Inhibition of monocyte chemotaxis to C-C chemokines by antisense oligonucleotide for cytosolic phospholipase A2.

Monocyte chemotactic protein (MCP)-1, a member of the C-C (or beta) branch of the chemokine superfamily, at chemotactic concentrations, induced a rapid release of [3H]arachidonic acid but not of [14C]oleic acid from prelabeled human monocytes. This effect was associated with an increase in the intensity of the immunoreactive band corresponding to the phosphorylated form of cytosolic phospholipase A2, (cPLA2). To address the role of cPLA2 in the induction of monocyte chemotaxis, cells were treated with a specific antisense oligonucleotide. Monocytes cultured in the presence of 10 microM antisense oligonucleotide for 48 h showed a marked decrease (57 +/- 5%; n = 4) of cPLA2 expression, as evaluated by Western blot analysis and a nearly complete inhibition (81.8 +/- 4.2%; n = 3) of [3H]arachidonic acid release in MCP-1-stimulated cells. Monocyte chemotaxis in response to MCP-l also was inhibited in a concentration-dependent manner by cPLA2 antisense oligonucleotide (IC50 = 1.9 +/- 1.1 microM; n = 3), with complete inhibition observed between 3 and 10 microM. No inhibition of chemotactic response was observed in monocytes treated with a control oligonucleotide. Monocyte migration in response to MCP-3, RANTES (regulated on activation normal T cells expressed and secreted), and MIP-1 alpha/LD78 also was inhibited (>70%) in antisense oligonucleotide-treated cells. On the contrary, the chemotactic response elicited by formyl-methionyl-leucyl-phenylalanine and C5a, two "classical" chemotactic agonists, was minimally affected (<20%) by antisense oligonucleotide treatment. These data show that cPLA2 plays a major role in [3H]arachidonic acid release by MCP-1 in human monocytes and provide direct evidence for the involvement of cPLA2 in C-C chemokine-induced monocyte chemotaxis.

The recruitment of leukocytes from the blood compartment to the site of inflammation represents one of the characteristic elements of the inflammatory process (1). Locally produced chemotactic agonists are believed to play a crucial role in the "multistep paradigm" of leukocyte accumulation in tissues (2,3).
In the past few years a new superfamily of chemotactic cytokines, named chemokines, was described. The hallmark of this family is a four conserved cysteine motif (4 -7). According to the relative position of the first two cysteines it is possible to distinguish two families: the C-X-C (or ␣) chemokines, active on neutrophils and T lymphocytes (4 -7), and the C-C (or ␤) chemokines that exert their action on multiple leukocyte populations, including monocytes, basophils, eosinophils, T lymphocytes, natural killer, and dendritic cells (4 -9). Recently, a protein that may define a third family (the C or ␥ chemokines) was described. This protein is characterized by the absence of the first and third cysteines and is active on T lymphocytes (10). Chemokines, as well as classical chemotactic agonists, such as formylated peptides (of which fMLP 1 is the prototype) and C5a, bind to and activate a family of rhodopsin-like, GTPbinding protein-coupled seven-transmembrane domain receptors (11)(12)(13). Activated chemotactic receptors induce remodeling of membrane phospholipids by the action of phospholipases (C, D, and A 2 ) and these events ultimately lead to the induction of different biological responses: chemotaxis, activation of the oxidative burst, and release of lysosomal enzymes (12)(13)(14). The role of individual second messengers in the generation of different biological responses is still unclear.
In previous studies aimed at better clarifying the molecular bases for monocyte migration in response to monocyte chemotactic protein (MCP)-1, a prototypic C-C chemokine (8,(15)(16)(17), we reported that MCP-1 induces a rapid (Ͻ15 s) and transient (ϳ15 min) release of [ 3 H]arachidonic acid from labeled human monocytes (18,19). This effect was inhibited by Bordetella pertussis toxin treatment, was dependent on the influx of extracellular Ca 2ϩ , and was increased in a synergistic fashion by platelet-activating factor. Similar results were obtained with other proteins of the C-C chemokine family (e.g. MCP-3, RANTES, and MIP-1␣/LD78) (18 -20). In parallel, also the chemotactic response to MCP-1, MCP-3, RANTES, and MIP-1␣/LD78 were increased (18). These results, together with the finding that PLA 2 inhibitors block monocyte chemotaxis (18), suggest a role for arachidonic acid as a second messenger in monocyte migration to chemokines.
Cells of the monocytic lineage posses at least three different types of PLA 2 (21,22): a low molecular mass (ϳ14 kDa) secreted form that requires for its catalytic activity millimolar concentrations of Ca 2ϩ and does not show a selectivity for the fatty acid esterified at the sn-2 position (23, 24); a 85-kDa cytosolic PLA 2 (cPLA 2 ) that shows a certain degree of specific-ity for arachidonic acid and that translocates to the membrane fraction by a Ca 2ϩ (nanomolar)-dependent mechanism upon receptor stimulation (25)(26)(27); and a Ca 2ϩ -independent ATPregulated cytosolic PLA 2 that does not show a preference for the fatty acid at the sn-2 position (28). Because of the lack of specific inhibitors, the relative contribution of these enzymes in arachidonic acid metabolism in human monocytes is still uncertain.
In this paper we report that MCP-1-stimulated monocytes selectively released [ 3 H]arachidonic acid; no detectable release of [ 14 C]oleic acid was observed. This effect paralleled the phosphorylation of cPLA 2 evaluated by Western blot analysis. In addition, by the use of a specific antisense oligonucleotide, we show that cPLA 2 plays a crucial role in the chemotactic response of human monocytes to C-C chemokines. Human Monocytes Purification-Human monocytes were obtained from buffy coats of normal blood donors through the courtesy of Centro Trasfusionale Ospedale Sacco (Milan, Italy) and Centro Trasfusionale Ospedale Caduti Bollatesi (Bollate, Italy) as described previously (18). To reduce platelet contamination, monocytes were isolated according to the procedure described by Pawlowski et al. (29) with minor modifications (18). Briefly, anticoagulated whole blood was diluted 1:4 with cold phosphate-buffered isotonic saline without Ca 2ϩ and Mg 2ϩ (PBS; Life Technologies, Inc.) and centrifuged at 150 ϫ g at 4°C for 20 min. The supernatant was discarded, and cell pellet was washed in PBS in the same conditions. Cells were resuspended in PBS containing 0.3 mM EDTA (Merck, Darmstadt, Germany), layered on top of Ficoll (Biochrom, Berlin, Germany), and centrifuged at 800 ϫ g at room temperature for 25 min. Mononuclear cells were recovered, diluted, and washed twice in PBS at 4°C. To remove platelets specifically adherent to monocytes, mononuclear cells were resuspended in fetal calf serum (FCS; Hyclone, Logan, UT) containing 5 mM EDTA and subjected to two sequential incubations (15 min) at 37°C. Platelet-free mononuclear cells were recovered by centrifugation at 400 ϫ g at room temperature for 15 min. Monocytes were further purified (Ͼ90% pure) by centrifugation at 600 ϫ g on a 46% isoosmotic Percoll (Pharmacia Biotech Inc.) gradient, as described previously (30). The monocyte preparation obtained did not release [ 3 H]arachidonic acid when challenged with 10 units/ml thrombin (Sigma).
Release of Labeled Fatty Acids-Monocytes (10 6 /ml) were labeled with 1 Ci/ml [ 3 H]arachidonic acid (200 Ci/mmol) and/or 1 Ci/ml [ 14 C]oleic acid (60 mCi/mmol) (Amersham) during the last 18 h of oligonucleotide treatment. Incubation did not affect cell viability (Ͼ95%, by trypan blue dye exclusion) nor the ability of monocytes to migrate in response to MCP-1. At the end of the incubation, cells were washed twice and resuspended in RPMI 1640 medium supplemented with 0.2% fatty acid free bovine serum albumin (Sigma). Monocytes (10 7 /ml) were prewarmed at 37°C for 5 min and then stimulated. The reaction was terminated by the addition of 2 ml of chloroform/methanol/ formic acid (1:2:0.2, v/v/v) followed by agitation. Cell extracts were transferred to centrifuge tubes, and 1 ml of water and 2 ml of chloroform were added. Chromatographic separation of lipids was performed by evaporating the organic phase under a stream of nitrogen, redissolving the residue in chloroform, and loading the extract on silica gel G plates (Merck). Fatty acids were separated by thin layer chromatography using hexane/ethyl ether/formic acid (15:10:1, v/v/v) as a solvent system for 30 min as reported previously (18). Free fatty acids position on TLC plates was determined as comigration with commercially available standards after exposure to iodine vapors. Quantitative determination was obtained by scraping portions of the silica gel into scintillation vials followed by liquid scintillation spectrometry. The results are expressed as the percentage of radioactivity in the fatty acid band on the total radioactivity recovered from each lane. For phospholipid analysis, TLC plates (silica gel H) were resolved with solvent system of chloroform/ methanol/acetic acid/water (50:25:8:2, v/v/v) for 15 cm. Phospholipids were identified based on comigration with commercially available standards, and quantitative evaluation was performed by liquid scintillation spectrometry (18).
Migration Assay-Monocyte migration was evaluated using a microchamber technique (34) as described previously (18,30). 27 l of chemoattractant diluted in RPMI 1640 medium with 1% FCS were seeded in the lower compartment of the chemotaxis chamber (Nucleopore Corp., Pleasanton, CA), and 50 l of cell suspension (1.5 ϫ 10 6 /ml) were seeded in the upper compartment. The two compartments were separated by a 5-m pore size PVP membrane (Nucleopore). Chambers were incubated at 37°C in air with 5% CO 2 for 90 min. At the end of the incubation, filters were removed, fixed, and stained with Diff-Quik (Baxter s.p.a., Rome, Italy). Migrated monocytes in five high power oil immersion fields were counted.
Enzyme-linked Immunosorbent Assay for MCP-1-MCP-1 levels in monocyte culture supernatants were evaluated by a specific sandwich enzyme-linked immunosorbent assay exactly as described previously (35). The lower limit of sensitivity for this assay is 40 pg/ml.

Effect of MCP-1 on the Release of [ 3 H]Arachidonic Acid and [ 14 C]Oleic Acid from Labeled Human
Monocytes-Initial studies were designed to examine the selectivity of MCP-1-induced PLA 2 activity(s) for arachidonic acid. Monocytes cultured in nonadherent conditions in the presence of both [ 3 H]arachidonic acid and [ 14 C]oleic acid for 18 h showed a comparable uptake of the two labels (47.4 Ϯ 2.5 and 46.9 Ϯ 1.9%, respectively; n ϭ 4) with more than 90% of the labels incorporated in the phospholipid pools (data not shown). Distribution of the labels in phospholipids was: 60 Ϯ 3.6 and 72 Ϯ 2.6% phosphatidylcholine, 11 Ϯ 2.0 and 9 Ϯ 1.8% phosphatidylinositol/phosphatidyletanolamine, and 31 Ϯ 4.7 and 12 Ϯ 2.9% phosphatidylserine for [ 3 H]arachidonic acid and [ 14 C]oleic acid, respectively (n ϭ 3). Maximal chemotactic activity of MCP-1 is observed at 50 -100 ng/ml (6 -12 nM; 30) and these concentrations were used throughout this study. As shown in Fig. 1, 100 ng/ml MCP-1 induced a rapid accumulation of [ 3 H]arachidonic acid that peaked between 3 and 10 min and, at 3 min, corresponded to 196 Ϯ 18% (n ϭ 4) of control group activity (Fig. 1). MCP-1 activation appeared to be specific for arachidonic acid because no detectable release of [ 14 C]oleic acid was observed up to 30 min of stimulation. In the same experimental conditions, 10 Ϫ7 M fMLP for 3 min resulted in the release of 577 Ϯ 128% and 163 Ϯ 11% of control cells for [ 3 H]arachidonic acid and [ 14 C]oleic acid, respectively (n ϭ 4).
Effect of MCP-1 on cPLA 2 Phosphorylation-Agonist-triggered cPLA 2 activation, including that mediated by seven transmembrane domain receptors, is associated with increased phosphorylation of the protein on serine residues resulting in a stable 3-4-fold increase of cPLA 2 catalytic activity (36 -39). cPLA 2 phosphorylation correlates with the appearance of a more slowly migrating electrophoretic form of the protein. In agreement with previous reports (40,41), Western blot analysis of resting human monocytes shows that in these cells cPLA 2 migrates as a doublet (Fig. 2). Three min of stimulation with 100 ng/ml MCP-1 caused a decrease of the intensity of the faster migrating band and an increase in the intensity of the slower migrating species. Similar results were obtained with fMLP ( Fig. 2) and with MCP-3 (data not shown). A short (1 min) preincubation of monocytes with 100 nM platelet-activating factor, before MCP-1 stimulation, resulted in the complete loss of the faster migrating band (Fig. 2). These results parallel the synergism observed between platelet-activating factor and MCP-1 in terms of [ 3 H]arachidonic acid release from prelabeled cells (18,20). The relative changes in the two immunoreactive bands could be quantified by densitometry and expressed as the ratio of the slower over the faster migrating band: 1.5 Ϯ 0.2 and 4.1 Ϯ 0.6 for control and MCP-1-stimulated cells, respectively (n ϭ 3; p Ͻ 0.05 by paired Student's t test). Fig. 3 shows that a shift in the ratio of cPLA 2 immunoreactive bands was present in both cytosolic and membrane fractions. The effect was time-dependent, detectable 1 min after stimulation, reaching statistical significance in both cytosol and membrane between 3 and 5 min and declining to basal levels thereafter (data not shown).
Inhibition of cPLA 2 Expression by Antisense Oligonucleotide-Because of the lack of specific inhibitors, it is difficult to correlate cell functions to activation of different PLA 2 forms. Antisense technology provides a unique approach to this problem and was successfully used in monocytic cells (42,43) also to inhibit PLA 2 isoforms (23,43). Table I reports that monocytes treated with 10 M cPLA 2 antisense oligonucleotide for 48 h showed a marked decrease of the immunoreactive bands detected by quantitative Western blot, when compared with cultured untreated cells or with monocytes exposed to a similar concentration of control oligonucleotide in the same experimental conditions. Table I also shows that [ 3 H]arachidonic acid release was almost completely blocked in antisense oligonucleotide-treated monocytes challenged with 100 ng/ml MCP-1. Inhibition was not due to a toxic effect of the treatment because cell viability was higher than 90% (data not shown), and inhibition was not the result of homologous desensitization by MCP-1 released in the culture medium, because at the end of the incubation MCP-1 levels in untreated, antisense, and control oligonucleotide-treated cultures were similar (0.37 Ϯ 0.15, 0.44 Ϯ 0.15, and 0.39 Ϯ 0.12 ng/ml, respectively; n ϭ 3; p Ͼ 0.05 of oligonucleotide-treated versus untreated groups).
Effect of cPLA 2 Antisense Oligonucleotide on Chemotaxis-The effect of cPLA 2 antisense oligonucleotide treatment on monocyte chemotactic response to MCP-1 was investigated. fMLP was used as reference chemoattractant. Fig. 4 shows that antisense oligonucleotide treatment (0.3-3 M for 48 h) did not significantly alter the spontaneous migration of monocytes when compared with cells treated with control oligonucleotide or to untreated cultured monocytes. On the contrary, the number of cells migrated across polycarbonate filters in response to 50 ng/ml of MCP-1 was inhibited in a concentration-dependent manner (IC 50 ϭ 1.9 Ϯ 1.1 g/ml; n ϭ 3), with complete inhibition observed between 3 and 10 M antisense oligonucleotide (Fig. 4, upper panel, and data not shown). Control oligonucleotide did not affect basal or activated cell migration. Inhibition of MCP-1 response could partially be overcome by the use of higher concentrations of the agonist (Fig. 5A). In parallel experiments, the same monocyte population migrated normally in response to both optimal (10 Ϫ8 M) and suboptimal (10 Ϫ9 -10 Ϫ10 M) chemotactic concentrations of fMLP, indicating that inhibition was not the result of toxicity (Figs. 4, lower panel,  and 5B).
Effect of cPLA 2 Antisense Oligonucleotide on Chemotaxis to C-C Chemokines and C5a-Because of the discrepancy of effect of cPLA 2 antisense oligonucleotide treatment on monocyte chemotaxis to MCP-1 and fMLP, other C-C chemokines and a second classical chemoattractant were tested in the chemotaxis assay. All the agonists were used at their optimal chemotactic concentrations. In the same experimental condi-   tions, all the agonists tested induced the release of arachidonic acid from human monocytes (18,19). 2 6). On the contrary, the chemotactic responses to C5a and fMLP were only minimally affected (20.2 Ϯ 17.6; n ϭ 7 and 13.5 Ϯ 11.2; n ϭ 21, respectively), and inhibition never reached statistical significance (p ϭ 0.152 and p ϭ 0.123 by paired Student's t test for C5a and fMLP, respectively; Fig. 6).

DISCUSSION
The release of arachidonic acid and the production of eicosanoids is an early event in the activation of phagocytic cells by several inflammatory agonists including chemotactic factors (14,44,45). PLA 2 activation represents the most direct and the main mechanism of arachidonic release from the sn-2 position of membrane phospholipids. Thus, activation of PLA 2 is the rate-limiting step in arachidonic acid mobilization (21,22,46).
In the present study we report that chemotactic concentrations of MCP-1, a prototypic C-C chemokine, induced [ 3 H]arachidonic acid release and phosphorylation of cPLA 2 in a timedependent manner (Figs. 1-3). Similar results (not shown) were obtained with MCP-3, another member of the C-C chemokine family that shows 72% homology to (8) and shares binding sites (19,47) with MCP-1, in human monocytes. The kinetics of cPLA 2 phosphorylation after MCP-1 and MCP-3 stimulation were fast and correlated with arachidonic acid release from labeled monocytes. Both release and phosphorylation were already detectable 1 min after stimulation, peaked between 3 and 10 min, and returned to baseline within the following 10 min (Refs. 18 and 19 and Figs. 1 and 2). cPLA 2 is a 85-kDa protein that preferentially hydrolyzes phospholipids containing arachidonic acid at the 2 position and that was recently purified and cloned from the cytosol of myelomonocytic cell lines (25-27, 48 -50). Ca 2ϩ is not required for cPLA 2 catalytic activity (51,52), but nanomolar concentrations of Ca 2ϩ are needed for interfacial association with the lipid bilayer (25,26). In ionophore-permeabilized human monocytes it was shown that maximal arachidonic acid release by MCP-1 was observed in the presence of 300 -700 nM free Ca 2ϩ concentration (18). These concentrations are compatible with MCP-1-activated intracellular Ca 2ϩ levels in monocytes (30,53) and with the calcium concentrations required for cPLA 2 membrane association (25,26). In the experimental conditions used, MCP-1 did not release oleic acid from labeled monocytes, suggesting that the activated phospholipase(s) is specific for arachidonic acidlabeled phospholipid pools (Fig. 1). Finally, cPLA 2 antisense oligonucleotide-treated monocytes released only a minute fraction (Ͻ20% of control oligonucleotide-treated cells) of [ 3 H]arachidonic acid when challenged with MCP-1 (Table I). Taken together, these data indicate that cPLA 2 plays a major role in the mobilization of arachidonic acid in MCP-1-stimulated monocytes.
Monocytes treated with a specific antisense oligonucleotide were used to address the role of cPLA 2  complete inhibition of cell migration was observed (Ͼ85%) in response to an optimal concentration of MCP-1 (Figs. 4 and 5). Inhibition by the antisense oligonucleotide was concentrationdependent and -specific, because it was not observed in cells treated with a control oligonucleotide (Figs. 4 and 5) or with a c-myb-specific antisense oligonucleotide (data not shown). Inhibition of chemotactic response was not caused by toxicity of the treatment because: (i) cell viability was always higher than 90% by trypan blue dye exclusion, and treated monocytes were similar to untreated cells in terms of morphology (data not shown) and spontaneous migration (Fig. 4); (ii) cells exposed to control oligonucleotides showed a normal migration to MCP-1 (Fig. 4); and (iii) cPLA 2 antisense oligonucleotide-treated cells migrated normally to fMLP (Figs. 4, 5 and 6). Finally, inhibition was not caused by homologous receptor desensitization (8) because comparable levels of MCP-1 were present in the supernatants of untreated and control or antisense oligonucleotide-treated cells.
A more extensive analysis showed that monocyte chemotaxis to all the C-C chemokines tested was strongly (Ͼ70%) inhibited by the antisense oligonucleotide treatment, whereas monocyte migration to fMLP or to C5a was only minimally (Ͻ20%) affected (Fig. 6). Thus, according to their requirement for cPLA 2 , it is possible to divide the chemotactic agonists tested in two groups, a first one, highly sensitive to the action of the antisense oligonucleotide that includes all the C-C chemokines investigated, and a second one that was poorly sensitive to this treatment and that comprises classical chemotactic factors. At the moment, the reason for this difference is unknown. It is possible that fMLP and C5a but not chemokine receptors might have access to the surviving cPLA 2 molecules. fMLP and C5a receptors could also induce the required levels of free arachidonic acid through the activation of other types of PLA 2 that are not efficiently coupled to chemokine receptors. Alternatively, fMLP and C5a receptors could bypass cPLA 2 inhibition through the stronger activation of signaling pathways alternative to arachidonic acid mobilization. A similar hypothesis can be formulated to explain the ability of sopraoptimal concentrations of MCP-1 to overcome oligonucleotide inhibition. The optimal chemotactic concentration (50 ng/ml MCP-1) is similar to or less than the K d value of MCP-1 receptors (47,54). A higher degree of receptor occupancy could activate residual cPLA 2 or trigger alternative signaling pathways.
A direct role for arachidonic acid and its metabolites in cell movement was recently suggested in different cell types. Both 5-lypoxygenase and cyclooxygenase products were found to regulate epidermal growth factor-induced actin remodeling in A431 cells (55) and neutrophil migration in vivo (56). cPLA 2mediated arachidonic acid release was found to be required for basic fibroblast growth factor-stimulated migration of endothelial cells (57). A direct role for arachidonic acid in monocyte and macrophage adherence, expression of adhesion molecules, and chemotaxis was suggested (58 -60). Recently, three chemotactic factors for phagocytic cells: macrophage colony-stimulating factor (40,41), transforming growth factor-␤ (61), and fMLP (62) were shown to activate cPLA 2 in human monocytes, elicited guinea pig macrophages, and human neutrophils, respectively.
In a previous study we found a strict correlation between C-C chemokine-induced arachidonic acid release and monocyte migration (18). In the present study, we show that cPLA 2 appears to be the main effector enzyme for chemokine-elicited arachidonic acid release in human monocytes. In addition, by the use of specific antisense oligonucleotides, we provide evidence that arachidonic acid by itself or through its metabolites is strictly implicated in the induction of monocyte migration to C-C chemokines.