP2–SNAP23 complex could have been derived from various to secretory stimuli (Figs. 1B and 3A), suggesting that these com- intracellular membranes, including endosomal membranes and plexes could have been derived from various intracellular mem- secretory granule membranes in addition to the plasma membrane branes [12,16,26]. Furthermore, most of the SNARE complexes [12,16,26]. The SNARE complex on endosomal membranes and would have been hardly recognized by the anti-SNAP23 (Fig. 3C) secretory granule membranes would be less sensitive to the because of steric hindrance around the Qc SNARE motif, as men- secretory stimuli than that on the plasma membrane. These results tioned above. Based on these findings we speculate that the anti- suggest that the portion of the complex from the plasma mem- SNAP23 exclusively immunoprecipitated the stimulation-depen- brane would be relatively low compared with that of other intra- dent SNAP23–VAMP2 complex on the plasma membrane and that cellular membranes and that the monoclonal anti-VAMP2 and this complex would have been a transient ternary complex, where polyclonal anti-SNAP23 antibodies would have immunoprecipi- the steric hindrance for the binding of anti-SNAP23 might have tated VAMP2–SNAP23 from different pools of SNARE complexes. been loosened by the binding of additional, as yet unidentified, fac- tors, including syntaxins. Further study is necessary to demon- 4.2. Stability of binary and ternary SNARE complexes strate this speculation.

The present immunoprecipitation studies revealed that most of Acknowledgment the antibodies against SNARE proteins greatly concentrated their corresponding antigens but did not do so in the case of other This study was supported in part by HAITEKU from the Ministry SNARE proteins coprecipitated (Figs. 1A, B, 3A, C and E), suggesting of Education, Culture, Sports Science and Technology of Japan. that these antibodies immunoprecipitated much less efficiently SNARE complexes, either binary or ternary ones. The efficiency of SNARE complex formation is theoretically dependent on the con- References centration of each SNARE protein and the stability of the complex. The NEM treatment markedly increased the amount of SNARE [1] Butcher, F.R. and Putney Jr., J.W. (1980) Regulation of parotid gland function by cyclic nucleotides and calcium. Adv. Cyclic Nucleotide Res. 13, 215–249. complexes containing VAMP2 and syntaxin 4 (Fig. 1B), suggesting [2] Harper, J.F. (1988) Stimulus-secretion coupling: second messenger-regulated that the interaction between VAMP2 and syntaxin 4 was relatively exocytosis. Adv. Second Messenger Phosphoprotein Res. 22, 193–318. unstable in the presence of intact NSF, which dissociates cis-SNARE [3] Takuma, T. and Ichida, T. (1988) Amylase secretion from saponin- permeabilized parotid cells evoked by cyclic AMP. J. Biochem. 103, 95–98. complexes. In the present study and in previous ones [16,18] the [4] Takuma, T. and Ichida, T. (1991) Cyclic AMP antagonist Rp-cAMPS inhibits data showed that the binary t-SNARE complexes containing amylase exocytosis from saponin-permeabilized parotid acini. J. Biochem. 110, SNAP23 and various syntaxins were readily detectable (Figs. 1A, 292–294. [5] Takuma, T. and Ichida, T. (1994) Catalytic subunit of protein kinase A induces C, 3C and E), suggesting that the t-SNARE complexes are relatively amylase release from streptolysin O-permeabilized parotid acini. J. Biol. Chem. abundant and stable in rat parotid acinar cells. On the other hand, 269, 22124–22128. SNARE complexes containing both syntaxins (t-SNAREs) and [6] Shimomura, H., Imai, A. and Nashida, T. (2004) Evidence for the involvement of VAMPs (v-SNAREs) can hardly be isolated except in the case of cAMP-GEF (Epac) pathway in amylase release from the rat parotid gland. Arch. Biochem. Biophys. 431, 124–128. the complex containing syntaxin 3 and VAMP8 [16]. To our sur- [7] Hong, W. (2005) SNAREs and traffic. Biochim. Biophys. Acta 1744, 493–517. prise, however, the complex of SNAP23–VAMP8 was relatively [8] Jahn, R. and Scheller, R.H. (2006) SNAREs—engines for membrane fusion. Nat. abundant and stable in this study (Figs. 1A, C and 3C), as well as Rev. Mol. Cell Biol. 7, 631–643. [9] Deitcher, D.L., Ueda, A., Stewart, B.A., Burgess, R.W., Kidokoro, Y. and Schwarz, in a previous one [16]. Although gene knockout studies clearly T.L. (1998) Distinct requirements for evoked and spontaneous release of demonstrated that VAMP8 plays an important role in exocrine neurotransmitter are revealed by mutations in the Drosophila gene neuronal- secretion [12,13], the ISO and/or CyD treatment did not signifi- synaptobrevin. J. Neurosci. 18, 2028–2039. [10] Schoch, S., Deak, F., Konigstorfer, A., Mozhayeva, M., Sara, Y., Sudhof, T.C. and cantly increase the level of SNAP23-VAMP8 complex (Figs. 1C Kavalali, E.T. (2001) SNARE function analyzed in synaptobrevin/VAMP and 3C), suggesting the following two possibilities; (1) a large knockout mice. Science 294, 1117–1122. T. Takuma et al. / FEBS Letters 587 (2013) 583–589 589

[11] Yang, C., Mora, S., Ryder, J.W., Coker, K.J., Hansen, P., Allen, L.A. and Pessin, J.E. [19] Takuma, T., Arakawa, T., Okayama, M., Mizoguchi, I., Tanimura, A. and Tajima, (2001) VAMP3 null mice display normal constitutive, insulin- and exercise- Y. (2002) Trafficking of green fluorescent protein-tagged SNARE proteins in regulated vesicle trafficking. Mol. Cell. Biol. 21, 1573–1580. HSY cells. J. Biochem. 132, 729–735. [12] Wang, C.C., Ng, C.P., Lu, L., Atlashkin, V., Zhang, W., Seet, L.F. and Hong, W. [20] Okayama, M., Shitara, A., Arakawa, T., Tajima, Y., Mizoguchi, I. and Takuma, T. (2004) A role of VAMP8/endobrevin in regulated exocytosis of pancreatic (2012) SNARE proteins are not excessive for the formation of post-Golgi acinar cells. Dev. Cell 7, 359–371. SNARE complexes in HeLa cells. Mol. Cell. Biochem. 366, 159–168. [13] Wang, C.C., Shi, H., Guo, K., Ng, C.P., Li, J., Gan, B.Q., Chien Liew, H., Leinonen, J., [21] Takuma, T., Yajima, T. and Ichida, T. (1996) Induction of amylase release from Rajaniemi, H., Zhou, Z.H., Zeng, Q. and Hong, W. (2007) VAMP8/endobrevin as rat parotid acinar cells by cooling in vitro. Arch. Oral Biol. 41, 547–551. a general vesicular SNARE for regulated exocytosis of the exocrine system. [22] Segawa, A. and Yamashina, S. (1989) Roles of microfilaments in exocytosis: a Mol. Biol. Cell 18, 1056–1063. new hypothesis. Cell Struct. Funct. 14, 531–544. [14] Fujita-Yoshigaki, J., Dohke, Y., Hara-Yokoyama, M., Kamata, Y., Kozaki, S., [23] Okumura, K., Tojyo, Y. and Kanazawa, M. (1990) Changes in microfilament Furuyama, S. and Sugiya, H. (1996) Vesicle-associated membrane protein 2 is distribution during amylase exocytosis in rat parotid salivary glands in vitro. essential for cAMP-regulated exocytosis in rat parotid acinar cells. The Arch. Oral Biol. 35, 677–679. inhibition of cAMP-dependent amylase release by botulinum neurotoxin B. J. [24] Tojyo, Y., Okumura, K., Kanazawa, M. and Matsumoto, Y. (1989) Effect of Biol. Chem. 271, 13130–13134. cytochalasin D on acinar cell structure and secretion in rat parotid salivary [15] Takuma, T., Tagaya, M. and Ichida, T. (1997) Evidence for the putative docking/ glands in vitro. Arch. Oral Biol. 34, 847–855. fusion complex of exocytosis in parotid acinar cells. FEBS Lett. 404, 34–36. [25] Nashida, T., Yoshie, S., Imai, A. and Shimomura, H. (2004) Presence of [16] Imai, A., Nashida, T., Yoshie, S. and Shimomura, H. (2003) Intracellular cytoskeleton proteins in parotid glands and their roles during secretion. Arch. localisation of SNARE proteins in rat parotid acinar cells: SNARE complexes on Oral Biol. 49, 975–982. the apical plasma membrane. Arch. Oral Biol. 48, 597–604. [26] Weng, N., Thomas, D.D. and Groblewski, G.E. (2007) Pancreatic acinar cells [17] Imai, A., Yoshie, S., Nashida, T., Shimomura, H. and Fukuda, M. (2004) The express vesicle-associated membrane protein 2- and 8-specific populations of small GTPase Rab27B regulates amylase release from rat parotid acinar cells. J. zymogen granules with distinct and overlapping roles in secretion. J. Biol. Cell Sci. 117, 1945–1953. Chem. 282, 9635–9645. [18] Takuma, T., Arakawa, T. and Tajima, Y. (2000) Interaction of SNARE proteins in rat parotid acinar cells. Arch. Oral Biol. 45, 369–375.