Expression and role of a2 vacuolar-atpase (a2v) in trafficking of human neutrophil granules and exocytosis

Transcription

1 Article Expression and role of a2 vacuolar-atpase (a2v) in trafficking of human neutrophil granules and exocytosis Alice Gilman-Sachs, 1 Anjali Tikoo, Leyla Akman-Anderson, Mukesh Jaiswal, Evangelos Ntrivalas, and Kenneth Beaman Department of Microbiology and Immunology, Rosalind Franklin University of Medicine and Science, North Chicago, Illinois, USA RECEIVED DECEMBER 22, 2014; REVISED MARCH 19, 2015; ACCEPTED MARCH 20, DOI: /jlb.3A RR ABSTRACT Neutrophils kill microorganisms by inducing exocytosis of granules with antibacterial properties. Four isoforms of the a subunitofv-atpase a1v,a2v,a3v,anda4v have been identified. a2v is expressed in white blood cells, that is, on the surface of monocytes or activated lymphocytes. Neutrophil associated-a2v was found on membranes of primary (azurophilic) granules and less often on secondary (specific) granules, tertiary (gelatinase granules), and secretory vesicles. However, it was not found on the surface of resting neutrophils. Following stimulation of neutrophils, primary granules containing a2v as well as CD63 translocated to the surface of the cell because of exocytosis. a2v was also found on the cell surface when the neutrophils were incubated in ammonium chloride buffer (ph 7.4) a weak base. The intracellular ph (cytosol) became alkaline within 5 min after stimulation, and the ph increased from 7.2 to 7.8; this ph change correlated with intragranular acidification of the neutrophil granules. Upon translocation and exocytosis, a2v on the membrane of primary granules remained on the cell surface, but myeloperoxidase was secreted. V-ATPase may have a role in the fusion of the granule membrane with the cell surface membrane before exocytosis. These findings suggest that the granule-associated a2v isoform has a role in maintaining a ph gradient within the cell between the cytosol and granules in neutrophils and also in fusion between the surface and the granules before exocytosis. Because a2v is not found on the surface of resting neutrophils, surface a2v may be useful as a biomarker for activated neutrophils. J. Leukoc. Biol. 97: ; Introduction Neutrophils are the first line of defense against many pathogens, such as bacteria or fungi, and are major mediators of innate Abbreviations: a2v = isoform of the a2 subunit of V0 domain of vacuolar ATPase, Abcam = antimyeloperoxidase antibody, AM = acetoxymethyl ester, ACK = ammonium chloride buffer, ph 7.4, Baf = bafilomycin A1, CYTD = cytochalasin D, DHR = dihydrorhodamine, KRBC = Krebs Ringer phosphate buffer with bicarbonate and calcium, MMP-9 = metalloproteinase-9, MPO = myeloperoxidase, ROS = reactive oxygen species, SNARE = SNAP receptor, V-ATPase = vacuolar ATPase immunity [1, 2]. Various types of complex mechanisms are used by neutrophils to defend the host against these pathogens. These mechanisms consist of phagocytosis of extracellular pathogens, generation of ROS, exocytosis of granules containing antimicrobial molecules or integration with phagosomes, and formation of neutrophil extracellular traps (NETS). Some of these mechanisms are ROS dependent, and some are ROS independent [3, 4]. The neutrophil granules that are exocytosed have been isolated and characterized and can be distinguished by specific matrix and membrane proteins [2, 5, 6]. V-ATPases have been identified in neutrophils [7, 8], but their functional role in trafficking and exocytosis of neutrophil granules is not well understood. V-ATPases are multisubunit enzymatic complexes that regulate the intracellular ph within vesicles, in the cytoplasm, or in the extracellular environment [9 12]. They are found on lysosomes, endosomes, selected compartments of the Golgi apparatus, clathrin-coated vesicles, and transport and secretory vesicles. In mammalian cells, the V-ATPase is a very large molecular complex of about 800 kda consisting of 2 domains: the V0 domain, which is embedded in the plasma membrane of vesicles or the cell membrane (depending on the cell), and the V1 domain, which is exposed to the cytoplasm of the cell or the extracellular environment [13, 14]. The 2 domains interact such that the V1 domain converts ATP to ADP and, in the process, generates protons that, in turn, causes the rotation of the ring embedded in the membrane (V0); this process delivers H + from the cytoplasmic to the endosomal or extracellular side of the membrane, probably via channels formed by subunit V0a [10]. The reversible dissociation of V1 and V0 from the V-ATPase is a mechanism of physiologic regulation that appears to be widely conserved from yeast to animal cells [15, 16]. The V0a subunit links these 2 domains and is responsible for the subcellular location of the V-ATPase. Four isoforms of subunit V0a have been identified; they are a1v, a2v, a3v, and a4v, but their distribution is different in various types of eukaryotic cells. The isoforms of this 116-kDa V0a subunit contain information that targets V-ATPases toward different subcellular membranes [17]. 1. Correspondence: Department of Microbiology and Immunology, Rosalind Franklin University of Medicine and Science, 3333 Green Bay Road, N. Chicago, IL 60064, USA. alice.gilmansachs@rosalindfranklin.edu /15/ Society for Leukocyte Biology Volume 97, June 2015 Journal of Leukocyte Biology 1121

2 We have studied the a2 isoform of the V-ATPase (ATP6V0a2) for many years and have shown that it is a key component in inflammation, pregnancy outcome, and apoptosis [18 21]. We have produced and characterized a unique monoclonal antibody, anti-a2v(2c1), which reacts with the N-terminal region of the a2 isoform of subunit V0a in vacuolar ATPase (amino acids ) [22]. Anti-a2V does not react with a1, a3, or a4 isoforms of V-ATPases. For hematopoietic cells, a2v-atpase is expressed in lymphocytes and monocytes [23, 24]. It is present in the cytoplasm of most cells and in many vesicles; it is not found on the plasma membrane of lymphocytes until they are activated, but it is constitutively expressed on the plasma membrane of monocytes. However, characterization of the a2v expression in neutrophils has not been well studied. Therefore, the purpose of these studies was to determine whether a2v was expressed in neutrophils and to understand its role in the exocytosis function of these cells. V-ATPases have been shown to have a role in neutrophil function; they move from inside the cell to the surface of neutrophils after incubation with GM-CSF for 24 h [8]. In these studies, we determined the cellular location of the a2v isoform in neutrophils, specifically in the neutrophil granules, and investigated the relationship between stimulation of neutrophils and expression of a2v on the plasma membrane. In addition, ph changes following neutrophil activation were investigated to obtain a better understanding of the events that occur during inflammatory responses related to ph. MATERIALS AND METHODS Neutrophils Whole blood was collected in heparin tubes from healthy volunteers, according to a protocol approved by the institutional review board, and was used within 4 hours. In some experiments, cells in whole blood were treated with stimulators of neutrophils, washed, and then stained with antibodies; the red blood cells were then lysed with ACK. In other experiments, the red blood cells were lysed first with ACK, and then, the remaining white blood cells were stimulated. Isolation of neutrophils was performed by 2 methods: the first method was a combination of Ficoll-Hypaque (GE Healthcare BioSciences, Uppsala, Sweden) differential centrifugation, followed by lysis of remaining red blood cells with ACK; and the second method was by dextran sedimentation (6% final concentration) of the red blood cells for 1 h, Ficoll- Hypaque differential centrifugation to separate the mononuclear cells from the polynuclear neutrophils, followed by a 20 s hypotonic lysis of the remaining red blood cells at 4 C in polystyrene tubes [25]. All other reagents were maintained at room temperature. In all experiments, viability was.98% as tested by trypan blue dye exclusion. In flow cytometry analysis experiments, DAPI was used to exclude any dead cells. Reagents and antibodies Monoclonal anti-a2v-atpase (2C1) antibody was generated as previously described [18, 20, 26] and is specific for amino acids in the transmembrane region of polypeptide a2 of a2v-atpase (ATP6V0A2). This mouse antibody was used as either directly conjugated to Alexa Fluor 647 (AF647) (Covance, Denver, PA, USA) or biotinylated and then reacted with Streptavidin-AF594 (formerly Invitrogen, now Life Technologies, Carlsbad, CA, USA) or used unconjugated and then detected with AF488 or AF594 conjugated goat anti-mouse IgG. The specificity of this antibody was tested against peptides unique to each of the isoforms of the V0a subunit [27]. To identify granule subsets in combination with anti-a2v (2C1), the following antibodies were used: FITC- mouse monoclonal anti-cd63 [Becton-Dickinson (BD), Franklin Lakes, NJ, USA] or AF594-rabbit Abcam (Cambridge, Cambridgeshire, UK) to identify azurophilic (primary) granules; FITC-anti- CD66b (BD) (membrane) or rabbit anti-lactoferrin antibody (matrix) (Abcam) to identify specific (secondary) granules; AF488-mouse anti-cd11b (BD) and rabbit anti-gelatinase antibody (MMP-9) (Abcam) to identify gelatinase (tertiary) granules; and FITC-mouse anti-cd35 (BD) to identify secretory granules. AF488-goat anti-mouse IgG (green), AF594-goat antimouse IgG (red), AF594-donkey anti-mouse IgG, AF594 goat anti-rabbit IgG, and AF488-goat anti-rabbit IgG were obtained from Life Technologies. Stimulation of whole blood with neutrophil stimulators Stimulation of whole blood was performed in KRBC at ph 7.4 (Sigma-Aldrich, St. Louis, MO, USA) (1.26 g/l sodium bicarbonate and 161 mg/l calcium chloride). The ACK consisted of 8.26 g ammonium chloride, 1 g potassium bicarbonate, and g EDTA per 1 L double-distilled water (ph 7.4). Red blood cell lysis was performed at room temperature. DHR 123 was obtained from Invitrogen (now Life Technologies). fmlp, PMA, and LPS were obtained from Sigma-Aldrich. Whole blood from volunteers was diluted (1:4) with KRBC for 5 minutes in the CO 2 incubator. Preliminary experiments indicated that this buffer produced the maximum up-regulation of a2v following stimulation. Then, 100 ml of whole blood was stimulated with fmlp (1 3 mm) for 3 to 5 minutes at 37 C. After 5 min, the whole blood was washed with PBS by centrifugation, and then stained with a predetermined amount of anti-a2v [2C1 conjugated with Alexafluor-647 (AF647)] at room temperature for 1 h in 1% fetal calf serum-pbs. After 2 washes with PBS, the red blood cells were lysed with ACK, and the neutrophils were analyzed. Surface expression of a2v or other proteins was determined by flow cytometry with a BD LSR II flow cytometer. Dose-response curves were used to determine optimal conditions for surface expression of a2v with other neutrophil stimulators. Maximum expression was at 5 min with 3 mm fmlp, 30 min with 40 ng PMA, and 30 min with 50 ng LPS. In some experiments, CYTD was added to blood for actin depolymerization and to enhance exocytosis before stimulation with fmlp or PMA. In those experiments, whole blood was diluted 1:4 in KRBC buffer (ph 7.2) and treated with 4 mg of CYTD for 5 min at 37 C in the CO 2 incubator. The microbial peptide fmlp was then added (3 mm) for an additional 5 min. The blood was then processed as described above. Cells were stained with mouse monoclonal anti-a2v-af647 and analyzed by flow cytometric analysis. Red blood cells were lysed with ACK buffer, and DAPI was added immediately before running the sample to ensure exclusion of dead cells. Flow cytometric analysis To stain freshly isolated neutrophils, 100 ml of cells/tube were incubated with antibody or isotype control (mouse IgG 1 ) and/or other antibodies in staining buffer (1% FCS/PBS) for 1 h at room temperature. All experiments were performed in triplicate and, in some cases, were performed 5 to 9 times using freshly drawn blood and/or isolated neutrophils. Detection was performed using an LSR II flow cytometer (BD). Becton-Dickenson FACS Diva software or FlowJo software (Tree Star, Ashland, OR, USA) was used for data analysis. Respiratory burst The formation of ROS following the functional respiratory burst activity was determined with a DHR assay and flow cytometry (Life Technologies). In brief, whole blood was diluted in KRBC (1:4) containing calcium and bicarbonate and then incubated with 1 mm DHR for 5 to 15 min at 37 C in aco 2 incubator. To stimulate the neutrophils and induce ROS, either fmlp or PMA was added for 5, 10, or 15 min at 37 C in a CO 2 incubator. Red blood cells were then lysed with ACK buffer. Cells were analyzed at regular intervals of 5 min each with a BD LSR II flow cytometer to measure the increase in green fluorescence (DHR). The mean channel of fluorescence intensity (MCI) was measured on a log scale to determine activity. In some experiments, the stimulated neutrophils were also stained with anti-a2v-af647 after ROS formation [28] Journal of Leukocyte Biology Volume 97, June

3 Gilman-Sachs et al. a2 Vacuolar-ATPase and neutrophil exocytosis Fluorescent microscopy Neutrophils were stimulated in polystyrene tubes as described above with fmlp or LPS and then fixed with 2% paraformaldehyde for 15 min at room temperature, washed, and permeabilized with 0.05% Triton-X for 30 s. The cells were then washed, blocked with 3% BSA in PBS for 30 min at room temperature and then stained with the appropriate antibodies. After staining, the cells were then washed with PBS, cytospun onto polylysine-coated slides using a Shandon cytocentrifuge, followed immediately by mounting in Prolong Gold antifade reagent with DAPI (Life Technologies). In some experiments, cells were fixed but not permeabilized before they were stained with the antibodies. Neutrophils were identified based on the multilobed nuclear staining pattern. Cells were imaged with a Eclipse 80i fluorescence microscope (Nikon, Chiyoda, Tokyo, Japan) with the Metamorph digital imaging system (Molecular Devices, Sunnyvale, CA, USA). Signal-intensity line-scan analysis was performed on regions of cells where colocalization appeared to occur using the Linescan Metamorph digital system. For confocal microscopy, the same protocol was followed, except 1.5 mm cover slips were used. The slides were imaged with an Olympus 300 with Fluoview software (Olympus, Shinjuku, Tokyo, Japan). Experiments were performed at least 3 times. Intracellular ph measurements To determine the intracellular ph, neutrophils were isolated by dextran sedimentation, and the cells were resuspended in PBS (ph 7.2) at cells/ml. The cells were then incubated with SNARF-1 acetoxymethyl ester (AM; 5 mm) (Life Technologies) for 1 h at 37 C in the CO 2 incubator [29]. After the cells were washed once with PBS (ph 7.2), the SNARFcontaining neutrophils were resuspended in KRBC (ph 7.2) for 5 min at 37 C and then activated with fmlp for 5 min. To measure the change in ph with the LSR II flow cytometer, a gate was placed around the neutrophils; the cells were excited at 488 nm, and the ratio of the mean fluorescence intensity at 640:580 was obtained over time. In some experiments, the neutrophils were preincubated with bafilomycin (25, 50, or 100 nm for 30 min at 37 in the CO 2 incubator) before activating with fmlp. A standard curve for ph was obtained by suspending SNARF-loaded neutrophils in HEPES buffer at ph 6.8, 7.1, 7.4, or 7.8 for 1 h and then measuring the ph ratios by flow cytometry. Determination of intragranular ph LysoSensor probes (Life Technologies) are cell-permeant ratiometric dyes that can be used to determine the ph of acidic organelles, such as lysosomes. Three dyes were used in these experiments, DND-160, DND-189, and DND The pka of LysoSensor yellow/blue DND-160 is ;4.2; it produces blue fluorescence in neutral environments (ph.4.2), but changes to yellow fluorescence in more acidic granules. The pka of LysoSensor Green DND-189 is ;5.2, and it becomes bright green in acidic granules. The DND-153 has apka of 7.5 and is bright green when granular ph is,7.5 and becomes dim above that ph. Isolated neutrophils were incubated with each LysoSensor probe (final concentration, 1 mm) in PBS buffer (ph 7.2) for 30 min at 37 C in a CO 2 incubator, at /ml in mm tubes. The cells were then stimulated with either fmlp or PMA at /ml. Cells were imaged with a Nikon fluorescence microscope using a Metamorph digital-imaging system. Excitation was at 364 nm. Emission was collected at either nm (blue) or nm (green-yellow) over time. Images were shown as overlays of blue and green-yellow emission images [30] for DND-160. For DND-189 or DND-153, images were collected at nm. Bafilomycin in PBS buffer (ph 7.2) was added to the cells in test tubes for 20 min in CO 2 before stimulation to inhibit the V-ATPase activity. Cells were then transferred to slides and imaged after stimulation with agonists. Control was DMSO vehicle. Calcium influx Freshly isolated neutrophils were incubated with 4 mm Fluo-3 (Life Technologies) for 30 min at room temperature, washed, and then resuspended in PBS at cells/ml [31]. Cells were stimulated with fmlp (3 mm) and immediately analyzed with the LSRII flow cytometer. The mean fluorescent intensity (MFI) at 525 nm over time was obtained to determine the calcium influx. In some experiments, neutrophils were preincubated with bafilomycin before calcium influx was measured. Quantitation of secretion of MPO and elastase into supernatants upon stimulation with fmlp Neutrophils were isolated by dextran sedimentation and Ficoll-Hypaque sedimentation and stimulated with fmlp or PMA as described above. Supernatant fluid was collected after 5 min or 30 min, respectively, and then analyzed for MPO or elastase in the presence or absence of bafilomycin using a Quantikine ELISA kit for human MPO (Cayman Chemicals, Ann Arbor, MI, USA; or R&D Systems, Minneapolis, MN, USA) or an ELISA kit for human elastase protein (Hycult Biotech, Uden, Noord-Brabant, The Netherlands). A standard curve was used to measure the secretion levels of each enzyme in picomoles per liter for MPO and nanograms per milliliter for elastase. Statistics For comparison of 2 sample means, the paired Student s t tests were used. RESULTS Localization, expression, and exocytosis of a2v in neutrophil granules by microscopy Initially experiments were performed to determine which of the 4 subsets of neutrophil granules contained the a2 isoform of the V-ATPase (a2v). Freshly isolated neutrophils were permeabilized and costained with an antibody to a2v (2C1) and antibodies to either membrane or matrix biomarkers, respectively. Antibodies to CD63 or MPO were used to identify primary granules, antibodies to CD66b or lactoferrin for secondary granules, antibodies to CD11b or for tertiary granules or antibodies to CD35 for secretory granules (Fig. 1). By confocal microscopy, a2v appeared to be associated with the primary granules that expressed membrane CD63 and, to a lesser extent, with secondary, tertiary, and secretory granules; however, the criteria for colocalization suggested the microscopy pattern did not indicate colocalization but rather coassociation (Fig. 1A). The matrix protein MPO also appeared to be in primary granules with a2v. When the expression of a2v with membrane or matrix proteins in secondary or tertiary granules was analyzed by microscopy, there were some regions of coassociation but to a lesser extent. These results are consistent with the location of MPO, lactoferrin, and MMP-9 inside the granule and not on the granule membrane (Fig. 1B). Thus, a2v was most highly expressed on the primary granule membrane or within the matrix, less in secondary granules, and even less in tertiary granules. Fluorescent microscopy experiments with anti-cd35 (secretory granules) and anti-a2v also suggested some coassociation with membrane proteins. Matrix proteins were not assessed for secretory granules. Magnification of the micrograph suggested that CD63 and a2v were found in primary granules close to one another but did not actually colocalize (Fig. 1C). Conditions for expression of a2v on neutrophil surface upon stimulation with fmlp, PMA, or LPS By fluorescent microscopy, nonpermeabilized neutrophils that had not been activated with fmlp (resting neutrophils) did not show any surface staining with anti-a2v but did show expression of a2v in the intracellular granules of permeabilized cells Volume 97, June 2015 Journal of Leukocyte Biology 1123

4 Figure 1. Localization of the a2 isoform of V-ATPase (a2v) with neutrophil granules. Representative photomicrographs of permeabilized neutrophils stained with anti-a2v and antibodies to granule markers (n = 5). (A) Freshly isolated neutrophils were permeabilized and then stained with antibodies to membrane proteins of primary (CD63), secondary (CD11b), tertiary (CD66b), or secretory (CD35) granules and antibody to a2v. Cells were visualized using either confocal or Nikon fluorescent microscopes. (B) Localization of a2v with matrix proteins in primary (MPO), secondary (lactoferrin), or tertiary (MMP) granules of permeabilized neutrophils. Enlargement of the indicated areas is shown and obtained using Metamorph software to analyze colocalization of a2v with the corresponding granule markers. (C) a2v and granule proteins exhibit a typical cupping pattern; a2v-atpase and the neutrophil markers display a close association with each other in the granules. Original magnification (Fig. 2A), surface and intracellular). When the neutrophils were stimulated with fmlp, a2v was found on the surface of those cells; in contrast, expression of the a2v intracellular granules decreased or disappeared (Fig. 2B, surface and intracellular). These results suggest that neutrophils do not express a2v on their surface but that fmlp activation of neutrophils did induce the exocytosis of a2v-containing granules from the inside of the cell to the cell surface. Similar results were obtained in neutrophils that had been stimulated with LPS (Fig. 2C and D). Flow cytometric analysis also confirmed the above finding that a2v was not detected on the surface of unstimulated neutrophils stained with the monoclonal antibody to a2v (Fig. 3A). However, upon stimulation with fmlp (3 mm) for 5 min, a2v was found on the surface of approximately 15 to 55% of the neutrophils from various donors (n = 5). Similar percentages were found after activation with LPS (50 ng) for 30 min (40 60%) or 100 ng PMA (30 70%) for 5 min (Fig. 3A). All experiments were performed in KRBC (ph 7.2). To determine whether incubation in KRBC buffer alone was sufficient to bring a2v to the surface of the neutrophils, resting neutrophils were primed for 5 min in KRBC buffer, but a2v was not detected on the surface of the neutrophils (not shown). Therefore, it appeared that most, but not all, neutrophils up-regulated a2v to their surface upon activation with these commonly used neutrophil stimulators but not when primed with KRBC buffer in the absence of common neutrophil stimulators. a2v translocation to the surface is associated with ROS production To determine whether translocation of a2v to the cell surface could be related to the respiratory burst, DHR was used to measure the production of ROS. DHR-containing neutrophils were stimulated with fmlp or PMA, and an increase in ROS was observed by flow cytometry analysis. Two-color flow cytometric analysis was used to correlate the formation of ROS (DHR) with the surface expression of the a2v following fmlp or PMA Figure 2. Trafficking of a2v in neutrophils stimulated with either fmlp (top) or LPS (bottom) by microscopic analysis. Representative photomicrographs (n = 5). (A) Neutrophils were incubated with vehicle (no treatment) or for 5 min with fmlp (B), then either not permeabilized (surface) or permeabilized (intracellular), and stained with anti-a2v (2C1)-AF488. Microscopy indicates the intracellular location of a2v granules before treatment (A and C) and trafficking to the cell surface after stimulation with fmlp (B) or LPS (D) Journal of Leukocyte Biology Volume 97, June

5 Gilman-Sachs et al. a2 Vacuolar-ATPase and neutrophil exocytosis Figure 3. Stimulation of neutrophils induces trafficking of a2v to the surface with neutrophil granules. (A) Neutrophils were stimulated with fmlp, LPS or PMA, stained with anti-a2v (2C1-AF647), fixed, but not permeabilized. The expression of a2v on the plasma membrane surface was analyzed by flow cytometric analysis. Representative flow cytometric histograms indicate V0a2 was detected on the surface of the stimulated neutrophils (open histograms, treatment) compared to unstimulated neutrophils (filled histograms, zero time). DAPI was added to each tube, and only viable cells were analyzed (n = 9). (B) ROS generation by fmlp or PMA also induces trafficking of a2v to the surface of neutrophils. DHR 123 was used to measure ROS production by flow cytometry. Two-parameter histograms were obtained for DHR (x axis) versus a2v (y axis) (n =3). (C) Coexpression of a2v and CD63 on the surface of neutrophils activated with fmlp or the control vehicle, washed but not permeabilized, and stained with FITCanti-CD63 and biotinylated anti-a2v and then Streptavidin 594 (n = 3). (D) Representative flow cytometric histograms indicating coexpression of CD63, CD11b, or CD66b (primary, secondary, or tertiary granule biomarkers, respectively) with a2v on the surface of fmlp-stimulated neutrophils. Neutrophils were stimulated with fmlp for 5 min, then stained with anti-cd63 (primary granules), anti-cd11b (secondary granules), or anti-cd66b (tertiary granules) but were not permeabilized. Two-parameter histograms of untreated (top histograms) or treated (lower histograms) were obtained by flow cytometric analysis (n = 3). stimulation. An increase in the a2v was observed on the surface of the stimulated neutrophils producing ROS (Fig. 3B). This finding suggested activation of the neutrophils caused a2v- ATPase to move from an intracellular location to the plasma membrane, which was associated with ROS production. Exocytosis of a2v containing granules upon neutrophil activation To determine whether a2v and CD63, a primary granule membrane protein, were coexpressed on the surface of the neutrophils following stimulation, neutrophils were incubated with fmlp but not permeabilized and were then analyzed by fluorescent microscopy. The stimulator fmlp caused the coexpression of both CD63 and a2v onto the surface of these cells; in contrast, coexpression was not found on the surface of unstimulated control cells (Fig. 3C). In other experiments using flow cytometric analysis, 2-parameter flow cytometry indicated that, upon stimulation with fmlp, a2v and CD63 were highly upregulated and coexpressed on the cell surface. Similar findings showed the coexpression of a2v and CD66b or CD11b (Fig. 3D). These findings suggest that exocytosis of granules led to the translocation of a2v to the plasma membrane. This effect was more pronounced when CD63 (biomarker for primary granules) was used than it was with C66b or CD11b (biomarkers for secondary and tertiary granules). Translocation of a2v-containing granule to the cell surface is dependent on actin depolymerization and correlates with calcium influx and alkaline ph It has been shown that disruption of actin can enhance exocytosis of granules in neutrophils [32]. To show that translocation of a2v + granules is dependent on actin depolymerization, neutrophils were treated with cytochalasin D before stimulation with fmlp. The percentages of a2v on the surface of the stimulated neutrophils, as determined by flow cytometry, were greatly enhanced; the addition of cytochalasin D resulted in an almost 2 3 times increase in the percentages of a2v beyond that found in stimulated neutrophils that were not pretreated with cytochalasin D (Fig. 4A). A primary event following stimulation of neutrophils is a change in cytoplasmic calcium concentration. Calcium influx following fmlp activation was, therefore, measured using an intracellular Fluo-3 AM indicator and flow cytometry. Calcium influx was found in neutrophils stimulated with fmlp within seconds. Interestingly, bafilomycin alone seemed to enhance the calcium influx (Fig. 4B). In contrast, when neutrophils were incubated with bafilomycin alone, a2v was not translocated to the surface of the cell; however, stimulation of neutrophils pretreated with bafilomycin and then stimulated with fmlp caused more a2v to traffic to the surface than did stimulation with fmlp alone (Fig. 4C). Incubation of mast cells in ammonium chloride buffer (ph 7.2), a buffer that causes the intracellular ph to become alkaline, can induce exocytosis of granules [33]. To determine whether ammonium chloride buffer (ph 7.4) could induce translocation of a2v to the surface of the neutrophils, whole blood was incubated in NH 4 Cl buffer (ph 7.4) (ACK) to lyse the red blood cells and then stained with anti-a2v; translocation of a2v to the cell surface was observed (Fig. 4D, left). No stimulators of neutrophils were added, but the exposure of the neutrophils to the ACK alkaline buffer was sufficient to cause up-regulation of a2v to the surface. In contrast, when lysis of the whole blood with NH 4 Cl buffer was performed after staining with anti-a2v, no such translocation occurred (Fig. 4D, right). The cells were viable following these manipulations because DAPI was used to gate only with live cells by flow cytometry. These findings suggest that an alkaline intracellular ph can induce translocation of a2v + granules to the surface of neutrophils. Volume 97, June 2015 Journal of Leukocyte Biology 1125

6 Figure 4. Actin depolymerization, calcium influx, bafilomycin, and extracellular alkaline ph affects trafficking of a2v to the neutrophil cell surface. (A) Cytochalasin enhances the trafficking of a2v in fmlp-activated neutrophils, supporting the role of actin depolymerization in exocytosis of granules. Neutrophils revealed enhanced trafficking of a2v to the surface when cytochalasin was added before fmlp stimulation (n = 3; error bars = means 6 SEM; *significant difference between fmlp and fmlp treatment with cytochalasin, P # 0.05). (B) Calcium influx increases in neutrophils after treatment with bafilomycin, fmlp, or bafilomycin and fmlp together. Calcium influx was measured using Fluo-3 AM by flow cytometric analysis. Following excitation at 488 nm, emission at 525 nm was collected for 3 min. (C) Enhanced translocation of a2v to the surface of neutrophils upon activation with fmlp in the presence of bafilomycin. (D) Incubation of neutrophils in ACK (ph 7.4) induces translocation of a2v to the surface of resting neutrophils. Representative flowcytometric histograms comparing the expression of surface a2v on resting neutrophils (zero time) suspended in ACK (ph 7.4) for 5 min before (left) staining with AF647-anti-a2V or after (right) staining with AF647-anti-a2V (y axis) (n = 3). BAF, bafilomycin; FSC, forward light scatter. Effect of priming extracellular, intracellular, and intragranular ph on translocation of a2v to the cell surface V-ATPases may affect extracellular, intracellular, and intragranular ph [34].To determine whether an acidic extracellular environment could induce translocation of intracellular a2v to the surface, isolated neutrophils were primed in KRBC buffer at 5.5, 6.2, or ph 7.2 for 0, 1, or 2 h at room temperature or at 37 C. Expression of a2v was not observed on the surface of these neutrophils by flow cytometric analysis regardless of the extracellular ph in which they had been primed ranging from an acidic to neutral ph (not shown). In addition, neutrophils were activated with fmlp (3 mm) for 5 min after incubation in ph 7.2 or 6.2 buffer for 1 hour. After washing away the fmlp, the neutrophils were stained for the surface a2v and then analyzed by flow cytometry. Results indicated that a2v was found on the surface of the neutrophils that had been stimulated with fmlp under all these conditions (ranging from 31.4 to 49.4% of all the neutrophils), but the percentages were not different with either ph 7.2 or with other acidic ph buffers (not shown). These findings suggest that incubation in the acidic extracellular environment does not cause the translocation of a2v to the surface. Changes in the intracellular ph are known to occur following metabolic reactions inside cells. Changes in intracellular ph following stimulation with fmlp was determined in neutrophils preincubated with SNARF-1 for 30 min in PBS, primed with KRBC (ph 7.2) for 5 min and then activated with fmlp for 5 min. Flow cytometry was used to determine changes in the ratio at 640:580 nm over time and ratios were compared to the ratios in a standard curve to determine intracellular ph [35]. The intracellular ph in the untreated control neutrophils was between 7.0 and 7.2. The ph became more alkaline upon incubation of the neutrophils with fmlp and increased from 7.0 to 7.6 within 5 min (Fig. 5A). This increased ph correlates with our previous finding that a2v translocated to the surface of these cells following stimulation with fmlp for 5 min. The increase in ph indicates that, within 5 min after fmlp stimulation, the intracellular ph becomes alkaline rather than acidic. When increasing amounts of bafilomycin A1 (25, 50 and 100 nm) were added to unstimulated neutrophils, the intracellular ph ranged from 7.1 to 6.8, slightly less than unstimulated control neutrophils. When fmlp was added to neutrophils previously incubated with bafilomycin, the ph ranged from 7.4 to 7.8, suggesting that greater concentrations of bafilomycin (100 nm) did not affect the alkaline ph observed following fmlp activation (Fig. 5A). Therefore, apparently an alkaline intracellular ph, but not an acidic extracellular ph, correlated with the translocation of a2v to the plasma membrane of neutrophils. To determine changes in the intragranular ph of neutrophils upon stimulation, LysoSensor dyes were used. Neutrophils were first incubated with LysoSensor DND-160 (pka ;4.2) for 30 min and were then stimulated with PMA. Overlays of images of individual neutrophils were obtained by fluorescent microscopy at wavelengths of 540 nm (green) and 384 nm (blue). Merged images indicated the cells appeared blue at first (5 min); however, at 10 min, the granules were a yellow-green color (indicating the granules were becoming more acidic). After 25 min, the granules became blue again (Fig. 5B). These findings suggest that the granules had an initial ph.4.2, became more acidic after stimulation with PMA at 10 min, and then became alkaline at 25 min (ph.4.2). Prior incubation of these cells with bafilomycin inhibited ph changes following PMA stimulation; the granules in the cell remained blue, indicating that V-ATPases were necessary to maintain the intragranular ph of these granules (Fig. 5B). Similar results were found when fmlp was used to stimulate the cells (Fig. 5C). In addition, in this experiment, anti-a2v antibody was added to the live cell suspension to trace the granules that were exocytosed. At 10 min, a slight amount of translocation of a2v (red) to the surface was evident. However, after min, merged images of yellow/blue DND Journal of Leukocyte Biology Volume 97, June

7 Gilman-Sachs et al. a2 Vacuolar-ATPase and neutrophil exocytosis Figure 5. Effect of stimulation on changes in intracellular or intragranular ph of neutrophils. (A) Intracellular ph becomes alkaline following fmlp activation for 5 min. Neutrophils containing SNARF-1 AM were stimulated with fmlp and the cytosolic ph was determined by flow cytometric analysis using a standard curve as described in Materials and Methods (n = 3; error bars = means 6 SEM; *statistically significant difference between control and fmlp treatment, P # 0.05). (B D) Representative photomicrographs of the determination of intragranular ph in neutrophils using LysoSensor dyes with different pka values. Purified neutrophils containing LysoSensor dyes were stimulated with PMA (B) or fmlp (C and D) to determine changes in ph within granules (intragranular ph). (B) Neutrophils, loaded with LysoSensor yellow/blue DND-160 (pka 4.2) and stimulated with PMA, were photographed by microscopy for 25 min at room temperature, and blue or yellow fluorescence was collected for each cell with a Nikon Eclipse 80i fluorescent microscopy; overlays of the blue and yellow fluorescence of each image are shown (n = 10). The zero point images (the first obtained by microscopy) were blue, similar to the image shown at 10 min. Neutrophil granules become more acidic (green) in a time-dependent manner (minutes indicated) after stimulation with PMA (B) or fmlp (C). In the presence of bafilomycin, the granules do not appear acidic (B, bottom). Note the granular appearance of the dye, consistent with accumulation of the dye within the acidic granules of the cell. Original magnification of 380. (C) Colocalization of a2v and acidic granules on the surface after activation with fmlp for 40 min. Neutrophils were incubated with LysoSensor DND-160 for 30 min, then activated with fmlp, and incubated with anti-a2v-af647. The mixture was then transferred to a slide and observed by microscopy. The movement of the a2v-containing granules, which was associated with the decrease in ph, was imaged using a Nikon Eclipse 80i microscope. The granules that contain DND-160 (pka 4.2) and a2v (2C1-AF594) appear to move toward the cell surface. (D) Photomicrographs of neutrophils incubated with a different LysoSensor dye, DND-189 (pka 5.2) and treated with either fmlp, bafilomycin (Baf), or fmlp and Baf. The zero point images were similar to the pale green images shown when the LysoSensor DND-189-loaded neutrophils were incubated with bafilomycin alone (center image). granules (now green) and a2v (red) indicated a2v translocated to the surface, as evidenced by the presence of yellow granules on the surface of the neutrophil (Fig. 5C). Resting neutrophils were also incubated with LysoSensor green DND-189 (pka ;5.2); this dye partitions into acidic organelles and appears dim green in organelles with a ph.5.2 but becomes bright green when the organelles ph becomes more acidic. As a control, microscopy indicated the granules were nonfluorescent (dim green) in the presence or absence of bafilomycin, indicating the intragranular ph was.5.2 (Fig. 5D). In contrast, stimulation of these neutrophils by fmlp caused the granules to turn bright green within a few minutes, suggesting that fmlp stimulation caused the intragranular ph to become more acidic (Fig. 5D). a2v has a role in maintaining the ph in neutrophil granules. In data not shown, a third LysoSensor dye, DND-153, indicated that the ph of the granules was,7.5, but in Volume 97, June 2015 Journal of Leukocyte Biology 1127

8 the presence of bafilomycin the ph of the granules was more alkaline. These results are consistent with the finding that the intragranular ph is between 5.2 and 7.5 before stimulation, ph,4.2 after stimulation, and becomes alkaline when bafilomycin inhibits the V-ATPase. Exocytosis of matrix and membrane proteins Experiments were performed to determine the role of V-ATPase in exocytosis of matrix proteins (MPO or elastase) present in the a2v + primary granules. Neutrophils were isolated and stimulated with PMA in the presence or absence of bafilomycin. When the supernatant fluid was assayed for MPO (MPO), the matrix protein was found secreted in the supernatant fluid, as expected (Fig. 6A). Bafilomycin inhibited the secretion of some, but not all, of the MPO. Similar findings were seen for elastase (Fig. 6B). In similar experiments, fmlp was used to stimulate neutrophils, which were then analyzed by fluorescent microscopy to determine how fmlp affected the translocation of primary granules containing a2v and MPO (Fig. 6C). The MPO was detected within primary granules of unstimulated neutrophils (Fig. 6C, left) but was barely detected on the surface of fmlpstimulated neutrophils (Fig. 6C, middle). In contrast, a2v was found associated with MPO in granules before stimulation (Fig. 6C, left) but was found on the surface (Fig. 6C, right) of stimulated neutrophils. When neutrophils were pretreated with bafilomycin and then stimulated with fmlp, granules containing MPO alone or both MPO and a2v were found on the surface. These findings are consistent with the suggestion that the a2vcontaining granules were exocytosed to the cell surface upon stimulation with fmlp, but bafilomycin disrupts the proton gradient induced by fmlp stimulation, and some, but not all, of the granules remain on the surface of the neutrophil plasma membrane; this result accounts for the ELISA data, which indicated that fmlp in bafilomycin-treated neutrophils released less MPO to the extracellular environment than did fmlp alone. The activity of MPO from activated neutrophils was also slightly decreased in the presence of bafilomycin, suggesting that a2v V- ATPase has a role in the secretion and functional activity of this enzyme (Fig. 6D). DISCUSSION Our studies were aimed at determining the association of a2v, which is 1 of 4 isoforms of the V0a subunit of the V-ATPase complex, with exocytosis of antibacterial granules in neutrophils. The a2v molecule is not found on the surface of resting neutrophils; however, stimulation of neutrophils with fmlp, PMA, or LPS induced the translocation of intracellular a2v to the surface because of the trafficking of primary granules, which have a2v and CD63 on their membranes and, to a lesser extent, trafficking of a2v expressing secondary or tertiary granules to the surface. Extracellular buffers with a slightly acidic ph (5.5 or 6.2) did not induce trafficking of a2v to the surface, whereas extracellular buffers with a slightly alkaline ph (ammonium chloride, ph 7.4) did. The intracellular ph became alkaline (changing from ph 7.2 to 7.8 within 5 min) after stimulation with fmlp, and the intragranular ph of acidic granules in resting neutrophils, which was between ph 5.2 and 7.5, decreased to a ph of,4.2 upon stimulation with fmlp or PMA. These changes in ph correlate with the translocation of acidic granules containing a2v and CD63 to the surface and are consistent with studies that suggest a ph gradient induces the exocytosis of secretory granules in various types of cells [36 38]. In neutrophils, it appears that the a2v isoform of V-ATPase is responsible for this function. Figure 6. Inhibition of a2v activity by bafilomycin affects secretion of myeloperoxidase or elastase from stimulated neutrophils but does not prevent up-regulation of a2v to the surface. (A) Neutrophils were pretreated with bafilomycin (Baf) for 30 min, and then, PMA was added for 30 min. Supernatant fluids were collected, and ELISA was used to measure the concentration of MPO (A) or elastase (B) in these supernatant fluids; these experiments were performed 3 times (n = 3; error bars = means 6 SEM). MPO and elastase proteins were decreased when neutrophils were pretreated with Baf and then stimulated with PMA; *P # 0.05; statistically significant difference between treatment with PMA and treatment with PMA and bafilomycin. (C) Microscopic analysis of intracellular and surface expression of a2v (green) and MPO (red) in activated neutrophils preincubated in the presence or absence of Baf and then activated with fmlp. V0a2V remains on the neutrophil membrane after fmlp stimulation, but MPO is not detected on the membrane; however, pretreatment with Baf followed by stimulation with fmlp caused partial exocytosis, and granules with MPO and/or a2v are found on the surface (n = 3). (D) MPO enzyme activity assay (Cayman) of fmlp-stimulated neutrophils indicates a slight reduction in enzyme activity at higher intracellular ph, i.e., when neutrophils were pretreated with Baf before stimulation with fmlp (n = 3; error bars = means 6 SEM) Journal of Leukocyte Biology Volume 97, June

9 Gilman-Sachs et al. a2 Vacuolar-ATPase and neutrophil exocytosis These experiments were initially designed to determine conditions under which a2v can be induced to move to the surface of neutrophils. In our study, extracellular ATP did not induce up-regulation of a2v to the surface (not shown). When we analyzed the exocytosis of primary granule components, we found that fmlp or PMA induced the trafficking of a2v/cd63 + primary granules to the surface of activated neutrophils. However, we did not detect MPO (a matrix protein) on the surface [39] but found it secreted into the external environment. We also found that PMA or fmlp stimulation in the presence of bafilomycin did not prevent the trafficking of a2v to the surface but decreased the secretion of MPO and elastase into the external milieu and inhibited the complete exocytosis of all primary granules. Our findings are consistent with the suggestion that a2v, the V-ATPase membrane protein of primary granules (and probably other neutrophil granules), traffics to the surface upon stimulation with fmlp, fuses with the plasma membrane, and enhances the secretion of MPO or elastase outside of the cell but is not itself secreted because it remains in the neutrophil cell membrane [40]. Understanding the conditions under which the V-ATPase is regulated may have important clinical implications. There are 4 isoforms of the V0a subunit, and each isoform is thought to have a role in the trafficking of the V-ATPase to different organelles in various types of cells as well as in the regulatory uncoupling of V0 and V1 and in membrane fusion [9, 10, 41, 42]. The a1v isoform is involved in acidification of endomembrane organelles, especially in the microglia and other secretory cells of the brain and is prevalent in secretory granules of neurons or other cells in the brain; it has a role in the exocytosis of neurotransmitters in those granules [36, 43]. The a2v isoform is found in vesicles in the thymus, decidua, hematopoietic cells, dividing cells, epididymal clear cells [11, 34, 44, 45], and on the plasma membranes of macrophages [18, 20]. The a3v isoform is present in liver, lung, heart, brain, spleen, and kidney [34]. It is also found in the plasma membrane of osteoclasts, and mutations in the gene encoding the a3v isoform TCIRG1a cause the disease osteopetrosis [34, 45]. The a4v isoform is expressed in the plasma membrane of kidney renal intercalated cells, and certain mutations cause renal tubule acidosis [46]. The a4v isoform is not as well characterized as the other isoforms [34]. In recent years, other studies [17, 42, 47] have indicated that vacuolar ATPase has a role in the growth and spread of cancer cells. How the 4 isoforms are differentially regulated is not well-known. Neutrophil granules can be characterized by specific proteins in the membrane or the matrix [48]. Primary granules contain lytic enzymes in the matrix, such as MPO or elastase and membrane CD63 (tetraspanin) [49, 50]. Specific or secondary granules are rich in antimicrobial peptides, such as lactoferrin [51], and express CD66b on the membranes. Gelatinase or tertiary granules are more easily exocytosed and are the main reservoir of tissue-degrading enzymes, such as MMP-9; these granules express membrane CD11b. Secretory granules differ from the other 3 granules in that they are formed by endocytosis and contain plasma proteins and membrane-associated receptors, such as CD35 (CR1), proteinase 3, or CD11b/CD18 (CR3) [52, 53]. Secretory granules are most easily mobilized, followed by gelatinase granules, then specific granules, and finally, azurophilic granules [2, 52]. In these studies, a2v was coassociated most strongly with primary granules; therefore, we focused on that association in functional assays. In preliminary experiments, antibodies specific for the a1v and a3v isoforms were used to determine whether they were also in neutrophil granules. Although these isoforms were found in neutrophils, they were not clearly coexpressed with the biomarkers we used to detect primary, secondary, or tertiary granules. This finding suggests that the regulation of intragranular ph within these granules is due to the a2v-v-atpase enzyme complex. The role of the other isoforms needs to be further studied. V-ATPases have been identified in neutrophil granules [6, 7, 51]. A proteomic analysis of genes in each of the 3 major subsets of neutrophil granules [7, 25] identified proteins for ATP6V1C2, ATP6V0A1, and ATP6V0A2 (a2v), which are subunits of the V-ATPases in gelatinase granules, and proteins for ATP6V0D1 or TCIRG1 in gelatinase granules, specific granules [51], and azurophilic granules. Although we found that activation of neutrophils with known stimulators, such as fmlp, PMA, and LPS, causes the a2v-atpase on primary granules to move from inside the cell to the plasma membrane, the expression of a2v was not found to occur on 100% of all neutrophils by flow cytometric analysis or microscopy. This finding may be consistent with the heterogeneity suggested to occur during the differentiation of neutrophils and the formation of the different subsets of granules [51]. Heterogeneity occurs because granules contain proteins that are synthesized at different stages of development [2, 51, 52]. Subsets of granules contained within each cell may have different amounts of a2v, which vary in their ability to be up-regulated. We found the percentage of neutrophils expressing a2v on their surface varied per individual but was always,100%. The a2v surface-negative neutrophils may have very low levels of a2v on the surface, or it may be that those neutrophils do not exocytose all of the primary granules containing a2v [54]. Calcium influx and actin depolymerization either correlated with or enhanced the trafficking of a2v. In the presence of cytochalasin more a2v was up-regulated than in the absence of cytochalasin, suggesting depolymerization was important for the exocytosis of a2 + granules. The actin cytoskeleton regulates exocytosis [8, 32], and the amount of actin associated with each type of granule varies inversely to the rate of mobilization or exocytosis [32, 55]. Johnson et al. [55] suggested there was an inverse correlation between actin polymerization and exocytosis. They showed that Rab27a and its effector JCF1 colocalize in the primary granule membrane and have a role in actin remodeling. However, the role of ph in this process has not been established. Exocytosis of the granule contents is a consequence of the fusion of granule-subset membrane with the cell-surface plasma membrane and subsequent incorporation of the granule membrane into the plasma membrane [40]. In many cells, this membrane is rapidly recycled for reuse (adipocytes, neurons), but in neutrophils, the membrane of mobilized granules largely remains part of the plasma membrane, resulting in new receptors and other proteins necessary for neutrophil functions, such as migration to the tissues or expression of FcR [52]. The V0 domain of the V-ATPase is directly implicated in secretory-vesicle exocytosis by mediating membrane fusion of the Volume 97, June 2015 Journal of Leukocyte Biology 1129