89
Biochimica et Biophysics Acta, 1046 (1990) 89-96 Elsevier
BBALIP
53471
Production and characterization of monoclonal antibodies that specifically bind to phosphatidylcholine Kyung Soo Nam ‘, Kouji Igarashi
2, Masato Umeda
’
and Keizo Inoue ’
’ Department of Health Chemistry Faculty of Pharmaceutical Science, The University of Tokyo and ’ Tosoh Corporation, Tokyo Research Center, Tokyo (Japan) (Received
Key words:
Phosphatidylcholine;
5 February
Anti-phospholipid
1990)
antibody;
Lipid-protein
interaction
A series of monoclonal antibodies (mAbs) that react with phosphatidylcholine (PC) were established. All mAbs were highly specific to PC and no cross-reaction with other phospholipids was observed. The results obtained with two typical monoclonal anbodies, JE-1 and JE-8, were described. The analysis using synthetic PC analogs with modified polar head groups showed that the methyl groups on the quaternary nitrogen of the choline moiety were important for the binding. Each mAbs showed distinct acyl chain specificities of the PC molecules, and JE-1 showed considerable reactivity with PC with saturated fatty acids, whereas JE-8 could not react with the PC. Both mAbs bound to PC with unsaturated fatty acids, but showed distinct reactivity profiles. Both mAbs reacted only weakly with water-soluble haptens such as phosphorylcholine and L-a-glycerophosphocholine, suggesting that the hydrophobic moiety of the PC molecule is important for the maximum affinity. The interaction between the mAbs and the hydrophobic moieties of PC molecules was further studied by analyzing the effect of the mAbs on the activities of phospholipase A, and phospholipase C. JE-1 inhibited both enzyme activities, while JE-8 inhibited only the phospholipase C activity, indicating that JE-1 interacts more thoroughly with the hydrophobic region of the PC molecule than JE-8 does.
Introduction Recently, lipids and lipid-derived metabolites have been recognized as functioning as a second messenger by binding to specific membrane receptors and activating a signal cascade within the cell. Typical examples are arachidonic acid and its metabolites [1,2], the hydrolysis products of the phosphoinositides [3], and the platelet activating factor (PAF, identified primarily as 1-0alkyl-2-acetyl-sn-glycero-3-phosphocholine) [4]. Re-
Abbreviations: mAb, monoclonal antibody; PC, phosphatidylcholine; lyso-PC, lyso-phosphatidylcholine; SM, sphingomyelin; monomethylPE, 1,2-dipalmitoyl-sn-~ycero-3-phosphomonomethylethanola~ne; dimethyl-PE, 1,2-dipalmitoyl-sn-~ycero-3-phosphodimethylethanolamine; PE, 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine; di14:0 PC, 1,2-dimyristoyl-sn-glycero-3-phosphocholine; di-16:O PC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; di-18:0 PC, 1,2-distearoyl-sn-glycero-3-phosphocholine; di-14:l PC, 1,2-dimyristoleoylsn-glycero-3-phosphocholine; di-16:l PC, 1,2-dipalmitoleoyl-snglycero-3-phosphocholine; di-18:l PC, 1,2-dioleoyl-sn-glycero-3-phosphocholine. Correspondence: M. Umeda, Department of Health Chemistry, Faculty of Pharmaceutical Sciences, the University of Tokyo, Bunkyo-ku, Tokyo 113, Japan. 0005-2760/90/$03.50
0 1990 Elsevier Science Publishers
B.V. (Biomedical
ceptors bind these lipid-derived mediators in a highly specific manner and precise configurations of the mediator molecules were required for the binding. Although much effort has been focused on understanding the nature of specific lipid-protein interactions, the precise nature of the lipid-protein interactions is not known in any system. Concerning the interactions between choline-containing glycerophospholipids and proteins, many proteins have been shown to interact specifically with the lipids. Typical examples are the platelet activating factor receptor [5] and the perform (cytolysin) molecule which is responsible for the lysis of target cells by cytolytic T-lymphocytes [6]. Other proteins involved in the metabolism of choline-containing glycerophospholipids, such as the phosphatidylcholine-specific transfer protein [7], the phosphatidylcholine-specific lipid translocator protein which is responsible for the transmembrane movement of the lipid [8], 1-alkyl-2-lyso-sn-glycero-3phosphocholine:acetyltransferase [9], and phospholipase C specific for phosphatidylcholine [lO,ll] were also shown to interact specifically with the choline-containing glycerophospholipids. The difficulties of handling these proteins because of their limited aqueous solubility and scarcity in the biological membranes have Division)
90 hampered the progress in understanding the mechanisms of lipid-protein interactions between these molecules. Our approach to this problem has involved the production of the monoclonal antibodies which specifically recognize the choline-containing glycerophospholipids with binding profiles similar to those of other binding proteins. Recent findings showed that anti-idiotypic antibodies could recognize cross-reactive structures shared by the antibody and the receptor molecules, indicating the presence of a structural similarity between the antibody and the ligand binding receptors [12]. Monoclonal antibodies with definite specificity to the phospholipids may not only provide valuable information about the lipid-protein interactions, but may also represent a structural temperate for the production of anti-idiotypic antibodies which can cross-react with the actual receptor molecules. It is generally accepted that naturally abundant antigens such as DNA and membrane lipids are less immunogenic, and it has been difficult to obtain a monoclonal antibody with definite specificity against a certain phospholipid antigen. Although antiphospholipid antibodies have been frequently detected in patients with autoimmune, infectious, and other disorders, the majority of the antiphospholipid antibodies so far reported cross-reacted extensively with other phospholipids [13-151. These antiphospholipid antibodies formed a family of poorly characterized antibodies and no information has been available at present regarding the precise nature of the interaction between antiphospholipid antibodies and lipid antigens. In the previous study, we established an effective method for the production of monoclonal antibodies against phospholipid antigens [16,17] and succeeded in producing monoclonal antibodies which could recognize the stereospecific configuration of phosphatidylserine 1171. In the present study we estab~shed a series of mon~lon~ antibodies which specifically recognize phosphatidylcholine and we studied the interaction between the monoclonal antibodies and phosphatidylcholine molecules. Materials and ~eth~s Lipids Egg yolk phosphatidylcholine and sphingomyelin from bovine brain were prepared by chromatography on Aluminum Oxide Neutral and Iatrobeads. Phosphatidic acid was prepared from egg yolk phosphatidylcholine by treatment with cabbage phospholipase D (EC 3.1.4.4). Phosphatidylethanolamine was purified from Escherichia coii as described previously [18]. Phosphatidylserine was prepared from bovine brain white matter by the method of Folch 1191 and further purified by CM-cellulose column chromatography 1201.
Phosphatidylinositol was prepared from yeast by the method of Trevelyan [21]. Cardiolipin from beef heart was prepared by the method of Faure and Marechal [22]. Lysophosphatidylcholine was prepared by treatment of egg yolk phosphatidylcholine with phospholipase A, from snake venom (habu, Trimeresurus jlavoviridis) and was further purified by Iatrobeads column. 1,2-distearoyl, -dipalmitoyl and -dimyristoylso-glycero-3-phosphocholine, 1,2-dipalmitoyl-snglycero-3-phosphomonomethylethanola~ne and 1,2-dipalmitoyl-svr-glycero-3-phosphodimethylethanolamine were purchased from Calbiochem (San Diego, CA, U.S.A.). 1,2-dioleoyl-, 1,2-dipalmitoleoyland 1,2-dimyristoleoyl-sn-glycero-3-phosphocholine were purchased from Avanti Polar-Lipids ~~irmin~am, AL). 1-O-Octadecyl-2-O-( IV, ~-dimethylcarbamoyl)-glycero3-phosphocholine was kindly provided by Takeda Chemical. All these lipids showed a single spot on the thin-layer chromatography. Other chemicals. l-acyl-2-[14C]linoleoylglycerophosphocholine was purchased from the Amersham International (U.K.). Purified phospholipase A, of porcine pancreas and phospholipase C of Bacillus cereus were purchased from Boehringer Mannheim (F.R.G.). Choline chloride, phosphorylcholine chloride, L-W glycerophosphocholine were purchased from Sigma Chemical (St. Louis, MO). Production of monoclonal antibody against phosphatidvlcholine Salmonella Minnesota, strain R595 was kindly provided by Dr. S. Kanagasaki (The Institute of Medical Science, The University of Tokyo). The bacteria were washed twice with distilled water, twice with acetone and once with diethyl ether and dried in vacua. KDOlinked oligosaccharides were removed by heating in 1% aqueous acetic acid solution at 100 o C for 2 h as described previously [16]. The ~a~munefIa were coated with l-0-octadecyl-2-O-t ~,~-dimethylcarbamoyl)glycero-3-phosphocholine and the suspension containing 5 I-18 of the phospholipid and 50 pg of Sulmonella in 0.2 ml of 20 mM phosphate buffer (pH 7.4) and 150 mM NaCl (PBS) were prepared as described previously 1231. Immunization of mice was performed by the intrasplenic immunization as described previously [14]. Briefly, the Balb/c mouse is anesthetized by intraperitoneal injection of 1 mg of Nembutal (Abbott Laboratories, North Chicago, IL, U.S.A.) and the suspension containing 5 pg of phospholipid and 50 yg of ~a~rnone~~~ is injected with a 25-gauge butterfly needle into spleen which was exposed after an incision of the abdominal wall and the peritoneum was made. After the injection, the spleen was returned into the peritoneal cavity and the peritoneal walls were sutured with staples. 21 d after the immu~ation, the second intrasplenic injection was performed and the fusion with P3-X63-Ag.653 cells
91 (obtained from the Japanese Cancer Research Resources Bank) was performed 3 d later by the method of Galfre et al [24] and Kohler and Milstein [25]. Hybridomas were cultured in a synthetic medium (GIT medium, Nippon Pharm., Tokyo, Japan) without supplement of serum. Hybridoma supernatants were screened for antibodies that bound to phosphatidylcholine by an enzyme-linked immunosorbent assay (ELISA). The hybridomas were cloned three times by limiting dilution, and monoclonal antibodies were purified from culture supematants by ammonium sulfate precipitation. The heavy chain class of the mAbs was determined by using biotinylated rabbit anti-mouse immunoglobulins specific to IgM, IgGl, IgG2a, IgG2b, or IgG3 and peroxidase conjugated streptavidin (ZYMED laboratories, San Francisco, CA). Binding of the monoclonal antibodies to phospholipids The binding of the monoclonal antibodies to phospholipids was measured by enzyme-linked immunosorbent assay (ELISA) [26]. In brief, the wells of the microtiter plates were coated with 50 ~1 of phospholipid antigens in ethanol (10 PM) by evaporation at room temperature. The wells were blocked with a solution containing BSA (30 mg/ml and incubated with hybridoma supematants. The antibody bound was detected by biotinylated antimouse Igs (ZYMED Laboratories, San Francisco, CA) followed by incubation with peroxidase-conjugated streptavidin or alkaline phosphatase-conjugated streptavidin (ZYMED Laboratories). Optical density at 490 nm was determined by the addition of o-phenylenediamine substrate, and the relative fluorescence intensity was measured by the addition of 0.25 mM 4-methyl umbelliferyl phosphate substrate solution in a fluorescent ELISA reader (MTP-32 fluorescence microplate reader, Corona Electric, Tokyo, Japan). Throughout this study we used a 20 mM Hepes buffer containing 150 mM NaCl (pH 7.4) (Hepes-NaCl buffer) instead of phosphate-buffered saline. Inhibition of antibody binding to phospholipid Inhibition of the binding of anti-phosphatidylcholine antibodies by water-soluble haptens or choline-containing glycerophospholipids was performed as follows. 50 ~1 of the mixture containing monoclonal antibody and the various inhibitors were preincubated for 1 h at room temperature and the mixture was transferred to the microtiter wells coated with egg yolk phosphatidylcholine. The amounts of antibody bound to the plate were measured as described above. Choline chloride, phosphorylcholine chloride, L-a-glycerophosphocholine (Sigma Chemical, St. Louis, MO), and lyso-phosphatidylcholine were dissolved in Hepes-NaCl buffer and used for the experiment. The small unilamellar vesicles of phosphatidylcholine with various fatty acids were prepared by the sonication of the multilamellar vesicle
using the Branson Sonifier 185 (Branson Cleaning Equipment, Shelton, CT, U.S.A.). In brief, the dried lipid film containing 2.0 pmol of phosphatidylcholine was swollen in 1 ml of Hepes-NaCl buffer and was sonicated for 10 min under a nitrogen flow at the temperature above the phase-transition temperature of the lipids. Assay of phospholipase A, and phospholipase C Both phospholipase A, (porcine pancreas) and phospholipase C (B. cerew) assay were performed using 1-acyl-2-[‘4C]linoleoylglycerophosphocholine as substrate as previously described [27]. In phospholipase A, assay, the standard reaction mixture contained 20 PM labeled substrate, 100 mM Tris-HCl (pH 7.4) and 4 mM CaCl, in a total volume of 0.25 ml. The reaction was carried out for 10 min at 37’C and was stopped by adding 1.25 ml of Dole’s reagent [28]. The [‘4C]linoleate released was extracted by the method previously described [29]. In phospholipase C assay, the standard reaction mixture contained 20 FM labeled substrate, 4 mM CaCl,, and 50 mM Tris-HCl (pH 7.4) in a total volume of 0.4 ml. After incubation for 45 min at 37 o C, the reaction was stopped by adding 1 ml of cold methanol. All lipids were extracted by the methods of Bligh and Dyer [30] and analyzed on thin-layer chromatography (DC-Fertigplatten Kieselgel 60, Merck) with the solvent system of petroleum ether/ethyl ether/acetic acid (80 : 30 : 1, v/v). The radioactive spot corresponding to diacylglycerol was scraped off from the plate into vials and counted in a Packard Liquid Scintillation Counter. For the inhibition assay, various amounts of the anti-phosphatidylcholine monoclonal antibodies were included in the reaction mixture. Results Specificities of monoclonal antibodies against phosphatidylcholine A series of monoclonal antibodies that bound specifically to phosphatidylcholine were established after the direct immunization of the antigen into the spleens of Balb/c mice. The binding of the monoclonal antibodies to various lipid haptens were analyzed by direct binding of the monoclonal antibodies to the lipids coated on the microtiter plates (ELISA) and by inhibition analysis of the ELISA by aqueous suspensions of either lipids or water-soluble haptens. Among seven monoclonal antibodies obtained, results obtained with two typical monoclonal antibodies named JE-1 and JE-8, are described in this paper. Both monoclonal antibodies were highly specific to phosphatidylcholine and no cross-reaction with other phospholipids such as phosphatidylethanolamine, phosphatidylserine and phosphatidylinositol was observed (Fig. 1). Although JE-8 showed some weak binding to other phospholipids such as
92 OD 492
OD492
2.5
2.5
2.0
1.0
0
I
,
JE-1 2.0
1.0
. E’ ox
-
.
100
10
Antibody
0
1 Antibody
dilution
dilution
Fig. 1. Reactivities of anti-phosphatidylchohne mAbs. JE-1 and JE-8, with phospholipids. Microtiter plates were coated with 10 PM of phospholipid Ags. The mAb bound were detected with biotinylated anti-mouse Ig and streptavidin-conjugated peroxidase. Phospholipid Ags used were phosphatidylcholine (0). phosphatidylserine ( 0), phosphatidylinositol (A), phosphatidic acid (A). phosphatidylethanolamine (o), cardiolipin (H), and phosphatidylglycerol ( x ).
phosphatidylethanolamine and cardiolipin, the binding was not significant and it did not reach statistical significance. Reactivities of the monoclonal antibodies with synthetic phosphatidylcholine analogs with modified polar head groups are shown in Fig. 2. The order of the reactivity of both JE-1 and JE-8 was phosphatidylcholine Z+ dimethyl-PE > monomethyl-PE, indicating that the methyl groups on the quaternary nitrogen of the choline moiety of the phospholipids were responsible for the binding. ELISA inhibition assay In order to further analyze the interaction between the monoclonal antibodies and the phosphatidylcholine membranes, an inhibition analysis of the ELISA by small unilamellar vesicles composed of synthetic phosphatidylcholine with various fatty acid compositions was performed (Fig. 3 and Fig. 4). JE-1 showed considerable reactivity with phosphatidylcholine molecules carrying saturated fatty acids, and 50% inhibition of the
binding was observed at 100 pmol of di-18:O PC and 600 pmol of di-16:O PC, and only weak inhibition was observed with di-14:O PC (Fig. 3A). In contrast, phosphatidylcholine molecules with these saturated fatty acids failed to inhibit the binding of JE-8 (Fig. 3Bj. The reactivities of the monoclonal antibodies with phosphatidylcholine molecules with fatty acids containing one double bond in each acyl chain were next examined. Di-14:1-, di-16:1- and di-18:l PC significantly inhibited the binding of both monoclonal antibodies, and 50% inhibition of the binding of JE-1 was observed at less than 2 pmol of both di-18:1- and di-16:l PC, and at 600 pmol of di-14:l PC (Fig. 4A). In contrast to JE-1, the inhibition of the JE-8 binding increased with the decreasing acyl chain length of phosphatidylcholine molecules, and 50% inhibition of the binding of JE-8 was observed at 75 pmol of di-14:l PC, 300 pmol of di-16:l PC and 2500 pmol of di-18:l PC (Fig. 4B). The reactivities of the mAbs with the water-soluble head groups of phosphatidylcholine were also examined by inhibition of ELISA (Fig. 5). No inhibition was observed with choline chloride, while a certain extent of inhibition was observed with phosphorylcholine and L-a-glycerophosphocholine when higher concentrations of the inhibitors were employed. No appreciable inhibition was observed with other choline-containing phospholipid, lysophosphatidylcholine and sphingomyelin. Similar results were also obtained with JE-8 (data not shown).
Effect of the monoclonal antibodies on the activities of phospholipases In order to obtain further information on the interaction between anti-phosphatidylcholine antibodies and the hydrophobic moieties of the phosphatidylcholine molecule, the effect of the antibody binding on the activities of phospholipase A, (Fig. 6) and phospholipase C (Fig. 7) was studied. JE-1 inhibited the phospholipase A, (from porcine pancreas) activity dose-dependently, whereas no appreciable inhibition was observed with JE-8. Fig. 6B (Inset) shows double-recipro-
JE-1
JE-8
m
l,CH3
PC
-“:$3
DMPE
-N:$
MMPE
-NH-CH3
z
-NH2
I
1
I 3
m
PE
w
2.0
1.0 OD492
0
0
1.0
2.0
OD492
Fig. 2. Reactivities of JE-1 and JE-8 with synthetic phosphatidylcholine and its analogs. Microtiter plates were coated with 10 PM of the phospholipid Ags and the mAb bound were detected with biotinylated anti-mouse Ig and streptavidin-conjugated peroxidase. Phospholipid Ags used were phosphatidylcholine (PC), dimethyl-PE (DMPE), monomethyl-PE (MMPE), and 1.2-dipalmitoylphosphoethanolamine (PE).
93 100
JE-1
JE-8 6
C16:O
c18:0
C14:O A
0 0
2.5
9.8
39
156
2500
625
9.8
39
fp molf
156
625
2500
(p mol) Amounts
of synthetic
PC added
Fig. 3. Inhibition of mAb binding by phosphatidylcholine molecules composed of various saturated fatty acyl chains. JE-1 or JE-8 was preincubated with the vesicles composed of di-18:O PC (o), di-16:O PC (A), or di-14:O PC (H) and the mixtures were transferred to the microtiter wells coated with phosphatidylcholine. After incubation, the mAb bound was detected with biotinylated anti-mouse Igs and streptavidin-conjugated peroxidase.
JE-1 and JE-8 dose-dependently and a 50% inhibition was observed at 2 pg of JE-1 and 4 pg of JE-8 (Fig. 7).
cal plots of the kinetic data on hydrolysis of phosphatidylcho~ne in the absence or presence of various amounts of JE-1. The K, value (9.0 PM) of the enzyme increased in the presence of the antibody, whereas the Vmax value (68 pmol/min per mg) was not affected, suggesting that JE-1 competed with the active site of the enzyme for the substrate. The enzyme activity of phospholipase C (from B. cereur) was inhibited by both
Discussion
A series of mAbs that specifically react with phosphatidylcholine were established and the reactivity profiles of the two typical mAbs, designated JE-1 and
100
JET
JE-8
B
5
0 0
9.8
39
156
625
2500
9.8
(p mob
39
156
625
2500
(p mol) Amounts
of synthetic
PC added
Fig. 4. Inhibition of mAb binding by phosphatidylcholine molecules composed of various fatty acyl chains with one doubIe bond. JE-1 or JE-8 was preincubated with the vesicles composed of di-l&l PC (O), di-16:l PC (A), or di-14:1 PC (B) and the mixtures were transferred to the microtiter wells coated with phosphatidylcho~e. After incubation, the mAb bound was detected as described in Fig. 3.
94 100
Control
IgM
0
(p mol) Amounts
0
of Inhibitors
added
2.0 Amounts
4.0 of mAbs added
6.0 l,ag)
Fig. 5. Inhibition of mAb binding by various haptens. JE-1 was preincubated with soluble haptens: 0, phosphatidylcholine: t, choline: A, phosphorylcholine: n, L-a-glycerophosphocholine: A, lyso-PC: 0, SM and the mixtures were transferred to the microtiter wells coated with phosphatidylcholine. After incubation, the mAb bound was detected as described in Fig. 3.
Fig. 7. Effect of anti-phosphatidylcholine mAbs on phospholipase C activity. Phospholipase C (E. cereur) (0.04 unit) activity was measured in the presence of various amounts of JE-1 (a), JE-8 (A) or control IgM (0). The results are expressed as percentage inhibition of the enzyme activity measured in the absence of mAb.
JE-8, have been described. Both mAbs were highly specific to phosphatidylcholine and did not cross-react with other phospholipids. In the analysis using synthetic phosphatidylcholine analogs, the order of the reactivities to the mAbs is d&16:0 PC x=-dimethyl-PE > monomethyl-PE > PE, indicating the methyl groups on the quatemary nitrogen of the choline residue may play an important role in the interaction. The reactivities of mAbs which recognize phosphorylcholine have been extensively studied [31-351. Although antiphos-
phorylcholine mAbs, which were usually elicited by immunizing mice with phosphorylcholine-protein conjugates, such as phosphorylcholine-keyhole limpet hemocyanin, have also been shown to recognize a negatively charged phosphate and the trimethyl structure of the choline residue [41], the following lines of evidence indicate that the anti-phosphatidylcholine mAbs obtained in this study may be classified into the other family of the anti-phospholipid antibody. First, neither of the mAbs could react significantly with water-soluble haptens such as phosphorylcholine and L-W glycerophosphocholine. Second, the mAbs were highly specific to phosphatidylcholine and neither of them reacted appreciably with sphingomyelin when phosphorylcholine residue was substituted by ceramide. Third, the reactivities of the mAbs were highly dependent on the compositions of the fatty acyl chains of the phosphatidylcholine molecules. Fourth, anti-idiotypic antibody against JE-1 could not cross-react with the anti-phosphorylcholine mAb, TEPC 15 (unpublished observation). The interaction between mAbs and the membranes composed of various phosphatidylcholine molecules were examined by inhibition assay of ELISA. Both mAbs were shown to exhibit distinct acyl chain specificities and JE-1 interacts more strongly with the phosphatidylcholine molecules containing longer fatty acyl chains, while JE-8 interacts preferentially with the molecules containing short fatty acyl chains. JE-8 seems to prefer the phosphatidylcholine molecules in the fluid phase membranes and it could not interact with the phosphatidylcholine molecules in the tightly packed membranes. Both mAbs could interact with the water soluble head groups such as phosphorylcholine and L-a-glycerophosphocholine, indicating that both mAbs may interact with a single phosphatidylcholine mole-
100
JE-1 c
0 z a E
50
5 ae JE-8 *
control
0 0
1.0
Amounts
of mAbs
2.0 added
IgM
0
3.0 (,ug)
Fig. 6. Effect of anti-phosphatidylcholine rn4bs on phospholipase A, activity. (A) Phospholipase A, (porcine pancreas) (0.25 ng) activity was measured in the presence of various amounts of JE-1 (O), JE-8 (A) or control IgM (0). The results are expressed as percentage inhibition of the enzyme activity measured in the absence of mAb. (B) Lineweaver-Burk analysis of phospholipase A, activity incubated with JE-1. Phospholipase A, activity against increasing concentrations of phosphatidylcholine was measured in the presence of 6 pg (A), 3 pg (0) of JE-1 or in the absence of the mAb.
95
JEl
5 C”3
Fig. 8. Schematic
model of the interaction between phosphatidylcholine molecule.
anti-PC
mAbs and
cule. Since the interaction of the mAbs with soluble haptens was so weak, it is likely that a secondary interaction between JE-1 and the glycerol backbone or the fatty acyl chains of the phosphatidylcholine molecule affects the stability of the antigen-antibody complex. The interaction between the mAbs and the glycerol backbone or fatty acyl moiety of the phosphatidylcholine molecule was further supported by the inhibition analysis of the phospholipase activities by the mAbs. It is generally accepted that phospholipase C may have less of a hydrophobic character than the phospholipase A, since its site of action is in the aqueous region of the substrate aggregate and phospholipase A, needs to penetrate more deeply into the hydrophobic region of the phospholipid [36]. In the present analysis, JE-1 inhibited both the phospholipase A, and phospholipase C activity, while JE-8 inhibited only the phospholipase C activity. This observation might indicate that JE-1 may interact more thoroughly with the hydrophobic region of the phosphatidylcholine molecule than JE-8 does (Fig. 8). Further studies will be necessary for understanding the precise mechanisms of the interaction between the mAbs and the membranes. Recently, Mercolino et al [37,38] reported that mouse autoantibodies against bromelain-treated mouse erythrocytes (BrMRBC), which were mainly produced by Ly-1+ B cells in the peritoneal cavity, recognize phosphatidylcholine. Though the fine specificities of these anti-BrMRBC have not yet been studied, our anti-phosphatidylcholine mAbs showed similar characteristics with those of the anti-BrMRBC mAbs. Our mAbs not only showed strong lytic activity against the phosphatidylcholine membrane measured by the com-
plement dependent liposome lysis assay [16], but could also hemolyze intact mouse and human erythrocytes in a complement dependent manner (unpublished observation). Lytic activities were enhanced by pretreatment of the erythrocytes with trypsin or sialidase. Our antiphosphatidylcholine mAbs did not react with sphingomyelin and could not hemolyze sheep erythrocytes which the dominant phospholipid was sphingomyelin instead of phosphatidylcholine. Some of the anti-BrMRBC were reported to react only with BrMRBC and not with sheep erythrocytes, showing a similar reactivity profile with our anti-phosphatidylcholine mAbs [39]. Further genetic and idiotypic analyses of the anti-phosphatidylcholine mAbs [40,41] may clarify the relationship between these two families of mAbs. In our preliminary experiments, anti-idiotypic antibodies which were raised against JE-1 cross-reacted with phosphatidylcholine-specific lipid transfer protein purified from bovine liver (manuscript in preparation). The results may suggest that the anti-phosphatidylcholine mAbs obtained in this study will provide useful tools for the identification and purification of a family of choline-containing glycerophospholipid binding proteins which have not been purified by the conventional procedures of the protein purification. Acknowledgements The authors wish to Metropolitan Institute the monoclonal IgM ported by Grants-in-Aid Ministry of Education,
thank Dr. Akemi Suzuki (Tokyo of Medical Science) for offering antibody. This work was supfor Science Research from the Science and Culture of Japan.
References 1 Hammarstrom, S. (1983) Annu. Rev. B&hem. 52, 355-377. 2 Needleman, P., Turk, J., Jakschik, B.A., Morrison, A.R. and Lefkowith, J.B. (1986) Annu. Rev. B&hem. 55, 69-102. 3 Berridge, M.J. and Irvine, R.F. (1984) Nature 312, 315-321. 4 Braquet, P. and Rola-Pleszczynski, M. (1987) Immunology Today 8, 345-352. 5 Braquet, P., Touqui, L., Shen, T.Y. and Vargaftig, B.B. (1987) Pharmacol. Rev. 39, 97-145. 6 Tschopp, J., Schafer, S., Masson, D., Peitsch, M.C. and Heusser, C. (1989) Nature 337, 272-274. 7 Kamp, H.H., Wirtz, K.W.A., Baer, P.R., Slotboom, A.J., Rosenthal, A.F., Paltauf, F. and Van Deenen, L.L.M. (1977) Biochemistry 16, 1310-1316. 8 Backer, J.M. and Dawidowicz, E.A. (1987) Nature 327, 341-343. 9 Lee, T.C. (1985) J. Biol. Chem. 260.10952-10955. 10 Grillone, L.R., Clark, M.A., Godfrey, R.W., Stassen, F. and Crooke, S.T. (1988) J. Biol. Chem. 263, 2658-2663. 11 Pelech, S.L. and Vance, D.E. (1989) Trends in B&hem. Sci. 14, 28-30. 12 Gaulton, G.N. and Green, M.I. (1986) Annu. Rev. Imrnunol. 4, 253-280. 13 Alving, C.R. (1986) Chem. Phys. Lipids 40, 303-314.
96 14 Janoff, A. and Rauch, J. (1986) Chem. Phys. Lipids 40, 315-332. 15 Harris, E.N., Asherson, R.A. and Hughes, G.R.V. (1988) Annu. Rev. Med. 39, 261-271. 16 Miyazawa, A., Umeda, M., Horikoshi, T., Yanagisawa, K., Yoshioka, T. and Inoue, K. (1988) Mol. Immunol. 25, 1025-1031. 17 Umeda, M., Igarashi, K., Nam, KS. and Inoue, K. (1989) J. Immunol. 143, 2273-2279. 18 Doi, 0. and Nojima, S. (1971) B&him. Biophys. Acta 248, 234-244. 19 Folch, J. (1942) J. Biol. Chem. 146, 35-44. 20 Comfurius, P. and Zwaal. R.F.A. (1977) Biochim. Biophys. Acta 488, 36-42. 21 Trevelyan, W.E. (1966) J. Lipid Res. 7, 445-447. 22 Faure, M. and Marechal, J. (1962) Compt. Rend. 254, 4518-4520. 23 Umeda, M., Diego, I. and Marcus, D.M. (1986) J. Immunol. 137, 3264-3269. 24 Galfre, G.S., Howe, C.. Milstein, C., Butcher, G.W. and Howard. J.C. (1977) Nature 266, 550-552. 25 Kohler, G. and Milstein, C. (1975) Nature 256, 495-497. 26 Umeda, M., Diego, I., Ball, E.D. and Marcus, D.M. (1986) J. Immunol. 136, 2562-2567. 27 Horigome, K., Hayakawa, M., Inoue, K. and Nojima, S. (1987) J. B&hem. 101, 625-631. 28 Dole, V.P. and Meinertz, H. (1960) J. Biol. Chem. 235, 2595-2599. 29 Arai, H.. Inoue, K., Natori, T., Banno, Y., Nozawa, Y. and Nojima, S. (1985) J. Biochem. 97, 1525-1532.
30 Bligh, E.G. and Dyer. 911-917.
W.J. (1959) Can. J. Biochem.
31 Niedieck, chemistry
U.
B., Kuck, 15, 471-475.
and
Gardemin,
H.
(1978)
Physiol.
37,
Immuno-
32 Andres, C.M., Maddalena, A.. Hudak, S.. Young, N.M. and Claflin, J.L. (1981) J. Exp. Med. 154, 1584-1589. 33 Novotny, J., Bruccoleri, R., Newell, J., Murphy, D., Haber, E. and Karplus. M. (1983) J. Biol. Chem. 258, 14433-14437. 34 Lai, E.H.C., Kabat, E.A., Meienhofer, J.. Heimer, E.P., Olson, A.J. and Lemer, R. (1987) Nature 325, 168-171. 35 Bruderer, U.. Stenzel-Poore, M.P., Bachinger, H.P., Fellman, J.H. and Rittenberg, M.B. (1989) Mol. Immunol. 26, 63-71. 36 Waite. M. (1987) in The Phospholipases (Waite, M., ed.), pp. 191-241, Plenum Press. New York. 37 Mercolino, T.J., Arnold. L.W. and Haughton. G. (1986) J. Exp. Med. 163, 155-165. 38 Mercolino, T.J., Arnold, L.W., Hawkins, L.A. and Hat&ton, G. (1988) J. Exp. Med. 168. 687-698. 39 Pennell, CA., Arnold, L.W., Lutz, P.M.. LoCascio, N.J., Willoughby, P.B. and Hat&ton. G. (1985) Proc. Natl. Acad. Sci. USA 82. 3799-3803. 40 Mercolino, T.J., Locke, A.L., Afshari, A., Sasser, D.. Travis, W.W., Arnold. L.W. and Haughton, G. (1989) J. Exp. Med. 169, 18691877. 41 Kimura, H., Umeda, 142, 3477-3481.
M. and Marcus,
D.M.
(1989) J. Immunol.
Biochimica et Biophysics Acta, 1046 (1990) 89-96 Elsevier
BBALIP
53471
Production and characterization of monoclonal antibodies that specifically bind to phosphatidylcholine Kyung Soo Nam ‘, Kouji Igarashi
2, Masato Umeda
’
and Keizo Inoue ’
’ Department of Health Chemistry Faculty of Pharmaceutical Science, The University of Tokyo and ’ Tosoh Corporation, Tokyo Research Center, Tokyo (Japan) (Received
Key words:
Phosphatidylcholine;
5 February
Anti-phospholipid
1990)
antibody;
Lipid-protein
interaction
A series of monoclonal antibodies (mAbs) that react with phosphatidylcholine (PC) were established. All mAbs were highly specific to PC and no cross-reaction with other phospholipids was observed. The results obtained with two typical monoclonal anbodies, JE-1 and JE-8, were described. The analysis using synthetic PC analogs with modified polar head groups showed that the methyl groups on the quaternary nitrogen of the choline moiety were important for the binding. Each mAbs showed distinct acyl chain specificities of the PC molecules, and JE-1 showed considerable reactivity with PC with saturated fatty acids, whereas JE-8 could not react with the PC. Both mAbs bound to PC with unsaturated fatty acids, but showed distinct reactivity profiles. Both mAbs reacted only weakly with water-soluble haptens such as phosphorylcholine and L-a-glycerophosphocholine, suggesting that the hydrophobic moiety of the PC molecule is important for the maximum affinity. The interaction between the mAbs and the hydrophobic moieties of PC molecules was further studied by analyzing the effect of the mAbs on the activities of phospholipase A, and phospholipase C. JE-1 inhibited both enzyme activities, while JE-8 inhibited only the phospholipase C activity, indicating that JE-1 interacts more thoroughly with the hydrophobic region of the PC molecule than JE-8 does.
Introduction Recently, lipids and lipid-derived metabolites have been recognized as functioning as a second messenger by binding to specific membrane receptors and activating a signal cascade within the cell. Typical examples are arachidonic acid and its metabolites [1,2], the hydrolysis products of the phosphoinositides [3], and the platelet activating factor (PAF, identified primarily as 1-0alkyl-2-acetyl-sn-glycero-3-phosphocholine) [4]. Re-
Abbreviations: mAb, monoclonal antibody; PC, phosphatidylcholine; lyso-PC, lyso-phosphatidylcholine; SM, sphingomyelin; monomethylPE, 1,2-dipalmitoyl-sn-~ycero-3-phosphomonomethylethanola~ne; dimethyl-PE, 1,2-dipalmitoyl-sn-~ycero-3-phosphodimethylethanolamine; PE, 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine; di14:0 PC, 1,2-dimyristoyl-sn-glycero-3-phosphocholine; di-16:O PC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; di-18:0 PC, 1,2-distearoyl-sn-glycero-3-phosphocholine; di-14:l PC, 1,2-dimyristoleoylsn-glycero-3-phosphocholine; di-16:l PC, 1,2-dipalmitoleoyl-snglycero-3-phosphocholine; di-18:l PC, 1,2-dioleoyl-sn-glycero-3-phosphocholine. Correspondence: M. Umeda, Department of Health Chemistry, Faculty of Pharmaceutical Sciences, the University of Tokyo, Bunkyo-ku, Tokyo 113, Japan. 0005-2760/90/$03.50
0 1990 Elsevier Science Publishers
B.V. (Biomedical
ceptors bind these lipid-derived mediators in a highly specific manner and precise configurations of the mediator molecules were required for the binding. Although much effort has been focused on understanding the nature of specific lipid-protein interactions, the precise nature of the lipid-protein interactions is not known in any system. Concerning the interactions between choline-containing glycerophospholipids and proteins, many proteins have been shown to interact specifically with the lipids. Typical examples are the platelet activating factor receptor [5] and the perform (cytolysin) molecule which is responsible for the lysis of target cells by cytolytic T-lymphocytes [6]. Other proteins involved in the metabolism of choline-containing glycerophospholipids, such as the phosphatidylcholine-specific transfer protein [7], the phosphatidylcholine-specific lipid translocator protein which is responsible for the transmembrane movement of the lipid [8], 1-alkyl-2-lyso-sn-glycero-3phosphocholine:acetyltransferase [9], and phospholipase C specific for phosphatidylcholine [lO,ll] were also shown to interact specifically with the choline-containing glycerophospholipids. The difficulties of handling these proteins because of their limited aqueous solubility and scarcity in the biological membranes have Division)
90 hampered the progress in understanding the mechanisms of lipid-protein interactions between these molecules. Our approach to this problem has involved the production of the monoclonal antibodies which specifically recognize the choline-containing glycerophospholipids with binding profiles similar to those of other binding proteins. Recent findings showed that anti-idiotypic antibodies could recognize cross-reactive structures shared by the antibody and the receptor molecules, indicating the presence of a structural similarity between the antibody and the ligand binding receptors [12]. Monoclonal antibodies with definite specificity to the phospholipids may not only provide valuable information about the lipid-protein interactions, but may also represent a structural temperate for the production of anti-idiotypic antibodies which can cross-react with the actual receptor molecules. It is generally accepted that naturally abundant antigens such as DNA and membrane lipids are less immunogenic, and it has been difficult to obtain a monoclonal antibody with definite specificity against a certain phospholipid antigen. Although antiphospholipid antibodies have been frequently detected in patients with autoimmune, infectious, and other disorders, the majority of the antiphospholipid antibodies so far reported cross-reacted extensively with other phospholipids [13-151. These antiphospholipid antibodies formed a family of poorly characterized antibodies and no information has been available at present regarding the precise nature of the interaction between antiphospholipid antibodies and lipid antigens. In the previous study, we established an effective method for the production of monoclonal antibodies against phospholipid antigens [16,17] and succeeded in producing monoclonal antibodies which could recognize the stereospecific configuration of phosphatidylserine 1171. In the present study we estab~shed a series of mon~lon~ antibodies which specifically recognize phosphatidylcholine and we studied the interaction between the monoclonal antibodies and phosphatidylcholine molecules. Materials and ~eth~s Lipids Egg yolk phosphatidylcholine and sphingomyelin from bovine brain were prepared by chromatography on Aluminum Oxide Neutral and Iatrobeads. Phosphatidic acid was prepared from egg yolk phosphatidylcholine by treatment with cabbage phospholipase D (EC 3.1.4.4). Phosphatidylethanolamine was purified from Escherichia coii as described previously [18]. Phosphatidylserine was prepared from bovine brain white matter by the method of Folch 1191 and further purified by CM-cellulose column chromatography 1201.
Phosphatidylinositol was prepared from yeast by the method of Trevelyan [21]. Cardiolipin from beef heart was prepared by the method of Faure and Marechal [22]. Lysophosphatidylcholine was prepared by treatment of egg yolk phosphatidylcholine with phospholipase A, from snake venom (habu, Trimeresurus jlavoviridis) and was further purified by Iatrobeads column. 1,2-distearoyl, -dipalmitoyl and -dimyristoylso-glycero-3-phosphocholine, 1,2-dipalmitoyl-snglycero-3-phosphomonomethylethanola~ne and 1,2-dipalmitoyl-svr-glycero-3-phosphodimethylethanolamine were purchased from Calbiochem (San Diego, CA, U.S.A.). 1,2-dioleoyl-, 1,2-dipalmitoleoyland 1,2-dimyristoleoyl-sn-glycero-3-phosphocholine were purchased from Avanti Polar-Lipids ~~irmin~am, AL). 1-O-Octadecyl-2-O-( IV, ~-dimethylcarbamoyl)-glycero3-phosphocholine was kindly provided by Takeda Chemical. All these lipids showed a single spot on the thin-layer chromatography. Other chemicals. l-acyl-2-[14C]linoleoylglycerophosphocholine was purchased from the Amersham International (U.K.). Purified phospholipase A, of porcine pancreas and phospholipase C of Bacillus cereus were purchased from Boehringer Mannheim (F.R.G.). Choline chloride, phosphorylcholine chloride, L-W glycerophosphocholine were purchased from Sigma Chemical (St. Louis, MO). Production of monoclonal antibody against phosphatidvlcholine Salmonella Minnesota, strain R595 was kindly provided by Dr. S. Kanagasaki (The Institute of Medical Science, The University of Tokyo). The bacteria were washed twice with distilled water, twice with acetone and once with diethyl ether and dried in vacua. KDOlinked oligosaccharides were removed by heating in 1% aqueous acetic acid solution at 100 o C for 2 h as described previously [16]. The ~a~munefIa were coated with l-0-octadecyl-2-O-t ~,~-dimethylcarbamoyl)glycero-3-phosphocholine and the suspension containing 5 I-18 of the phospholipid and 50 pg of Sulmonella in 0.2 ml of 20 mM phosphate buffer (pH 7.4) and 150 mM NaCl (PBS) were prepared as described previously 1231. Immunization of mice was performed by the intrasplenic immunization as described previously [14]. Briefly, the Balb/c mouse is anesthetized by intraperitoneal injection of 1 mg of Nembutal (Abbott Laboratories, North Chicago, IL, U.S.A.) and the suspension containing 5 pg of phospholipid and 50 yg of ~a~rnone~~~ is injected with a 25-gauge butterfly needle into spleen which was exposed after an incision of the abdominal wall and the peritoneum was made. After the injection, the spleen was returned into the peritoneal cavity and the peritoneal walls were sutured with staples. 21 d after the immu~ation, the second intrasplenic injection was performed and the fusion with P3-X63-Ag.653 cells
91 (obtained from the Japanese Cancer Research Resources Bank) was performed 3 d later by the method of Galfre et al [24] and Kohler and Milstein [25]. Hybridomas were cultured in a synthetic medium (GIT medium, Nippon Pharm., Tokyo, Japan) without supplement of serum. Hybridoma supernatants were screened for antibodies that bound to phosphatidylcholine by an enzyme-linked immunosorbent assay (ELISA). The hybridomas were cloned three times by limiting dilution, and monoclonal antibodies were purified from culture supematants by ammonium sulfate precipitation. The heavy chain class of the mAbs was determined by using biotinylated rabbit anti-mouse immunoglobulins specific to IgM, IgGl, IgG2a, IgG2b, or IgG3 and peroxidase conjugated streptavidin (ZYMED laboratories, San Francisco, CA). Binding of the monoclonal antibodies to phospholipids The binding of the monoclonal antibodies to phospholipids was measured by enzyme-linked immunosorbent assay (ELISA) [26]. In brief, the wells of the microtiter plates were coated with 50 ~1 of phospholipid antigens in ethanol (10 PM) by evaporation at room temperature. The wells were blocked with a solution containing BSA (30 mg/ml and incubated with hybridoma supematants. The antibody bound was detected by biotinylated antimouse Igs (ZYMED Laboratories, San Francisco, CA) followed by incubation with peroxidase-conjugated streptavidin or alkaline phosphatase-conjugated streptavidin (ZYMED Laboratories). Optical density at 490 nm was determined by the addition of o-phenylenediamine substrate, and the relative fluorescence intensity was measured by the addition of 0.25 mM 4-methyl umbelliferyl phosphate substrate solution in a fluorescent ELISA reader (MTP-32 fluorescence microplate reader, Corona Electric, Tokyo, Japan). Throughout this study we used a 20 mM Hepes buffer containing 150 mM NaCl (pH 7.4) (Hepes-NaCl buffer) instead of phosphate-buffered saline. Inhibition of antibody binding to phospholipid Inhibition of the binding of anti-phosphatidylcholine antibodies by water-soluble haptens or choline-containing glycerophospholipids was performed as follows. 50 ~1 of the mixture containing monoclonal antibody and the various inhibitors were preincubated for 1 h at room temperature and the mixture was transferred to the microtiter wells coated with egg yolk phosphatidylcholine. The amounts of antibody bound to the plate were measured as described above. Choline chloride, phosphorylcholine chloride, L-a-glycerophosphocholine (Sigma Chemical, St. Louis, MO), and lyso-phosphatidylcholine were dissolved in Hepes-NaCl buffer and used for the experiment. The small unilamellar vesicles of phosphatidylcholine with various fatty acids were prepared by the sonication of the multilamellar vesicle
using the Branson Sonifier 185 (Branson Cleaning Equipment, Shelton, CT, U.S.A.). In brief, the dried lipid film containing 2.0 pmol of phosphatidylcholine was swollen in 1 ml of Hepes-NaCl buffer and was sonicated for 10 min under a nitrogen flow at the temperature above the phase-transition temperature of the lipids. Assay of phospholipase A, and phospholipase C Both phospholipase A, (porcine pancreas) and phospholipase C (B. cerew) assay were performed using 1-acyl-2-[‘4C]linoleoylglycerophosphocholine as substrate as previously described [27]. In phospholipase A, assay, the standard reaction mixture contained 20 PM labeled substrate, 100 mM Tris-HCl (pH 7.4) and 4 mM CaCl, in a total volume of 0.25 ml. The reaction was carried out for 10 min at 37’C and was stopped by adding 1.25 ml of Dole’s reagent [28]. The [‘4C]linoleate released was extracted by the method previously described [29]. In phospholipase C assay, the standard reaction mixture contained 20 FM labeled substrate, 4 mM CaCl,, and 50 mM Tris-HCl (pH 7.4) in a total volume of 0.4 ml. After incubation for 45 min at 37 o C, the reaction was stopped by adding 1 ml of cold methanol. All lipids were extracted by the methods of Bligh and Dyer [30] and analyzed on thin-layer chromatography (DC-Fertigplatten Kieselgel 60, Merck) with the solvent system of petroleum ether/ethyl ether/acetic acid (80 : 30 : 1, v/v). The radioactive spot corresponding to diacylglycerol was scraped off from the plate into vials and counted in a Packard Liquid Scintillation Counter. For the inhibition assay, various amounts of the anti-phosphatidylcholine monoclonal antibodies were included in the reaction mixture. Results Specificities of monoclonal antibodies against phosphatidylcholine A series of monoclonal antibodies that bound specifically to phosphatidylcholine were established after the direct immunization of the antigen into the spleens of Balb/c mice. The binding of the monoclonal antibodies to various lipid haptens were analyzed by direct binding of the monoclonal antibodies to the lipids coated on the microtiter plates (ELISA) and by inhibition analysis of the ELISA by aqueous suspensions of either lipids or water-soluble haptens. Among seven monoclonal antibodies obtained, results obtained with two typical monoclonal antibodies named JE-1 and JE-8, are described in this paper. Both monoclonal antibodies were highly specific to phosphatidylcholine and no cross-reaction with other phospholipids such as phosphatidylethanolamine, phosphatidylserine and phosphatidylinositol was observed (Fig. 1). Although JE-8 showed some weak binding to other phospholipids such as
92 OD 492
OD492
2.5
2.5
2.0
1.0
0
I
,
JE-1 2.0
1.0
. E’ ox
-
.
100
10
Antibody
0
1 Antibody
dilution
dilution
Fig. 1. Reactivities of anti-phosphatidylchohne mAbs. JE-1 and JE-8, with phospholipids. Microtiter plates were coated with 10 PM of phospholipid Ags. The mAb bound were detected with biotinylated anti-mouse Ig and streptavidin-conjugated peroxidase. Phospholipid Ags used were phosphatidylcholine (0). phosphatidylserine ( 0), phosphatidylinositol (A), phosphatidic acid (A). phosphatidylethanolamine (o), cardiolipin (H), and phosphatidylglycerol ( x ).
phosphatidylethanolamine and cardiolipin, the binding was not significant and it did not reach statistical significance. Reactivities of the monoclonal antibodies with synthetic phosphatidylcholine analogs with modified polar head groups are shown in Fig. 2. The order of the reactivity of both JE-1 and JE-8 was phosphatidylcholine Z+ dimethyl-PE > monomethyl-PE, indicating that the methyl groups on the quaternary nitrogen of the choline moiety of the phospholipids were responsible for the binding. ELISA inhibition assay In order to further analyze the interaction between the monoclonal antibodies and the phosphatidylcholine membranes, an inhibition analysis of the ELISA by small unilamellar vesicles composed of synthetic phosphatidylcholine with various fatty acid compositions was performed (Fig. 3 and Fig. 4). JE-1 showed considerable reactivity with phosphatidylcholine molecules carrying saturated fatty acids, and 50% inhibition of the
binding was observed at 100 pmol of di-18:O PC and 600 pmol of di-16:O PC, and only weak inhibition was observed with di-14:O PC (Fig. 3A). In contrast, phosphatidylcholine molecules with these saturated fatty acids failed to inhibit the binding of JE-8 (Fig. 3Bj. The reactivities of the monoclonal antibodies with phosphatidylcholine molecules with fatty acids containing one double bond in each acyl chain were next examined. Di-14:1-, di-16:1- and di-18:l PC significantly inhibited the binding of both monoclonal antibodies, and 50% inhibition of the binding of JE-1 was observed at less than 2 pmol of both di-18:1- and di-16:l PC, and at 600 pmol of di-14:l PC (Fig. 4A). In contrast to JE-1, the inhibition of the JE-8 binding increased with the decreasing acyl chain length of phosphatidylcholine molecules, and 50% inhibition of the binding of JE-8 was observed at 75 pmol of di-14:l PC, 300 pmol of di-16:l PC and 2500 pmol of di-18:l PC (Fig. 4B). The reactivities of the mAbs with the water-soluble head groups of phosphatidylcholine were also examined by inhibition of ELISA (Fig. 5). No inhibition was observed with choline chloride, while a certain extent of inhibition was observed with phosphorylcholine and L-a-glycerophosphocholine when higher concentrations of the inhibitors were employed. No appreciable inhibition was observed with other choline-containing phospholipid, lysophosphatidylcholine and sphingomyelin. Similar results were also obtained with JE-8 (data not shown).
Effect of the monoclonal antibodies on the activities of phospholipases In order to obtain further information on the interaction between anti-phosphatidylcholine antibodies and the hydrophobic moieties of the phosphatidylcholine molecule, the effect of the antibody binding on the activities of phospholipase A, (Fig. 6) and phospholipase C (Fig. 7) was studied. JE-1 inhibited the phospholipase A, (from porcine pancreas) activity dose-dependently, whereas no appreciable inhibition was observed with JE-8. Fig. 6B (Inset) shows double-recipro-
JE-1
JE-8
m
l,CH3
PC
-“:$3
DMPE
-N:$
MMPE
-NH-CH3
z
-NH2
I
1
I 3
m
PE
w
2.0
1.0 OD492
0
0
1.0
2.0
OD492
Fig. 2. Reactivities of JE-1 and JE-8 with synthetic phosphatidylcholine and its analogs. Microtiter plates were coated with 10 PM of the phospholipid Ags and the mAb bound were detected with biotinylated anti-mouse Ig and streptavidin-conjugated peroxidase. Phospholipid Ags used were phosphatidylcholine (PC), dimethyl-PE (DMPE), monomethyl-PE (MMPE), and 1.2-dipalmitoylphosphoethanolamine (PE).
93 100
JE-1
JE-8 6
C16:O
c18:0
C14:O A
0 0
2.5
9.8
39
156
2500
625
9.8
39
fp molf
156
625
2500
(p mol) Amounts
of synthetic
PC added
Fig. 3. Inhibition of mAb binding by phosphatidylcholine molecules composed of various saturated fatty acyl chains. JE-1 or JE-8 was preincubated with the vesicles composed of di-18:O PC (o), di-16:O PC (A), or di-14:O PC (H) and the mixtures were transferred to the microtiter wells coated with phosphatidylcholine. After incubation, the mAb bound was detected with biotinylated anti-mouse Igs and streptavidin-conjugated peroxidase.
JE-1 and JE-8 dose-dependently and a 50% inhibition was observed at 2 pg of JE-1 and 4 pg of JE-8 (Fig. 7).
cal plots of the kinetic data on hydrolysis of phosphatidylcho~ne in the absence or presence of various amounts of JE-1. The K, value (9.0 PM) of the enzyme increased in the presence of the antibody, whereas the Vmax value (68 pmol/min per mg) was not affected, suggesting that JE-1 competed with the active site of the enzyme for the substrate. The enzyme activity of phospholipase C (from B. cereur) was inhibited by both
Discussion
A series of mAbs that specifically react with phosphatidylcholine were established and the reactivity profiles of the two typical mAbs, designated JE-1 and
100
JET
JE-8
B
5
0 0
9.8
39
156
625
2500
9.8
(p mob
39
156
625
2500
(p mol) Amounts
of synthetic
PC added
Fig. 4. Inhibition of mAb binding by phosphatidylcholine molecules composed of various fatty acyl chains with one doubIe bond. JE-1 or JE-8 was preincubated with the vesicles composed of di-l&l PC (O), di-16:l PC (A), or di-14:1 PC (B) and the mixtures were transferred to the microtiter wells coated with phosphatidylcho~e. After incubation, the mAb bound was detected as described in Fig. 3.
94 100
Control
IgM
0
(p mol) Amounts
0
of Inhibitors
added
2.0 Amounts
4.0 of mAbs added
6.0 l,ag)
Fig. 5. Inhibition of mAb binding by various haptens. JE-1 was preincubated with soluble haptens: 0, phosphatidylcholine: t, choline: A, phosphorylcholine: n, L-a-glycerophosphocholine: A, lyso-PC: 0, SM and the mixtures were transferred to the microtiter wells coated with phosphatidylcholine. After incubation, the mAb bound was detected as described in Fig. 3.
Fig. 7. Effect of anti-phosphatidylcholine mAbs on phospholipase C activity. Phospholipase C (E. cereur) (0.04 unit) activity was measured in the presence of various amounts of JE-1 (a), JE-8 (A) or control IgM (0). The results are expressed as percentage inhibition of the enzyme activity measured in the absence of mAb.
JE-8, have been described. Both mAbs were highly specific to phosphatidylcholine and did not cross-react with other phospholipids. In the analysis using synthetic phosphatidylcholine analogs, the order of the reactivities to the mAbs is d&16:0 PC x=-dimethyl-PE > monomethyl-PE > PE, indicating the methyl groups on the quatemary nitrogen of the choline residue may play an important role in the interaction. The reactivities of mAbs which recognize phosphorylcholine have been extensively studied [31-351. Although antiphos-
phorylcholine mAbs, which were usually elicited by immunizing mice with phosphorylcholine-protein conjugates, such as phosphorylcholine-keyhole limpet hemocyanin, have also been shown to recognize a negatively charged phosphate and the trimethyl structure of the choline residue [41], the following lines of evidence indicate that the anti-phosphatidylcholine mAbs obtained in this study may be classified into the other family of the anti-phospholipid antibody. First, neither of the mAbs could react significantly with water-soluble haptens such as phosphorylcholine and L-W glycerophosphocholine. Second, the mAbs were highly specific to phosphatidylcholine and neither of them reacted appreciably with sphingomyelin when phosphorylcholine residue was substituted by ceramide. Third, the reactivities of the mAbs were highly dependent on the compositions of the fatty acyl chains of the phosphatidylcholine molecules. Fourth, anti-idiotypic antibody against JE-1 could not cross-react with the anti-phosphorylcholine mAb, TEPC 15 (unpublished observation). The interaction between mAbs and the membranes composed of various phosphatidylcholine molecules were examined by inhibition assay of ELISA. Both mAbs were shown to exhibit distinct acyl chain specificities and JE-1 interacts more strongly with the phosphatidylcholine molecules containing longer fatty acyl chains, while JE-8 interacts preferentially with the molecules containing short fatty acyl chains. JE-8 seems to prefer the phosphatidylcholine molecules in the fluid phase membranes and it could not interact with the phosphatidylcholine molecules in the tightly packed membranes. Both mAbs could interact with the water soluble head groups such as phosphorylcholine and L-a-glycerophosphocholine, indicating that both mAbs may interact with a single phosphatidylcholine mole-
100
JE-1 c
0 z a E
50
5 ae JE-8 *
control
0 0
1.0
Amounts
of mAbs
2.0 added
IgM
0
3.0 (,ug)
Fig. 6. Effect of anti-phosphatidylcholine rn4bs on phospholipase A, activity. (A) Phospholipase A, (porcine pancreas) (0.25 ng) activity was measured in the presence of various amounts of JE-1 (O), JE-8 (A) or control IgM (0). The results are expressed as percentage inhibition of the enzyme activity measured in the absence of mAb. (B) Lineweaver-Burk analysis of phospholipase A, activity incubated with JE-1. Phospholipase A, activity against increasing concentrations of phosphatidylcholine was measured in the presence of 6 pg (A), 3 pg (0) of JE-1 or in the absence of the mAb.
95
JEl
5 C”3
Fig. 8. Schematic
model of the interaction between phosphatidylcholine molecule.
anti-PC
mAbs and
cule. Since the interaction of the mAbs with soluble haptens was so weak, it is likely that a secondary interaction between JE-1 and the glycerol backbone or the fatty acyl chains of the phosphatidylcholine molecule affects the stability of the antigen-antibody complex. The interaction between the mAbs and the glycerol backbone or fatty acyl moiety of the phosphatidylcholine molecule was further supported by the inhibition analysis of the phospholipase activities by the mAbs. It is generally accepted that phospholipase C may have less of a hydrophobic character than the phospholipase A, since its site of action is in the aqueous region of the substrate aggregate and phospholipase A, needs to penetrate more deeply into the hydrophobic region of the phospholipid [36]. In the present analysis, JE-1 inhibited both the phospholipase A, and phospholipase C activity, while JE-8 inhibited only the phospholipase C activity. This observation might indicate that JE-1 may interact more thoroughly with the hydrophobic region of the phosphatidylcholine molecule than JE-8 does (Fig. 8). Further studies will be necessary for understanding the precise mechanisms of the interaction between the mAbs and the membranes. Recently, Mercolino et al [37,38] reported that mouse autoantibodies against bromelain-treated mouse erythrocytes (BrMRBC), which were mainly produced by Ly-1+ B cells in the peritoneal cavity, recognize phosphatidylcholine. Though the fine specificities of these anti-BrMRBC have not yet been studied, our anti-phosphatidylcholine mAbs showed similar characteristics with those of the anti-BrMRBC mAbs. Our mAbs not only showed strong lytic activity against the phosphatidylcholine membrane measured by the com-
plement dependent liposome lysis assay [16], but could also hemolyze intact mouse and human erythrocytes in a complement dependent manner (unpublished observation). Lytic activities were enhanced by pretreatment of the erythrocytes with trypsin or sialidase. Our antiphosphatidylcholine mAbs did not react with sphingomyelin and could not hemolyze sheep erythrocytes which the dominant phospholipid was sphingomyelin instead of phosphatidylcholine. Some of the anti-BrMRBC were reported to react only with BrMRBC and not with sheep erythrocytes, showing a similar reactivity profile with our anti-phosphatidylcholine mAbs [39]. Further genetic and idiotypic analyses of the anti-phosphatidylcholine mAbs [40,41] may clarify the relationship between these two families of mAbs. In our preliminary experiments, anti-idiotypic antibodies which were raised against JE-1 cross-reacted with phosphatidylcholine-specific lipid transfer protein purified from bovine liver (manuscript in preparation). The results may suggest that the anti-phosphatidylcholine mAbs obtained in this study will provide useful tools for the identification and purification of a family of choline-containing glycerophospholipid binding proteins which have not been purified by the conventional procedures of the protein purification. Acknowledgements The authors wish to Metropolitan Institute the monoclonal IgM ported by Grants-in-Aid Ministry of Education,
thank Dr. Akemi Suzuki (Tokyo of Medical Science) for offering antibody. This work was supfor Science Research from the Science and Culture of Japan.
References 1 Hammarstrom, S. (1983) Annu. Rev. B&hem. 52, 355-377. 2 Needleman, P., Turk, J., Jakschik, B.A., Morrison, A.R. and Lefkowith, J.B. (1986) Annu. Rev. B&hem. 55, 69-102. 3 Berridge, M.J. and Irvine, R.F. (1984) Nature 312, 315-321. 4 Braquet, P. and Rola-Pleszczynski, M. (1987) Immunology Today 8, 345-352. 5 Braquet, P., Touqui, L., Shen, T.Y. and Vargaftig, B.B. (1987) Pharmacol. Rev. 39, 97-145. 6 Tschopp, J., Schafer, S., Masson, D., Peitsch, M.C. and Heusser, C. (1989) Nature 337, 272-274. 7 Kamp, H.H., Wirtz, K.W.A., Baer, P.R., Slotboom, A.J., Rosenthal, A.F., Paltauf, F. and Van Deenen, L.L.M. (1977) Biochemistry 16, 1310-1316. 8 Backer, J.M. and Dawidowicz, E.A. (1987) Nature 327, 341-343. 9 Lee, T.C. (1985) J. Biol. Chem. 260.10952-10955. 10 Grillone, L.R., Clark, M.A., Godfrey, R.W., Stassen, F. and Crooke, S.T. (1988) J. Biol. Chem. 263, 2658-2663. 11 Pelech, S.L. and Vance, D.E. (1989) Trends in B&hem. Sci. 14, 28-30. 12 Gaulton, G.N. and Green, M.I. (1986) Annu. Rev. Imrnunol. 4, 253-280. 13 Alving, C.R. (1986) Chem. Phys. Lipids 40, 303-314.
96 14 Janoff, A. and Rauch, J. (1986) Chem. Phys. Lipids 40, 315-332. 15 Harris, E.N., Asherson, R.A. and Hughes, G.R.V. (1988) Annu. Rev. Med. 39, 261-271. 16 Miyazawa, A., Umeda, M., Horikoshi, T., Yanagisawa, K., Yoshioka, T. and Inoue, K. (1988) Mol. Immunol. 25, 1025-1031. 17 Umeda, M., Igarashi, K., Nam, KS. and Inoue, K. (1989) J. Immunol. 143, 2273-2279. 18 Doi, 0. and Nojima, S. (1971) B&him. Biophys. Acta 248, 234-244. 19 Folch, J. (1942) J. Biol. Chem. 146, 35-44. 20 Comfurius, P. and Zwaal. R.F.A. (1977) Biochim. Biophys. Acta 488, 36-42. 21 Trevelyan, W.E. (1966) J. Lipid Res. 7, 445-447. 22 Faure, M. and Marechal, J. (1962) Compt. Rend. 254, 4518-4520. 23 Umeda, M., Diego, I. and Marcus, D.M. (1986) J. Immunol. 137, 3264-3269. 24 Galfre, G.S., Howe, C.. Milstein, C., Butcher, G.W. and Howard. J.C. (1977) Nature 266, 550-552. 25 Kohler, G. and Milstein, C. (1975) Nature 256, 495-497. 26 Umeda, M., Diego, I., Ball, E.D. and Marcus, D.M. (1986) J. Immunol. 136, 2562-2567. 27 Horigome, K., Hayakawa, M., Inoue, K. and Nojima, S. (1987) J. B&hem. 101, 625-631. 28 Dole, V.P. and Meinertz, H. (1960) J. Biol. Chem. 235, 2595-2599. 29 Arai, H.. Inoue, K., Natori, T., Banno, Y., Nozawa, Y. and Nojima, S. (1985) J. Biochem. 97, 1525-1532.
30 Bligh, E.G. and Dyer. 911-917.
W.J. (1959) Can. J. Biochem.
31 Niedieck, chemistry
U.
B., Kuck, 15, 471-475.
and
Gardemin,
H.
(1978)
Physiol.
37,
Immuno-
32 Andres, C.M., Maddalena, A.. Hudak, S.. Young, N.M. and Claflin, J.L. (1981) J. Exp. Med. 154, 1584-1589. 33 Novotny, J., Bruccoleri, R., Newell, J., Murphy, D., Haber, E. and Karplus. M. (1983) J. Biol. Chem. 258, 14433-14437. 34 Lai, E.H.C., Kabat, E.A., Meienhofer, J.. Heimer, E.P., Olson, A.J. and Lemer, R. (1987) Nature 325, 168-171. 35 Bruderer, U.. Stenzel-Poore, M.P., Bachinger, H.P., Fellman, J.H. and Rittenberg, M.B. (1989) Mol. Immunol. 26, 63-71. 36 Waite. M. (1987) in The Phospholipases (Waite, M., ed.), pp. 191-241, Plenum Press. New York. 37 Mercolino, T.J., Arnold. L.W. and Haughton. G. (1986) J. Exp. Med. 163, 155-165. 38 Mercolino, T.J., Arnold, L.W., Hawkins, L.A. and Hat&ton, G. (1988) J. Exp. Med. 168. 687-698. 39 Pennell, CA., Arnold, L.W., Lutz, P.M.. LoCascio, N.J., Willoughby, P.B. and Hat&ton. G. (1985) Proc. Natl. Acad. Sci. USA 82. 3799-3803. 40 Mercolino, T.J., Locke, A.L., Afshari, A., Sasser, D.. Travis, W.W., Arnold. L.W. and Haughton, G. (1989) J. Exp. Med. 169, 18691877. 41 Kimura, H., Umeda, 142, 3477-3481.
M. and Marcus,
D.M.
(1989) J. Immunol.
