Identification of a human cDNA clone for lysosomal type Ca2+-independent phospholipase A2 and properties of the expressed protein.

A Ca2+-independent phospholipase A2 (PLA2) maximally active at pH 4 and specifically inhibited by the transition-state analogue 1-hexadecyl-3-trifluoroethylglycero-sn-2-phosphomethanol (MJ33) was isolated from rat lungs. The sequence for three internal peptides (35 amino acids) was used to identify a 1653-base pair cDNA clone (HA0683) from a human myeloblast cell line. The deduced protein sequence of 224 amino acids contained a putative motif (GXSXG) for the catalytic site of a serine hydrolase, but showed no significant homology to known phospholipases. Translation of mRNA produced from this clone in both a wheat germ system and Xenopus oocytes showed expression of PLA2 activity with properties similar to the rat lung enzyme. Apparent kinetic constants for PLA2 with dipalmitoylphosphatidylcholine as substrate were Km = 0.25 mM and Vmax = 1.89 nmol/h. Activity with alkyl ether phosphatidylcholine as substrate was decreased significantly compared with diacylphosphatidylcholine. Significant lysophospholipase, phospholipase A1, or 1-O-alkyl-2-acetyl-sn-glycero-3-phosphocholine acetylhydrolase activity was not observed. Enzyme activity was insensitive to p-bromophenacyl bromide, bromoenol lactone, trifluoromethylarachidonoyl ketone, mercaptoethanol, and ATP, but was inhibited by MJ33 and diethyl p-nitrophenyl phosphate, a serine protease inhibitor. SDS-polyacrylamide gel electrophoresis with autoradiography of the translated [35S]methionine-labeled protein confirmed a molecular mass of 25.8 kDa, in good agreement with the enzyme isolated from rat lung. By Northern blot analysis, mRNA corresponding to this clone was present in both rat lung and isolated rat granular pneumocytes. These results represent the first molecular cloning of a cDNA for the lysosomal type Ca2+-independent phospholipase A2 group of enzymes.

Phospholipase A 2 (PLA 2 ) 1 represents a diverse family of enzymes that hydrolyze the sn-2-fatty acyl or alkyl bond of phospholipids. The best characterized member of this family is a group of low molecular mass (ϳ15 kDa) enzymes that require Ca 2ϩ for catalytic activity and show maximal activity in an alkaline (pH 8.5) medium (1). This group of enzymes, called secreted PLA 2 (sPLA 2 ; types 1-3), include snake and bee venoms and mammalian pancreatic and synovial PLA 2 enzymes. The more recently characterized 87-kDa cytoplasmic PLA 2 (cPLA 2 ) requires only M Ca 2ϩ for binding to the interface and shows maximal activity at neutral pH (2). This enzyme is widely distributed in cell types and may be specifically linked to eicosanoid metabolism. Finally, there is a group of enzymes (iPLA 2 ) that do not require Ca 2ϩ for maximal activity. As recently reviewed (3), few of these enzymes have been purified, and in contrast to sPLA 2 and cPLA 2 , little molecular information is available. Therefore, although the iPLA 2 enzymes are widely distributed, relatively little is known about their intragroup relationships and specific functions. Based on biochemical characteristics, Ackermann and Dennis (3) have divided the iPLA 2 enzymes into three subgroups: lysosomal iPLA 2 , brush-border membrane iPLA 2 , and intracellular (cytosolic/ membrane) iPLA 2 . A distinguishing characteristic of the lysosomal iPLA 2 group is maximal activity in acidic medium, and accordingly, we have designated the enzyme in this report as aiPLA 2 . iPLA 2 activity in rat lung homogenates was first described ϳ25 years ago (4), and activity at acid pH was subsequently demonstrated in rabbit lung (5), rat granular pneumocytes (lung alveolar type II cells) (6), and rat and human alveolar macrophages (7,8). Activity has been localized further to the lung lamellar bodies (the surfactant secretory organelle) and the lysosomal fraction (5,7,9). Under our assay conditions, lung aiPLA 2 activity is inhibited by a phospholipid transitionstate analogue, 1-hexadecyl-3-trifluoroethylglycero-sn-2-phosphomethanol (MJ33), while Ca 2ϩ -dependent PLA 2 activity in the lung homogenate is insensitive (7). Using MJ33 as a probe, we previously purified a protein with aiPLA 2 activity to appar-ent homogeneity (10). This protein had a molecular mass of ϳ15 kDa and a unique N-terminal amino acid sequence. Subsequent attempts to reproduce the isolation yielded similar enzyme activity, although the molecular mass of the isolated protein was ϳ26 kDa. Furthermore, the amino acid sequence of the larger enzyme, as described in this report, does not contain the previously determined N-terminal sequence. Consequently, we believe that aiPLA 2 described in this report is different from the previously isolated protein.
Amino acid sequencing (35 residues) of tryptic digests of the 26-kDa protein isolated from rat lung revealed 100% identity to deduced amino acids from the nucleotide sequence (previously unpublished) of a clone isolated from a human myeloblast cDNA library (11). This report provides evidence that the protein expressed by this cDNA is aiPLA 2 and provides the first molecular cloning for this group of enzymes.

EXPERIMENTAL PROCEDURES
Materials-Lipids were obtained from Avanti Polar Lipids (Birmingham, AL). All radiochemicals and x-ray film were purchased from DuPont NEN, and bisbodipy-C 11 -PC was from Molecular Probes, Inc. (Eugene, OR). Trifluoromethylarachidonoyl ketone (AACOCF 3 ) was from Calbiochem; bromoenol lactone was from Cayman Chemical Co., Inc. (Ann Arbor, MI); and p-bromophenacyl bromide (pBPB), 2-mercaptoethanol, diethyl p-nitrophenyl phosphate (DENP), Naja naja PLA 2 , Sephacryl S-100, and phenyl-Sepharose CL-4B were from Sigma. DE52 was purchased from Whatman (Maidstone, United Kingdom). Molecular mass standards for SDS-PAGE and Transblot membrane were from Bio-Rad. Nitrocellulose membrane was from Schleicher & Schuell. Superscript reverse transcriptase and T4 DNA polymerase were from Life Technologies, Inc. Klenow enzyme was from Boehringer Mannheim. pBluescript SK ϩ vector was from Stratagene (La Jolla, CA). In vitro transcription and wheat germ in vitro translation kits were from Ambion Inc. (Austin, TX). Male Sprague-Dawley rats weighing ϳ200 g were obtained from Charles River Breeding Laboratories (Kingston, NY).
Isolation of Rat Lung aiPLA 2 -Rats were anesthetized with sodium pentobarbital (50 mg/kg intraperitoneally). Lungs were ventilated through the trachea, perfused through the pulmonary artery in situ to clear them of blood, removed from the thorax, and stored at Ϫ80°C. Frozen lung tissue (ϳ80 g) was extensively homogenized with Polytron PT-10 and Potter-Elvehjem homogenizers in 200 ml of ice-cold 50 mM Tris-HCl buffer (pH 7.4) containing 1 mM EDTA, 0.2 mM 4-(2-aminoethyl)benzenesulfonyl fluoride, and 10% glycerol. All subsequent steps of purification were performed at 4°C. Assays for PLA 2 activity in the absence and presence of MJ33 utilized the fluorescence assay with acidic (pH 4.0) Ca 2ϩ -free buffer (see below).
The lung homogenate was spun at 10,000 ϫ g to remove cell debris and then centrifuged at 100,000 ϫ g for 1 h. The supernatant was subjected to 55% ammonium sulfate fractionation. The resultant precipitate was dissolved in 80 ml of buffer and dialyzed extensively against 50 mM Tris-HCl (pH 7.4) containing 1 mM EDTA and 5% glycerol. This buffer was used in all further steps of purification. The dialyzed sample was centrifuged at 1500 ϫ g for 15 min., and the supernatant was applied to a DE52 column (38 x 2.2 cm), which was washed with 2 bed volumes of the same buffer at a flow rate of 60 ml/h and then eluted with a linear gradient of 0 -50 mM NaCl. The major aiPLA 2 activity was recovered after the second bed volume wash and was retarded with respect to the unbound protein peak (Fig. 1). The fractions containing aiPLA 2 activity that was inhibited by MJ33 were pooled, concentrated to 7 ml on an Amicon YM-10 membrane, applied to a Sephacryl S-100 HR column (120 ϫ 2.2 cm), and eluted with buffer at a flow rate of 17 ml/h (Fig. 1). Fractions with MJ33-sensitive aiPLA 2 activity (3.5 ml each) were pooled and applied to a second DE52 column (7 ϫ 1.3 cm). The column was washed with 5 bed volumes of buffer and then eluted with a linear gradient of 10 -60 mM NaCl. aiPLA 2 activity eluted between 20 and 40 mM NaCl (Fig. 1). The eluted protein fraction was applied to a 2-ml phenyl-Sepharose column equilibrated with buffer containing 1.5 M KCl. The column was eluted with a step gradient of 1.5 to 0 KCl, followed by buffer and then water. PLA 2 activity was detected in the H 2 O wash, which was analyzed by SDS-PAGE.
Amino Acid Sequencing of Rat Lung Protein-Protein samples were analyzed by SDS-PAGE (15% acrylamide) under reducing conditions using the Laemmli buffer system (12). Bands were first visualized with Coomassie Blue and then with silver staining. The protein was electro-blotted onto polyvinylidene difluoride membrane (Transblot) in order to analyze for internal amino acid sequences. Selected protein bands were digested in situ with trypsin and then separated by high pressure liquid chromatography using a Zorbax column. Ten peaks were selected and subjected to mass spectral analysis. Based on this analysis, three peptides were selected and analyzed for their amino acid sequences by Edman degradation (13). These analyses were carried out in the Protein Microchemistry Laboratory of the Wistar Institute (Philadelphia, PA). The sequences were compared with the National Center for Biotechnology Information data base for similarity to known sequences.
Isolation and Sequencing of the cDNA Clone-The human myeloid cell line KG-1 (CCL246 from the American Type Culture Collection, Rockville, MD) was used as a source of mRNA. RNA preparation, cloning, and sequencing of the HA0683 clone (referred to as the KIAA0106 gene) have been described previously (11,14). Briefly, cytoplasmic RNA was prepared using a non-ionic detergent (15), followed by chromatography on oligo(dT)-cellulose to isolate the poly(A) ϩ RNA. First-strand synthesis was primed with a poly(dT)-NotI primer oligonucleotide (5Ј-CTCTAGAGGCGGCCGC(T) 34 -3Ј) using Superscript reverse transcriptase. Second-strand synthesis was performed as described (16), repaired with T4 DNA polymerase, digested with NotI, size-fractionated on a sucrose gradient, and ligated into a NotI-EcoRVcut pBluescript SK ϩ vector. Automated sequencing was performed by the dideoxy method (15) using Chemical Robot DSP-240 (Seiko Instruments Inc., Tokyo) or CATALYST800 (Applied Biosystems, Inc., Foster City, CA), and analysis was performed using an Applied Biosystems 373A sequencer and the Applied Biosystems sequence analysis system INHERIT. Large-scale preparations of plasmid were obtained by centrifugation on cesium chloride-ethidium bromide gradients (15).
In Vitro Transcription-The cDNA clone HA0683 was linearized by digestion with NotI, and capped mRNA was then synthesized by transcription using T7 RNA polymerase and 7 mGpppG. The transcripts were purified by phenol/chloroform extraction and ethanol precipitation, quantitated by comparing intensity after ethidium bromide staining with control RNA with known concentration, checked for integrity by electrophoresis on a formaldehyde-agarose (1%) gel, and subsequently used as a template for translation in wheat germ or oocytes.
Wheat Germ in Vitro Translation-Translation of the HA0683 mRNA transcript was performed in the wheat germ system according to the manufacturer's protocol. In brief, RNA samples of up to 2 g were translated in a reaction mixture containing 50 mM potassium acetate, amino acid master mixture (0.16 M creatine phosphate, 0.5 mM methionine, 0.5 mM leucine, and a 1 mM concentration of the other amino acids), and wheat germ extract. The total reaction volume was 50 l. Each experiment included a negative control in which no RNA was added and a positive control in which XeF-1 RNA, encoding Xenopus elongation factor-1␣ (M r 50,300) provided by the supplier, was added. Reaction mixtures were incubated at 27°C for 60 min. The translated products were then assayed for aiPLA 2 activity.
Autoradiography of Translated Protein with [ 35 S]Methionine-Labeling of the HA0683 translation product in wheat germ extract was performed with the addition of 29.7 Ci of translation-grade [ 35 S]methionine (1174 Ci/mmol) and a 1 mM concentration of the other amino acids. Otherwise, the reaction mixture was as described above. Small aliquots of translated proteins were heated at 95°C for 2 min in the sample buffer containing 5% 2-mercaptoethanol and subjected to 15% SDS-PAGE together with prestained molecular mass standards. Following electrophoresis, the gels were destained, dried, and subjected to autoradiography.
PLA 2 Assay Using Liposomes-Enzyme activity was measured at pH 4 (40 mM acetate buffer with 5 mM EDTA) using either a liposome-based radiochemical assay as described previously (9) or a fluorescence assay (bisbodipy-C 11 -PC) for rapid screening. For the radiochemical assay, the liposomal substrate was labeled with 1-palmitoyl-2-[9,10-3 H]palmitoylsn-glycerol-3-phosphocholine ([ 3 H]DPPC)/egg PC/egg phosphatidylglycerol/cholesterol in a molar ratio of 10:5:2:3. The specific activity of [ 3 H]DPPC was 4400 dpm/nmol. Lipids were dried under N 2 , resuspended in isotonic saline, repeatedly freeze/thawed by alternating liquid N 2 and warm H 2 O, and then extruded through a 100-m membrane to generate unilamellar liposomes. The 250-l PLA 2 assay reaction FIG. 1. Column chromatography profiles for rat lung aiPLA 2 . Protein in buffer (50 mM Tris-HCl (pH 7.4), 1 mM EDTA, and 5% glycerol) was applied sequentially to DE52, Sephacryl S-100, and DE52 columns. Fractions from each column were collected and measured by the fluorescence assay for Ca 2ϩ -independent PLA 2 activity at pH 4 in the absence (q) and presence (Ç) of MJ33 and for protein content (E). A, DE52. The column was washed with 2 bed volumes of buffer (fractions 0 -42) and then eluted by a linear gradient of 0 -50 mM NaCl at a flow rate of 60 ml/h. B, Sephacryl S-100. MJ33-inhibitable active fractions from the DE52 columns (bracketed in A) were applied to a Sephacryl-100 column. The column was eluted with buffer at a flow rate of 17 ml/h. C, second DE52. MJ33-inhibitable active fractions from the Sephacryl-100 column (bracketed in B) were applied to a second DE52 column. The column was washed with 5 bed volumes of buffer and then eluted with a linear gradient of 10 -60 mM NaCl at a flow rate of 60 ml/h (dashed line). MJ33-inhibitable fractions (bracketed) to a phenyl-Sepharose column (see text). volume contained 0.226 mol of total lipid, generally in 50 mM sodium acetate buffer (pH 4) plus 1 mM EGTA. Incubation was at 37°C generally for 1 h. Enzyme activity in the assay showed linear increase with incubation time between 15 and 120 min with both wheat germ-and oocyte-expressed proteins. The reaction was stopped by the addition of CHCl 3 /CH 3 OH (2:1, v/v); lipids were extracted by the method of Bligh and Dyer (18); and radiolabeled free fatty acids were separated by a two-step TLC procedure using hexane/ether/acetic acid as a solvent system (9). Authentic palmitic acid was co-chromatographed. The free fatty acid spots were identified using I 2 vapor, scraped from the plate, and analyzed by scintillation counting using an internal standard for quench correction. Recovered dpm was corrected for blank values obtained in the absence of enzymes. PLA 2 activity was calculated based on the specific radioactivity of DPPC. In some experiments, lipids were analyzed by TLC using CHCl 3 /CH 3 OH/NH 4 OH/H 2 O (65:35:2.5:2.5 by volume) (19) to separate lyso-PC in order to assay for PLA 1 activity. In other experiments, liposomes were labeled with [choline-3 H]DPPC in addition to the usual label, and the generation of labeled free fatty acid and lyso-PC was compared.
For the PLA 2 fluorescence assay, the liposomal substrate was DPPC/ bisbodipy-C 11 -PC/phosphatidylglycerol/cholesterol in a molar ratio 10: 0.05:2:3, and total lipid was 0.171 mol in 250 l of sodium acetate (50 mM) plus EGTA (1 mM) buffer at pH 4. To stop the reaction, the medium was diluted to 2 ml with the assay buffer, and the fluorescent product was measured at 490 nm excitation and 520 nm emission. Standard curves at pH 4 and 8.5 were linear with bodipy-C 11 fatty acid concentration up to 4 M and were used to calculate PLA 2 activity.
To test substrate specificity of PLA 2 , liposomes containing PC with various radiolabeled fatty acyl or alkyl constituents were prepared with lipids in a molar ratio of 10:2:3 for PC/phosphatidylglycerol/cholesterol. To test H ϩ dependence, PLA 2 activity with the standard or a micellar assay (see below) was measured in buffers with varying pH plus or minus 10 mM CaCl 2 (all with 1 mM EGTA). Varying pH buffers contained 50 mM glycine (pH 3), 50 mM acetate (pH 4 -5), 50 mM MES (pH 6), or 50 mM Tris (pH 7-9). For evaluation of MJ33, the inhibitor was added (usually at 3 mol %) to the lipid mixture prior to generation of the liposomal substrate. The effects of potential PLA 2 inhibitors (AA-COCF 3 , BEL, and DENP) were determined following preincubation of enzyme with inhibitor at pH 4 for 30 min at 37°C. To determine the effect of pBPB, protein was preincubated with inhibitor at pH 7.4 in 25 mM Tris containing 0.5 mM EDTA. The concentrations selected for testing of inhibitors have been shown previously to produce maximal effect in other systems (9,20,21). PLA 2 Assay Using Mixed Micelles-Mixed micelles were prepared with 10 mM [ 3 H]DPPC (specific activity, 4400 dpm/nmol) and 4 mM Triton X-100 in saline by sonication using a cuphorn sonicator (Heat Systems Ultrasonics, Plainview, NY) at 70% of maximum power with 3 ϫ 20-s bursts at 30-s intervals. For PLA 2 assay, 50 l of enzyme, 850 l of buffer, and 100 l of micelles (final DPPC concentration, 1 mM) were incubated at 37°C for 60 min in the presence or absence of Ca 2ϩ . Lipid radioactivity was assayed as described above for the liposomal assay.
Lysophospholipase Assay-To measure lysophospholipase activity, the [ 3 H]lyso-PC substrate was generated by treating liposomes labeled with [choline-3 H]DPPC with N. naja PLA 2 at pH 8.5 plus 10 mM Ca 2ϩ to achieve complete hydrolysis. [ 3 H]Lyso-PC was separated from the incubation mixture by TLC, and the lyso-PC spot was scraped and extracted with CHCl 3 /CH 3 OH (20:1) (19). Enzyme was incubated with labeled lyso-PC as substrate at pH 4 or 8.5 for 1 h at 37°C. Lipids were extracted by the method of Bligh and Dyer (18), and the aqueous fraction was assayed for radioactivity (representing [ 3 H]glycerophosphorylcholine or its degradation products).
PAF  Northern Blot Analysis-Granular pneumocytes were prepared by elastase digestion of rat lungs as described previously (23). Total RNA was extracted from freshly isolated granular pneumocytes or from homogenized rat lungs using the acid guanidinium thiocyanate/phenol/ chloroform extraction method (24). The RNA was dissolved in diethyl pyrocarbonate-treated H 2 O, quantitated by absorbance at 260 nm, and stored frozen at Ϫ80°C. Extracted RNA samples were electrophoresed on 1% agarose gels containing formaldehyde (15). The size-separated RNA was capillary-transferred onto nitrocellulose membranes using 20 ϫ SSC (1 ϫ SSC ϭ 150 mM NaCl and 15 mM sodium citrate). After an overnight transfer, the RNA was fixed by baking for 2 h at 80°C in a vacuum oven. [ 32 P]dCTP (3000 Ci/mmol) was used to generate 32 Plabeled HA0683 cDNA probes by random priming using Klenow enzyme. Membranes were prehybridized in a solution consisting of 5 ϫ SSC, 5 ϫ Denhardt's solution (15), 50% formamide, 1% SDS, and 10 g/ml denatured salmon sperm DNA at 42°C for at least 4 h. Hybridization with the probe (5 ϫ 10 5 cpm/ml) was performed by overnight incubation in the same buffer at 42°C. Membranes post-hybridization were washed once with 2 ϫ SSC and 0.1% SDS for 15 min at room temperature and twice with 0.1 ϫ SSC and 0.1% SDS at 65°C for 30 min and then exposed to x-ray film.

RESULTS
Isolation of Rat Lung PLA 2 -By the assay used to identify fractions during the isolation procedure, the isolated rat lung protein was active at acid pH in the absence of Ca 2ϩ and was inhibited by MJ33. The sequential use of the DE52, Sephacryl S-100, and repeat DE52 columns resulted in a 163-fold increase in PLA 2 specific activity, although 99% of the starting activity was lost (Table I). The subsequent phenyl-Sepharose column resulted in further purification as determined by SDS-PAGE (Fig. 2), although insufficient protein was recovered for enzymatic characterization. Enzymatic properties of the preparation were analyzed with the partially purified protein from the second DE52 column.
Identification of the Human cDNA Clone Corresponding to Rat Lung aiPLA 2 -SDS-PAGE of the fraction with aiPLA 2 activity that eluted from the phenyl-Sepharose column showed two proteins bands with apparent molecular masses of 26.3 and 25.0 kDa (Fig. 2), and these were used for internal amino acid sequencing. Fragments produced by tryptic digests of these two proteins were subjected to amino acid sequence analysis. For the 25.0-kDa protein, two fragments (25 amino acids total) were sequenced and found to have significant homology to rat brain thiol-specific antioxidant (25). Therefore, this protein was not further evaluated. For the 26.3-kDa protein, three internal peptides comprising 35 amino acids were sequenced and showed 100% identity (Fig. 3) to a reported translated open reading frame of unknown function from a human male myeloblast cell line, KG-1 (11). Fig. 3 illustrates the nucleotide sequence of this 1653-base pair cDNA (HA0683) and its deduced amino acid sequence comprising 224 residues (calculated M r ϭ 25,032). In addition to the putative protein coding sequence, this clone contains 44 base pairs of upstream and 934 base pairs of downstream sequence (Fig. 3).
Expression in Wheat Germ in Vitro Translation System-To determine if the human cDNA encodes a translatable protein with a molecular mass similar to that of rat lung protein, HA0683 cDNA was transcribed in vitro with T7 polymerase and translated in vitro using a wheat germ system. SDS-PAGE with autoradiography of the translated [ 35 S]methionine-labeled protein (Fig. 4) showed an apparent molecular mass of 25.8 kDa, similar to the mass of the predicted protein and of the PLA 2 enzyme isolated from rat lung (Fig. 2). [ 35 S]Methioninelabeled protein expressed by the wheat germ system increased as a function of RNA concentration, with saturation at ϳ1.0 g in a 50-l reaction volume (Fig. 4). Saturation of the wheat germ system at higher concentrations of RNA has been shown previously (26).
Ten separate in vitro translations were carried out with the wheat germ expression system; although these used varying concentrations of cRNA, in each we found the expression of an enzyme activity with the characteristics of aiPLA 2 (assay at pH 4 in Ca 2ϩ -free buffer). PLA 2 activity of the expressed protein measured by the radiochemical assay increased linearly as a function of the cRNA concentration used for expression up to 1 g (Fig. 5). For translation with 1 g of cRNA, expressed PLA 2 activity was 1556 Ϯ 20 pmol/h (mean Ϯ S.E.; n ϭ 7). Control incubation with the wheat germ system in the absence of cRNA showed zero PLA 2 activity. To confirm PLA 2 activity, generation of free fatty acid from [ 3 H]palmitoyl-labeled DPPC and of [ 3 H]lyso-PC from [ 3 H]choline-labeled DPPC was measured in the same assay using protein expressed in the wheat germ system. Liberation of the free fatty acid was 1288 Ϯ 32 pmol/h, and generation of lyso-PC was 1273 Ϯ 27 pmol/h (mean Ϯ range; n ϭ 2). Blank values for PLA 2 assays were ϳ250 dpm, and the usual value for expressed enzyme was ϳ6000 dpm, giving a satisfactory signal-to-noise ratio of ϳ24.
Characterization of aiPLA 2 Activity-The standard PLA 2 assay was carried out in the absence of Ca 2ϩ . The activities of both the expressed PLA 2 at pH 3-8.5 as well as the partially purified rat lung enzyme at pH 4 were not affected by the addition of Ca 2ϩ (Fig. 6). The pH versus activity profile for the isolated lung enzyme showed maximal activity at pH 4 and essentially no activity at pH 6 and above (Fig. 6). The pH profile for the expressed protein was similar (Fig. 6). The pH profile and Ca 2ϩ independence for the expressed protein were similar using either the liposomal or micellar assay system (Fig. 6). aiPLA 2 expressed in vitro showed similar activity using dipalmitoyl-PC or palmitoylarachidonoyl-PC as substrate (Fig. 7 and Table II). The apparent K m for the substrates was ϳ250 M. PLA 2 activity using the alkyl ether phospholipid, O-hexadecylarachidonoyl-PC, as substrate at 1 mM was lower by ϳ50% compared with DPPC (Table II), although the apparent K m was similar (Fig. 7). In vitro translated aiPLA 2 did not show any PLA 1 or lysophospholipase activity, and PAF acetylhydrolase activity was negligible (Table II).
Potential inhibitors (AACOCF 3 (100 M), BEL (100 M), pBPB (20 M), and 2-mercaptoethanol (5 mM)) were tested for their effects on the activity of the isolated rat lung enzyme and in vitro translated aiPLA 2 (Table III). ATP (10 mM), an activator of myocardial iPLA 2 (27), was also evaluated. None of these agents had a significant effect on the activity of the rat lung or expressed enzyme. MJ33 significantly inhibited the activity of FIG. 6. pH dependence of aiPLA 2 from rat lung compared with human aiPLA 2 expressed in vitro using the wheat germ system. A, PLA 2 activity of partially purified enzyme from rat lung was measured using the fluoresence assay at varying pH in the absence or presence (at pH 4) of Ca 2ϩ . B, PLA 2 activity of in vitro translated enzyme using the wheat germ system was measured using [ 3 H]DPPC in the liposomal and micellar assays at varying pH in the absence or presence of Ca 2ϩ . All buffers contained 1 mM EGTA. 10 mM CaCl 2 was added for the ϩCa 2ϩ incubations.

FIG. 7.
Effect of substrate concentration on aiPLA 2 activity. 50 l of aiPLA 2 expressed in vitro using the wheat germ system was incubated for 1 h at 37°C with varying concentrations of DPPC, palmitoylarachidonoyl-PC (PAPC), or O-hexadecylarachidonoyl-PC (HAPC). aiPLA 2 activity expressed in vitro was measured using the standard assay. The apparent K m and apparent V max for each substrate were calculated by standard methods from double-reciprocal plots. aiPLA 2 (Table III), with a maximal effect at 1 mol % (Fig. 8). Because of the "lipase" motif ( Fig. 3), the effect of the serine protease inhibitor DENP was evaluated and was found to significantly inhibit aiPLA 2 activity, with 80% inhibition at 0.5 mM (Fig. 8 and Table III).
Expression of Human aiPLA 2 in Xenopus Oocytes-To further evaluate the protein product of the HA0683 clone, cRNA generated from the clone was expressed in Xenopus oocytes and assayed for aiPLA 2 activity. Expressed aiPLA 2 activity increased as a function of time after oocyte injection, reached a maximum at 30 h, and maintained this level at 48 h postinjection (Fig. 5). Further studies were carried out using oocytes at 48 h after injection. In each of seven separate injection experiments, we observed an increase in aiPLA 2 activity compared with oocytes injected with deionized H 2 O, with a mean increase of ϳ40% (Table IV). Similar results were obtained using both the fluorescence and radiochemical assays. For the radiochemical assay, the blank value (ϳ250 dpm representing Յ50 dpm/oocyte) was subtracted from the measured PLA 2 activity; this gave a signal-to-noise ratio of ϳ10 for the expressed PLA 2 activity or ϳ4 for the difference between cRNA-and deionized H 2 O-injected oocytes. To confirm PLA 2 activity, liberation of free fatty acid and generation of lyso-PC were compared in the same assay; liberation of free fatty acid was 590 Ϯ 13 dpm/oocyte, and generation of lyso-PC was 580 Ϯ 6 dpm/ oocyte (mean Ϯ range; n ϭ 2). Activity was not changed in the presence of 10 mM Ca 2ϩ , but was decreased to the basal value in the presence of 3 mol % MJ33 (Table V). When assayed at pH 8.5 in the presence of Ca 2ϩ , activity decreased by 90% compared with pH 4. These results parallel those obtained with the wheat germ expression system and the isolated lung enzyme.
Northern Analysis-Since the cDNA clone was isolated from a human myeloblast cell line, it was of interest to test whether this mRNA is present in rat lung, the source of the original protein isolation. Northern analysis of aiPLA 2 RNA demonstrated high levels of expression of a hybridizing mRNA, ϳ1.7 kilobase pairs in size, in both rat lung and granular pneumocytes isolated from rat lung (Fig. 9). DISCUSSION iPLA 2 enzymes have been shown to be widely distributed and ubiquitously expressed in most mammalian tissue, underlining their potential importance in cellular functions (3). Although mammalian iPLA 2 enzymes have been isolated and characterized from various sources, there is scant molecular information due to difficulties associated with purification and insufficient yield. To date, the only iPLA 2 -type enzyme that has been sequenced is a PAF hydrolase (28), which also has properties of a low density lipoprotein-associated PLA 2 (29). Therefore, little is known about the structure or mechanisms of iPLA 2 enzymes or their relationship to other PLA 2 enzymes. Here, we describe the cloning, sequencing, and characterization of a human iPLA 2 (aiPLA 2 ) that shows maximal activity in an acidic medium.
Evidence for aiPLA 2 Activity of Expressed Protein-Expression of the HA0683 cDNA using both the wheat germ and Xenopus oocyte systems demonstrated PLA 2 activity. In vitro translated aiPLA 2 did not show any PLA 1 or lysophospholipase activity, excluding the possibility that the combined activities of those two enzymes could have accounted for the measured PLA 2 activity. Furthermore, assays of activity with both wheat germ-and oocyte-expressed enzyme based on recoveries of labeled free fatty acid from sn-2-fatty acyl-labeled DPPC and labeled lyso-PC from choline-labeled DPPC were nearly identical, indicating PLA 2 for the lipolytic activity. The pH 4 optimum for activity and the Ca 2ϩ independence indicated that the  protein encoded by the HA0683 cDNA clone has enzymatic properties that correspond to the activity profile for the partially purified rat lung enzyme as well as the activity previously demonstrated in homogenates of rat lungs and granular pneumocytes (7,30). Northern analysis demonstrated that this gene is highly expressed in rat lung and granular pneumocytes (Fig. 9). aiPLA 2 activity was significantly inhibited by MJ33, a competitive inhibitor of acidic Ca 2ϩ -independent PLA 2 activity in lung homogenates and subcellular organelles (lysosomes and lamellar bodies) (7,30), and by a serine protease inhibitor, DENP. Unlike other iPLA 2 enzyme activities that have been characterized, the activity of the expressed protein was not affected by the presence of detergent (Triton X-100), and neither it nor the isolated rat lung enzyme was affected by the inhibitor BEL or the activator ATP (27,31). Thus, it appears that aiPLA 2 presented in this report is distinct from the iPLA 2 enzymes previously purified from the P388D 1 macrophage cell line (20) and from cardiac muscle (27). The relative activity of PAF acetylhydrolase, a Ca 2ϩ -independent enzyme that can cleave short-chain acyl groups at the sn-2-position (22), was negligible. The isolated rat enzyme and the expressed protein are differentiated from sPLA 2 and cPLA 2 by their Ca 2ϩ independence and pH profile as well as by insensitivity to inhibitors (pBPB, 2-mercaptoethanol, and AACOCF 3 ).
Properties of the Predicted Protein-The HA0683 cDNA encodes a mature protein of 224 amino acids with a calculated molecular mass of 25.0 kDa. We have shown in a wheat germ in vitro translation system that mRNA transcribed from this clone results in the expression of a protein of 25.8 kDa in size, in reasonable agreement with the predicted mass of the deduced amino acid sequence and with the estimated mass of aiPLA 2 isolated from rat lung. The predicted protein has 32 (14.3%) negatively charged and 30 (13.4%) positively charged residues, with no predicted charge clusters and a predicted pI of 6.0 (32). Nonpolar residues account for 107 (48%) of the 224 amino acids in the predicted sequence. A hydrophobicity plot of the predicted protein did not indicate any extended regions of FIG. 9. Northern analysis of aiPLA 2 in rat lung and granular pneumocytes. Total RNA from rat lung and granular pneumocytes (10 g) and cRNA (10 pg), transcribed in vitro from the putative aiPLA 2 clone using T7 RNA polymerase, were electrophoresed on a formaldehyde-agarose gel, transferred to a nitrocellulose membrane, and hybridized with the aiPLA 2 probe labeled with [ 32 P]dCTP by random priming. A single band of ϳ1.7 kilobase pairs was detected. The expression of aiPLA 2 was seen in lung and granular pneumocytes. The positions of 18 S and 28 S ribosomal RNA are indicated.   high hydrophobicity, consistent with the fact that aiPLA 2 was isolated as a soluble protein.
Sequence Homology to Other Proteins-Searches for similarity to the HA0683 cDNA sequence at the protein (SWISSPROT protein data base) and DNA (GenBank TM data base) levels did not identify any phospholipases. Therefore, aiPLA 2 is apparently not related to any of the phospholipases previously sequenced. The predicted aiPLA 2 enzyme contains a 5-amino acid sequence, GXSXG (double underline in Fig. 3), that has been described as a lipase motif and may represent the active serine of the catalytic triad SDH (33). Inhibition by DENP supports a serine-based mechanism for aiPLA 2 activity. The motif has been described in serine proteases and in neutral lipases. Of note, it has also been described in PAF acetylhydrolase (low density lipoprotein-associated PLA 2 ) (28,29). Thus, the motif is present in both iPLA 2 enzymes (aiPLA 2 and PAF acetylhydrolase) that have been cloned, although it is not observed in sPLA 2 or cPLA 2 .
From the protein data base, the amino acid sequence with the most similarity to an HA0683-encoded protein was a hypothetical 29.5-kDa protein from yeast (SWISSPROT ID YBG4) (34), which matched at 99 out of 261 amino acid positions. A search for similarity to the deduced N-terminal sequence of aiPLA 2 revealed homology to two human unknown proteins isolated by two-dimensional gel electrophoresis from liver (SWISSPROT ID P30041) (35) and red blood cells (SWISSPROT ID P32077) (36). These sequences, consisting of 14 and 12 amino acids, respectively, showed 100% homology to the predicted aiPLA 2 sequence following its deduced start site. Electrophoresis showed molecular masses of 23.5 kDa for the liver protein (35) and 26 kDa for the red blood cell protein (36), with an identical pI of 6.2. The similarities in N-terminal sequence, size, and pI to the parameters predicted for the HA0683-encoded protein suggest that these human proteins of unknown function may represent aiPLA 2 , indicating widespread distribution for the enzyme. Although originally isolated from a human myeloid cell line, HA0683 hybridized to RNA from all tissues and cell lines tested (11). Since aiPLA 2 may be a lysosomal enzyme, its widespread distribution would be expected.