Ral Isoforms Are Implicated in FcγR-Mediated Phagocytosis: Activation of Phospholipase D by RalA

Phagocytosis is a complex, evolutionarily conserved process that permits the internalization and degradation of large particles (>0.5 μm) in diverse cellular functions, ranging from nutrition in unicellular organisms and remodeling of tissues during development to the maintenance of homeostasis and elimination of pathogens in the immune response of higher organisms. The phagocytic signal is triggered by a variety of receptors, including the extensively studied FcγR, which recognizes the constant fragment of Ig opsonized on particles (1). Following activation and clustering of the FcγR, signal transduction leads to rearrangements of the actin cytoskeleton that drive the extension of pseudopods around the particle to form the phagosome, the vacuole in which the particle is engulfed. To maintain homeostasis of the cell surface, this step also requires the insertion of endomembranes by focal exocytosis (2), which depends on actomyosin contractile activity (3) and membrane-fusion events (4). After internalization, the phagosome matures into a phagolysosome through interactions with the endosomal–lysosomal pathway, leading to acidification of the vacuole, degradation of the ingested material, and its eventual recycling for Ag presentation (5).

Membrane trafficking is essential for phagocytosis during the formation and maturation of the phagosome. A number of intracellular compartments, including early and late endosomes, lysosomes (6), and the endoplasmic reticulum (7), have been proposed as membrane sources for focal exocytosis at phagocytic sites. SNARE complexes (8, 9) and different phospholipids (10, 11) are involved in these membrane-trafficking events. Phospholipase D (PLD) activity, which generates phosphatidic acid (PA) from phosphatidylcholine, was reported to increase during activation of several phagocytosis receptors, including FcγR (12). Recently, we demonstrated that PA is produced at phagocytic sites and that the PLD isoforms PLD1 and PLD2 are necessary for efficient phagocytosis (13). PLD2 is present at the plasma membrane, whereas the isoform PLD1 is associated with the late endosome/lysosome compartment that is recruited to phagocytic sites, suggesting a role for PLD in focal exocytosis during phagosome formation (13). How the PLD isoforms are regulated and their cellular functions during phagocytosis remain to be elucidated.

The monomeric Ral GTPases are potential regulators of PLD. RalA and RalB form a subgroup of the superfamily of Ras proteins and share ∼85% homology in their amino acid sequences. The importance of the Ral GTPases in membrane-trafficking events and organization of the actin cytoskeleton is well established in a number of cellular functions, such as cell migration, neurite branching, receptor endocytosis, and regulated exocytosis (14). In this study, the functional importance of Ral proteins in phagocytosis was investigated, in particular the relationship between RalA and PLD isoforms during phagosome formation. Ral proteins were observed to have opposing roles in mediating the internalization of 3-μm IgG-opsonized particles: RalA acts as a positive modulator, whereas RalB plays a negative role. Additional results provide evidence that RalA interacts with the PLD isoforms and regulates PLD activity, suggesting that a pathway involving RalA and PLD modulates the efficiency of FcγR-dependent phagocytosis in macrophages.

Culture conditions for the murine macrophage RAW 267.4 cell line in RPMI 1640-glutamax medium supplemented with 10% FBS were as described previously (13).

RPMI 1640 and FBS were purchased from Invitrogen. Latex beads (3 μm) from Sigma-Aldrich (St. Louis, MO) were coated with human IgG from Zymed (San Francisco, CA). Rat anti-hemagglutinin (HA) affinity matrix was purchased from Roche Diagnostics (Indianapolis, IN). The following monoclonal Abs were used: anti-myc (clone 4A6; Upstate Biotechnology, Lake Placid, NY), anti-RalA (BD Biosciences, San Jose, CA), anti-HA (HA.11; Covance, Emeryville, CA), and anti-actin (clone AC-15; Sigma-Aldrich). Goat anti-RalB Abs were from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit polyclonal Abs against PLD1 (C-terminal) and PLD2 (N-terminal) were obtained from Cell Signaling Technology (Beverly, MA) and BD Biosciences, respectively. Goat secondary Abs against mouse Fc were from Sigma-Aldrich, and those coupled to peroxidase and directed against mouse and rabbit IgG were from Pierce (Rockford, IL). Donkey peroxidase Abs against goat IgG were from Santa Cruz Biotechnology. Transferrin-Alexa Fluor 568, LysoTracker DN Red99, and secondary goat Abs coupled to Alexa Fluor conjugates (488 or 555) from Molecular Probes (Eugene, OR) were used as previously described (13).

RalA wild type (RalA-WT; X04328.1) and the variants, constitutively active RalA G23V, RalA 23VΔN11 (11 aa deleted from the N-terminal eliminates the PLD-binding site), and catalytically inactive RalA G26A, cloned in pRK5 plasmid were described previously (15, 16). RalB wild type (RalB-WT; X15015.1) and the variants, constitutively active RalB G23V and catalytically inactive RalB S28N in pRK5-myc vector, were kindly provided by G. Lalli and A. Hall (University College of London, London, U.K.). RalA-WT was subcloned by PCR into the BamHI and XbaI sites of the pRK5-myc, enhanced GFP plasmid 1, and dsRedmonomerC1 vectors, and RalB-WT was similarly subcloned into enhanced GFP plasmid 1 (17). The other plasmids, cathepsinD-RFP, pCGN-PLD1 (HA-PLD1), pCMV3-PLD1-GFP (GFP-PLD1), pCGN-PLD2 (HA-PLD2), and enhanced GFP (pEGFP)-PLD2 (GFP-PLD2), as well as plasmid transfection by electroporation, were as previously described (13).

Macrophages were transfected with stealth interference RNA (iRNA) duplexes specific for mouse RalA, RalB, or a control nontargeted oligonucleotide (medium CG) using Liopofectamine iRNA max, according to the manufacturer’s instructions (Invitrogen, Cergy Pontoise, France). The following iRNA duplexes corresponding to the sense strands were selected for RalA: 5′-gguaacaagucagaucuagaagaua-3′ (A-1) and 5′-acagagcugaccaguggaacguuaa-3′ (A-2) and for RalB: 5′-gcccugacgcuucaguucauguaug-3′ (B-1) and 5′-gccaauguggacaagguauucuuu-3′ (B-2). Using a control oligo-Alexa Fluor 488 (Invitrogen), the transfection efficiency of stealth iRNA was estimated by flow cytometry to be >85%. After 48 h, the reduction in the Ral proteins was determined by Western blots, and the effects of the iRNAs on phagocytosis and PLD activity were analyzed. For immunoblots, cells were solubilized in 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 2 mM EDTA, 0.5% Triton X-100, 0.5% deoxycholate, and a protease-inhibitor mixture (Sigma-Aldrich).

Phagocytosis was carried out in serum-free complete RPMI 1640 medium and stimulated using 3-μm latex particles opsonized with human IgG (20 beads/cell), essentially as previously described for phagocytosis assays and confocal microscopy (13). For Ral-activation assays, PLD activity tests, and immunoprecipitation experiments, phagocytosis was initiated by adding the particles directly to cell suspensions.

For these assays, phagocytosis was synchronized by adding IgG-opsonized particles to cells grown on glass coverslips, centrifuging cells for 2 min at 100 × g at 18°C, and then initiating phagocytosis by placing them at 37°C, as previously described (13). Briefly, phagocytosis was stopped 5 or 30 min later by washing twice in cold PBS. At these time points, the uptake of nonopsonized particles was negligible. After fixation, external beads were labeled with goat anti-human IgG coupled to Alexa Fluor 555. Cells transfected with Ral plasmids were identified by permeabilizing cells in Triton X-100 and labeling with anti-myc Ab. Unlabeled internalized beads were visualized with phase-contrast optics. The number of internalized beads per cell was determined for randomly chosen fields (≥80 cells for each condition) using superimposed fluorescent and phase-contrast images (Adobe Photoshop 9). After determining the mean number of beads internalized per cell, the phagocytic index calculated for transfected cells was divided by the index obtained for nontransfected control cells or by the index obtained for cells transfected with the control iRNA (for the iRNA experiments). These values are presented as a percentage of control cells. The Student t test was used to evaluate the differences between the different transfection conditions.

Phagocytosis was initiated at 37°C by adding IgG-opsonized particles to 10⁷ cells in suspension, and it was stopped at different times (0, 5, 10, or 20 min) by adding ice-cold TBS and collecting the cells by centrifugation. To test for the specificity of RalA activation by FcγR, macrophages were incubated for 5 min at 37°C with noncoated particles, BSA-coated particles (prepared as described above for IgG-opsonized particles), or heat-aggregated IgG (18). For each time point, Ral-GTP was precipitated from lysates (2 mg total protein) with the Ral-binding domain of RalBP1, according to the instructions in the Ral Activation Assay Kit (Upstate Biotechnology). Ral-GTP precipitates and 10–20 μg protein aliquots of total cell lysates were analyzed on Western blots using Abs against RalA, RalB, and actin.

Forty-eight hours after transfection with iRNAs, macrophages (6 × 10⁶) in suspension were incubated at 37°C in the absence (resting) or presence (stimulated) of IgG-opsonized particles for 5 min. Cell lysates were prepared in ice-cold 50 mM Tris-HCl (pH 8) by three freeze–thaw cycles. Aliquots of the lysates (corresponding to 10⁶ cells) were assayed for PLD activity using the Amplex Red PLD Assay Kit (Molecular Probes), and PLD activity was estimated after a 1-h incubation at 37°C with a Mithras LB940 Fluorometer (Berthold Technology, Yvelines, France). For each cell preparation, a standard curve was established using purified PLD from Streptomyces chromofuscus (Sigma-Aldrich), and an average activity (mU/ml) was calculated from six determinations made for each condition.

Macrophages were transfected with myc-RalA and HA-PLD1 or HA-PLD2 and were collected 24 h later (5 × 10⁶ cells/condition). Phagocytosis was stimulated by adding IgG-opsonized beads to cell suspensions at 37°C for different times, and cell lysates were prepared in 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1 mM MgCl_2, 0.1 mM DTT, 0.5% Triton X-100, 0.5% deoxycholate and protease-inhibitor mixture (Sigma-Aldrich). PLD proteins were immunoprecipitated from lysates (0.5 mg protein) using rat anti-HA affinity beads, according to the manufacturer’s instructions (Roche).

Total-cell lysates were homogenized in 20 mM HEPES (pH 7.5), 320 mM sucrose, 2 mM EDTA, and protease-inhibitor mixture, and subcellular fractions corresponding to a soluble cytoplasmic supernatant and a membrane fraction composed of organelles and large-membrane fragments were obtained by centrifugation (10,000 × g for 20 min), as previously described (13). Proteins were separated on 4–12% Nupage gradient gels in MES buffer (Invitrogen) and transferred to nitrocellulose. Immunoblots were revealed using the SuperSignal Chemiluminescent Kit (Pierce). For the iRNA experiments, blots were probed for actin to control for equal protein loadings. Images were acquired using a Chemi-Smart 5000 and the Chemi-Capt program (Vilber Lourmat, Marne la Vallée, France). Images were processed with Photoshop 9, and protein bands were quantified using the Bio1D program (Vilber Lourmat).

Live cells were observed 18–24 h after transfection in the absence or presence of IgG-opsonized particles. Cells destined for immunofluorescent labeling were fixed for 10 min at 4°C with 4% paraformaldehyde in 0.125 M phosphate buffer; for intracellular labeling, this was followed by a 5–10-min permeabilization step in fixative containing 0.2% Triton X-100. Samples were then blocked with 10% goat serum before incubating with primary Abs and subsequently revealed with appropriate secondary goat Abs. Images of labeled cells were obtained using a Zeiss LSM 510 inverted microscope equipped with a planapo oil (×63) immersion lens (numerical aperture = 1.4), using Ar 488-nm and He/Ne 543-nm lasers and emission filters of 505–530 and 560 nm. Images were recorded with the same parameters and optimal pinhole and were processed using LSM 5 Image Browser and Adobe Photoshop 9. Using Zeiss CLSM software 2.8, the proportion of dsRed-RalA colocalized with GFP-PLD1 was determined in nonstimulated cells and in cells during particle ingestion, using the weighted colocalization percentages generated for double-labeled pixels (x ± SEM; n = 15 cells/condition).

The presence and localization of the Ral proteins in murine RAW264.7 macrophages were first studied by Western blotting (Fig. 1A). RalA and RalB were detected in cell homogenates as single bands migrating at 28 kDa. Subcellular fractionation of these total-cell lysates by differential centrifugation indicated that both of the endogenous Ral isoforms are concentrated (>90%) in the membrane fraction corresponding to the plasma membrane and organelles. Because Abs for immunocytochemistry are not available for RalB, the distribution of the Ral isoforms was compared in macrophages transfected with RalA-GFP or RalB-GFP. Confocal images of live cells expressing these tagged proteins indicated that RalA and RalB are similarly distributed, principally at the cell periphery, but also on cytoplasmic vesicles (Fig. 1B). However, in fixed cells, the Ral-labeled cytoplasmic compartments appeared more often as cytoplasmic puncta rather than as vesicles (data not shown).

Like other small GTPases, the Ral proteins cycle between a GTP-bound active form and a GDP-inactive form, and the active conformation influences their interaction with various effectors (14). To answer the question of whether the Ral proteins are activated during phagocytosis, the quantities of RalA-GTP and RalB-GTP were evaluated using the Ral-binding domain of RalBP1 to pull down the active forms from cell lysates prepared from macrophages maintained in suspension under resting conditions or following phagocytic stimulation. When macrophages were incubated with IgG-opsonized particles, the amounts of activated RalA and RalB increased transiently, and their activation profiles were similar (Fig. 1C). The amount of the active forms increased after only 2.5 min of stimulation, reaching a maximum at 5 min before returning to the basal level observed in nonstimulated macrophages at 20 min. Quantification of Western blots revealed a maximal 2–4-fold increase at 5 min for RalA-GTP and RalB-GTP with IgG-opsonized particles. Moreover, a 5-min incubation with aggregated IgG, which stimulates all classes of FcγRs (18), also activated RalA to a similar extent (Fig.1D). In contrast, noncoated particles and BSA-coated particles had only a slight effect on the quantity of RalA-GTP (0.3-fold increase; Fig.1D). These data suggest that Ral activation is linked to stimulation of FcγRs.

Next, the functional importance of the Ral proteins in the regulation of FcγR-induced phagocytosis was determined by overexpressing wild type and mutants, either constitutively active or catalytically inactive forms of RalA and RalB, and testing their effects on the capacity of transfected macrophages to internalize IgG-opsonized particles. In fixed cells expressing Ral protein tagged with myc, no marked differences in the distribution of the Ral proteins at the cell periphery and in cytoplasmic puncta were observed between wild type and mutants (data not shown). For the phagocytosis assays, a comparison of the number of beads ingested by transfected cells and nontransfected cells in the same fields served as a measure of phagocytic capacity (13). Control macrophages transfected with pEGFP plasmid displayed a phagocytic capacity similar to that of nontransfected cells, indicating that the transfection protocol itself did not alter the internalization process (Fig. 2). After a 5-min stimulation, overexpression of RalA-WT caused an increase of 34.2 ± 6.4% in the phagocytic capacity, and the constitutively active mutant RalA G23V caused an increase of 35.4 ± 3.0%, whereas the dominant negative mutant RalA G26A reduced the phagocytic capacity by 28.4 ± 1.6% (Fig. 2A). In contrast, the overexpression of the wild type or the mutant forms of RalB, both active (G23V) and inactive (S28N), reduced the phagocytic capacity of macrophages by 20–35% (Fig. 2B). Similar results were obtained when phagocytic capacities were assayed after a 30-min incubation (Fig. 2). These results suggest a positive role for RalA in the regulation of phagocytosis and, eventually, a negative role for RalB, although the observation that active and inactive forms of RalB inhibit phagocytosis is difficult to interpret.

To confirm a role for the Ral GTPases in phagocytosis, the expression of endogenous RalA and RalB was diminished by transfecting macrophages with iRNA duplexes specific for RalA (A-1 and A-2) and RalB (B-1 and B-2). After a 48-h treatment with iRNA duplexes, the depletion of RalA and RalB was assessed on Western blots (Fig. 3A). The expression of RalA and RalB was specifically decreased with their respective iRNAs, and neither the level of expression of the other isoform nor that of actin were affected. In comparison with a control nontargeted iRNA duplex, a quantification of the Ral protein bands revealed that the iRNAs A-1 and A-2 reduced the expression of RalA by 80 and 60%, respectively, and that the iRNAs B-1 and B-2 diminished the expression of RalB by 90 and 75%, respectively (Fig. 3B). Quantification of their effects on the phagocytic capacity of macrophages is summarized in Fig. 3C. RalA iRNA A-1 and A-2 inhibited the uptake of IgG-opsonized particles by 28 ± 2.3% and 21.9 ± 4.4%, respectively. In contrast, reducing the expression of the isoform RalB with iRNA duplexes enhanced the phagocytic capacity of macrophages by 25.9 ± 4.4% and 43.6 ± 10.3%, respectively. When the expression of both Ral isoforms was reduced, the phagocytic capacity was inhibited by 30.2 ± 3.2% (Fig. 3C), suggesting that the effects of the RalA isoform predominate. Because the expression levels of the FcγRs were unaffected in cells with reduced levels of endogenous RalA and RalB (Supplemental Fig. 1), these results implicate endogenous Ral GTPases in signaling pathways linked to the phagocytosis of IgG-opsonized particles. In agreement with our observations on the effects of overexpressing wild type and mutant Ral proteins, endogenous RalA apparently plays a positive role and endogenous RalB plays a negative role in the ingestion process.

Given that RalA seemed to be a positive regulator of phagocytosis and seemed to override the opposing negative effects of RalB, we focused our investigation on RalA. We next characterized the localization of RalA in live macrophages transfected with GFP-RalA during phagocytosis by confocal microscopy under resting conditions and during stimulation with IgG-coated particles. As shown in Fig. 1B, RalA was observed at the plasma membrane and on intracellular vesicles of varying sizes in nonstimulated macrophages (Fig. 4). The nature of the intracellular vesicular compartment was investigated by incubating transfected cells with transferrin, a marker of early and recycling endosomes, or LysoTracker, which labels late endosomes and lysosomes. RalA showed very little association with the transferrin compartment (Fig. 4A), but it was clearly located on some vesicles labeled with LysoTracker (Fig. 4B). In agreement, RalA was observed, but not exclusively, on lysosomal membranes in transfected macrophages coexpressing GFP-RalA and cathepsin-RFP, a specific lysosomal enzyme (Fig. 4C). Thus, in RAW 264.7 macrophages, RalA is localized principally at the plasma membrane, as well as on a vesicular population, corresponding to late endosomes and lysosomes in nonstimulated cells.

Following stimulation of phagocytosis, the distribution of RalA between the plasma membrane and intracellular vesicles was still evident (Fig. 5). However, as shown in the time-series of confocal images in Fig. 5, enrichment of RalA was observed at phagocytotic sites as the phagosomal cup forms and the pseudopods begin to extend around the particle. This was particularly evident in the plane of the plasma membrane (e.g., panels taken at 5 and 10 min; Fig. 5). When the phagosome closes and is internalized, the intense GFP-RalA labeling was lost from the membrane of the phagosome, although a weak labeling of the internalized phagosome and adjacent lysosomes persisted (Supplemental Fig. 2). Thus, RalA is enriched at phagocytotic sites, suggesting a role for RalA in the initial steps of phagosome formation.

Given that PLD isoforms also positively regulate the initial steps of phagocytosis (13, 19) and that PLD is a known downstream target of Ral GTPases (20), we next investigated whether there was a link between these proteins during phagocytosis. As a first step in this direction, the phagocytic capacity of macrophages expressing a catalytically active RalA (23VNT11) mutant that cannot associate with PLD (16) was examined after 5- and 30-min stimulations. As shown in Fig. 6A, after a 5-min stimulation, the phagocytic capacity of these cells was reduced by 28.2 ± 5.2% compared with cells transfected with a pEGFP plasmid, and similarly after a 30-min stimulation, suggesting that an interaction between RalA and PLD influences the phagocytotic process.

To further explore this possibility, endogenous RalA levels were reduced using an iRNA approach, and PLD activity was assayed in these macrophages maintained in resting conditions and after a 5-min stimulation with IgG-opsonized particles (Fig. 6B). Reduced RalA or RalB levels did not affect basal PLD activity in nonstimulated macrophages. In contrast, during stimulation, the 2-fold increase in PLD activity observed in macrophages transfected with the control iRNA was diminished by ∼25% in cells with reduced RalA levels. Furthermore, in macrophages transfected with iRNAs directed against RalB, PLD activity was not affected (Fig. 6B). Taken together, these results indicate that RalA positively influences PLD activity during FcγR-mediated phagocytosis.

Next, the localization of RalA with the PLD isoforms was examined in live cells cotransfected with dsRed-RalA and GFP-PLD2 (Fig. 7A) or GFP-PLD1 (Fig. 7B) using confocal microscopy. In resting conditions, dsRed-RalA largely colocalized with GFP-PLD2 at the plasma membrane; during the uptake of IgG-opsonized particles, RalA and PLD2 were observed at phagocytotic sites where they colocalized on the phagosomal membrane (Fig. 7A). In contrast, the distribution of dsRed-RalA with GFP-PLD1 appeared to change during particle ingestion (Fig. 7B). In the absence of phagocytic stimulation, GFP-PLD1 was expressed principally on a cytoplasmic vesicular population, previously identified as late endosomes and lysosomes (13), and was expressed weakly at the plasma membrane, whereas dsRed-RalA was concentrated at the plasma membrane and on a population of cytoplasmic vesicles, which was only partially colocalized with the PLD1⁺ vesicles. However, around sites of particle ingestion (Fig. 7B), RalA was evident on PLD1⁺ vesicles, suggesting that RalA had been recruited to these vesicles. A colocalization analysis was done to confirm the apparent increased association of RalA with PLD1 during a phagocytic stimulation. Under resting conditions, this analysis revealed that 22.9 ± 4.2% of dsRed-RalA was colocalized with GFP-PLD1. When phagocytosis was stimulated with IgG-coated particles, the colocalization of dsRed-RalA with GFP-PLD1 increased to 54.9 ± 3.3%. These results suggest that RalA may interact with both PLD isoforms and that the RalA and PLD1 association is enhanced during phagosome formation.

An interaction of RalA with the PLD isoforms was then assessed by immunoprecipitating PLD1 and PLD2 from macrophages maintained in suspension under resting conditions or stimulated with IgG-opsonized particles for different time periods (5, 10, or 20 min). For these experiments, cells were cotransfected with HA-PLD1 or HA-PLD2 and myc-RalA, and PLD was immunoprecipitated with anti-HA Abs. The Western blot analysis of the immunoprecipitates is presented in Fig. 8. For the PLD1 immunoprecipitates, the quantity of PLD1 detected with anti-HA and anti-PLD1 Abs was similar in resting and stimulated conditions, whereas the quantity of RalA detected with anti-myc and anti-RalA Abs varied with the duration of stimulation (Fig. 8A). In comparison with nonstimulated cells, the quantity of myc-RalA precipitated with PLD1 increased at 5 and 10 min of stimulation, and it returned to a level similar to resting conditions after 20 min of stimulation. A quantification of the myc-RalA band after 5 min of stimulation indicated a 2–3-fold augmentation compared with the quantity present in the PLD1 immunoprecipitates at 0 and 20 min. For PLD2 immunoprecipitates (Fig. 8B), the quantity of PLD2 detected with anti-HA and anti-PLD2 was similar at the different time points. However, in these immunoprecipitates, the quantity of RalA-myc increased in stimulated conditions, but it did not vary from 5 to 20 min of stimulation. These results provide evidence of an interaction between RalA and the PLD isoforms during phagocytosis and, more importantly, suggest a correlation between the transient association of RalA with PLD1 and the activation of RalA, which occurs during stimulation of the FcγR.

Several members of the small GTPase families, including the Rho and ARF families (21, 22), have emerged as key elements regulating cytoskeleton rearrangements and membrane trafficking during FcγR-mediated phagocytosis. This report provides evidence that 3-μm IgG-opsonized particles activate the Ral GTPases and that these proteins can modulate phagocytic efficiency. Activation of the Ral proteins is mediated, at least in part, by the engagement of FcγR, although simultaneous cross-activation of other phagocytosis receptors (23, 24) cannot be excluded. However, our findings are in line with recent observations indicating that the GTPase Rap1 and Ral guanine-dissociation stimulator, a guanosine exchange factor that can activate Ral proteins, participate in FcγR-dependent phagocytosis (25). Interestingly, the phagocytosis assays of macrophages expressing reduced endogenous levels of Ral proteins revealed that RalA facilitated, whereas RalB impaired, phagocytosis efficiency and that the RalA effect was dominant. In other cellular activities, distinct and overlapping functions of the Ral isoforms have been described. In the case of neurite arborization, the Ral isoforms apparently cooperate, and both positively regulate different aspects of the same process (17). In contrast, cancer cell migration depends on RalB and not RalA (26), whereas tumor proliferation and survival were proposed to depend on the balance between antagonistic activities of RalA and RalB (27).

The opposing effects of the Ral isoforms may be linked to their association with different Ral effector proteins. For example, RalBP1 (RLIP76), which binds Ral-GTP, also possesses a GTPase-activating domain for Cdc42 and Rac1 (28). By inactivating these Rho GTPases, RalBP1 could inhibit actin polymerization and, thereby, reduce phagocytic efficiency, because dynamic rearrangements of the actin cytoskeleton are required during phagosome formation (21). In contrast, another potential effector of the Ral proteins, the exocyst, was recently shown to facilitate phagocytosis (29). The exocyst is an octomeric protein complex, which serves to tether vesicles at target membranes prior to membrane fusion (27) in a number of membrane-trafficking events (30, 31). Activated forms of both Ral isoforms were shown to promote assembly of the exocyst complex through their interaction with the exocyst proteins Exo84 and Sec5 (32). Although the functional link between the exocyst and Ral proteins has not been addressed during the course of phagocytosis, the positive effect that RalA exerts on phagocytosis may well involve exocyst-mediated membrane incorporation at sites of phagosome formation.

We showed in this article that the positive effect of RalA on phagocytosis can be explained, in part, by its interaction with PLD, which itself positively regulates phagocytosis (13, 19). A relationship between the Ral GTPases and PLD was also described during regulated exocytosis (16), receptor endocytosis (33), and neurite branching (17). In the current study, RalA localized with PLD1 and PLD2 on nascent phagosomes, and the coimmunoprecipitation of RalA with the PLD isoforms under resting and stimulated conditions provides evidence that these proteins can interact in macrophages. More specifically, the transient activation of RalA following stimulation of the FcγR was correlated with an increase in the amount of RalA immunoprecipitated with PLD1. This observation is in line with in vitro studies of RalA–PLD1 interactions, which showed that the N-terminal 11 aa of RalA can bind PLD1 (20, 34) and that RalA-GTP enhanced this association (34). Hence, formation of RalA-PLD complexes seem to be critical for the positive regulation of phagocytosis because phagocytic efficiency was reduced when the constitutively active RalA mutant with a deleted PLD-binding site was expressed.

During phagocytosis of IgG-coated particles, regulation of PLD activity apparently depends on the RalA isoform, because reduced expression of RalA, but not that of RalB, partially inhibited the total cellular PLD activity in stimulated macrophages, although neither affected PLD activity in nonstimulated cells. Because PLD2 has a higher basal activity than PLD1 in mammalian cells, and PLD1 is more sensitive to diverse activators and constitutes the major producer of PA following stimulation (35), this result implies that RalA is likely to modulate PLD1 activity during FcγR-mediated phagocytosis, undoubtedly in concert with other regulators. Formation of RalA-PLD1 complexes was described with protein kinase C (36) and the ARF GTPases, ARF1 (20, 34) in vitro and ARF6 in vivo (16), and these activators can cooperate with RalA to synergistically stimulate PLD1. Although their potential roles in activating PLD during phagocytosis have not been resolved, protein kinase C regulates phagocytosis (1), and ARF6 and ARF1 are activated upon stimulation of the FcγR and have been implicated in the focal exocytosis of endomembranes during phagosome formation (37–39).

Determining the cellular functions regulated by this RalA–PLD interaction during phagocytosis remains an important unresolved issue. Because lipids and lipid-anchored proteins coordinate spatial and temporal patterns within the nascent phagosomal membrane (40), modulation of PA production by PLD at phagocytotic sites (13) is likely to contribute to the organization of the phagosomal cup and could affect multiple functions. PA can be converted to other bioactive lipids (e.g., diacylglycerol) (23), or it can stimulate lipid production, as for PI(4,5)P₂ (41), both of which are implicated in the phagocytotic process. Alternatively, PA may activate or recruit regulators of the actin cytoskeleton and membrane trafficking (42). In the case of focal exocytosis, PA could participate directly in the fusion process because membrane domains of PA (due to its small polar head) promote negative membrane curvatures favoring membrane hemifusion (43) or, indirectly, by serving as a lipid anchor for proteins involved in the fusion process [e.g., 11FIP, located on recycling endosomes (44) or syntaxin, one of the SNARE proteins (45)]. In function of the size and number of particles to be internalized, phagosomal formation has been proposed to require more or less focal exocytosis of internal membranes (7), and this may mean that the requirements for PA at phagocytic sites differ. Further investigation along this line should clarify the importance of the RalA–PLD pathway in the delivery of endomembranes to phagocytotic sites.

In conclusion, the present results indicate that RalA and RalB are implicated in the regulation of phagocytosis, most likely by controlling diverse downstream effectors, as seen by their distinct and opposed effects on phagocytic activity. Among these effectors, our data reveal a novel regulatory pathway involving RalA and PLD in the production of PA during FcγR-mediated phagocytosis. These results reinforce the view that PA represents an important cog in the phagocytotic machinery, and the next challenge will be to unravel its actual function during phagosome formation.

We thank Drs. G. Lalli and A. Hall for kindly providing the RalB-myc plasmids, Dr. F. Darchen for the cathepsin-RFP plasmid, and Dr. S. Gasman for advice on the preparation of the manuscript. We also thank Tam Thahouly and Gaby Ulrich for technical assistance and Plateforme d’Imagerie In Vitro of the Centre National de la Recherche Scientifique, Instituts Fédératifs de Recherche 37 en Neurosciences for the use of the confocal microscopy facilities.

Disclosures The authors have no financial conflicts of interest.