Conformational differences of human IgE on hydrophobic and hydrophilic solid phases detected by monoclonal antibodies

MolecularImmunology, Vol. 24,No. 11,pp. 1129-I134,1987 Printedin Great Britain.

0161-5890/87 $3.00+ 0.00 0 1987PergamonJournalsLtd

CONFORMATIONAL DIFFERENCES OF HUMAN IgE ON HYDROPHOBIC AND HYDROPHILIC SOLID PHASES DETECTED BY MONOCLONAL ANTIBODIES FRANCINE ANFOSSO,* CHRISTIAN DEMEZURE,MICHEL DELAAGE, ROSE CHEBALLAH, FRANCOISE BELLOT~

and ALAIN BOURG~IS$

Immunotech, case 915, 13288 Marseille Cedex 9, France (First received 10 October

1986; accepted in revised form 3 April 1987)

Abstract-Very sensitive assays of IgE are required for determining prevalence of allergic reactions in children. In order to develop a sensitive two-site IRMA two kinds of mAb were produced. Antibodies specific for D, 1 determinants were derived from immunization with a 40 kDa papain Fc fragment. They bound equally native and 56°C heated IgE. D,2 specific mAb were obtained after immunization with IgE anti-D, 1 complex and were selected on the basis of their inability to bind heated IgE. In a two-site assay on plastic plates, D, 1 specific mAb led to the binding of IgE but always prevented further binding of anti-D,1 n&b, anti-human K chain mAb or allergen on bound IgE. This was not true when CNBr activated cellulose was used. The influence of the nature of the solid phase disappeared when D,2 specific mAb were coated on plastic tubes. In this case, the binding of a second mAb with identical or different fine specificity was observed. The best matched pair was E 164 (anti-D,2) on the solid phase and 6HlO (anti-D, 1) as a tracer. As little as 0.2 UI/ml of IgE could be detected in a 2 hr test.

INTRODUaION

Immunoglobulins E (IgE) display a central role in immediate hypersensitivity reactions (Bennich and Johansson, 1971). They bind with high affinity to Fc receptors on basophil granulocytes and mast cells through their Fc portion. In this region two antigenic areas, D, 1 and D,2, have been described. They differ

by their sensitivity to heat denaturation. D,2, shared by the C, 3 and C,4 domains of the Fc portion of IgE antibodies is involved in the biological activity and is sensitive to heating at 56°C during 4 hr in contrast to D,l (Ishizaka et al., 1970; Geha et al., 1985; Rousseaux-Prevost et al., 1986). The IgE level in plasma is an important marker of allergy. Since IgE rarely exceeds 5 pg/ml in serum of highly allergic patients (Pecoud et al., 1982) the IgE assay requires sensitive methods. The most commonly used are the paper radioimmunosorbent test (PRIST, Pharmacia) or enzymo-immunoassays (Stein et al., 1983; Bradley-Eisenbray et al., 1983; Ceska and *Author to whom reprint requests should be sent at present address: INSERM U270, Faculte de Medecine Nord, Bd Pierre Dramard, 13326 Marseille Cedex 15, France. tPresent address: REVLON Biotechnology Research Center, MELOY Laboratories. 2501 Research Bd, Rockville, Maryland, U.S.A. $Present address: Centre d’Immunologie de Marseille Luminv. case 906. 13288 Marseille Cedex 9, France. Abbreviations: mAb, ‘monoclonal antibodies; Ig; immunoglobulin; IgE, immunoglobulin E; Fc, IgE Fc fragment; individual human myeloma IRE having K light IaE,,,. chams; IgE, individual human myioma IgE having I light chains; i.p., intraperitoneally; FPLC, fast protein liquid chromatography; IRMA, immunoradiometric assay. )..,~

,

Lundqvist, 1972; Subba Rao et al., 1983; Walsh et al., 1984; Magnusson et al., 1985). In this report we describe production and characterization of monoclonal antibodies (mAb) against the two different Fc antigenic regions. A very sensitive two-sites radioimmunometric assay of IgE was produced with two antibodies, one being directed towards the D, 1 antigenic region, and the other towards the D,2 region. Studies presented here showed that the binding of mAb to IgE is dependent on their fine specificity and on the nature of the solid phase. MATERIALS AND METHODS Animals, immunoglobulins

BALB/c female mice were obtained from IFA-CREDO. Purified human myeloma IgE (K) and IgE (1) were kindly provided by Dr H. Bazin (Brussels, Belgium) and Dr Capron (Lille, France) respectively. Labelled anti E-specific antiserum was purchased from Pharmacia (Sweden). Rat mAb against mouse K chain, mouse IgG, and IgG,, were given by Immunotech (Marseille, France). Culture media

All cell cultures were in Dulbecco’s modified Eagle medium (DMEM) supplemented with non essential amino acids (0.1 mM) sodium pyruvate (1 mM), penicillin-streptomycin glutamine (2 mM), (lOOU/ml) and lO--20% fetal calf serum (FCS). Hybridoma production Fusion 1. 25 mg of human IgE (rc) were digested with 250 pg of crystallized papain (Sigma) for 4 hr

1129

1130

FRANCME ANPOSWet al.

(Ishizaka et al., 1970). The reaction was stopped with iodoacetamide. In a preliminary experiment, the extent of the digestion was investigated by a 5% SDS polyacrylamide gel electrophoresis using human IgE, and L chains as markers. The Fc fragment was isolated by FPLC (Pharmacia) using a preparative mono Q column with a triethylamine (TEA)-NaCl gradient. Buffer A (TEA 20 mM, pH = 7.6) was run for 2 min. Then NaCl (1 M) was progressively introduced in TEA buffer in order to reach 0.35 M NaCl at 13 min and 1 M NaCl at 15 min. The antigenic reactivity in each fraction was determined by IRMA using the PRIST assay. Fraction 29 was used for the immunization. BALB/c mice were i.p. injected with 6Opg of this fraction in complete Freund’s adjuvant (CFA) on days -90, -60, - 30 and intravenously on day - 3. On day 0 mouse spleen cells were hybridized with P3 X63 Ag8 mouse myeloma cells at a 5 to 1 ratio as described by Galfrt and Milstein (1981). Fusion 2. On day -30, BALB/c mice were immunized with Ag-Ab immune complexes consisting of 7Opg of IgE(I) incubated overnight at 4°C with 200 pg of 6H10 mAb obtained from Fusion 1. Immunization was carried out by multiple sites id. injec-

Table 1. Characteristics

Clones

Subclass

6HlO 5Gll 2E6 El64 El27 El24

IgG2a IgG*, :gp IgG, IeG.

Direct” IgE binding (%) 53.1 55.8 50.2 57.3 55.7 53.2

tions. Animals were boosted on day - 3. On day 0 the fusion was performed. Hybridoma screening tests Direct binding assay. IgE(IZ) (10 pg/ml) was coated on 96 wells polystyrene flexible plates (Falcon) and incubated with 50 ~1 of culture supernatants for 2 hr at room temperature. Plates were washed with phosphate buffer saline (PBS) 0.1 A4, pH = 7.2, Tween 20 O.Ol%, and then 50 ,~l of ‘251-labelled goat anti-mouse Ig (GAMIG) were added for 2 hr. After three washes with PBS-Tween, bound radioactivity was counted in a gamma counter. Indirect binding assay. Rat mAb against mouse rc chain (10 pg/ml) was coated on 96 wells flexible plates, and incubated with 50 ~1 of hybridoma supernatants for 2 hr. After washing, 50 ~1 of ‘ZSI-labelled IgE(k) either native or heated (4 hr at 56°C) were incubated for 2 hr. The plates were washed and the radioactivity counted. Specific IgE binding hybridomas were cloned by limiting dilutions, screened and again subcloned and injected into mice previously sensitized with pristane (Aldrich). Titers of ascitic fluid were determined by analytical FPLC and IRMA.

of IgE specific mono&ma1 antibodies Heated’ IgE binding W) 51.5 52.3 48.7 5.3 8.5 7.5

IgG, or 1 Bence-Jones binding

Affinity’ constant M-1

0 0 0 0 0 0

2.5 x IO9 2.2 x 109 not done 3.3 x 109 0.15 x 109 2.5 x lo9

“‘251-Labelled IgE (K) was added (50,OOOcpm) to hybridroma supematant complexed to anti-rat K chain mAb coated on flexible plastic plates. “2SI-Labelled IgE(x) was heated for 4 hr at 56°C and added (50,000 cpm) as referred above. ‘Affinity constants were determined by Scatchard’s plot.

Table 2. Different kE bindine assaw Binding of radioiodinated Coated” solid phases A anti-D, 1 on plastic 6H10 5Gll anti-D,1 on BrCN activated paper discs 6HlO 5Gll B anti-D,2 on plastic El64 El24 anti-D,2 on BrCN activated paper discs El64 El24

tracer (B/T%)

IgE direc? binding assay W)

Anti-D, 1 (6HlO)

Anti-D,2 (El64)

53.1 55.8

6.0 5.4

58.2 49.4

2.2 2.4

2.6 3.1

48.7 41.6

33.6 27.5

42.1 39.5

23.2 21.4

27.6 32.3

57.3 53.2

55.3 48.7

58.5 51.3

22.7 21.9

28.2 22.7

50.3 43.5

47.4 39.6

51.2 44.6

25.2 22.6

35.8 30.9

Anti-human K chain

Grass‘+ allergen

__

‘Polyvinyl plates or paper discs were coated with 10 pg/ml of mAb. “2sI-Labelled IgE, was added to mAb bound to the solid phases. ‘IgE, (250 ng/ml) in PBSBSA was incubated with mAb solid phase for 2 hr, washed and then radioiodinated tracer (50,000 cpm) was added for 2 hr. adioiodinated Phleum pratense extract.

Anti-human IgE monoclonal antibodies Isotype determinations

They were performed by incubating hybridoma supernatants with IgE coated plates. Binding was revealed by addition of “‘1 rat mAb specific for mouse IgG, , IgG,, and IgGr,, . PuriJication

Murine IgG, from ascitic fluids were isolated by Protein-A Sepharose (Pharmacia) affinity chromatography. They were eluted with glycine-HCl buffer, 0.1 &f, pH = 2.8. IgG, mAb were purified by a modified procedure of the ammonium sulfate precipitation technique. At 4°C one volume of the sample was diluted with one volume of PBS, and two volumes of 3.2 M ammonium sulfate (pH = 7.2) added drop by drop and then immediately centrifuged for 5 min. The precipitate was rapidly solubilized with PBS and extensively dialyzed. Cross competition experiments

Biotinylated IgE (100 ng/ml) was coated on Immunotech solid phase (Patent No. 844 30010.3) then incubated with radio-iodinated mAb (20,000 cpm/ml) plus increasing doses of unlabelled mAb. After overnight incubation, the tubes were washed and the radioactivity counted. Immunoradiometric assays Coated tubes. Biotinylated

r

1131

V

T I

Fraction number

Fig. 1. FPLC purification of a Fc fragment from human myeloma IgE (c, K) after papain digestion. The antigenic reactivity was tested using the PRIST assay.

is only a part of the Fc portion. It was the immunogen used in the Fusion 1. Anti-human IgE mAb selection

anti-D,1 or D,2 mAb After screening of 423 hybridoma supematants (0.5 pg/ml) were coated on Immunotech solid phase. from the first fusion, 10 different clones were subVarious concns of human IgE (100 p 1) in PBSBSA 0.2% were added for 1 hr. After washing with PBS 100 ~1 of lZ51-mAb (specific for human D, 1 or D,2) were added for 1 hr and bound radioactivity was determined. CNBr activated cellulose paper. Cellulose paper discs (Schleicher-Schull No. 589, 0.5 cm diameter) were CNBr activated (Ceska and Lundqvist, 1972) and coated with 10 pg/disc of mAb. IgE levels were determined as described above. Coated plates. Anti-IgE mAbs (lOpg/ml) were coated on 96 wells polystyrene flexible plates (Falcon) and incubated with 50 ~1 of human IgE solutions for 2hr at room temp. After washing as previously described, plates were incubated with 50 ~1 of 1251 mAb (specific for human D, 1, D, 2 or K chain) or ‘25I allergen for 2 hr. After three washes, bound radioactivity was counted. RESULTS

Preparation of Fc fragment

Digestion of human IgE with papain for 4 hr produced several fragments isolated by FPLC. Results in Fig. 1 showed that fractions 27 and 29 firmly bound to polyclonal antiserum specific for human Fc,. Fraction 29 which possessed the highest binding activity was submitted to SDS gel electrophoresis. It migrated as a 40 kDa fragment (Fig. 2) suggesting it

Fig. 2. SDS polyacrylamide gel electrophoresis of native (A) or reduced (B) human IgE (e : 70,000 Da; K: 23,000 Da) and of fraction 29 of the FPLC uurification (0.

FRANCINE ANFOSSO er al.

assay of human IgE (IgE-IRMA), different pairs of D, 1 or D,2 specific mAbs were studied. When D, 1 specific mAbs were coated on plastic plates, they firmly bound radiolabelled human IgE. Nevertheless after incubation of cold IgE, the binding of a radioiodinated second antibody was very low when anti-D,1 or anti-K chain mAbs were used (Table 2-A). Furthermore when IgE from allergic sera were added to the coated tubes, no binding of radioiodinated specific allergen could be observed. In contrast when D, 1 specific mAb was coated to CNBr activated cellulose discs, IgE two sites IRMA could be established, whatever the specificity of the tracer. When IgE-IRMA was attempted with D,2 specific mAb coated on a plastic solid phase, the assay has always been established. Radioiodinated tracers specific for D,2, D,l or human K chain bound with a high avidity (Table 2-B) and a radioiodinated allergen (Phleum pratense) also bound to the specific IgE linked to the immune solid phase. Similar results were observed when D,2 specific mAbs were coupled to CNBr activated cellulose discs. noradiometric

20 t._ 0

l-•w

I 05

I I

.-.m.w I 2

I 3

Unlabelled mAb (W/ml)

Fig. 3. Cross-competition experiments between D, 1 and D,2 specific mAbs and IgE coated tubes: (a) Radioiodinated D, 1 specific6HlOmAb in competition with D,l specific mAb: 6H10 (O-O), 5Gll (m-m) and 2E6 (O-_O), and with D,2 specific mAb El64 (0-O). (b) Radiolabelled D,2 specific El64 mAb in competition with D,2 specific mAb: El64 (O-O), El24 (A--&, El27 (A-A) and with a D,l specific mAb. 6HlO (0-O).

cloned and three of them (5Gl1, 2E6, 6HlO) were selected for their high capacity to selectively bind to human IgE (K or 1). This binding was not affected by heating IgE at 56°C for 4 hr, suggesting that these mAb were specific for D, 1 epitopes (Table 1). They did not bind human IgG or Bence-Jones proteins. In order to obtain mAb directed against other sites on the Fc fragment, another fusion was performed. D,l sites on IgE (K) were blocked by an excess of 6HlO mAb obtained from Fusion 1. This Ag-Ab complex was injected into BALB/c mice. Selection of anti-light chain and anti-idiotype antibodies was avoided during screening tests by using a distinct myeloma IgE protein (K). Out of 13 clones producing

Topographic relationship of antigenic determinants defined by mAbs

The study of native and heat denatured IgE binding to the mAb produced in both fusions distinguished two types of mAb (anti-D,1 and -D,2). This classification study was completed by crosscompetitive experiments. Each individual mAb was labelled and incubated with all unlabelled mAbs at increasing concentrations in IgE coated tubes. Figure 3 presents the results obtained for radiolabelled anti-

highly specific human anti-IgE mAb, seven recognized determinants borne by the D,2 region as shown by their inability to bind heated IgE (bound heated

IgE < lo%), two others showed an intermediate pattern and three expressed the anti-D, 1 specificity (bound heated IgE> 40%). Three clones (E164, E127, and E124) producing anti-D,2 antibodies were selected. They did not bind human IgG or Bence-Jones light chains (Table 1). The affinity constants of the anti-D,2 and anti-D,1 mAbs selected were determined by equilibrium inhibition assay. From all the mAbs, El64 gave the highest direct binding of intact IgE (KA = 3.3 x 109MA1) and the lowest one when IgE was heat-denatured. D,l and D,2 specific mAb behave differently on plastic solid phases In an attempt

to perform

a two-site

immu-

IgE cone. (kIU/L)

Fig. 4. Two-sites IgE IRMA using different combinations of D,1(6HlO, SGll)andD,2(E164,E124)specificmAb: El64 on tlte tubes and 6HlO (A-A) or El24 (l-J--t-~) as tracers, SGll on the tubes and radioiodinated El24 (0-O). IgE content was expressed in International Units (1 \unit = 2.4 ng).

Anti-human IgE monoclonal antibodies

1133

previously observed with passive immunization (Thalamer and Freund, 1985). Assays for total IgE measurement have been performed until now with anti-D,2 specific antibodies (Chandler et al., 1983; Sanchez Madrid et al., 1984) which also gave better RAST results than anti-l), 1 (Pecoud et al., 1982). Interesting observations have been obtained when we introduced D, 1 specific mAb in the assay. IgE seems to have different spatial Two-sites immunoradiometric IgE assay conformations depending on the nature of the solid In order to establish an IgE-IRMA, several pairs phase and the specificity of mAb. When D, 1 specific mAb were adsorbed on plastic plates, the recognition of antibodies were tested (Fig. 4). Pairs constituted of IgE by a second mAb possessing a similar with anti-D, 1 and anti-D,2 antibodies gave the best specificity (anti-D,l) was hardly observed. This was sensitivity; El64 and 6HlO mAb were thus selected. not true when CNBr activated cellulose discs were Immunotech solid phase was coated with El64 mAb. Binding of IgE was revealed by adding 1251-labelled used. Incubation of IgE on plastic solid phase coupled with anti-D, 1 mAb results in the inaccessibility 6HlO mAb. This combination allowed a very sensitive two-steps assay detecting 0.2 III/ml. Addition of of the symmetrical D, 1 domain and the antigen binding fragments (radioiodinated anti-human K serum did not modify the background nor the slope chain mAb or allergen were unable to bind to the of the curve (data not shown). immune solid-phase IgE complex). The influence of the nature of the solid phase disappears when D,2 specific mAb were either coated DISCUSSION to the plastic or used as tracers. Immobilization of The Fc part of IgE contains two distinct antigenic proteins on solid phase has been widely used for regions D,l and D,2. D,2 determinants, shared by developing RIA procedures (Hermann, 1981). In the C,3 and C,4 domains have been shown to be some cases the solid phase may preferentially obscure sensitive to heat denaturation and to disulfide bonds some epitopes (Stahli et al., 1983; Stevens et al., reduction and to be involved in the binding of IgE to 1986). In the present study, hydrophilic and hydroreceptors on mast cells and basophils (Ishizaka et al., phobic phases were compared. The hydrophilic phase 1970; Ishizaka and Ishizaka, 1978). Recent studies did not obscure the epitopes recognized by our mAb. (Rousseaux-Prevost et al., 1986) have demonstrated On the other hand the hydrophobic phase in certain that these properties are restricted to the C,3 domain cases (anti-D,1 mAb on the solid phase) completely in the rat. The heat sensitivity of IgE was a useful test abolished the reactivity of large parts of the molecule for screening supernatant hybridomas in this work. (D,l determinant, K chain determinant, antibody binding site). We thus defined two kinds of mAb: Several hypotheses could explain these differences: -those directed against D, 1 sites, which equally (a) The immunogen used for Fusion 1 (Fc, papainic bind native or heated IgE fragment) could be partially denatured (Kochwa et -those, specific for D,2 sites, which bind native al., 1971; Dorrington and Bennich, 1978; RousseauxIgE but display almost no reactivity toward heated Prevost et al., 1984) and could then produce antiIgE. Combination of D, 1 and D,2 specific mAb allows bodies able to induce conformational changes on IgE, such as externalization of hydrophobic regions interthe development of a very sensitive two-sites IRMA for human IgE and offers two main advantages acting with the plastic phase but not with the hydrocompared to currently available immunoassays: time philic phases. (b) Data from segmental flexibility (Oi et al., 1984) required for performing the assay is reduced to less suggested that murine IgE is the least flexible immuthan 2 hr; the sensitivity is lowered to 0.2 UI/ml. Such noglobulin. Nevertheless conformational studies low levels of serum IgE are usually found in newborns and their measurement is very important to (Dorrington and Bennich, 1978; Pumphrey et al., 1986) indicated that the C,3 and C,4 domains are predict and prevent the development of allergic reacloose. On the opposite the C,2 domain is much more tions (Croner et al., 1982; Michel et al., 1980). The rigid and when an anti-D, 1 mAb is bound to this affinity constants of the mAb produced in these domain the Fab regions and the basal part of the Fc fusions and their fine specificities can explain this region could be set tightly on the solid phase. This high sensitivity. could be emphasized when the hydrophobic plastic The first fusion gave only anti-D, 1 mAb. In order phase is used, thus inhibiting the interaction of IgE to obtain anti-D,2 mAb, the immunogen used in the second fusion was total IgE complexed with anti-D, 1 with another anti-D,1 mAb, an anti-K mAb or an allergen. mAb. More than 50% of selected mAb are D,2 (c) The localization of the D, 1 antigenic sites could specific. Thus this procedure seems to favour the allow the simultaneous binding of both D, 1 sites of expression of chosen determinants in the same way

D, 1 (6HlO) and anti-D,2 (E164) mAb. The anti-D, 1 (6HlO) binding to IgE was not inhibited by the anti-D,2 mAb but was substantially inhibited by all the anti-D, 1 monoclonal antibodies tested. On the other hand the anti-D,2 (E164) binding to IgE was inhibited by El27 anti-D,2 but not by El24 anti-D,2 clearly suggesting that the D,2 region comprises at least two different epitopes.

FRANCINEANFOSSO et al.

1134

one IgE by both anti-D, 1 sites of a single monoclonal

antibody leading to a total stabilization of the IgE molecule. If, nevertheless, the hydrophilic support to which the antibody is attached preciudes this double binding, the IgE moleucle could keep enough freedom to allow further attachment of either another antibody (anti-D, 1, anti+) or an allergen. The fusions presented here allowed the production of anti-D, 1 mAb directed toward a single epitope and anti-D,2 mAb directed toward two different epitopes. Those antibodies will be useful to study the interaction of IgE with its specific receptors.

release from rat mast cells by divalent antibodies against IgE receptors. J. Zmmun. 120, 800-805. Kochwa S., Terry W., Capra J. and Yang N. (1971) Structural studies of immunogIobulin E: physicochemical studies of the IgE molecule. Ann. A? Y. Acad. Sci. 190, 49-70.

Magnusson C. and Masson P. (1985) Immunoglobulin E assayed after pepsin digestion by an automated and highly sensitive particle counting immunoassay: application to human cord blood. J. Allergy Clin. Ztnmun. 75, 513-524. Michel F. B., Bousquet J., Greilier P., Robinet Levy M. and Coulomb Y. (1980) Comparison of cord blood immunoglobulin E concentrations and maternal allergy to the prediction of atopic diseases in infancy. J. Allergy Clin. Zmmun. 65, 422-428.

Acknowledgemen&--The authors are grateful to Drs H. Bazin (Brussels, Belgium) and A. Capron (Lille, France) for the gift of myeloma IgE, to Drs M. Him, C. Labit, G. Legal1 la Salle for skilful contribution to the fusions and to Mm M. Delezay for expert technical assistance.

REFERENCES Bennich H. and Johansson S.G.O. (1971) Structure and function of human immunoglobulin E. Adv. Zmmun. 13, 1-55.

Bradley-Eisenbray A., Krebs T., Dunnette S. and Gleich G. (1983) Comparison of ultrasensitive methods for the measurement of IgE. J. Zmmun. Meth. 58, 365-373. Ceska M. and Lundkvist U. (1972) A new and simple radioimmunoassay method for the determination of IgE. Immunochemistry

9, 1021-1030.

Chandler H., Coulter A., Healey K., Komitschuk M,, McGregor A. and Hurrel J. (1983) Mon~lonal hvbridoma antibodies against human IgE and their use in a rapid and sensitive enzyme immunoassay for their semiquantitative assessment of total 1gE levels in human blood. Znt. Archs Allergy appl. Zmmun.12, 267-272. Croner S., Kjellman N., Eriksson B. and Roth A. (1982) IgE screening in 1701 newborn infants and the development of atopic disease during infancy. Arch. Dis. Child !!7, 364-370.

Dorrington K. and Bennich H. (1978) Structurefunction relationships in human immunoglobulin E. Zmmun. Reo. 41, 1-25.

Galfrb G. and Milstein C. (1981) Preparation of monoclonal antibodies: strategies and procedures. In Methods in Enzymology {Edited by Colowick S. and Kaplan N.), 73, 3-46. Academic Press, New York. Geha R., Heim B. and Gould H. (1985) Inhibition of the Praustnitz-Kiistner reaction by an immunoglobulin-chain fragment synthesized in E. coli. Nature 315, 577-578. Hermann J. E. (1981) Quantitation of antibodies immobilized on plastics. Meth. Enzym. 73, 239-244. Ishizaka K., Ishizaka T. and Lee E. (1970) Biologic function of the Fc fragments of E myeloma protein. Zmmw?ochemistry 7, 687-702.

Ishizaka T. and Ishizaka K. (1978) Triggering histamine

Oi V. T., Vuong T. M., Hardy R., Reidler J., Daugl J., Herzenberg C. and Stryer L. (1984) Correlation between segmental flexibility and effector function of antibodies. Nature 307, 136240. Pecoud A., Ochsner M., Arrendahl H. and Frei C. (1982) Improvement of the RAST sensitivity by using an antibody specific for the determinant D,,. Clin. Allergy 7, 195-199.

Pumphrey R. S. H. (1986) Computer models of the human immunoglobulins. Zmmun. Today 7, 206211. Rousseaux-Prevost R., Rousseau J., Bazin H. and Biserte G. (1984) Differential reduction of the inter chain disutfide bonds of rat immunoglobulin E. Molec. Zmmun. 21, 2333241. Rousseaux-Prevost R., Rousseaux J. and Bazin H. (1986) Studies of the IgE binding sites to rat mast cell receptor with protelytic fragments and with a monoclonal antibody against epsilon heavy chain. (in press). S~chez-Madrid F., Morago G., Corbi A. and Carreira J. (1984) Monoclonai antibodies to three distinct epitopes on human IgE their use for determination of allergenspecific IgE. .Z. Zmmun. Meth. 73, 367-378. Stahli C., Miggiano V., Stocker J., Staehelin Th., Hiiring P. and Takacs B. (1983) Distinction of epitopes by monoclond antibodies. Meth Enzvm. 92. 242-253. Stein R., Evans S., Milner R.: Rand C. and Dolovich J. (1983) Isotopic and en~matic IgE assays in non-allergic subjects. Allergy 38, 389-398. Stevens F., Jwo J., Carperos W., Kiihler H. and Schiffer M. (1986) Relationships between liquid and solid-phase antibody association characteristics: implications for the use of competitive ELISA techniques to map the spatial location of idiotopes. J. Zmmun. 1937-1944. Subba Rao P., McCartney-Francis N. and Metcalfe D. (1983) An avidin-biotin micro ELISA for rapid measurement of total and allergen specific human IgE. J. Zmmun. Meth. 57, 71-85.

Thallamer J. and Freund J. (1985) Passive immunization: a method of enhancing the immune response against antigen mixtures. J. Zmmun. Meth. 80, 7-13. Walsh B., Wrigley C. and Baldo B. (1984) Simultaneous detection of IgE binding to several allergens using a nitr~ellulo~ “polydisc”. Znr. Archs Allergy appl. Zmmun. 66, 99-102.