Apilimod

The Lipid Kinase PIKfyve Coordinates the Neutrophil Immune Response through the Activation of the Rac GTPase

Neutrophils rapidly arrive at an infection site because of their unparalleled chemotactic ability, after which they unleash numerous attacks on pathogens through degranulation and reactive oxygen species (ROS) production, as well as by phagocytosis, which se- questers pathogens within phagosomes. Phagosomes then fuse with lysosomes and granules to kill the enclosed pathogens. A com- plex signaling network composed of kinases, GTPases, and lipids, such as phosphoinositides, helps to coordinate all of these processes. There are seven species of phosphoinositides that are interconverted by lipid kinases and phosphatases. PIKfyve is a lipid kinase that generates phosphatidylinositol-3,5-bisphosphate and, directly or indirectly, phosphatidylinositol-5-phosphate [PtdIns(5)P]. PIKfyve inactivation causes massive lysosome swelling, disrupts membrane recycling, and, in macrophages, blocks phagosome maturation. In this study, we explored for the first time, to our knowledge, the role of PIKfyve in human and mouse neutrophils. We show that PIKfyve inhibition in neutrophils does not affect granule morphology or degranulation, but it causes LAMP1+ lysosomes to engorge. Additionally, PIKfyve inactivation blocks phagosome–lysosome fusion in a manner that can be rescued, in part, with Ca2+ ionophores or agonists of TRPML1, a lysosomal Ca2+ channel. Strikingly, PIKfyve is necessary for chemotaxis, ROS production, and stimulation of the Rac GTPases, which control chemotaxis and ROS. This is consistent with observations in nonleukocytes that showed that PIKfyve and PtdIns(5)P control Rac and cell migration. Overall, we demonstrate that PIKfyve has a robust role in neutrophils and propose a model in which PIKfyve modulates phagosome maturation through phosphatidylinositol-3,5-bisphosphate–dependent activation of TRPML1, whereas chemotaxis and ROS are regulated by PtdIns (5)P-dependent activation of Rac.

Neutrophils are the first responders to an infection and, thus, play an essential role in coordinating the innate immune response (1–4). They do this because of theirunmatched chemotactic ability, sensing and tracking the chemical trail to sites of infection (2). Once in contact with pathogens, they unleash a variety of attacks, including degranulation to secrete cytokines, hydrolytic enzymes, and antibacterial peptides; acti- vation of the NADPH oxidase, to generate reactive oxygen species (ROS); and phagocytosis, to engulf and sequester the pathogens into phagosomes (1–3, 5). Phagosomes then mature by fusing with granules and lysosomes to kill and digest the pathogens (1, 6). All of these responses are coordinated through a variety of re- ceptors and intracellular signals, including small GTPases (7) and the phosphoinositide (PtdInsP) lipids (1, 3, 8). Based on the phos- phorylation pattern of the head group, there are seven species of PtdInsPs that typically function by differentially distributing to in- tracellular membranes and then recruiting a set of protein effectors specific to that PtdInsP species (9, 10). Through this general process, phosphatidylinositol-4,5-bisphosphate and phosphatidylinositol- 3,4,5-trisphosphate help to coordinate actin and membrane dy- namics to direct chemotaxis and phagocytosis (11–14). In com- parison, phosphatidylinositol-3-phosphate regulates endosomal membrane trafficking, phagosome maturation, and activation ofknowledge about the importance of phosphatidylinositol-5-phosphate [PtdIns(5)P] and phosphatidylinositol-3,5-bisphosphate [PtdIns(3,5)P2] in neutrophil function.PtdIns(3,5)P2 is synthesized by the lipid kinase PIKfyve by phos- phorylating phosphatidylinositol-3-phosphate (19–21). In contrast, controversy remains about the source of PtdIns(5)P (19, 22).

In one model, PIKfyve synthesizes PtdIns(5)P directly by phos- phorylating phosphatidylinositol (22, 23), whereas in another model, PtdIns(3,5)P2 is converted to PtdIns(5)P via action of the myotubularin lipid phosphatases (19, 24, 25). Regardless, loss of PIKfyve function causes multiple defects, including embryonic lethality in PIKfyve2/2 mice (26), swollen endolysosomes (27), hindered membrane recycling (28, 29), impaired lysosomal Ca2+ signaling (30), and defective autophagic flux (31, 32), attesting to its importance (20). Interestingly, PIKfyve has an emerging role in the immune response. For example, PIKfyve inhibition disrupts TLR and cytokine signaling; in fact, the PIKfyve inhibitor apili- mod was used to suppress IL-12/IL-13 signaling before it was discovered to be a selective inhibitor of PIKfyve (33, 34). In addition, mice carrying a platelet-specific PIKfyve2/2 genotype suffer from massive macrophage activation and inflammation (35). Lastly, in- hibition of PIKfyve blocks phagosome maturation in macrophages (36). This likely occurs because PtdIns(3,5)P2 is needed to activate TRPML1, a PtdIns(3,5)P2-gated lysosomal Ca2+ channel (30, 37). When TRPML1 is silenced or Ca2+ is chelated, phagosomes and lysosomes dock but fail to fuse (37). It remains possible that some of these defects are due to the concomitant loss of PtdIns(5)P when PIKfyve is inhibited. Indeed, PIKfyve and MTMR3, a myotubu- larin, are implicated in cell migration via the PtdIns(5)P-dependent activation of the Rac GTPase, a critical coordinator of actin remodeling (25, 38).Activation of the Rac GTPases is critically important for neu- trophil chemotaxis, phagocytosis, and stimulation of the NADPH oxidase (39–41).

Given all of this, we postulated that PIKfyve activity, through the synthesis of PtdIns(3,5)P2 and/or PtdIns(5)P, is essential for coordinating various neutrophil functions. Indeed, we found that PIKfyve inhibition blocks phagosome maturation, chemotaxis, and the NADPH oxidase but not degranulation. In part, PIKfyve regulates these functions by modulating the lyso- somal TRPML1 Ca2+ channel and activation of the Rac GTPases.Bone marrow from the femur and tibia of C57BL/6 mice was extracted by flushing the bones with complete DMEM plus 10% FBS using a 27-gauge needle. Cells were centrifuged at 1000 3 g for 5 min to yield a pellet that was then resuspended in 1 ml of complete DMEM and centrifuged at 1000 3 g for 30 min over a Percoll gradient containing 55, 65, and 80% Percoll. The band of neutrophils between 80 and 65% Percoll was collected, washed with PBS, and resuspended in complete DMEM. To plate cells, glass cover- slips were coated with 3% BSA at room temperature for 30 min, followed by a PBS wash. One million neutrophils were then plated by incubating for 30 min at 37˚C and at 5% CO2. Animal handling and treatment was done according to guidelines established by the Institutional Animal Care Ethics Board.Blood-derived human neutrophil isolation and stimulationSelect assays were performed with human neutrophils isolated by layering4.5 ml of citrate-buffered human peripheral blood onto 4 ml of 1-Step Polymorphs buffer in a 15- ml tube (Accurate Chemical & Scientific, Westbury, NY). This was spun for 35 min at 500 3 g at room temperature. Neutrophils were recovered from the bottom layer. RBCs were lysed with13 BD Pharm lyse buffer (BD Biosciences, ON, Canada).

The total number of neutrophils was diluted in 13 PBS and counted with a Beck- man Coulter Z2 cell counter. Isolation of human blood was done according to guidelines established by the Institutional Research Ethics Board.PIKfyve inhibition and vacuolationTo inhibit PIKfyve, neutrophils were treated with the indicated concen- trations of apilimod (Toronto Research Chemicals, Toronto, ON, Canada) or YM201636 (Adooq Bioscience, Irvine, CA) for the indicated periods of time. To quantify vacuolation, cells were imaged live in the continuous presence of the inhibitor for no more than 20 min using differential inter- ference contrast microscopy. Vacuoles were defined as being .1 mm. Al- ternatively, cells were stimulated and processed for live cell imaging or fixed.Phagocytosis and phagosome maturation assaysFor phagocytosis and phagosome maturation, we used polymer beads with a diameter of 2.08 mm (Bangs Laboratories, Fishers, IN) opsonized with human IgG, as described (37). Particles were added to neutrophils and synchronized by centrifugation of the cells for 5 min at 400 3 g. Subse- quently, cells were washed three times with PBS and incubated with complete DMEM at 37˚C and 5% CO2 for 15 min to allow internalization of the beads. For quantification of phagocytic index and efficiency, cells were fixed with 4% PFA for 20 min at room temperature, followed by quenching with 100 mM glycine for 20 min. To identify external versus internal beads, cells were incubated with fluorescently labeled goat anti- human Abs at 1:1000, as described (42).

For phagosome maturation, cells were incubated for a chase time of 1 h at 37˚C and 5% CO2 before staining external beads as above; this was followed by fixation and processing for immunofluorescence as described previously (43) and below.Stimulation of degranulationNeutrophils were plated and pretreated with 30 nM apilimod for 30 min at 37˚C prior to the addition of 1 mM latrunculin A (Abcam, Cambridge, MA) for 30 min. Degranulation was induced by adding 300 nM fMLF (Sigma-Aldrich, Oakville, ON, Canada) for 3 min, followed by fixation with 4% paraformaldehyde (PFA) for 20 min and quenching with 100 mM glycine. To assess degranulation for primary granules, cells were stained with anti-CD63 and anti-myeloperoxidase (MPO) Abs (see below). To assess degranulation of secondary and tertiary granules, cells were stained with anti-lactoferrin and MMP9, respectively (see below).ImmunofluorescenceAfter the required manipulation, neutrophils were washed and fixed with 4% PFA for 20 min, followed by three washes in PBS and quenching us- ing 100 mM glycine in PBS for 20 min. For intracellular immunostaining, cells were permeabilized with 0.5% Triton X-100 for 10 min at room tem- perature to stain with rabbit anti-mouse polyclonal Abs to MPO, MMP9, lactoferrin (all used at 1:100; Bioss, Boston, MA), or CD63 (H-193, 1:200; Santa Cruz Biotechnology, Paso Robles, CA). Alternatively, cells were permeabilized with 100% ice-cold methanol for 5 min to stain with rat anti- mouse LAMP1 mAbs (clone 1D4B, 1:100; Developmental Studies Hybridoma Bank, Iowa City, IA). Human neutrophils were fixed with 4% PFA and permeabilized for 3–5 min with ice-cold 100% methanol before staining with rabbit anti-human LAMP1 mAbs (1:200; Cell Signaling Technology, Danvers, MA). For extracellular staining, no permeabilization was performed before immunostaining. All primary Abs were incubated for 1 h at room temperature in 0.5% BSA, followed by three washes in PBS and a 1-h incubation with fluorescently labeled secondary Abs at 1:1000 (Jackson ImmunoResearch Laboratories, West Grove, PA).

Lastly, cells were washed every 5 min with 0.5% BSA for 30 min to remove the excess secondary Abs. The coverslips were mounted with Dako mounting media and visualized. For staining ruptured cells, after the desired treatment, neutrophils were washed three times with cold 13 PBS and scraped into 1 ml of cold ho- mogenization buffer (20 mM Tris [pH 7.4], 1 mM AEBSF, 1 mM MgCl2, 1 mM CaCl2, 1 mg/ml each RNase and DNase and supplemented with mammalian protease inhibitor mixture [Sigma-Aldrich]). Cells were then lysed by passaging six to eight times through a 25-gauge needle, followed by centrifugation for 10 min at 10,000 3 g at 4˚C to collect lysed cell bodies. The pellet was then resuspended and fixed in 400 ml of 4% PFA for20 min, followed by a 20-min incubation at room temperature with 100 mM glycine to quench PFA. Cells were then pelleted and stained with anti-LAMP1 and anti-CD63 Abs, as described above.For scoring vacuolation, live cell imaging was done using an inverted Olympus IX83 microscope (Olympus, Richmond Hill, ON, Canada) with a Hamamatsu ORCA-flash 4.0 digital camera. Cells plated on coverslips were placed in a chamber with HEPES-buffered RPMI 1640 medium at 37˚C and imaged using differential interference contrast microscopy.For fluorescence imaging, we used a Quorum spinning disk confocal microscope (Quorum, Guelph, ON, Canada) equipped with a Hamamatsu C9100-13 EM-CCD camera and a 1003 oil objective (NA 1.4) to obtain single-plane images. Images were analyzed using ImageJ (v. 1.47 bundled with 64-bit Java) and processed with Adobe Photoshop (v. 7.0.1; Adobe Systems, San Jose, CA) without altering the relative fluorescence intensity of the images. To quantify phagosome maturation, images were converted to 8-bit and into false color, as described (36, 37). Briefly, phagosome- associated LAMP1 fluorescent intensities were clustered into three groups based on false-color range: white-yellow color indicated a high intensity of LAMP1 with grayscale intensities of 225–180, orange to red indicated a partial intensity of LAMP1 (grayscale intensities of 180–80), and purple to blue indicated the absence of LAMP1 (grayscale intensities of 80–1) around the phagosomes (43).

To quantify fluorescence associated with granule markers, regions of interest were drawn around individual cells, and total fluorescence was acquired, followed by background correction.Zigmond chamber chemotaxis assayOne million bone marrow neutrophils in 100 ml of HBSS with 1% gelatin were incubated with different concentrations of apilimod for 30 min at 37˚C and then plated onto a 5% BSA–coated microscope cover glass (22 3 40 mm) for 10 min. The cover glass was inverted onto a Zigmond chamber with 100 ml of HBSS medium, and 100 ml of HBSS containing 1 mM fMLF was added to the right and left chambers. Time-lapse video microscopy was used torecord neutrophil movements in the Zigmond chambers for 15 min (one frame per 20 s). Captured images were analyzed for cell direction and speed using cell-tracking software (Retrac version 2.1.01 freeware). Data were collected from five independent experiments.To measure ROS generation, we used reduction of cytochrome C. One million bone marrow neutrophils in 100 ml of PBS with 10 mM D-glucose were incubated with different concentrations of apilimod for 30 min. They were then mixed with 880 ml of PiCM-G (138 mM NaCl, 2.7 mM KCl,0.6 mM CaCl2, 1 mM MgCl2, 5 mM glucose, 10 mM NaH2PO4/Na2HPO4 [pH 7.4]) supplemented with 0.1 mM cytochrome C and incubated for an additional 10 min at 37˚C. Cells were then stimulated with 1 mM fMLF or 1 mM PMA for 30 min at 37˚C. The absorbance of reduced cytochrome C at 550 nm was recorded and background corrected (reaction lacking cell lysates). Data were collected from five independent experiments.Preparation of recombinant fusion protein of GST and p21-binding domain proteinTo quantify Rac GTPase activation, we used affinity chromatography and a fusion protein of GST and the p21-binding domain (PBD) of PAK, as de- scribed previously, with a few modifications (44). Briefly, recombinant pro- teins were induced in BL21* Escherichia coli in the presence of 0.4 mM IPTG for 3 h at 30˚C. Fifty ODs of bacterial culture were centrifuged before the addition of 50 ml of bacterial lysis buffer (10 mM Tris [pH 8], 1 mM EDTA, 150 mM NaCl, 100 mg/ml lysozyme, 5 mM DTT, 1% Triton X-100, and supplemented with bacterial protease inhibitor mixture [Bio Basic, Markham, ON, Canada]) and 1 mM PMSF.

Bacteria were ground using a mortar and pestle with 1 g of Celite (Sigma-Aldrich). Lysates were cleared by centrifugation, and supernatant was added to reduced glutathione- Sepharose (Invitrogen, Carlsbad, CA) and incubated at 4˚C for 1 h with agitation, followed by three washes with bacterial lysis buffer.Affinity precipitation of GTP-bound Rac GTPase and Western blottingOne million neutrophils were treated at 37˚C for 30 min with 50 nM apilimod or DMSO, followed by the addition of 1 mM fMLF or vehicle for 1 min. Cells were placed immediately on ice and lysed with 100 ml of ice-cold 53 MLB lysis buffer (125 mM HEPES [pH 7.5], 25 mM EDTA,1% Triton X-100, 750 mM NaCl, 25 mM MgCl2, and 50% glycerol, supplemented with mammalian protease inhibitors [Sigma-Aldrich]). Cell lysates were clarified by centrifugation at 10,000 3 g for 5 min at 4˚C. Ten percent of cell lysates were removed to measure protein levels across each sample. The remaining cell lysates were incubated with 50 ml of glutathione- Sepharose beads attached to GST–PBD or GST (50% suspension) and incubated for 1 h at 4˚C with agitation. The samples were centrifuged for 2 min at 10,000 3 g, and the supernatant was removed. The pellets were washed three times with 13 MLB lysis buffer before protein elution with 23 Laemmli buffer containing 2-ME. Protein eluants were loaded and separated in a 12% SDS-PAGE, transferred to a polyvinylidene difluoride membrane, and processed for Western blotting with mouse anti-Rac1 Abs (clone 23A8, 1:2500; GeneTex, Irvine, CA), rabbit polyclonal anti-Rac2 Abs (EMD Millipore, ON, Canada), and HRP-linked goat anti-mouse or rabbit secondary Abs used at 1:10,000 (CEDARLANE, ON, Canada). ECL was detected and analyzed by band densitometry using a Gel Documentation System (Bio-Rad, Mississauga, ON, Canada).Statistical analysesAll experiments were repeated at least three times, and all data were subjected to statistical analysis using an unpaired or paired Student t test for single-parameter experiments or using ANOVA and the Tukey post hoc test for multiparameter experiments. Statistical significant was drawn at p , 0.05.

ResultsLysosomes, but not granules, vacuolate in PIKfyve-inhibited neutrophilsThe importance of PIKfyve activity in neutrophils has not been ex- amined previously. To investigate this, we used a pharmacological ap- proach to acutely block PIKfyve activity by apilimod or YM201636, two selective inhibitors of PIKfyve (34, 45). In fact, apilimod was recently shown to have exquisite selectivity for PIKfyve (46). First, we examined the sensitivity of murine neutrophils to apilimod or YM201636 by testing different concentrations and incubation times and scoring the number of vacuoles larger than 1 mm in diameter. By incubating cells for 1 h, we noted a gradual rise in the number of vacuoles in neutrophils exposed to increasing amounts of the in- hibitors (Fig. 1A, 1C, Supplemental Fig. 1A, 1C). We then used an intermediate concentration of 20 nM for apilimod and 10 nM for YM201636 to examine the rate of vacuolation. Neutrophils began to vacuolate significantly within 30 min of drug exposure and be- came highly vacuolated at 90 min of exposure (Fig. 1B, 1D, Supplemental Fig. 1B, 1D). Thus, to minimize off-target and indirect effects of PIKfyve inhibition, we generally treated neutrophils for,1 h at ,50 nM apilimod or YM201636, unless otherwise noted. We next attempted to identify the nature of the vacuoles in neu- trophils. Neutrophils are not only equipped with lysosomes, they also possess lysosome-related primary granules (or azurophilic granules), secondary granules, and tertiary granules (2, 47, 48), which can be labeled with Abs to LAMP1, MPO, lactoferrin, and MMP9, respectively. Although we clearly observed vacuolation of LAMP1+ lysosomes in neutrophils treated for 1 h with 20 nM apilimod, we did not discern vacuolation of organelles that labeled with the other markers (Fig. 2).

This suggests that LAMP1+ lysosomes are suscep- tible to swelling, whereas primary, secondary, and tertiary granulesresist enlargement in neutrophils acutely inhibited for PIKfyve.PIKfyve controls phagosome–lysosome fusion in neutrophilsWe previously showed that PIKfyve has an important role in phagosome maturation in macrophages (36). Because neutrophils are also professional phagocytes, we assessed the role of PIKfyve in phagocytosis and phagosome maturation. First, we evaluated the ability of neutrophils to engulf IgG-coated beads by measuring the phagocytic index and efficiency before and after treatment with apilimod or YM201636. We observed a significant reduction in phagocytic appetite, as measured by index and efficiency in neutrophils treated with $30 nM apilimod or $10 nM YM201636 (Fig. 3A, 3B), which was reminiscent of our prior observations with macrophages (36).To investigate the impact on phagosome maturation, we then treated murine neutrophils with 20 nM apilimod or 5 nM YM201636, which is sufficient to vacuolate neutrophils but is permissive for phagocytosis. After waiting 1 h to elicit phagosome maturation, we processed, stained, and quantified phagosomal acquisition of LAMP1 to track phagosome–lysosome fusion, as previously de- scribed (36, 37, 49). We observed a remarkable inhibition of phagosome maturation in cells blocked for PIKfyve. In vector-treated murine neutrophils, ∼60% of phagosomes labeled strongly with LAMP1 (LAMP1+), whereas only ∼10% were negative(LAMP12, Fig. 3C, 3D). In striking comparison, neutrophils inhibited for PIKfyve had ,5% LAMP+ phagosomes, whereas∼70% were LAMP12 (Fig. 3C, 3D). Similar trends were ob-tained in murine neutrophils treated with 5 nM YM201636 (Supplemental Fig. 2). In contrast, control and PIKfyve-inhibited neutrophils had similar levels of CD63 and MPO associated with phagosomes, suggesting that PIKfyve activity is not necessary for phagosome fusion with primary granules (Fig. 3E, 3F, Supplemental Fig. 3).We then assessed the mechanism by which PIKfyve mightcontrol phagosome maturation in neutrophils.

Given our prior work in macrophages, we postulated that PIKfyve is necessary to stimulate TRPML1 to release lysosomal Ca2+ and trigger phag- osome–lysosome fusion. To test this, we exposed apilimod-treated cells to ionomycin, a Ca2+ ionophore, or to MLSA1, a TRPML1 agonist, to determine whether these agents could rescue phag- osome–lysosome fusion in PIKfyve-inhibited cells. First, and as a control, we showed that ionomycin or MLSA1 alone did not im- pact LAMP1 staining of phagosomes (Fig. 3C, 3D). Second, and most important, ionomycin and MLSA1 were able to decrease thenumber of phagosomes devoid of LAMP1 staining; it decreased from ∼70% in apilimod-alone neutrophils to ∼30% in apilimod- treated cells exposed to ionomycin or MLSA1 (Fig. 3C, 3D). However, the rescue was partial, and most phagosomes became partially stained with LAMP1, as defined in Materials andMethods. Consistent with a role for TRPML1 and lysosomal Ca2+ in phagosome–lysosome fusion in neutrophils, we showed that Ca2+ chelation with BAPTA-AM potently hindered LAMP1 la- beling of phagosomes (Fig. 3C, 3D). Overall, these data suggest that PIKfyve controls phagosome–lysosome fusion in neutrophils, in part by stimulating TRPML1 and releasing lysosomal Ca2+ to trigger fusion. This is consistent with our previous work in mac- rophages (37), as well as a previously observed periphagosomal increase in cytosolic Ca2+ in neutrophils (50, 51).PIKfyve is not necessary for fMLF-induced degranulationIn addition to phagocytosis and phagosome maturation, neutrophils rely on degranulation to eliminate pathogens (5). Because PIKfyve and TRPML1 are linked to regulated exocytosis (52–54), we postulated that PIKfyve activity might govern some aspects of degranulation. Given that primary granules are lysosome-related organelles, and their secretion can depend on Ca2+ (5, 55–57), we examined the appearance of CD63 on the cell surface and the disappearance of MPO from within cells. We also tested the se- cretion of secondary and tertiary granules by quantifying deple- tion of lactoferrin and MMP9 from within cells after stimulation with fMLF.

First, we observed that resting neutrophils and those treated with only apilimod had comparable surface levels of CD63 (Fig. 4A, 4B) and similar cell-associated levels of MPO, lacto- ferrin, and MMP9, suggesting that PIKfyve does not affect basal levels of degranulation (Fig. 4). Second, we showed that fMLF exposure enhanced the levels of cell surface CD63 and abated cell-associated MPO, lactoferrin, and MM9 relative to resting neutrophils (Fig. 4). Lastly, the fMLF-induced degranulation was unabated in neutrophils pretreated with apilimod (Fig. 4). Overall, these data show that acute loss of PIKfyve activity does not impair fMLF-induced exocytosis of granules, although we cannot rule out an effect during chronic PIKfyve loss.PIKfyve activity is necessary for fMLF-directed chemotaxisPIKfyve and MTMR3 cooperate to synthesize PtdIns(5)P, which, together, activate the Rac GTPase to coordinate cell migration (25, 38, 58). Hence, we postulated that PIKfyve activity might regulate neutrophil chemotaxis. To test this hypothesis, we examined the ability of murine neutrophils to move toward an fMLF chemical gradient by quantifying their speed and directionality toward the fMLF source at different concentrations of apilimod. As expected, vector-exposed neutrophils exhibited a remarkable capacity to orient and move toward the fMLF gradient. By mapping the po- sition of neutrophils relative to their starting position, we foundthat ∼80% had moved toward the fMLF gradient (within the rightmost quadrants in Fig. 5A) and traveled at an average speedof 8.8 6 0.2 mm/min (Fig. 5B). Strikingly, at apilimod concen- trations as low as 10 nM, neutrophils became disoriented, with only ∼50% of the cells moving toward the gradient (Fig. 5A). At 35 and 70 nM apilimod, neutrophils were effectively spread equally across each quadrant and traveled a shorter overall dis- tance (Fig. 5A), which effectively suggests randomized and slower movement by neutrophils.

Indeed, neutrophil speed was signifi- cantly abated at 10 nM apilimod (5.1 6 0.5 mm/min) and brought to a near standstill at 70 nM (Fig. 5B). Overall, these experiments reveal an important function for PIKfyve activity in neutrophil chemotaxis.PIKfyve activity is necessary for fMLF-induced ROS productionWe then considered the possibility that PIKfyve activity may be necessary for ROS synthesis, which primarily occurs through activation of the NADPH oxidase (4). To test this, we measured ROS production in response to fMLF in control and apilimod- treated neutrophils using a cytochrome reduction assay. Resting neutrophils and those treated with apilimod alone (10, 35, or 70 nM) had similar levels of ROS production (Fig. 6). As expected, neutrophils exposed to stimulants like fMLF or phorbol esters exhibited a large increase in ROS (Fig. 6). Strikingly, even with pretreatment with 10 nM apilimod, ROS production was signifi- cantly subdued in neutrophils exposed to fMLF or PMA (Fig. 6). At 35 and 70 nM apilimod, ROS synthesis was essentially thwarted in stimulant-exposed neutrophils (Fig. 6). Overall, this suggests that PIKfyve activity is important for ROS generation, likely by stimulating the NADPH oxidase.PIKfyve activity is necessary for fMLF-induced Rac activationOur observations so far indicate that PIKfyve activity affects chemotaxis and activation of the NADPH oxidase and can impact phagocytosis. A common factor in all of these processes is that they depend on the activation of the Rac GTPases (16, 17, 40, 41, 59). Moreover, PIKfyve activity, through the action of MTMR3, is linked to Rac GTPase activation to catalyze migration of non- leukocytes (25, 38, 58). Finally, PtdIns(5)P was shown to bind and stimulate Tiam1, a guanyl exchange factor for Rac (58). Given all of this, we postulated that PIKfyve activity is necessary to stim- ulate Rac GTPases in neutrophils.

To test this hypothesis, we used an affinity chromatography assay that uses a GST–chimeric pro- tein of the PBD of p21-activated kinase to precipitate GTP-bound Rac, followed by Western blotting against the Rac1 and Rac2 GTPases. As expected, GST alone did not recover any Rac1 or Rac2 from cells stimulated with fMLF (Fig. 7). In addition, GST–PBD recovered little Rac1 or Rac2 in resting cells or cells treated with only apilimod or YM201636. In striking contrast, fMLF treatment led to a strong recovery of both Rac GTPases, which was abolished by pretreatment with apilimod or YM201636 (Fig. 7). Overall, these results support a model in which PIKfyve activity is necessary for Rac GTPase activation during neutro- phil stimulation to control chemotaxis, ROS production, and phagocytosis.PIKfyve activity is necessary for human neutrophil functionLastly, we tested whether the role of PIKfyve in neutrophil function was applicable to human neutrophils by examining two specific aspects: phagosome maturation and chemotaxis. Using interme- diate concentrations of apilimod and YM201636, we showed that phagosomes in human neutrophils were impaired for acquisition of LAMP1, suggesting a defect in phagosome–lysosome fusion (Fig. 8A, 8B). Lastly, human neutrophils treated with apilimod became disoriented and migrated more slowly than vector-treated cells, showing a defect in chemotaxis (Fig. 8C, 8D). Overall, these data suggest that PIKfyve also plays an important role in the immune response of human neutrophils.

Discussion
Neutrophils are exceptionally important for the immune response, rapidly targeting sites of infection by chemotaxis and unleashing a series of attacks on pathogens, including ROS generation, secretion of antibacterial peptides, and engulfment and digestion of patho- gens by phagocytosis (1, 2, 4). These processes are dependent on and coordinated by a complex signaling network that uses small GTPases and PtdInsP signals. However, the importance of PIK- fyve activity, which synthesizes PtdIns(3,5)P2 and directly or in- directly, PtdIns(5)P, for neutrophil function remained unexplored. We investigated that in this study and found that PIKfyve ac- tivity is critical for neutrophils to perform chemotaxis, generate ROS, and undertake phagosome fusion with lysosomes, but not for degranulation. Moreover, we propose a model by which phagosome maturation in neutrophils uses the PtdIns(3,5)P2 ef- fector, TRPML1, to mediate lysosomal Ca2+ release and trigger phagosome–lysosome fusion, whereas chemotaxis and ROS pro- ceed through PIKfyve-dependent activation of the Rac GTPases, likely through PtdIns(5)P.PIKfyve is well established to regulate lysosome morphology. In the absence of its activity, lysosomes in a multitude of cells, in- cluding the yeast vacuole, undergo massive enlargement (19, 20). However, to our knowledge and with the exception of platelet dense granules (35), lysosome-related organelles have not been examined for their susceptibility to enlarge in the absence of PIKfyve activity. In this article, we show that neutrophils contain a distinct LAMP1 compartment that becomes engorged by acute impairment of PIKfyve activity. In contrast, none of the neutrophil granules that we examined underwent perceptible changes in mor- phology under acute inhibition of PIKfyve. In addition, we showed that PIKfyve-inhibited neutrophils were still competent for fMLF-induced secretion of primary, secondary, and tertiary gran- ules, as well as for phagosome–primary granule fusion.

The de- granulation of primary granules in PIKfyve-inhibited cells is perplexing because fMLF-induced Rac2 activation is blunted by PIKfyve inhibitors but Rac22/2 murine neutrophils are defective for chemoattractant-induced primary granule secretion (60). We speculate that this inconsistency may be caused by differences in chronic (genetic) versus acute (pharmacological) disruption of Rac2 and PIKfyve function.Overall, these results suggest that PIKfyve has little direct control of granule function in neutrophils and seem consistent with the normal biogenesis and stimulus-dependent secretion of alpha- granules and dense granules found in platelets in mice carrying a platelet-specific PIKfyve gene deletion (35). Nevertheless, it re- mains possible that chronic deficiency of PIKfyve in neutrophils could impair biogenesis and function of the various neutrophil granules; this will require the generation of mice carrying a neutrophil-specific PIKfyve gene deletion.Neutrophils and macrophages have distinct phagosome-matura- tion processes. Although both use phagosome–lysosome fusion, phagosomes in neutrophils also merge with primary granules, acquire a high concentration of ROS, and do not acidify signifi- cantly (1, 57). In this article, we showed that phagosomes enclosing IgG-coated particles require PIKfyve activity and Ca2+ to fuse with LAMP1+ lysosomes in neutrophils. Importantly, ionomycin or the TRPML1 agonist MLSA1 partially rescued phagosome acquisition of LAMP1 in PIKfyve-inhibited cells. This suggests that a PtdIns(3,5)P2-dependent activation of TRPML1 releases lysosomal Ca2+ and triggers phagosome–lysosome fusion. These observations are consistent with our earlier findings in macrophages, which also require the PIKfyve–TRPML1–Ca2+ axis to trigger fusion between docked phagosomes and lysosomes (36, 37).

In contrast, PIKfyve was not necessary for phagosome fusion with primary granules, suggesting a different mechanism for this process.PIKfyve activity is necessary for Rac activation in neutrophilsAblation of PIKfyve activity strongly impaired Rac1 and Rac2 activation in response to fMLF signaling. Interestingly, Rac1 and Rac2 exhibit distinct functions in neutrophils (61). For example, Rac2, but not Rac1, is essential for activation of the NADPH oxidase and ROS production (41, 62). In comparison, Rac2 and Rac1 are both necessary for chemotaxis but play distinct roles; neutrophils require Rac2 for chemokinesis, whereas Rac1 is needed for neutrophil orientation toward the gradient (41, 62–65). Based on these and our own observations, the inability of PIKfyve- inhibited neutrophils to activate Rac1 and Rac2 GTPases likely underpins the impaired ROS production and chemotaxis in these cells. In addition, it is noteworthy that there is cross-talk between Rac1 and Rac2 (66). Thus, it will be important to investigate whether PIKfyve acts on Rac1 and Rac2 independently using parallel pathways or whether it is part of a common regulatory mechanism that mediates Rac1–Rac2 cross-talk.The importance of PIKfyve for chemotaxis fits the observa- tions by Oppelt et al. (25, 38) showing that PIKfyve regulates cell migration of fibroblasts and cancer cells. These investigators suggested that PIKfyve collaborates with MTMR3 to generate PtdIns(5)P, which then stimulates Rac (25, 38). More recently, Viaud et al. (58) showed that PtdIns(5)P binds preferentially to the C-terminal PH domain of Tiam1 that is adjacent to the canonical DH domain, which serves as a hallmark for Rho family GEF proteins. We envision that a similar process may be occurring in stimulant-exposed neutrophils. In fact, neutrophils express various GEFs that contain a DH–PH domain that could conceivably be targeted by PtdIns(5)P, including Tiam1, Tiam2, P-Rex, and Vav (7, 67–69).

Overall, we provide evidence that PIKfyve activity is necessary for a multitude of neutrophil functions. We posit that PIKfyve function is split between two signaling branches: a PtdIns(3,5)P2- dependent arm that governs phagosome maturation and lysosome activity through TRPML1 and a PtdIns(5)P-dependent arm that regulates Rac, thus modulating the NADPH oxidase and chemo- taxis. Unfortunately, attempts to biochemically measure PtdIns (3,5)P2 and PtdIns(5)P in neutrophils were not successful because of low [3H]myo-inositol incorporation (70). Thus, we could not test whether these PtdInsPs change during stimulation with fMLF or treatment with apilimod. However, it will be worthwhile to revisit the measurement of PtdInsPs in neutrophils upon improved sensitivity of flow scintillation and/or differentiation of PtdInsP isomers by mass spectrometry (71). Finally, it is noteworthy that apilimod was initially discovered as an anti–IL-12/13 inhibitor and was clinically tested against Crohn’s disease (33, 34); it is enticing to consider whether apilimod-dependent neutrophil sup- pression contributes to these clinically relevant Apilimod outcomes.