Inflanunopharmacology. 1997; 5 : 247-260. © 1997 Kluwer Academic Publishers. Printedin the Netherlands

Q U A N T I T A T I O N OF H U M A N T I S S U E A N D I M M U N E CELL T Y P E I I 14 k D a P H O S P H O L I P A S E A z B Y E N Z Y M E IMMUNOASSAY B.J. BOLOGNESE l, S.D. HOLMES3, L.J. McMILLAN 2, K.F. KAISER2 A N D L.A. MARSHALL 1° 1Department of Immunopharmacology, 2Department of Gene Expression Sciences, 709 Swedeland Road, King of Prussia, PA 19406-0939, USA; 3Department of Biotechnology, Great Burgh, Epsom, Surrey, KT18 5XQ, UK *Correspondence ABSTRACT Bolognese BJ, Holmes SD, McMillan LJ, Kaiser KF, Marshall LA. Quantitation of human tissue and immune cell type II 14 kDa phospholipase A2 by enzyme immunoassay. Inflammopharmacology. 1997;5:247-260. T h e metabolism of arachidonic acid into inflammatory mediators (e.g. prostaglandin, leukotrienes) is dependent upon the rate-limitingenzymephospholipase A2. Localization and quantification of type II 14 kDa phospholipase A2 (PLA2) in cells or tissue preparations has historically been accomplished through activity measurements, a process that can provide variable results due t o interference by exogenous substances with hydrolysis assessment. Othershave reported on the use of sandwich enzyme immunoassays (EIA) t o measure 14 kDa PLA2 by mass in serum and exudate fluids, e.g. synovial fluid. Herein, we report the utilization of a human recombinant type II 14 kDa PLA2 sandwich EIA t o directly measure cell or tissue-residing 14 kDa PLA2. It is known that type II 14 kDa PLA2 resists acid treatment, and this technique was applied t o cell fractions which liberated the enzyme from cellular membrane components prior t o quantitation by EIA. Two human immune cell populations were assessed and shown t o contain measurable levels of 14 kDa PLA2. Neutrophil or monocyte cytosolic fractions contained no measurable levels whereas the respective 100000g particulate fractions contained 2.6-t-0.8 pg (neutrophil) and 2.1-t-0.6 pg (monocyte) 14 kDa PLA2/lag protein. Human placenta cytosolic fractions contained no measurable levels while 100 000g particulate contained ~25 ng 14 kDa PLA2/mg protein. This EIA, in conjunction with acid extraction, provides an easy and reproducible assay t o identify and quantify this enzymein cells and whole tissues, expanding our ability t o study the relationship of this enzymet o inflammatory processes.

Keywords." phospholipase A 2 , enzyme immunoassay, neutrophils, monocyte, human placenta, human synovial fluid

INTRODUCTION Phospholipase A2 (PLA2) enzymes (EC 3. I. 1.4) are a class of sn-2 acylhydrolases that remove the sn-2 fatty acid from phospholipids producing free fatty acids and lysophospholipids. Mammalian non-pancreatic type II, 14 kDa PLA2 exists as a n extracellular form which has been found in the synovial fluid (SF) of patients suffering from rheumatoid arthritis [1,2], serum of patients with septic shock [3], and in the medium of cultured macrophages [3,4]. This enzyme also exists in a cell-associated form exemplified by that isolated from a number of cells and tissues, e.g. spleen [5], placenta 247

248

Quantitation of Human Tissue and Immune CellType II 14kDa PLA2by EIA

[2], h u m a n monocytes [6,7] o r neutrophils [8,9] as identified through catalytic activity o r protein purification. A structurally and biochemically distinct 85 kDa PLA2 has been identified [10--14]. This enzyme prefers arachidonic acid (AA) in substrate phospholipid and has been found only as a cell-associated cytosolic form. The localization o f both enzymes initially was described using acylhydrolytic activity [3,9,15-18]. Indeed, discriminating acylhydrolytic analysis has been used to d e m o n strate the co-existence o f the 85 kDa PLA2 as well as the type II 14 kDa PLA2 in h u m a n platelets, neutrophils and monocytes [9] o r m o u s e macrophages [19,20]. Acylhydrolysis is best utilized as a quantitative measure on purified enzyme, while its use on cell o r tissue fractions can only be semiquantitative at best. Analyses in crude samples, such as cell fractions, are often complicated by interfering agents that may bind enzyme a n d / o r substrate, disrupting phospholipid substrate vesicle structure [21]. Direct protein measurement would offer a superior quantitation measurement. However, the 14 kDa PLA2 is present at such low levels in most tissues that partial isolation and concentration are required p r i o r to Western analysis [22]. Protein concentrations o f the extracellular type II 14 kDa PLA2 have been successfully measured in blood o r exudate fluids by a number o f investigators using a double monoclonal sandwich enzyme immunoassay [3,15,16]. W e have developed an EIA to quantitatively measure type II 14 kDa PLA2 employing mouse monoclonal antibodies raised against native recombinant h u m a n (rh) non-pancreatic type II 14 kDa PLA2 which is specific [4,23] and can measure enzyme in the femtomolar range (7-280 fmol). This 14 kDa PLA2 EIA was validated by assessing the levels o f type II 14 kDa PLA2 in h u m a n rheumatoid synovial fluid and then the use o f this assay was extended to quantification o f cell-associated type II 14 kDa PLA2. Acid extraction o f sample to solubilize enzyme has been previously described by Marki and Franson [8]. Employing acid extraction p r i o r to direct EIA analysis, we confirmed the existence o f and quantified an immunologically related cell-association type II 14 kDa PLA2 in h u m a n neutrophils, monocytes and placenta which predominantly occurs in the particulatemembrane containing fraction.

METHODS

Materials and chemicals Hanks balanced salt's solution (HBSS) with and without Ca2+/Mg2+, RPMI-1640, fetal bovine serum (FBS, mycoplasma-free), and phosphate-buffered saline (PBS) was obtained fromGibco (Grand Island, NY, USA). Percoll solution, bovine y-globulin, p nitro phenyl phosphate (PNO), ethyleneglycol-bis-(13-aminoethyl ether)-N,N, NJ,Nltetraacetic acid (EGTA). Tris buffer (Tris), phenylmethylsulphonyl fluoride (PMSF), leupeptin, soybean trypsin inhibitor and aprotinin were purchased from Sigma Chemical Co. (St Louis, MO, USA). Streptavidin-alkaline phosphatase conjugate was purchased from A m e r s h a m (Arlington Heights, IL, USA).

Bolognese et al.

249

Phospholipase A 2 enzymes Recombinant h u m a n (rh) type II 14 kDa PLA2 cloned from a h u m a n placenta cDNA library was expressed as a n authentic processed enzyme in CHO cells and purified essentially by literature methods as previously described [24].

H u m a n synovialf l u i d collection Human synovial fluid from joints o f patients with rheumatoid arthritis was collected and generously donated by Dr A.S. Huppert, Arthritis and Rheumatology Clinical Assistant Professor at Hahnemann University, Philadelphia, PA, USA.

H u m a n immune cell isolation and fractionation Human neutrophils o r monocytes were isolated (85-90% pure) as previously described [9,25]. F o r fractionation, the cells (1-2 x 108 cells/ml) were resuspended in cell fractionation (CF) buffer containing 0.34 mol/L sucrose, 10 m m o l / L Hepes, p H 7.4, 1 m m o l / L EGTA, 1 mmol/L PMSF, 200 limol/L leupeptin, 20 p.g/ml soybean trypsin inhibitor, 20 lag/ml aprotinin as previously described [25]. Inclusion o f EGTA localizes the 85-kDa PLA2 predominantly in the cytosolic fraction [9,26]. The cell suspension was homogenized by nitrogen cavitation (900 psi, 10-15 min, 4°C). The homogenate was centrifuged at 400g for 10 min at 4°C to remove unbroken cells, and the supernatant was centrifuged a t 100 000g for 60 min at 4°C to obtain a supernatant (cytosol) and pelleted particulate fraction. The particulate fraction was resuspended in CF buffer at 20% o f the volume o f cytosol. All fractions and homogenates were flash frozen with liquid N2 and stored at -70°C for analysis.

Human placenta collection Human placenta was collected from a single donor. The placenta was hand h o m o genized in 10 ml CF buffer at 4°C per g r a m o f tissue. Cytosolic and particulate fractions were then prepared from the homogenates in the same m a n n e r as that employed for h u m a n cells described above.

Quantification o f type H 14 kDa PLA2 by EIA Monoclonal antibody development and characterization Mouse anti-rh type II 14 kDa PLAz monocional antibodies (mAb) SK088-3C6.16.2 (3C6) and SK097-1E8.5.2 (1E8) were prepared as previously described [7,27]. C A F l female mice (Charles River Portage, Wilmington, MA, USA) were immunized by three

250

Quantitation of Human Tissue and Immune Cell Type II 14 kDa PLA2by EIA

subcutaneous injections o f rh type II 14 kDa PLA2 (100 lag, 50 ~tg, and 25 lag, respectively), in Freunds complete adjuvant, at 4-week intervals. Immune mice, selected on the basis o f s e r u m titre to the immunizing antigen, were boosted intravenously with rh type II 14 kDa PLA2 for 3 consecutive days p r i o r to removal o f their spleens. Spleen cells were fused with SP2/0-AG14 myeloma cells, and hybrids were selected using H A T (hypoxanthine, aminopterin, thymidine)-containing medium, as described in Reference 28. Culture supernatants o f hybrids were screened for anti-PLA2 activity using a standard solid-phase EIA. The mAbs were purified by protein A chromatography from supernatant o f doubly cloned subline, and were isotyped as an IgGlk. Monoclonal antibodies SK088-3C6.16.2 and SK097-1E8.5.2 were tested for cross-reactivity to inflammatory mediators as well as porcine pancreatic type I 14 kDa PLA2 using a standard solid-phase Elisa. Microtitre plates (Nunc, Naperville, IL, USA) were coated with the antigens specified in Table 1 at concentrations ranging from 0.1-1.0 lag/well/ 0.1 ml. Bound test antibodies and controls were detected by incubating the samples with peroxidase-conjugated specific anti-immunoglobulin (Boehringer M a n n h e i m Biochemicals, Indianapolis, IN, USA) and then developing for peroxidase activity using the chromogenic substrate ortho-phenylenediamine (Fluka, Ronkonkoma, NY, USA). Optical density was read on a Multiskan plate reader. None o f the mAbs demonstrated cross-reactivity with either type I pancreatic 14 kDa PLA2, 85-kDa PLA2 animal o r h u m a n albumin o r various inflammatory mediators, such as t u m o u r necrosis factor, interleukin 1, platelet-activating factor as assessed by competitive binding studies (Table 1).

EIA methodology Microtitre plates ( N u n c Immuno Plate Maxisorp F96, Roskilde, Denmark) were coated with m A b 3C6 (100 lal, 2 ~tg/ml) in 50 m m o l / L sodium phosphate, 150 m m o l / L NaCI, 0.02% NAN3, p H 7.4 for 18 h at 4°C. The plates were washed 4 times with wash buffer (containing 10 mmol/L Tris, 150 mmol/L NaCI, 0.02% NAN3, 0.05% Tween 20, p H 7.4). Non-specific sites were blocked with 200 lal 1% BSA in 50 m m o l / L Tris, 150 m m o l / LNaCI, 0.02% NAN3, p H 7.4 for 5-60 min at 37°C. Samples and the corresponding standards (50 lal) were diluted in assay buffer or appropriate vehicles representing the samples under analysis, e.g. culture medium (RPMI 1640, 5% FBS), or CF buffer (non-treated or acid extracted). Monoclonal l E8 was biotinylated using RP1234 biotinylation kit from Amersham (Arlington Heights, IL, USA). Samples o r standards were coincubated with 50 lal conjugate (2 lag/ml biotinylated m A b 1E8) in assay buffer (0.5% bovine ~,-globulin, 50 mmol/L Tris, 150 m m o l / L NaC1, 0.02% NAN3, 0.05% Tween 20) for 1 h at 37°C. The plates were washed 4 times with wash buffer, followed by the addition o f 100 lal/well o f the streptavidin-alkaline phosphatase conjugate (diluted 1:2000 in streptavidin buffer, 0.5% bovine ,/-globulin, 50 m m o l / L Tris, 150 m m o l / LNaC1, 0.02% NaN3, 1 mmol/L MgCI2, p H 7.4) and incubated for 30 min at 37°C. The plates were washed again followed by the addition o f 100 Ixl substrate/weU (p-nitrophenyl phosphate, 1 mg/ml) and incubated for 30 min at 37°C

Bolognese et al.

251

TABLE I Analysis for cross-reactivity to other related inflammatory proteins Monoclonal antibodies (neat supernatant) OD value

Assays Coating con.

SK097-1 E8

SK088-3C6

Human albumin Murine albumin Rabbit albumin Rat albumin

1 lag/well 1 lag/well 1 lag/well 1 lag/well

0.070 0.077 0.069 0.054

0.078 0.077 0.061 0.064

2.349 ND ND 2.028

TNF IL-Ict IL-113 PAF

1 lag/well 1 lag/well 1 lag/well 1 ~tg/well

0.071 0.074 0.072 0.079

0.083 0.066 0.063 0.061

2.152 4.000 4.000 ND

rHSF-PLA2 Porcine pancreatic PLA2 C. AtroxPLA2

0.1 lag/well 0. I lag/well 0.1 lag/well

2.617 0.068 0.063

1.025 0.072 0.068

2.801 4.000 ND

Antigen

Positive control

Positive control valueswereobtained by incubation ofantigenwith corresponding antibody directedto the specific antigen. ND =not determineddue to lack of availability of positive control antibodies

after which the plate was read at 405 nm using a MR7000 plate reader (Dynatech Laboratories Inc., Chantilly, VA, USA). Standard curves were generated from purified rh type II 14 kDa PLA2 solubilized in the appropriatevehicle and delivered as 50 Ixl at concentrations ranging from 0.1-4 ng/ml. Extrapolation from the standard curve analysis was used t o calculate unknown sample values using Delta Soft v2.12 (Biomettalics Inc., Princeton, NJ, USA). Data was expressed as ng or p g 14 kDa PLA2/ml or p.g protein.

Solubil&ation

o fcellular 14 kDa PLA2 by acid extraction

The 14 kDa PLA2 is known to resist acid treatment, which has been utilized as an initial purification step t o solubilize 14 kDa PLA2 and reduce overall total protein [2,8,9,29]. Cell fractions or h u m a n synovial fluid were subjected t o acid extraction by exposure to an equal volume of 0.36 N H2SO4 for 1 h at 4°C [8,29]. After treatment, sample p H was adjusted t o 7.4 by the addition of 2 m o l / L Tris (pH 10) and analysed immediately by EIA to measure type II 14 kDa PLA2 concentration.

252

Quantitation of Human Tissue and Immune CellType II 14kDa PLA2by EIA

Protein determination All protein concentrations were determined by Bradford [30] protein analysis kits (Biorad, Richmond, CA, USA).

Calculations and statistics D a t a is expressed as mean___standard deviation (SD) o f 3 determinations (n = 3) unless otherwise stated and subjected to Student's t-test All cell experiments were performed 2-6 times using cells obtained from 2-6 donors.

RESULTS A N D DISCUSSION Characterization o fthe EIA assay This sandwich EIA assay was shown to have the capability o f measuring soluble rh type II 14 kDa PLA2 in protein-containing s e r u m or inflammatory exudates at femtomolar concentrations. Figure 1A shows that rh type II 14 kDa PLA2 solubilized in assay buffer, phosphate-buffered saline (PBS) o r cell culture media (RPMI 1640 with 5% FBS) was detected over 0.1-4 ng/ml (7-280 fmol) in a linear fashion. The assay sensitivity was equal to or greater than those previously reported. For example, this EIA was 3000-fold more sensitive (7-280 fmol) compared with the standard curve values reported by Pruzanski et al. [16] (0.02-10 nmol) and similarto those o f Stoner et al. [31] (7-1750 fmol). As cell and tissue fractions are prepared in CF buffer, standard curves using rh type II 14 kDa PLA2 solubilized in CF buffer, with and without acid treatment, were evaluated to assess vehicle interference (Figure IB). Utilization o f 100% CF buffer alone had a negative effect on the performance o f the EIA when compared with the assay buffer standard curve (Figure 1A). The reason for this negative result is possibly the high sucrose concentration. Dilution o f the CF buffer (1:2) with w a t e r was required to produce standard curves comparable with that obtained using EIA assay buffer. CF buffer, subjected to acid extraction produced a standard curve comparable with using assay buffer (Figure 1B). Recovery o f standard from the biological matrix was conducted by the addition of rh type II 14 kDa PLA2 (1 ng/ml) to a dilute sample o fh u m a n synovial fluid. This was assayed by EIA with o r without acid extraction. Table 2 demonstrates a l m o s t total recovery o f exogenously added enzyme from the biological matrix. Acid extraction of rh type II 14 kDa PLA2 standard alone did not a l t e r measurement by EIA. The assay therefore provides a sensitive quantitative analysis of type II 14 kDa PLA2 with essentially 100% recovery in diluted synovial fluid samples.

Bolognese et al.

253 1.5

A .

1.0 O

ci ci

o.s

0 . C

.

.

.

0.0 2.0

.

'

0.5

.

.

.

.

'

.

1.0

.

.

.

'

.

.

.

1.5

.

'

.

.

.

2.0

.

'

.

.

.

2.5

.

'

.

3.0

.

.

.

"

3.5

. . . .

'

4.0

.

.

.

.

'

.

4.5

.

.

.

'

5.0

B.

~)_

C,~l fmc~ondon (CF) b~f~" add e x t n i c t ~ C F buffer 1:2 Diluted C F buffet

oE

1.5

~"

1.0

O 0.5 , m r

0.0 0.0

. . . . . . 0.5

1.0

1.5

"

'

2.0

"

'

2.5

"

. . . . . . . . . 3.0

3.5

4.0

4.5

5.0

rh 14kDa PLA2 (ng/ml) F i g u r e 1 . T h e effect of v a r i o u s a s s a y buffers o n the type I I 14 k D a PLA2 E I A s t a n d a r d curve. A s h o w s a typical s t a n d a r d c u r v e ( 0 . 1 ~ , n g / m l ) p r e p a r e d i n a s s a y b u f f e r ( O ) , P B S ( O ) , R P M I 1640 5 % F B S ( A ) . B s h o w s the response of the E I A when the s t a n d a r d c u r v e i s p r e p a r e d i n C F b u f f e r ( D ) , acid-extracted C F b u f f e r ((3) a n d C F buffer d i l u t e d 1:2 ( A ) TABLE 2 Recovery o f exogenously a d d e d r h type II 14 k D a PLA2 from a biological matrix

Sample SFA SFB SFC

Acid-extracted s a m p l e alone ( n g / 5 lag protein)

Acid-extracted s a m p l e plus 1 n g / m l PLA2 ( n g / 5 lag protein)

1.5 __+0.21 1.84-t-0.01 2.21 _+0.14

2.56-t-0.37 2.31 _+0.07 3.65_+0.32

Human synovial fluid was assayed at equal protein levels (5 pg/test) after acid extraction as described in Methods. In brief, synovial fluid samples from three different donors (A, B and C) were incubated with an equal volume of acid for 1 h at 4°C, then either water vehicle or exogenous rh type II 14 kDa PLA2 (equivalent t o I ng/ml/assay) was added. Samples were readjusted t o pH 7.4 by the addition of 2 mol/LTris and directly measured by EIA. Measurement of C F buffer alone yielded 0 + 0 ng while the C F buffer containing 1 ng/ml contained 1.06+0.01 ng/ml 14 kDa PLA2. Each data point represents m e a n + S D

(n=3)

254

Quantitation of Human Tissue and Immune CellType II 14 kDa PLA2 by EIA

TABLE 3 Measurement of immunoreactive type II 14 kDa PLA2 measured in human synovial fluid with and without acid extraction Donor

Non-extracted Acid-extracted (ltg/ml HSF)

p value

A B C

2.6±0.0 5.4±0.7 3.2±0.2

3.2±0.4 6.6±0.0 4.2±0.3

NS NS NS

Mean

3.8±1.3

4.7±1.6

NS

Human synovial fluid was assayed at equal protein levels (5 /ag/test) directly or after acid extraction as described in Methods. In brief, synovial fluid samples fromthree different donors were incubatedwith an equal volume of acid for 1 h at 4°C. Samples were readjusted to pH7.4by the addition of 2 mol/LTris and directly measured by EIA. Each data point represents the mean +SD (n= 3). Statistics were calculated using Student's t-test (NS= not significant at p <0.05)

Utilization o f EIA to measure synovialf l u i d type H 14 kDa PLA2 Others have successfully used sandwich EIAs to directly measure 14 kDa PLA2 mass in the SF of individuals suffering from rheumatoid arthritis [16,31]. In these cases, no modifications were performed on the samples. Pruzanski et al. [16] reported between 0.45 and 7.62 Ixg/ml 14 kDa PLA2 in h u m a n synovial fluid, testing undiluted and serially diluted samples, while Stoner et al. [31] reported 0.13--0.38 lag/ml in diluted samples. In our assays, the h u m a n synovial fluid samples were analysed at equal total protein concentrations (5 ktg protein). Samples required significant dilution (1:50 t o 1:90) t o allow the values t o fall on the linear (1--4 ng/ml) portion of the standard curve. Table 3 summarizes the type II 14 kDa PLA2 mass measured in the synovialfluid from three different patients, which ranged from 2.7 to 5.4 ~tg 14 kDa PLA2/ml synovial fluid. The levels measured in acid-extracted h u m a n synovial fluid (3.2-6.6 p,g/ml) were identical t o those in untreated SF samples. This is consistent with enzyme existing in soluble form in SF and agrees with published reports detailing no potentiation of measurable PLA2 activity in h u m a n synovial fluid after acid extraction [2]. However, it was observed that this procedure was important in reducing the viscosity of the sample.

Utilization o f EIA to measure cell-associated type H 14 kDa PLA2 Quantitation of type II 14 kDa PLA2 in h u m a n immune cells We have taken this assay one step further, applying it t o the measurement of type II 14 kDa PLA2 levels in the cellular fractions of h u m a n cells or tissues. We have previously

Bolognese et al.

255

TABLE 4 Immunoreactive type II 14 kDa PLA2 measured in human immune cell cellular fractions with and without acid extraction 100 000g particulate

Donor

Neu~ophilcellul~

Cytosol (pg/lag protein)

Non-extracted (pg/lig protein)

Acid extracted (pg/lig protein)

ND ND ND ND ND ND ND

I.I±0.I 0.9±0.0 NT NT NT NT 1.0±0.1

2.1±0.1 1.2±0.I 1.6±0.7 2.6±0.6 3.7±0.2 3.3±0.2 2.4±1.0

ND ND ND ND ND

NT NT NT NT NT

2.9±0.2 i.7±0.4 1.5±0.1 2.2±0.1 2.1±0.6

~actions

A B C D E F Mean

Mon~ecellularffacfions G H

I J Mean

Cellular fractionsfrom each donor were assayed (300 i~g/test) after being subjected to acid extraction as described in Methods. ND= levels belowlimit of detection; NT=not tested. Data represent mean +SD, n=3

reported that h u m a n neutrophil or monocyte particulatefractions contain acylhydrolytic activity possessing the biochemical and pharmacological characteristics of a 14 kDa PLA2. The localization of a type II 14 kDa PLA2 in resting or unstimulated neutrophils was demonstrated by electron microscopy t o be in intracellular granules [32]. Upon activation w i t h serum-opsonized Staphylococcus aureus, the labelling indicated substantial localization of 14 kDa PLA2 on or in phagolysosomes. Fractions of h u m a n cells were prepared and untreated cell fractions were compared w i t h those subjected to acid extraction as described in Methods. All samples were equalized for protein (100-300 ttg/sample) and CF buffer volume prior to the acid extraction procedure. Preliminary studies indicated that at least 300 ttg of total neutrophil particulate protein was required t o obtain values that would fall above the lower limits of the EIA assay standard curve. Table 4 shows the total immunoreactive type II 14 kDa PLA2 material measured from different h u m a n cell particulate fraction preparations. In two individuals, the values obtained from analysis of non-acid-treated neutrophil particulates range from 0.9 to 1.I pg/lag particulate fraction protein. This was approximately one half of that measured in the same sample after acid extraction.

256

Quantitation of Human Tissue and Immune CellType II 14kDa PLA2 by EIA

The mean type II 14 kDa PLA2 measured in neutrophil particulate fractions o f 6 donors was 2.6 + 0.8 pg/lag protein. The cytosolic fractions (300-500 lag total protein) were tested for type II 14 kDa PLA2 employing acid extraction and did not produce values that were above the confident detectable limits o f the assay. Rosenthal et al. [32] reported that acid-extracted h u m a n neutrophils contained 160 pg 14 kDa PLA2/107 cells as assessed by EIA which is in relative agreement with the amounts reported herein (338 pg 14 kDa PLA2/107 cells; equivalent to 370 lag particulate protein). The difference observed may be due, in p a r t , to the use o f cell fractions versus the use o f homogenates. Table 4 shows the total immunoreactive type II 14 kDa PLA2 material measured from monocytes particulate fraction preparations o f 4 different h u m a n donors. The fractions were acid extracted p r i o r to EIA analysis and contained 1.5-2.9 pg 14 kDa PLA2/lag protein with a mean concentration o f 2.1-t-0.59 pg/lag fraction. All fractions were analysed at 300 lag protein per fraction which is equivalent to 1.8 x 107 cells. The recovery o f exogenous h r type II 14 kDa PLA2 added as an aqueous solution to P M N particulate fraction p r i o r to acidification (1 ng/ml h r type II 14 kDa PLA2/ sample) was examined. The exogenous type II PLA2 (1 ng/ml) was totally recovered in these samples (acid extracted particulate, 1.1 ng/ml, vs. acid-extracted particulate + exogenous PLA2, 2.1 ng/ml, n = 3), demonstrating that acid extraction o f biological matrix does not appear to negatively effect the recovery o f and quantitation o f enzyme present. Marki and Franson [8] demonstrated that h u m a n neutrophil fractions contained neutral-active and Ca2+-dependent phospholipase A2 activity that was largely suppressed by endogenous inhibitors. In these studies, acid treatment significantly enhanced activity, presumably by removing potential acylhydrolase inhibitors, such as free fatty acids. Acid extraction breaks up membrane structure, to an extent freeing up the protein o r enzyme which might be physically bound to the membrane, resulting in a reported 40% increase in activity from h u m a n platelets prepared in the same m a n n e r [2]. Marshall and Roshak [9] reported a 400% increase in 14 kDa PLA2 activity measured in acid-extracted neutrophil 100000g particulate fractions when compared with non-extracted controls; also, these studies demonstrated the ability to measure 14 kDa-like activity in the cytosolic fraction as a result o f acid extraction while no activity was detected in non-extracted samples. These observed increases in activity, while greater than those observed in mass as a result o f this treatment, could be due in part to the differences and variations in interfering phenomena when activity versus mass is assessed.

Quantitation o f placenta type II 14 kDa PLA2 To evaluate the ability o f the EIA to measure the type II 14 kDa PLA2 in tissues, h u m a n placenta was homogenized and fractionated as described in Methods. Table 5 shows that no type II 14 kDa PLA2 was detected in the cytosolic fraction o f placenta, whereas the microsomal fraction contained measurable levels (24.6+6.9 ng/mg protein). Table 5 also demonstrates the linear detection o f type II 14 kDa PLA2 by

Bolognese et al.

TABLE 5 Immunoreactive fraction)

257

type II 14 k D a PLAz m e a s u r e d in h u m a n placenta (100000g particulate

Placenta fraction u t i l i z e d(p.g)

T y p e II 14 k D a PLA2 (ng/assay)

100 50 30 10 3 I Average n g / m g

2.0+0.09 1.0+0.01 0.8 __+0.03 0.2 + 0.02 0.1 __+0.01 ND

T y p e II 14 k D a PLA2 n g / m g fraction 20.0-t-0.9 18.9-t-0.3 24.9 -t- 0.8 24.6 -t- 2.4 37.0__ 1.4 ND 24.5 -t- 6.9

Cellular fractions of human placenta were prepared as described in Methods. Cytosolic samples assayed at 100 p.g protein and subjected t o acid extraction did not contain detectable levels of enzyme as assayed by EIA. T h e particulate fraction was acid extracted and assayed over a number of protein concentrations (1100 I~g). N D = levels below limit of detection. Data represent mean ± SD, n = 3

TABLE 6 Reproducibility

o f assay: s t a n d a r d c u r v e a n d s a m p l e analysis from four different experiments

Sample Standards 0 ng/ml 0.1 ng/ml 0.4 ng/ml 1 ng/ml 2 ng/ml 4ng/ml

Experiment 1 Mean O D i S D

Experiment2 Mean O D ± S D

Experiment3 Mean O D ± S D

Experiment4 Mean O D ± S D

0.199+0.01 0.321 ±0.01 0.737 ±0.03 0.859:1:0.03 1.254±0.03 2.45 ±0.04

0.231±0.01 0.346±0.01 0.723 +0.01 0.884±0.01 1.281- t - 0 . 0 4 2.45 ±0.13

0.199+0.02 0.324±0.01 0.733±0.03 0.857-t-0.03 1.262+0.03 2.45 ±0.04

0.27 ±0.01 0.444___0.01 0.781 +0.07 1.012±0.11 1.371 ±0.04 2.35 ±0.01

0.854-t-0.01

0.921 -t-0.05

0.855- t - 0 . 0 2

0.899___0.01

Placenta

50 lxg acid-extracted microsome

Cellular fractions of human placenta were prepared as described in Methods. T h e particulate fraction was acid extracted and assayed in 4 different EIA plates. T h e mean O D (405 nm)± SD oftbe samples as well as the standard curve are shown. Data represent mean + SD, n = 2

258

Quantitation of Human Tissue and Immune CellType II 14kDa PLA2 by EIA

examining increasing concentrations of acid-treated microsomal fractions. It is important to note that we began to obtain more variable data when samples were diluted to a point where the optical density (OD) measurement coincided with the standard values below 0.2 ng/ml. Finally, our results concur with those o f Aitken et al. [33] who reported that the h u m a n placenta contained 26.0+__7.0 ng PLA2/mg tissue. These values were obtained by homogenization o f tissue in 16-25-fold higher ionic strength media (1 m o l / LNaCI). This was shown to be better than when they extracted in low-ionic strength media (0.32 mol/L sucrose-10 m m o l / L Hepes). High salt compromises antigen-antibody interaction and therefore needs to be removed or samples greatly diluted before ELISA can be performed. Alternatively, we used a lowionic strength buffer (0.34 mol/L sucrose, 10 m m o l / L Helves, 1 mmol/L EGTA) to homogenize the tissue, then treated the homogenate with acid. Acid extraction and readjustment o f p H o f samples for 14 kDa PLA2 can be accomplished in smaller volumes with low salt which will not affect ELISA. The reproducibility o f the ELISA was examined using placenta particulate fractions. Human placenta was homogenized and fractionated as described in Methods and stored as aliquots a t -20°C. Four different aliquots were assayed for 14 kDa PLA2 mass following acid extraction in four different experiments. Table 6 demonstrates good agreement in the OD measurement o f the standard curve and o f the placenta samples between the four experiments.

SUMMARY Sandwich EIAs have been successfully used to measure 14 kDa PLA2 in s e r u m and exudate fluids. The utility o f this assay was expanded to quantify 14 kDa PLA2 in cellular o r tissue fractions. The type II 14 kDa PLA2 was predominantly found in the 100 000g particulate fraction o f all cells or tissue evaluated. Acid extraction p r i o r to analysis improved the measurement of enzyme by EIA, presumably through its solubilization from the membranes in a fashion that can be readily analysed by ELISA with n o further processing. The EIA provides a more sensitive and convenient way to quantitate this enzyme. Taken together, EIA utilized in conjunction with sample acid extraction p r i o r to analysis provides a method to identify and quantify this enzyme in cells and whole tissues. This may be particularly useful in continuing studies o f the role o f type II 14 kDa PLA2 in inflammatory disorders.

ACKNOWLEDGEMENTS W e would like to t h a n k Amy Roshak and Drs James Winkler and Jeffrey Jackson for their help in reviewing this manuscript. W e also t h a n k Frances Bolognese for placenta tissue donation.

Bolognese et al.

259

REFERENCES 1. Seilhamer JJ, Pruzanski W, Vadas P e t al. Cloning and recombinant expression of phospholipase A2 present in rheumatoid arthritic synovial fluid. J Biol Chem. 1989;264:5335-8. 2 . Kramer RM, Hession C, Johansen B e t al. Structure and properties of a human non-pancreatic phospholipase A2. J Biol Chem. 1989;264:5768-75. 3 . Vadas P, Scott K, Smith G e t al. Serum phospholipase A2 enzyme activity and immunoreactivity in a prospective analysis of patients with septic shock. Life Sci. 1992;50:807-11. 4 . Bolognese B, McCord M , Marshall LA. Differential regulation of elicited-peritoneal macrophage 14 kDa and 85 kDa phospholipase A2(s) by transforming growth factor-13. Biochim Biophys Acta. 1995;1256:201-9. 5 . Kanda A, Ono T , Yoshida M . T h e primary structure of a membrane-associated phospholipase A2 from human spleen. Biochem Biophys Res Commun. 1989; 163:42-8. 6 . Gailbraith W , Paschetto KA, StevensT M , Kerr JS. PhospholipaseA2 activityin undifferentiated U937 cells. Agents Actions. 1989;27:422--4. 7 . Roshak A, Sathe G, Marshall LA. Suppression of monocyte 85kDa phospholipase Az by antisense and effects on endotoxin-induced prostaglandin biosynthesis. J Biol Chem. 1994;269:25999-6005. 8 . Marki F, Franson R. Endogenous suppression of neutral-active and calcium-dependent phospholipase A2 in human polymorphonuclear leukocytes. Biochem BiophysActa. 1986;87:149-56. 9 . Marshall LA, Roshak A. Coexistence of two biocbemically distinct phospholipase A2 activities in human platelet, monocyte and neutrophil. Biochem Cell Biol. 1993;71:331-9. 10. Leslie CC, Voelker DR, Channon JY, Wall M M , Zelarney PT. Properties and purification of an arachidonyl-hydrolyzingphospholipase A2 from a macrophage cell line, RAW 264.7. Biochem Biophys Acta. 1988;963:476-92. 11. Diez E, Mong S. Purification ofa phospholipase A2 from human monocyte leukemic U937 cells. J Biol Chem. 1990;265:14654q51. 12. Clark JD, Milona N, Knopf JL. Purification of a 110-kilodalton cytosolic phospholipase A2 from the human monocytic cell line U937. Proc Natl Acad Sci USA. 1990;87:7708-12. 13. Clark JD, Lin L, Kriz RW et al. A novel arachidonic acid-selective cytosolic PLA2 contains C a2+dependent translocation domain with homology t o PKC and GAP. Cell. 1991 ;65:1043-51. 14. Kramer RM, Roberts EF, Manetta J, Putnam JE. T h e Ca:+-sensitive cytosolic phospholipase A2 is a 100-kDa protein in human monoblast U937 cells. J Biol Chem. 1991 ;266:5268-72. 15. Green JA, Smith G, Buchta R et al. Circulating pbospholipase A2 activity associated with sepsis and septic shock is indistinguishable from that associated with rheumatoid arthritis. Inflammation. 1991;15:355~7. 16. Pruzanski W , Scott K, Smith G, Rajkovic I, Stefanski E, Vadas P. Enzymatic activity and immunoreactivity of extracellular phospholipase A2 in inflammatory synovial fluids. Inflammation. 1992;16:451-7. 17. Bolognese B, Marshall LA. Use of radiolabeled E. coli as a substrate t o assess phospholipase activity. DuPont NEN Biotech Update, vol. 8 , number 3 , Fall 1993. 18. Pruzanski W , Albin-Cook K, Laxer R M et al. Phospholipase A2 in juvenile rheumatoid arthritis: Correlation t o disease type and activity. J Rheum. 1994;21(10):1951~1. 19. Barbour SE, Dennis EA. Antisense inhibition of group II phospholipase A2 expression blocks the production of prostaglandin E2 by P388DI cells. J Biol Chem. 1993;268:21875-82. 2 0 . Lister MD, Glaser KB, Ulevitch RJ, Dennis EA. Inhibition studies on the membrane-associated pbospholipase A2 in vitro and prostaglandin E2 production in vivo of the macrophage-like P388Dr cell. J Biol Chem. 1989;264:8520-8. 2 1 . Marshall LA, Mayer RJ. Phospholipase A2 isoforms as novel drug targets. IN: Ruttolo Jr R R , Hollinger MA, eds. Inflammation; Mediators and Pathways. New York: C R C Press; 1995:1-22. 2 2 . Hulkower KI, Werheimer SJ, Levin Wet al. lnterleukin-ll~ induces cytosolic phospholipase A2 and prostaglandin H synthase in rheumatoid synovialfibroblasts: Evidence for their roles in the production of prostaglandin E2. Arthritis Rheum. 1994;37:653-61. 2 3 . McCord ME, Bolognese B, Marshall LA. Prostaglandin E2 is required for transforming growth factor I]1 inhibition of elicited macropbage 14-kDa phospholipase A2 release. Br J Pharm. 1997; 116:2575 -8I. 2 4 . Stadel JM, Jones C, Livi G e t al. Recombinant human secretory phospholipase Az: Purification and characterization of the enzymefor active site studies. J M o l Recog. 1993;5:145-53. 2 5 . Marshall LA, Winkler JD, Griswold DE et al. Effects of Scalaradial, a type II phospholipase A2 inhibitor, on human neutrophil arachidonic acid mobilization and lipid mediator formation. J Pharm Exp Ther. 1994;268:709-17.

260

Quantitation of Human Tissue and Immune Cell Type II 14 kDa PLA2 by EIA

2 6 . Channon JY, Leslie CC. A calcium dependent mechanism for associating a soluble arachidonylhydrolyzing phospholipase A2 with membrane in the macrophage cell line RAW 264.7. J Biol Chem. 1990;265:5409-13. 2 7 . Marshall LA, Bolognese B, Roshak A. Characterization of phospholipase A2 release by elicitedperitoneal macrophage and its relationship t o eicosanoid production. J Lipid Med. 1994;10:295-313. 2 8 . McKeam TJ, Fitch F W , Smiek DE, Samiento M , Stuart F P . Properties of rat anti-MHC antibodies produced by clones rat-mouse hybridomas. Im munol Rev. 1979;47:91-115. 29. Marshall LA, Bauer J, Sung ML, Chang JY. Evaluation of antirheumatic d r u g s for their effect in vitro on purified human synovial fluid phospholipase A2. J Rheumatol. 1991;18:59--65. 30. Bradford M M . A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976;71:248-54. 31. Stoner CR, Reik L, Donohue M , Levin W , C r o w l RM. H u m a n group II phospholipase A2: Characterization of monoelonal antibodies and immunoehemical quantitation of the protein in synovial fluid. J Immunol Meth. 1991;145:127-136. 32. Rosenthal MD, Gordon MN, Buescher ES, Slusser JH, Harris LK, Franson RC. Human neutrophils store type II 14kDa phospholipase A2 in granules and secrete active enzyme in response t o soluble stimuli. Biochem Biophys Res Commun. 1995;208(2):65ff6. 33. Aitken MA, Farrugia W, W o n g MH, Scott KF, Brennecke SP, Rice GE. Type II phospholipase A2 in human gestational tissues: extractable immuno- and enzymatic activity in fetal membranes. Biochem BiophysActa. 1993;1170:314-20. Manuscript received21 Feb. 9 7 . Accepted for publication 28 May 9 7 .