Calcium-Independent Phospholipase A2 Is Required for Lysozyme Secretion in U937 Promonocytes1

As a part of their surveillance functions in the immune system, monocytes/macrophages secrete large amounts of the bactericidal enzyme lysozyme to the extracellular medium. We report here that lysozyme secretion in activated U937 promonocytes depends on a functional calcium-independent phospholipase A2 (iPLA2). Inhibition of the enzyme by bromoenol lactone or by treatment with a specific antisense oligonucleotide results in a diminished capacity of the cells to secrete lysozyme to the extracellular medium. Calcium-independent PLA2 is largely responsible for the maintenance of the steady state of lysophosphatidylcholine (lysoPC) levels within the cells, as manifested by the marked decrease in the levels of this metabolite in cells deficient in iPLA2 activity. Reconstitution experiments reveal that lysoPC efficiently restores lysozyme secretion in iPLA2-deficient cells, whereas other lysophospholipids, including lysophosphatidic acid, lysophosphatidylserine, and lysophosphatidylethanolamine, are without effect. Arachidonic acid mobilization in activated U937 cells is under control of cytosolic phospholipase A2 (cPLA2). Selective inhibition of cPLA2 results in a complete abrogation of the arachidonate mobilization response, but has no effect on lysozyme secretion. These results identify iPLA2-mediated lysoPC production as a necessary component of the molecular machinery leading to lysozyme secretion in U937 cells and rule out a role for cPLA2 in the response. Collectively, the results demonstrate distinct roles in inflammatory cell signaling for these two intracellular phospholipases.

L ysozyme degrades bacterial cell walls of Gram-positive bacteria and the chitinous components of fungal cell walls. The enzyme occurs in many body fluids, such as tears, saliva, and mucus, and is produced and secreted by phagocytic cells and a variety of cells of epithelial origin (1). Stimuli that induce lysozyme secretion from phagocytic cells also induce the phospholipase A 2 (PLA 2 ) 3 -mediated mobilization of free arachidonic acid (AA). Whether these two responses are causally related has been the subject of recent research (2)(3)(4). PLA 2 enzymes are frequently classified into three main classes on the basis of whether the enzyme is secreted (sPLA 2 ), cytosolic Ca 2ϩ -dependent (cPLA 2 ), or cytosolic Ca 2ϩ -independent (iPLA 2 ) (5,6). The sPLA 2 s are low molecular mass, secreted enzymes that require millimolar concentrations of calcium for their catalytic activity and do not show fatty acid selectivity (7). The cPLA 2 is an 85-kDa protein that requires nanomolar to micromolar concentrations of calcium, is specific for AA-containing phospholipids, and appears to play a crucial role in agonist-induced AA mobilization (8). The iPLA 2 has been shown to play in important role in reg-ulating basal phospholipid deacylation/reacylation reactions in phagocytic cells (9).

Treatment of the cells with antisense oligonucleotides
The antisense oligonucleotides used in these studies were derived from prior publications reporting their effects (16 -18). The iPLA 2 antisense sequence corresponded to nt 59 -78 in the murine group VI iPLA 2 sequence, which is conserved in human group VI iPLA 2 (19,20). The antisense Institute of Molecular Biology and Genetics, University of Valladolid School of Medicine, Valladolid, Spain sequence was 5Ј-CTC CTT CAC CCG GAA TGG GT-3Ј. As a control, the iPLA 2 sense sequence was 5Ј-ACC CAT TCC GGG TGA AGG AG-3Ј. Phosphorotioate-modified oligonucleotides were used to limit degradation. The antisense and sense oligonucleotides were mixed with Lipofectamine (Invitrogen, Carlsbad, CA), and complexes were allowed to form at room temperature for 15-30 min. The complexes were then added to the cells, and the incubations were allowed to proceed under standard cell culture conditions. The final concentrations of oligonucleotide and Lipofectamine were 1 M and 10 g/ml, respectively. Oligonucleotide treatment and culture conditions were not toxic for the cells as assessed by the trypan blue dye exclusion assay and by quantitating adherent cell protein.

Lysozyme release assay
The cells were stimulated with PMA, Con A, or platelet-activating factor (PAF) for the indicated times. After centrifugation, the supernatants were removed, and the cell pellets were overlaid with 1 ml of PBS and homogenized. Lysozyme in the supernatants and the cell pellets was measured spectrophotometrically as follows. Briefly, 1 ml of sample was mixed with 1 ml of a Micrococcus lysodeikticus suspension (0.3 mg/ml in 0.1 M sodium phosphate buffer, pH 7.0). The decrease in absorbance at 450 nm was measured at room temperature. A calibration curve was constructed using chicken egg white lysozyme as a standard. Lysozyme release is expressed according to the formula: (S/S ϩ P) ϫ 100, where S is the amount of lysozyme measured in the supernatant, and P is the amount of lysozyme measured in the cell pellets.

Measurement of AA release
The cells were labeled with 0.5 Ci/ml [ 3 H]AA for 18 h. After this period, the cells were washed an placed in serum-free medium for 1 h before the addition of 100 nM PMA in the presence of 0.5 mg/ml BSA. The supernatants were removed, cleared of cells by centrifugation, and assayed for radioactivity by liquid scintillation counting.

Lysophospholipid determination
Cells labeled with 0.5 Ci/ml [ 3 H]choline or [ 14 C]ethanolamine for 2 days were used. After the incubations, lipids were extracted with ice-cold nbutanol and separated by TLC with chloroform/methanol/acetic acid/water (50/40/6/0.6) as a solvent system. Spots corresponding to lysophosphatidylcholine (lysoPC) or lysophosphatidylethanolamine (lysoPE) were scraped into scintillation vials, and the amount of radioactivity was estimated by liquid scintillation counting (16).

Data presentation
Assays were conducted in duplicate or triplicate. Each set of experiments was repeated at least three times with similar results. Unless otherwise indicated, the data presented are from representative experiments.

Lysozyme release in U937 cells
Lysozyme is one of the proteases involved in nonspecific immune defense against bacterial infection. Fig. 1 shows that the U937 cells released large quantities of lysozyme to the incubation medium when activated with the phorbol ester PMA. More than 60% of total enzyme was found in the supernatant after 2-h incubation with PMA. Fig. 1 also shows that lysozyme secretion was strongly blunted by the PLA 2 inhibitors BEL and MAFP, suggesting the possible involvement of iPLA 2 and/or cPLA 2 in this response.
To test the above suggestion more rigorously, the effect of an iPLA 2 antisense oligonucleotide on lysozyme secretion was evaluated. The antisense oligonucleotide used is the human counterpart of the murine one that we and others have successfully employed previously (16 -18). In these experiments this antisense produced a 70-80% decrease in both immunoreactive iPLA 2 protein and cellular iPLA 2 activity (data not shown, but see Ref. 21). Under these conditions, lysozyme release in PMA-treated cells was strongly inhibited (Fig. 2), thus providing strong evidence for the involvement of iPLA 2 in this process.
The cPLA 2 inhibition was achieved by the use of pyrrophenone, a highly selective inhibitor of cPLA 2 vs iPLA 2 in cells (15,22). Pyrrophenone exerted no significant effect on the release of lysozyme ( Fig. 1).
Studies were conducted next to evaluate the effect of iPLA 2 inhibition on lysozyme release in response to the receptor-mediated agonists Con A and PAF. The significant release of lysozyme induced by both agonists was strongly abrogated by BEL (Fig. 3). These data indicate that iPLA 2 inhibition also leads to modulation of receptor-mediated lysozyme release in U937 promonocytes.

AA mobilization in activated U937 cells is not involved in lysozyme release
[ 3 H]AA-labeled U937 cells were stimulated with the phorbol ester PMA (100 nM) for different periods of time, and the release of radiolabel in the supernatant was measured. After a time lag of ϳ15 min, PMA-treated cells showed a modest, but significant, release of radiolabel (Fig. 4). This sustained release was completely blocked by MAFP, but was unaffected by BEL (Fig. 4), suggesting that, in agreement with previous data (13,21), cPLA 2 , not iPLA 2 , mediates AA release in activated U937 promonocytic cells.
The cPLA 2 -mediated AA release was sensitive to the unspecific kinase inhibitor quercetin (23). Fifty percent inhibition was observed at a quercetin concentration of 5 M, while higher concentrations of the inhibitor were required to inhibit lysozyme secretion (Fig. 5). The lack of correspondence between the concentrationresponse effects of quercetin on AA release and lysozyme secretion suggests that both responses are unrelated.

LysoPC levels are decreased in iPLA 2 -deficient cells
The iPLA 2 plays a major role in a number of cells in the regulation of basal phospholipid deacylation reactions by providing the bulk of lysoPC acceptors used in these reactions (9,24). Fig. 6A shows that U937 cells made deficient in iPLA 2 by antisense treatment exhibited considerably lesser amounts of lysoPC than control cells, and this was readily observable in control unstimulated cells as well as in PMA-stimulated U937 cells. Changes in cellular lysoPC due to activation with PMA were too small to be detected (not shown). No significant effect of the iPLA 2 antisense on cellular levels of lysoPE in [ 14 C]ethanolamine-labeled cells was detected.
Importantly, addition of lysoPC (25 M) restored lysozyme release in the iPLA 2 -deficient cells (Fig. 6B). Other lysophospholipids tested, i.e., lysoPE, lysophosphatidylserine, and lysophosphatidic acid were ineffective. LysoPAF was as effective as lysoPC in restoring lysozyme release (Fig. 6B). Together, these results suggest that it is the choline headgroup of the lysophospholipid that is necessary for biological activity under these settings. The dose response of the effect of lysoPC on the restoration of lysozyme    Fig. 6C. Significant effects of lysoPC were already observed at concentrations between 5-10 M. Lysophospholipid concentrations Ͼ25 M induced significant lysozyme release on their own and thus were not suitable for these reconstitution experiments.
Addition of free fatty acids such as AA, palmitic acid, or linoleic acid (up to 10 M) failed to restore lysozyme release in iPLA 2deficient U937 cells (not shown). Collectively, the results suggest that lysoPC, and not other putative PLA 2 -derived metabolites such as a free fatty acid or lysoPE, is required for U937 cells to release lysozyme to the incubation medium.

Studies of the regulation of iPLA 2 sctivity
Homogenates of U937 cells, either untreated or treated with PMA, were prepared, and assays were conducted to assess iPLA 2 activity using a vesicle substrate assay. Under these conditions we failed to detect any change in the iPLA 2 specific activity of homogenates from PMA-treated cells vs untreated cells. As an alternative approach, iPLA 2 was measured using the mammalian membrane assay system described by Diez and colleagues (25). We have previously used this system to detect iPLA 2 activity changes in homogenates from H 2 O 2 -treated U937 cells (21). In this system, purified [ 3 H]AA-labeled mammalian membranes are used as a substrate. Using this assay, again no differences in the iPLA 2 activity of untreated cells vs that of PMA-treated cells could be demonstrated.

Discussion
The current study addresses the possible involvement of two intracellular PLA 2 s in lysozyme secretion in activated U937 cells. In particular, the data identify iPLA 2 as an important player in the secretion process and rule out a significant role for cPLA 2 . These conclusions are based on two different approaches that yield cells deficient in either iPLA 2 activity or cPLA 2 activity, namely the use of chemical inhibitors and antisense oligonucleotides. Thus, experimental approaches leading to the blockade of cellular iPLA 2 result in abrogation of the cell's capacity to secrete lysozyme. On the contrary, strategies that lead to the blockade of cPLA 2 activity do not significantly affect lysozyme release.
The iPLA 2 appears to play an important role in regulating phospholipid fatty acid recycling in a variety of cell types by providing the lysophospholid acceptors used in the acylation reaction (9). Thus, inhibition of either iPLA 2 activity by chemical inhibitors or iPLA 2 expression by antisense oligonucleotides results in a marked decrease in the steady state levels of lysoPC, the primary acceptor of free AA for its incorporation into membrane phospholipids (24,26). The contribution of iPLA 2 to the steady state level of cellular lysoPC appears to largely depend on cell type. Thus, it is estimated that the iPLA 2 contribution to cellular lysoPC levels ranges from as much as 80 -90% in rat submandibular ductal cells (27) to 50 -60% in phagocytic cells (17,28,29) and to 20 -25% in rat uterine stromal cells and rat pancreatic islets (30,31). We confirmed that in common with the aforementioned cell types, U937 cells deficient in iPLA 2 also exhibit significantly lower lysoPC levels. Importantly, the low lysoPC level found in iPLA 2 -deficient cells appears to be related to the diminished capacity of these cells to secrete lysozyme after activation. This conclusion is based on the finding that exogenous supplementation with lysoPC fully restores the capacity of the cells to release lysozyme to the extracellular medium. Interestingly, iPLA 2 depletion does not substantially change the steady state level of lysoPE, and exogenous lysoPE lacks the ability to restore lysozyme secretion in iPLA 2 -deficient cells, indicating that the lysoPC effect is specific. In support of this contention, structure-function relationship studies revealed that the choline headgroup appears to be important for the biological activity of lysoPC, since lysoPAF was the only lysophospholipid tested that was able to restore lysozyme release. In turn, these studies indicate that the type of linkage present at the sn-1 position of the lysophospholipid is unimportant.
LysoPC is a natural amphiphile; it incorporates into membranes and affects membrane fluidity and permeability (32)(33)(34). For instance, lysoPC, but not lysoPE, enhances the exocytosis of ram spermatozoa treated with Ca 2ϩ (35), which is in accord with the results of this study. Thus, we hypothesize that the continuing iPLA 2 -mediated phospholipid fatty acid recycling of membranes and concomitant generation of choline-containing lysophospholipid are important for secretion to take place.
It is interesting to note that surface receptors for lysoPC that mediate some of the biological actions of this phospholipid have recently been described (36). Signaling through lysoPC receptors involves rapid activation of the mitogen-activated protein kinase cascade as well as increased intracellular Ca 2ϩ levels (36). However, exogenous lysoPC neither activates the mitogen-activated protein kinase cascade nor increases the intracellular Ca 2ϩ levels in U937 cells (M. Balboa and J. Balsinde, unpublished observations), thus ruling out a role for lysoPC receptor signaling in the current study.
It was somewhat unexpected that exogenous AA and other fatty acids failed to restore lysozyme secretion in iPLA 2 -deficient cells. This suggests that stimulus-triggered increases in free AA levels have no role in regulating lysozyme secretion. In keeping with this observation the data have shown that both responses can be dissociated by the use of the unspecific kinase inhibitor quercetin. Moreover, AA mobilization in the activated U937 cells appears to be under the control of cPLA 2 , an enzyme that has no role in lysozyme secretion. In turn, iPLA 2 plays no role in stimulus-induced AA mobilization in these cells. It is interesting to note that cPLA 2 activation transiently elevates cellular lysoPC levels in activated cells (37). Since cPLA 2 plays no discernible role in lysozyme release, it is the steady state level of lysoPC (iPLA 2 -mediated), not the transient increases in this metabolite that occur as a consequence of cellular activation (cPLA 2 -mediated), that are important for lysozyme secretion. In agreement with this observation, no changes in the specific activity of iPLA 2 were detected after activation of the cells with PMA.
An interesting aspect of these results, showing the importance of lysoPC in lysozyme secretion, is the possibility of functional redundancy with exogenous secreted PLA 2 s that might act on the lysozyme-secreting cells in a paracrine fashion. Certain PLA 2 forms that are secreted by immunoinflammatory cells are able to attack the outer membrane phosphatidylcholine very well. These include group X PLA 2 (38,39) and also group V PLA 2 (40). The temporal accumulation of lysoPC at discrete sites on the plasma membrane induced by these enzymes might facilitate the exocytotic process Recently, two important cellular functions that, like enzyme secretion, require profound membrane rearrangement, have been suggested to involve the participation of the iPLA 2 . These are chemotaxis (16) and cell spreading (41). Coincident with the results of this study, it was the constitutive activity of the iPLA 2 that was found to be necessary to sustain both these functions, and in addition, the contribution of iPLA 2 was dissociated from cPLA 2 activation (16,41). Collectively, these studies underscore the importance of iPLA 2 in regulating processes that require changes in membrane phospholipid homeostasis and support the growing idea that the iPLA 2 and the cPLA 2 play separate and often unique roles in inflammatory cell signaling.