Synaptotagmin-like protein 4 and Rab8 interact and increase dense granule release in platelets

Authors


Albert Smolenski, UCD Conway Institute, School of Medicine and Medical Science, University College Dublin, Belfield, Dublin 4, Ireland.
Tel.: +353 1 716 6746; fax: +353 1 716 6701.
E-mail: albert.smolenski@ucd.ie

Abstract

Summary.  Background: Platelets are highly specialized cells that regulate hemostasis and thrombosis in the vasculature. Upon activation, platelets release various granules that impact on platelets, the coagulation system, other blood cells and the vessel wall; however, the mechanisms controlling granule release are only partially known. We have shown previously that synaptotagmin-like protein (Slp)1 decreases dense granule release in platelets. Objectives: To determine the role of other Slps and their binding partners on platelet dense granule release. Methods: RT-PCR and immunoblotting were used to identify Slps in human platelets. Interaction between Slp4 and Rab8 was investigated with pull-down assays, coimmunoprecipitation, and confocal microscopy. Secretion assays on permeabilized platelets were performed to investigate the effects of Slp4 and Rab8 on dense granule release. Results: Slp4 mRNA and protein are expressed in human platelets. Slp4 interacts with Rab8 in transfected cells and at endogenous protein levels in platelets. We mapped the Rab interaction site to the Slp-homology domain of Slp4, and showed preferential binding of Slp4 to the GTP-bound form of Rab8. Live microscopy showed colocalization of green fluorescent protein–Slp4 and mCherry–Rab8 at the plasma membrane of transfected cells. Endogenous platelet Slp4 and Rab8 colocalized in the center of activated platelets, where granule secretion takes place. Secretion assays revealed that Slp4 and Rab8 enhance dense granule release and that the Slp4 effect is dependent on Rab8 binding. Conclusions: Slp4 and Rab8 are expressed and interact in human platelets, and might be involved in dense granule release.

Introduction

Platelets play a central role in hemostasis, and are critical components in the development of atherothrombosis [1]. Platelets contain dense granules and α-granules, and, upon activation, release their granule contents to promote thrombus formation. Dense granules contain ADP, 5-hydroxytryptamine (5-HT, serotonin), calcium ions, and pyrophosphate, whereas α-granules contain chemokines, coagulation factors, growth factors, and other proteins [2]. Inherited defects in platelet granule formation lead to bleeding disorders, e.g. Hermansky–Pudlak or Chediak–Higashi syndromes, affecting dense granules, or the gray platelet syndrome for α-granules [3]. Platelet granule release is thought to depend on mechanisms that are similar to synaptic vesicle secretion [4,5]. Granules are transported to the plasma membrane, where they are tethered; this is followed by docking and fusion of granule and plasma membranes, resulting in the release of granule contents into the extracellular space. The fusion of granules with the plasma membrane requires soluble N-ethylmaleimide-sensitive attachment protein receptors (SNAREs). SNARE family members that have been suggested to play a role in platelet granule release include VAMP-2, VAMP-3, VAMP-8, syntaxin 2, syntaxin 4, syntaxin 7, syntaxin 11, SNAP-23, and SNAP-29 [6]. In addition to the SNAREs, membrane fusion requires members of the Sec1/Munc18 family, which direct SNARE function [7], and platelets have been shown to express various isoforms of Munc18 [8]. Another group of proteins that are emerging as important regulators of the membrane fusion process are the synaptotagmins and related multiple C2 domain proteins. Synaptotagmins are characterized by two tandem C-terminal C2 domains, which interact with membrane phospholipids and are involved in membrane curvature induction [9,10]. Synaptotagmins have also been reported to interact with SNAREs and to act as Ca2+ sensors. Synaptotagmin-like proteins (Slps) constitute a related group of proteins containing tandem C2 domains and an N-terminal Slp-homology domain (SHD) that binds small GTP-binding proteins of the Rab family [11–13]. Slp1 has been shown to inhibit dense granule release [14], and Munc13-4, a protein with separate N-terminal and C-terminal C2 domains, positively regulates α-granule and dense granule release in platelets [15]. Slp1 and Munc13-4 interact with Rab27, a small G-protein involved in dense granule biogenesis and release [16]. Rab proteins control vesicle trafficking by binding to internal membranes as well as to numerous effector proteins [17].

In this article, we describe the identification and characterization of Slp4 (also called granuphilin) in platelets. We show that Slp4 interacts with Rab8, and that both proteins might play a role in the regulation of dense granule release.

Materials and methods

Antibodies, constructs, and materials

A polyclonal rabbit antibody against human Slp4 was developed by OpenBiosystems (Lafayette, CO, USA), with the peptide AEGTLQLRSSMAKQKLGL at position 654 of Slp4. Other antibodies used in this study included: anti-Slp4 (HPA001475, rabbit; Atlas Antibodies, Stockholm, Sweden), anti-Slp1 [14], anti-Rab8A human (H00004218-M02, mouse; Tebu-Bio, Peterborough, UK), anti-FLAG tag (M2; Sigma-Aldrich, Arklow, Ireland), Alexa Fluor 488-conjugated donkey anti-rabbit IgG (A-21206; Life Technologies, Dun Laoghaire, Ireland), Alexa Fluor 568-conjugated donkey anti-mouse IgG (A10037; Life Technologies), and horseradish peroxidase-conjugated donkey anti-rabbit and donkey anti-mouse (Jackson ImmunoResearch Europe, Newmarket, Suffolk, UK).

Constructs and protein purification

Full-length human Slp4-a was obtained from Origene (Rockville, MD, USA, SC120377, SYTL4). Slp4 was FLAG-tagged or green fluorescent protein (GFP)-tagged at the N-terminus, and expressed by use of the mammalian expression vector pcDNA4/TO (Invitrogen). Hemagglutinin (HA)-tagged Rab8A (N-terminus) was purchased from the University of Missouri-Rolla, and mCherry–Rab8 was expressed in pmCherryC1 (Life Technologies). Site-directed mutagenesis was performed by PCR amplification of Rab8 constructs with mutagenic primer pairs, Pfu DNA polymerase (Fermentas, Thermo Fisher Scientific, Dublin, Ireland), digestion with DpnI (Fermentas), and transformation into TOP10 bacteria (Life Technologies). All constructs were verified by DNA sequencing (Eurofins MWG Operon, Ebersberg, Germany). Slp4, Slp1 and Rab8 glutathione-S-transferase (GST) fusion proteins were generated with the pGEX-4T3 vector (GE Healthcare, Little Chalfont, UK) expressed in Escherichia coli BL21, and purified as described previously [14,18].

Cell preparation, transfection, lysis, immunoprecipitation, and pull-down experiments

HEK293T and HeLa cells were cultured in DMEM supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin, at 37 °C and 5% CO2 in air. Cells were transfected with Metafectene (Biontex, Martinsried, Germany) or Fugene (Promega, Southampton, UK), according to the manufacturer’s instructions. Venous blood was drawn from healthy volunteers taking no medications who gave their informed consent according to the declaration of Helsinki. Forty milliliters of freshly drawn venous blood from healthy volunteers was collected into 10 mL of prewarmed CCD-EGTA buffer (100 mm trisodium citrate, 7 mm citric acid, 140 mm glucose, 15 mm EGTA), and centrifuged at 150 ×g and room temperature for 15 min. Platelet-rich plasma (PRP) was recovered, and platelets were pelleted at 600 ×g for 10 min and resuspended in resuspension buffer (145 mm NaCl, 5 mm KCl, 1 mm MgCl2, 10 mm Hepes, 10 mm glucose, pH 7.4) to a final concentration of 2 × 108 platelets mL–1. Cell lysis, immunoprecipitation and pull-down assays were performed as previously described [19], with 5 μL of Anti-FLAG M2 Affinity Gel (Sigma-Aldrich), 10 μL of anti-Slp4 polyclonal antibody or 5 μL of gluthatione–Sepharose 4B suspension (GE Healthcare) saturated with GST fusion proteins.

RT-PCR analysis

Total cellular RNA was isolated from 1 mL of freshly washed platelets (approximately 1 × 109 platelets) with 40 mL of Trizol reagent, as described by the manufacturer (Life Technologies). Total RNA from human brain was obtained from BD Biosciences (Oxford, UK). Total RNA was reverse transcribed with RevertAid H Minus Reverse Transcriptase (Fermentas), and amplified with random hexamer primers (Fermentas). The cDNA was then used for RT-PCR amplification with specific sense and antisense primers for Slp2 (forward, 5′-CAGCACCAGCAAGCCCGAGT-3′; reverse, 5′-CCCTCTCGGAGCCCCTCTCG-3′), Slp3 (forward, 5′-GGCAGCTGCAGGTCTCGGTG-3′; reverse, 5′-CAAGGTGCCATCTGGCCGCA-3′), Slp4 (forward, 5′-GCCTGGGCCGTTTGAGTCCC-3′; reverse, 5′-CACTGGGCTCCTTCTGCCGC-3′), and Slp5 (forward, 5′-AGGCCCCAATGGCAGCTGGA-3′; reverse, 5′-CGTGTGCTTGGGGCTGGTGA-3′). PCR amplification was performed for 25 cycles at an annealing temperature of 55–65 °C. Positive bands were excised and verified by sequencing.

Confocal microscopy

HeLa cells were grown on eight-well microslides (Ibidi, Planegg, Germany), and transfected with GFP–Slp4 and mCherry–Rab8. After 18–24 h, living cells were visualized with an Andor (Belfast, UK) spinning disk microscope, with 488-nm and 594-nm lasers and a Plan Fluor × 100/1.30 oil objective. The obtained images were analyzed with andor iq2 software. Platelets were allowed to spread on glass coverslips, fixed and stained as described previously [20], with rabbit Slp4 and mouse Rab8 primary antibodies followed by anti-rabbit Alexa 488-conjugated and anti-mouse Alexa 564-conjugated secondary antibodies, and mounted onto slides with Fluoromount (Sigma-Aldrich). Stained platelets were visualized as described above. Cross-reactivities of secondary antibodies were ruled out by swapping of secondary antibodies in single-labeling experiments. For unactivated platelets, an equal volume of CCD-EGTA blood was mixed with 3.7% of room temperature paraformaldehyde prepared in phosphate-buffered saline (PBS), and fixed for 2 h at room temperature. Subsequent steps were carried out at 4 °C. PRP supernatant was obtained by centrifugation at 300 ×g for 15 min. The PRP was centrifuged at 1300 ×g for 12 min to pellet the platelets. The pellet was resuspended in resuspension buffer, and an equal volume of 4% paraformaldehyde in PBS was added to the final platelet suspension for 30 min at room temperature. Platelets were then diluted in 990 μL of PBS, dropped onto coverslips, permeabilized, and stained as described above.

Dense granule secretion assay

Washed platelets were permeabilized with 0.6 μg mL−1 streptolysin-O, kindly provided by S. Bhakdi (Mainz) [21], and incubated with purified GST fusion proteins on ice; secretion assays were then performed as previously described [14], with 18 mm Ca2+, which corresponds to a final concentration of ∼ 1.3 μm free Ca2+ [22], at 30 °C for 2 min to stimulate granule release. Recombinant proteins had no effect on 5-HT secretion in the absence of Ca2+ (Fig. S1). The average efficacy of 5-HT release ranged between 30% and 50% of total 5-HT. The secreted levels of 5-HT were normalized against samples incubated with GST only and stimulated with Ca2+. Data are the mean of at least three independent experiments performed in triplicate. The statistical significance of the means was analyzed by analysis of variance and the Bonferroni post hoc test (95% confidence interval) (Figs. 1 and 4A,B) and by a two-tailed t-test (Fig. 4C) with graphpad prism software, Version 5.0 (GraphPad Software, San Diego, CA, USA). P-values of < 0.05 were considered to be statistically significant.

Results

Slp4 is present in human platelets and increases dense granule release

We have shown previously that Slp1 controls dense granule release in platelets [14]. To investigate a potential role of other Slp family members in platelets, we isolated RNA from washed human platelets, and performed RT-PCR analysis. In addition to Slp1, only Slp4, and not Slp2, Slp3, or Slp5, was found to be expressed in platelets (Fig. 1A and data not shown). Following the detection of Slp4 mRNA, an Slp4 antibody was used to confirm the presence of endogenous Slp4 in platelets at the protein level by immunoblotting. The antibody recognized Slp4 both in cells transfected with FLAG–Slp4 and in human platelet lysate, with no Slp4 being detected in non-transfected HEK293T cells (Fig. 1B). Our previous work had suggested that Slp1 inhibits dense granule release [14]. To analyze a possible role of Slp4 in granule release, we performed a secretion assay with streptolysin-O-permeabilized platelets and purified GST–Slp4. Slp4 significantly enhanced the Ca2+-stimulated release of 5-HT, a marker for dense granules (Fig. 1C), whereas GST alone had no significant effect (Fig. S2). The enhancing action of Slp4 was dose-dependent (Fig. S3). In the absence of Ca2+, Slp4 did not stimulate 5-HT release (Fig. S1). Deletion of the SHD of Slp4 abolished the enhancing effect of Slp4 (Fig. 1C, 2nd bar). As SHDs of Slps have been shown to interact with Rab proteins, these data suggest that Slp4 might require its SHD to interact with a Rab protein to increase dense granule release.

Figure 1.

 Synaptotagmin-like protein (Slp)4 is expressed in human platelets, and enhances dense granule release. (A) Expression of Slp4 mRNA in human platelets. RNA was isolated from washed platelets, and RT-PCR was carried out with specific primers for Slp4. Brain RNA was used as a positive control, with the negative control containing no template. The indicated band corresponds to the expected size of the PCR product of 336 bp, and DNA sequencing confirmed the identity of this band as Slp4. The lower band in the brain sample might represent a splice isoform of Slp4. (B) Expression of Slp4 protein in human platelets. Non-transfected and FLAG–Slp4-transfected HEK293T cells and human platelets were lysed, and expression of Slp4 was analyzed by SDS-PAGE and immunoblotting with a specific anti-Slp4 antibody (rabbit, peptide purified; OpenBiosystems). A band of the expected size was detected in transfected cells and in platelets, but not in non-transfected cells. (C) Effects of Slp4 on dense granule release. Platelets were permeabilized with streptolysin-O, and incubated with 1 μM (final concentration) purified recombinant glutathione-S-transferase (GST) alone, GST–Slp4 fusion protein lacking the Slp-homology domain (SHD), and GST–wild-type Slp4. Granule release was stimulated by addition of Ca2+, and released 5-hydroxytryptamine (5-HT) was measured as an indicator of dense granule release, as described in Materials and methods. Data were normalized to stimulated platelets incubated with GST. The results shown are expressed as means of three independent experiments performed in triplicate. *P < 0.05 and ***P < 0.001 (statistically significant).

Slp4 interacts with Rab8 in platelets

To identify Rab proteins that might bind to Slp4 in platelets, we performed GST pull-down assays with purified GST–Rab fusion proteins. Previous extensive screening experiments had shown that Slp4 can bind to Rab8A, Rab27A, and Rab27B, and weakly to Rab3A [23,24], and human platelets are known to express Rab8 and Rab27 [2]. Therefore GST–Rab8A, GST–Rab27A and GST–Rab27B were generated, and Slp4 binding was analyzed. Endogenous Slp4 in platelet lysates bound strongly to GST–Rab8, whereas almost no binding to Rab27 isoforms could be detected (Fig. 2A, upper panels). As Slp1 had previously been shown to bind to Rab8 [25], we investigated the interaction of Slp1 with Rab8 and Rab27. Slp1 was able to bind to Rab8 and Rab27A and B equally well (Fig. 2A, lower panels). Next, we compared binding of Slp1 and Slp4 to endogenous platelet Rab8. Pull-down assays with GST–Slp4 and GST–Slp1 indicated a stronger interaction of Slp4 and Rab8 than of Slp1 and Rab8 (Fig. 2B). To verify the binding of Slp4 to Rab8 at completely endogenous levels, we performed immunoprecipitation with an Slp4 antibody. Analysis of the precipitates by immunoblotting revealed that Rab8 was present in Slp4 antibody samples but not in IgG controls, indicating that endogenous Slp4 and Rab8 interact in human platelets (Fig. 2C). This binding was further confirmed by coimmunoprecipitation from HEK293T cells transfected with FLAG–Slp4 and HA–Rab8. Rab8 could only be detected in precipitates from cells coexpressing both proteins (Fig. 2D). To verify the role of the SHD of Slp4 in interaction with Rab8, GST–Slp4ΔSHD and GST–Slp4 were used in pull-down experiments from human platelets and transfected HEK293T cells. Subsequent immunoblotting revealed that the Slp4 mutant lacking the SHD did not bind to Rab8 (Fig. 2E,F). From these data, we conclude that Slp4 binds to Rab8 in platelets, and that the SHD of Slp4 is required for Rab8 binding.

Figure 2.

 Synaptotagmin-like protein (Slp)4 interacts with Rab8 in platelets and transfected cells. (A) Pull-down assay of endogenous Slp4 and Slp1. Equal amounts of purified recombinant glutathione-S-transferase (GST), GST–Rab8, GST–Rab27A and GST–Rab27B coupled to glutathione–Sepharose beads were used to pull down endogenous Slp4 and Slp1 from lysates of washed human platelets. The presence of Slp4 and Slp1 in the precipitates was analyzed by SDS-PAGE and immunoblotting. Total platelet lysates were analyzed in parallel, to verify equal levels of Slp4 and Slp1 in the lysate samples. (B) Pull-down of endogenous Rab8 from human platelet lysate with GST–Slp1 and GST–Slp4. Equal amounts of GST as control, GST–Slp1 and GST–Slp4 coupled to glutatione–Sepharose beads were used for precipitation. Bound Rab8 was visualized by immunoblotting with an anti-Rab8 antibody. Total platelet lysates were analyzed to verify equal loading. (C) Coimmunoprecipitation of endogenous Slp4 and Rab8. Washed human platelets were lysed, and a specific antibody against Slp4 (Atlas) was used to immunoprecipitate endogenous Slp4. Non-specific IgG was used as the negative control. The precipitates were analyzed for the presence of bound Rab8 by immunoblotting with anti-Rab8 antibody. (D) Coimmunoprecipitation of transfected FLAG–Slp4 and hemagglutinin (HA)–Rab8. HEK293T cells were transfected with FLAG–Slp4, HA–Rab8, or FLAG–Slp4, together with HA–Rab8. Empty vector was transfected as a negative control. After lysis, Slp4 was precipitated with an anti-FLAG antibody. The precipitates were analyzed for the presence of bound Rab8 by immunoblotting with anti-HA antibody, and the lysates were analyzed for the expression of equal levels of Slp4 and Rab8 (totals). (E) Pull-down of endogenous Rab8 with GST–Slp4 and GST–Slp4ΔSHD from human platelets. Equal amounts of GST, GST–Slp4 and GST–Slp4ΔSHD coupled to glutathione–Sepharose beads were incubated with human platelet lysate. Bound endogenous Rab8 protein was visualized with an anti-Rab8 antibody. (F) Pull-down of transfected HA–Rab8. Lysates of HeLa cells overexpressing HA–Rab8 were subjected to pull-down experiments with equal amounts of GST–Slp4 and GST–Slp4ΔSHD beads. The precipitates were analyzed for the presence of bound Rab8 by immunoblotting with anti-Rab8 antibody. Non-transfected and pcDNA4T0-transfected cells were used as controls. The data shown are representative of independent experiments performed at least three times. SHD, Slp-like protein-homology domain; wt, wild-type.

Subcellular localization of Slp4 and Rab8

To investigate the subcellular localization of Slp4 and Rab8, HeLa cells were transfected with either GFP–Slp4 or mCherry–Rab8. In living cells, GFP–Slp4 showed strong localization to the plasma membrane, with no nuclear or cytoplasmic staining being present (Fig. 3A1). Most mCherry–Rab8 was observed on cytosolic vesicle-like structures, and there was no membrane or nuclear staining (Fig. 3A2). We next determined the effect of coexpression of GFP–Slp4 and mCherry–Rab8 on their localization. As in singly transfected cells, cotransfected Slp4 was present predominantly in the plasma membrane (Fig. 3B1). Interestingly, in cotransfected cells there was an enrichment of mCherry–Rab8 at the plasma membrane (Fig. 3B2). Overlaid images confirmed colocalization of GFP–Slp4 and mCherry–Rab8 at the plasma membrane (Fig. 3B3). Next, we investigated the localization of endogenous Slp4 and Rab8 in human platelets by using specific antibodies. Immunofluorescence staining revealed partial colocalization of Slp4 and Rab8 in unactivated platelets (Fig. 3C1–3). In activated platelets, Slp4 was present predominantly in the center and, to some extent, in the outer plasma membrane. Rab8 was present in the center of activated platelets, where it colocalized with Slp4 (Fig. 3C4–6). Quantitation of the colocalization of Slp4 and Rab8 in platelets and transfected cells indicated approximately 30–40% colocalization of both proteins (Fig. S4). These data support our findings that Slp4 and Rab8 interact in transfected cells and in platelets.

Figure 3.

 Synaptotagmin-like protein (Slp)4 and Rab8 colocalize in transfected cells and in platelets. (A) Localization of green fluorescent protein (GFP)–Slp4 and mCherry–Rab8 in singly transfected cells. HeLa cells were transfected with either GFP–Slp4 (A1) or mCherry–Rab8 (A2) fusion constructs. One day after transfection, living cells were imaged with a confocal microscope. Green indicates Slp4 staining; red indicates Rab8 staining. (B) Colocalization of cotransfected Slp4 and Rab8. HeLa cells were transfected with both GFP–Slp4 and mCherry–Rab8, and the localization of the proteins was analyzed by live-cell microscopy. Colocalization of Slp4 and Rab8 is shown in yellow (B3). (C) Localization of endogenous Slp4 and Rab8 in non-activated and activated platelets. Resting human platelets were fixed in solution (C1–C3) or platelets were allowed to spread on glass (C4–C6), and this was followed by fixation. Platelets were then permeabilized and dual-labeled with primary antibodies directed against Slp4 (green, rabbit; OpenBiosystems) or Rab8 (red, mouse; Tebu-Bio) followed by Alexa Fluor-conjugated secondary antibodies. Overlay of green/red images indicate colocalization of fluorochromes (yellow; C3 and C6). Scale bar: 10 μm. The images shown are representative of independent experiments performed three times.

Rab8 increases dense granule release

We had already shown that Slp4 is able to augment platelet dense granule release (Fig. 1C). To investigate the potential role of Rab8 and of a combination of Slp4 and Rab8 on dense granule release, we performed secretion assays with permeabilized platelets. Both Rab8 alone and Slp4 and Rab8 combined were able to increase Ca2+-induced dense granule release significantly (Fig. 4A). No effects of Rab8 were observed in the absence of Ca2+ stimulation (Fig. S1).

Figure 4.

 GTP–Rab8 interacts with synaptotagmin-like protein (Slp)4 and enhances dense granule release. (A) Effects of Rab8 and Slp4 on dense granule release. Permeabilized platelets were incubated with 1 μM glutathione-S-transferase (GST), 1 μM GST–Rab8, or 0.5 μM GST–Slp4 together with 0.5 μM GST–Rab8. Ca2+-induced dense granule secretion was measured by the use of 5-hydroxytryptamine (5-HT). Data were normalized to the GST control. (B) Secretion assays as described in (A) were performed with 1 μM of a constitutively active GTP-bound mutant of Rab8 (Q67L) or an inactive GDP-bound mutant (T22N). (C) Effects of combinations of GST–Slp4 with active or inactive mutants of Rab8 were analyzed in secretion assays, as described in (A). Data were normalized to the combination of Rab8Q67L with Slp4. The results shown in (A)–(C) are expressed as mean of seven (A), four (B) and seven (C) independent experiments performed in triplicate. *P < 0.05 and **P < 0.01 (statistically significant). (D) Pull-down of endogenous Slp4 from human platelet lysate with GST–Rab8Q67L and GST–RabT22N coupled to glutathione–Sepharose beads. Bound Slp4 was analyzed by SDS-PAGE and immunoblotting with an anti-Rab8 antibody. Total platelet lysates were analyzed to verify equal loading. (E) Coimmunoprecipitation of transfected Slp4 and Rab8 mutants. HEK293T cells were transfected with FLAG–Slp4 together with wild-type and mutants of hemagglutinin (HA)–Rab8. Cells were lysed, Rab8 was precipitated with anti-HA antibodies, and precipitates and lysates were analyzed by immunoblotting. The data shown in (D) and (E) are representative of three independent experiments.

Rab GTPases function as molecular switches that can alternate between GTP-bound active and GDP-bound inactive forms. To investigate whether the Rab8 effects on granule release were dependent on the GTP-bound or GDP-bound form, two mutants were created, a constitutively active GTP-binding mutant, Rab8Q67L, and a constitutively inactive GDP-binding mutant, Rab8T22N, as previously described [26]. On their own, both Rab8 mutants were able to enhance dense granule release similarly to the wild-type form of Rab8 (Fig. 4B). We next tested the ability of Rab8 mutants to enhance Slp4-induced granule release. The GTP–Rab8 mutant (Q67L) had a significantly stronger effect on granule release than the GDP–Rab8 mutant (T22N) (Fig. 4C). To investigate the possible reason for this difference in more detail, we studied the binding of Rab8 mutants to Slp4. The GTP–Rab8 mutant was capable of pulling down endogenous Slp4, whereas the inactive mutant was not (Fig. 4D). These differences in binding were confirmed by coimmunoprecipitation from HEK293T cells transfected with FLAG–Slp4 and either HA–Rab8Q67L or HA–Rab8T22N (Fig. 4E). We conclude that Slp4 interacts preferentially with the GTP-bound form of Rab8, resulting in enhanced dense granule release. Rab8 might also be able to increase granule release independently of Slp4 binding (Fig. 4B).

Discussion

We have identified Slp4 as new platelet protein involved in dense granule release. Slp4 interacts with the GTP-bound form of the small G-protein Rab8. Studies of living cells suggest interaction of Slp4 and Rab8 at the plasma membrane, and both proteins localize to the granule-containing center of activated spread platelets, indicating a possible role for Slp4 and Rab8 in the granule release process. In secretion assays, Slp4 and Rab8 were able to augment dense granule release. Preliminary experiments using platelet factor 4 as a marker of α-granules did not show any significant role for Slp4 in α-granule release (Fig. S5).

The exact mechanism of Slp4 action remains to be determined. Previously, Slp4 has been described as a regulator of the release of insulin granules from pancreatic β-cells [27], amylase granules from parotid gland acinar cells [28], and dense core granules frpm PC12 neuronal cells [23]. In pancreatic β-cells, Slp4 is thought to act as a bridging molecule that enables the docking of secretory granules containing membrane-bound Rab27 to SNAREs such as syntaxin in the plasma membrane [29]. Slp4 was shown to interact with syntaxin 2 in parotid acinar cells [28]. Very recently, both Munc18-2 and syntaxin-11 have been reported to have important functions in platelet secretion [30,31], and possible links with Slp4 need to be investigated. Slp4 might be involved in granule transport by recruitment of the motor protein myosin Va [32]. Slp4 could also facilitate the membrane fusion event, as described for synaptotagmin and other multiple C2 domain proteins [10]. A recently generated mRNA expression database suggests that Slp4 is the most highly expressed protein of the C-type tandem C2 protein family in human platelets, followed by Slp1 as the second most highly expressed protein, whereas classical synaptotagmins are expressed at much lower levels [33]. Slp4 function differs from that of the closely related Slp1 in a number of ways. Slp1 attenuates dense granule release and interacts with Rap1GAP2, a GTPase-activating protein of Rap1 [14], whereas Slp4 enhances granule release and does not bind to Rap1GAP2 (data not shown). Furthermore, Slp1 interacts with Rab27, whereas Slp4 preferentially binds to Rab8 (Fig. 2).

Our experiments suggest that Slp4 requires Rab8 to increase granule release in platelets (Fig. 1C). Granule-attached Rab8 and plasma membrane-bound Slp4 could help to dock granules to the plasma membrane, as has previously been proposed for Rab27 [34]. This concept is supported by our findings that Rab8 alone localizes to vesicular structures, whereas coexpression of Slp4 leads to membrane targeting of Rab8 (Fig. 3B). Slp4 appears to be a classic effector of Rab8, as Slp4 interacts preferentially with GTP–Rab8. Interestingly, Rab8 is probably already in the GTP-bound state in resting platelets, as is indicated by the interaction of Slp4 and Rab8 in resting platelets (Fig. 2C). Thrombin treatment does not affect the Slp4–Rab8 interaction (data not shown), although thrombin has been shown to induce phosphorylation of Rab8. [35]. Also Rab27 is present in its GTP-bound form in resting platelets [36]. Our data further indicate that GTP–Rab8 is more effective in stimulating granule release than GDP–Rab8, although only in combination with Slp4 (Fig. 4C). In addition, Rab8 might enhance granule release independently of its nucleotide binding state (Fig. 4B), which indicates Slp4-independent functions of Rab8. Studies in other cells have shown that Rab8 regulates membrane recycling and the docking and fusion of exocytotic vesicles [37]. Rab8 facilitates the export of cholesterol from macrophages [38], and regulates translocation of glucose transporter type 4-containing vesicles in muscle cells [39]. Rab8 also plays an essential role in ciliogenesis by binding to the exocyst complex of secretory vesicles [40] The exocyst complex has been shown to regulate dense granule release in platelets [41], suggesting possible links between Rab8, the exocyst complex, and Slp4. Further studies using isolated reconstituted liposomes, cellular models and genetically modified platelets are required to define the exact mechanisms through which Slp4 and Rab8 control platelet granule secretion.

Disclosure of Conflict of Interests

This work was supported by a grant from Science Foundation Ireland to A. Smolenski (08/IN.1/B1855). The authors state that they have no conflict of interest.

Ancillary