- Top of page
Background: Granulocyte exocytosis is proposed to be critically dependent on the interaction of soluble N-ethylmaleimide-sensitive factor attachment protein (SNAP) receptors (SNAREs) located on granules/vesicles (v-SNAREs) and plasma membrane (t-SNAREs). Previous studies indicated that the v-SNARE, vesicle-associated membrane protein (VAMP)-2, as well as t-SNAREs (SNAP-23, syntaxin-4 and -6) are implicated in exocytosis from human granulocytes. Vesicle-associated membrane proteins-7 and -8 have been implicated in endosome/lysosome trafficking, however, their role in granulocyte exocytosis remains obscure.
Objective: We sought to investigate the expression and functional role of SNARE isoforms in the secretion of different granule-derived mediators in human eosinophils and neutrophils.
Methods: The expression of SNAREs was determined by subcellular fractionation and flow cytometry. SNARE-specific antibodies were examined for their ability to impair mediator release from permeabilized eosinophils and neutrophils.
Results: Vesicle-associated membrane proteins-7 and -8 were localized to granule and membrane-enriched fractions in eosinophils and neutrophils, whereas syntaxin-6 was not detectable. In permeabilized cells, anti-VAMP-7, but not anti-VAMP-8, antibody impaired the secretion of all mediators examined (in eosinophils, eosinophil peroxidase and eosinophil-derived neurotoxin; in neutrophils, myeloperoxidase, lactoferrin and matrix metalloprotease-9) in a dose-dependent manner. In contrast, anti-VAMP-2 modestly and selectively impaired secretion from small granules and vesicles. Syntaxin-4, but not syntaxin-6, was found to interact with SNAP-23 and was partially involved in mediator secretion from multiple compartments.
Conclusion: Our observations indicate for the first time a critical role for VAMP-7 in both eosinophil and neutrophil mediator release.
The release of preformed granule-derived mediators from eosinophils and neutrophils is critical in the manifestation of inflammatory responses in airway diseases such as asthma and chronic obstructive pulmonary disease (1, 2). Eosinophils exhibit a rapid mobilization of secretory vesicles, termed piecemeal degranulation, which may or may not coincide with large crystalloid granule exocytosis depending on culture conditions or activating agents (3, 4). Neutrophils have four distinct granules: azurophilic granules, small (secondary) granules, tertiary granules and secretory vesicles. Similar to eosinophils, mediator release from neutrophils is characterized by a rapid and sequential release of secretory vesicles, secondary granules and/or tertiary granules, followed by the slower release of azurophilic granules (5).
Exocytosis of membrane-bound vesicles/granules has been shown to be critically dependent on the interaction of one vesicle (v-) soluble N-ethylmaleimide-sensitive factor attachment protein (SNAP) receptors (SNARE) isoform, vesicle-associated membrane protein (VAMP), with two isoforms of target (t-) SNAREs: a syntaxin isoform and either synaptosome-associated-protein of 25 or 23 kDa (SNAP-25 or SNAP-23) (6, 7). It has been hypothesized that distinct SNARE isoforms may, in part, determine the specificity of vesicle/granule trafficking. In accordance with this paradigm, it may be anticipated that different granule populations express nonoverlapping sets of VAMPs. However, studies have supported the promiscous interactions of SNAREs in vitro (8) and that single SNAREs are capable of participating in multiple trafficking steps in vivo (7).
We previously showed VAMP-2 was localized to secretory vesicles, but not on crystalloid granules, in human eosinophils and implicated in IFN-γ-induced exocytosis (9). In separate studies, VAMP-2 was implicated in exocytosis of neutrophil secretory vesicles, secondary and tertiary granules (10). Both human eosinophils (11) and neutrophils (12) express the t-SNAREs, SNAP-23 and syntaxin-4, which are localized, although not exclusively, to plasma membranes. In addition, syntaxin-6 has been implicated in exocytosis of neutrophil mediators (12). Vesicle-associated membrane protein(s) localized to eosinophil crystalloid granules and neutrophil azurophilic granules have not been identified. Recent investigations indicated that the isoforms, VAMP-7 and -8, are involved in endocytic and/or exocytic trafficking (13–16). In this study, we sought to determine whether there is a differential role of v-SNAREs (VAMP-2, -7 and -8) in the release of compartment-specific stored mediators. We also examined the functional role of syntaxin-4 and -6, previously shown to be involved in human neutrophils (10, 12), in the release of multiple granule-derived mediators. Our data demonstrate, for the first time, a critical role for VAMP-7 in the secretion of stored mediators from multiple granule populations in human granulocytes.
- Top of page
The SNARE isoforms localized to eosinophil crystalloid granules and neutrophil azurophilic granules have not been previously characterized. We have shown the v-SNAREs, VAMP-7 and -8, but not VAMP-2, are localized to these granule compartments. Antibody-mediated inhibition of VAMP-7, but not VAMP-8, impaired the release of crystalloid granule mediators, EPO and EDN, and the azurophilic granule-specific mediator, MPO, in a dose-dependent manner. Our observations are consistent with previous studies that have shown VAMP-7 is localized to CD63+ lysosomes, which are similar to CD63+ crystalloid and azurophilic granules (14, 27, 28). In human neutrophils, we also found VAMP-7 mAb dose-dependently impaired the release of mediators from secondary (LF) and tertiary granules (MMP-9). These findings indicate that VAMP-7 and VAMP-8 expression was not restricted to distinct granule populations, but were also detected in low-density membrane-enriched fractions. Other studies have similarly supported that these v-SNAREs are not strictly compartment-specific. For example, VAMP-7 has been implicated in exocytosis of numerous granule populations in secretory cells (17, 25, 27) and VAMP-8 in either endosome or granule trafficking (15, 16). Our data indicate that exocytosis of separate granule populations cannot be attributed to differences in SNARE isoform expression alone. This finding is consistent with previous studies that have shown SNAREs can interact with multiple partners and may perform more than one trafficking role in vivo (29).
In contrast to VAMP-7 mAb, we found VAMP-8-specific antibodies did not demonstrate any inhibitory effect on the release of any of the examined eosinophil or neutrophil mediators. Both antibodies used in the current study recognize the cytoplasmic NH2-terminal region required for SNARE–SNARE interactions, and specifically label native proteins (13, 18) which was confirmed in this study by flow cytometry. Our observations suggest that VAMP-8 is not critical for mediator release from human granulocytes. However, our data do not conclusively rule out a functional role for VAMP-8. To date, the only ascribed function of SNAREs is to catalyse membrane fusion. It is possible that VAMP-8 may play a role in specific patterns of mediator release, such as granule–granule fusion during compound exocytosis. Alternatively, the expression of VAMP-8 on the surface of granules could be indicative of a role in the fusion of granule with either endosomes or small vesicles.
In both eosinophils and neutrophils, secretory vesicles are rapidly mobilized and exocytosed following agonist-induced cell activation as a first line of defence (4, 5). In eosinophils, secretory vesicles have also been postulated to account for the release of mediators from cytoplasmic crystalloid granules, which exhibit evidence of content loss by microscopic analysis (termed piecemeal degranulation) (7, 30). It has been postulated that membrane budding of crystalloid granules may be a potential mechanism for piecemeal degranulation. Alternatively, secretory vesicles could comprise a separate population that function as transport vesicles for the release of crystalloid granule cargo (3, 4). In this study and in a previous report (9), we have shown that VAMP-2 is localized to secretory vesicles, but is not detected on large crystalloid granules. This observation suggests that at least a proportion of the secretory vesicle pool comprises a separate population that is not derived from crystalloid granules.
In human eosinophils, VAMP-2 vesicles were co-localized with RANTES and implicated in vesicle-mediated release of this cytokine following treatment with interferon-γ (9). In the present study, we have provided evidence that VAMP-2 is functionally implicated in exocytosis of EPO, but not EDN (2). Both of these mediators are predominantly localized to crystalloid granules, but EPO has also been reported in secretory vesicles by electron microscopy (3). However, the inhibitory effect of VAMP-2 mAb was not as pronounced as that of VAMP-7 mAb for EPO. This suggests that, under our experimental conditions, the bulk of secreted EPO and EDN released by crystalloid granule fusion is dependent on VAMP-7, but not on VAMP-2. A study by Hoffmann et al. (31) similarly demonstrated that VAMP-2 was expressed in secretory vesicles, but not crystalloid granules, by subcellular fractionation of human eosinophils. They reported that tetanus toxin (TeNT) impaired the release of ECP from SLO-permeabilized eosinophils, although this was not compared with an inactivated toxin control. We attempted to confirm our findings with VAMP-2 mAb by examining BoNT-B-LC (which cleaves VAMP-2 at the same site as TeNT). Although reagent incompatabilty prevented our analysis of EPO, we found no inhibitory effect of BoNT-B-LC on EDN release. Unlike the assays for MPO and EPO, the assay for EDN was not sensitive to BoNT-B-LC-related reagents. Our observations suggest that VAMP-2 is predominantly involved in exocytosis of a preformed small secretory vesicle pool that is distinct from crystalloid granules.
Previous studies have provided evidence that VAMP-2 is localized to multiple compartments in neutrophils, including small vesicles, secondary granules and tertiary granules (10, 32). Mollinedo et al. (10) showed that VAMP-2 mAb impaired the surface upregulation of the common secondary/tertiary granule marker, CD66b, but not the azurophilic granule marker, CD63, in electropermeabilized neutrophils. Consistent with this study, we observed that the same VAMP-2 mAb had a negligible effect on release of the specific azurophilic granule mediator, MPO. However, we also found that VAMP-2 did not significantly impair exocytosis of LF or MMP-9 from neutrophils. This discrepancy may be due to differences in experimental methods because we have evaluated secreted mediators from secondary and tertiary granules, rather than upregulation of surface markers. Our observation that VAMP-2 is detected only in concentrated high-speed membrane pellets from human neutrophils is consistent with previous study by Brumell et al. (32) that suggested a significant amount of this v-SNARE was localized to small secretory vesicles that pellet only at high-centrifugation speeds.
The t-SNAREs, SNAP-23 and syntaxin-4 are each capable of interacting with VAMP-2, -7 and -8 in vitro (8, 13, 33) which suggests that they have the potential to support exocytosis of multiple granule compartments. SNAP-23 and syntaxin-4 are localized to the plasma membrane in eosinophils (11), neutrophils (10, 12), but are also detected on other intracellular organelles. We detected a t-SNARE complex of SNAP-23 and syntaxin-4 in membrane fractions of resting and Ca2+ and GTPγS-activated granulocytes, but we were unable to find immunoreactivity for v-SNAREs (VAMP-2, -7 and -8) in syntaxin-4 immunoprecipitates (with or without N-ethylmaleimide). It is possible that rapid recycling of SNARE complexes limited the ability to detect v/t-SNARE pairings by immunoprecipitation. For example, in neurons it has been shown that v/t-SNARE complexes are rapidly disassembled and VAMPs recycled within seconds following mediator release (34). Alternatively, it is also possible that v/t-SNARE interactions may have been obscured by additional regulatory molecules that are recruited to the granule-docking site for assembly of SNARE complexes and granule-membrane fusion. We anticipate that such epitope obstruction may have limited the inhibitory effect of SNAP-23 and syntaxin-4 antibodies on mediator release. The inhibitory effect was moderate for both the t-SNARE antibodies, whereas VAMP-7 and VAMP-2 mAbs (for EPO) were more effective inhibitors of secretion in permeabilized cells. In contrast to a previous report, which indicated a role for syntaxin-6 in neutrophil secretion (12), we were unable to detect this t-SNARE in either subcellular fractions or high-speed membrane pellets from eosinophils and neutrophils. We also confirmed that doses up to 20 μg/ml of syntaxin-6 mAb did not exert any inhibitory effect on the secretion of granule mediators from either cell type.
In summary, we have identified VAMP-7 as a predominant SNARE involved in granule-derived mediator release from eosinophils and neutrophils. However, this conclusion is limited by the model of secretion used in this study. In support of our findings, other studies suggest VAMP-7 and/or VAMP-8 are implicated in secretion from haematopoietic cell types. Both VAMP-7 and VAMP-8 were previously localized to granules of rat basophilic cells (RBL-2H3), although a functional role for either isoform in exocytosis has not yet been clearly demonstrated (13). In human platelets, the addition of a recombinant cytoplasmic domain of VAMP-8 was shown to selectively impair the release of dense-granule mediators from permeabilized cells (16). Finally, VAMP-7 was recently implicated in exocytosis of late endocytic vesicles from RAW.264 macrophages (35). Taken together, these findings suggest that a conserved family of fusion proteins may regulate the secretion of inflammatory mediators from different cell types. Further studies are required to determine the signalling molecules involved in the recruitment of granule populations to the plasma membrane and their role in SNARE assembly. These studies are necessary in the pursuit of candidate proteins that may prove to be selective targets for the modulation of secretory function of inflammatory cells.