The cationic cluster of group IVA phospholipase A2 (Lys488/Lys541/Lys543/Lys544) is involved in translocation of the enzyme to phagosomes in human macrophages.

Group IVA cytosolic phospholipase A(2)alpha (cPLA(2)alpha) plays a role in the microbicidal machinery of immune cells by translocating to phagosomes to initiate the production of antimicrobial eicosanoids. In this work, we have studied the involvement of the cationic cluster of cPLA(2)alpha (Lys(488)/Lys(541)/Lys(543)/Lys(544)) in the translocation of the enzyme to the phagosomal cup in human macrophages responding to opsonized zymosan. Phagocytosis was accompanied by an increased mobilization of free arachidonic acid, which was strongly inhibited by pyrrophenone. In transfected cells, a catalytically active enhanced green fluorescent protein-cPLA(2)alpha translocated to the phagocytic cup, which was corroborated by frustrated phagocytosis experiments using immunoglobulin G-coated plates. However, a cPLA(2)alpha mutant in the polybasic cluster that cannot bind the anionic phospholipid phosphatidylinositol 4, 5-bisphosphate (PIP(2)) did not translocate to the phagocytic cup. Moreover, an enhanced yellow fluorescent protein (EYFP)-cPLA(2)alpha and an enhanced cyan fluorescent protein-pleckstrin homology (PH) domain of the phospholipase Cdelta1 (PLCdelta(1)) construct that specifically recognizes endogenous PIP(2) in the cells both localized at the same sites on the phagosome. High cellular expression of the PH domain inhibited EYFP-cPLA(2)alpha translocation. On the other hand, group V-secreted phospholipase A(2) and group VIA calcium-independent phospholipase A(2) were also studied, but the results indicated that neither of these translocated to the phagosome. Collectively, these data indicate that the polybasic cluster of cPLA(2)alpha (Lys(488)/Lys(541)/Lys(543)/Lys(544)) regulates the subcellular localization of the enzyme in intact cells under physiologically relevant conditions.

Phospholipase A 2 (PLA 2 ) enzymes cleave membrane phospholipids at the sn-2 position of the glycerol backbone, releasing a free fatty acid and a lysophospholipid ( 1 ). One of the better studied roles of PLA 2 enzymes is their involvement in infl ammatory processes due to their ability to liberate free arachidonic acid (AA) from membrane phospholipids (2)(3)(4)(5). Free AA will in turn be oxygenated by specifi c enzymes to generate a wide variety of compounds with potent pro-and anti-infl ammatory actions, collectively called the eicosanoids ( 6,7 ). Some of these compounds have also been described as important bactericidal agents ( 8 ).
Of the many existing PLA 2 enzymes, only one, the group IVA PLA 2 [also known as cytosolic phospholipase A 2 ␣ (cPLA 2 ␣ )], manifests specifi city for AA-containing phospholipids ( 9 ). Today, it is widely accepted that cPLA 2 ␣ is the key enzyme in AA release leading to physiological and/or pathophysiological eicosanoid production (9)(10)(11). Structurally, cPLA 2 ␣ possesses a C2 domain containing the binding site for Ca 2+ [and a binding site for ceramide-1 phosphate as well ( 12 )] and a catalytic domain where the residues involved in enzymatic activity are located ( 9 ). The catalytic domain also contains a cluster of cationic amino acids (Lys 488 , Lys 541 , Lys 543 , and Lys 544 ) that may mediate Abstract Group IVA cytosolic phospholipase A 2 ␣ (cPLA 2 ␣ ) plays a role in the microbicidal machinery of immune cells by translocating to phagosomes to initiate the production of antimicrobial eicosanoids. In this work, we have studied the involvement of the cationic cluster of cPLA 2 ␣ (Lys 488 / Lys 541 /Lys 543 /Lys 544 ) in the translocation of the enzyme to the phagosomal cup in human macrophages responding to opsonized zymosan. Phagocytosis was accompanied by an increased mobilization of free arachidonic acid, which was strongly inhibited by pyrrophenone. In transfected cells, a catalytically active enhanced green fl uorescent protein-cPLA 2 ␣ translocated to the phagocytic cup, which was corroborated by frustrated phagocytosis experiments using immunoglobulin G-coated plates. However, a cPLA 2 ␣ mutant in the polybasic cluster that cannot bind the anionic phospholipid phosphatidylinositol 4, 5-bisphosphate (PIP 2 ) did not translocate to the phagocytic cup. Moreover, an enhanced yellow fl uorescent protein (EYFP)-cPLA 2 ␣ and an enhanced cyan fl uorescent protein-pleckstrin homology (PH) domain of the phospholipase C ␦ 1 (PLC ␦ 1 ) construct that specifi cally recognizes endogenous PIP 2 in the cells both localized at the same sites on the phagosome. High cellular expression of the PH domain inhibited EYFP-cPLA 2 ␣ translocation. On the other hand, group V-secreted phospholipase A 2 and group VIA calcium-independent phospholipase A 2 were also studied, but the results indicated that neither of these translocated to the phagosome. Collectively, these data indicate that the polybasic cluster of cPLA 2 ␣ (Lys  Abbreviations: AA, arachidonic acid; BEL, bromoenol lactone; bis-BODIPY FL C11-PC, 1,2-bis-(4,4-difl uoro-5,7-dimethyl-4-bora-3a, 4a-diaza-s indecane-3-undecanoyl)-sn -glycero-3-phosphocholine; cPLA 2 ␣ , group IVA cytosolic phospholipase A 2 ; ECFP, enhanced cyan fl uorescent protein; EGFP, enhanced green fl uorescent protein; EYFP, enhanced yellow fl uorescent protein; iPLA 2 , calcium-independent phospholipase A 2 ; PIP 2 , phosphatidylinositol 4,5-bisphosphate; PLA 2 , phospholipase A 2 ; PLC ␦ 1 , phospholipase Cdelta1; PLC ␦ 1 -PH, pleckstrin homology domain of PLC ␦ 1 ; sPLA 2 , secreted phospholipase A 2 ; sPLA 2 -V, group V secreted phospholipase A 2.
1 To whom correspondence should be addressed. e-mail: jbalsinde@ibgm.uva.es; mbalboa@ibgm.uva.es decreases AA release in cells activated by serum, pointing to an important role for that cluster in cPLA 2 activation under physiological stimulation ( 18 ). Phagocytosis is a specialized process of cells of the innate immune system to engulf invading microorganisms, foreign particles, and apoptotic cells and debris. Phagocytosis is initiated at the site of particle attachment, producing a polarized region within the membrane that enlarges as a pseudopod to engulf the particle and drive it into the cytoplasm where it will be degraded ( 19 ). It has been previously shown that cPLA 2 ␣ translocates to the phagosomal cup during the ingestion of particles, and such a translocation appears to be important for in situ eicosanoid generation and effi cient elimination of microbes (20)(21)(22)(23). enzyme binding to anionic phospholipids in membranes ( 13 ). In fact, in vitro, anionic phospholipids such as phosphatidylinositol 4,5-bisphosphate (PIP 2 ), PI(3,4,5)P 3 , and PI(3,4)P 2 strongly increase the cPLA 2 ␣ specifi c activity when incorporated into the vesicle substrate ( 14,15 ). Mutation of the aforementioned polybasic 4-Lys cluster eliminates the activating effect of PIP 2 on cPLA 2 ␣ activity ( 13 ). Also, by plasmon resonance experiments, it has been demonstrated that PIP 2 activates cPLA 2 ␣ by increasing the catalytic effi ciency of the enzyme as a result of increasing membrane penetration ( 16 ). In live cells, exogenous PIP 2 and PI(3,4)P 2 induce the translocation of cPLA 2 ␣ to the intracellular membranes ( 17 ). More recently, it has been described that mutation of the Lys 488 , Lys Fluorescence arising from the EGFP-cPLA 2 ␣ (green) or zymosan (red) is shown on the left column. Middlecolumn panels show a pseudocolored fl uorescence intensity from the EGFP-cPLA 2 ␣ . Panels on the right column show a detailed amplifi cation of the phagosomes framed in the middle panels. Scale bar = 10 m.

Cells
Human macrophages were obtained from Buffy coats of healthy volunteer donors obtained from the Centro de Hemoterapia y Hemodonación de Castilla y León (Valladolid, Spain). Briefl y, blood cells were diluted 1:1 with PBS, layered over a cushion of Ficoll-Paque and centrifuged at 750 g for 30 min. The mononuclear cellular layer was then recovered and washed three times with PBS, resuspended in RPMI supplemented with 2 mM L-glutamine, 40 g/ml of gentamicin, and allowed to adhere to plastic in sterile dishes for 2 h. Nonadherent cells were then removed by extensively washing with PBS. Macrophage differentiation was achieved by incubating the adhered monocytes in RPMI supplemented with 2 mM L-glutamine, 40 g/ml of gentamicin, and 5% human serum for 2 weeks in the absence of exogenous cytokine mixtures.

Plasmid transfection
Human macrophages were transfected by the Nucleofection technique (Amaxa), following the kit specifi cations for human macrophages. Briefl y, cells were harvested by treatment with trypsin for 90 min and then by gentle scraping. After washing, the cells were resuspended in 100 l Human Macrophage Nucleofector solution plus 5 g of plasmid. Nucleofection was carried out using the program Y-010 and the cells were resuspended in 400 l of macrophage serum free medium (GIBCO) plus 5% heat inactivated human serum.

Synchronized phagocytosis
These experiments were carried out as described elsewhere with minor modifi cations ( 27 ). Macrophages were seeded over glass coverslips, allowed to adhere, and then washed with RPMI and resuspended in this medium. Cells were then kept at 4°C for 5 min and opsonized zymosan was then added. After 15 min incubation, coverslips were transferred to plates with RPMI at 37°C and the phagocytosis was allowed to proceed for different periods of time. Reactions were stopped by fi xation with 3% paraformaldehyde and 3% sacarose for 15 min if they were to be analyzed by microscopy. After three washes with PBS, the coverslips were mounted in glass slides with antifade medium. Samples were then analyzed by epifl uorescent or confocal microscopy.

Confocal microscopy
Transfected cells were seeded in MatTek dishes and allowed to adhere for 24 h in RPMI supplemented with 2 mM L-glutamine, 40 g/ml of gentamicin, and 5% human serum. Medium was then changed by HBSS with 10 mM HEPES and 1.3 mM CaCl 2 and fl uorescence was monitored by confocal a Bio-Rad Radiance 2100 laser-scanning system coupled to a Nikon TE-2000U with a thermostatized chamber (Warner Instruments). The objective was CFI Plan Apo 60×, 1.4 numerical aperture, oil immersion. The fl uorescence of enhanced cyan fl uorescent protein (ECFP) was monitored at 457 nm argon excitation using the combination of a long pass barrier fi lter HQ470LP and a short pass fi lter HQ520SP. The fl uorescence of EGFP was monitored at 488 nm Argon excitation using the combination of a long pass fi lter HQ500LP and a short pass fi lter HQ560SP. The EYFP was monitored at 514 nm Argon excitation and the fi lters HQ520LP and HQ560 SP. The Alexa Fluor 594 fl uorescence was monitored at 543 nm HeNe excitation using a long band pass fi lter HQ570LP.

AA release
Cells were labeled with 0.5 Ci/ml of [ 3 H]AA overnight. Afterward, the cells were washed three times with PBS supplemented with 0.5 mg/ml fatty acid free-BSA and incubated with RPMI supplemented with 0.5 mg/ml fatty acid-free BSA and 1% ITS. Cells were then stimulated and supernatants were removed at different time periods. Cells monolayers were overlaid with ice-cold phosphate buffer containing 0.05% Triton X-100 and scraped. Radioactivity was quantifi ed by liquid scintillation counting and AA release in the supernatants was referred to total radioactivity for each condition.

Frustrated phagocytosis
This was carried out as described by Marshall et al. ( 28 ). Briefl y, human macrophages were dislodged from the culture plates by trypsin treatment, and were resuspended in RPMI containing 10 mM HEPES and 2 mM EDTA. Cells were gently stirred for 2-3 h to allow for the reexpression of membrane receptors. Cells were then washed, resuspended in RPMI containing 10 mM HEPES and 2 mM MgCl 2 , and plated over MatTek dishes treated or not with 10 mg/ml pure IgG. After 30 min at 37°C, the cells were monitored by confocal microscopy. Some pictures were obtained in the XZ axis to have a better view of the cell membranes attached to the glass. In some experiments, the cells were labeled with 5 M bis-BODIPY FL C 11 -PC for 30 min, washed twice, and processed for frustrated phagocytosis. Fluorescence was monitored by confocal microscopy using 488 argon excitation and the combination of a HQ500 long band pass fi lter and HQ560 short band pass fi lter.

Translocation of cPLA 2 ␣ to the phagosome in zymosan-stimulated human macrophages
Human macrophages are capable of recognizing yeastderived zymosan particles and engulf them. By using transfected cells, we detected the translocation of a chimeric construct EGFP-cPLA 2 ␣ from the cytosol to the phagocytic cups ( Fig. 1 ). Translocation of the enzyme was particularly prominent in nonsealed phagocytic cups ( Fig.  1, 10-15 min). Once the phagosome was sealed and internalized, the EGFP-cPLA 2 ␣ separated from it ( Fig. 1 , 60  min). Experiments were performed next to rule out the possibility that the increased fl uorescence arising from EGFP-cPLA 2 ␣ in the forming phagosome was due to an increased volume of cytoplasm imaged in the plane. This was addressed by imaging in the same cell the construct EGFP-cPLA 2 ␣ and a monomeric form of DsRed (to correct for local variations in cytoplasmic volume). Figure 2 shows a clear increase in EGFP-cPLA 2 ␣ fl uorescence in the phagosomes that does not correspond with an increase in cytoplasmic volume (ratio EGFP-cPLA 2 ␣ /DsRed) at any time, thus indicating that enzyme translocation actually occurs.

Phagocytosis of opsonized zymosan activates cPLA 2 ␣ in human macrophages
Zymosan induced a signifi cant release of AA (and metabolites) to the extracellular medium, suggesting the activation of a PLA 2 ( Fig. 3A ). This PLA 2 was identifi ed as cPLA 2 ␣ on the basis of complete inhibition of the response by a low concentration of pyrrophenone, a cPLA 2 ␣ inhibitor ( 29 ) ( Fig. 3B ).
To confi rm that the cPLA 2 ␣ that translocates to the phagocytic cup is functionally active, an experiment of frustrated phagocytosis was performed, using glass plates coated with IgG ( 28 ). Macrophages exposed to IgGcoated glass surfaces responded by translocating the EGFP-cPLA 2 ␣ to the membranes more proximal to the glass surface but not to other cellular membranes ( Fig.  4A , B, E ). In this experiment, the IgG-coated glass would represent the phagocytosable particle ( 28 ). Next, we loaded the cells with the fl uorogenic phospholipase substrate bis-BODIPY FL C11-PC ( 30 ), and subjected them to the frustrated phagocytosis assay. A dramatic increase in fl uorescence was observed, especially in the proximity of the IgG-coated glass ( Fig. 4C, D and F ), indicating that an A-type phospholipase is acting at that place (where the "phagosome" is being initiated). To confi rm that such a phospholipase is actually cPLA 2 ␣ , we conducted experiments in the presence of pyrrophenone ( Fig. 4D and F ). As expected, pyrrophenone at doses as low as 1 µM strongly blocked the fl uorescence increase in the  at 15 and 30 min). B: Cells transfected with the construct EGFP-cPLA 2 ␣ or EGFP-4KE/A-cPLA 2 ␣ were subjected to a frustrated phagocytosis assay and plated over noncoated glasses (Ctrl) or IgG-coated glasses (IgG) as indicated, and analyzed by confocal microscopy. Pictures of the XZ axis of the cells were taken and the intensity of the fl uorescence obtained was analyzed in pseudocolor. The fi gure is representative of more than 40 cells that were analyzed per experiment and the experiment was repeated four times. White bar = 10 m. cells exposed to IgG-coated glass, indicating that such a fl uorescence increase is mainly due to cPLA 2 ␣ activation.

Mutation of a four-lysine cluster in cPLA 2 ␣ that is involved in binding of anionic phospholipids suppresses EGFP-cPLA 2 ␣ translocation to the phagocytic cup
We have previously described that mutations on Lys488, Lys541, Lys543, and Lys544 of cPLA 2 ␣ result in a defective translocation of cPLA 2 ␣ to intracellular membranes in response to exogenous PIP 2 ( 17 ). Because it is well described that PIP 2 increases in the phagosome ( 31 ), we next studied the behavior of the 4-Lys mutant (EGFP-4KE/A-cPLA 2 ␣ ) in macrophages exposed to opsonized zymosan. The results, as shown in Fig. 5A , clearly indicated that this mutant does not translocate to the phagocytic cup in activated cells at any time tested, suggesting that a functional binding site in cPLA 2 ␣ for anionic phospholipids is necessary for such a translocation to be observed. To further substantiate this observation, we performed experiments of frustrated phagocytosis, similar to those shown in Fig. 4 , using cells transfected with the EGFP-4KE/A-cPLA 2 ␣ mutant. Fig. 5B shows that, unlike the wild-type enzyme, the mutant did not translocate to the membranes that are closer to the IgG-coated plate. These data highlight the importance of the 4-Lys cluster for proper binding of cPLA 2 ␣ to the phagosomal membranes.
Experiments were next conducted to study a possible role for anionic phospholipids in cPLA 2 ␣ translocation to the phagocytic cup. To this end, we took advantage of the PIP 2 -binding properties of a fl uorescent chimeric protein of EGFP with the pleckstrin homology domain of the PLC ␦ 1 (EGF-PLC ␦ 1 -PH) ( 28 ). Figure 6A shows that, when transfected into human macrophages, the chimera labels the main cellular reservoir of PIP 2 , i.e., the plasma membrane ( 24 ). Immediately after promoting PIP 2 hydrolysis by activating the cells with a calcium ionophore, the fl uorescence disappears from the plasma membrane and accumulates in the cytoplasm as would be expected from a functional PH domain ( Fig. 6A ). By using this chimera, we confi rmed in human macrophages previous observations by Botelho et al. ( 31 ), indicating that PIP 2 levels increase at the phagocytic cup while the phagosome is being formed and decrease when it seals ( Fig. 6B ).
Subsequently, the cells were cotransfected with both constructs, namely ECFP-PLC ␦ 1 -PH and EYFP-cPLA 2 ␣ . In resting cells, ECFP-PLC ␦ 1 -PH was present only in the plasma membrane whereas EYFP-cPLA 2 ␣ was found primarily in the cytoplasm ( Fig. 7 ). However, after exposure of the cells to opsonized zymosan, both chimeric proteins localized at the forming phagocytic cups ( Fig. 7B ). There was no colocalization in the cytoplasm of resting or stimulated cells. We noticed also that in those cells where ECFP-PLC ␦ 1 -PH construct was expressed at higher levels the translocation of the cPLA 2 ␣ to the phagosomes was inhibited ( Fig. 7 C ).
The behavior of EGFP-PLC ␦ 1 -PH and EGFP-cPLA 2 ␣ were analyzed in more detail by confocal analysis of z-stack series from cells that were engulfi ng particles at similar stages of phagocytosis ( Figs. 8, 9 ). We observed that both constructs were enriched in the same sites of the phagosomes, especially along the particles and in the base. It is worth noting that translocation of the EGFP-cPLA 2 ␣ to the base of some of the phagosome was more prominent than that of EGFP-PLC ␦ 1 -PH. It is possible that in the base of the phagosome other factors, in addition to or independently of PIP 2, contribute to the translocation of cPLA 2 ␣ . transfected with either sPLA 2 -V-EGFP or iPLA 2 -VIA-EYFP. In agreement with our previous studies in murine macrophages ( 25,32 ), sPLA 2 -V-EGFP was found in resting cells associated with secretory granules and Golgi-like structures ( Fig. 10A ). iPLA 2 -VIA-EYFP had a mitochondrial localization in unstimulated cells ( Fig. 10B ), which is in accordance with previous estimates ( 33 ). After the cells where challenged with opsonized zymosan, localization was studied between the chimeric proteins and the fl uorescent particles. We failed to detect association of the fl uorescence arising from either sPLA 2 -V-EGFP or iPLA 2 -VIA-EYFP to the phagocytic cups under any condition ( Fig. 10 ). Thus, of the three major PLA 2 classes potentially capable of generating lipid mediators during infl ammation (2)(3)(4)(5)34 ), only the cPLA 2 ␣ translocates to the phagocytic cup in human macrophages.

DISCUSSION
A major regulatory mechanism of cPLA 2 ␣ activity in cells is the Ca 2+ -dependent control of the physical state of the enzyme. In resting cells, the enzyme resides in the cytosol and hence, having no access to its substrate in the membrane, has no activity. In stimulated cells, cPLA 2 ␣ translocates to the membrane in a Ca 2+ -dependent process resulting in phospholipid hydrolysis and free AA release ( 9 ). Most of the research carried out to date has thoroughly documented the translocation of cPLA 2 ␣ to intracellular membranes such as those of the nuclear envelope, Golgi complex, or endoplasmic reticulum ( 9 ). A few instances have been reported of translocation of cPLA 2 ␣ to cellular membranes different to those indicated above ( 20,35,36 ). This study adds to these studies by showing that the phagosomal membrane of human macrophages is also a site for cPLA 2 ␣ translocation during activation conditions. Particularly relevant to our report is the work by Girotti et. al. ( 20 ) showing translocation of cPLA 2 ␣ to forming phagosomes in murine macrophages. Although the molecular mechanism was not investigated, the authors noted that chelation of extracellular calcium decreased the total number of phagocytic events, but cPLA 2 ␣ still remained associated to the lasting phagosomes at intracellular calcium levels equaling those of resting cells ( 20 ). In previous work from our laboratory, we showed that introducing exogenous shortchain PIP 2 into cells promotes the translocation of cPLA 2 ␣ from cytosol to perinuclear membranes at basal levels of intracellular Ca 2+ ( 17 ). Moreover, other studies have demonstrated that PIP 2 transiently increases in the forming phagosome and disappears after phagosome sealing, being undetectable in the sealed phagosome ( 31 ). Intrigued by these observations, we speculated that translocation of cPLA 2 ␣ to the forming phagosome could require the cationic cluster of four Lys (Lys488/Lys541/ Lys543/Lys544) that is present in cPLA 2 ␣ and has been shown to bind anionic phospholipids such as PIP 2 ( 13 ). Using confocal microscopy techniques, we document the localization of cPLA 2 ␣ and a PIP 2 -binding construct,

Secreted group V and cytosolic calcium-independent group VIA PLA 2 s do not translocate to the phagocytic cup
To address whether other PLA 2 s in addition to cPLA 2 ␣ could also translocate to the phagocytic cup during phagocytosis of opsonized zymosan, we used human macrophages Fig. 7. Localization of ECFP-PLC ␦ 1 -PH and EYFP-cPLA 2 ␣ during phagocytosis in human macrophages . Human macrophages, cotransfected with the ECFP-PLC ␦ 1 -PH and the EYFP-cPLA 2 ␣ constructs, were subjected to synchronized phagocytosis using Alexa Fluor 594-labeled opsonized zymosan, fi xed at 0 (Control) and 15 min, and analyzed by confocal microscopy (A). B: A cell with high expression of the construct ECFP-PLC ␦ 1 -PH is shown. In A and B, middle panels, fl uorescence intensities are shown in pseudocolor, and detailed fl uorescence in forming phagosomes is shown in panels to the right.
ECFP-PLC ␦ 1 -PH, in forming phagosomes. Also, high level expression of the ECFP-PLC ␦ 1 -PH inhibits the translocation of the cPLA 2 ␣ to phagosomes. Moreover, mutation of the cationic cluster of Lys488, Lys541, Lys543, and Lys544 that serves as a PIP 2 -binding site eliminates the ability of the enzyme to translocate to the phagosomal membrane. Collectively, these results provide the fi rst example of a physiologically-relevant condition where the cationic cluster of cPLA 2 ␣ participates in regulating enzyme association to membranes and hence, its activity in intact cells.
The presence of cPLA 2 ␣ at the membrane of the nascent phagosome may serve important pathophysiological roles because during phagocytosis large quantities of eicosanoids are produced, which could be involved in the kill-ing of the ingested microorganism at the phagosome ( 8,(21)(22)(23). In the mouse model, other PLA 2 s in addition of cPLA 2 ␣ have been suggested to be related to phagocytic events. Particularly relevant to this work, it has been shown that group V secreted PLA 2 also translocates to the phagosome in zymosan-stimulated murine peritoneal macrophages ( 23 ). Intrigued by this report, we also studied the possible movement of sPLA 2 -V to phagosomes in our human macrophage cell system. We failed to detect translocation of a chimeric sPLA 2 -V-EGFP protein to phagosomes in our systems. We had previously demonstrated that this chimeric sPLA 2 -V behaves the same as the native sPLA 2 -V protein in terms of biochemical properties, enzymatic activity, and subcellular localization ( 25,32 ). There are many differences between the murine and human   macrophage models of zymosan phagocytosis that may explain these different results. For instance, in murine macrophages, zymosan induces abundant AA release (37)(38)(39)(40)(41)(42)(43) and may be internalized primarily via dectin-1 receptors ( 44 ). However, human macrophages do not respond readily to zymosan by releasing AA, and opsonization of the particle appears to be required for full responses, which may occur primarily via Fc receptors. Thus, it appears likely that the different mechanisms of internalization of phagocytosable particles in mouse versus human may account, at least in part, for the remarkable differences in the translocation ability of sPLA 2 -V to the phagosomes.
The one other PLA 2 that we investigated is iPLA 2 -VIA, a calcium-independent enzyme. Unlike cPLA 2 ␣ and sPLA 2 enzymes, the involvement of iPLA 2 -VIA in receptormediated AA mobilization appears not to be a general one but to depend on cell type and stimulation conditions ( 2,(45)(46)(47)(48). Using a chimeric construct, iPLA 2 -VIA-EYFP, we detected no appreciable change in the subcellular localization of this enzyme during opsonized zymosan challenge; the enzyme always remained associated to mitochondria, which is consistent with previous data ( 33 ). Thus, from the three major PLA 2 families potentially capable of effecting AA release for eicosanoid production, only one, cPLA 2 ␣ , translocates to the phagosome in human cells.