The actin cytoskeleton regulates exocytosis in all secretory cells. In neutrophils, Rac2 GTPase has been shown to control primary (azurophilic) granule exocytosis. In this report, we propose that Rac2 is required for actin cytoskeletal remodeling to promote primary granule exocytosis. Treatment of neutrophils with low doses (≤10 μM) of the actin-depolymerizing drugs latrunculin B (Lat B) or cytochalasin B (CB) enhanced both formyl peptide receptor- and Ca2+ ionophore-stimulated exocytosis. Higher concentrations of CB or Lat B, or stabilization of F-actin with jasplakinolide (JP), inhibited primary granule exocytosis measured as myeloperoxidase release but did not affect secondary granule exocytosis determined by lactoferrin release. These results suggest an obligatory role for F-actin disassembly before primary granule exocytosis. However, lysates from secretagogue-stimulated neutrophils showed enhanced actin polymerization activity in vitro. Microscopic analysis showed that resting neutrophils contain significant cortical F-actin, which was redistributed to sites of primary granule translocation when stimulated. Exocytosis and actin remodeling was highly polarized when cells were primed with CB; however, polarization was reduced by Lat B preincubation, and both polarization and exocytosis were blocked when F-actin was stabilized with JP. Treatment of cells with the small molecule Rac inhibitor NSC23766 also inhibited actin remodeling and primary granule exocytosis induced by Lat B/fMLF or CB/fMLF, but not by Ca2+ ionophore. Therefore, we propose a role for F-actin depolymerization at the cell cortex coupled with Rac-dependent F-actin polymerization in the cell cytoplasm to promote primary granule exocytosis.
- Rac guanosine triphosphatase
human neutrophils are prominent white blood cells that play an important role in defense against microbial infection. Intense activation by bacterial or inflammatory stimuli triggers neutrophil degranulation, releasing a range of antibacterial factors that can exert collateral tissue damage. Degranulation involves granule translocation to the plasma membrane followed by exocytosis, a tightly regulated process (33). Excessive degranulation from neutrophils is a central feature in numerous inflammatory disorders such as severe asthma, emphysema, and rheumatoid arthritis (45).
Neutrophils contain four different granule subtypes: primary (azurophilic), secondary (specific), and tertiary (gelatinase) granules, as well as secretory vesicles. The signaling pathways that control exocytosis of granules appear to be unique to each subtype. Granules are released in a hierarchical fashion in response to secretagogues by first releasing secretory vesicles, followed by tertiary, secondary, and finally primary granules (44). Granule subtypes contain overlapping as well as unique luminal contents. Primary granules contain some of the most potent cytolytic enzymes that aid in digestion of pathogens, such as elastase and myeloperoxidase, which also significantly contribute to host tissue damage (23, 26), which emphasizes the importance of studying the regulation of their exocytosis mechanism.
Rab and Rho GTPases have been shown to regulate exocytosis in a variety of secretory cells, including mast cells (5, 39, 40), cytotoxic T lymphocytes (3, 20, 29, 48), and eosinophils (31, 35). Recent studies have defined the need for these two classes of small monomeric GTPases in regulating exocytosis of primary granules in neutrophils. Both Rab27a- and Rac2-deficient mice show impaired secretion of the primary granule enzyme myeloperoxidase (1, 37). Rab27a, via its effector protein JFC1/Slp1, may act to discriminate between primary granules destined for exocytosis from those which preferentially fuse intracellularly with phagosomes (37). The precise role that Rac2 plays in exocytosis remains unclear. In many cell types, Rho GTPases such as Rac are known to be key regulators of cytoskeletal remodeling. This was demonstrated in their ability to activate cytoskeletal remodeling and contribute to cell motility and chemotaxis (39, 41). Importantly, neutrophils deficient in Rac2 have defects in filamentous (F)-actin assembly, which prevents cell migration, as distinct from Rac1 deficiency leading to inhibition of cell spreading (14, 19). Rac2 is also the predominant Rac protein expressed in neutrophils and is the main GTPase required for activation of the superoxide-generating NADPH oxidase complex (14, 22).
Myeloid cells contain an F-actin-rich cortical region that is proposed to act as a barrier against granule docking and fusion at the plasma membrane. A proteomic analysis of neutrophil granule subtypes revealed that actin associates with all granule populations (28). Indeed, exocytosis of all granules is enhanced by preincubation of neutrophils with the actin depolymerization drug cytochalasin B, which favors the actin-barrier hypothesis (7, 28). However, other studies have shown that actin depolymerization inhibits exocytosis, suggesting that F-actin formation instead facilitates exocytosis (13, 16, 27, 34, 36). This is plausible, since neutrophil chemotaxis is triggered by polarized F-actin assembly, which could similarly drive polarized mobilization of granules on this actin network (18, 39). Therefore, a role for both actin depolymerization and polymerization during exocytosis is feasible.
In this study, we propose that Rac-mediated actin polymerization is necessary for directing granules in the cell cytoplasm to the plasma membrane, whereas actin depolymerization must occur concurrently at the cell cortex to allow exocytosis. We found that drugs which promote actin depolymerization stimulated receptor-mediated exocytosis at low dosage. Stabilization of F-actin specifically inhibited primary granule exocytosis but had little effect on secondary granule exocytosis. Microscopic analyses of neutrophils stimulated with the actin drug cytochalasin B together with f-Met-Leu-Phe showed that cortical actin was remodeled to a polarized state with primary granule marker colocalization. The actin drugs latrunculin and jasplakinolide affected both granule distribution and the polarization of actin remodeling; however, only jasplakinolide blocked exocytosis. A recently discovered small molecule inhibitor of Rac (15) was also found to block actin remodeling and primary granule exocytosis. Surface exposure of the primary granule membrane marker CD63 was used to further quantify the effects of actin and Rac-directed drugs. Our findings suggest that Rac-mediated F-actin formation is necessary for primary granule movement to the cell membrane, whereas concurrent actin depolymerization at the cell cortex stimulates granule exocytosis in general.
Isolation of human peripheral blood neutrophils.
Human polymorphonuclear neutrophils were isolated from healthy donors in accordance with protocols approved by the University of Alberta Health Research Ethics Board. Whole blood (50–100 ml) was drawn from donors, mixed with 6% dextran in RPMI 1640 (Invitrogen, Burlington, ON, Canada), and incubated at room temperature for 30 min to allow sedimentation of red blood cells. The upper leukocyte-rich phase was layered onto 15 ml of 5.7% Ficoll and centrifuged at room temperature at 400 g for 30 min to separate leukocytes and monocytes from granulocytes. The granulocyte pellet was exposed to 1.5 ml of sterile deionized water for 20 s to lyse any remaining red blood cells and then quickly placed into excess buffer A (RPMI 1640 and 5 mM EDTA) and centrifuged at room temperature at 300 g for 5 min. After centrifugation, the cell pellet was resuspended in buffer B (RPMI 1640, 5 mM EDTA, and 2% FBS). Cells were then allowed to rest on ice for 1 h before experiments.
Secretion assays were performed by analyzing levels of granule proteins secreted into cell supernatants and by analyzing surface expression of CD63 via flow cytometry (1, 28). Resting cells were resuspended at 1 × 106 cells/ml in phenol red-free RPMI 1640. For biochemical analysis of granule marker exocytosis, 50 μl of cell suspension were added to each well of a 96 V-well plate containing actin drugs and stimuli in RPMI 1640 to a final volume of 250 μl. After stimulation, microplates were centrifuged at 300 g at 4°C for 6 min, and the levels of myeloperoxidase (MPO) and lactoferrin (LTF) in supernatants were determined as a measurement of primary and secondary granule exocytosis, respectively. MPO was assayed using tetramethylbenzidine (TMB), and LTF was assayed using quantitative immunoblot analysis. For flow cytometry analysis, 1 ml of cell suspension was aliquoted into microfuge tubes containing actin drugs and stimuli in RPMI. After stimulation, cells were pelleted, fixed in 5% formalin, blocked in PBS containing 5% nonfat milk, and stained with FITC-conjugated anti-CD63 (Serotec, Raleigh, NC). Cytochalasin B (CB; destabilizes F-actin; Sigma-Aldrich, Mississauga, ON, Canada), latrunculin B (Lat B; destabilizes F-actin; Calbiochem, San Diego, CA), jasplakinolide (JP; stabilizes F-actin; Calbiochem), and the small molecule Rac inhibitor NSC23766 (Calbiochem) were prepared as 10 mM stock solutions in DMSO and diluted before use. Neutrophils were pretreated with these drugs for 5–15 min at 37°C before stimulation with 2.5 μM Ca2+ ionophore (A-23187) or 5 μM f-Met-Leu-Phe (fMLF) (Sigma-Aldrich) for 15 min at 37°C to induce degranulation. In some cases, neutrophils were preincubated with 10 μM CB for 5 min before addition of 5 μM fMLF for 15 min (CB/fMLF; typical secretagogue). Cells showed >95% viability as determined by Trypan blue exclusion at the end of all incubations.
Rac activation assays.
Activated (GTP bound) Rac1 and Rac2 were affinity precipitated from neutrophil lysates using glutathione S-transferase-p21 binding domain (GST-PBD) (2). Lysates were prepared from 8 × 106 cells by sonication in 400 μl of H buffer [20 mM HEPES-KOH, pH 7.5, 1 mM DTT, 5 mM MgCl2, 60 mM NaCl, 1% Triton X-100 plus protease inhibitor cocktail (PIC): 1 μg/ml each leupeptin, pepstatin, antipain, and aprotinin, 1 mM phenylmethylsulfonyl fluoride]. Cell debris was removed by centrifugation, and 300 μg of lysate were incubated with 30 μg of immobilized GST-PBD in 400 μl of H buffer for 30 min at 4°C. The beads were recovered, washed four times in H buffer, and resuspended in 45 μl of Laemmli sample buffer. Each sample (15 μl) was analyzed by immunoblot using antibodies specific for Rac1 (ARC03; Cytoskeleton, Denver, CO) or Rac2 (07-604; Upstate, Waltham, MA). Immunoreactive bands were detected using IRDye800 secondary antibodies (Rockland Immunochemicals, Gilbertsville, PA) and an Odyssey image analysis system (LiCor, Seattle, WA).
Measurement of O2− release from neutrophils.
Generation of extracellular O2− from cells in suspension was measured as previously described (32). Briefly, cells (1–2 × 106) were suspended in 1-ml microcuvettes containing supplemented PBS (PBS+; or PBS, pH 7.4, with 1.2 mM MgCl2, 5 mM KCl, 0.5 mM CaCl2, 5 mM glucose, and 0.1% BSA) and 50 μM ferricytochrome c at 25°C. The mixture was blanked at 550 nm in a Beckman DU 640 spectrophotometer (Beckman Instruments, Mississauga, ON, Canada) before addition of 16 nM PMA (Calbiochem) or 5 μM fMLF. To test the effects of NSC23766 on O2− production, we added 160 μM NSC23766 to 2 × 107 cells/ml in RPMI 1640 and incubated at 37°C for 15 min before treatment with PMA or fMLF.
Stimulated cells were fixed in freshly prepared 2% paraformaldehyde in 0.25 M sucrose-PBS while still in suspension. Fixed cells were adhered to poly-l-lysine-coated glass slides, permeabilized with 0.5% Triton X-100 in PBS, and stained with 10 μg/ml anti-CD63 (Serotec) conjugated to Alexa Fluor 488 (Invitrogen) to detect primary granules and 0.3 μM rhodamine-phalloidin (Invitrogen) to detect F-actin. Images were acquired on an Olympus FV1000 confocal laser scanning microscope (Olympus Canada, Markham, ON, Canada) with a ×63/1.4 NA plan apochromat objective and processed using Olympus Fluoview software.
Stimulated cells were fixed in 2.5% glutaraldehyde in 0.1 M cacodylate, pH 7.2, at 4°C and stained with DAB solution (2.5 mM diaminobenzidine and 0.02% hydrogen peroxide in 0.1 M cacodylate, pH 7.2) for 5 min to enhance the electron density of peroxidase-containing vesicles. Cell pellets were washed three times in 0.1 M cacodylate, pH 7.2, embedded in 1% ultrapure agarose, and post-fixed in 0.2% aqueous osmium tetroxide for 1 h at 4°C. Samples were embedded in Epon resin after serial dehydration; ultrathin sections were cut and mounted on copper-coated grids and post-stained with saturated uranyl acetate and lead citrate. Sections were viewed on a Philips 410 transmission electron microscope (Philips Electron Optics, Eindhoven, The Netherlands), and images were acquired using a SIS Megaview III charge-coupled device digital camera and AnalySIS software (Olympus).
Actin polymerization assay.
To assay cellular activity inducing actin polymerization, we used an established pyrene-actin polymerization assay (9, 25). Briefly, 12 μg of neutrophil lysate in 70 μl of lysis buffer (5 mM Tris-Cl, pH 8, 50 mM KCl, 0.2 mM CaCl2, 0.2 mM ATP, 0.17% Nonidet P-40, 0.35 mM MgCl2 and PIC) was mixed with 50 μl of actin polymerization stock mixture containing 35% pyrene-labeled actin (12 μM) in G buffer (5 mM Tris-Cl, pH 8, 0.2 mM CaCl2, and 0.2 mM ATP; Cytoskeleton). Fluorescence intensity readings were taken every 18 s using a QM-4SE spectrofluorometer (360-nm excitation/407-nm emission, 10-nm bandwidth, 2-s integration) with a four-position heated sample holder set to 30°C (Photon Technologies, London, ON, Canada). The pyrene-actin stock mixture was equilibrated for 10 min, and then test samples were added and fluorescence measurements taken for 30 min. Actin polymerization activity (APA) was calculated from polymerization curves by determining the average rate of fluorescence intensity (FI) increase for 30 min of reaction time divided by the number of micrograms of test sample protein (ΔFI/μg). APA values were normalized to values for untreated resting cells for each experiment with the APA of lysis buffer subtracted.
Calculations and statistical analysis.
Data were analyzed using one-way statistical analysis of variance (ANOVA), and post hoc analysis was determined using Tukey's post test. The data are presented as means ± SE.
Effect of actin drugs on primary granule exocytosis.
Bacterially derived N-formyl methionyl derivatives (fMLF) are potent stimuli for neutrophil activation. Isolated neutrophils can degranulate in response to fMLF stimulation in vitro, with modest secretion of most granule types except for primary granules, which fail to undergo exocytosis (1). However, primary granule exocytosis may be induced in vitro by “priming” via pretreatment with the actin-depolymerizing drug CB. This drug blocks the barbed ends of actin filaments, leading to cortical actin meshwork disassembly and enhancing fMLF-induced primary granule release (21, 46). This well-known effect of CB in neutrophils suggests that F-actin depolymerization from the cell cortex is specifically required for primary granule exocytosis.
To explore the role of actin remodeling in the regulation of human neutrophil primary granule exocytosis, we tested several actin drugs alone or in combination with fMLF stimulation. Incubation of neutrophils with the actin-depolymerizing drugs CB or Lat B alone did not stimulate primary granule exocytosis, as measured by MPO secretion, a luminal marker of primary granules (data not shown). However, both CB and Lat B induced a dose-dependent increase in MPO secretion in response to fMLF at concentrations ≤10 μM, whereas higher doses inhibited primary granule exocytosis (Fig. 1). Lat B, which is more specific for actin than CB and promotes actin depolymerization by sequestering actin monomers (47), enhanced fMLF-induced secretion at doses 100 times lower than CB. JP, which binds to and stabilizes F-actin (6), did not stimulate exocytosis in response to fMLF (Fig. 1). These results suggest that partial actin depolymerization is needed to promote exocytosis, whereas complete F-actin depolymerization is inhibitory.
We next determined whether JP affected granule exocytosis when neutrophils were stimulated with the typical secretagogues CB/fMLF and A-23187 (Fig. 2, A and B, respectively). For both conditions, JP inhibited primary granule exocytosis (Fig. 2, MPO) but had little effect on secondary granule exocytosis (Fig. 2, LTF). This confirms an obligatory requirement for actin depolymerization in initiating exocytosis specifically for primary granules.
Since Lat B is a more potent actin depolymerization drug than CB (47), we hypothesized that addition of Lat B to CB/fMLF- or A-23187-stimulated cells might enhance exocytosis. However, we found that Lat B had a bimodal effect that was specific for primary granule exocytosis. At a low dosage (1.25 μM), Lat B enhanced both CB/fMLF- and A-23187-stimulated primary granule exocytosis but had little effect on secondary granule exocytosis (Fig. 2, A and B, MPO and LTF, respectively). Higher doses of Lat B (50 μM) inhibited primary granule exocytosis, whereas secondary granule exocytosis was slightly enhanced under these conditions. High doses of Lat B did not result in loss of cell viability as determined by Trypan blue exclusion. This suggests that the mode of action by which Lat B enhanced exocytosis was by facilitating moderate actin depolymerization. This regulatory mechanism was not apparent for secondary granule exocytosis, and therefore further studies focused on primary granule exocytosis.
F-actin and primary granule distribution in stimulated neutrophils.
To confirm our biochemical findings and to visualize the effects of actin-altering drugs, we examined human neutrophils using confocal microscopy. Neutrophil primary granules were labeled with anti-CD63 antibodies conjugated to Alexa Fluor 488 and F-actin with rhodamine-phalloidin. Cells were treated similarly to those in secretion assays and then fixed while in suspension and adhered to poly-l-lysine-coated glass slides for staining. Resting neutrophils exhibited diffuse primary granule staining and cortical actin staining in a ringlike structure (Fig. 3, resting). Intensity profiling (Fig. 3, right) showed sharp peaks at the cell periphery for rhodamine-phalloidin, corresponding to the actin ring, but relatively even distribution of the primary granule marker CD63. Upon stimulation with CB/fMLF, F-actin polarization to the cell's edge occurred along with a redistribution of primary granules to the same sites; this is evident from colocalizing peaks on one side of the intensity profile. CB or fMLF alone did not induce these rearrangements (Supplemental Fig. S1). (Supplemental data for this article is available at the American Journal of Physiology-Cell Physiology website.) Pretreatment with JP followed by CB/fMLF stimulation resulted in primary granule and F-actin staining throughout the cell with diffuse cortical staining (Fig. 3, JP CB/fMLF). Interestingly, Lat B increased primary granule translocation to the cell periphery; however, it also reduced cytoplasmic F-actin to a diffuse, nonpolarized state (Fig. 3, Lat B CB/fMLF). These findings confirm biochemical results showing that Lat B enhances CB/fMLF-induced primary granule exocytosis, although it disrupts the polarization mechanism.
Although confocal microscopy showed that CD63 accumulated at more peripheral regions of the cell following stimulation, the resolution of this technique did not allow for quantification of exocytosis. We used flow cytometry to quantitatively analyze surface upregulation of CD63, which occurs following primary granule exocytosis. Unstimulated neutrophils or incubation with either CB or fMLF alone showed little surface CD63 labeling (Fig. 4A, hatched bars). Stimulation with CB/fMLF triggered elevated levels of CD63 at the plasma membrane that could be further enhanced by 30% when preincubated with Lat B (Fig. 4A, solid bars). JP effectively blocked exocytosis to resting cell levels (Fig. 4A, JP).
Morphological analysis of neutrophils by electron microscopy.
We used electron microscopy to obtain high-resolution images of neutrophils and further characterize neutrophil morphology after treatment with actin drugs. Resting or fMLF-treated cells showed numerous electron-dense granules evenly distributed throughout the cytoplasm, whereas CB/fMLF-stimulated cells showed significantly fewer vesicles (Fig. 5, compare resting, fMLF, and CB/fMLF). Pretreatment with JP before CB/fMLF stimulation resulted in clustering of primary granules in the center of the cell (Fig. 5D) with a granule count similar to that of resting cells (Fig. 5F). Interestingly, pretreatment with Lat B before CB/fMLF stimulation decreased the number of primary granules within the neutrophil, and those that remained were at the cell periphery (Fig. 5E). These findings confirm our initial conclusions that moderate actin depolymerization is needed for primary granule exocytosis.
Rac1 and Rac2 activation in stimulated human neutrophils.
We have recently shown that primary granule exocytosis was reduced in bone marrow neutrophils isolated from rac2−/− mice, which was associated with a lack of primary granule translocation to the cell membrane during stimulation (1). Therefore, we investigated whether Rac regulates primary granule exocytosis by altering actin cytoskeleton dynamics in human neutrophils. To do this, we used the small molecule Rac inhibitor NSC23766, which has been shown to inhibit GDP/GTP exchange for both Rac1 and Rac2, a requirement for small G protein activation and signaling (8, 15). NSC23766 blocks Rac function by binding to Trp56 and specifically inhibiting the binding of guanine nucleotide exchange factors (GEFs) Trio and Tiam1 (15). However, it has yet to be shown to inhibit GTP binding to Rac1 and Rac2 in human neutrophils. Therefore, we examined the effects of this inhibitor on stimulated neutrophils using a Rac activation assay (2). This assay uses GST conjugated to the Rac-binding domain of p21 activated kinase (PAK) to affinity isolate GTP-bound Rac. Stimulation with fMLF showed high levels of GTP-bound Rac1 (Fig. 6A) and Rac2 (Supplemental Fig. S2) within 1 min of treatment, which was reduced to basal levels by 15 min. Interestingly, the addition of actin drugs resulted in sustained levels of activated Rac (Fig. 6A and Supplemental Fig. S2, compare 1 min and 15 min). These observations show that CB and Lat B prolong the activation of Rac1 and Rac2 initiated by fMLF. Stimulation with actin drugs alone (CB or Lat B) showed no activation of Rac.
We next tested the effects of the small molecule Rac inhibitor NSC23766 on Rac in pull-down assays. Preincubation of neutrophils with NSC23766 effectively inhibited fMLF- and CB/fMLF-triggered Rac activation (Fig. 6A and Supplemental Fig. S2, compare NSC23766 + and − lanes). The inhibitory effect of NSC23766 was less evident on Lat B/fMLF-stimulated neutrophils (see discussion). We investigated this further, via titration of Lat B, to define the critical concentration of Lat B at which the inhibitory effect of NSC23766 on Rac activation could be overcome. Cells stimulated for 1 min required Lat B concentrations above 0.1 μM to enhance exocytosis as well as initiate the reversal of NSC23766 inhibition; however, complete reversal was not obtained until 100 μM Lat B (Fig. 6B, left). At 15 min of stimulation, 0.1 μM Lat B was adequate to initiate reversal of NSC23766 inhibition, and complete reversal of NSC23766 inhibition was observed between 1 and 10 μM Lat B (Fig. 6B, right). Since the action of Lat B is well defined (binds to actin monomers), we conclude that the effects of Lat B on enhancing Rac activation may be due to a secondary cellular response to the shortage of F-actin.
NSC23766 inhibits primary granule exocytosis and actin polymerization.
To link Rac activation to the exocytosis mechanism in human neutrophils, we examined the effects of NSC23766 on primary granule secretion. Human neutrophils were pretreated with NSC23766 for 15 min and then stimulated with CB/fMLF. This brief pretreatment significantly reduced the secretion of primary granule MPO (Fig. 7A) and upregulation of CD63 as determined by flow cytometry (Fig. 7B), which correlates with the inhibition of Rac activation (Fig. 6). Inhibition was evident at 10 μM NSC23766 and maximal at >40 μM. Secretion of MPO and upregulation of CD63 in response to A-23187 was unaffected by NSC23766, even at doses up to 160 μM. Lat B/fMLF-stimulated cells also showed reduced primary granule exocytosis when pretreated with NSC23766 (Fig. 7), even though Rac remained activated (GTP bound) under these conditions (10 μM Lat B, see Fig. 6).
To link the role of Rac to F-actin formation, we investigated whether inhibition of Rac by NSC23766 would affect F-actin. We used an in vitro actin polymerization assay that measures the capacity of a sample to stimulate the polymerization of exogenously added pyrene-actin. Pyrene-actin undergoes a fluorescence intensity increase when incorporated into F-actin (9). Lysates prepared from fMLF- or CB/fMLF-stimulated cells exposed to NSC23766 showed reduced capacity for F-actin formation compared with those unexposed (Fig. 8 and Supplemental Fig. S3). Interestingly, there were negligible differences in lysates prepared from cells that were stimulated with A-23187.
We also examined the effects of NSC23766 on respiratory burst induced by PMA or fMLF. This was done to determine the specificity of NSC23766 on Rac-mediated functions in neutrophils. Rac2 is essential for O2− release from neutrophils (17). Pretreatment of neutrophils with NSC23766 for 15 min had no effect on high levels of O2− release from neutrophils stimulated with PMA (Fig. 9A) or the much reduced fMLF-stimulated levels (Fig. 9B). Although it has been proposed that activated Rac regulates respiratory burst (reviewed in Ref. 11), our observation, combined with Rac activation results (Fig. 6), correlates with previous findings in which Rac-GTP formation was not required for activation of the NADPH oxidase system in human neutrophils (17) or macrophages (4). Stimuli that activate G protein-coupled receptors, such as fMLF, have been shown to activate the GEFs Vav1 and P-Rex1, leading to Rac activation required for respiratory burst (30, 50). Since NSC23766 was unable to block O2− release from PMA-stimulated neutrophils, this suggests that a different GEF, such as Trio or Tiam1, may be responsible for activating Rac-mediated exocytosis of primary granules. Together, these results show that the Rac inhibitor NSC23766 specifically targets Rac-regulated actin remodeling and exocytosis in response to CB/fMLF and Lat B/fMLF.
Morphological analysis of NSC23766-treated neutrophils.
Our findings indicate that primary granule exocytosis was inhibited by NSC23766 in a dose-dependent fashion when stimulated with CB/fMLF or Lat B/fMLF. To confirm these results, we examined NSC23766-treated neutrophils via confocal and electron microscopy. Stimulation of neutrophils with CB/fMLF or Lat B/fMLF resulted in primary granule translocation to the cell periphery (Figs. 10 and 11), which was also observed in intensity profiles as large peaks located at the peripheries (Fig. 10, right). Pretreatment of cells with NSC23766 before stimulation with CB/fMLF modestly reduced F-actin polarization at cell edges (Fig. 10), inhibited primary granule translocation to the cell membrane (Fig. 11), and inhibited exocytosis as determined by granule counts of electron microscopy sections (Fig. 11E). Similar results were obtained for NSC23766-pretreated Lat B/fMLF-stimulated neutrophils. These results suggest that NSC23766 inhibited primary granule movement to the cell membrane in cells in suspension. Together, our results suggest that Rac plays a vital role in primary granule exocytosis via stimulation of actin polymerization to aid in granule mobilization to the cell periphery.
Role of actin remodeling during exocytosis.
Actin remodeling is considered to be integral to exocytosis of neutrophil granules (7, 28, 42), although the specific mechanism has not been fully elucidated. In the present study we have shown a requirement for Rac-mediated actin polymerization as well as cortical actin depolymerization in primary granule exocytosis. In resting cells, we observed an F-actin cortical ring, similar to that observed in resting mast cells (38), which dispersed after stimulation with CB/fMLF and Lat B/fMLF. This is in agreement with the concept of an F-actin barrier preventing inadvertent granule docking and fusion in resting cells. However, the presence of the F-actin ring in resting cells is distinct from previous studies, which showed diffuse F-actin in resting neutrophils that assembled into a cortical ring upon stimulation by fMLF (12, 14). This discrepancy may be due to differences in neutrophil isolation techniques. In earlier studies, human neutrophils were prepared using dextran sedimentation and discontinuous plasma-Percoll gradients, whereas we used Ficoll to separate neutrophils from dextran-generated plasma and then rested the cells in EDTA-containing medium on ice to lower neutrophil activity (i.e., to achieve “resting cells”) before experiments.
We also examined neutrophil morphology by electron microscopy to obtain high-resolution images. These images showed a clear difference in the effects of actin drugs (Fig. 5, D vs. E). Stabilization of actin with JP resulted in the central accumulation of granules, whereas depolymerization allowed for significant plasma membrane localization, especially in the case of Lat B (Fig. 5E). Biochemical assays confirm that low concentrations of Lat B increase granule translocation and exocytosis, indicating a requirement for limited actin depolymerization, which is similar to that reported for exocytosis in neuroendocrine cells (16, 36).
Signaling for actin remodeling during exocytosis could be linked to other neutrophil functions. For example, when neutrophils encounter a chemoattractant, they undergo actin remodeling for chemotaxis (14, 49, 51), a mechanism clearly linked to Rho protein function (41), and it is likely that this may also facilitate polarized granule translocation and exocytosis. This apparently conflicts with the need for depolymerization of F-actin during primary granule exocytosis, since depolymerizing agents are required to stimulate exocytosis in vitro (1, 21, 46). Results from confocal analysis of resting neutrophils suggest that the cortical actin ringlike structure acts as a physical barrier to prevent uncontrolled granule docking and fusion as previously shown in mast cells (38, 39).
Our results show that upon stimulation with CB/fMLF, there is a polarized reorganization of the F-actin to sites of primary granule translocation (Figs. 3 and 10, CB/fMLF). This is seemingly incompatible with the uniform application of stimulus in our experiments. However, it has been previously shown that neutrophils contain an inherent “sidedness” and produce a polarized response, even when placed in an environment of uniform stimuli (51, 52). We reproducibly observed similar polarized responses in all image analyses performed. Polarized actin was not as evident in Lat B/fMLF-stimulated neutrophils; therefore, Lat B treatment likely disrupts the polarized exocytosis mechanism but not exocytosis itself.
Role of Rac signaling during exocytosis.
Rho proteins such as Rac control actin remodeling during the neutrophil response to chemoattractant (41, 51). Rac is also activated in response to the secretagogue fMLF (Ref. 8, Fig. 6); however, fMLF alone is not sufficient to stimulate primary granule exocytosis (1, 28). Recent studies have defined the need for the Rab27a small GTPase in regulating exocytosis of primary granules in immune cells and secretory granules in neuronal cells (10, 37). Furthermore, these studies also link Rab function to actin-based transport of secretory vesicles. It is interesting to speculate these two GTPases act coordinately in their regulation of exocytosis, since Rab GTPases have been shown to tether vesicles to actin-based motors, whereas Rac may act to remodel actin filaments (33).
We showed that combinations of CB/fMLF or Lat B/fMLF are potent secretion stimuli, and interestingly, these combinations also caused sustained Rac activation for up to 15 min (Fig. 6). This finding suggests that actin depolymerization by CB or Lat B feeds back into the Rac activation pathway to maintain a predominantly GTP-bound form. Studies of sustained activation of small GTPases including Rab27a have been previously observed to enhance exocytosis (10). For example, studies using guanosine 5′-O-(3-thiotriphosphate) (GTPγS) as a secretagogue have concluded that sustained activation of GTP-binding proteins may be a sufficient stimulus for membrane fusion; however, the GTPase class involved has not been confirmed (i.e., Arf, Rab, or Rho) (5, 40, 43). Our result would suggest that Rac may be a primary target of sustained activation by GTPγS in secretion. This is further supported by studies of the small molecule Rac inhibitor NSC23766. We have shown for the first time that NSC23766 inhibits primary granule exocytosis in response to CB/fMLF or Lat B/fMLF stimulation. This effect was specific for primary granules, since LTF secretion, a marker for secondary granule exocytosis, was not inhibited by NSC23766 (data not shown). We were unable to distinguish whether NSC23766 mainly inhibited Rac1 or its closely related homolog, Rac2, in neutrophil primary granule exocytosis. However, neutrophils mainly express Rac2, and Rac2 is essential for primary granule release (1). Therefore, these results support a role for Rac2 in F-actin-mediated human neutrophil primary granule exocytosis.
Biochemical analysis showed that NSC23766 effectively blocked fMLF- and CB/fMLF-stimulated Rac-GTP formation but not that of Lat B/fMLF (Fig. 6). These results suggest that the more potent actin depolymerization reagent (Lat B over CB) may act to sustain Rac signaling via a feedback loop, similar to that observed with CB/fMLF at 15 min of stimulation, but that additionally overrides the inhibitory effects of NSC23766. This could be explained by the existence of a molecular sensor of increased G-actin that promotes continued activation of Rac via a GEF that is not inhibited by NSC23766. Currently, the association of two Rac GEFs, Trio and Tiam1, are known to be blocked by NSC23766, whereas this inhibitor has no effect on Vav1 association (15).
Our findings from imaging and biochemical analyses of granule exocytosis indicated that NSC23766 inhibited cytoplasmic F-actin polymerizing activity, as well as primary granule translocation for subsequent exocytosis, in response to CB/fMLF or Lat B/fMLF. Flow cytometry experiments confirmed that fMLF or CB/fMLF stimulation caused little if any increase in F-actin, and similarly, NSC23766 pretreatment did not affect overall levels of F-actin (data not shown). In addition, NSC23766 had no effect on actin depolymerization by CB or Lat B combined with fMLF (Fig. 10). This suggests that inhibition of Rac does not result in actin depolymerization and instead defines a role for Rac as a mediator of F-actin remodeling to facilitate granule translocation to the plasma membrane. It was recently shown that Rac controls actin remodeling by generating free barbed ends and new filament assembly, which is in accord with our theory (49).
In Ca2+ ionophore-induced exocytosis, NSC23766 showed no effect, suggesting that Ca2+ acts downstream of Rac to induce granule translocation and secretion. It has been shown in RBL-2H3 cells that Ca2+ signaling is highly important in granule exocytosis and that Ca2+ levels are reduced when cells are exposed to dominant negative constructs of Rac proteins and restored when exposed to constitutively active forms of Rac (24). Therefore, Rac-mediated signaling in neutrophils may exploit Ca2+ signaling as a mode of primary granule translocation and exocytosis.
Together, our results illustrate a possible mechanism by which Rac regulates exocytosis of primary granules, via facilitating F-actin formation necessary for granule translocation to sites of exocytosis. Depolymerization of actin at the cell cortex must occur concurrently with cytoplasmic F-actin formation for exocytosis of primary granules. These results contribute to our understanding of neutrophil biology, with relevance to their role in degranulation and release of cytotoxic mediators in disease.
This work was supported by grants from AllerGen NCE, the Canadian Lung Association, and the CIHR.
We thank Honey Chan and Emily MacLean for expert technical assistance. P. Lacy is a Canadian Lung Association/Canadian Institutes of Health Research (CIHR) New Investigator, and G. Eitzen is a CIHR New Investigator and Alberta Heritage Foundation for Medical Research Scholar.
↵* P. Lacy and G. Eitzen shared senior author.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society