TI-VAMP/VAMP7 is required for optimal phagocytosis of opsonised particles in macrophages
2004; Springer Nature; Volume: 23; Issue: 21 Linguagem: Inglês
10.1038/sj.emboj.7600427
ISSN1460-2075
AutoresVirginie Braun, Vincent Fraisier, Graça Raposo, Ilse Hurbain, Jean‐Baptiste Sibarita, Philippe Chavrier, Thierry Galli, Florence Niedergang,
Tópico(s)Inhalation and Respiratory Drug Delivery
ResumoArticle7 October 2004free access TI-VAMP/VAMP7 is required for optimal phagocytosis of opsonised particles in macrophages Virginie Braun Virginie Braun Membrane and Cytoskeleton Dynamics group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Vincent Fraisier Vincent Fraisier Digital Imaging Platform, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Graça Raposo Graça Raposo Electron Microscopy Group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Ilse Hurbain Ilse Hurbain Electron Microscopy Group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Jean-Baptiste Sibarita Jean-Baptiste Sibarita Digital Imaging Platform, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Philippe Chavrier Philippe Chavrier Membrane and Cytoskeleton Dynamics group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Thierry Galli Thierry Galli Membrane Traffic and Neuronal Plasticity, INSERM U536, Institut du Fer-à-Moulin, Paris, France Search for more papers by this author Florence Niedergang Corresponding Author Florence Niedergang Membrane and Cytoskeleton Dynamics group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Virginie Braun Virginie Braun Membrane and Cytoskeleton Dynamics group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Vincent Fraisier Vincent Fraisier Digital Imaging Platform, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Graça Raposo Graça Raposo Electron Microscopy Group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Ilse Hurbain Ilse Hurbain Electron Microscopy Group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Jean-Baptiste Sibarita Jean-Baptiste Sibarita Digital Imaging Platform, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Philippe Chavrier Philippe Chavrier Membrane and Cytoskeleton Dynamics group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Thierry Galli Thierry Galli Membrane Traffic and Neuronal Plasticity, INSERM U536, Institut du Fer-à-Moulin, Paris, France Search for more papers by this author Florence Niedergang Corresponding Author Florence Niedergang Membrane and Cytoskeleton Dynamics group, UMR144 CNRS-Institut Curie, Paris, France Search for more papers by this author Author Information Virginie Braun1, Vincent Fraisier2, Graça Raposo3, Ilse Hurbain3, Jean-Baptiste Sibarita2, Philippe Chavrier1, Thierry Galli4 and Florence Niedergang 1 1Membrane and Cytoskeleton Dynamics group, UMR144 CNRS-Institut Curie, Paris, France 2Digital Imaging Platform, UMR144 CNRS-Institut Curie, Paris, France 3Electron Microscopy Group, UMR144 CNRS-Institut Curie, Paris, France 4Membrane Traffic and Neuronal Plasticity, INSERM U536, Institut du Fer-à-Moulin, Paris, France *Corresponding author. UMR144 CNRS-Institut Curie, 26, rue d'Ulm, 75248 Paris cedex 05, France. Tel.: +33 1 42 34 63 67; Fax: +33 1 42 34 63 77; E-mail: [email protected] The EMBO Journal (2004)23:4166-4176https://doi.org/10.1038/sj.emboj.7600427 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Phagocytosis relies on extension of plasmalemmal pseudopods generated by focal actin polymerisation and delivery of membranes from intracellular pools. Here we show that compartments of the late endocytic pathway, bearing the tetanus neurotoxin-insensitive vesicle-associated membrane protein (TI-VAMP/VAMP7), are recruited upon particle binding and undergo exocytosis before phagosome sealing in macrophages during Fc receptor (FcR)-mediated phagocytosis. Expression of the dominant-negative amino-terminal domain of TI-VAMP or depletion of TI-VAMP with small interfering RNAs inhibited phagocytosis mediated by Fc or complement receptors. In addition, inhibition of TI-VAMP activity led to a reduced exocytosis of late endocytic vesicles and this resulted in an early blockade of pseudopod extension, as observed by scanning electron microscopy. Therefore, TI-VAMP defines a new pathway of membrane delivery required for optimal FcR-mediated phagocytosis. Introduction In mammals, phagocytosis is a hallmark of specialised cells such as macrophages, dendritic cells and polymorphonuclear neutrophils, while it can also be accomplished by other cells, such as thyroid and bladder epithelial cells or mesangial cells in the kidney. It is induced by the recognition of the particle by a large variety of cell surface receptors, among which are receptors for components of microorganisms surface or receptors specific for opsonins, like immunoglobulins (Fc receptors, FcRs) or complement proteins (complement receptors, CRs). Entry is followed by the maturation of the phagosome in a phagolysosome that becomes competent for the degradation of the internalised particle and eventually the presentation of derived peptidic antigens (Aderem and Underhill, 1999; Underhill and Ozinsky, 2002). Upon receptor ligation, signalling cascades (Greenberg and Grinstein, 2002) lead to actin polymerisation, which is considered to be the driving force allowing the plasma membrane to be extended around the particle. Concomitant with actin polymerisation, and to accommodate the extension of the plasma membrane, focalised exocytosis of membrane from internal pools takes place at the site of the phagocytic cup formation (Holevinsky and Nelson, 1998; Booth et al, 2001). The nature and origin of the membranes contributing to phagocytosis is still an open question. Recycling vesicles containing the transferrin receptor (TfR) or cellubrevin/vesicle-associated membrane protein 3 (VAMP3) are exocytosed at the site of phagocytosis (Bajno et al, 2000; Niedergang et al, 2003). More recently, the endoplasmic reticulum (ER) has been observed in association with forming phagosomes in J774.1 macrophages (Gagnon et al, 2002). In addition, calreticulin and calnexin, two resident proteins of the ER, are required for phagocytosis in Dictyostelium discoideum (Muller-Taubenberger et al, 2001). More specifically, the role of vesicle-associated, soluble N-ethylmaleimide-sensitive factor attachment protein receptors (v-SNAREs) in focal exocytosis during phagocytosis was revealed in studies where phagocytosis was inhibited by tetanus neurotoxin (TeNT), which cleaves and inactivates synaptobrevin/VAMP2 and cellubrevin/VAMP3 (Hackam et al, 1998). However, macrophages of mice lacking VAMP3 were still able to phagocytose opsonised particles or latex beads, although a delay was observed in the uptake of zymosan (Allen et al, 2002). These observations suggested that other SNAREs, presumably TeNT insensitive, could be implicated in phagocytosis. Here we show that the TeNT-insensitive v-SNARE protein TI-VAMP/VAMP7 plays a crucial role in the onset of phagocytosis in macrophages. Results TI-VAMP is associated to a late endosomal compartment in macrophages In order to gain insight into the function of TI-VAMP in macrophages, we first analysed its subcellular localisation. Immunofluorescence analysis in RAW264.7 macrophages showed a high degree of colocalisation of endogenous TI-VAMP with Lamp1-positive late endosomes/lysosomes, and often the TI-VAMP/VAMP7 vesicles were found on the periphery of the Lamp1 compartment (Figure 1A, panels e–h). By contrast, TI-VAMP is only poorly colocalised with the TfR, a marker of early and recycling endosomes (Figure 1A). The same observations were made in primary bone marrow-derived macrophages and in RAW264.7 cells transiently expressing a GFP-tagged TI-VAMP (data not shown). In addition, immunogold labelling of ultrathin cryosections of RAW264.7 macrophages with an antibody against TI-VAMP detected multivesicular compartments (Figure 4). All these results indicate that TI-VAMP/VAMP7 is associated with a late endocytic compartment in macrophages. Figure 1.Endogenous TI-VAMP/VAMP7, colocalised with the lysosomal marker Lamp1, is recruited early and accumulates at the site of phagocytosis. (A) RAW264.7 cells were fixed, permeabilised and stained with anti-TI-VAMP and Cy3-anti-mouse IgG (a, c, e, g) and either anti-TfR (b, d) or anti-Lamp1 (f, h) followed by Cy2-anti-rat IgG. (c, g) Combined images; (d, h) insets in (c, g). Cells were analysed by wide-field fluorescence microscopy with deconvolution. Medial optical sections are shown. Bar, 5 μm. An average of 3.7±0.5 (n=10) TI-VAMP-positive and TfR-positive structures were colocalised per cell representing 8.7±1.1% of total TfR-positive structures while an average of 12±1.6 (n=10) TI-VAMP-positive structures were found to overlap with Lamp1-positive structures per cell which corresponded to 41±5.8% of total Lamp1-positive structures. (B) RAW264.7 cells were incubated for 10 min at 37°C with IgG-SRBCs, and then fixed and stained with Cy3-anti-rabbit IgG (right panel). This staining reveals particles that are still accessible to antibodies and therefore in nonclosed phagosomes, whereas internal particles were observed by phase contrast. The cells were then permeabilised, labelled with anti-TI-VAMP followed by Cy2-anti-mouse IgG (left panel) and analysed by confocal microscopy. One optical section is shown. The arrowheads point to external particles positive for TI-VAMP. Bar, 5 μm. (C) GFP-TI-VAMP is recruited early to phagosomes. RAW264.7 cells transiently transfected to express GFP-TI-VAMP were incubated for 10 min at 37°C with IgG-SRBCs, then placed on ice and, without fixation, stained with Cy3-anti-rabbit IgG to detect external particles. Then, cells were fixed, permeabilised and polymerised actin was labelled with phalloïdine-Alexa350. The number of accumulations of polymerised actin and GFP-TI-VAMP were scored for 50 GFP-TI-VAMP-expressing cells and expressed as an index of accumulation per cell. Then, the index obtained for TI-VAMP recruitment was expressed as a percentage of the index of actin cups. Data are the mean±s.e.m. of three independent experiments. Download figure Download PowerPoint TI-VAMP is recruited to phagosomes We next analysed TI-VAMP/VAMP7 dynamics during phagocytosis in macrophages allowed to phagocytose IgG-opsonised sheep red blood cells (IgG-SRBCs) at 37°C. Endogenous TI-VAMP, as well as the GFP fusion chimaera, accumulated around particles internalised for 60 min, as described for late endosomal/lysosomal markers during phagosome maturation (data not shown; Desjardins et al, 1994; Oh and Swanson, 1996; Bajno et al, 2000; Henry et al, 2004). This recruitment was specific, as endobrevin/VAMP8, which labels vesicles along the endocytic pathway (Antonin et al, 2000; Mullock et al, 2000 and our unpublished observations), was not accumulated around SRBCs in the same conditions (data not shown). When we analysed early time points, such as 10 min, we noticed that TI-VAMP/VAMP7 was also present around the particles that were still at the beginning of the phagocytic process and engaged in a nonclosed phagosome, defined by their accessibility to fluorescent antibodies directed against the opsonised SRBCs (arrowhead in Figure 1B at 10 min, pointing to external particles accessible to extracellular labelling). To make sure that only external SRBCs were stained, we labelled the cells with the antibodies against SRBCs prior to fixation and staining with phalloidin. We then scored the presence of TI-VAMP recruitment and polymerised actin. Actin cups form in the very first minutes of phagocytosis and then depolymerise (Swanson et al, 1999; Coppolino et al, 2002; Araki et al, 2003; Henry et al, 2004). We observed that TI-VAMP was present in 40% of phagocytic cups defined by F-actin (Figure 1C). The accumulation of TI-VAMP in actin cups was also observed in primary macrophages after 5 min of phagocytosis (data not shown). Therefore, TI-VAMP/VAMP7 recruitment can be detected as early as filamentous actin during the process of phagosome formation. To further analyse the kinetics of recruitment of TI-VAMP during phagocytosis, the internalisation of IgG-SRBCs was followed by video microscopy in RAW264.7 cells expressing GFP-TI-VAMP (Figure 2 and Movie 1 in Supplementary data). TI-VAMP-positive vesicles migrated towards the site of phagocytosis during the first minutes of phagocytosis and thereafter, and were progressively incorporated around the particles as the phagosome was formed. Figure 2.TI-VAMP/VAMP7-positive compartments are recruited early during phagocytosis. RAW264.7 cells transfected to express GFP-TI-VAMP were put into contact with IgG-SRBCs and recorded at 37°C using 4D deconvolution video microscopy. A time-stack was built with maximum intensity projections of deconvolved image stacks (see Movie 1, Supplementary data). Selected images are shown (top left corner, time in s). Download figure Download PowerPoint Together, these results indicate that TI-VAMP/VAMP7 translocates to the phagosomal membrane early during phagosome formation in macrophages. TI-VAMP-positive vesicles are exocytosed during phagosome formation The early recruitment of TI-VAMP at the onset of phagocytosis prompted us to assess whether TI-VAMP/Lamp1-positive vesicles were exocytosed before phagosome sealing. Exocytosis was analysed by electron microscopy, after loading of the TI-VAMP-positive compartments with a fluid-phase marker, horseradish peroxidase (HRP), and detection of the enzymatic activity by conventional diaminobenzidine tetrahydrochloride (DAB) cytochemistry. Fluorescence microscopy confirmed that internalised HRP accumulated in the Lamp1-positive compartment after a 30 min pulse and 1 h chase (data not shown). Electron microscopy showed that DAB was present in multivesicular late endosomes (Figure 3Aa). After 10 min of contact between RAW264.7 cells and IgG-SRBCs, we observed DAB-positive compartments recruited near the cell surface in 40% of the cells (n=22 cells and 342 structures counted in total) (arrowheads in Figure 3Ab), whereas only 25% of the DAB-positive structures were close to the cell surface in control nonphagocytosing cells (n=41 cells and 591 structures counted in total) (Figure 3Aa). In addition, DAB-positive membranes (thin arrows in Figure 3Ac and d) reminiscent of an exocytic event could be found in the extracellular space under or very close to bound particles. Figure 3.TI-VAMP/Lamp1-positive compartments are exocytosed early during phagocytosis. (A) RAW264.7 cells were allowed to internalise HRP (50 mg/ml) for 30 min at 37°C. After a 1 h chase, phagocytosis was performed for 10 min. The cells were then fixed and processed for DAB cytochemistry and conventional electron microscopy. The arrowheads point to HRP-loaded vesicles recruited to sites of red blood cell attachment and the arrows to exocytosed DAB-positive membranes. The white stars label external SRBCs. Bar, 2 μm. (B) RAW264.7 cells were incubated with IgG-SRBCs for 10 min at 37°C, then placed on ice and, without fixation, stained with anti-Lamp1 antibodies followed by Cy3-anti-rat IgG. External red blood cells were detected with Cy2-anti-rabbit IgG antibodies. Finally, the cells were fixed and analysed by confocal microscopy. One medial optical section is shown. Bar, 5 μm. (C) RAW264.7 cells were treated as in (B), except that secondary antibodies were RPE-coupled. Live cells were then analysed by flow cytometry (bold line). As a control, the cells were incubated with IgG-SRBCs on ice before labelling (dotted histogram). Download figure Download PowerPoint In order to quantify exocytosis of the TI-VAMP compartment during phagocytosis, we monitored the phenomenon by fluorescence microscopy and flow cytometry. We set up an assay to follow the exposure to the cell surface of a luminal epitope of Lamp1. For this purpose, RAW264.7 macrophages were incubated with IgG-SRBCs for 10 min at 37°C, then placed on ice and labelled to detect only surface-exposed Lamp1. The microscopic analysis revealed that Lamp1 was detected under opsonised particles bound to the macrophages (Figure 3B). To make sure that this staining was not due to clustering of Lamp1 molecules already present at the plasma membrane, cells were analysed by flow cytometry (Figure 3C). Compared to cells incubated with IgG-SRBCs at 4°C, a population of phagocytosing macrophages (around 15% of the total cell population) showed an increase in surface Lamp1 levels, further demonstrating exocytosis of Lamp1-positive membranes from internal stores. We then sought to detect this exocytotic phenomenon by direct opening of the TI-VAMP/VAMP7 compartment at the site of attachment of particles after immunogold labelling of ultrathin cryosections (Figure 4A and B). Antibodies directed against TI-VAMP as well as Lamp1 decorated multivesicular compartments (asterisks in Figure 4A and B). Direct contacts between the Lamp1-positive vesicles and the plasma membrane were observed at sites of particle attachment (see arrows in Figure 4A). Importantly, similar observations were made for TI-VAMP-positive compartments (arrows in Figure 4B). Taken together, all these results demonstrate that exocytosis of TI-VAMP/Lamp1-positive compartments occurs during phagosome formation. TI-VAMP fusion activity is required for optimal phagocytosis via FcR and CR TI-VAMP belongs to the Longin family of v-SNAREs characterised by the presence of a long amino-terminal autoinhibitory domain (Martinez-Arca et al, 2003), called the Longin domain (Filippini et al, 2001). The Longin domain negatively regulates the ability of TI-VAMP to participate in SNARE complexes (Martinez-Arca et al, 2003) and has a dominant-negative effect in neuronal cells, where its overexpression leads to an inhibition of neurite outgrowth (Martinez-Arca et al, 2000). In RAW264.7 cells, while full-length GFP-TI-VAMP had no effect on phagocytosis of IgG-SRBCs, expression of GFP-Longin led to a diminution of phagocytosis efficiency that was comparable to the inhibition observed with dominant-negative Cdc42, known to inhibit phagocytosis (Figure 5A) (Caron and Hall, 1998; Massol et al, 1998). In contrast, the association of IgG-opsonised particles was not strongly affected. The same results were obtained after 10 min (data not shown) or 60 min of phagocytosis (Figure 5). These results show that the fusion activity of TI-VAMP is important for FcR-mediated phagocytosis. Figure 4.Exocytosis of Lamp1- and TI-VAMP-positive compartments. (A) RAW264.7 macrophages were allowed to phagocytose IgG-SRBCs for 10 min at 37°C and then fixed and processed for ultrathin cryosectioning. Cryosections were immunogold labelled with anti-Lamp1 antibody followed by Protein A-Gold. Bar, 1 μm. Staining of SRBCs was due to a nonspecific reaction. The asterisks indicate Lamp1-positive compartments. The arrows show Lamp1 that was found on membrane ruffles under particles, facing the extracellular medium. (B) Cells were treated as described in (A). Cryosections were immunogold labelled with anti-TI-VAMP antibody followed by Protein A-Gold. Bar, 1 μm. A TI-VAMP-positive compartment is indicated by an asterisk and the arrows show TI-VAMP facing the extracellular medium. Download figure Download PowerPoint Figure 5.The amino-terminal domain of TI-VAMP inhibits FcR and CR-mediated phagocytosis. (A) RAW264.7 macrophages transiently expressing GFP-TI-VAMP, GFP-Longin or GFP-Cdc42N17 were incubated with IgG-SRBCs for 60 min at 37°C. The cells were then fixed and external SRBCs stained with Cy3-anti-rabbit IgG antibodies. The efficiencies of association (left) and phagocytosis (right) were calculated on 50 transfected and 50 control cells. The mean±s.e.m. of nine independent experiments is plotted. Control, GFP-negative cells. (B) RAW264.7 macrophages transiently expressing GFP-TI-VAMP (grey bars) or GFP-Longin (black bars) were allowed to phagocytose zymosan or C3bi-SRBCs for 60 min at 37°C. The samples were processed as in (A). The mean±s.e.m. of three independent experiments is plotted. Control, GFP-negative cells. Download figure Download PowerPoint We next analysed whether the role of TI-VAMP could be generalised to other types of phagocytosis. As shown in Figure 5B, CR-mediated phagocytosis was inhibited in cells expressing the dominant-negative Longin domain, whereas the uptake of zymosan was not affected. This result indicates that TI-VAMP activity is required in both type I FcR-mediated and type II CR-mediated phagocytosis that rely on different signalling cascades (Caron and Hall, 1998). By contrast, it is not implicated in the entry of zymosan, which involves various surface receptors, including mannose receptors, CR3 and the recently described Dectin1 (Brown and Gordon, 2001). TI-VAMP siRNA depletion leads to a defect in FcR-mediated phagocytosis In order to impair the function of TI-VAMP by another approach, we designed siRNA targeted to the sequence of murine TI-VAMP and analysed the consequences of RNAi depletion of this protein on FcR-mediated phagocytosis. With two different TI-VAMP siRNA, we observed a significant decrease in total level of the protein after 24 h of treatment in RAW264.7 cells (Figure 6B). By immunofluorescence, the depletion appeared to affect clusters of cells, while in some cells the fluorescence was still bright (Figure 6A). Therefore, phagocytosis of IgG-opsonised SRBCs was performed and the efficiency of phagocytosis was scored in cells exhibiting a clear decrease of TI-VAMP fluorescent signal and compared to control cells (treated either with a control siRNA or with oligofectamine alone) (Figure 6). The depletion of TI-VAMP led to an inhibition of phagocytosis comparable to that observed in cells expressing the inhibitory Longin domain (Figure 6C). Therefore, blocking or depleting TI-VAMP inhibited FcR-mediated phagocytosis by approximately 50% and this further shows that TI-VAMP activity is crucial for receptor-mediated phagocytosis. Figure 6.TI-VAMP siRNA treatment inhibits phagocytosis. RAW264.7 macrophages were transfected with siRNA directed against TI-VAMP (RNA.1 or RNA.2). After 24 and 48 h of treatment, cells were analysed by immunofluorescence (A), Western blot (B) or allowed to phagocytose IgG-SRBCs (C). (A) Cells were fixed, permeabilised, then labelled with anti-TI-VAMP followed by Cy3-anti-mouse IgG and analysed by wild-field fluorescence microscopy. Control cells were mock transfected. Bar, 10 μm. (B) Lysates were prepared and Western blotting was performed with anti-TI-VAMP (lower panel) and, after stripping of the membrane, with anti-clathrin HC (upper panel). Control cells were mock-transfected cells. (C) Transfected cells were incubated for 60 min at 37°C with IgG-SRBCs, then fixed and stained with Cy2-anti-rabbit IgG. The cells were then permeabilised and labelled with anti-TI-VAMP followed by Cy3-anti-mouse IgG. The efficiencies of association (left) and phagocytosis (right) were calculated on 50 cells that presented a decrease of TI-VAMP signal and 50 control cells (mock-transfected cells or cells transfected with siRNA GFP). The mean±s.e.m. of three independent experiments is plotted. Download figure Download PowerPoint Inhibition of both TeNT-sensitive VAMPs and TI-VAMP The implication of TeNT-sensitive SNAREs in phagocytosis has been reported earlier (Hackam et al, 1998). Therefore, in order to further analyse both the TeNT-sensitive-VAMP- and TI-VAMP-dependent pathways during phagocytosis, we sought to inhibit both sets of proteins simultaneously. For this purpose, macrophages were transfected to express TeNT, which cleaves cellubrevin/VAMP3 (McMahon et al, 1993 and our unpublished results), together with the Longin domain of TI-VAMP. None of the treatment led to a strong inhibition of particle association to the cells (Figure 7A). As reported earlier (Hackam et al, 1998), expression of the TeNT inhibited phagocytosis by 50%, whereas an inactivated toxin did not show any effect (Figure 7B). The simultaneous expression of TeNT and GFP-Longin did not lead to a higher inhibition. This was not due to an inability to further block phagocytosis in RAW264.7 cells, because in cells treated with cytochalasin D, phagocytosis was completely abolished (Figure 7B). Therefore, in contrast to inhibiting actin polymerisation, blocking endomembrane delivery does not completely prevent phagocytosis, and this most probably reflects the fact that the plasma membrane might account for particle engulfment without the need for internal membrane recruitment. These experiments could also suggest that TeNT-sensitive VAMPs, like VAMP3, and TI-VAMP function in a coordinated manner. To address this possibility, we followed simultaneously cellubrevin/VAMP3 and TI-VAMP/VAMP7 during the first steps of IgG-SRBC internalisation in living cells (Figure 7C and Movie 2 in Supplementary data). The dynamics of CFP-cellubrevin and YFP-TI-VAMP indicated that both proteins relocalised to the same phagosome (see arrowheads in Figure 7C, indicating the presence of the probes in extending pseudopods). It appeared that the recruitment of TI-VAMP immediately followed the accumulation of cellubrevin/VAMP3 (arrowheads in Figure 7C) and that the kinetics was slightly shifted in some cases (see arrows for cellubrevin/VAMP3 in Figure 7C). In a parallel set of experiments, Cy5-coupled secondary antibodies were used to detect external particles on cells prior to fixation. At 5, 10 and 20 min, VAMP3 is accumulated on 51, 45 and 51% of the phagosomes still accessible to Cy5-anti-rabbit antibodies, respectively, and TI-VAMP on 13, 20 and 20% of these phagosomes (data not shown). Therefore, there was no strong variation with time of the results for nascent phagosomes. It is worth noting, however, that the presence of a pool of VAMP3 already at the plasma membrane accounts for the signal detected very early around the phagosomes. Interestingly, there is a certain degree of colocalisation of CFP-cellubrevin and YFP-TI-VAMP in phagocytosing macrophages, suggesting that the two pathways of focal exocytosis might cooperate in time and space during phagocytosis. Figure 7.VAMP3- and TI-VAMP-mediated membrane delivery. (A) RAW264.7 macrophages transiently expressing GFP-TI-VAMP, GFP-Longin, TeNTE234Q, TeNT or both GFP-Longin and TeNT, or nontransfected cells pretreated for 30 min at 37°C with cytochalasin D (Cyto D) were processed for phagocytosis as described in Figure 5. Controls for transfection were negative cells and for Cytochalasin D, DMSO-treated cells. The mean efficiency of association±s.e.m. of four independent transfection experiments is plotted. (B) RAW264.7 cells were treated as in (A). The mean efficiency of phagocytosis±s.e.m. of four independent transfection experiments is represented. (C) RAW264.7 cells cotransfected to express YFP-TI-VAMP and CFP-VAMP3 were put into contact with IgG-SRBCs and followed at 37°C under a Zeiss LSM510 Meta confocal microscope. Images were recorded simultaneously in the two channels with a line average of eight every 3 s (see Movie 2, Supplementary data). Selected images are shown (top left corner, time in s). The arrowheads point to accumulation of YFP-TI-VAMP and the arrows to CFP-VAMP3. Download figure Download PowerPoint TI-VAMP controls exocytosis and membrane extension during phagocytosis In order to further dissect the role of TI-VAMP in focal exocytosis, we analysed the effect of the GFP-Longin inhibitory construct on the early exposure of Lamp1 on the surface of phagocytosing cells (Figure 8A). The surface level of Lamp1 was measured by flow cytometry after gating on live, GFP-positive cells, expressing either the full-length GFP-TI-VAMP protein or the GFP-Longin domain and placed into contact with IgG-SRBCs for 10 min at 37°C. Whereas 18% of the phagocytosing cells expressing GFP-TI-VAMP showed an increase in the surface expression of Lamp1 (see marker M1 in histogram in bold, Figure 8A), only 6% of the cells expressing GFP-Longin exhibited an increased surface level of Lamp1 (marker M1, filled histogram in Figure 8A). Therefore, exocytosis of Lamp1 is inhibited when the fusion activity of TI-VAMP is impaired. Figure 8.TI-VAMP controls exocytosis and membrane extension during phagocytosis. (A) RAW264.7 cells transfected to express GFP-TI-VAMP or GFP-Longin were incubated with IgG-SRBCs for 10 min at 37°C, then placed on ice and, without fixation, stained with anti-Lamp1 antibodies followed by RPE-anti-rat IgG. At least 5000 live, GFP-positive cells were gated and analysed by flow cytometry. Control cells incubated with secondary antibodies alone (dotted histogram), GFP-Longin-expressing cells (filled histogram), GFP-TI-VAMP-expressing cells (bold line) are shown. M1: Lamp1-positive cells, 18% for GFP-TI-VAMP- and 6% for GFP-Longin-expressing cells. (B) RAW264.7 cells expressing GFP-TI-VAMP (a) or GFP-Longin (b–d) were grown on CeLLocate coverslips, incubated with IgG-SRBCs for 60 min at 37°C, fixed, located on coverslips by fluorescence microscopy and then imaged by SEM. Red blood cells were artificially coloured under Adobe Photoshop 7.0. Bar, 1 μm. Download figure Download PowerPoint To gain further insight into the stage of FcR-mediated phagocytosis that is affected by the inhibition of TI-VAMP, RAW264.7 cells were transfected to express the GFP-Longin construct and analysed by scanning electron microscopy (SEM) after identification by immunofluorescence of the transfected positive c
Referência(s)