HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 7, 5061-5073, February 15, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/7/5061    most recent M109083200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Degousee, N. Articles by Rubin, B. B. Search for Related Content PubMed PubMed Citation Articles by Degousee, N. Articles by Rubin, B. B. Social Bookmarking                      What's this?

Groups IV, V, and X Phospholipases A₂s in Human Neutrophils

ROLE IN EICOSANOID PRODUCTION AND GRAM-NEGATIVE BACTERIAL PHOSPHOLIPID HYDROLYSIS^*

From the ^a Division of Vascular Surgery, Max Bell Research Center, Toronto General Hospital, University Health Network, Toronto M5G 2C4, Canada, the ^b Departments of Chemistry and Biochemistry, University of Washington, Seattle, Washington 98195, ^c Granulocyte Research Laboratory, Department of Hematology, Copenhagen University Hospital, Copenhagen DK-2100, Denmark, ^d Medical Research Council Group in Membrane Biology, Departments of Medicine and Biochemistry, Medical Sciences Building, University of Toronto, Toronto, Ontario M5S 1A8, Canada, ^e Clinic of Internal Medicine IV, Department of Gastroenterology and Hepatology, University of Vienna, Vienna A-1090, Austria, the ^f Institut de Pharmacologie Moleculaire et Cellulaire, CNRS-UPR 411, 660 Route des Lucioles, Sophia Antipolis, 06560 Valbonne, France, ^g Department of Allergy and Immunology, Harvard Medical School, Boston, Massachusetts 20115, and the ^h Department of Genetics, Rutgers, the State University of New Jersey, Piscataway, New Jersey 08854-8082

Received for publication, September 20, 2001, and in revised form, December 5, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The bacterial tripeptide formyl-Met-Leu-Phe (fMLP) induces the secretion of enzyme(s) with phospholipase A₂ (PLA₂) activity from human neutrophils. We show that circulating human neutrophils express groups V and X sPLA₂ (GV and GX sPLA₂) mRNA and contain GV and GX sPLA₂ proteins, whereas GIB, GIIA, GIID, GIIE, GIIF, GIII, and GXII sPLA₂s are undetectable. GV sPLA₂ is a component of both azurophilic and specific granules, whereas GX sPLA₂ is confined to azurophilic granules. Exposure to fMLP or opsonized zymosan results in the release of GV but not GX sPLA₂ and most, if not all, of the PLA₂ activity in the extracellular fluid of fMLP-stimulated neutrophils is due to GV sPLA₂. GV sPLA₂ does not contribute to fMLP-stimulated leukotriene B₄ production but may support the anti-bacterial properties of the neutrophil, because 10-100 ng per ml concentrations of this enzyme lead to Gram-negative bacterial membrane phospholipid hydrolysis in the presence of human serum. By use of a recently described and specific inhibitor of cytosolic PLA₂- (group IV PLA₂), we show that this enzyme produces virtually all of the arachidonic acid used for the biosynthesis of leukotriene B₄ in fMLP- and opsonized zymosan-stimulated neutrophils, the major eicosanoid produced by these pro-inflammatory cells.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Neutrophils (polymorphonuclear leukocytes) are one of the principal effector cells in the inflammatory response. Following activation, neutrophils release a barrage of cytotoxic products, such as reactive oxygen species, degradative enzymes, and phospholipases, including phospholipase A₂ (PLA₂)¹ (1-3). PLA₂ enzymes catalyze the hydrolysis of phospholipids, yielding free fatty acids and lysophospholipids. The identity of the PLA₂ enzyme(s) that are contained in and secreted by human neutrophils has not been defined.

A family of 10 secreted PLA₂ (sPLA₂) enzymes has been described in mammals that currently includes group (G) IB, IIA, IIC, IID, IIE, IIF, III, V, X, and XII sPLA₂ (4, 5). sPLA₂ enzymes exert biological effects through multiple mechanisms, including the release of arachidonic acid, which may be metabolized to leukotrienes and prostaglandins (6, 7), bactericidal activity via hydrolysis of the outer membrane of Gram-positive bacteria (8, 9), and through binding to specific sPLA₂ receptors (10-12). The different sPLA₂s are not close isoforms of each other because their amino acid sequences are ~30-50% identical among the paralogs. This plus the fact that the sPLA₂ enzymes have distinct tissue distributions argue for different physiological functions for each enzyme (5, 13). Thus, GIB sPLA₂ has been identified in the pancreas and functions in phospholipid digestion (14) but is also found in non-digestive tissues, where its functions remain unknown (5). GIIA sPLA₂ is expressed at high levels during inflammatory reactions (15, 16) and, until recently (4, 5, 13), was thought to be the principal sPLA₂ isoform in the immune system. Human GIIA sPLA₂ mRNA has been identified in myocardium, skeletal muscle, lung, liver, placenta, prostate, and the small and large intestine but was not detected in peripheral blood leukocytes (17). GIIC sPLA₂ is present in the mouse but appears as a pseudo gene in humans (4). GIID sPLA₂ mRNA was detected in human spleen, thymus, small intestine, colon, lung, pancreas, and placenta (17, 18), whereas GIIE sPLA₂ expression in humans is restricted to the brain, heart, lung, and placenta (19). mRNA coding for GIIF sPLA₂ has been identified in many human tissues, with the highest levels detected in the placenta, thymus, prostate, testes, kidney, liver, and thyroid (17). GIII sPLA₂, which is homologous to bee venom PLA₂, is expressed in the pancreas, kidney, liver, lung, skeletal muscle, and myocardium (17, 20). GV sPLA₂ has been identified in myocardium, placenta, mast cells, and macrophages (21-24). Although GV sPLA₂ was shown to contribute to platelet-activating factor-mediated PGE₂ production by a murine macrophage-like cell line (24), GV sPLA₂ mRNA was not identified in human peripheral blood leukocytes (17). GX sPLA₂ is expressed in organs associated with the immune response (25) and induces cyclooxygenase-2-dependent PGE₂ synthesis by adherent mammalian cells (26-28). One study identified GX sPLA₂ mRNA in peripheral blood leukocytes (25), whereas two others did not (17, 29). Transcripts for GXII sPLA₂ were identified in multiple tissues, including myocardium, skeletal muscle, kidney, pancreas, and type 2 helper T cells (30, 31).

sPLA₂ enzymes share a number of structural characteristics, including several intramolecular disulfide bridges, a Ca²⁺-binding loop, and the requirement for sub-millimolar to millimolar concentrations of Ca²⁺ for catalytic activity (32-34). In addition, some of the sPLA₂ enzymes have N-terminal prepropeptide sequences of varying lengths (5, 13). The mechanism by which prepropeptide sPLA₂ enzymes mature into secreted proteins has not been fully defined. Some sPLA₂ enzymes have dibasic motifs at the C terminus of their signal sequence, such as the arginine doublet in GX sPLA₂ (25), that may be efficiently cleaved by subtilisin-like protein convertases in the Golgi apparatus (35). Some sPLA₂ enzymes, including GIB and GX sPLA₂, are secreted as proenzymes and then cleaved to yield mature, catalytically active proteins, whereas others, like GIIA and GV sPLA₂, do not have an extra N-terminal peptide. GX sPLA₂ also contains a single consensus sequence for post-translational modification, the N-glycosylation acceptor site (Asn-X-Ser/Thr) at Asn-113 (25), and N-glycosylation of GX sPLA₂ was demonstrated when this enzyme was expressed in HEK293 cells (36).

PLA₂ enzymatic activity has been detected in intracellular granules in human neutrophils (37-39). Subcellular fractionation experiments have shown that neutrophils have a heterogenous population of granules that have distinct intra-granular and membrane-bound proteins (40). Thus, neutrophils contain azurophilic, specific, and gelatinase granules, as well as secretory vesicles, that function as regulated storage organelles. When neutrophils are stimulated, the granules may release their protein contents into the extracellular environment, or may fuse with phagosomes to form phagolysosomes, where the contents of the granules cooperate in the killing of microbes (41). Fusion of granule membranes, which are important reservoirs of membrane-bound proteins, with the plasma membrane and phagolysosomes may also participate in the eradication of microbes (40). The identity of the granules that contain sPLA₂ enzymes in human neutrophils has not been established.

A large body of experimental evidence supports the hypothesis that agonist-stimulated release of arachidonic acid from phospholipids is mediated, at least in part, by cPLA₂ (reviewed in Ref. 42). First, cPLA₂ selectively hydrolyzes phospholipids with arachidonic acid in the sn-2 position (43). Second, exposing neutrophils to fMLP results in a decrease in the electrophoretic mobility of cPLA₂, a finding consistent with cPLA₂ phosphorylation, and stimulates the translocation of cPLA₂ from cytosolic to microsomal and nuclear compartments (44, 45). Third, co-incubating neutrophils with the cPLA₂ inhibitor methylarachidonyl fluorophosphonate (MAFP) decreases fMLP-stimulated arachidonic acid mass release (44, 45). Fourth, peritoneal macrophages from mice subjected to targeted disruption of the cPLA₂ gene (cPLA₂^/) produce less PGE₂, LTB₄, and platelet-activating factor following exposure to inflammatory stimuli than macrophages from wild type mice (cPLA₂^+/+) (46, 47). Three isoforms of cPLA₂ have been identified as follows: cPLA₂, cPLA₂, and cPLA₂ (43, 48, 49). Although cPLA₂ is thought to mediate arachidonic acid release by a variety of cells (42), the physiologic roles of cPLA₂ and cPLA₂ have not been defined.

The purpose of this study was to identify the sPLA₂ enzymes that are expressed by neutrophils, define the subcellular localization of these enzymes, determine which sPLA₂ enzymes are released by neutrophils, and evaluate the role of the extracellular sPLA₂ enzymes in neutrophil LTB₄ biosynthesis and Gram-negative bacterial phospholipid hydrolysis. We identified GV and GX sPLA₂ mRNA in a homogeneous population of human neutrophils, showed that these cells contain GV and GX sPLA₂ protein, demonstrated that GV and GX sPLA₂ are both present in azurophilic granules, and found that GV sPLA₂ is also a component of specific granules. Furthermore, we showed that GV sPLA₂ was released into the extracellular environment following exposure to formyl-Met-Leu-Phe (fMLP) or opsonized zymosan (OZ), whereas mature GX sPLA₂ was not. Inhibition of extracellular sPLA₂ activity with the active site-directed, tight-binding inhibitor indoxam (50), which was found to inhibit GV sPLA₂ activity 125-fold more efficiently than GX sPLA₂ activity, had no effect on fMLP-stimulated neutrophil LTB₄ synthesis. In contrast, pretreatment of neutrophils with pyrrolidine-1 (51), a highly specific inhibitor of cPLA₂ activity that has no effect on the catalytic activity of recombinant GIIA, V, or X sPLA₂ (52), abrogated fMLP- and OZ-induced neutrophil LTB₄ biosynthesis. Finally, we showed that recombinant GIIA or GV sPLA₂, but not GX sPLA₂, efficiently hydrolyzed [³H]oleic acid from [³H]oleic acid labeled live Escherichia coli in the presence of serum.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Ficoll-Paque was from Amersham Biosciences and was used for neutrophil isolation. DNase I RNase-free was from Qiagen (Mississauga, Ontario, Canada). Thermus aquaticus (Taq) polymerase was from MBI Fermentas (Burlington, Ontario, Canada). Rhodamine-phycoerythrin (R-PE)-conjugated anti-CD16 monoclonal antibody (clone 3G8; Monosan, Uden, Netherlands), fluorescein isothiocyanate (FITC)-conjugated anti-CD19 (Immunotech, Luminy, France), FITC-conjugated anti-CD3 (Beckman Coulter Inc., Fullerton, CA), and FITC-conjugated anti-HLA-DR (Immunotech, Luminy, France) were used for cell sorting procedures. [9,10-³H]Oleic acid (NFT 289, 5 Ci/mmol) was from PerkinElmer Life Sciences. Myocardial RNA was from Ambion (Austin, TX). DNA sequence analysis was performed at the "DNA Sequencing Facility" of The Hospital for Sick Children (Toronto, Ontario, Canada). Isolation of the IgG fraction from serum was done with the ImmunoPure Immobilized Protein G Plus Orientation Kit (Pierce). The E. coli strain pldA was kindly provided by Dr. Peter Elsbach (New York University School of Medicine, New York). Zymosan was obtained from Sigma, and OZ was prepared as described previously (53). Indoxam and pyrrolidine-1 were synthesized as described elsewhere (50, 52).

Neutrophil Isolation-- After obtaining informed consent from healthy donors that were not taking any medications, 100 ml of blood was obtained and anticoagulated in citrate. Neutrophils were then isolated by dextran sedimentation and Ficoll-Paque density gradient centrifugation, exactly as described (44).

Neutrophil Purification by Fluorescence-activated Cell Sorting (FACS) Analysis-- Neutrophils (1 × 10⁷) enriched by Ficoll centrifugation were incubated for 30 min at 4 °C with 20% heat-inactivated AB serum prepared in KRPD buffer (where KRPD is Krebs-Ringers phosphate dextrose), washed twice with KRPD, and simultaneously labeled for 30 min at 4 °C with anti-CD16 R-PE, anti-CD19 FITC, anti-CD3 FITC, and anti-HLA-DR FITC, according to the manufacturer's instructions. After two washes with cold KRPD buffer, cells were subjected to dual channel FACS analysis on a MoFlo cytometer (Cytomation Inc., Fort Collins, CO) using a 150-megawatt Coherent Innova enterprise II ion laser (Coherent Inc., Santa Clara, CA) tuned at 488 nm and equipped with CyCLOPS Summit software. Fluorescence was measured using 570/40 (R-PE) and 530/40 (FITC) band pass filters. CD16⁺, CD19, CD3, and HLA-DR cells were separated by sorting in the "sort purify" mode setting with a flow rate of 10,000 cells/s. Gates were set to exclude debris and nonviable cells on the basis of light scatter properties. Neutrophils were defined by a combination of forward, pulse width, and side scatter characteristics as well as the fluorescence intensity of anti-CD16 R-PE. Aliquots of the Ficoll-enriched, CD16⁺-sorted cells were reanalyzed on a FACscan using the CellQuest program (Becton Dickinson, San Jose, CA) and were routinely >99.9% pure. Flow cytometric and FACS analysis were performed at the "Flow Cytometry Facility" (Princess Margaret Hospital, Toronto, Ontario, Canada). Cell viability was determined by trypan blue exclusion and was always more than 90%. Sorted cells were placed on ice and immediately processed for RNA isolation.

RNA Extraction and Reverse Transcription-- Ficoll-enriched or FACS-sorted CD16⁺ cells (1 × 10⁶) were directly lysed in 1 ml of TRIzol Reagent for 5 min at room temperature. After addition of 0.2 ml of CHCl₃ and vigorous shaking, tubes were centrifuged at 12,000 × g for 15 min at 4 °C. The aqueous phase was aspirated and supplemented with glycogen (10 µg/ml), and total RNA was precipitated by mixing with 0.5 ml of isopropyl alcohol. After 10 min at room temperature, samples were centrifuged at 12,000 × g for 10 min at 4 °C. RNA pellets were washed with 75% ethanol, briefly air-dried, dissolved in 55 µl of RNase-free water, and incubated with 5 µl of DNase I RNase-free (10 units/ml) at 37 °C for 30 min. DNase I was then inactivated by heat treatment for 5 min at 70 °C. The absence of genomic DNA in the RNA preparations was confirmed by performing PCR analysis for "minus-RT controls" (using RNA that was not reverse-transcribed as the template).

cDNA mixtures were prepared in a 20-µl reaction using a first strand cDNA synthesis kit (MBI Fermentas, Burlington, ON) according to the manufacturer's instructions. Briefly, 5 µl of the DNA-free RNA preparations from Ficoll-enriched or FACS-sorted CD16⁺ human neutrophils or 2 µg of DNA-free RNA extracted from myocardial tissue (Ambion, Austin, TX) was reverse-transcribed using 40 units of Moloney murine leukemia virus-reverse transcriptase in the presence of 0.5 µg of oligo(dT₁₈) primers, 50 mmol/liter Tris-HCl, pH 8.3, 50 mmol/liter KCl, 4 mmol/liter MgCl₂, 10 mmol/liter dithiothreitol, deoxynucleotide (dNTP) mix (1 mmol/liter each), and 20 units of RNase inhibitor. The reaction mixture was incubated for 60 min at 39 °C (transcription) and 10 min at 70 °C (inactivation of RT). The cDNA mixture was then diluted with RNase-free water to a final volume of 100 µl.

PCR-- A total of 5 µl of the diluted first strand cDNA was amplified in a PCR that included the cDNA from 4166 cells in the case of Ficoll-enriched or FACS-sorted CD16⁺ neutrophils or 100 ng of total RNA from human myocardium. PCR was performed in 50-µl reactions containing 20 mmol/liter Tris-HCl, pH 8.4, 50 mmol/liter KCl, 1.5 mmol/liter MgCl₂, dNTP mix (0.2 mmol/liter each), 2 units of Taq polymerase, and 0.5 µmol/liter of the specific primers. After 4 min at 95 °C, 40 cycles of amplification with a PCR processor (PTC-100 Thermal Cycler, MJ Research, Waltham, MA) was carried out as follows: 30 s at 95 °C, 45 s at 60 °C, and 50 s at 72 °C, followed by 10 min at 72 °C to ensure a complete extension of the amplified DNA. Hot start PCR was employed to increase the specificity of the amplification.

PCR Primers-- Primers were selected that showed insignificant homology to other genes in the EMBL DNA sequence data base. When gene sequence data were available, primer pairs were chosen to span introns in their genomic sequences, thus ensuring mRNA-specific amplification. In addition, primers were selected to have a G + C content between 45 and 65%, a size between 18- and 23-mer, and to exclude primer-dimer structures. The sequences of the sPLA₂ primers and predicted molecular weight of the PCR products for GIB, GIIA, GIID, GIIF, GIII, GV, and GX sPLA₂ and for CD52, c-Fms, HLA-DR, and -actin are listed in Table I.

Negative controls were performed by omitting reverse transcriptase from cDNA synthesis or by omitting cDNA from the PCR amplifications. PCR products were analyzed by electrophoresis through 3% agarose gels and viewed under UV light after ethidium bromide staining. PCR product specificities were confirmed by DNA sequence analysis using an ABI Prism 377 DNA Sequencer (Applied Biosystems, Foster City, CA).

Generation of Recombinant sPLA₂ Proteins-- Recombinant human GIIA, GIIF, GX, and GXII sPLA₂ were produced by refolding the inclusion body protein obtained from expression in E. coli, as described previously (17, 26, 30, 54). The preparation of recombinant human GIB, GIIE, and GV sPLA₂ will be reported elsewhere.² All recombinant sPLA₂s were judged to be pure by Laemmli gel electrophoresis. Electrospray mass spectrometry analysis of recombinant human sPLA₂s was carried out on a Bruker/Hewlett-Packard Esquire LC ion trap high performance liquid chromatography/mass spectrometer. The observed mass of all sPLA₂s was within 0.5 atomic mass units of the theoretical mass, indicating that all disulfides were formed.

Generation of Anti-sPLA₂ Antisera-- Antiserum to each recombinant human sPLA₂ was prepared by Cocalico Biologicals (Reamstown, PA). Rabbits were immunized with 100 µg of antigen with Complete Freund's Adjuvant by multiple subcutaneous and intramuscular injections. On days 14 and 21, rabbits received a booster injection with 50 µg of antigen with Incomplete Freund's Adjuvant. A test bleed was collected at day 35, and a third boost with 50 µg of antigen in Incomplete Freund's adjuvant on day 49 was carried out. The second test bleed was on day 56, and after a final boost with 50 µg of antigen in Incomplete Freund's adjuvant on day 60, exanguination bleeds were obtained on day 67.

All anti-sPLA₂ antisera were tested for specificity toward the various human sPLA₂s (hGIB, hGIIA, hGIIE, hGIIF, hGV, hGX, and hGXII), as described below. Whereas each antiserum readily detected 1 ng of recombinant sPLA₂, no signal was detected when 50 ng of each of the other human sPLA₂s were analyzed (ECL detection, Amersham Biosciences). Thus, the individual antisera were highly specific for each of the human sPLA₂ enzymes. The anti-GV sPLA₂ and anti-GX sPLA₂ antisera were further purified by passage through a protein G-agarose column, as described by the manufacturer (Pierce).

Preparation of Soluble and Microsomal Fractions for Western Blot Analysis-- Cells were washed twice with phosphate-buffered saline and centrifuged. The pellet was suspended in 66 mM HEPES buffer, pH 7.5, 1 mM EDTA, 1 mM EGTA, 1 mM Na₃VO₄, 25 mM NaF, 1 mM diisopropylfluorophate, and 10 µg/ml leupeptin and aprotinin, sonicated on ice with 3 bursts (20% maximum power) of 15 s each, and centrifuged for 5 min at 14,000 × g to remove nuclei. The supernatant was then centrifuged at 150,000 × g for 20 min at 4 °C to resolve the soluble and microsomal fractions, which were resuspended in Laemmli loading buffer.

Preparation of Neutrophil Granule Fractions-- Neutrophils (1.1 × 10⁹) were resuspended in 18 ml of KRPD + 5 mM glucose, pelleted by centrifugation, resuspended in 18 ml of KRPD + 5 mM glucose (relaxation buffer), and incubated with 5 mM diisopropylfluorophate for 15 min on ice. Cells were then pelleted by centrifugation, resuspended in 13 ml of relaxation buffer, and subjected to nitrogen cavitation, as described (55). 10 ml of the post-nuclear supernatant was then put on a 3-layer Percoll gradient (generated in relaxation buffer containing 1 mM phenylmethylsulfonyl fluoride) and centrifuged (55). Samples were collected in fractions of 1 ml. Fractions 1-6 were pooled and designated as the -band, fractions 7-12 the ₁-band, fractions 13-18 the ₂-band, and fractions 19-24 the -band. Analysis for the granule markers MPO, NGAL, lactoferrin, gelatinase, human serum albumin, and HLA was carried out exactly as described (56) and showed that less than 0.5% of the MPO, lactoferrin, and gelatinase was identified in the soluble fraction (not shown), thereby confirming that nitrogen cavitation left the granules largely intact (55).

Western Blot Analysis-- Cell lysates, extracellular fluid, granule fractions, and soluble or microsomal fractions were analyzed by SDS-PAGE using 16.5% gels at a constant 100 V in 100 mM Tris, 100 mM Tricine, and 0.1% SDS as running buffer. Proteins were transferred to a PVDF membrane (PerkinElmer Life Sciences) in 25 mM Tris-HCl, 192 mM glycine, 20% methanol (0.1% SDS for GIIA sPLA₂, 0.01% SDS for GV sPLA₂ and GX sPLA₂, and no SDS for the other sPLA₂ enzymes), pH 8.3 ±0.1 at 100 V for 1 h, followed by overnight blocking in 5% milk and 1% goat serum. The blots were then incubated with primary antibody (see legends under figures for specific antibodies) for 2 h at room temperature, washed, and incubated with horseradish peroxidase-conjugated secondary antibody for 1 h. Detection of immunoreactive bands was carried out using enhanced chemiluminescence.

Measurement of PLA₂ Activity-- Neutrophils (2 × 10⁷) were resuspended in KRPD with 1 mM CaCl₂ and 0.25% BSA, warmed to 37 °C for 10 min, and treated with 5 µM cytochalasin B for 2 min and either 0.1% Me₂SO or 1 µM fMLP for 10 min. Stimulations were terminated by centrifugation at 14,000 × g for 1 min. The PLA₂ activity in 50 µl of extracellular fluid was determined by measuring the amount of free [³H]oleic acid released from [³H]oleic acid-labeled autoclaved E. coli membranes, according to the protocol developed by Elsbach (57). The reaction was carried out in a total volume of 1.5 ml of 0.1 M Tris buffer, pH 7.5, containing 7 mM CaCl₂, 10 mg of fatty acid-free BSA, and 2.8 × 10⁸ radiolabeled, autoclaved E. coli (corresponding to 5.6 nmol of phospholipid). After a 30-min incubation at 37 °C, the reaction was terminated by filtration through a 0.45-µm Millipore filter, and the released [³H]oleic acid bound to the BSA carrier was measured by liquid scintillation counting (58). All cpm measurements were corrected for nonenzymatic hydrolysis. One unit of PLA₂ activity is defined as the amount of enzyme that hydrolyzes 56 pmol of phospholipid substrate in 30 min at 37 °C, which corresponds to 1% of the total E. coli substrate.

LTB₄ Production-- In some experiments, neutrophils were preincubated with increasing concentrations of indoxam or pyrridoline-1 for 10 min at 37 °C, as indicated in the figure legends. Following exposure to vehicle, cytochalasin B and fMLP, or OZ (5 mg/ml), cells were centrifuged at 1000 × g for 5 min at 4 °C, and LTB₄ release was subsequently determined by enzyme-linked immunosorbent assay, as described by the manufacturer (Cayman).

Radiolabeling of Live E. coli-- An inoculum of E. coli pldA (lacking the principal envelope phospholipase) was diluted 1:10 in fresh LB medium and grown for 2 h at 37 °C to mid-log phase in a shaking water bath. Bacteria were harvested at 3000 rpm for 10 min, resuspended in 0.2% lactalbumin medium supplemented with 3 µCi/ml [9,10-³H]oleic acid (PerkinElmer Life Sciences catalog number NFT 289, 5 Ci/mmol) complexed with 0.02% bovine serum albumin (fatty acid-free), and incubated for 2 h while being shaken. After harvesting, bacteria were suspended in fresh 0.2% lactalbumin medium containing 1% bovine serum albumin and incubated for 30 min at 37 °C. After centrifugation, cells were washed three times in 150 mM NaCl containing 1% BSA to remove unincorporated [9,10-³H]oleic acid. The labeled bacteria were then resuspended in 150 mM NaCl, adjusted to a concentration of 1 × 10⁹/ml by measuring the absorbance at 550 nm, and kept on ice until ready for use.

Measurement of Bacterial Phospholipid Degradation in Live Radiolabeled E. coli by Recombinant GIIA, GV, and GX sPLA₂-- Two ml of serum from the blood of healthy donors was pooled and passed over a 1-ml heparin column (Amersham Biosciences), resulting in the removal of detectable PLA₂ activity (data not shown). For the assay, [9,10-³H]oleic acid-labeled live E. coli (2.5 × 10⁷) were resuspended in 250 µl of 40% (v/v) Hanks' balanced salts solution, 1.5% (w/v) BSA, 120 mM HEPES, pH 7.4, and 2% heparin column-purified normal human serum. After a 15-min preincubation at 37 °C, human recombinant GIIA, V, or X sPLA₂ was added to a final concentration ranging from 0 to 500 ng/ml and incubated for 60 min at 37 °C. The reaction was terminated by adding 250 µl of ice-cold 150 mM NaCl and centrifugation at 14,000 × g for 4 min. Radioactivity in the recovered supernatants was measured by liquid scintillation counting, and results were corrected for the nonspecific hydrolysis of labeled E. coli.

Statistical Analysis-- Results are expressed as the mean ± S.D. of triplicate determinations. Comparisons between groups were made by repeated measures analysis of variance, followed by post-hoc analysis with paired t tests, where indicated. A p value < 0.05 was considered significant. When multiple comparisons were made, a Bonferroni correction factor was applied.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Identification of GV and GX sPLA₂ mRNA in Human Neutrophils-- To determine which of the sPLA₂ mRNA species were expressed by neutrophils, sets of primers for GIB, GIIA, GIID, GIIF, GIII, GV, and GX sPLA₂ were designed (Table I). The ability of each of these primer sets to amplify their respective sPLA₂ mRNA by RT-PCR was confirmed with mRNA that had been isolated from human myocardium, as shown in Fig. 1A. Multiple primer sets were also generated for GIIE sPLA₂. However, we were unable to detect GIIE sPLA₂ mRNA in myocardium. Hanasaki and co-workers (19) reported that GIIE sPLA₂ mRNA was detected only after multiple rounds of PCR amplification, indicating that this protein is expressed at very low levels, and we have not been able to detect GIIE sPLA₂ using a variety of human tissue cDNAs (Invitrogen multiple tissue panel). RT-PCR analysis of neutrophils that had been prepared by centrifugation on a discontinuous Ficoll-Paque density gradient identified GIIA, GIID, GV, and GX sPLA₂ mRNA species (Fig. 1B) but failed to identify GIB, GIIF, or GIII sPLA₂ mRNA.

View larger version (59K): [in this window] [in a new window]   Fig. 1.   RT-PCR analysis of sPLA2 mRNA species expressed in neutrophils isolated by Ficoll density gradient centrifugation. A, DNA-free RNA extracted from human cardiac myocytes was reverse-transcribed and subjected to PCR analysis with primer sets designed to amplify specific regions of GIB, GIIA, GIID, GIIF, GIII, GV, and GX sPLA2 cDNA. For each sPLA2 enzyme a PCR product corresponding to the predicted size of the segment of cDNA to be amplified was detected. For GV sPLA2, a second PCR product with a molecular weight of ~250 bp was also detected. B, neutrophils were isolated by dextran sedimentation and Ficoll density gradient centrifugation. Following RNA extraction, RT-PCR analysis was carried out with each set of sPLA2 primers, as described under "Materials and Methods." The blot shows that GIIA, GIID, GV, and GX sPLA2 mRNA were detected in this population of neutrophils. RT, negative control in which PCR was carried out without RT; act, actin; mw, molecular weight standards. Blots are representative of four studies.

Caution must be exercised in the interpretation of these results, as neutrophils isolated on a Ficoll-Paque density gradient are contaminated by macrophages, eosinophils, and lymphocytes (Fig. 2A), cells that could be a source of sPLA₂ mRNA (60, 61). Indeed, mRNA from the Ficoll-Paque neutrophil preparation contained c-fms, a monocyte-specific antigen (62), HLA-DR, which is expressed by lymphoid cells, and CD52, which is expressed by monocytes, eosinophils, and lymphoid cells but not by neutrophils (Fig. 2B) (63, 64). Neutrophils express exceptionally high levels of CD16 on their surface in comparison with other leukocytes (65) and do not express CD3 or CD19. To generate a preparation of highly purified neutrophils, cells isolated on a Ficoll-Paque density gradient were simultaneously incubated with anti-CD3, anti-CD19, and anti-HLA-DR antibodies conjugated to FITC, as well as a PE-conjugated anti-CD16 antibody. Dual label fluorescence-activated cell sorting (FACS) was then carried out to separate CD16 positive cells from cells that express CD3, CD19, or HLA-DR (66). With this protocol, a homogeneous population of cells (>99.9% neutrophils, Fig. 2C) was generated. Following extraction of the RNA from this purified population of neutrophils, RT-PCR failed to identify c-fms, HLA-DR, or CD52 mRNA, thereby demonstrating that the neutrophil population generated by cell sorting was free of contaminating macrophages, eosinophils, and lymphocytes (Fig. 2D). When RNA from neutrophils that had been purified by cell sorting was subjected to RT-PCR analysis, GV and GX sPLA₂ mRNA species were identified, whereas GIIA and GIID mRNA species were not detected (Fig. 2E).

View larger version (48K): [in this window] [in a new window]   Fig. 2.   Identification of GV and GX sPLA2 mRNA in neutrophils purified by four-label FACS analysis. A, neutrophils were isolated by dextran sedimentation and Ficoll density gradient centrifugation. Cells were then incubated with a PE-conjugated anti-CD16 antibody and FITC-conjugated anti-CD3, anti-CD19, and anti-HLA-DR antibodies and analyzed by dual channel fluorescence. The scattergram shows that a readily detectable fraction of the cells in the preparation expressed CD3, CD19, and/or HLA-DR on their cell surface and that some cells had very low levels of CD16 surface expression. B, RNA extracted from neutrophils was treated with DNase I, reverse-transcribed, and analyzed by PCR with primer sets designed to amplify specific regions of c-fms, HLA-DR, and CD52 as described under "Materials and Methods." The detection of c-Fms, HLA-DR, and CD52 mRNA confirmed that this cell preparation contained macrophages, eosinophils, and/or lymphocytes. C, neutrophils were labeled as in A and subjected to FACS analysis in which PE-labeled cells (i.e. CD16-positive cells) were separated from FITC-labeled cells (i.e. CD3-, CD19-, or HLA-DR-positive cells). In addition, only cells with the forward and side light scattering characteristics of neutrophils were retained. With this protocol, a population of cells that was >99.9% neutrophils was generated. D, RNA was extracted from FACS-purified neutrophils, treated with DNase I, reverse-transcribed, and analyzed by PCR. No c-fms, HLA-DR, and CD52 mRNA was detected in this cell preparation. E, RNA was isolated from neutrophils purified by FACS, treated with DNase I, and reverse-transcribed. PCR analysis identified GV and GX sPLA2 mRNA but failed to identify GIIA or GIID sPLA2 mRNA. Blots are representative of three studies. act, actin; mw, molecular weight standards.

Sequence Analysis of the GV and GX sPLA₂ RT-PCR Products-- RT-PCR analysis with a primer set designed to amplify GV sPLA₂ mRNA yielded two distinct products when human myocardium (Fig. 1A) or FACS-purified human neutrophils (Fig. 2E) were used as the source of RNA. Following amplification of the putative GV sPLA₂ PCR products, nucleotide sequence analysis demonstrated that the 358-bp PCR product corresponds to nucleotides 24-381 of human GV sPLA₂ mRNA. Sequence analysis of the 251-bp PCR product also confirmed a 100% match with GV sPLA₂ mRNA from nucleotides 24-381, in which nucleotides 186-292, which correspond to exon 4 of the hGV sPLA₂ gene, have been deleted. Nucleotide sequence analysis also demonstrated that the 370-bp PCR product generated with the primer set for GX sPLA₂ mRNA corresponded to nucleotides 569-938 of hGX sPLA₂ mRNA. No sequence data were obtained from the lower molecular weight product that was generated with the primer set for GX sPLA₂ mRNA. These results provide direct evidence that circulating neutrophils express GV and GX sPLA₂ mRNA species.

Identification of GV and GX sPLA₂ Enzymes in Circulating Human Neutrophils-- To determine which sPLA₂ enzymes are present in circulating human neutrophils, Western blot analysis was carried out for GIB, IIA, IIE, IIF, V, X, and XII sPLA₂. The antisera used in these studies were specific for each of the sPLA₂ enzymes and did not cross-react with any other known human sPLA₂s (Fig. 3). To evaluate neutrophils for the presence of individual sPLA₂ enzymes, and to define the subcellular location of these enzymes, neutrophils were sonicated in disruption buffer, resolved into soluble and microsomal fractions, incubated with vehicle or recombinant sPLA₂ enzyme (as a positive control), and evaluated by Western blotting. The corresponding sPLA₂ enzymes were also run alone, in separate lanes, as additional positive controls. Therefore, we were able to determine whether proteins identified by individual anti-sPLA₂ antisera migrated at the same apparent molecular weight as recombinant sPLA₂ proteins following SDS-PAGE. With this approach, we identified proteins that were detected with the anti-GV sPLA₂ and anti-GX sPLA₂ antisera, in whole cell lysates and in the soluble fraction of neutrophils, that co-migrated with the respective recombinant enzymes (Fig. 4, E and F). When the soluble fraction of neutrophils was spiked with recombinant GV sPLA₂ and evaluated by Western blotting, a single band was identified that co-migrated with recombinant GV sPLA₂ (Fig. 4E). Similarly, when the soluble fraction of neutrophils was spiked with recombinant GX sPLA₂, Western blotting studies identified a single band that co-migrated with recombinant GX sPLA₂ (Fig. 4F). In contrast, proteins that co-migrated with recombinant GIB, GIIA, GIIE, GIIF, or GXII sPLA₂ were not detected by Western blot analyses using highly specific antisera (Fig. 4). For each of these later sPLA₂s, we confirmed that the addition of the respective recombinant sPLA₂ protein to the soluble or microsomal fraction of neutrophils resulted in the appearance of a band that co-migrated with the recombinant sPLA₂ protein. These positive controls indicate that GIB, GIIA, GIIE, GIIF, or GXII sPLA₂ are not being degraded in neutrophil lysates. Furthermore, the detection limit for sPLA₂ proteins using these antisera in Western blots was in the range of 0.05-0.5 ng with enhanced chemiluminescence detection (data not shown). Taken together, these results provide direct evidence that only GV and GX sPLA₂ mRNAs and proteins are present in circulating human neutrophils.

View larger version (75K): [in this window] [in a new window]   Fig. 3.   Specificity of individual anti-sPLA2 antisera. Recombinant GIB, GIIA, GIIE, GIIF, GV, GX, or GXII sPLA2 (1 ng each) were applied to a PVDF membrane, washed, and probed with anti-GIB, GIIA, GIIE, GIIF, GV, GX, or GXII sPLA2 antisera, as described under "Materials and Methods." The blot demonstrates that each of the anti-sPLA2 antisera identified the respective sPLA2 protein and did not cross-react with other sPLA2 proteins. The blot is representative of three studies.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Neutrophils contain GV and GX sPLA2. Neutrophils (106) were disrupted by sonication and resolved into soluble (S) and membrane fractions (MF). The soluble or microsomal fractions were then incubated in the presence (+) or absence () of 1 ng of recombinant GIB or GIIA sPLA2 or 1.5 ng of recombinant GIIE, GIIF, GV, GX, or GXII sPLA2 on ice for 10 min. Intact neutrophils (PMN) and the soluble (50 µg) or microsomal fraction (8 µg) of these cells, plus or minus the spiked recombinant sPLA2 proteins, were then resolved by SDS-PAGE, transferred to PVDF membrane, and probed with the following: A, anti-GIB; B, anti-GIIA; C, anti-GIIE; D, anti-GIIF; E, anti-GV; F, anti-GX; or G, anti-GXII sPLA2 antisera as described under "Materials and Methods." For each blot, the recombinant sPLA2 enzyme was run alone as a control. Representative results for 4-8 separate experiments are shown.

Distribution of GV and GX sPLA₂ in the Granules of Circulating Human Neutrophils-- To determine whether GV and GX sPLA₂ were present in neutrophil granules, cells were disrupted by nitrogen cavitation and resolved on a 3-layer Percoll density gradient into fractions that contain azurophil, specific and gelatinase granules, and secretory vesicles (55). Analysis of granule marker enzyme content in the , ₁, ₂, and  fractions demonstrated that this protocol resulted in the separation of the azurophil, specific and gelatinase granules, and secretory vesicles (Table II). As shown in Fig. 5A, we identified a protein in the , ₁, and ₂ fractions that co-migrated with recombinant GV sPLA₂. Slightly less GV sPLA₂ was detected in the ₁ fraction than in the  fraction, with minimal GV sPLA₂ in the ₂ fraction. No proteins that co-migrated with recombinant GV sPLA₂ were identified in the  fraction, which contains both secretory vesicles and plasma membranes (67). Similarly, we identified a protein in the , ₁, and ₂ fractions that co-migrated with recombinant GX sPLA₂ (Fig. 5B). The majority of the GX sPLA₂ was identified in the  fraction, with comparatively little GX sPLA₂ in the ₁ or ₂ fractions. No proteins that co-migrated with recombinant GX sPLA₂ were identified in the  fraction. The distribution of GX sPLA₂ follows the distribution of the azurophil granule marker, MPO (Table II). In contrast, the distribution of GV sPLA₂ follows the distribution of the azurophil granule marker MPO and the specific granule markers lactoferrin and NGAL. This indicates that GX sPLA₂ is localized to azurophil granules solely, whereas GV sPLA₂ is localized to both azurophil and specific granules.

View larger version (32K): [in this window] [in a new window]   Fig. 5.   Distribution of GV and GX sPLA2 in neutrophil granules. Cells were disrupted by N2 cavitation. The post-nuclear supernatant was loaded on a three-layer Percoll gradient, centrifuged, and collected in 1-ml aliquots. The , 1, 2, and  fractions were formed by pooling fractions 1-6, 7-12, 13-18, and 19-24, respectively, as described under "Materials and Methods." Aliquots of individual granule fractions (corresponding to 106 neutrophils) were then incubated with vehicle or recombinant GV or GX sPLA2, as indicated, resolved by SDS-PAGE, transferred to PVDF membrane, and probed with anti-GV sPLA2 (A) or anti-GX sPLA2 (B) antisera. For each blot, the recombinant sPLA2 enzyme was run alone as a control. Representative results for granule preparations from two different neutrophil donors are shown.

Neutrophils Release GV sPLA₂ following Exposure to fMLP or OZ-- sPLA₂ enzymes may be released by cells into their extracellular environment (37). Thus, when neutrophils were exposed to the bacterial tripeptide fMLP, sPLA₂ activity (as measured with radiolabeled bacterial membranes) in the extracellular fluid increased by 3-4-fold (Fig. 6A), a finding consistent with previous results (44). To determine whether the increase in extracellular PLA₂ activity was associated with the release of GV or GX sPLA₂, cells were treated with fMLP, and the extracellular fluid was evaluated by Western blot analysis. As shown in Fig. 6B, extracellular GV sPLA₂ was detected that co-migrated with recombinant GV sPLA₂, and addition of recombinant GV sPLA₂ to the supernatant of fMLP-stimulated cells resulted in the appearance of a single, more intense band on Western blots. A protein with an apparent molecular mass of ~18-19 kDa that cross-reacted with the anti-GX sPLA₂ antiserum was also identified in the supernatant of vehicle (Me₂SO) and fMLP-stimulated cells, but this protein did not co-migrate with recombinant GX sPLA₂ (mature form, lacking an N-terminal extension) (Fig. 6C). Similarly, exposure to OZ resulted in the release of GV sPLA₂ by neutrophils but did not result in the release of any proteins that were identified by the anti-GX sPLA₂ antiserum (Fig. 6, D and E). No sPLA₂ enzymatic activity was detected in the extracellular fluid of OZ-stimulated cells (not shown). The failure to detect sPLA₂ activity in the extracellular fluid of OZ-stimulated cells appears to be due to inhibition of sPLA₂ enzymatic activity by OZ, as co-incubation of recombinant GV sPLA₂ or the supernatant of fMLP-stimulated cells with OZ decreased sPLA₂ enzymatic activity by >95% (not shown). These results demonstrate that circulating neutrophils release GV sPLA₂ but release little or no mature GX sPLA₂ into their extracellular environment following exposure to fMLP or OZ (<0.2 ng per 10⁶ neutrophils).

View larger version (36K): [in this window] [in a new window]   Fig. 6.   fMLP and OZ both stimulate the release of GV sPLA2 from neutrophils. A, 1 ml of cells (2 × 107/ml) was resuspended in KRPD with 1 mM CaCl2 and 0.25% BSA, treated with cytochalasin B and Me2SO (open bars) or fMLP (filled bars), and centrifuged, and the supernatant was assayed for PLA2 activity as described under "Materials and Methods." Results are the mean ± S.D. of four experiments. B, following an identical experimental protocol, 25 µl of the supernatant of Me2SO- or fMLP-treated cells (corresponding to 5 × 105 cells) was incubated in the presence (rV) or absence (control) of 1.5 ng of recombinant GV sPLA2 on ice for 10 min, resolved by SDS-PAGE, transferred to PVDF membrane, and probed with the anti-GV sPLA2 antisera. C, the supernatant of Me2SO- or fMLP-treated cells was incubated in the presence (rX) or absence (control) of 1 ng of recombinant GX sPLA2 on ice for 10 min, resolved by SDS-PAGE, transferred to PVDF membrane, and probed with the anti-GX sPLA2 antisera as described under "Materials and Methods." D, cells (2 × 107) were resuspended in KRPD with 1 mM CaCl2 and 0.25% BSA and treated with vehicle or OZ for 10 min. The supernatant (25 µl, 5 × 105 cells) was then incubated in the presence (rV) or absence (control) of 0.5 ng of recombinant GV sPLA2 on ice for 10 min, resolved by SDS-PAGE, transferred to PVDF membrane, and probed with the anti-GV sPLA2 antisera. E, the supernatant of vehicle or OZ treated cells was incubated in the presence (rX) or absence (control) of 1 ng of recombinant GX sPLA2 on ice for 10 min, resolved by SDS-PAGE, transferred to PVDF membrane, and probed with the anti-GX sPLA2 antisera. Results are representative of eight separate experiments.

Indoxam Exhibits Selectivity for the Inhibition of Recombinant GIIA and GV sPLA₂-- LY311727 and indoxam are potent and active site-directed inhibitors of GIIA sPLA₂ (50, 68). Prior to evaluating the ability of LY311727 or indoxam to inhibit the PLA₂ activity released by fMLP-stimulated cells, we evaluated the ability of these compounds to inhibit recombinant GV and recombinant GX sPLA₂. The effect of LY311727 or indoxam on recombinant GIIA sPLA₂ activity was also assessed for comparison. Assays were carried out under conditions where sPLA₂ activity was linear with respect to the amount of enzyme used (Fig. 7A) and time (Fig. 7B). Under these conditions, the specific activity of recombinant GIIA sPLA₂ (dotted line) was ~4-fold higher than recombinant GV sPLA₂ (dashed line) and 14-fold higher than recombinant GX sPLA₂ (solid line). As shown in Fig. 7C, co-incubation with LY311727 inhibited recombinant GIIA sPLA₂ activity (IC₅₀ 50 nM) more potently than recombinant GV sPLA₂ activity (IC₅₀ 2 µM) or recombinant GX sPLA₂ activity (IC₅₀ 0.75 µM). In contrast, co-incubation with indoxam (Fig. 7D) inhibited recombinant GIIA sPLA₂ activity (IC₅₀ 10 nM) and recombinant GV sPLA₂ activity (IC₅₀ 40 nM) to a similar degree but had a much less potent effect on recombinant GX sPLA₂ activity (IC₅₀ 5 µM). Therefore, indoxam exhibited a 125-fold selectivity for inhibition of recombinant GV sPLA₂ in comparison with recombinant GX sPLA₂ in vitro. The IC₅₀ value for inhibition of human GX sPLA₂ by indoxam is similar to that measured for mouse GX sPLA₂ (28).

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Effect of LY311727 or indoxam on the catalytic activity of recombinant GIIA, GV, and GX sPLA2. A, increasing quantities of recombinant GIIA, GV, or GX sPLA2 were incubated with [3H]oleate-labeled E. coli membranes for 30 min, and PLA2 activity was measured as described under "Materials and Methods." B, 0.25 ng of GIIA sPLA2, 2.5 ng of GV sPLA2, or 5 ng of GX sPLA2 were incubated with [3H]oleate-labeled E. coli membranes for 15, 30, or 45 min, and PLA2 activity was measured as described under "Materials and Methods." The mean specific activity for recombinant GIIA, GV, and GX sPLA2 was 45.6 ± 3.8, 10.8 ± 0.3, and 3.2 ± 0.5 units/ng, respectively. 0.25 ng of GIIA sPLA2, 2.5 ng of GV sPLA2, or 5 ng of GX sPLA2 was incubated with increasing concentrations of LY311727 (C) or indoxam (D) for 10 min. Following addition of [3H]oleate-labeled E. coli membranes for 30 min, PLA2 activity was measured. All results are the mean ± S.D. of at least four separate experiments.

Co-incubation with Indoxam Inhibits Extracellular sPLA₂ Activity but Does Not Attenuate LTB₄ Production by Neutrophils Treated with fMLP-- Following exposure of neutrophils (2 × 10⁷) to 1 µM fMLP for 10 min, PLA₂ activity in the extracellular medium was 202 ± 4 units/ml (Fig. 8A). Co-incubation of the extracellular fluid from fMLP-treated neutrophils with indoxam decreased PLA₂ activity in a dose-dependent manner (Fig. 8A). Because fMLP-stimulated neutrophils released a protein that cross-reacted with the anti-GV sPLA₂ antiserum (cf. Fig. 6B), these results are consistent with the notion that circulating neutrophils release catalytically active GV sPLA₂ following exposure to fMLP. The concentration of indoxam required to inhibit the major portion of PLA₂ activity secreted from neutrophils (IC₅₀ ~50 nM) is similar to the IC₅₀ of indoxam measured with recombinant GV sPLA₂ using the assay with radiolabeled E. coli membranes (Fig. 7D). The difference between the PLA₂ activity released by untreated neutrophils (40 ± 3 units/ml) and the PLA₂ activity in the supernatant of fMLP-treated neutrophils following incubation with 10 µM indoxam (45 ± 3 units/ml) was negligible. This indicates that indoxam inhibited essentially all of the PLA₂ activity that was released from neutrophils in response to fMLP. The effect of indoxam on OZ-induced PLA₂ release by neutrophils was not studied, because OZ inhibited the activity of the PLA₂ released by these cells. Because the concentrations of indoxam used in these studies do not inhibit GX sPLA₂, these results are consistent with the Western blotting data indicating that fMLP-stimulated neutrophils do not release catalytically active GX sPLA₂.

View larger version (17K): [in this window] [in a new window]   Fig. 8.   Effect of inhibition of extracellular GV sPLA2 activity on LTB4 production by fMLP-stimulated neutrophils. A, cells (2 × 107/ml) were resuspended in KRPD with 1 mM CaCl2 and 0.25% BSA, treated sequentially with cytochalasin B and fMLP, and centrifuged as described under "Materials and Methods." The supernatant from these cells was then incubated with increasing concentrations of indoxam (0-10 µM), and PLA2 activity was measured. B, cells (20 × 106/ml) were resuspended in KRPD with 1 mM CaCl2 and 0.25% BSA, incubated with increasing concentrations of indoxam (0-10 µM), and treated sequentially with cytochalasin B and Me2SO (open bars) or fMLP (filled bars). Following centrifugation, the supernatant was assayed for LTB4 as described under "Materials and Methods." Results are the mean ± S.D. of four experiments.

To evaluate the role of extracellular GV sPLA₂ in neutrophil LTB₄ biosynthesis, cells were pretreated with indoxam and stimulated with fMLP. As shown in Fig. 8B, inhibition of extracellular GV sPLA₂ activity with indoxam did not inhibit fMLP-stimulated LTB₄ biosynthesis by human neutrophils. Therefore, GV sPLA₂ released from neutrophils in response to fMLP does not participate in neutrophil LTB₄ biosynthesis. In contrast, preincubating neutrophils with the selective cPLA₂ inhibitor, pyrrolidine-1 (51, 52), resulted in a dose-dependent inhibition of both fMLP- and OZ-induced neutrophils LTB₄ biosynthesis (IC₅₀ 0.1-0.5 µM, Fig. 9, A and B). These findings indicate that cPLA₂ participates in fMLP- and OZ-induced LTB₄ synthesis by neutrophils. Preincubation with pyrrolidine-1 had no effect on the fMLP-induced release of GV sPLA₂, as assessed by Western blotting, or on PLA₂ activity released by fMLP-stimulated cells (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 9.   Effect of inhibition of cPLA2 on LTB4 production by fMLP- or OZ-stimulated neutrophils. Cells (2 × 107/ml) were resuspended in KRPD with 1 mM CaCl2 and 0.25% BSA, preincubated with increasing concentrations of pyrrolidine-1 for 10 min, and treated with cytochalasin B and Me2SO (open bars) or fMLP (filled bars) (A) or vehicle (open bars) or OZ (filled bars) (B). Following centrifugation, the supernatant was assayed for LTB4 as described under "Materials and Methods." Results are the mean ± S.D. of three independent experiments.

Recombinant GIIA and GV sPLA₂, but Not Recombinant GX sPLA₂, Catalyze the Hydrolysis of Phospholipids in Live E. coli-- A clear link between E. coli envelope phospholipid degradation and overall bacterial destruction by rabbit neutrophils has been established (69). We have demonstrated that human neutrophils contain GV sPLA₂ and release this enzyme upon exposure to fMLP or OZ, so we evaluated the ability of recombinant GV sPLA₂ to hydrolyze bacterial phospholipids. Studies with GIIA and GX sPLA₂s were also carried out for comparison. Live E. coli were labeled with [³H]oleic acid during the log growth phase, washed extensively in buffer with 1% BSA, and co-incubated with increasing concentrations of recombinant GIIA, GV, or GX sPLA₂. For these studies, human serum that had been eluted from a heparin column (to remove all PLA₂ activity) was added at a final concentration of 2% serum. Co-incubation of serum purified over the heparin column with live [³H]oleic acid-labeled E. coli did not result in measurable phospholipid hydrolysis (data not shown). Co-incubation of live [³H]oleic acid-labeled E. coli with 10 ng of recombinant GIIA sPLA₂ for 60 min resulted in the hydrolysis of ~25% of the labeled bacterial phospholipids, and this value did not change significantly as the amount of recombinant GIIA sPLA₂ was increased to 500 ng (Fig. 10). Co-incubation of live [³H]oleic acid-labeled E. coli with recombinant GV sPLA₂ resulted in a dose-dependent increase in bacterial phospholipid hydrolysis that reached a plateau at 100 ng of recombinant GV sPLA₂, whereas co-incubation of up to 500 ng of recombinant GX sPLA₂ with live [³H]oleic acid-labeled E. coli resulted in virtually no hydrolysis (Fig. 10). When recombinant GIIA sPLA₂ or recombinant GV sPLA₂ was co-incubated with live [³H]oleic acid-labeled E. coli in the absence of serum, no bacterial phospholipid hydrolysis was detected (data not shown). These results provide direct evidence that recombinant GIIA and recombinant GV sPLA₂ can hydrolyze phospholipids present in the outer membrane of live E. coli and that this phospholipid hydrolysis is dependent on the presence of serum.

View larger version (14K): [in this window] [in a new window]   Fig. 10.   Recombinant GIIa and GV sPLA2, but not recombinant GX sPLA2, hydrolyze Gram-negative bacterial phospholipids. E. coli was metabolically labeled with [3H]oleic acid, washed extensively with 150 mM NaCl and 1% BSA, and co-incubated with 0-500 ng of recombinant GIIA (open bars), GV (filled bars), or GX sPLA2 (dashed bars). Bacterial phospholipid hydrolysis was estimated by measuring the release of [3H]oleic acid into the supernatant as described under "Materials and Methods." Results are the mean ± S.D. of four experiments.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

GIIA sPLA₂ has been purified from rabbit inflammatory exudates (2, 71), and immunohistochemical studies identified GIIA sPLA₂ in human neutrophils (37). These seminal findings led to the notion that human neutrophils express and release GIIA sPLA₂ during an inflammatory response (72). However, recent studies (4) indicate that multiple sPLA₂ enzymes, in addition to GIIA sPLA₂, may play important roles in the generation of an inflammatory response. Thus, some myeloid cells express more than one sPLA₂ enzyme (73), and sPLA₂ enzymes other than GIIA sPLA₂, including GV and GX sPLA₂, have been shown to participate in the generation of pro-inflammatory mediators (5, 24). Furthermore, nine human sPLA₂ enzymes have now been described (5). The existence of multiple sPLA₂ enzymes raises the possibility that the anti-GIIA antisera used in earlier studies (37) could have cross-reacted with other sPLA₂ enzymes. Based on these considerations, we decided to evaluate systematically which sPLA₂ mRNA species and sPLA₂ proteins exist in circulating human neutrophils and to define the subcellular localization of these enzymes. Such information is necessary as a prelude for studying the biological functions of sPLA₂ enzymes in these cells. In this study, we also focused on the identification of the enzyme(s) that are released by neutrophils that were stimulated with two distinct agonists, fMLP and OZ, and on the role that these sPLA₂ enzymes play in neutrophil LTB₄ biosynthesis and Gram-negative bacterial phospholipid hydrolysis. We also studied the role of cPLA₂ in fMLP- and OZ-stimulated LTB₄ biosynthesis by incubating cells with pyrrolidine-1 (52), because this inhibitor is more specific than previously used inhibitors of cPLA₂.

Identification of GV and GX sPLA₂ mRNAs and Proteins in Neutrophils-- Extensive experimental evidence has documented the existence of sPLA₂ enzyme(s) in human neutrophils. Thus, neutrophils contain an acid-stable enzyme with PLA₂ activity (74); sPLA₂ has been localized to the granules of resting human neutrophils and shown to translocate to phagolysosomes following exposure to agonists (37), and neutrophils stimulated with fMLP release sPLA₂ activity (44). In the present study we have convincingly shown by the combined use of RT-PCR and Western blot analysis with highly specific anti-sPLA₂ antisera that circulating and non-stimulated human neutrophils contain GV and GX sPLA₂, whereas GIB, GIIA, GIID, GIIE, GIIF, GIII, and GXII sPLA₂ were undetectable in these cells. Our results underscore the importance of using highly purified neutrophils for RT-PCR analysis, as GIIA and GIID sPLA₂ mRNA species were detected in neutrophils isolated on a Ficoll-Paque density gradient but were not identified in FACS-purified cells that were free of contaminating macrophages, eosinophils, and lymphocytes. For GIID and GIII sPLA₂, only RT-PCR was used for detection, with successful execution of positive controls, because antisera for these proteins are not yet available. Although GIIE sPLA₂ mRNA could not be detected by RT-PCR (Hanasaki and co-workers (19) reported that GIIE sPLA₂ cDNA could be detected by PCR only after multiple rounds of extensive amplification), GIIE sPLA₂ protein was not detected by Western blot analysis even though spiking the neutrophil lysate with 1.5 ng of recombinant GIIE sPLA₂ gave a readily detected band. In contrast to our results, a previous study (75) reported that GV sPLA₂ is not present in human neutrophils. This discrepancy may be due to the fact that we used a high affinity antiserum that could detect <1 ng of recombinant GV sPLA₂, while the earlier study made use of a lower affinity monoclonal antibody that was only able to detect ~30 ng of this antigen (76). Despite earlier reports that circulating human neutrophils contain GIIA sPLA₂ based on immunohistochemical analysis (37), we could not detect GIIA sPLA₂ in these cells either by RT-PCR or Western blot. For the latter analyses, the GIIA sPLA₂ Western blot band was easily seen when the neutrophil lysate was spiked with 1 ng of recombinant GIIA sPLA₂, thus the amount of GIIA sPLA₂ is 1 ng per 10⁶ neutrophils. Our findings underscore the cell type specificity of sPLA₂ expression; for example, GIIA, GIIC, GIID, GIIE, GIIF, and GV sPLA₂ were detected in bone marrow-derived mast cells from BALB/cJ mice, whereas transcripts for GIB and GX sPLA₂ were not identified in these cells (73). As neutrophils express some mRNA species during myelopoiesis that are not expressed in mature cells (77), we could not exclude the possibility that other sPLA₂ mRNA species besides GV and GX sPLA₂ are expressed at an earlier phase of neutrophil maturation.

Sequence analysis demonstrated that the 358-bp GV sPLA₂ RT-PCR product was identical to the corresponding sequence in GV sPLA₂, whereas the 251-bp GV sPLA₂ product had a 107-bp deletion. Deletion of this 107-bp sequence, which corresponds to exon 4 of the human GV sPLA₂ gene, would result in a frameshift that would introduce a premature TGA stop codon. Therefore, translation of this alternatively spliced GV sPLA₂ mRNA would be predicted to give rise to a catalytically inactive, truncated GV sPLA₂ protein.

We identified proteins in neutrophils that were recognized by anti-GV and anti-GX sPLA₂ antisera and that co-migrated exactly with recombinant GV sPLA₂ (Fig. 4E) and recombinant GX sPLA₂ (Fig. 4F). These recombinant proteins were generated in E. coli and lack N-terminal signal sequences. Therefore, our findings are consistent with the notion that the N-terminal signal sequence of GV and GX sPLA₂ in neutrophils had been cleaved.

The identification of multiple sPLA₂ enzymes in neutrophils is consistent with the results of studies of other types of cells, in which multiple sPLA₂ enzymes were expressed. Thus, GIIA and GV sPLA₂ were identified in mouse bone marrow mast cells (23), cultured rat astrocytes (78), and rat neonatal cardiomyocytes,³ whereas GIB and GIIA sPLA₂ were both identified in epidermis (79) and rat splenic macrophages (80, 81). In mast cells, GIIA sPLA₂ was recognized in secretory granu