The Ca2+-sensitive Cytosolic Phospholipase A2 Is a 100-kDa Protein in Human Monoblast U937 Cells*

Human monoblast US37 cells contain a soluble phos- pholipase Az (PLAz) that is activated over the range of 150-600 nM Ca2+ and is stable only at neutral pH. We have purified this PLAz over 34,000-fold to near ho- mogeneity using sequential ion exchange, hydrophobic interaction, and gel filtration chromatography steps. The protein has a M, of - 100,000 (by sodium dodecyl sulfate-polyacrylamide gel electrophoresis) and an iso- electric point of 5.1. Four lines of evidence indicate that this 100-kDa polypeptide represents the PLA2. (i) The intensity of staining of the 100-kDa protein was proportional to the degree of purification of PLAz ac- tivity, (ii) the relative staining intensity of the 100-kDa protein precisely paralleled the elution profile of PLAz activity during chromatography steps, (iii) the PLAz activity recovered from a nondenaturing gel (>60% of the total activity applied) coincided exactly with the major high molecular weight protein detected by silver staining, and (iv) monoclonal antibodies against the 100-kDa protein immunoprecipitated the PLAz. We conclude that the cytosolic PLAz isolated from US37 cells represents a novel, high molecular weight PLAz responding to physiological (intracellu-lar) changes in Ca2+ concentration and therefore may play a critical role in cellular signal transduction processes and the biosynthesis of lipid mediators. Phospholipases beads were washed with 1 ml of 150 mM NaC1, 10 mM Tris/HCl, pH 7.5, containing 1 mg/ml BSA and incubated with 125 ng of phenyl-Superose 10/10 purified PLA, added in 100 pl of the same buffer for 4 h at room temperature. PLA, assays were performed on supernatants and pelleted beads to estimate mouse antibody-mediated binding of PLA, to anti-(mouse-1gG)-coated Pandex beads. Other Methods-Protein measurements in fractions were made using the BCA protein assay (Pierce Chemical Co.). The protein content of highly purified PLA, preparations was estimated using staining intensity on SDS-polyacrylamide gels and/or absorbance at 280 nm. Free Ca2+ was measured using the Ca2+ fluorescence probe Fluo-3 (Molecular Probes) and a SLM model 48000 spectrofluori- meter.

cellular cytosolic PLA, involved in signal transduction processes and receptor-mediated liberation of arachidonic acid represents a different enzyme with distinct molecular and functional properties (7). Such a cytosolic PLAz has been characterized in platelets (8,9), renal mesangial cells (lo), macrophages (ll), and leukemia cell lines (12, 13).' It has an apparent M, of 60,000-70,000 (by gel filtration), is unstable, requires less than millimolar concentrations of Ca2+ for activity, and exhibits a preference for arachidonate esterified at the sn-2 position of phospholipid substrates. The purification to homogeneity and definite identification of this PLAz have been difficult due to the low levels (<0.01%) of protein present in cells and the loss of activity resulting from proteolysis, aggregation, and denaturation. Consequently, a widespread discrepancy exists regarding the molecular weight and identity of this intracellular PLAz (8-13). In this report we describe the successful purification of a cytosolic PLAz from U937 cells responding to physiological increments of Ca2+ and present biochemical and immunochemical evidence that unambiguously demonstrates that this PLA, is a 100-kDa protein.

EXPERIMENTAL PROCEDURES
Materials-Q Sepharose Fast Flow and Superose 12 resins, as well as fast protein liquid chromatography Mono Q HR 5/5, phenyl-Superose H R 10/10, and HR 5/5 chromatographic columns, were from Pharmacia LKB Biotechnology Inc. BSA was from Sigma, and prestained protein molecular weight standards were from Bethesda Research Laboratories.
Culture of U937 CelIs"U937 cells (obtained from the American Type Culture Collection) were cultured a t 37 "C in modified Dulbec-CO'S modified Eagle's medium/F12 medium (GIBCO), 5% fetal bovine serum (HyClone), 4 mM L-glutamine, 0.02% Pluronic F68 (BASF Corp.), 50 pg/ml tobramycin (Lilly), 20 mM Hepes, 25 mM NaHC03, and 100 p~ ethanolamine using a 40-liter Biotech overhead drive system (Bellco Biotechnology). The culture was stirred at 6-7 rpm and sparged first with 70% Nz, 25% 02, and 5% Conand finally, when the cell density had reached approximately 1 X lo6 cells/ml, with 100% 02. When cell density was 2-3 X lo6 /ml, the cell suspension was transferred to a sterile 50-liter carboy (Nalgene) and stored at 4 "C for 48 h to allow settling of cells. All subsequent steps were performed at 4 "C. After removal of the cell-free medium by aspiration, the remaining cell suspension was centrifuged (800 X g, 30 min), and cell pellets were suspended in 140 mM NaCl, 5 mM KC1, 2 mM EDTA, 25 mM Tris/HCl, pH 7.4. The cells were washed with the same buffer and resuspended a t 108/ml in buffer containing 100 p M leupeptin and 50 FM pepstatin A.
Purification of PLA2-After addition of phenylmethanesulfonyl After completion of our manuscript Clark et al. (32) reported on the purification of a 110-kDa PLA, from U937 cells. It is noteworthy that taking into account all differences in the respective PLAz assay systems (e.g. our incorporation of dioleoyl glycerol into phosphatidylcholine liposomes resulting in a 4-fold enhancement of PLA, activity versus their addition of 70% glycerol to incubations stimulating the PLA2 activity >&fold) the cytosolic PLA, prepared by Clark et al. and the one described here exhibit similar specific activities. fluoride to 1 mM, cells (-10") were lysed by nitrogen cavitation using a Parr cell disruption bomb (600 p.s.i. for 20 min at 4 "C). The lysate was centrifuged (150,000 X g, 60 min, at 4 "C), and the supernatant media were collected and filtered through a Millipak (0.22 pm) filter (Millipore). The resulting clear filtrate was applied a t a flow rate of 6 ml/min to a column of Q Sepharose (3.2 X 14 cm), pre-equilibrated with 2 mM 2-mercaptoethanol, 1 mM EGTA, 25 mM Tris/HCl, pH 8 (buffer A) containing 150 mM NaC1. After extensive washing, a 1500ml linear salt gradient was developed from 150-600 mM NaCl in buffer A. Fractions were collected (12 ml), and those with PLAz activity were pooled (120 ml) and concentrated to 40 ml. The concentrate was applied to a phenyl-Superose HR 10/10 column, preequilibrated in buffer A containing 750 mM NaCl, at a flow rate of 1 ml/min. After extensive washing of the column at the same flow rate, a 120-ml linear gradient of ethylene glycol was generated by mixing buffer A containing 750 mM NaCl and buffer A containing 50% ethylene glycol. The flow rate was 0.75 ml/min, and 4-ml fractions were collected. The active fractions were pooled (60 ml), concentrated to <1 ml, resuspended in 15 volumes of buffer A containing 30% ethylene glycol and 150 mM NaCl, and reconcentrated to 1 ml (to give a final ethylene glycol concentration of 30%). T h e concentrate was chromatographed on tandem Superose 12 columns (each 1.6 X 50 cm) pre-equilibrated with buffer A containing 30% ethylene glycol and 150 mM NaCI. The flow rate was 0.15 ml/min, and 1.5-ml fractions were collected. The fractions containing PLA, activity were pooled (15 ml), concentrated to <0.65 ml, resuspended in 20 ml of buffer A containing 150 mM NaCl (to give a final ethylene glycol concentration of <1%), and reconcentrated to <0.65 ml. The concentrate was loaded a t a flow rate of 0.25 ml/min onto a phenyl-Superose H R 5/5 column pre-equilibrated with buffer A containing 150 mM NaCI. The column was washed with the same buffer collecting 0.5ml fractions. The PLA2 activity eluted after 2 column volumes in 7 ml. In some preparations combined active fractions from the phenyl-Superose column were applied directly to a Mono Q H R 5/5 column previously equilibrated with buffer A containing 150 mM NaCl. PLA, activity was eluted a t 0.5 ml/min with a linear salt gradient (150-600 mM NaCl) collecting 0.5-ml fractions.
Polyacrylamide Gel Electrophoresis (PAGE)-SDS-PAGE was performed according to Laemmli (14). For native PAGE, samples were mixed with 1 volume of 5 mM Tris acetate, pH 7.5, containing 20% glycerol, loaded onto a 6% polyacrylamide minigel prepared in 50 mM Tris acetate, pH 7.5, and electrophoresed at 30 mA for 2 h a t 4 "C using 50 mM Tris acetate, pH 7.5, as electrophoresis buffer. For elution of PLA,, gel slices (1 X 0.3 mm) were transferred into 150 p1 of 150 mM NaC1, 3 mM 2-mercaptoethanol, 1 mM EGTA, 25 mM Tris/HCl, pH 8, containing 1% CHAPS (Sigma) and 2 mg/ml BSA and incubated overnight at 4 "C. Proteins were visualized on the gels by silver staining (15). Isoelectric focusing was performed with a Bio-Rad 111 mini isoelectric focusing cell following the instructions of the manufacturer.
Zmmunoblotting-Mono Q-purified PLA, (80 ng/lane) was subjected to SDS-PAGE electrophoresis on 8-16% gels (Novex) and transferred to nitrocellulose (Novex) using the Sartoblot electroblotter system (Sartorius). The nitrocellulose sheets were treated with 0.1% Triton X-100, 150 mM NaC1, 1 mM EDTA, 10 mM Tris/HCl, pH 7.5 (buffer B) containing 3% BSA and then incubated with immune serum or hybridoma supernatant media in the presence of 0.1% Triton X-100. The sheets were washed with buffer B, incubated for 4 h a t room temperature with lZ5I-labeled sheep anti-mouse IgG F(ab')2fragment (Amersham Corp.), washed again, and exposed to xray film for 7.5 h a t -70 "C.
Assay of PLA,Actiuity-PLA2activity was assayed using sonicated liposomes containing l-palmitoyl-2-[14C]arachidonoyl-sn-glycero-3phosphocholine (52 mCi/mmol, from Du Pont-New England Nuclear) and sn-1.2-dioleoyl glycerol (Avanti Polar Lipids) a t a molar ratio of 2:l as previously described (8) and modified as follows. The assay mixture contained 1 mM CaC12, 2 mM 2-mercaptoethanol, 150 mM NaC1, 50 mM Hepes, pH 7.4, and 1 mg/ml BSA. T h e substrate consisted of 2 p~ radiolabeled phosphatidylcholine liposomes (50,000 dpm per incubation) containing 1 p~ dioleoyl glycerol. As reported earlier (16), dioleoyl glycerol incorporated into the phosphatidylcholine substrate stimulates the activity of cytosolic PLA, approximately 4-fold. The pH optimum for the cytosolic PLA2is 9, but in an attempt to assay under more physiological conditions, incubations were performed at pH 7.4. At this pH the PLA, activity is 2-fold less than at p H 9. To probe for secretory PLA, in acid-extracted U937 cells, the Escherichia coli assay system was used (17).
Preparation of Antibodies-Immunization was performed by in-jecting 8 pg of antigen (phenyl-Superose H R 5/5-purified PLA,) in Freund's complete adjuvant intraperitoneally into BALB/c mice. Two additional injections with 8 pg of antigen and a third with 4 pg (all emulsified in Freund's incomplete adjuvant) were given at an interval of 3 weeks. A final injection of 9 pg (without adjuvant) was administered 3 days prior to the fusion. Fusion of spleen cells was performed as described by Goding (18). Splenocytes (2 X 10') were fused with 2 X lo7 HL-1 Friendly Myeloma-653 cells (Ventrex) using 43% poly(ethy1ene glycol) 4000 (ATCC). Cells were then plated into 960 wells and grown in hypoxanthine/aminopterin/thymidine medium (Sigma) containing 10% hybridoma enhancing media (Sigma), 20% fetal bovine serum (HyClone), 35% HL-1 (Ventrex), and 35% RPMI-1640 with Hepes (GIBCO). After 2 weeks the supernatants of the hybridomas were tested for production of antibodies in an enzymelinked immunosorbent assay using Mono Q-purified PLA, for screening (0.04 pg/well) according to Ausubel (19). Immunoprecipitation of PIA-Mouse antiserum, hybridoma supernatant media, or control solutions (preimmune serum, hybridoma medium, or supernatant from hybridoma producing antibodies against protein distinct from PLA,) were incubated with 100 p1 Flurotec anti-(mouse-IgG) coated beads (Pandex MS-00-1) overnight at 4 "C with gentle rocking. The beads were washed with 1 ml of 150 mM NaC1, 10 mM Tris/HCl, pH 7.5, containing 1 mg/ml BSA and incubated with 125 ng of phenyl-Superose 10/10 purified PLA, added in 100 pl of the same buffer for 4 h a t room temperature. PLA, assays were performed on supernatants and pelleted beads to estimate mouse antibody-mediated binding of PLA, to anti-(mouse-1gG)-coated Pandex beads.
Other Methods-Protein measurements in fractions were made using the BCA protein assay (Pierce Chemical Co.). The protein content of highly purified PLA, preparations was estimated using staining intensity on SDS-polyacrylamide gels and/or absorbance a t 280 nm. Free Ca2+ was measured using the Ca2+ fluorescence probe Fluo-3 (Molecular Probes) and a SLM model 48000 spectrofluorimeter.

RESULTS
Purification of PLA2 from U937 Cells"u937 cells do not express detectable levels of the 14-kDa secretory PLAz as determined by extracting U937 lysates with acid and assaying PLAz activity with the E. coli substrate (6). However, as demonstrated below, U937 cells provide an excellent source for isolation of the cytosolic PLA2. The purification protocol used to first isolate and identify the PLA, is summarized in Table I and Fig. 1. Upon ultracentrifugation of disrupted cells, the PLAz activity was recovered in the soluble fraction and the membrane fraction contained less than 10% of the lysate PLAz activity. The recovery of PLAz in the soluble fraction was critically dependent on the presence of Ca2+-chelating agents and protease inhibitors. The apparent increase in total PLAz activity observed in the cytosolic compared with the lysate fraction could be due to the removal of an endogenous inhibitor of PLAz and/or membrane phospholipid competing with the radiolabeled substrate in the PLAz assay. The cytosolic PLA, was purified by sequential anion exchange, hydrophobic interaction (high salt), gel filtration, and hydrophobic interaction (low salt) chromatography. First, the 150,000 x g supernatant (fraction I) was applied to an anion exchange column that was developed with a gradient of NaC1. The PLAz activity eluted as a single peak with 380-450 mM salt (fraction 11) and was subsequently bound to a phenyl hydrophobic interaction column equilibrated with buffer containing 750 mM NaCl. Under these conditions the PLAz strongly bound and could only be eluted with ethylene glycol (15-35%) (fraction 111). The concentrated active fractions were chromatographed on a Superose 12 column. The PLAz activity eluted from the gel filtration column as a single peak with an apparent molecular mass of 70 kDa (fraction IV). The active fractions were concentrated, resuspended, and reconcentrated (to lower the ethylene glycol to 4 % ) and applied to a phenyl hydrophobic interaction column equili- " PLA, activity was determined using the standard assay as detailed under "Experimental Procedures." The phosphatidylcholine concentration used (2 p~) was subsequently found to be nonsaturating (see "Results"), and the values indicated represent -67% of the activity attainable with saturating substrate concentrations. PLA? activity is expressed as milliunits, where 1 milliunit is defined as the amount of enzyme required to hydrolyze 1 nmol of phospholipid/min. * Phenyl-Superose columns were equilibrated and PLA, activity loaded in buffer containing 750 mM NaCl (111) and 150 mM NaCl (V).
As described in the text and demonstrated in Fig. 3, further purification (-2-fold) was obtained, when fraction V was subjected to chromatography on a Mono Q HR 5/5 column. brated in buffer containing 150 mM NaCl. Here, the PLAz activity bound only weakly and eluted during the washing step after 2 column volumes (fraction V), while most of the contaminating proteins either flowed through immediately or bound strongly to the column. The isolated PLA, after this purification step contained a major band of 100 kDa as well as a minor band of 55 kDa (Fig. 1). Overall, the procedure yielded less than 100 pg of protein, and the PLA, activity was purified >34,000-fold with a 23% yield relative to the lysate fraction.
Electrophoresis under Native Conditions-In order to investigate whether the 100-kDa protein represented the PLA,, the final enzyme preparation (fraction V) was subjected to electrophoresis under native conditions as described under "Experimental Procedures" applying 50 ng of protein in duplicate lanes. The two lanes were cut from the gel. One was stained with silver and the other sliced into small pieces that were eluted with buffer containing 1% CHAPS and assayed for PLA, activity. We found that greater than 60% of the applied PLA, activity was recovered after elution and that the PLA2 activity profile coincided with the major, darkly stained high molecular weight band indicating that this band represents the PLAZ protein (Fig. 2). Isoelectric focusing of the purified PLA, revealed that the 100-kDa protein has an isoelectric point of 5.1.
Correlation of PLA2 Activity and 100-kDa Protein upon Mono Q Chromatography-In some PLA, preparations the active fractions from the phenyl-Superose HR 515 column were directly loaded onto a Mono Q HR 515 column that was subsequently eluted with a NaCl gradient. Fractions from this column containing protein and PLAz activity were analyzed Detection of PLA2 activity after native PAGE of purified PLA2. Aliquots of purified PLA, (50 ng, fraction V) and a standard protein (BSA) were applied in parallel lanes and subjected to electrophoresis under nondenaturing conditions. One lane was sliced into 20 pieces that were eluted with buffer containing 1% CHAPS, and 5-pl aliquots of these extracts were assayed for PLA, activity as detailed under "Experimental Procedures." The remaining gel was silver-stained. BSA by SDS-PAGE. As demonstrated in Fig. 3, only a single major intense band at 100 kDa was observed in the fractions containing PLA, activity. Furthermore, the relative intensity of the 100-kDa band precisely paralleled the elution profile of PLA, activity from the Mono Q column. It should be noted that this chromatography not only provided an excellent means for concentration of the PLA2 but also resulted in a further 2-fold increase in the specific activity.
Immunoadsorption of PLA2-In a complementary approach to verify that the 100-kDa protein was indeed the PLA2, antibodies were raised against the 100-kDa protein and tested for immunoreactivity with the PLA2 (monitored via its enzymatic activity). Antibodies against the 100-kDa protein were identified in an enzyme-linked immunosorbent assay using Mono Q-purified 100-kDa protein for selection as detailed under "Experimental Procedures." As demonstrated in Fig.  4A all antibodies were able to immunoprecipitate PLA,, albeit with distinct affinities. The antibody-mediated removal of PLA, activity from the solution was accompanied by a concomitant appearance of PLA2 activity in the immunoprecipitate. In addition, the polyclonal antibody P 132 and the monoclonal antibody M 3-1 also reacted strongly with the 100-kDa band in a Western blot indicating that they were able to recognize the protein in its denatured form (Fig. 4B).
To examine the Ca2+ sensitivity of the cytosolic PLA2 from U937 cells more closely, purified enzyme was assayed in EGTA/CaC12 buffers with nanomolar free Ca2+ concentrations. As shown in Fig. 5 there was a low basal PLA, activity at 100 nM Ca", followed by a sharp 4-fold increase in PLA, activity achieving maximal activity at 600 nM ca2+.
Functional and Kinetic Properties-The cytosolic PLA, was active in the absence of added Ca2+ and was apparently Ca2+independent. However, it was inhibited by 5 mM EGTA suggesting a Ca2+ requirement in the submicromolar range.
using <2 ng of Mono Q-purified enzyme in assays with increasing amounts (0.8-10 PM) of substrate revealed linear kinetics in a Lineweaver-Burk plot and gave an apparent K,,, of 0.9 p M and a V,,, of 12.4 pmol/min corresponding to >6.2 pmol/min/mg?

DISCUSSION
We report here the purification to near homogeneity of a cytosolic PLA, from U937 cells that responds to Ca2+ concentrations similar to those associated with cytoplasmic free Ca2+ transients observed during cell activation. The conclusion that the PLA, is a 100-kDa protein is based on several lines of evidence. (i) The intensity of staining of the 100-kDa protein was proportional to the degree of purification of PLA2 activity ( Fig. 1 and Table I). (ii) When resolved from minor protein contaminants by native gel electrophoresis, the isolated 100-kDa protein band contained all the recovered PLA, activity. (iii) The 100-kDa band co-fractionated precisely with the PLA, activity on the final purification steps, as shown for the Mono Q column (Fig. 3). (iv) Monoclonal antibodies raised against the 100-kDa protein recognized and were able to immunoprecipitate the PLA,.
Convincing evidence has been presented for the existence of Ca2+-independent PLA2s in intestinal brush-border membranes ( M I 98,000) (20) and myocardial cytosol ( M I 45,000) (21). Other studies reported the presence of a Ca2+-dependent PLA2in the cytosolic fraction of various cells elucidating some unique functional properties of this enzyme (8-13). Although purification schemes and tentative molecular weights have been reported, the complete characterization of this PLAz has been hampered by the low abundance and apparent lability upon purification. Recently, Diez and Mong (13) reported the purification to homogeneity of a Ca2+-dependent cytosolic PLA, from U937 cells. Our results differ dramatically from the results of these authors. They purified the PLA2 activity only 1900-fold over the cell lysate and concluded, based solely on the SDS-PAGE analysis of the purification process, that the cytosolic PLA2 has a molecular mass of 56 kDa. However, the intensity of this 56-kDa band was not proportional to the degree of purification in the final purification steps, and the specific activity of the purified PLA2 was substantially less than determined in the present study. Therefore, in the final enzyme preparation obtained by Diez and Mong the cytosolic PLA, still comprised only a small fraction of the total protein, and the 56-kDa band represented a major contaminating protein and not the PLA2.
An important physiological role of the Ca'+-sensitive cytosolic PLA2 may be the release of arachidonic acid and generation of lysophospholipids for the biosynthesis of eicosanoids and platelet-activating factor in response to hormones, neurotransmitters, growth factors, autacoids, and other receptor ligands (10,22,23). The mechanism(s) responsible for receptor-mediated activation of intracellular PLA2 have not been elucidated, but there is indirect evidence to indicate that various factors may be involved, including elevated cytoplasmic free Cap+ (24), influx of extracellular Cap+ (25,26),diacylglycerol (16,27), and protein kinase C (28-30). More recently, it was suggested that GTP-binding proteins may directly couple receptors to PLA, (23, 31), although the specific GTP-binding proteins involved have not yet been identified. The availability of reasonable quantities of cytosolic human PLA,, as well as PLAp antibodies that may serve as immunoprobes, will allow for the molecular studies that are needed to further elucidate the structure and function of this novel intracellular PLAp. Such knowledge will facilitate delineation of the molecular events that control the activity of the cytosolic PLAPand provide a new insight into the involvement of PLA, in cellular signal transduction processes and the generation of lipid mediators.