Secretory phospholipase A2 (sPLA2) produces lipids that stimulate polymorphonuclear neutrophils (PMNs). With the discovery of sPLA2 receptors (sPLA2-R), we hypothesize that sPLA2 stimulates PMNs through a receptor. Scatchard analysis was used to determine the presence of a sPLA2ligand. Lysates were probed with an antibody to the M-type sPLA2-R, and the immunoreactivity was localized. PMNs were treated with active and inactive (+EGTA) sPLA2 (1–100 units of enzyme activity/ml, types IA, IB, and IIA), and elastase release and PMN adhesion were measured. PMNs incubated with inactive, FITC-linked sPLA2-IB, but not sPLA2-IA, demonstrated the presence of a sPLA2-R with saturation at 2.77 fM and aK d of 167 pM. sPLA2-R immunoreactivity was present at 185 kDa and localized to the membrane. Inactive sPLA2-IB activated p38 MAPK, and p38 MAPK inhibition attenuated elastase release. Active sPLA2-IA caused elastase release, but inactive type IA did not. sPLA2-IB stimulated elastase release independent of activity; inactive sPLA2-IIA partially stimulated PMNs. sPLA2-IB and sPLA2-IIA caused PMN adhesion. We conclude that PMNs contain a membrane M-type sPLA2-R that activates p38 MAPK.
- secretory phospholipase A2 receptor
- p38 mitogen-activated protein kinase
secretory phospholipase a2 (sPLA2) has been implicated in diverse inflammatory states including arthritis, pancreatitis, acute chest syndrome in patients with sickle cell anemia, and multiple organ failure (MOF) following traumatic injury (1, 4, 24, 37, 38, 56,69). Moreover, circulating levels of sPLA2 have been identified as a sensitive marker of mortality following pancreatitis, traumatic injury, acute chest syndrome, and sepsis (4, 62, 67,70, 71, 74, 75, 79). Traditionally, the biological activity of these enzymes has been attributed to their ability to cleave lipids, causing release of arachidonate and resulting in eicosanoid generation (1, 4, 24, 37, 38, 56, 67, 69, 83). These lipid mediators then may precipitate cellular activation or even the systemic inflammatory response and predispose the patient to organ injury (55).
The presence of sPLA2 receptors on a variety of cells, including smooth muscle, fibroblasts, Swiss 3T3 cells, and astrocytes, has been documented (2, 3, 27, 28, 32, 43-45, 48, 49, 59,60, 85). Two distinct sPLA2 receptors have been identified: the M-type, present on smooth muscle, and the N-type, found on cells of neural lineage (2, 3, 27, 28, 32, 43-45,47-49, 59, 60, 85). These different receptors display selective affinities for the various groups of mammalian and reptile sPLA2 (2, 3, 27, 28, 30, 32, 43-45, 47-49,59, 60, 85). In addition, these receptors serve important physiological functions as ligand occupancy affects cell physiology, manifested by migration of vascular smooth muscle cells, inhibition of acetylcholine release, proliferation of Swiss 3T3 cells, and tumor invasion (6, 22, 34, 40). In concordance with their association with inflammation, a number of investigators have documented the proinflammatory effects of receptor occupancy (19,20, 73, 82).
The identification of sPLA2 receptors raises the possibility that ligand occupancy on leukocytes may produce inflammation through receptor activation. This study seeks to ascertain whether granulocytes, which are important cellular effectors of the systemic inflammatory response (10, 64, 80), possess sPLA2 receptors. We hypothesize that sPLA2 can directly affect polymorphonuclear neutrophil (PMN) function through occupancy of a membrane receptor. Previous studies from our laboratory demonstrated that sPLA2-IB elicited the release of elastase in the presence of EGTA, but in the presence of EGTA, sPLA2-IA did not cause elastase release (84). Therefore, we focused our investigations to determine the presence of sPLA2 receptors on the PMN membrane, the type of receptor based on immunoreactivity, the changes in PMN function elicited by receptor ligation, and activation of mitogen-activated protein (MAP) kinases conferred by ligand occupancy.
MATERIALS AND METHODS
All chemical reagents, including a fluorescein isothiocyanate (FITC) protein labeling kit and sPLA2-IA (Naja naja) and sPLA2-IB (porcine, specific activity 600 U/mg), were obtained from Sigma Chemical (St. Louis, MO) unless otherwise specified. All reagents employed in these experiments were endotoxin free and were made from sterile water (USP) for intravenous administration, purchased along with sterile 0.9% saline for intravenous injection (USP), from Baxter Healthcare (Deerfield, NY). All buffers were made from injection-grade USP solutions obtained from the following manufacturers: 10% CaCl2, Fujisawa USA (Deerfield, IL); 23.4% NaCl, 20 Meq/ml KCl, and 50% MgSO4, American Regent Laboratories (Shirley, NY); and sodium phosphates (278 mg/ml monobasic and 142 mg/ml dibasic) and 50% dextrose, Abbott Laboratories (North Chicago, IL). In addition, all solutions were sterile-filtered before use by employing Nalgene MF75 series disposable sterilization filter units, purchased from Fischer Scientific (Pittsburgh, PA). Human sPLA2-IIA was the kind gift of Dr. Hubertus Stockinger (Department of Biochemicals, Boehringer Mannheim, Mannheim, Germany), and the units of enzymatic activity were determined via a commercial assay done in Germany employing a specific sPLA2-IIA phosphatidylserine substrate (results not shown). Cy3-labeled goat anti-rabbit IgG was obtained from Jackson ImmunoResearch Laboratories (West Grove, PA); fluorescein-labeled wheat germ agglutinin (WGA) and mannose-BSA were procured from Molecular Probes (Eugene, OR); and Ficoll-Paque and protein A-Sepharose columns were purchased from Pharmacia Biotech (Uppsala, Sweden). Serum-free P1000 medium was obtained from Eiken Chemical (Tokyo, Japan). Polyacrylamide gel electrophoresis (PAGE) was performed on a Bio-Rad mini-gel system by using 4–20% Tris · HCl gels from Bio-Rad (Hercules, CA). Tris-SDS solution was obtained from Owl Scientific (Woburn, MA). Rabbit polyclonal antibodies to activated dual-phosphorylated (Thr180/Tyr182) p38 and activated dual-phosphorylated (Thr202/Tyr204) p42/44 MAPKs as well as a goat anti-rabbit horseradish peroxidase (HRP)-labeled antibody were obtained from Cell Signaling Technology (Beverly, MA). A Gilford Response II UV spectrophotometer was obtained from Ciba Corning Diagnostics (Medfield, MA). A Leica DRM mechanized fluorescence microscope, equipped with a movable stage with a custom Zeiss ×63 water-immersion lens, was purchased from Leica Microsystems (Exton, PA). Four epifluorescence cubes, FITC, Cy3, Cy5, and 7-amino-4-methyl-3-coumarinylacetic acid were obtained from Bioptechs (Butler, PA) and Chroma Technology (Brattleboro, VT). A cooled charge-coupled device camera and Slidebook software for computer operation were purchased from The Cooke (Tonawanda, NY) and Intelligent Imaging Innovations (Lakewood, CO), respectively.
sPLA2 activity assay.
To determine the efficacy of the inhibitor strategy employed, we developed an assay of sPLA2 activity by using a quality control test procedure developed by Sigma Chemical. Briefly, a 2% emulsion of purified l-αphosphatidylcholine from soybeans was prepared in a solution containing 150 mM NaCl and 5 mM CaCl2, pH 8.9 (10 mM NaOH). Ten milliliters of this emulsion were placed in a test tube, and the change in pH was monitored at room temperature. A solution containing 2.5 U/ml sPLA2or buffer was added to the reaction medium and incubated for 10 min. During the reaction, 50-μl aliquots of 10 mM NaOH were added to maintain a pH of 8.9, and the total volume of NaOH required and the exact reaction time (min) were recorded. The dilution factor corresponds to the initial dilution of the sPLA2. sPLA2 activity was calculated from the equation where NaOHRXN is the volume of NaOH added to maintain the pH of the reaction mixture at pH 8.9 and NaOHBL is the volume of NaOH added to the “blank” to reach and maintain a pH of 8.9. This assay was used to assess the ability of 5 mM EGTA or 1–100 μM p-bromophenacyl bromide (BPB) to inhibit enzymatic activity (25). In these assays, 1 unit of enzyme activity was defined as the amount of enzyme needed to hydrolyze 1 μM l-phosphatidylcholine intol-α-lysophosphatidylcholine and fatty acid per minute. These assays were not employed to determine the units of activity for the respective sPLA2 enzymes, which were performed by the manufacturer and confirmed via electronic mail.
PMNs were isolated from heparinized whole blood of healthy volunteers after informed consent was obtained under a protocol approved by the Colorado Multiple Institutional Review Board. Briefly, the isolation procedure consisted of dextran sedimentation, Ficoll hypaque gradient centrifugation, and hypotonic lysis of contaminating red blood cells (63). The final cell population was >99% PMNs as determined by differential staining and was >99% viable as determined by trypan blue exclusion.
FITC labeling of sPLA2.
The instructions for use of the FITC protein labeling kit (Sigma) were followed exactly. Briefly, types sPLA2-IA and sPLA2-IB were suspended in 0.1 M sodium carbonate (pH 9.0). FITC at ratios of 20:1, 10:1, and 5 μl (FITC:sPLA2) dissolved in dimethyl sulfoxide (DMSO) was added per 2 mg of sPLA2. The solution of sPLA2 and FITC was stirred in the dark for 2 h. Running the solution through a Sephadex 20 column separated the unbound FITC. The labeling index was determined for each ratio by calculating the FITC-to-protein molar ratio for optimal labeling by measuring the maximal absorbance of fractions at wavelengths of 280 and 495 nm, respectively. The optimum labeling indices were 38.5 and 37.8% for the 10:1 FITC: sPLA2 ratio, respectively, for two different kits.
Cellular association of sPLA2 with PMNs.
To determine whether exogenously added sPLA2 was associated with PMNs, isolated PMNs were incubated with 10 μg/ml FITC-labeled or unlabeled sPLA2 for 10 min at 37°C and examined by digital microscopy. To examine the effects of enzymatic activity on the cellular association of exogenous sPLA2, the reaction mixture, which contained PMNs, was pretreated for 1 min with 5 mM EGTA or 100 μM BPB before the addition of type IA or type IB FITC-labeled sPLA2, which was dissolved in sterile saline. The associations of labeled, enzymatically inactive type IA and type IB sPLA2 with PMNs were examined by digital microscopy. Because digital microscopy can only examine a small number of cells per sample, we measured the number of sPLA2-labeled PMNs by using flow cytometry. PMNs (106) were warmed to 37°C or kept at 4°C, incubated with EGTA for 30 s, and incubated with 10 μg/ml FITC-labeled sPLA2-IA or sPLA2-IB for 10 min at 37°C or for 30, 60, and 120 min of incubation at 4°C. The samples were assayed by flow cytometry and 5,000 events counted per sample. Controls for these experiments included unlabeled sPLA2-IA or sPLA2-IB and unbound FITC. These experiments were repeated by using PMNs from three different healthy donors.
SPLA2 receptor saturation.
Saturation of the putative sPLA2 receptor was accomplished as a necessary preliminary experiment before we embarked upon formal Scatchard analysis. The reaction mixture, with PMNs, was incubated with EGTA for 1 min to bind Ca2+, and then increasing concentrations of labeled sPLA2-IB were added until the PMN-associated fluorescence was maximally intense as determined by digital microscopy. The quantification of brightness per PMN for all experiments was done by masking FITC signals, calculating the number of FITC pixels per 25 cells, and averaging the mean intensity in pixels/PMN for each sPLA2 concentration. The saturation concentrations were confirmed by flow cytometry.
The existence of a putative sPLA2 receptor was investigated by using standard techniques for Scatchard analysis with fluorescently labeled ligands, except that digital microscopy rather than flow cytometry was employed (5, 26). Briefly, the reaction mixture, containing the isolated PMNs, was pretreated with 5 mM EGTA for 1 min to bind extracellular Ca2+ and was then incubated with differing concentrations of FITC-labeled sPLA2-IB for 10 min at 37°C. The PMNs were fixed with fresh 4% paraformaldehyde and examined by digital microscopy, and brightness was calculated by masking the individual PMNs as explained in sPLA2receptor saturation. Analysis of the curves was accomplished by using standard techniques as previously reported (5, 7, 26, 35). These experiments were repeated three times for all concentrations of FITC-labeled sPLA2, employing two different lots of FITC-labeled sPLA2-IB. An estimate of total brightness was repeated for both aliquots of FITC-labeled of sPLA2 by using the labeled protein alone at the highest concentration employed and serial dilutions of two orders of magnitude. Last, to determine whether unlabeled sPLA2could competitively compete with the labeled compound, isolated PMNs, pretreated with EGTA, were exposed to various concentrations of unlabeled sPLA2 just before (5 s) the addition of a saturating concentration of labeled sPLA2, and the brightness was quantified by digital microscopy. These competitive assays were performed by using methods previously published (46).
Preparation of anti-sPLA2 receptor antibody.
A polyclonal antibody against the recombinant soluble form of the mouse M-type sPLA2 receptor was prepared as follows. cDNA encoding the signal peptide and the presumed extracellular domains (amino acids 1–1365) of the mouse sPLA2 receptor were placed in a mammalian expression vector that utilizes the SRa promoter for the recombinant transcription. This soluble sPLA2receptor expression plasmid was introduced into CHO-K1 cells (American Type Culture Collection) with the calcium phosphate poration method, and cell clones were selected on the basis of acquired resistance to G418 (1 mg/ml). Selected cell lines were then cultured in serum-free P1000 medium, and the conditioned medium was collected. By using a sPLA2-IB-affinity column, soluble mouse M-type sPLA2 receptor was purified as a single 180-kDa band on SDS-PAGE as described previously (28). After immunization in rabbits, antiserum was prepared, and the antibodies were purified by using a protein A-Sepharose column. Specificity of the antibody was demonstrated by immunohistochemistry detection of antibody signal in the glomerular region of mouse kidney that was abolished by the addition of soluble sPLA2 receptor and was not detected in kidneys of sPLA2 receptor deficient mice. In addition, Western blot detected a single band in mouse lung, uterus, and spleen that completely disappeared in the tissues of the sPLA2receptor-deficient mice (results not shown). Furthermore, the antibody did display immunoreactivity in the membranes of a human smooth muscle cell line, cells that express the M-type sPLA2 receptor on their membranes (results not shown).
Separation of proteins from neutrophil lysates and immunoblotting for the sPLA2 receptor.
The proteins from PMN lysates (1 × 106 cell equivalents) were separated by PAGE, transferred to polyvinylidene difluoride (PVDF) membrane, and incubated with a rabbit polyclonal antibody to the M-type sPLA2 receptor. A secondary incubation was performed with a goat anti-rabbit HRP polyclonal antibody, and the immunoreactivity was visualized by using an enhanced chemiluminescence system with exposure of X-ray film.
sPLA2 receptor staining.
Immunofluorescence staining was applied to detect and localize M-type sPLA2 receptors on PMNs. PMNs were isolated and suspended in Krebs-Ringer phosphate (pH 7.35) with 2% dextrose (KRPD). A rabbit polyclonal antibody (see above) against the M-type sPLA2 receptor was added to the PMN suspension in a 1:10 volume dilution and incubated at 4°C for 30 min. After two washes with KRPD, the PMNs were suspended and treated with Cy3-labeled goat anti-rabbit IgG (1:400 dilution) at 4°C for 45 min. Cells were washed and then fixed in 2% paraformaldehyde for 20 min. After being thoroughly washed with PBS, the Cy3-labeled PMNs were smeared onto glass slides, counterstained with fluorescein-labeled WGA (5 μg/ml, for cell surface staining) and bis-benzimide (1 μg/ml, for nuclear staining), and then mounted with aqueous antiquenching medium. To assess the specificity of the immunostaining, an aliquot of cells were suspended in KRPD at 4°C without antibody and treated with Cy-3-labeled goat anti-rabbit IgG (1:400 dilution) for 45 min at 4°C. Fixation and counterstaining were processed under identical conditions. The PMNs were analyzed by digital microscopy.
Separation of proteins from neutrophil lysates and immunoblotting for MAPK activation.
Isolated PMNs (1.25 × 106) were incubated at 37°C over a time course of 0.5–5 min with sPLA2. The cells were then pipetted into sample buffer (Tris-SDS-Eagle's basal medium) and fresh inhibitor mix (40 mM sodium orthovanadate, 1 M nitrophenylphosphate, 100 mM PMSF, and 1 mg/ml leupeptin) and boiled for 5 min. The proteins were then separated by 4–20% gradient PAGE and transferred to PVDF membranes. The membranes were blocked with 5% BSA and then incubated with either a rabbit anti-dual-phosphorylated (Tyr180/Thr182) p38 MAPK antibody or a rabbit anti-dual-phosphorylated (Tyr202/Thr204) ERK1/2 antibody. The membranes were then washed with Trizma-buffered saline (TBS) plus 0.1% Tween and incubated with an HRP-conjugated goat anti-rabbit antibody. Immunoreactivity was visualized by enhanced chemiluminescence and subsequent exposure to X-ray film. Density of the bands was measured by using a Hewlett Packard model 6201C scanner and Scion 4.02 software downloaded from the National Institutes of Health web site.
PMN elastase release assay.
PMNs (1.5 × 106) were warmed in 1.5-ml tubes to 37°C in a shaking water bath, preincubated with saline or 5 mM EGTA, and then stimulated with sPLA2-IA (10 and 100 U/ml), sPLA2-IB (10 and 100 U/ml), and sPLA2-IIA (1–10 U/ml) for 5 min at 37°C. PMNs were also incubated with mannose-BSA (1–10 μM) for 5 min at 37°C, and in selected experiments these PMNs were then activated with 1 μMN-formyl-methionyl-leucyl-phenylalanine (fMLP). The PMNs were then pelleted at 400 g for 5 min, and the supernatant was removed. The supernatants (50 μl) were added to an equal volume of reaction buffer: 0.1 M HEPES and 0.4 M NaCl, pH 7.35. Elastase release was determined spectrophotometrically in the supernatant by the reduction of the specific substrate methoxy-succinyl-alanyl-alanyl-prolyl-valyl p-nitroanilide (AAPVNA) at 405 nm in duplicate. To ensure that the reduction of AAPVNA was secondary to elastase, identical wells containing 5 μM of the specific elastase inhibitor methoxy-succinyl-alanyl-alanyl-prolyl-valyl chloromethyl ketone (AAPVCK) were run in conjunction with each treatment. Elastase release is reported as the percentage of total cellular elastase as determined by 0.1% Triton-X paired treatment of an identical number of PMNs (63). Both fMLP (1 μM)-activated and platelet-activating factor (PAF; 2 μM)-primed PMNs activated with fMLP were employed as controls, and these assays were undertaken in the presence or absence of 5 mM EGTA.
Inhibition of sPLA2-mediated elastase release.
PMNs were incubated for 30 min in the dark at 37°C with DMSO, 0.1–10 μM SB-203580, or 0.1–10 μM PD-98059 to inhibit p38 MAPK activity and p42/44 MAPK activation, respectively (12,18). The reaction mixture with the PMNs was incubated with 5 mM EGTA for 1 min and then stimulated with 100 U/ml sPLA2-IB. Elastase release was measured as previously described (9).
PMN adherence assay.
PMN adherence to fibrinogen-coated plates was completed by using inactive sPLA2 (+EGTA) as previously described (9). The results are expressed as the percentage of adherent PMNs from the leukocytes isolated from eight healthy donors.
The means, standard deviations, and standard errors of the mean were calculated by using standard techniques. Statistical differences among groups were determined by using a paired or independent analysis of variance followed by Scheffé's post hoc analysis for multiple comparisons. Statistical significance was determined at theP < 0.05 level.
Inhibition of sPLA2 activity.
As a necessary preliminary, an in vitro assay based on the change in pH was employed to determine the effects of EGTA Ca2+chelation and BPB incubation on sPLA2 activity. Chelation of Ca2+ with EGTA (5 mM) inhibited the activity of sPLA2-IA, sPLA2-IB, and sPLA2-IIA by 98 ± 3, 98 ± 4, and 99 ± 2%, respectively, supporting previous reports that Ca2+ is required for sPLA2 enzymatic activity (25). In addition, pretreatment with 100 μM BPB also inhibited sPLA2-IA, sPLA2-IB, and sPLA2-IIA activity by 98.5 ± 4 to 99 ± 3%. Lesser concentrations of EGTA and BPB did not completely inhibit sPLA2 activity for any of the types tested (results not shown). Thus EGTA preincubation is an effective inhibitor of sPLA2 activity, which was confirmed by the use of a second antagonist. These assays were not employed to ascertain the units of sPLA2 activity but to determine whether Ca2+ chelation and BPB treatment were efficacious methods to inhibit sPLA2 activity (25).
Membrane binding of sPLA2.
Because sPLA2 may act as a ligand to provoke elastase release, we examined the cellular association of sPLA2 with PMNs (69). These studies used FITC-labeled sPLA2-IA and sPLA2-IB and both flow cytometry and digital microscopy. Digital microscopy revealed strong cellular association of sPLA2-IB but not sPLA2-IA (Fig.1). Because enzymatically inactive sPLA2-IB was avidly associated with the PMN membrane, and digital microscopy is only able to examine a limited number of cells from a given sample, the number of PMNs labeled with inactive sPLA2 was measured by flow cytometry (Fig.2). The incubations with labeled sPLA2 were either identical to those of the digital microscopy, 10 min at 37°C, or the PMNs were incubated for 30, 60, and 120 min at 4°C. Controls for these experiments included unlabeled sPLA2 and unattached “free” FITC. Compared with PMNs treated with unlabeled sPLA2-IB (Fig. 2, peak A), PMNs incubated with FITC-linked sPLA2-IB for 30 min at 4°C demonstrated almost a 2-log shift in mean fluorescence intensity at 30 min (Fig. 2, peak B) without marked increases with longer incubation times (results not shown), and ∼98% of the PMNs evidenced cellular association of labeled-sPLA2. In addition, incubation of PMNs with FITC-labeled sPLA2 for 10 min at 37°C yielded a 10-fold shift in mean fluorescence intensity (unlabeled sPLA2: 8.4 ± 1.2 vs. labeled sPLA2: 127 ± 25) and a 24.4 ± 1.0 shift compared with buffer with free FITC-treated PMNs with labeling of 70% of the cells (results not shown). In addition, FITC-labeled sPLA2-IA did demonstrate minimal cell association of this sPLA2 when EGTA was deleted, implying that the enzymatic activity of type IA sPLA2 may occur in the PMN plasma membrane (results not shown). These findings provide supportive evidence that a receptor for sPLA2 resides on the PMN membrane.
Scatchard analysis of the sPLA2 receptor on PMNs.
Before Scatchard analysis was initiated, it was important to determine saturation of the putative receptor. FITC-labeled sPLA2-IB was added in increasing concentrations of isolated PMNs, and the fluorescent intensity was calculated. Saturation of the putative sPLA2 receptor was found to be 2.77 fM (Fig.3 A, n = 3). After the receptor saturation was determined, Scatchard analysis was completed (37°C with a 10-min incubation time) and the graphical relationship was revealed (Fig. 3 B). From the shape of the curve, the receptor is heterogeneous and there is negative cooperativity. From standard calculation, the dissociation constant (K d) is 167 pM, and the calculated number of receptors per cell is ∼1,672 (5, 7, 35). To ensure that the labeled sPLA2 affinity for the PMN membrane was specific for the protein itself and not due to the labeling of sPLA2 to FITC, isolated PMNs, pretreated with 5 mM EGTA, were incubated with FITC-labeled sPLA2-IB in the presence of unlabeled sPLA2-IB (Fig. 3 C). As the concentration of unlabeled sPLA2 was increased, the amount of cell-associated FITC-tagged sPLA2 decreased, demonstrating that the unlabeled sPLA2 was able to compete for the sPLA2 receptor sites on the PMN.
Presence and location of the M-type sPLA2 receptor.
To determine the presence of the M-type sPLA2 receptor on PMNs, the proteins from whole cell lysates were separated, transferred to nitrocellulose, and probed with a polyclonal antibody to the M-type sPLA2 receptor. A single band of immunoreactivity was identified at ∼185 kDa (Fig. 4). To confirm the membrane location of this receptor, we employed a triple stain technique and digital microscopy. Isolated PMNs were incubated with the polyclonal antibody, fixed with paraformaldehyde, and smeared onto slides. After appropriate washing, the membranes were labeled with fluorescein-labeled WGA (green), the nuclei were labeled with bis-benzimide (blue), and the receptor immunoreactivity was labeled with a Cy3-labeled secondary antibody (red) (Fig.5). The bright red signal shown in Fig.5 C (anti-sPLA2 receptor antibody + Cy3-labeled secondary antibody) demonstrates the presence of the sPLA2 receptor on the PMNs. The lack of red signal in Fig.5 A (Cy3 secondary antibody alone) confirms that the red shown in Fig. 5 C is not due to nonspecific binding of the Cy3 secondary antibody. The receptor labeling correlated directly with the membrane labeling. The sPLA2 receptor plus Cy3 secondary antibody-produced area of red staining overlapped the green membrane staining as shown by the yellow color (green overlapping red) in Fig. 5 D. The Cy3-only stained cells had no evidence of any yellow intensity. Thus the receptor localizes to the PMN membrane, because the intact, live PMNs were incubated with the polyclonal antibody to the M-type receptor before fixation.
sPLA2-mediated elastase release.
To confirm that sPLA2 was able to affect degranulation, PMNs were treated with sPLA2-IA, sPLA2-IB, and sPLA2-IIA, and elastase release was measured (Fig.6). To inhibit sPLA2activity, the reaction mixture, which includes the PMNs, was treated with 5 mM EGTA for 1 min before the addition of sPLA2. All three types of sPLA2 caused the release of significant amounts of elastase from PMNs compared with buffer-treated control cells [sPLA2-IA: 100 U/ml = 16.0 ± 0.2%, 10 U/ml = 13.5 ± 3.0%; sPLA2-IB: 100 U/ml = 32.0 ± 6.0%, 10 U/ml = 12.5 ± 2.0%; and sPLA2-IIA: 10 U/ml = 30.0 ± 4.0%, 1 U/ml = 15.0 ± 3.0% vs. control: 5.1 ± 0.1% (n = 8, P < 0.05)]. Type IB sPLA2 had a response at 100 U/ml that was equivalent to that of PAF (200 nM)-primed and fMLP (1 μM)-activated elastase release (46.1 ± 4.5%).
Having confirmed that “active” sPLA2 directly stimulated PMNs to release elastase, we tested the effects of enzymatically inactive sPLA2 on PMN elastase release. The reaction mixture, which included the PMNs, was pretreated with 5 mM EGTA for 1 min before the addition of sPLA2. In the presence of EGTA, sPLA2-IA did not cause release of elastase, indicating that sPLA2-IA requires enzymatic function to facilitate elastase release (Fig. 6). In contrast, EGTA preincubation of sPLA2-IB did not affect its ability to cause elastase release from PMNs (Fig. 8). Finally, sPLA2-IIA was able to stimulate elastase release in the presence of EGTA, but its activity was attenuated by 39.7 ± 0.3% (Fig. 6). This attenuation was statistically significant at 10 U/ml (P < 0.05, n = 8). These data indicated that sPLA2-IA, from N. naja venom, caused PMN elastase release through its enzymatic function but not secondary to ligand occupancy. Type IB sPLA2 demonstrated significant PMN elastase release regardless of its enzymatic activity, implying stimulation of a membrane receptor. Type IIA sPLA2induced PMN elastase release that appeared to be the result of both enzymatic activity and receptor activation. Inclusion of sPLA2-IIA was important because of its reported role in systemic inflammation in humans (23, 67, 71, 76, 77). As controls, we determined the role of extracellular Ca2+ in fMLP-mediated elastase release in both buffer-pretreated and PAF-primed PMNs. Chelation of extracellular Ca2+ moderately decreased fMLP-elicited elastase (fMLP: 10.3 ± 0.5% vs. fMLP + EGTA: 8.3 ± 0.7%); however, chelation of extracellular Ca2+ did significantly decrease PAF priming of fMLP-mediated elastase release (PAF/fMLP: 48 ± 3.2% vs. PAF/fMLP + EGTA: 32 ± 6.8%), though not nearly to the levels of fMLP alone (10.3 ± 0.5%) or the buffer-treated controls (4 ± 0.5%). These results were concordant with previous data that did not find a strict requirement for extracellular Ca2+ in PMN exocytosis (50, 51, 61, 68).
Because it has been postulated that the sPLA2 receptor may represent the mannose receptor on PMNs and other leukocytes, we employed mannose bound to BSA to determine its effects on PMN elastase release in the presence or absence of enzymatically inactive sPLA2-IIA (73). As demonstrated in Table1, mannose-BSA did not cause a significant release of elastase from PMN, nor did it affect the sPLA2-IIA-mediated elastase release in the presence of EGTA. Furthermore, BSA alone did not affect elastase release from control or sPLA2-IIA stimulated PMNs (Table 1, results not shown). In addition, mannose-BSA primed fMLP-mediated elastase release 5.1 ± 1.0-fold, whereas BSA alone did not affect PMN elastase release in response to fMLP (results not shown).
sPLA2-mediated PMN adhesion.
To ensure that sPLA2 affected other PMN functions in a nonenzymatic manner, we assayed the ability of enzymatically inactive (EGTA preincubation) sPLA2-IB and sPLA2-IIA to affect PMN adhesion to fibrinogen-coated plates. It is important to note that β2-mediated integrin adhesion of PMNs to RGD ligands is not dependent on extracellular Ca2+; rather, PMNs require extracellular Mg2+ for firm adhesion (11, 15, 33). SPLA2-IB and sPLA2-IIA were employed because they both elicited PMN elastase independent of their enzymatic activity. Both sPLA2-IB and sPLA2-IIA caused PMN adhesion to fibrinogen-coated plates in a concentration-dependent fashion from 1–100 U/ml [IB: 100 U/ml, 19 ± 0.8%, 10 U/ml, 15 ± 1.5%, and 1 U/ml, 12 ± 1.4%; IIA: 100 U/ml, 22 ± 0.3%, 10 U/ml, 18 ± 1.2%, and 1U/ml, 12 ± 1.4% vs. buffer control: 4 ± 0.5% (P < 0.05, n= 9)] (Fig. 7). Lesser concentrations of any type of sPLA2 and exclusion of EGTA pretreatment did not affect PMN adhesion (results not shown). In addition, the higher concentrations of inactive sPLA2 elicited PMN adhesion similar to the positive controls [fMLP (1 μM), 22 ± 3.1%; PAF (2 μM), 19 ± 2.0%].
Examination of receptor-dependent sPLA2 signaling pathways.
MAPKs are a family of receptor-linked enzymes that are integrally involved in PMN signaling (39, 58). Our previous work has confirmed the pivotal role of p38 MAPK in receptor-provoked elastase release (84). Furthermore, p38 MAPK has been implicated in sPLA2 receptor activation in human astrocytoma cells (31). Activation of p38 MAPK (dual phosphorylation) was measured by using an antibody that recognizes only the dual-phosphorylated (Thr180/Tyr182) form of p38 MAPK, and previous work has demonstrated that dual phosphorylation of p38 MAPK correlates directly with activity (18). Rapid dual phosphorylation (3.1 ± 0.5-fold increase at 30 s) of p38 MAPK occurred with persistence of the signal for 2 min and diminution of the signal at 5 min following stimulation with inactive sPLA2-IB (Fig. 8). To confirm these data, a second inhibitor of sPLA2 enzymatic activity, BPB (100 μM), was employed and demonstrated identical results. Furthermore, EGTA-inactivated synovial sPLA2 caused a fourfold increase in activated p38 MAPK with diminution of the signal at 5 min. Neither BPB nor EGTA caused p38 MAPK activation (data not shown). In addition, because p42/44 MAPK is another important signaling molecule in PMNs, we tested the ability of sPLA2 to activate this MAPK. Incubation of PMNs with sPLA2 did not activate (dual phosphorylation) p42/44 MAPK (data not shown) (16,53).
To demonstrate that sPLA2-mediated elastase release was dependent on activation of p38 MAPK, PMNs were preincubated with the specific p38 MAPK inhibitor SB-203580 or the selective MAP kinase kinase (MEK)1/2 inhibitor PD-98059 (12). In these experiments, elastase release was measured after the addition of BPB-inactivated sPLA2. Inhibition of p38 MAPK decreased elastase release in a concentration-dependent fashion with maximal inhibition (36 ± 3.4%) at 10−5 M SB-203580 (P < 0.05, n = 8; Fig.9). In contrast, the specific MEK1 inhibitor PD-98059, which obviates p42/44 MAPK activity, did not inhibit elastase release.
The sPLA2 are members of a large family of enzymes that are capable of producing biologically active compounds and signaling molecules from membrane lipids (13, 14, 25). Such phospholipase activity has been delineated in the production of eicosanoids, the remodeling pathway of PAF biosynthesis, and the generation of other intracellular signaling molecules (13, 14,25, 67). Both eicosanoids and PAF have been implicated in a number of PMN-mediated syndromes of organ injury (1, 4, 24, 37,38, 56, 66, 69, 83). Increased circulating levels of sPLA2 have been demonstrated in a number of PMN-mediated syndromes, including acute lung injury, postinjury MOF, acute pancreatitis, and arthritis (4, 62, 67, 70, 71, 74, 75,79). Moreover, the levels of sPLA2 and/or sPLA2 activity in these maladies correlate with patient outcome (56, 70, 71, 74, 75, 84). For example, the level of sPLA2 activity in sickle cell patients accurately predicts the patients who will develop acute chest syndrome, a life-threatening variant of acute lung injury that is particular to patients with sickle cell anemia (71). Few data with respect to the sPLA2 receptor in PMN-mediated organ injury exist; however, sPLA2-IB receptor-deficient mice are resistant to endotoxic shock (30). Although the precise mechanisms of protection against endotoxic shock is not known, these studies provide provocative evidence that sPLA2 receptors may have relevance in inflammatory diseases (30). Collectively, these clinical data and animal models strongly suggest that the activity of sPLA2 is associated with inflammation and the development of acute, multiple organ injury, a syndrome that is PMN mediated (55).
The current study has demonstrated that the M-type sPLA2receptor was present on the PMN membrane in numbers similar to the PAF receptor, ∼1,600 for sPLA2 vs. 1,100 for PAF (45). Receptor occupancy caused activation of p38 MAPK that was directly linked to PMN elastase release. The cellular association of inactive sPLA2-IB and sPLA2-IIA with PMNs and the inability of inactive sPLA2-IA to associate with the PMN membrane are entirely consistent with previous reports of the smooth muscle M-type receptor affinity for specific sPLA2 isoenzymes (40, 43-45, 49). Furthermore, PMN receptor occupancy by both sPLA2-IB and sPLA2-IIA caused PMN adhesion to RGD ligands, elastase release, and activation of p38 MAPK, all in the absence of phospholipase activity. Furthermore, pretreatment of PMNs with mannose-BSA had little effect on sPLA2-IIA-mediated elastase release and provides evidence that the M-type sPLA2 receptors and the mannose receptor on PMNs comprise distinct entities (73). SPLA2-IIA activation of p38 MAPK was rapid, within 30–60 s, and this time course was consistent with both elastase release, which begins at 3 min and is maximal at 5 min (results not shown and Fig. 8), and activation of p38 MAPK by other PMN agonists, including fMLP or PAF (18, 39, 57,58). To demonstrate that p38 MAPK activation leads to the release of elastase, we used the p38 inhibitor SB-203580 to block activated p38 MAPK. SB-203580 attenuated the release of elastase but did not eliminate it completely, similar to SB-203580 inhibition of fMLP activation of PMN oxidase (58). Although previous reports have documented activation of p42/44 MAPKs as the result of sPLA2 receptor occupancy in astrocytes (31, 41,54), we were unable to demonstrate p42/44 MAPK activation in PMNs following incubation with sPLA2. Moreover, specific MEK1 inhibition with PD-98059, which abrogates p42/44 MAPK activation, did not affect sPLA2-receptor mediated elastase release. The reason for the incomplete inhibition with SB-203580 may lie in parallel signaling pathways. There are multiple signaling pathways for elastase release in the PMN that involve the activation of a number of other kinases, including both protein kinase C and phosphatidylinositol 3-kinase (17, 65, 78). It is likely that there are redundant signaling pathways involved with sPLA2 receptors that merit further investigation.
The ability of active sPLA2-IA and sPLA2-IIA to affect PMN elastase release implies cleavage of membrane lipids. Previous studies have demonstrated that sPLA2 induces the surface expression of β2-integrins on the PMN, an activity ascribed to its phospholipase activity; moreover, sPLA2-IB can also cause the release of β-glucuronidase from macrophages (72, 73, 82). However, to date, there is little evidence to suggest that sPLA2 can digest the membranes of healthy PMNs, though their ability to cleave the membranes of apoptotic cells and cancerous or transformed cell lines has been reported (8, 19, 42, 81). It is important to note that there are no plasma lipid substrates for sPLA2 to cleave into eicosanoids or other active compounds that could then produce the observed degranulation. Human plasma and bone marrow, however, contain a number of lipid classes, which are carried on human albumin that sPLA2 may cleave to generate active metabolites (24,52, 66, 72, 82). Lipid extraction of the supernatant of isolated PMNs does not contain a detectable amount of chloroform-soluble compounds as assayed by diode array spectroscopy at 235 nm or by gas chromatography mass spectroscopy (results not shown). The activation of PMNs by exogenous secretory phospholipase activity was not expected and merits further investigation.
To date, most of the sPLA2 binding data of the M-type receptor employed transfection of COS, 293, or CHO cells, with one transfection yielding only transient expression of the human M-type receptor (3, 32). Such transfection models are attractive but may have little correlation with native receptor ligand interactions due to differences in cellular processing of the receptor protein or the requirements of accessory proteins in the binding of ligand. Two reports have suggested the possibility of such accessory proteins for binding sPLA2 to the M-type receptor; moreover, a recent review stressed the need for sPLA2binding studies employing cells with “native” receptors (21,29, 36).
To our knowledge, the existence of sPLA2 receptors on PMNs has not been previously reported, and their presence may further explain the role of sPLA2 in inflammatory processes that implicate the release of inflammatory mediators. Recently, sPLA2-IA and sPLA2-IIA were shown to activate human pulmonary macrophages, causing both release of β-glucuronidase and production of IL-6; moreover, inhibition of sPLA2activity did not affect β-glucuronidase release; however, treatment of these macrophages with p-aminophenyl-mannopyranoside-BSA (mp-BSA), a mannose receptor agonist, caused similar release of β-glucuronidase (73). In addition, pretreatment with mp-BSA did not affect β-glucuronidase release by sPLA2-IIA but augmented release by sPLA2-IA (73). Taken together, these findings indicate that both the sPLA2-IA and sPLA2-IIA appear to activate human macrophages in lieu of their enzymatic activity and through either the mannose receptor or an sPLA2-specific receptor (73). In contrast, the current study characterizes the function of sPLA2 receptors on PMNs and provides evidence that these receptors could be involved in PMN-mediated inflammation, a role in which sPLA2 has historically been placed (1,4, 22, 24, 34, 37, 38, 40, 56, 67, 69, 74-77, 79). This proinflammatory effect from sPLA2 receptor stimulation may be clinically important in diverse arenas, including PMN-mediated organ dysfunction. Both acute lung injury and postinjury MOF have been associated with increased levels of sPLA2; however, the methods used to measured sPLA2 include ELISAs, or activity assays that employed varied substrates, making comparisons among these studies difficult (56, 62, 67, 74-77). Further work employing a uniform measurement of sPLA2 activity is required to identify clinically relevant levels of these enzymes. Thus modulation of sPLA2 receptor occupancy as well as its inherent phospholipase activity has the potential to lead to novel therapeutic strategies that could attenuate the inflammatory response seen in these clinical scenarios.
Address for reprint requests and other correspondence: E. E. Moore, Dept. of Surgery, Denver Health Medical Center, 777 Bannock St., Denver, CO 80204 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society