Proc. Natl. Acad. Sci. USAVol. 90, pp. 5818-5822, June 1993Immunology

Identification and functional characterization of platelet-activatingfactor receptors in human leukocyte populations using polyclonalanti-peptide antibodyELIANE MULLER*, PIERRE DAGENAISt, NAIMA ALAMI*, AND MAREK ROLA-PLESZCZYNSKI***Immunology Division, Department of Pediatrics, and tDepartment of Pharmacology, Faculty of Medicine, Universite de Sherbrooke, Sherbrooke, PQ,Canada, J1H 5N4

Communicated by Henry Metzger, February 5, 1993 (received for review October 21, 1992)

ABSTRACT Recently, the successful cloning of a receptorfor platelet-activating factor (PAF), a lipid mediator of inflam-mation, was reported. Here we investigated the distributionand potential diversity of human PAF receptors (hPAF-Rs)among individual leukocyte populations by (i) hPAF-R mRNAtranscription studies and (i") analysis of cell surface expressionof hPAF-R protein using a polyclonal anti-peptide antibody(anti-hPAF-R1"-173). Northern blot analysis, flow cytometry,and immunoblotting with anti-hPAF-R antibody indicated thatmonocytic, neutrophilic, and B-lymphocytic cell lines allshared a similar hPAF-R species, whereas resting T-cell andnatural killer cell lines failed to express detectable levels ofeither hPAF-R protein or mRNA. Peripheral blood leukocytepopulations showed a distribution of hPAF-R cell surfaceexpression similar to that of the corresponding cell lines.Furthermore, binding of anti-hPAF-R'"-173 antiserum, puri-fied IgG, or Fab and F(ab')2 fragments to the receptor of allinvestigated PAF-R-positive cell lines induced an increase inintracellular free calcium concentration. The characterizationof the expression of a lipid ligand receptor using antibodiesagainst an intrinsic portion of the receptor protein has, to ourknowledge, never been reported previously.

Platelet-activating factor (PAF) is a potent phospholipidautacoid with diverse and important physiological and patho-logical effects (1). It is produced by several cell types,including monocytes, neutrophils, basophils, eosinophils,mast cells, lymphocytes, platelets, and endothelium (2).Many of these cells that produce PAF as well as other celltypes can become targets ofPAF bioactions (for review, seerefs. 3 and 4). Most of the effects of PAF can be blocked byvarious structurally unrelated PAF-receptor (PAF-R) antag-onists, indicating that PAF acts through specific receptors.

In 1991, Honda et al. (5) successfully cloned and expresseda guinea pig lung PAF-R; they showed it to be a new memberof the G-protein-coupled receptor family with seven putativetransmembrane domains. Subsequently, the same group (6),as well as others (7, 8), have cloned the cDNA for a humanPAF-R (hPAF-R) from human leukocyte or myeloid cell linecDNA libraries. The hPAF-R is a 342-amino acid protein(estimated Mr 39,203) having a potential extracellular N-gly-cosylation site at residues 169-171.From binding studies in various cell types, apparent Kd

values of PAF-R were usually found between 0.1 nM and 10nM. However, rank order ofpotency for several antagonists,sensitivity to various cations, as well as signal transductionmechanisms were found to differ among cell populations,suggesting the existence of more than one hPAF-R subtype(9, 10). Whether an eventual diversity of PAF-R was never-theless associated with common features of one PAF-R

subtype shared among various cell populations remainedunknown. To address this question, we isolated a hPAF-RcDNA and compared different, representative leukocytic celllines for the mRNA expression pattern of this particularhPAF-R species. In addition, we developed polyclonal anti-bodies (Ab) to hPAF-R in order to investigate the cell surfaceexpression of the cloned hPAF-R. In this report, we presentevidence that human monocytes, neutrophils, platelets, andB cells express a shared hPAF-R mRNA species and cellsurface hPAF-R protein. We also describe the successfulproduction of specific antibody to a lipid ligand receptor-namely, the human PAF-R.

MATERIALS AND METHODSReagents. PAF (1-O-hexadecyl-2-acetyl-sn-glycero-3-

phosphocholine) was from Bachem and N-formyl-L-methionyl-L-leucyl-L-phenylalanine (fMLP) was from Sigma.WEB 2170 and WEB 2086 were gifts from H. Heuer (Boeh-ringer Mannheim). mAb to CD3, CD20, and CD14 were fromBecton/Dickinson.

Cell Cultures. U937, THP-1, HL-60, Raji, and Molt 4 cells(American Type Culture Collection) as well as Jurkat E 6.1(gift of G. Dupuis, University of Sherbrooke), Jurkat E77-6.8(gift of A. Weiss, University of California, San Francisco),and YT cells (gift of J. M. Castracane, Endogen, Boston)were cultured as recommended by the suppliers.

Isolation ofPeripheral Blood Leukocytes. Human peripheralblood leukocytes from healthy volunteers were isolated bydextran sedimentation and subsequent Ficoll/Hypaque cen-trifugation according to Boyum (11). Monocytes were iso-lated as described (12).hPAF-R cDNA Isolation and Northern Blot Analysis. The

895-bp Sma I fragment of the guinea pig PAF-R cDNA (ref.5; gift ofT. Shimizu, University ofTokyo) was used to screena AgtlO cDNA library prepared from human peripheral bloodmononuclear leukocytes (gift from S. Anderson, NationalInstitutes of Health/National Cancer Institute, Frederick,MD) under low stringency conditions. A 700-bp cDNAfragment that contained the translational start codon of thehPAF-R was isolated. Its sequence was identical to that ofseveral hPAF-R cDNAs isolated recently (6-8). Northernblot hybridization was performed at 68°C under conditionsdescribed previously (13). Blots were rehybridized with a1-kb Pst I cDNA probe (American Type Culture Collection)

Abbreviations: KLH, keyhole limpet hemocyanin; PAF, platelet-activating factor; PAF-R, PAF receptor; hPAF-R, human PAF-R;Ab, antibody(ies); mAb, monoclonal Ab; fMLP, N-formyl-L-methionyl-L-leucyl-L-phenylalanine; [Ca2+]i, intracellular free cal-cium concentration; GAPDH, glyceraldehyde phosphate dehydro-genase; CGRP, chorionic gonadotropin-releasing peptide; NK, nat-ural killer; FITC, fluorescein isothiocyanate.ITo whom reprint requests should be addressed.

5818

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Janu

ary

17, 2

022

Proc. Natl. Acad. Sci. USA 90 (1993) 5819

of the "housekeeping" gene glyceraldehyde phosphate de-hydrogenase (GAPDH).

Polyclonal Antipeptide Ab Preparation. A synthetic pep-tide, corresponding to the hPAF-R residues 164-173, waslinked (Sheldon Biotechnology, Montreal) to keyhole limpethemocyanin (KLH) for immunization or to bovine serumalbumin (BSA) for ELISA tests. This segment has little(<10%) homology with sequences of other relevant humanG-protein-coupled receptor sequences. New Zealand Whiterabbits were immunized s.c. with 0.5 mg of the peptide-KLHconjugate in complete Freund adjuvant. Booster injectionswere given at 4-week intervals in incomplete Freund adju-vant. Antigen recognition by anti-PAF-R164-173 Ab was testedby ELISA using horseradish peroxidase-conjugated goatanti-rabbit Ab. Ninety-six-well plates were coated with thepeptide (10 ,ug/ml) coupled to BSA or, as a control, with BSAalone. Nonspecific binding was blocked by 1% skim milk in0.01 M phosphate-buffered saline (PBS). Purification of theIgG fraction was achieved by affinity chromatography of theantiserum or preimmune serum of the same rabbit on HiTrapprotein G columns (Pharmacia). F(ab')2 or Fab fragmentswere generated by proteQlytic digestion of the IgG fractionusing the Immuno Pure F(ab')2/Fab preparation kit (Pierce).Contaminating IgG and entire or partially digested Fc frag-ments were removed from the preparations by protein Aaffinity purification (Pierce). Small Fc fragments were dis-carded by dialysis against PBS at pH 7.4 (Mr cutoff 30,000).FACScan Analysis and Immunoblotting. Cell surface ex-

pression of hPAF-R on leukocytes and leukocytic cell lineswas assessed by flow cytometry as described earlier (14).Western blots were performed according to Coffey et al. (15).

Intracellular Free Calcium Concentration [Ca2+]J Mobiliza-tion. [Ca2+], was determined as described earlier (13).

RESULTS AND DISCUSSIONhPAF-R Transcription in Leukocytic Cell Lines. Northern

blots with RNA isolated from the monocytic cell line U937,the promyelocytic cell line HL-60, differentiated into gran-ulocytes, as well as four lymphocytic cell lines-namely, Raji(B cells), Jurkat E 6.1 and Jurkat E 77-6.8 (T cells), and YT[natural killer (NK) cells]-were hybridized with the entirehPAF-R cDNA (Fig. 1). Monocytic, neutrophilic, and B-lym-phocytic cell lines all expressed a similar hPAF-R mRNAspecies of -4 kb. The T-cell and NK cell lines, which weanalyzed, did not transcribe detectable levels of this mRNA.hPAF-R Expression at the Cell Surface of Leukocytic Cell

Lines and Freshly Isolated Human Leukocytes. To character-

ize the expression of PAF-R at the protein level, we devel-oped polyclonal Ab against a decapeptide, hPAF-R164-173,corresponding to a segment in the second extracellular loopof the putative ternary structure of the hPAF-R. Antigenspecificity of the Ab was tested by competitive ELISA.Preincubation of the antiserum in the range of the half-maximal concentration (EC50 = 1:25,000) with increasingamounts of the free peptide progressively diminished theamount of bound Ab. Competition with 1 mM of the peptidedisplaced the antigen-antibody binding by 70%. Preincuba-tion of the antiserum with unrelated peptides as well as withthe carrier protein KLH had no effect (data not shown).The anti-hPAF-R164-173 Ab recognized the native receptor

at the surface of whole cells, as demonstrated by flowcytometry of U937 cells (Fig. 2). The significant labeling ofU937 cells (I) was not blocked by the PAF-R antagonist WEB2170 (II); it was, however, competitively inhibited by 1 mMfree hPAF-R164-173 peptide (III) but was not affected byunrelated peptides (IV). The anti-hPAF-R164-173 Ab, recog-nizing the native hPAF-R on the cell surface with highspecificity, provided the necessary tool for analyzing thehPAF-R expression on various cell types.The cell lines, tested previously for hPAF-R gene tran-

scription, were subjected to FACScan analysis (Fig. 3). Themonocytic cell lines U937 and THP-1, as well as the neutro-philic cell line HL-60, were all anti-hPAF-R Ab-positive.Although undifferentiated HL-60 cells transcribed only verylow levels ofPAF-R mRNA (compare Fig. 1), they expressedsome PAF-R at their cell surface. The B-cell line Raji was alsoanti-hPAF-R Ab-positive, whereas no significant labelingwas observed in the NK cell line YT or in the T-cell line Molt4. The T-cell line Jurkat was labeled at a very low level. Theselatter findings are consistent with recent observations thatindicate that T lymphocytes do not express PAF-R unlessthey are properly activated (16). Partial activation could bethe reason for the low level of surface labeling detected in the

100O 10(1hI.

1( 101 102 10' 10( 10 10 101FL11 il.

250

cell lines: Monocvtic

1937

Neutrophilic

IIL-60 (1.3% DMSO)

day: 0 5

28Sx9 28S_> _

t) -11()"

18S -

FIG. 1. hPAF-R expression in leukocytic cell lines. Northernblots with 4 Hg of poly(A)+ RNA isolated from U937 cells, 20 pg oftotal RNA isolated from undifferentiated and 1.3% (DMSO)-differ-entiated HL-60 cells, and 6 pg ofpoly(A)+ RNA isolated from Jurkat,YT, and Raji cell lines were hybridized with the hPAF-R cDNA.Amount and integrity of mRNA were probed with GAPDH cDNA.

10, 102 10) 1()0 10o 10

FLI F Ll1

FIG. 2. Specificity of polyclonal Ab. Labeling of U937 cells withanti-hPAF-R164 173 Ab. Shaded areas represent fluorescein isothio-cyanate (FITC)-conjugated control antibody. Conditions were asfollows: (I) labeling of untreated U937 cells with anti-hPAF-R164 173Ab at the half-maximal concentration of 1:75,000 (. ) or 1:600

); (II) U937 cells, untreated ( ) or pretreated (. ) with30 ,uM WEB 2170, 5 min prior to incubation with Ab 1:3000; (III)competition assay at half-maximal concentration of the antiserumpreincubated alone ( ) or with 1 mM free hPAF-RIl64173 peptide

) 30 min prior to addition to the cells; (IV) competition assayat half-maximal concentration of the antiserum (. ) preincubatedwith 1 mM unrelated peptides [angiotensin II (- ) or CGRP( )] or with 660 Hg of carrier-protein KLH per ml (----).

Lymphocytic

Immunology: MiiUer et al.

1.r-

0J96 .

= >. m

5

1-14

18S -

Dow

nloa

ded

by g

uest

on

Janu

ary

17, 2

022

Proc. Natl. Acad. Sci. USA 90 (1993)

Monocytic cell lines100

U937

100 101 102 103FL1

70-THP-1

0100 101 102 103

FL1

Monocytes

104

103

- 102

101

100 101 102 103 104 10' 101 102 10' 104

Neutrophilic cell line

104 -

_" 10

101

FL1 FL1

21

Neutrophils

-1 1- l04 -

' 103

Ut21lL4

1-- 5.6.. .97 .1

100 101 102 103 1043h

1 ~~~901

l . ... .

l1o 101 102 103 104

Lymphocytic cell lines

FL1 FL1

o0-e100 101 102 10' 100 10' 10 103

FL1 FL1

FIG. 3. FACScan analyses of leukocytic cell lines. The labelingofdifferent leukocytic cell lines by the hPAF-R Ab was tested by flowcytometry. Dotted curves represent FITC-conjugated control anti-body. The hPAF-R Ab dilution was 1:600.

Jurkat T-cell line. Previous reports ofPAF effects on T cells(e.g., ref. 17) may have been linked to the different activa-tional states of these cells or to interactions with otherhPAF-R subtypes.To correlate the hPAF-R expression on the cell surface of

peripheral blood leukocytes with the results obtained for thecorresponding cell lines, we isolated leukocytic subpopula-tions and analyzed them by two-color immunofluorescence(Fig. 4). Monocytes and neutrophils were labeled with themonocyte/neutrophil-specific anti-CD14 Ab as well as withthe anti-hPAF-R antiserum. When compared with the control(FITC alone), almost all CD14+ monocytes and CD14+neutrophils showed the presence of hPAF-R on the plasmamembrane. In addition, purified platelets analyzed by single-color fluorescence were also anti-hPAF-R Ab-positive. Wefurther analyzed a mixed population of monocyte-depletedlymphocytes, made up of T, B, and NK cells. Most CD3+cells, representing T cells, were negative for the anti-hPAF-RAb. In contrast, most of the B cells, stained with anti-CD20Ab, were hPAF-R Ab-positive. Similar results were obtainedusing purified hPAF-R IgG (data not shown).

3 10 .11. 10 4.510( 101 102 103 1

N 102I.

-1 1..I

100l 101 102 103 104

LymphocytcsT cells

- 104

103

101

10

L04 10(

LymphocytesB cclls

~104

ln

102-

10) 1(1' 10' 103 104FL1 (FITC)

85s] 11

Y1211 10

)t) 101 102 103 104

10

.... .

101

100 101 102 10' 104FLI (anti-hPAF-R Ab)

FiG. 4. Two-color FACScan analyses. Leukocytic subpopula-tions were incubated either with the FITC-conjugated goat anti-rabbit Ab alone (Left) or with the anti-hPAF-R Ab revealed byFITC-conjugated Ab (The Jackson Laboratory) (FL1) and a secondcell-specific phycoerythrin-conjugated Ab (FL2) (Right). Cells la-beled by both fluorochrome-conjugated Ab are found in the second(upper right) quadrant. Percentages of the labeled cells are indicatedby numbers in the corner of each quadrant.

hPAF-R Membrane Proteins. The size of the hPAF-Rprotein, present on the cell surface of PAF-R-positive cells,was investigated by immunoblotting of cell lysates from U937and Raji cell lines and compared to lysates from the Molt 4

Pla telcts

5820 Immunology: Miiller et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

17, 2

022

Proc. Natl. Acad. Sci. USA 90 (1993) 5821

A e0

69 kDa-106.0 kDa--80.0 kDt-

-49.5 kDa-

-32.5 kDa--27.5 kDa-

FIG. 5. Immunoblotting. Western blots with 30 Ag of totalproteins from lysates of Molt 4, Raji, and U937 cells were incubatedwith the anti-hPAF-R antiserum at a dilution of 1:250 (A) or thepreimmune serum at a dilution of 1:250 (B). The blots were revealedwith 1251-labeled protein A and exposed for 40 h.

cell line (Fig. 5). A single protein band of -69 kDa waslabeled by the antiserum but not the preimmune serum inlysates from U937 and Raji cells. The discrepancy betweenthe calculated mass of the hPAF-R of -39 kDa (6) and theprotein detected here may be due to the predicted glycosyl-ation in the second extracellular loop (8). Moreover, the sizeof the labeled protein corresponded roughly to that of a65-kDa PAF-binding protein isolated from human platelets byShen et al. (18).

[Ca2+] Mobilization by Anti-hPAF-R'"-173 Ab. To analyzewhether anti-hPAF-R Ab were able to compete with theligand PAF in [Ca2+], mobilization experiments, we stimu-lated differentiated HL-60 cells with PAF after treating thecells with and without the antiserum. In the absence ofanti-hPAF-R Ab, the addition ofPAF resulted in an increasein [Ca2+ij as previously demonstrated in other cell systems(ref. 3; Fig. 6). This effect was not blocked by the presenceof anti-hPAF-R Ab. However, the antiserum itself, at a

dilution of 1:200, induced [Ca2]1i mobilization (Fig. 6A).Furthermore, undifferentiated HL-60 cells, which failed torespond to PAF itself, also failed to respond to the antiserum(Fig. 6B). Additionally, anti-hPAF-R Ab induced a rise in[Ca2+]i in the hPAF-R Ab-positive U937 cells (data notshown) as well as in the Raji cell line (Fig. 6C). The Jurkatand YT cell lines responded neither to PAF nor to theantiserum (data not shown).To discriminate between [Ca2+], fluxes induced by the

hPAF-R Ab or by potentially bioactive lipids present in therabbit antiserum, the IgG fraction obtained by affinity puri-fication of the serum was tested for its ability to mobilizeintracellular calcium. The IgG fraction alone induced a[Ca2+], flux in U937 cells (Fig. 6D) that was inhibited uponaddition of the PAF-specific antagonist WEB 2170 (Fig. 6E).A marginal effect was observed with the purified IgG fractionof the preimmune serum (Fig. 6F). Since WEB 2170 blockedthe Ca2+ effect of the hPAF-R Ab without, however, affect-ing Ab binding to the cell (see Fig. 2 II), the rise in [Ca2+]seemed to be mainly mediated and induced by the PAF-R.The ability of receptor-specific Ab to mimic the effects of

the natural ligand and induce [Ca2+], fluxes is not a rarefinding. It was shown, for example, for the CD3/T-cell recep-tor complex that Ab binding induces a perturbation of themembrane receptors that triggers an increase in inositoltrisphosphates and induces the release of Ca2+ from intracel-lular stores (19). Although the anti-hPAFR164 173 Ab couldbind to the target receptor in the presence ofWEB 2170 (seeFig. 2 II), it was not able to induce [Ca2+]i fluxes under theseconditions. This suggests that WEB 2170 may have blocked achange in conformation induced by Ab binding to the PAF-Rmolecule (without affecting the binding itself) and therebyinhibited subsequent antibody-mediated [Ca2+]i influx.We further investigated whether desensitization of the

receptor by PAF resulted in the loss of responsiveness to theanti-hPAF-R Ab. This as well as the inverse situation-namely, the loss of responsiveness to PAF after desensiti-

600

500

400

300

200

100

0t

500

400

+^ 300

2UU b200%f

10u '

600

500

400

300

200

100

00

A ~~~~~HL-60day 6

t Ianti-hPAF-R Ab PAF. fMtLP

100 200 300 400 500

Raji 800

600]

I Ianti-hPAF-R AbPAF

400

200

0-

100 200 300 0I0

800 E U937

600 - (IgG)

400 WEB2170

200 t

PAF tanti-hPAF-R IgG0 10 100 200 300 400 D0 0

Time, sec

B HL-60day 0

_I -

anti-hPAF-R Ab PAF fMLPI 100 200

D U937(IgG)

FIG. 6. [Ca2+]1i mobilizationinduced by PAF or anti-hPAF-

t J R164173 Ab. Differentiated [1.3%tPAF anti-hPAF-R IgG tfMLP dimethyl sulfoxide (A)] or undif-

200 ~~~~~~ferentiated (B) HL-60 cells as well100 200 300 400 500 as the B-cell line Raji (C) werestimulated with 10 nM PAF or the

F & U937 anti-hPAF-R antiserum at a dilu-(IgG) tion of 1:200. Control stimulations

were done with 1 uM fMLP or 1jiM calcium ionophore A23187.U937 cells were stimulated with 90ng of anti-hPAF IgG per ml (D) inthe absence or presence of 10 ,uM

PI-IgG WEB 2170 (E) or with 90 ng ofPI-IgGIfMLP ~preummune serum IgG per ml [PI-100 200 300 400 500 IgG (F)]. fMLP was used at 10 nM

and 1 ,uM, respectively.

Immunology: Miiller et al.

lJi

Dow

nloa

ded

by g

uest

on

Janu

ary

17, 2

022

Proc. Natl. Acad. Sci. USA 90 (1993)

800

600

400

200

C

0-cMC)

100 sec

t t t tPAF fMLP

300 anti-hPAF-Ri 4-173 IgG

200

100

anti-hPAF-R164t73 IgG

C 800

600 -

400

PAF IMLP

Papain digestion-mix epsin digestion-mix(4h) (t 2h)

Fab/H

Fab

c

-.=. anti-hPAF-F

cmi800 - F(ab ) 2

600 .

400].

anti-hPAF-R

sec

R 64-173 IgG anti-hPAF-R164-173 Fab

164-173 G anti-hPAF-R 164-173F(ab)

-f\-

fMLP

fMLP

FIG. 7. Specificity of the anti-hPAF-R Abl63-174-induced [Ca2+jmobilization. (A) The PAF-Rs on U937 cells were desensitized witheither PAF (10 nM) or anti-hPAF-R IgG (165 ng/ml). Subsequently,the same cells were stimulated with anti-hPAF-R IgG (165 ng/ml) orPAF (10 nM), respectively. (B) Proteolytic digestion of anti-hPAF-RIgG with papain (Fab) or pepsin [F(ab')21. At the end of incubation(digestion-mix) and after affinity purification and dialysis, these frac-tions were tested by SDS/PAGE under reducing conditions and silverstained. The middle lanes [protein A-purified Fab and F(ab')2 frac-tions] were intentionally overloaded to monitor the presence ofcontaminating Fc fragments. H, monomeric heavy chains, originatingfrom reduced, undigested IgG; 1/2Fc, monomeric half of the Fcfragments generated by reduction after papain digestion; Fab/H,monomeric half of the reduced Fab fragments; Fab'/H, monomerichalfofthe reduced Fab' fragments. (C) Identical batches ofU937 cellswere stimulated with either anti-hPAF-R IgG (330 ng/ml) or purifiedFab fragments (100 ng/ml). A second batch of cells was stimulatedwith anti-hPAF-R IgG (165 ng/ml) or its purified F(ab')2 fragments(380 ng/ml). Control stimulations were done with 1 ,uM fMLP.

zation with the anti-hPAF-R Ab-occurred, as demonstratedfor U937 cells in Fig. 7A.

Finally, to further analyze the underlying mechanism oftheAb-induced [Ca2+]i fluxes, we generated F(ab')2 and Fab

fragments of anti-hPAF-R IgG. The integrity as well as thepurity of the Fab and F(ab')2 fractions were monitored bySDS/PAGE (Fig. 7B). When Fab fragments were used tostimulate U937 cells, these univalent immunoglobulins were,on a molar basis, as effective in calcium mobilization as thecorresponding amount ofentire IgG (Fig. 7C). This effect wasnot amplified by subsequent bridging of these fragments witheither goat anti-rabbit F(ab')2 fragments or whole goat anti-rabbit IgG (data not shown). However, bivalent F(ab')2fragments, which were also able to induce a rise in [Ca2+]i,were reproducibly about four to five times less immunore-active than their corresponding IgG fraction. It is possiblethat the pepsin digestion caused some damage to the antigenrecognition site, thereby reducing its binding efficiency.Since monovalent Fab fragments of anti hPAF-R IgG werequite effective in stimulating intracellular calcium influx bythemselves, dimerization of hPAF-R by bivalent receptor-specific Ab, such as, for example, reported for IgE receptorson mast cells (20), seemed therefore not necessary to triggera rise in [Ca2+]i.

In conclusion, the potential interest of the present work is3-fold: (i) it describes the successful production of Ab to aG-protein-coupled receptor, a hitherto difficult task; (ii) itcharacterizes the phenotypic and functional hPAF-R expres-sion on leukocyte populations and correlates it with hPAF-Rgene transcription; (iii) it provides direct evidence of signaltransduction mediated through the hPAF-R protein. Thesestudies thus provide new venues for studying the expressionand function of hPAF-R.

We thank Drs. Jana Stankova, Claire Dubois, and Gilles Dupuisfor helpful suggestions and discussions, Mrs. Sylvie Turcotte and M.Denis Gingras for technical assistance, and Mrs. Carole Jacques forsecretarial help. This work was supported by a postdoctoral fellow-ship (E.M.) from the Swiss National Foundation, a studentship(N.A.) from the Canadian Intemational Development Agency, and agrant (MA-6460) from the Medical Research Council of Canada.

1. Braquet, P., Touqui, L., Shen, T. Y. & Vargaftig, B. B. (1987)Pharmacol. Rev. 39, 97-145.

2. Braquet, P. & Rola-Pleszczynski, M. (1987) Immunol. Today 8,345-352.

3. Hwang, S.-B. (1990) J. Lipid Mediators 2, 123-158.4. Rola-Pleszczynski, M. (1991) in Cell-Cell Interactions in the Release

ofInflammatory Mediators, eds. Wong, P. Y. K. & Serhan, C. N.(Plenum, New York), pp. 205-221.

5. Honda, Z.-I., Nakamura, M., Miki, I., Minami, M., Watanabe, T.,Seyama, Y., Okado, H., Toh, H., Ito, K., Miyamoto, T. & Shimizu,T. (1991) Nature (London) 349, 342-346.

6. Nakamura, M., Honda, Z.-I., Izumi, T., Sakanaka, C., Mutoh, H.,Minami, M., Bito, H., Seyama, Y., Matsumoto, T., Noma, M. &Shimizu, T. (1991) J. Biol. Chem. 266, 20400-20405.

7. Ye, R. D., Prossnitz, E. R., Zou, A. & Cochrane, C. G. (1991)Biochem. Biophys. Res. Commun. 180, 105-111.

8. Kunz, D., Gerard, N. P. & Gerard, C. (1992) J. Biol. Chem. 267,9101-9106.

9. Hwang, S. B. (1988) J. Biol. Chem. 263, 3225-3233.10. Kroegel, C., Yakawa, T., Westwick, J. & Barnes, P. J. (1989)

Biochem. Biophys. Res. Commun. 162, 511-521.11. B6yum, A. (1968) Scand. J. Clin. Lab. Invest. 21, Suppl., 77-80.12. Rola-Pleszczynski, M. & Stankova, J. (1992) Blood 80, 1004-1011.13. Muller, E., Dupuis, G., Turcotte, S. & Rola-Pleszczynski, M. (1991)

Biochem. Biophys. Res. Commun. 181, 1580-1586.14. Stankova, J., Gagnon, N. & Rola-Pleszczynski, M. (1992) Immu-

nology 76, 258-263.15. Coffey, M., Peters-Golden, M., Fantone, J. C., III, & Sporn,

P. H. S. (1992) J. Biol. Chem. 267, 570-576.16. Calabresse, C., Nguer, M. C., Pellegrini, O., Benveniste, J., Rich-

ard, Y. & Thomas, Y. (1992) Eur. J. Immunol. 22, 1349-1355.17. Rola-Pleszczynski, M., Pouliot, C., Turcotte, S., Pignol, B., Bra-

quet, P. & Bouvrette, L. (1988) J. Immunol. 140, 3547-3552.18. Shen, T. Y., Hussaini, I., Hwang, S.-B. & Chang, M. N. (1989)

Adv. Prostaglandin Thromboxane Leukotriene Res. 19, 359-362.19. Imboden, J. B. & Stobo, J. D. (1985) J. Exp. Med. 161, 46-456.20. Ishizaka, T. (1979) in The Mast Cell, eds. Pepys, J. & Edwards,

A. M. (Pitman, London), pp. 21-29.

A

B

5822 Immunology: Miiller et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

17, 2

022