Epstein–Barr virus primes human polymorphonuclear leucocytes for the biosynthesis of leukotriene B4

Abstract

In the present study, we have investigated the effect of the short-term incubation of polymorphonuclear leucocytes (PMN) with infectious Epstein–Barr virus (EBV) on leukotriene B₄ (LTB₄) biosynthesis. Pre-exposure of PMN to EBV led to an increased production of LTB₄ upon stimulation with either the ionophore A23187, the chemotactic peptide fMLP, or phagocytic particles (zymosan). Experiments performed with viral particles pretreated with a neutralizing antibody raised against the gp350 of the viral envelope revealed that a specific interaction between the PMN surface and the viral glycoprotein gp350 is required for the priming effect of EBV. Preincubation of PMN with EBV resulted in an increased release of arachidonic acid upon stimulation with a second agonist. Moreover, LTB₄ biosynthesis in EBV/A23187-treated PMN was greatly diminished in the presence of an inhibitor of the cytosolic phospholipase A₂ (cPLA₂), suggesting that cPLA₂ plays a critical role in the priming effect of EBV. Accordingly, EBV by itself promoted Ser-505 phosphorylation of cPLA2 and strongly enhanced fMLP-induced phosphorylation of p38 MAP kinase, an enzyme known to phosphorylate cPLA₂ in human PMN. Furthermore, fMLP-induced translocation of cPLA₂ was strongly enhanced when PMN were previously exposed to EBV. These data indicate that binding of EBV to human PMN results in the activation of intracellular events involved in the release of pro-inflammatory lipid mediators.

INTRODUCTION

Epstein–Barr virus (EBV) is a highly pathogenic herpesvirus that causes infectious mononucleosis and has been linked to a variety of human malignancies such as Burkitt's lymphoma, undifferentiated nasopharyngeal carcinoma, Hodgkin's disease and immunoblastic lymphoma (reviewed in [1]). Although the major target cells for in vivo EBV infection are B lymphocytes and epithelial cells, it is now believed that EBV may interact with a wider spectrum of cell types. For instance, the presence of the EBV genome and viral proteins have been found in certain T cell lymphomas and immature thymocytes, and several T cell lines have been successfully infected in vitro with EBV [2–7].

Furthermore, we have shown previously that EBV binds specifically to human monocytic cells and PMN via an as yet unknown receptor distinct from the CD21 antigen expressed on B cells [8–10] and that such binding results in the synthesis of inflammatory mediators. In monocytic cells, binding of EBV to the cell membrane was found to induce the production of interleukin (IL)-1 and -6 [9,11]. In PMN, EBV attachment also stimulates the expression of inflammatory molecules such as IL-1, IL-1 receptor antagonist, IL-8 and macrophage inhibitory protein (MIP)-1α [12,13]. In both cell types, modulation of inflammatory mediators was found to involve an interaction of the viral envelope glycoprotein gp350 with the cell surface.

Leukotrienes (LTs) represent another group of molecules with regulatory functions on the immune response. LTs, and particularly LTB₄, constitute a family of potent proinflammatory lipidic compounds produced from the dioxygenation of arachidonic acid (AA) via the 5-lipoxygenase (5-LO) pathway [14,15]. The major sources of LTB₄ are cells involved in the inflammatory response, i.e. neutrophils, monocytes and macrophages [14]. One important biological property of LTB₄ is its ability to stimulate phagocyte locomotion and chemotaxis [16]. LTB₄ is also a potent modulator of lymphocyte (T and B cells) and phagocyte functions, and is involved in the regulation of cytokine production. LTB₄ also augments cytotoxic activities of monocytes, enhances target cell recognition by natural killer (NK) cells and improves their lytic activity against tumour cells (reviewed in [17,18]).

Numerous reports have described modulatory effects of viruses or proteins of the viral envelopes on the arachidonic acid (AA) cascade. For example, Behera et al. [19] reported that respiratory syncytial virus (RSV) infection of a bronchial epithelial cell line leads to an up-regulation of the biosynthesis of LTs. These results were consistent with other findings showing that increased LT synthesis correlated with the presence of RSV particles in patients affected with pneumonia or bronchiolitis [20,21]. Exposure of monocytic cells to the HIV-1 glycoprotein gp120 induced the expression of cyclooxygenase and lipoxygenase pathway enzymes [22], and it was reported recently that the cytotoxic effect of gp120 on a neuroblastoma cell line involved activation of the AA cascade together with enhanced membrane lipid peroxidation [23]. We have described previously the capacity of EBV to prime monocytes for an increased synthesis of LTB₄ and LTC₄ through modulation of 5-LO activity [24]. In the present study, we show that exposure of PMN to EBV enhances levels of LTB₄ biosynthesis upon stimulation with a second agonist through a mechanism implicating a direct stimulatory effect of EBV on cPLA₂ phosphorylation (Ser-505) and a priming effect of the viral particles on the translocation of cPLA₂ induced by neutrophil agonists. Since phagocytes play a crucial role in host defence against infectious agents, changes in the generation of proinflammatory mediators may be an important issue in the early events of viral infection.

MATERIALS AND METHODS

Materials

fMLP, zymosan and A23187 were purchased from Calbiochem (San Diego, CA, USA). The cPLA₂ inhibitor, methyl arachidonyl fluorophosphonate (MAFP), was obtained from Cayman Chemicals (Ann Arbor, MI, USA). MAFP, A23187 and fMLP were dissolved in DMSO. Zymosan was boiled 30 min in distilled water, washed twice and resuspended in 0·9% NaCl at 20 mg/ml.

Isolation of PMN

PMN were isolated from venous blood collected on heparin from normal healthy volunteers using dextran sedimentation of erythrocytes followed by standard techniques of centrifugation on Ficoll-Hypaque cushions (Pharmacia-Biotec Inc, Baie d'Urfé, Canada) as described previously [13]. PMN preparations contained less than 1% monocytes as determined by monoesterase staining, and less than 0·3% B and T lymphocytes as measured by flow cytometry using anti-CD2, anti-CD3, and anti-CD19 monoclonal antibodies (Becton Dickinson, San Jose, CA, USA). Cell viability was determined by the trypan blue-dye exclusion procedure and was greater than 99% in all preparations.

Epstein–Barr virus preparations

Preparations of EBV were obtained from TPA-treated B95-8 cells, as described previously [25]. Briefly, B95-8 cells (mycoplasma-free tested), which are latently infected by EBV, were grown in RPMI-1640 medium supplemented with 10% heat-inactivated FBS and exposed to tetradecanoyl phorbol acetate (TPA) (20 ng/ml) to promote virus production. When cell viability was 20% or less, as determined by trypan blue-dye exclusion, cell culture supernatants were collected and passed through a 0·45-mm pore size filter. Viral particles were then purified by ultracentrifugation. Virus stocks were resuspended in RPMI-1640 and stored at −80°C until use. Viral titres were determined as outlined previously [25] and adjusted to 10⁷ transforming units/ml (TFU/ml). For every experiment, cell-free supernatants collected from TPA-induced BJAB cells were processed as described above and served for mock infected controls. PMN were infected as outlined in the figure legends, using optimal doses of EBV determined by serial dilutions as described previously [9]. When indicated, viral preparations were UV-irradiated for 1 h at 265 nm, a treatment which renders the viral particles non-infectious without altering the native structure of EBV viral envelope. The neutralizing monoclonal antibody 72A1 (ATCC, Manassas, VA, USA) directed against the viral envelope protein gp350/220 [26], which blocks the EBV binding to the cell surface, and the non-neutralizing antibody 2L10, also reacting against gp350/220 but without inhibiting the infectious process [27], were utilized to evaluate the specificity of the response obtained with EBV. Saturating amounts of each monoclonal antibody preparation were incubated with the viral particles for 1 h at 37°C.

Incubation conditions

Ten million PMN were incubated in serum-free Hanks's balanced salt solution (HBSS) (gibco BRL, Burlington, Ontario, Canada) containing 10 mmol/l HEPES and 1·6 mmol/l CaCl₂. PMN were preincubated with or without infectious EBV (10⁵ TFU) and stimulated with fMLP (1 μmol/l), zymosan (1 mg/ml) or the ionophore A23187 (75 nmol/l) under conditions outlined in the figure legends. At the end of the incubation periods, reactions were stopped by addition of 2 volumes of a methanol/acetonitrile mixture (50/50, v/v) containing 12·5 ng/ml each of prostaglandin B₂ (PGB₂) and 19-hydroxy-PGB₂ which served as internal standards for RP-HPLC analysis. The denatured samples were then stored at −20°C. For whole blood incubations, 1 ml of freshly collected heparinized blood was preincubated with or without EBV (10⁵ TFU) for different periods of time, and blood samples were further stimulated (or not) with 1 μmol/l fMLP, as indicated in the figure legends. After 10 min, incubations were stopped by cooling the blood samples in an ice-water bath. Plasma were collected and stored at −20°C.

Reverse-phase HPLC analysis of arachidonic acid metabolites

The denatured samples (prepared as described above) were centrifuged at 2000 g for 20 min to remove the precipitated material, and the organic solvent content was reduced to ∼50% by evaporation under a stream of nitrogen. The amounts of LTB₄, LTC₄ or 5,15-diHETE (for the 5-LO activity assay) were then analysed by reverse-phase (RP) HPLC using an extraction procedure described previously [28] and UV detection. Plasma samples obtained from whole blood incubations were denatured with organic solvents and 5-LO products were analysed by RP-HPLC, as described previously [29].

Analysis of arachidonic acid release

PMN were incubated in serum-free HBSS in presence or absence of EBV for 15 min at 37°C prior to stimulation with 1 μm fMLP. Cells were then denatured by adding 2 volumes of ice-cold methanol containing 10 ng of D₈-arachidonic acid per millilitre as an internal standard. Samples were processed for HPLC analysis as described above and the HPLC fractions containing AA (determined by using ³H-AA) were collected, evaporated under reduced pressure and resuspended in 100 μl of acetonitrile. AA was measured by liquid chromatography-mass spectrometry (LC-MS) using a nebulizer-assisted electrospray (ion spray) interface coupled to a mass spectrometer (API-III; PE Sciex, Thornhill, Ontario, Canada), as described previously [30].

Analysis of phosphorylation and translocation of cPLA₂

Cell lysates were obtained from unstimulated cells or cells treated with EBV and/or the soluble agonist fMLP. Levels of phosphorylated (Ser-505) and non-phosphorylated cPLA₂ were assessed by immunoblot analyses as described previously [24]. Briefly, cells (5 × 10⁶/sample) were pelleted at 500 g for 8 min and resuspended in 300 μl of a sucrose buffer containing protease inhibitors (1 mmol/l phenylmethyl sulphonyl fluoride and 10 mg/ml aprotin and leupeptin) as described previously [31]. For measurements of cPLA₂ in subcellular fractions, cells (5 × 10⁶/sample) resuspended in 150 μl of the sucrose buffer were sonicated (three times for 20 s at 60% duty cycles, output control set at 3, using a Johns Scientific Inc. Ultrasonic Processor (Biotech Scientifique, Montréal, Canada) and the lysates were centrifuged at 12000 g for 10 min at 4°C. The pellets consisting of nuclei (nucleus-containing fraction [N]) were kept for subsequent SDS-PAGE analyses, and the supernatants were recentrifuged at 180000 g for 25 min at 4°C. The pellets thus obtained were composed of cellular membranes (M) (including nuclear membranes) [31] and the supernatants constituted the cytosolic fractions (C). The whole cell lysates and subcellular fractions were electrophoresed on a 5% to 15% gradient SDS-polyacrylamide gel in which the phosphorylated form of cPLA₂ (Ser-505) migrates slower than the non-phosphorylated form, as described previously [32]. Briefly, proteins were transferred for 2·5 h at 500 mA current setting into an Immobilon-P polyvinylidene difluoride (PVDF) blotting membrane; transfer efficiency was visualized by Ponceau Red staining. The membranes were first pretreated for 30 min at 25°C in Tris-buffered saline (TBS: 25 mmol/l Tris-HCl, pH 7·6, 0·2 mol/l NaCl) containing 0·15% Tween 20 and 5% dried milk (wt/vol) as a blocking agent before incubation for 1 h at 25°C in TBS-Tween solution containing the anti-cPLA₂ antiserum diluted 1 : 500 (antiserum MF-142, generously provided by Dr P. Weech, Merck-Frosst, Montréal, Canada). After two washes of 10 min each in TBS solution, the membranes were treated with a horseradish peroxidase-linked donkey antirabbit antibody and revealed with the enhanced chemiluminescence (ECL) reagent (Dupont-NEN, Boston, MA, USA) as outlined by the manufacturer.

Analysis of the expression of p38 MAPK

Neutrophils (10 × 10⁶ cells) were incubated with or without EBV (10⁵ TFU) for 30 min at 37°C in HBSS/Ca²⁺ and then stimulated or not with fMLP (1 µm) for the indicated time. The reaction was stopped by the addition of 100 µl of boiling 2X Laemmli buffer and the proteins were analyzed by electrophoresis on a 9% SDS-polyacrylamide gel. Proteins were transferred onto a PVDF membrane as described above and the membrane was immunoblotted with antibodies specific for the phosphorylated p38 MAPK or the phosphorylated and nonphosphorylated enzyme (Calbiochem, San Diego, CA, USA), as recommended by the manufacturer. Proteins were revealed with the enhanced chemiluminescence (ECL) detection system.

Measurement of intracellular calcium concentration

Neutrophils (10⁷ cells/ml) were preincubated with 1 µm of the fluorescent probe fura-2-acetoxymethyl ester (Molecular Probes, Junction City, OR, USA) for 30 min at 37°C. Cells were then washed twice to remove the extracellular probe and incubated at 37°C for an additional 30 min in presence or absence of EBV (10⁵ TFU) or GM-CSF (1 µm) prior to fMLP (1 µm) stimulation. The cells were then washed twice, resuspended at 5 × 10⁶ cells/ml in HBSS containing 1·6 mm calcium, and finally transferred to the thermally (37°C) regulated cuvette compartment of a spectrofluorometer (SLM 8000, Aminco, Urbana, IL, USA). The fluorescence of the cells in response to stimulation with fMLP was monitored for the indicated period of time at an excitation wavelength of 340 nm and emission wavelength of 510 nm. The intracellular calcium concentrations were calculated as described by Tsien et al. [33].

Statistical analysis

Statistical analysis was performed using Student's paired (two-tailed) t-test and differences were considered significant at P <0·05.

RESULTS

Priming effect of EBV on LTB₄ biosynthesis by PMN

PMN were preincubated with or without EBV for increasing periods of time before stimulation with the ionophore A23187. As seen in Fig. 1a, cells stimulated with a low concentration of A23187 (50 nm) or cells exposed to EBV alone did not produce detectable amounts of LTB₄ as measured by HPLC. In contrast, PMN preincubated with EBV and further stimulated with 50 nm A23187 released considerable amounts of LTB₄. Optimal LTB₄ release was obtained when PMN were pretreated for a period of 10–30 min with EBV before stimulation with the ionophore. A similar kinetic for the priming effect of EBV on LTB₄ biosynthesis was observed in whole blood, using as second stimuli the chemotactic peptide fMLP (Fig. 1b). The neutrophil agonist fMLP (1 μm) alone had only a slight stimulatory effect on LTB₄ biosynthesis, whereas preincubation of PMN with EBV for only 10 min resulted in enhanced release of LTB₄ when stimulated with fMLP. Maximal effects were obtained with a preincubation time of 30 min with the viral particles; at longer preincubation times, the priming effect of EBV declined but remained significant for up to 2 h, the longest time point tested. Whole blood exposed to EBV, but not further stimulated with fMLP, did not produce detectable amounts of LTB₄. We also assessed the ability of EBV to prime human PMN for LT biosynthesis upon stimulation with zymosan particles or GM-CSF and fMLP. PMN isolated from different healthy donors were all responsive to the priming effect of EBV as shown in Table 1.

Fig. 1.

Kinetics of the priming effect of EBV on LTB₄ biosynthesis by PMN. (a) PMN (10⁷ cells/0·1 ml) were preincubated at 37°C with or without EBV (10⁴ TFU/0·1 ml) for the indicated period of times, diluted 10-fold with HBSS, and stimulated with 50 nm A23187 (or treated with DMSO) for 10 min at 37°C. Reactions were stopped by addition of organic solvents and 5-LO products were analysed by RP-HPLC as described in Materials and methods. (b) Heparinized whole blood (1 ml) was preincubated with or without EBV (10⁵ TFU/ml) and stimulated with fMLP (1 μm) (or treated with DMSO) for 10 min at 37°C. Reactions were stopped on ice, samples were centrifuged and plasma were collected and analysed by RP-HPLC. In (a) and (b), the mock infected controls represent PMN that were exposed to cell-free supernatant obtained from resting B95-8 cells as described in Materials and methods. The values obtained with mock treated PMN (not shown) were subtracted from the corresponding data points. The mock treatment indicates PMN that were exposed to cell-free supernatant obtained from TPA-activated BJAB cells as described in Materials and methods. The data shown are the mean ±s.d. of triplicate incubations from one experiment representative of five separate experiments in each case. The indicated amounts of LTB₄ represent the sum of LTB₄ and its ω-oxidation products 20-hydroxy- and 20-carboxy-LTB₄. (a) ▵, EBV; □, A23187; ▪, EBV/A23187. (b) ▵, EBC; □, f-MLP; ▪, EBV/f-MLP.

Table 1.

Priming effect of EBV on LBT₄ biosynthesis in human PMN

Donor	Zymosan	EBV + zymosan	GM-CSF + fMLP	EBV/G-CSF + fMLP
1	20·8 ± 2·2	57·8 ± 6·1*	0·1 ± 0·1	11·0 ± 0·3*
2	47·4 ± 5·0	142·4 ± 11·7*	0·2 ± 0·1	17·1 ± 3·0*
3	24·2 ± 7·4	113·2 ± 16·0*	0·1 ± 0·1	13·0 ± 2·0*

 PMN (5 × 10⁶ cells/0·1 ml) were preincubated with or without EBV (10⁴ TFU/0·1 ml) for 10 min at 37°C, diluted 10-fold with HBSS, and further stimulated 30 min with zymosan (1 mg/ml); PMN (5 × 10⁶ cells/1 ml) were preincubated with GM-CSF (200 pm) for 30 min at 37°C with or without EBV (10⁵ TFU/1 ml), diluted 10-fold with HBSS and further stimulated 10 min with 1 μm fMLP. LTB₄ and its metabolites were measured by RP-HPLC as described in Materials and methods. The data indicated represent the sum of LTB₄ and its ω-oxidation products 20-hydroxy- and 20-carboxy-LTB₄ and are the mean of quadruplicates ± s.d. expressed in picomoles per 5 × 10⁶ cells. LTB₄ was not detectable in untreated cells or cells treated with EBV only. *Results are statistically different (P ≤ 0·01) from the corresponding incubations without EBV.

Specificity of the effect of EBV on LTB₄ biosynthesis

We next investigated the specificity of the priming effect of EBV and determined whether this activity requires the presence of infectious viruses or is due to the binding of intact particles to the cellular membrane. As shown in Fig. 2, UV-irradiation of EBV virions, a treatment that causes DNA damage and prevents viral transcription without altering the native particle and the viral binding, had no inhibitory effect on the ability of EBV to prime for LT synthesis. This suggests that neosynthesis of viral transcripts is not involved and that structural integrity of the components of the viral envelope and adsorption of EBV to the cell surface of the PMN are sufficient for a priming effect of EBV. The relevance of the interaction of PMN and the viral envelope protein gp350 to the priming effect of EBV was then assessed by preincubating EBV particles with the neutralizing antibody 72A1, which inhibits binding of the virus to target cells [26]. As shown in Fig. 2, neutralizing antibodies to EBV strongly diminished the ability of EBV to activate LT biosynthesis upon stimulation with the ionophore A23187. In contrast, the monoclonal antibody 2L10, which recognizes another epitope of the viral envelope protein gp350 without neutralizing or blocking EBV binding [27], did not interfere with the priming effect of EBV on LT release (data not shown).

Fig. 2.

Effect of infectious and neutralized EBV particles on LTB₄ biosynthesis in A23187-stimulated PMN suspensions.PMN (10⁷ cells/0·1 ml) were preincubated with or without infectious EBV (10⁴ TFU/0·1 ml), or UV-irradiated EBV (a treatment which renders the viral particles non-infectious without altering the native structure of the viral envelope), or inactivated virus (pretreated with the neutralizing antibody 72A1, which is known to block EBV binding) for 15 min at 37°C, diluted 10-fold with HBSS, and stimulated with A23187 (75 nm) for 10 min at 37°C. LTB₄ and its metabolites were measured by RP-HPLC as described in Materials and methods and the indicated amounts of LTB₄ represent the sum of LTB₄ and its ω-oxidation products 20-hydroxy- and 20-carboxy-LTB₄. The values obtained with mock treated PMN (not shown) were subtracted from the corresponding data points. The mock treatment indicates PMN that were exposed to cell-free supernatant obtained from TPA-activated BJAB cells as described in Materials and methods. Results (mean ±s.d.) are from one experiment representative of three separate experiments performed in triplicate. *P ≤0·05. □, EBV; , 72A1; , A23187; ▪, EBV/A23187; , UV-treated EBV/A23187; , EBV-72A1/A23187.

Effects of EBV on arachidonic acid release and cPLA₂ activation processes

The effect of EBV on AA release was assessed. Figure 3 shows that EBV primes PMN for the release of increased amounts of AA after stimulation with fMLP. A kinetic analysis revealed that this priming effect was seen after only 2 min of incubation with the agonist. Untreated cells or cells treated with fMLP (1 μm) or EBV alone did not show significant release of AA up to an incubation time of 10 min; accordingly LTB₄ biosynthesis was detectable only in neutrophils treated successively with EBV and fMLP (data not shown). To assess that the observed augmentation of AA release involved the cPLA₂, experiments were performed in presence of a cPLA₂ inhibitor, MAFP. As seen in Fig. 4, a strong inhibition of the priming effect by EBV on LTB₄ biosynthesis was observed in the presence of the cPLA₂ inhibitor. This inhibition was relieved when exogenous AA was added to the reaction mixture at the time of stimulation with the ionophore, demonstrating that the reduction in LTB₄ synthesis was due to depletion of AA rather than to an unspecific inhibitory effect of the drug. Since the activation of the cPLA₂ involves Ser-505 phosphorylation [32,34–36] and a calcium-dependent translocation from the cytosol to membranes where the substrate is localized [37–39], we next examined the effect of EBV on these two events related to cPLA₂ activation. As seen in Fig. 5a, PMN incubated with EBV for a period of 15 min clearly showed enhanced levels of phosphorylated cPLA₂ (upper band of the doublet). As reported previously [40], fMLP also stimulated cPLA₂ phosphorylation, which was further enhanced by pretreatment of the PMN with EBV particles. The nucleus-containing fraction (N) prepared by sonication of PMN contains only a very small amount of the nuclear membranes which are present in the membrane fraction (M) [31]. The apparent discrepancy seen in Fig. 5a (and 5b) in the total amount of cPLA₂ (which seems lower in non-stimulated PMN in comparison to PMN exposed to stimulatory agents) is explained on the basis that the antiserum used in these Western blot analyses has a greater affinity for the phosphorylated form of the cPLA₂, as noted previously [31]. It is worth noting that the relative amount of both forms of cPLA₂ (phosphorylated and non-phosphorylated) found in mock treated or unstimulated PMN varied between PMN preparations (compare Figs 5 and 6); however, in unstimulated PMN the nonphosphorylated cPLA₂ was always detectable. The effect of EBV on the subcellular distribution of cPLA₂ was next investigated. Figure 5b shows the distribution of cPLA₂ in three PMN fractions. As reported previously, in non-stimulated PMN, cPLA₂ was mainly cytosolic and was present in both phosphorylated and non-phosphorylated forms as shown by the presence of the doublet [31,40]. While EBV did not, on its own, alter the subcellular distribution of the cPLA₂, EBV and fMLP clearly stimulated the phosphorylation of cPLA₂ present both in the membrane and cytosolic fractions, in agreement with the data shown in Fig. 5a (cPLA₂ immunoblots of whole PMN). Most importantly, pretreatment of PMN with EBV particles resulted in a dramatic increase of the translocation of cPLA₂ (phosphorylated form) to the membrane fraction of the PMN upon stimulation with fMLP, demonstrating that exposure of PMN to EBV has a significant impact on both the phosphorylation and the translocation of cPLA₂ from cytosol to membrane structures.

Fig. 3.

Effect of EBV on AA release in fLMP-stimulated PMN. PMN (5 × 10⁶ cells/0·1 ml) were preincubated with or without EBV (10⁴ TFU/0·1 ml) for 15 min at 37°C, diluted fivefold with HBSS, and incubated with or without fMLP (1 μm) at 37°C for the indicated period of times. The mock treatment indicates PMN that were exposed to cell-free supernatant obtained from TPA-activated BJAB cells as described in Materials and methods. Samples were analysed by RP-HPLC and fractions containing AA were collected for quantification by mass spectrometry as described in Materials and methods. Results (mean ±s.d.) are from one experiment representative of three separate experiments performed in triplicate. *P ≤0·05. , mock; □, EBV; , fMLP; ▪, EBV/fMLP.

Fig. 4.

Effect of MAFP on LTB₄ biosynthesis in EBV-primed PMN. PMN (10⁷ cells/0·1 ml) were preincubated in the presence or absence of MAFP (3 μm) for 15 min at 37°C, further incubated with or without EBV (10⁴ TFU/0·1 ml) for 30 min, and then diluted 10-fold with HBSS and stimulated with A23187 (75 nm) for 10 min in the presence or absence of exogenous AA (3 μm). The mock treatment indicates PMN that were exposed to cell-free supernatant obtained from TPA-activated BJAB cells as described in Materials and methods. LTB₄ and its metabolites were measured by RP-HPLC as described in Materials and methods. Results (mean ±s.d.) are from one experiment representative of three separate experiments performed in triplicate. *P ≤0·05.

Fig. 5.

Effect of EBV on Ser-505 phosphorylation and translocation of cPLA₂ in PMN suspensions. (a) PMN (5 × 10⁶ cells/0·1 ml) were incubated with or without EBV (10⁴ TFU/0·1 ml) for 15 min at 37°C, diluted 10-fold with HBSS, and then stimulated or not with 1 μm fMLP. Whole cell extracts were processed for analysis of cPLA₂ by SDS-PAGE and immunoblotting as described in Materials and methods; 90 ml aliquots of each whole cell extracts (∼ 375 μl) were applied to the gels. (b) PMN (5 × 10⁶ cells/0·1 ml) were incubated with or without EBV (10⁴ TFU/0·1 ml) for 15 min at 37°C. Cells were then washed and fractionated into nucleus (N), cytosolic (C) and membrane (M) fractions, as indicated in Materials and methods; the membrane fraction obtained by sonication of PMN contains plasma membranes and most of the nuclear membranes [31]. The membrane and nuclear fractions obtained from 5 × 10⁶ PMN samples were resuspended in 300 and 125 μl of Laemmli buffer, respectively, whereas the cytosolic fractions (∼275 μl) were mixed with 0·2 volume of Laemmli buffer 5×; 90 μl aliquots of each fraction were applied on the gels. Each subcellular fraction was processed for SDS-PAGE analysis and immunoblotting of cPLA₂ as described in Materials and methods. In (a) and (b), the mock treatment indicates PMN that were exposed to cell-free supernatant obtained from TPA-activated BJAB cells as described in Materials and methods. The data shown are from one experiment representative of four separate experiments performed with PMN obtained from different donors.

Fig. 6.

Effects of EBV on fMLP-induced mobilization of calcium in human PMN. PMN (10⁷ cells) were incubated with Fura-2-AM for 30 min at 37°C and subsequently pretreated with mock control, EBV or GM-CSF (0·2 nm) for an additional 30 min prior to fMLP (1 µm) stimulation. The concentration of intracellular calcium in neutrophils treated with mock, EBV or GM-CSF alone was 125–150 nm throughout the 300 second incubations. The arrow indicates the time of addition of fMLP. The data shown are from one experiment representative of four separate experiments performed with PMN obtained from different donors. —, Mock + fMLP; ‐‐‐, EBV + fMLP; ···, GM-CSF + fMLP.

Effect of EBV on Ca²⁺ mobilization and MAP kinase activation in neutrophils

The translocation of cPLA₂ is a Ca²⁺-dependent process important for its activity [38,41]; cPLA₂ is also activated by phosphorylation of serine-505 by mitogen-activated protein kinase (MAP kinase) [39]. Recent evidence suggests that extracellular signal-regulated kinase (erk) 1 and 2 and p38 MAP kinase can phosphorylate cPLA₂ in human neutrophils [32,42–45]. These MAP kinases are themselves activated through phosphorylation processes. To understand better the cellular events involved in the activation of cPLA₂ in EBV-primed neutrophils, additional experiments were performed. We first investigated the effect of EBV on intracellular calcium concentration. As depicted in Fig. 6, stimulation of Mock-treated PMN with fMLP initiated a rapid and transient increase of intracellular calcium concentration that peaked at 30 s post-stimulation, and decreased gradually afterward. The general profile of the calcium concentration curves of EBV- and GM-CSF-treated PMN suspensions was similar to that of Mock-treated cells with the difference that peak calcium concentrations were markedly increased in EBV- and GM-CSF-treated PMN, which indicated a stronger response of PMN to fMLP challenge in these experimental conditions. Indeed, we have found that a 30-min preincubation of PMN with EBV prior to stimulation increased maximal Ca²⁺ concentration by more than 30% compared to Mock samples. In contrast, EBV by itself did not affect the intracellular concentration of calcium, which remained at basal level at all times during the course of the experiment (data not shown).

The ability of EBV to activate (phosphorylate) erk-1, -2 and p38 MAP kinase in primed PMN was investigated next. Kinetic analysis shows that EBV priming does not modulate the phosphorylation level of erk-1 and -2 in PMN (data not shown). However, levels of phosphorylated p38 MAP kinase in fMLP-treated PMN were significantly increased when PMN were pretreated with EBV (Fig. 7b) as compared to unprimed PMN stimulated with fMLP (Fig. 7a). These results suggest that the phosphorylation of cPLA₂ in EBV-primed PMN may involve the activation of p38 MAP kinase rather than that of erk-1 or -2.

Fig. 7.

Effect of EBV on the phosphorylation of p38 MAPK by fMLP. PMN (10 × 10⁶ cells) were incubated in the absence (a) or presence of EBV (b) for 30 min at 37°C prior to stimulation with fMLP (1 µm) for varying periods of time. Cell lysates from each sample were analysed by SDS-PAGE and Western blotting using antibodies specific for the phosphorylated (P-p38 MAPK) or the phosphorylated and non-phosphorylated p38 MAPK. Results shown are representative of three independent experiments.

EBV replication is not required for cPLA₂ phosphorylation

In the next series of experiments, the effects of the protein synthesis inhibitor cycloheximide and PAA, an inhibitor of viral DNA polymerase, were investigated on the phosphorylation of cPLA₂ in PMN exposed to EBV or fMLP or EBV/fMLP. Figure 8 shows that in control (mock infected) PMN, the non-phosphorylated form of cPLA₂ was clearly predominant and that fMLP, EBV and EBV-fMLP caused a complete upward shift of the band (indicating cPLA₂ phosphorylation, in agreement with data shown in Fig. 5a and b). While PAA appeared to cause a slight increase in the phosphorylated form of cPLA₂ in the control (mock infected) PMN, both cycloheximide and PAA did not prevent the stimulation of cPLA₂ phosphorylation induced under the three incubation conditions indicated. These data demonstrate that neither protein synthesis nor viral replication were necessary for the effects of EBV on cPLA₂ phosphorylation (Fig. 8).

Fig. 8.

Effect of cycloheximide (CHX) and phosphonoacetic acid (PAA) on EBV-induced phosphorylation of cPLA₂. PMN (5 × 10⁶ cells/0·1 ml) were preincubated with or without 10 μg/ml of cycloheximide or 200 μg/ml of PAA for 1 h at 37°C, washed and further incubated with or without EBV (10⁴ TFU/0·1 ml) for 15 min and finally diluted 10-fold with HBSS and stimulated or not for 10 min with 1 μm fMLP. Whole cell extracts were processed for analysis of cPLA₂ as described in Fig. 5. The mock treatment indicates PMN that were exposed to cell-free supernatant obtained from TPA-activated BJAB cells as described in Materials and methods. The data shown are from one experiment representative of four separate experiments performed with PMN obtained from different donors.

DISCUSSION

Activation of PMN is of critical importance in response to viral infection. Upon activation, PMN release a variety of immunoregulatory agents, one of which is the potent chemoattractant LTB₄. In the present study, we show that exposure of PMN to EBV particles initiates cellular events which render these cells highly responsive, in terms of LTB₄ biosynthesis, to a second stimuli. Our results suggest that the gp350 of the viral envelope could be involved in the mechanism of EBV priming of PMN for LTB₄ biosynthesis. Indeed, pretreatment of EBV with the neutralizing antibody 72A1, which is known to block the binding of gp350 to the EBV receptor, abrogated the priming effect, while the non-neutralizing antibody 2L10, which does not interfere with the binding of EBV to its receptor, had no effect. The cellular receptor involved in the interaction of the glycoprotein gp350 with the PMN surface has yet to be characterized. Previous data from our laboratory suggested that this receptor is distinct from the well-defined CD21 (CR2) antigen used by EBV for entry into B cells; indeed, the CD21 antigen is not expressed at the surface of PMN [9] and moreover anti-CD21 antibodies did not interfere with the attachment of EBV to PMN and monocytes [8,9]. At the present time, it is unknown if EBV uses the same alternative receptor on PMN and monocytes. However, a similar priming effect of EBV on LTB₄ biosynthesis was observed recently in monocytic cells [24], and as observed in these cells, expression of EBV viral proteins is not necessary for the priming of PMN since a priming effect of similar magnitude was observed with UV-inactivated EBV particles. Furthermore, protein synthesis and viral replication inhibitors did not prevent the stimulatory effect of EBV on cPLA₂ phosphorylation. Thus, on the basis of these observations, we speculate that viral adsorption (involving the gp350 of the envelope) to the cell membrane of PMN is sufficient to cause the priming effect reported herein.

The priming effect of EBV on LTB₄ biosynthesis was also observed in whole blood ex vivo, a model which mimics physiological conditions for virus/target cells interactions more effectively. Although in this more complex model LTB₄ generation cannot be attributed to PMN only, we have demonstrated previously that in whole blood stimulated with fMLP, neutrophils account for 80% of LTB₄ biosynthesis, and monocytes for 20% only [46]. Importantly, the priming effect of EBV was observed in the present study with various stimuli which are either receptor-dependent, such as the soluble agonist fMLP or phagocytic particles, or receptor-independent such as the unspecific stimulus ionophore A23187, implicating that the priming by EBV particles is independent of the mechanisms by which LTB₄ biosynthesis is activated and involves events downstream of PMN agonist receptor ligation. Our data also indicated that the interaction of EBV with the PMN triggers a cascade of events which leads ultimately to the enhanced activation of an enzyme involved in LTB₄ production, the cPLA₂. Indeed, EBV priming resulted in enhanced release of AA after PMN stimulation, and an inhibitor of cPLA₂, MAFP [47] inhibited LTB₄ biosynthesis in EBV/A23187-treated PMN, suggesting that cPLA₂ plays a critical role in the priming effect of EBV. This is in agreement with previous reports supporting the involvement of cPLA₂ in the release of AA utilized in the biosynthesis of LTB₄ in agonist-stimulated human PMN [40,48]. It is important to emphasize here that the inhibition of LTB₄ biosynthesis by MAFP does not rule out that other PLA₂; in particular the secreted PLA₂ may also be involved in AA release in EBV-treated PMN, acting in conjunction with the cPLA₂, as demonstrated recently in a macrophage cell line [49]. MAFP is also recognized as an inhibitor of the Ca²⁺-independent PLA₂ [50], an enzyme involved in phospholipid remodelling and reportedly not directly implicated in the release of AA and eicosanoid biosynthesis [51]. The putative mechanism(s) by which EBV primes PMN for enhanced cPLA₂ activation upon cell stimulation have been investigated. Previous studies have shown that phosphorylation of cPLA₂ on Ser-505 [32,35,51] and a calcium-dependent translocation from the cytosol to the cellular membrane are involved in cPLA₂ activation [37–39]. We have found that these two regulatory events are enhanced upon treatment of PMN with EBV, and that the effect on phosphorylation was detectable after only 15 min of PMN exposure to EBV, in agreement with the observed kinetics of priming of LTB₄ biosynthesis by EBV.

The activity of cPLA₂ is regulated by calcium and by phosphorylation by members of the MAP kinase family. Our results demonstrate clearly that EBV is not a calcium-mobilizing agent by itself, but primes neutrophils for increased intracellular calcium mobilization by agonists such as fMLP. We thus postulate that binding of EBV to a specific membrane antigen expressed on neutrophils (an unknown EBV receptor) rapidly activates cellular events causing increased intracellular calcium mobilization and increased translocation of cPLA₂ upon cell stimulation. This is supported by the enhanced rise of Ca²⁺ observed within 1 min of stimulation of EBV-primed PMN with fMLP, which correlates with our kinetic analysis of AA release showing a significant increase in the liberation of AA in EBV-primed PMN as early as 2 min post-stimulation. Furthermore, while it was reported that cPLA₂ can be phosphorylated by erk-1, -2 and p38 MAP kinases, only the latter was found to be putatively involved in EBV-induced priming of neutrophils, as indicated by the stimulatory effect of EBV on fMLP-induced phosphorylation of p38MAP kinase. Interestingly, a recent study has demonstrated that 5-LO is also phosphorylated by p38 MAP kinase [52]; this kinase may therefore exert a dual stimulatory effect at the level of the cPLA₂ and the 5-LO, which may both significantly contribute to the increased capacity of EBV-treated neutrophils to produce LTs. Taken together, these data suggest that the enhanced AA release observed in EBV-treated PMN is causally related to a direct stimulatory effect of EBV on cPLA₂ phosphorylation and to a priming effect of EBV on translocation. It cannot be excluded however, that the priming effect of EBV on LTB₄ biosynthesis in PMN may implicate other mechanisms involved in the regulation of the cPLA₂ and of other enzymes of the AA cascade, in particular the 5-LO itself; this possibility remains to be investigated. Interestingly, previous studies addressing the mechanism(s) of the priming effects of bacterial lipopolysaccharides (LPS) or the growth factor granulocyte-macrophage colony-stimulating factor (GM-CSF) on LTB₄ biosynthesis in human PMN also reported the involvement of enhanced cPLA₂ phosphorylation and translocation [40,48], as well as 5-LO activation in the priming of PMN [48].

Several previous reports have described virus/host cell interactions resulting in alterations at the level of the AA cascade. For instance, the envelope glycoprotein gp120 of HIV-1 was shown to induce the release of LTB₄ and LTC₄ in human mononuclear cells [22] and to activate the expression of 5-LO in human neuroblastoma cells [23]. Similarly, both RSV and SV-40 viruses were shown to up-regulate 5-LO activity in bronchial epithelial cells and fibroblasts, respectively [19,53]. It is noteworthy that unlike the other viruses cited above, EBV does not directly affect LTB₄ biosynthesis but primes PMN for enhanced response to a second stimulus. At present the signal transduction cascade, which ultimately leads to modulatory effects of EBV and other viruses on the AA cascade, is still largely unknown.

Increased LTB₄ biosynthesis could be a way for the virus to attract more PMN (and other leucocytes responsive to the chemotactic effect of LTB₄) at the site of infection. Previous results from our laboratory have shown that EBV induces expression of various chemokines and immunoregulatory molecules in monocytes and PMN [8,9,11,13] and it was suggested that up-regulation of MIP-1α, which is chemotactic for B cells, may enhance recruitment of B cells to the site of infection [13]. Thus, modulation of cytokines and leukotrienes in monocytes and PMN might be part of an important immune process which dictates the outcome of EBV infections.