Eur. J. Immunol. 1990. 20: 2089-2094
Frangoise RossiO, Brigitte Guilberta, CBcile Tonnelle+, ThBrkse TernynckA, Frangois Fumoux+, Stratis AvrameasA and Michael D. KazatchkineO Unit6 d’Immunopathologie and INSERM U 28O, H6pital Broussais, Pans, Unit6 d’hmunocytochimie URA CNRS 35P, Dhpartement d’Immunologie, Institut Pasteur, Pans and Centre d’Immunologie INSERM-CNRS de Marseille-Luminy , Marseille +
Idiotypic interactions between human IgG and natural IgM antibodies
Idiotypic interactions between normal human polyspecific IgG and natural IgM antibodies* Pooled normal human polyspecific IgG (IVIg) contain anti-idiotypes against a variety of autoantibodies from patients with autoimmune diseases and IgG autoantibodies present in IVIg. The present study indicates that IVIg may also react through idiotypidanti-idiotypic interactions with human natural IgM antibodies. Sixty-four percent of IgM secreted by B lymphoid cell lines derived from B cells of healthy elderly donors and 18% of IgM secreted by cloned EBV-transformed cord B cells that were tested, bound through their variable region to F(ab’)z fragments of 1VIg.The binding to 2,4,6-trinitrophenyl (TNP) of a polyreactive IgM with anti-TNP specificity, was inhibited by F(ab‘)z fragments from IVIg, indicating the presence in IVIg of anti-idiotypes that may interfere with the antibody-combining site of polyreactive IgM antibodies. The ability of IgM antibodies to interact with idiotypes on IVIg was not related to the degree of polyreactivity of natural antibodies. Our observations further document that IVIg contain antibody specificities against Ig from normal individuals and suggest that IgG originating from the physiologically expressed repertoire may modulate the expression of the potential B cell repertoire.The results may be relevant to the suppressive effect of IVIg in autoimmune diseases.
1 Introduction Sera from healthy individuals contain polyreactive IgM antibodies and IgG antibodies that recognize a variety of self-related antigens [11. B cells synthesizing polyreactive antibodies against self and non-self antigens are found in spleens from nonimmunized [2] and from immunized [3] animals. In mice, a subset of natural polyreactive IgM antibodies exhibits a high degree of autoreactivity [4] and V region connectivity [5]. Intravenous immunoglobulins (IVIg) are therapeutic preparations of IgG prepared from the plasma of a large number of healthy donors. Infusion of IVIg has resulted in the decrease of autoantibody titer in several autoimmune diseases including anti-factor VIII autoimmune disease [6], chronic demyelinating inflammatory polyneuropathy [7], vasculitis with circulating anti-neutrophil cytoplasmic antigen antibodies (D. R. W. Jayne, unpublished observations), and in patients with recurrent abortion and anticardiolipin antibodies [8]. IVIg contain anti-Id against a variety of autoantibodies associated with autoimmune disease [9]. We have, thus, postulated that the beneficial
[I 85131
*
2089
This work was supported by Institut National et de la Sant6 de la Recherche Mtdicale (INSERM), Centre National de la Recherche Scientifique (CNRS) and Sandoz SA (Basel, Switzerland).
Correspondence: Frangoise Rossi, Unit6 d’Immunopathologie et INSERM U28, Hapita1 Broussais, 96, rue Didot, F-75674 Pans C6dex 14, France Abbreviations: Act: Actin Alb: Human albumin Am-OVA: OVA-coupled arsonate Id: Idiotypehdiotypic IVIg: Intravenous Ig Mg: Myoglobulin Myo: Myosin PC-OVA: OVAcoupled phosphorylcholine Phox-OVA: OVA-coupled phenyloxazalone PO: Peroxidase Spe: Spectrin Tf: Transferrin Tg: Thyroglobulin ’hkTbbulin 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
effect of IVIg in autoimmune diseases depends on the ability of IVIg to interfere with the regulatory function of the Id network on autoimmune responses [ 101. The present study indicates that IVIg may interact through Idanti-Id recognition with human natural polyreactive IgM antibodies. Id interactions between IgG and natural IgM antibodies may play a role in modulating the expression of the B cell repertoire in normal and diseased individuals.
2 Materials and methods 2.1 Antigens BSA,whole skeletal muscle myoglobin (Mg), KLH, human thyroglobulin (Tg), ribonuclease, human microsomes, calf thymus native double-stranded DNA and trinitrobenzene sulfonic acid (Sigma Chemical Co., St. Louis, MO), human transferrin (Tf; Behringwerke AG., Marburg, FRG), human albumin (Alb; Schwartz Mann, Cambridge, MA) and horseradish peroxidase (grade I, R Z = 3) (PO; Boehringer-Mannheim, Mannheim, FRG) were purchased as indicated. Calf muscle actin (Act) [ll]and myosin (Myo) [12], pig brain tubulin (Tub) [13], TNP-25/BSA [14], OVA (Sigma), TNP-OVA [141, arsonate-OVA (Ars-OVA) [15], phosphorylcholine-OVA (PC-OVA) [161 and phenyloxazoloneOVA (Phox-OVA) [17] were obtained as described. Human RBC spectrin (Spe) was a kind gift from Dr. E. Lazarides (Caltech, Pasadena, CA). Hemoglobin was prepared from human RBC using a 0.83% ammonium chloride lysis buffer. 2.2 IgM antibodies from EBV-transformed human B cells We studied IgM antibodies secreted by human EBVtransformed lymphoid B cell lines and EBV-transformed 0014-2980/90/0909-2089$3.50 + .25/0
2090
F. Rossi, B. Guilbert, C. Tonnelle et al.
Eur. J. Immunol. 1990. 20: 2089-2094
cloned B cells. SN were obtained from cultures of 59 cloned EBV-transformed human B cell lines raised from PBL from normal elderly individuals [MI. All cell lines secreted exclusively or predominantly Ig of the IgM isotype. The specificity of secreted IgM was tested on a panel of 12 antigens: Act,Tub, Mg, Spe, DNA,TNP, Alb, PO, Myo, Tf,Tg and KLH. Six cell lines were found to secrete monospecific antibody that recognized only one antigen of the panel; 26 cell lines secreted polyreactive antibodies; 27 cell lines had no detectable antibody specificity.
Plates were coated with F(ab’)2 from IVIg at a concentration of 100 pg/ml. A positive control was included in each ELISA plate, which consisted of a purified human monoclonal IgM that was found to bind to F(ab‘)2 from 1VIg.The revealing reagent was anti-Fcp antibody that had been extensively adsorbed with human F(ab’)s fragments and showed no reactivity with F(ab’)z. Absorbance values of the binding of IgM to insolubilized IVIg were corrected for IgM concentrations.
B cell clones (n = 135) were obtained by immortalizing human cord B cells and peripheral blood B cells from healthy adult individuals with EBV [19]. The specificity of the monoclonal IgM secreted by cloned cells was tested on a panel of nine antigens: Act,Tub, Myo, DNA,TNF’, Mg,Tf, Alb and Tg. Of these clones 16 secreted “monospecific” IgM, 32 secreted polyreactive antibodies and 87 had no detectable antibody activity.
2.6 Statistics
The concentration of IgM in culture SN was determined by ELISA. Microtiter plates were coated with goat antihuman Fcp antibodies (Institut Pasteur Production, Marnes-la-Coquette, France) and incubated with dilutions of the SN to be tested. Bound antibodies were revealed using peroxidase-labeled goat anti-Fcp antibodies (Sigma). Absorbances at 495 nm were compared to those obtained with IgM in a known reference serum.
2.3 IVIg Pooled normal polyspecific human IgG for therapeutic use (IVIg; Sandoglobulin@),were a kind gift from the Blood Transfusion Department of the Swiss Red Cross, Bern, Switzerland. F(ab’)2 fragments were prepared from IVIg by pepsin digestion and chromatography on protein A-Sepharose (Phannacia Fine Chemicals, Uppsala, Sweden). F(ab‘)z fragments were free of detectable Fc fragments as assessed by ELISA. 2.4 Assessment of natural antibody activity of culture SN and IVIg, by ELISA Flat-blottom microtiter plates (CML, Nemours, France) were coated with 100 p1 of antigens of the panel diluted in 0.1 carbonate hi carbonate buffer, pH 9.6, for 1 h a t 37°C and overnight at 4°C. Optimal concentrations of the various antigens have previously been established [18]. Coated wells were saturated with 0.1% gelatin in PBS, pH 7.4 (PBS-gel).The plates were washed. SN to be tested for antibody activity diluted 1: 5 in PBS-gel containing 0.05% Tween20 (PGT) were then added and the plates were incubated for 1 h at 37°C. After a further washing, bound Ig were revealed with P-galactosidase-labeled goat anti-Fcp (Sigma) or anti-Fa (Byosis, Compikgne, France) antibodies.
We calculated correlation coefficients between the relative ability of natural IgM to bind F(ab’)2 fragments of IVIg and polyreactivity of the natural antibodies.The binding of IgM to each antigen of the panel was determined and expressed as % of the binding to each antigen of a 1:250 dilution of a normal reference serum. The degree of polyreactivity was then calculated by dividing the sum of relative binding values of IgM to each antigen by the number of antigens tested. 2.7 Inhibition of IgM binding to IVIg and to TNP
Titration of anti-TNP activity was performed on plates coated with TNP-BSA 5 pg/ml in carbonate-bicarbonate buffer, pH 9.6. Dilutions of test SN were incubated for 2 h at 37 “C. Binding of anti-TNP antibodies was revealed using peroxidase-labeled anti-p antibodies. The dilution of IgMcontaining SN that resulted in a 50% value of &95 in the linear portion of the ELISA titration curve was determined and subsequently used for comparative binding assays. Diluted SN were incubated with increasing concentrations of F(ab’)2 from IVIg for 1 h at 37 “C and overnight at 4 “C before addition to TNP-coated wells. Residual binding of IgM to TNP was then revealed with labeled anti-p antibodies. Results were expressed as: &95% value of test sample x loo % Inhibition = 1value of SN incubated with PBS alone
3 Results 3.1 Idanti-Id interactions between natural IgM and IVIg
2.5 Binding of IgM from culture SN to insolubilied IVIg
Of 59 IgM, 39 (64%)culture SN from B lymphoid cell lines and 24 IgM of the 135 (18%) SN from cloned EBVtransformed cord B cells bound to insolubilized F(ab’)z from IVIg, as assessed by ELISA. A strong binding was observed with low inputs of some of the IgM, whereas no binding was observed with other IgM that were present at high concentrations in culture SN. For those IgM that bound to F(ab’)2 fragments of IVIg, the binding reaction was dose-dependent as shown in Fig. 1for IgM form a B cell line (TUM) and for a monoclonal IgM (8-SC-111).The lack of binding to IVIg of a number of IgM that were tested indicated that the binding reaction did not involve the C region of the IgM molecule. The results strongly suggest the Idanti-Id nature of the interactions between natural IgM antibodies and IVIg.
Binding of IgM from culture SN to F(ab’)~fragments from IVIg was quantitated using an ELISA, as described above.
We next examined whether the binding of natural IgM antibodies to an antigen could be inhibited by F(ab’)2
Eur. J. Immunol. 1990.20: 2089-2094
Idiotypic interactions between human IgG and natural IgM antibodies
fragments from IVIg. Fig. 2 indicates that the binding to TNP of IgM from the B cell line TUM was inhibited in a dose-dependent fashion by F(ab’)2 from IVIg that had been adsorbed with TNP and did not contain anti-TNP activity. One hundred percent inhibition of binding t o TNP were achieved at a 103-foldmolar input of IVIg F(ab’)2 over that of IgM. The binding of IgM from B cell line TUM to TNF’ was also inhibited by TNP in a dose-dependent manner. In these experiments, TNP was present as a TNP-complex. BSA alone did not inhibit the interaction between IgM antibodies and IVIg (data not shown). The binding reaction to TNP-complex of monoclonal IgM 8-SC-111 was inhibited by TNP but was not inhibited by F(ab’)z fragments from IVIg (Fig. 2). The results indicate the presence in IVIg of anti-Id that specifically interfere with the antibody combining site of a polyreactive IgM antibody with natural anti-TNP specificity.
-E
6oo 400
I
p. 200
0
I
0
Table 1 shows the ability of the 39 IgM antibodies from B lymphoid cell lines to bind to F(ab’)z from IVIg and their relative degree of polyreactivity as assessed by the ability of the antibodies to recognize a panel of 12 antigens and haptens. Comparison between anti-IVIg F(ab’)2 activity and the calculated degree of polyreactivity of IgM antibodies yielded a correlation coefficient of 0.654 for the 39 IgM listed in Table 1. Among the 59 SN tested, 20 did not bind to F(ab‘h fragments from IVIg. Three of these IgM did not bind to F(ab’)2 from IVIg but recognized at least one of the antigens of the panel. Among the 39 IgM that bound to IVIg F(ab’)~,6 recognized only one antigen, 23 recognized more than one antigen and 10 did not recognize any of the 12 antigens tested.
300
200
100
3.2 Relationship between the binding of IgM to IVIg and polyreactivity of antibodies
Table 2 shows the binding of 25 mAb to IVIg F(ab’)2 and their mean antigen-binding activity assessed by the ability of the autoantibodies to bind to a panel of nine antigens and haptens. No correlation was found between anti-IVIg activity and polyreactivity. Of the 25 monoclonal IgM that bound t o IVIg F(ab‘)z 5 recognized only one antigen of the panel, 8 recognized more than one antigen and 12 did not bind to any antigens tested. Thirty-four of the monoclonal IgM exhibited no binding activity to IVIg F(ab‘)z and bound to one or more antigens of the panel. We also compared the mean binding activity to IVIg of monospecific IgM either from cell lines or monoclonal, with the mean binding activity to IVIg of IgM that recognized more than 50% of the antigens of the panel. No significant difference was found between these two groups of IgM.
5
B
2091
IgM Wml)
Figure 1. Binding of IgM from theTUM (0-0) B lymphoid cell line and of monoclonal IgM from clone 8-SC-111 (0-0) to insolubilized F(ab‘)2 fragments from IVIg.
100,
3.3 Reactivity of IVIg with target antigens of natural antibodies
The reactivity of IVIg with a panel of five haptens and nine self-related antigens that had previously been found to be often recognized by natural polyspecific human IgM [ 11was investigated by ELISA. Binding assays of IVIg t o haptens were performed using haptens coupled to OVA and IVIg that had previously been adsorbed with OVA by three successive chromatography steps on insolubilized antigen. As shown in Fig. 3, IVIg bound to all antigens of the panel in a dose-dependent fashion. The pattern of the binding curves was concave to the abcissa indicating saturable binding.
4 Discussion 1
100
TNP (nM)
10000
1
10
100
1000 10000
SAOL (nM)
Figure 2. Inhibition of binding to TNP of natural IgM antibodies by TNP and by F(ab’)2fragments from 1VIg.The left panel shows the dose-dependent inhibition by TNP-BSA of the binding of IgM from the TUM B lymphoid cell line (0-0) and of monoclonal IgM 8-SC-111 (0-0) to TNF! The binding was not inhibited by soluble BSA (data not shown).The right panel shows inhibition of the binding of IgM TUM and IgM 8-SC-111 to TNP by F(ab’)z fragments from IVIg. IVIg had been adsorbed withTNP-BSA and did not show anti-TNP activity. The abscissa of both figures indicates molar amounts of inhibitor.Theinput of IgM was 1.6 p~ and 2 p~ for TUM and IgM 8-SC-111, respectively.
The present study indicates that IVIg may react through Idanti-Id interactions with natural human IgM secreted by EBV-transformed B cells. The interaction between natural IgM antibodies and IVIg was investigated by assessingthe binding of culture SN from B lymphoid cell lines and from cloned EBV-transformed B cells to F(ab’)z fragments from IVIg, using an ELISA. Sixty-four percent of IgM secreted by B lymphoid cell lines and 18% of monoclonal IgM bound to IVIg. The binding reaction involved F(ab’)2 from IVIg and the V region of IgM.
2092
F. Rossi, B. Guilbert, C. Tonnelle et al.
Eur. J. Immunol. 1990.20: 2089-2094
Table 1. Summary of binding to IVIg and polyreactivity of IgM antibodies
B cell line
37 3 44
50 49
40 38 18 27 52 23 54 58 32 60 26 24 TUM 17 43 22 21 18 25 15 46 28 39 20 19 7 4 16 61 2 35 33 1 12
Anti-IVIgF(ab’)2 activitya, @4YS Pi?) 146.75 121.53 48.62 4.12 37.35 30.07 25.64 23.18 18.69 17.77 16.52 15.46 1s.12 14.48 13.65 13.36 12.83 12.00 11.16 10.44 8.86 8.53 8.34 7.98 7.85 7.58 7.39 7.35 6.85 6.50 5.19 4.92 3.74 3.64 2.46 2.44 1.62 0.69 0.19
Reactivity with antigensb) Act Tub Mg Spe DNATNP Alb PO Myo Tf
+ +
+ +
+
-
+ + -
+ + +
-
Tg Ker
+ + - + + + + + + - + + + + +
-
-
-
-
-
+
Mean antigen binding activity) (A495 P d 132.91 148.58 15.40 0.00 7.92 1.77 0.00 21.57 16.01 3.49 10.84 13.71 22.91 0.00 45.18 34.78 4.10 12.19 1.22 111.74 11.12 9.29 101.72 63.13 30.49 35.04 0.00 0.00 1.67 12.36 1.05
0.00 4.12 11.45 0.00 0.00 0.00 1.08 0.00
a) A4y5/pgof IgM obtained in a binding assay to F(ab‘)2 IVIg using a 1 :5 dilution of culture SN b) + depict reactivity with an antigen. - indicate lack of reactivity.The threshold of reactivity in the ELISAwas defined as the h Y 5 value obtained with a 1 :225 dilution of a normal human serum. c) Sum of &95 obtained in binding assays of a 1 : 5 dilution of culture SN to the 12 antigens divided by 12 and by IgM concentration.
The binding to TNP of polyreactive IgM with anti-TNP specificity was inhibited by the antigen in a dose-dependent manner. The binding to TNP of one of the IgM antibodies was inhibited by F(ab’)2 fragments from IVIg, indicating the presence in IVIg of anti-Id that may interfere with the antibody-combining site of polyreactive IgM antibodies. IgM antibodies secreted by EBV-transformed normal B cells are multireactive as shown by their ability to bind to a broad panel of self-related antigens, antigens and haptens, both in mice [20] and man [21].We examined whether there was a relationship between polyreactivity of natural IgM antibodies and their ability to bind to F(ab’)2 of IVIg. A relatively weak correlation was found in the case of IgM from B lymphoid cell lines; no correlation was found in the
case of IgM mAb.When taking all the IgM which bound to IVIg into consideration, no difference in the mean binding activity of IgM to F(ab’)z from IVIg was found between monospecific IgM and those IgM that were highly polyreactive. IgM mAb bound with higher affinity to IVIg than IgM antibodies secreted by B lymphoid cell lines. However, a higher proportion of IgM from B lymphoid cell lines recognized IVIg. The latter observations may relate to the fact that the cell lines were derived from B cells of elderly donors. Our observations indicate that the ability of IgM to recognize an idiotype on IVIg or to be recognized by anti-Id in IVIg does not increase with increased polyreactivity of natural antibodies. This reinforces the concept of specific interactions between some of these IgM antibodies and IVIg.
Idiotypic interactions between human IgG and natural IgM antibodies
Eur. J. Immunol. 1990. 20: 2089-2094
2093
Table 2. Summary of binding to IVIg and polyreactivity of IgM antibodies
B cell line
Anti-IVIgF( ab’)z activity) (‘4495 CLg)
Reactivity with antigensb) Act Tub Myo DNA TNP Mg
Alb
Tg 14.81 9.64 486.51 4.31 0.00 5.Y7 8.12 0.00 0.00 0.00 0.0 15.82 0.00 184.70 0.00 22.00 57.57 0.00 0.00 0.00 0.00 3.05 100.30 0.00 236.58
672.93 444.49 402.36 316.37 314.50 302.73 298.87 285.33 270.54 251.62 201.53 170.58 152.73 117.20 101.08 84.03 79.08 71.57 51.79 26.46 26.18 25.57 13.07 10.52 4.97
8-sc-111 7-SC-137 8-sc-57 7-SC-136 7-SC-127 7-SC-106 7-SC-105 7-SC-107 7-SC-108 7-sc-66 8-SC-107 2-SP-90 7-SP-72 7-SC-134 2-SP-44 2-SP-41 7-SC-120 8-sc-106 7-SP-46 8-sc-138 2-SP-86 2-SP-54 2-SP-80 8-SC-1.10 8-sc-74
Tf
Mean antigen binding activity)
a) A4&pg of IgM obtained in a binding assay to F(ab’)z using a 1 :5 dilution of culture SN b) + depict reactivity with an antigen. - indicate lack of reactivity.The threshold of reactivity was defined as the A value obtained with a 1 : 225 dilution of a normal human serum. c) Calculated as described in Sect. 2.7
IVIg are pools of IgG prepared from the plasma of a large number of donors, usually between 6000 and 15000. IVIg display a large number of antibody specificities against external antigens [22] and highly conserved mammal autoantigens [9]. The present study demonstrates that IVIg also react with various haptens. Previous studies have indicated that IVIg contain anti-Id against autoantibodies in the serum of patients with various autoimmune diseases [9, 23, 241 and against autoantibodies present in IVIg [9]. As the number of donors to the pool of IVIg increases, the ability of IVIg to express anti-Id activity against antibodies with the preparation is increased [25,26]. The present finding of Id/anti-Id interactions between IVIg and natural
IgM antibodies further documents that IVIg contain anti.body specificities against Ig from normal individuals. These interactions are indicative of an Id recognition between IgG and IgM in the serum of normal individuals who contribute to the pool of IVIg. Our observations also suggest that antibodies of the IgG isotype belonging to the actual B cell repertoire may participate in network interactions with natural polyreactive IgM antibodies. Such interactions may possibly influence the expression of the potential B cell repertoire which is analyzed through the study of EBVtransformed B cells. Intravenous administration of Ig into mice modifies the expressed B cell repertoire as assessed by a change in the preferential use of specificVH families [27].
200 500
-
150
0 %
100
E d C
T
3 50
O f 0
I
200
800 lVlQ
400
(PQW
800
1000
0
100
200
300 lvlg (pglm 1)
400
500
Figure 3. Binding of IVIg to a panel of antigens and haptens. Left panel: binding t o Alb (A),Act (OhDNA (O), Mg (I)My0 , (Oh microsomes (M),Tg (*),Tf (x) and Tub (A). Right panel: binding to TNP-OVA ( O ) , PCOVA (A),ARS-OVA (I) and Phox-OVA (*). IVIg had been preadsorbed with OYA .
2094
F. Rossi, B. Guilbert, C. Tonnelle et al.
Id interactions between IVIg and natural IgM antibodies could be relevant to the suppressive effect of IVIg in several autoimmune diseases in which IVIg may have modified the expressed repertoire by perturbing the Id network. Received April 26, 1990.
5 References 1 Guilbert, B., Dighiero, G. and Avrameas, S., J. Immunol. 1982. 128: 2779. 2 Dighiero, G., Lymberi, F?, Mazie, J. C., Rouyre, S., ButlerBrowne, G. S.,Whalen, R. G. and Avrameas, s.,J. lmmunol. 1983. 131: 2267. 3 Guilbert, B., Mahana, W., Gilbert, M., Mazie, J. C. and Avrameas, S., Immunology 1985. 56: 401. 4 Dighiero, G., Lymberi, F?, Holmberg, D., Lundquist, I., Coutinho, A. and Avrameas, S., J. Immunol. 1985. 134: 765. 5 Holmberg, D., Forsgren, S., Ivars, F. and Coutinho, A., Eur. J. Immunol. 1984. 14: 435. 6 Sultan,Y., Kazatchkine, M. D., Maisonneuve, F? and Nydegger, U. E.. Lancet 1984. ii: 765. 7 Vermeulen, M.,Van der Mesche, F. G. A., Speelman, J. D., Weber, A. and Busch, H. F. M., J. Neurol. Sci. 1985. 70: 317. 8 Carreas, L. O., Pkrez, G. N.,Vega, H. R. and Cazavilla, F., Lancet 1988. ii: 393. 9 Rossi, F. and Kazatchkine, M. D., J. Immunol. 1989. 143: 4104. 10 Rossi, F., Dietrich, G. and Kazatchkine, M. D., Immunol. Rev. 1989. 110: 135. 11 Spudich, J. A. and Watt, S., J. Biol. Chem. 1971. 246: 4866.
Eur. J. Immunol. 1990.20: 2089-2094
12 Whalen, R. G., Butler-Browne, G. S. and Gros, F., J. Mol. Biol. 1978. 126: 415. 13 Shelansky, M. L.,Gaskin, F. and Cantor, C., Proc. Natl. Acad. Sci. USA 1973. 70: 765. 14 Little, J. R., Eisen, H. N., Biochemistry 1966. 5: 3385. 15 Nisonoff, A., Methods Immunol. Immunochem. 1967. 1: 120. 16 Pery, F?, Luffau, G., Charley, J., Petit, A., Rouze, P. and Bernard, S., Annu. Immunol. (Inst. Pasteur) 1979. 130C: 517. 17 Makela, O., Kaartinen, M., Pelkonen, J. L.T. andKarjalainen, K., J. Exp. Med. 1978.148: 1644. 18 Seigneurin, M. J., Guilbert, B., Bourgeat, M. J. and Avrameas, S., Blood 1988. 3: 581. 19 Guigou,V., Guilbert, B., Moinier, D.,Tonnelle, C., Bloubly, C., Avrameas, S., Fougereau, M. and Fumoux, F., J. lmmunol. 1990, in press. 20 Lymberi, F?, Dighiero, G. ,Ternynck,T. and Avrameas, S., Eur. J. Immunol. 1985. 15: 702. 21 Garzelli, C., Taub, F. E., Scharff, J. E., Probhakar, B. S., Ginsberg-Fellner, F., Notkins, A. L., J. Virol. 1984. 52: 722. 22 Morel], A. and Nydegger, U. E. (Eds.), Clinical Use of Intravenous Immunoglobulins, Academic Press, London 1986. 23 Rossi, F., Sultan, Y. and Kazatchkine, M. D., Clin. Exp. Immunol. 1988. 74: 311. 24 Van Doom, F! A., Rossi, F., Brand, A., Van Lint, M., Vermeulen, M. and Kazatchkine, M. D., J. Neuroimmunol. 1990, in press. 25 Tankersley, D. L., Preston, M. S. and Finlay, J. S., Mol. Immunol. 1988. 25: 41. 26 Gronsky, F?, Barrer, R., Bodenbender, L., Kanzy, E. J., Schmidt, K. H., Zilag, H. and Seiler, E R., Behring lnst. Mitt. 1988. 82: 127. 27 Sundblat, A., Hauser, S., Holmberg, D., Cazenave, F? A. and Coutinho, A., Eur. J. lmmunol. 1989. 19: 1425.
Frangoise RossiO, Brigitte Guilberta, CBcile Tonnelle+, ThBrkse TernynckA, Frangois Fumoux+, Stratis AvrameasA and Michael D. KazatchkineO Unit6 d’Immunopathologie and INSERM U 28O, H6pital Broussais, Pans, Unit6 d’hmunocytochimie URA CNRS 35P, Dhpartement d’Immunologie, Institut Pasteur, Pans and Centre d’Immunologie INSERM-CNRS de Marseille-Luminy , Marseille +
Idiotypic interactions between human IgG and natural IgM antibodies
Idiotypic interactions between normal human polyspecific IgG and natural IgM antibodies* Pooled normal human polyspecific IgG (IVIg) contain anti-idiotypes against a variety of autoantibodies from patients with autoimmune diseases and IgG autoantibodies present in IVIg. The present study indicates that IVIg may also react through idiotypidanti-idiotypic interactions with human natural IgM antibodies. Sixty-four percent of IgM secreted by B lymphoid cell lines derived from B cells of healthy elderly donors and 18% of IgM secreted by cloned EBV-transformed cord B cells that were tested, bound through their variable region to F(ab’)z fragments of 1VIg.The binding to 2,4,6-trinitrophenyl (TNP) of a polyreactive IgM with anti-TNP specificity, was inhibited by F(ab‘)z fragments from IVIg, indicating the presence in IVIg of anti-idiotypes that may interfere with the antibody-combining site of polyreactive IgM antibodies. The ability of IgM antibodies to interact with idiotypes on IVIg was not related to the degree of polyreactivity of natural antibodies. Our observations further document that IVIg contain antibody specificities against Ig from normal individuals and suggest that IgG originating from the physiologically expressed repertoire may modulate the expression of the potential B cell repertoire.The results may be relevant to the suppressive effect of IVIg in autoimmune diseases.
1 Introduction Sera from healthy individuals contain polyreactive IgM antibodies and IgG antibodies that recognize a variety of self-related antigens [11. B cells synthesizing polyreactive antibodies against self and non-self antigens are found in spleens from nonimmunized [2] and from immunized [3] animals. In mice, a subset of natural polyreactive IgM antibodies exhibits a high degree of autoreactivity [4] and V region connectivity [5]. Intravenous immunoglobulins (IVIg) are therapeutic preparations of IgG prepared from the plasma of a large number of healthy donors. Infusion of IVIg has resulted in the decrease of autoantibody titer in several autoimmune diseases including anti-factor VIII autoimmune disease [6], chronic demyelinating inflammatory polyneuropathy [7], vasculitis with circulating anti-neutrophil cytoplasmic antigen antibodies (D. R. W. Jayne, unpublished observations), and in patients with recurrent abortion and anticardiolipin antibodies [8]. IVIg contain anti-Id against a variety of autoantibodies associated with autoimmune disease [9]. We have, thus, postulated that the beneficial
[I 85131
*
2089
This work was supported by Institut National et de la Sant6 de la Recherche Mtdicale (INSERM), Centre National de la Recherche Scientifique (CNRS) and Sandoz SA (Basel, Switzerland).
Correspondence: Frangoise Rossi, Unit6 d’Immunopathologie et INSERM U28, Hapita1 Broussais, 96, rue Didot, F-75674 Pans C6dex 14, France Abbreviations: Act: Actin Alb: Human albumin Am-OVA: OVA-coupled arsonate Id: Idiotypehdiotypic IVIg: Intravenous Ig Mg: Myoglobulin Myo: Myosin PC-OVA: OVAcoupled phosphorylcholine Phox-OVA: OVA-coupled phenyloxazalone PO: Peroxidase Spe: Spectrin Tf: Transferrin Tg: Thyroglobulin ’hkTbbulin 0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
effect of IVIg in autoimmune diseases depends on the ability of IVIg to interfere with the regulatory function of the Id network on autoimmune responses [ 101. The present study indicates that IVIg may interact through Idanti-Id recognition with human natural polyreactive IgM antibodies. Id interactions between IgG and natural IgM antibodies may play a role in modulating the expression of the B cell repertoire in normal and diseased individuals.
2 Materials and methods 2.1 Antigens BSA,whole skeletal muscle myoglobin (Mg), KLH, human thyroglobulin (Tg), ribonuclease, human microsomes, calf thymus native double-stranded DNA and trinitrobenzene sulfonic acid (Sigma Chemical Co., St. Louis, MO), human transferrin (Tf; Behringwerke AG., Marburg, FRG), human albumin (Alb; Schwartz Mann, Cambridge, MA) and horseradish peroxidase (grade I, R Z = 3) (PO; Boehringer-Mannheim, Mannheim, FRG) were purchased as indicated. Calf muscle actin (Act) [ll]and myosin (Myo) [12], pig brain tubulin (Tub) [13], TNP-25/BSA [14], OVA (Sigma), TNP-OVA [141, arsonate-OVA (Ars-OVA) [15], phosphorylcholine-OVA (PC-OVA) [161 and phenyloxazoloneOVA (Phox-OVA) [17] were obtained as described. Human RBC spectrin (Spe) was a kind gift from Dr. E. Lazarides (Caltech, Pasadena, CA). Hemoglobin was prepared from human RBC using a 0.83% ammonium chloride lysis buffer. 2.2 IgM antibodies from EBV-transformed human B cells We studied IgM antibodies secreted by human EBVtransformed lymphoid B cell lines and EBV-transformed 0014-2980/90/0909-2089$3.50 + .25/0
2090
F. Rossi, B. Guilbert, C. Tonnelle et al.
Eur. J. Immunol. 1990. 20: 2089-2094
cloned B cells. SN were obtained from cultures of 59 cloned EBV-transformed human B cell lines raised from PBL from normal elderly individuals [MI. All cell lines secreted exclusively or predominantly Ig of the IgM isotype. The specificity of secreted IgM was tested on a panel of 12 antigens: Act,Tub, Mg, Spe, DNA,TNP, Alb, PO, Myo, Tf,Tg and KLH. Six cell lines were found to secrete monospecific antibody that recognized only one antigen of the panel; 26 cell lines secreted polyreactive antibodies; 27 cell lines had no detectable antibody specificity.
Plates were coated with F(ab’)2 from IVIg at a concentration of 100 pg/ml. A positive control was included in each ELISA plate, which consisted of a purified human monoclonal IgM that was found to bind to F(ab‘)2 from 1VIg.The revealing reagent was anti-Fcp antibody that had been extensively adsorbed with human F(ab’)s fragments and showed no reactivity with F(ab’)z. Absorbance values of the binding of IgM to insolubilized IVIg were corrected for IgM concentrations.
B cell clones (n = 135) were obtained by immortalizing human cord B cells and peripheral blood B cells from healthy adult individuals with EBV [19]. The specificity of the monoclonal IgM secreted by cloned cells was tested on a panel of nine antigens: Act,Tub, Myo, DNA,TNF’, Mg,Tf, Alb and Tg. Of these clones 16 secreted “monospecific” IgM, 32 secreted polyreactive antibodies and 87 had no detectable antibody activity.
2.6 Statistics
The concentration of IgM in culture SN was determined by ELISA. Microtiter plates were coated with goat antihuman Fcp antibodies (Institut Pasteur Production, Marnes-la-Coquette, France) and incubated with dilutions of the SN to be tested. Bound antibodies were revealed using peroxidase-labeled goat anti-Fcp antibodies (Sigma). Absorbances at 495 nm were compared to those obtained with IgM in a known reference serum.
2.3 IVIg Pooled normal polyspecific human IgG for therapeutic use (IVIg; Sandoglobulin@),were a kind gift from the Blood Transfusion Department of the Swiss Red Cross, Bern, Switzerland. F(ab’)2 fragments were prepared from IVIg by pepsin digestion and chromatography on protein A-Sepharose (Phannacia Fine Chemicals, Uppsala, Sweden). F(ab‘)z fragments were free of detectable Fc fragments as assessed by ELISA. 2.4 Assessment of natural antibody activity of culture SN and IVIg, by ELISA Flat-blottom microtiter plates (CML, Nemours, France) were coated with 100 p1 of antigens of the panel diluted in 0.1 carbonate hi carbonate buffer, pH 9.6, for 1 h a t 37°C and overnight at 4°C. Optimal concentrations of the various antigens have previously been established [18]. Coated wells were saturated with 0.1% gelatin in PBS, pH 7.4 (PBS-gel).The plates were washed. SN to be tested for antibody activity diluted 1: 5 in PBS-gel containing 0.05% Tween20 (PGT) were then added and the plates were incubated for 1 h at 37°C. After a further washing, bound Ig were revealed with P-galactosidase-labeled goat anti-Fcp (Sigma) or anti-Fa (Byosis, Compikgne, France) antibodies.
We calculated correlation coefficients between the relative ability of natural IgM to bind F(ab’)2 fragments of IVIg and polyreactivity of the natural antibodies.The binding of IgM to each antigen of the panel was determined and expressed as % of the binding to each antigen of a 1:250 dilution of a normal reference serum. The degree of polyreactivity was then calculated by dividing the sum of relative binding values of IgM to each antigen by the number of antigens tested. 2.7 Inhibition of IgM binding to IVIg and to TNP
Titration of anti-TNP activity was performed on plates coated with TNP-BSA 5 pg/ml in carbonate-bicarbonate buffer, pH 9.6. Dilutions of test SN were incubated for 2 h at 37 “C. Binding of anti-TNP antibodies was revealed using peroxidase-labeled anti-p antibodies. The dilution of IgMcontaining SN that resulted in a 50% value of &95 in the linear portion of the ELISA titration curve was determined and subsequently used for comparative binding assays. Diluted SN were incubated with increasing concentrations of F(ab’)2 from IVIg for 1 h at 37 “C and overnight at 4 “C before addition to TNP-coated wells. Residual binding of IgM to TNP was then revealed with labeled anti-p antibodies. Results were expressed as: &95% value of test sample x loo % Inhibition = 1value of SN incubated with PBS alone
3 Results 3.1 Idanti-Id interactions between natural IgM and IVIg
2.5 Binding of IgM from culture SN to insolubilied IVIg
Of 59 IgM, 39 (64%)culture SN from B lymphoid cell lines and 24 IgM of the 135 (18%) SN from cloned EBVtransformed cord B cells bound to insolubilized F(ab’)z from IVIg, as assessed by ELISA. A strong binding was observed with low inputs of some of the IgM, whereas no binding was observed with other IgM that were present at high concentrations in culture SN. For those IgM that bound to F(ab’)2 fragments of IVIg, the binding reaction was dose-dependent as shown in Fig. 1for IgM form a B cell line (TUM) and for a monoclonal IgM (8-SC-111).The lack of binding to IVIg of a number of IgM that were tested indicated that the binding reaction did not involve the C region of the IgM molecule. The results strongly suggest the Idanti-Id nature of the interactions between natural IgM antibodies and IVIg.
Binding of IgM from culture SN to F(ab’)~fragments from IVIg was quantitated using an ELISA, as described above.
We next examined whether the binding of natural IgM antibodies to an antigen could be inhibited by F(ab’)2
Eur. J. Immunol. 1990.20: 2089-2094
Idiotypic interactions between human IgG and natural IgM antibodies
fragments from IVIg. Fig. 2 indicates that the binding to TNP of IgM from the B cell line TUM was inhibited in a dose-dependent fashion by F(ab’)2 from IVIg that had been adsorbed with TNP and did not contain anti-TNP activity. One hundred percent inhibition of binding t o TNP were achieved at a 103-foldmolar input of IVIg F(ab’)2 over that of IgM. The binding of IgM from B cell line TUM to TNF’ was also inhibited by TNP in a dose-dependent manner. In these experiments, TNP was present as a TNP-complex. BSA alone did not inhibit the interaction between IgM antibodies and IVIg (data not shown). The binding reaction to TNP-complex of monoclonal IgM 8-SC-111 was inhibited by TNP but was not inhibited by F(ab’)z fragments from IVIg (Fig. 2). The results indicate the presence in IVIg of anti-Id that specifically interfere with the antibody combining site of a polyreactive IgM antibody with natural anti-TNP specificity.
-E
6oo 400
I
p. 200
0
I
0
Table 1 shows the ability of the 39 IgM antibodies from B lymphoid cell lines to bind to F(ab’)z from IVIg and their relative degree of polyreactivity as assessed by the ability of the antibodies to recognize a panel of 12 antigens and haptens. Comparison between anti-IVIg F(ab’)2 activity and the calculated degree of polyreactivity of IgM antibodies yielded a correlation coefficient of 0.654 for the 39 IgM listed in Table 1. Among the 59 SN tested, 20 did not bind to F(ab‘h fragments from IVIg. Three of these IgM did not bind to F(ab’)2 from IVIg but recognized at least one of the antigens of the panel. Among the 39 IgM that bound to IVIg F(ab’)~,6 recognized only one antigen, 23 recognized more than one antigen and 10 did not recognize any of the 12 antigens tested.
300
200
100
3.2 Relationship between the binding of IgM to IVIg and polyreactivity of antibodies
Table 2 shows the binding of 25 mAb to IVIg F(ab’)2 and their mean antigen-binding activity assessed by the ability of the autoantibodies to bind to a panel of nine antigens and haptens. No correlation was found between anti-IVIg activity and polyreactivity. Of the 25 monoclonal IgM that bound t o IVIg F(ab‘)z 5 recognized only one antigen of the panel, 8 recognized more than one antigen and 12 did not bind to any antigens tested. Thirty-four of the monoclonal IgM exhibited no binding activity to IVIg F(ab‘)z and bound to one or more antigens of the panel. We also compared the mean binding activity to IVIg of monospecific IgM either from cell lines or monoclonal, with the mean binding activity to IVIg of IgM that recognized more than 50% of the antigens of the panel. No significant difference was found between these two groups of IgM.
5
B
2091
IgM Wml)
Figure 1. Binding of IgM from theTUM (0-0) B lymphoid cell line and of monoclonal IgM from clone 8-SC-111 (0-0) to insolubilized F(ab‘)2 fragments from IVIg.
100,
3.3 Reactivity of IVIg with target antigens of natural antibodies
The reactivity of IVIg with a panel of five haptens and nine self-related antigens that had previously been found to be often recognized by natural polyspecific human IgM [ 11was investigated by ELISA. Binding assays of IVIg t o haptens were performed using haptens coupled to OVA and IVIg that had previously been adsorbed with OVA by three successive chromatography steps on insolubilized antigen. As shown in Fig. 3, IVIg bound to all antigens of the panel in a dose-dependent fashion. The pattern of the binding curves was concave to the abcissa indicating saturable binding.
4 Discussion 1
100
TNP (nM)
10000
1
10
100
1000 10000
SAOL (nM)
Figure 2. Inhibition of binding to TNP of natural IgM antibodies by TNP and by F(ab’)2fragments from 1VIg.The left panel shows the dose-dependent inhibition by TNP-BSA of the binding of IgM from the TUM B lymphoid cell line (0-0) and of monoclonal IgM 8-SC-111 (0-0) to TNF! The binding was not inhibited by soluble BSA (data not shown).The right panel shows inhibition of the binding of IgM TUM and IgM 8-SC-111 to TNP by F(ab’)z fragments from IVIg. IVIg had been adsorbed withTNP-BSA and did not show anti-TNP activity. The abscissa of both figures indicates molar amounts of inhibitor.Theinput of IgM was 1.6 p~ and 2 p~ for TUM and IgM 8-SC-111, respectively.
The present study indicates that IVIg may react through Idanti-Id interactions with natural human IgM secreted by EBV-transformed B cells. The interaction between natural IgM antibodies and IVIg was investigated by assessingthe binding of culture SN from B lymphoid cell lines and from cloned EBV-transformed B cells to F(ab’)z fragments from IVIg, using an ELISA. Sixty-four percent of IgM secreted by B lymphoid cell lines and 18% of monoclonal IgM bound to IVIg. The binding reaction involved F(ab’)2 from IVIg and the V region of IgM.
2092
F. Rossi, B. Guilbert, C. Tonnelle et al.
Eur. J. Immunol. 1990.20: 2089-2094
Table 1. Summary of binding to IVIg and polyreactivity of IgM antibodies
B cell line
37 3 44
50 49
40 38 18 27 52 23 54 58 32 60 26 24 TUM 17 43 22 21 18 25 15 46 28 39 20 19 7 4 16 61 2 35 33 1 12
Anti-IVIgF(ab’)2 activitya, @4YS Pi?) 146.75 121.53 48.62 4.12 37.35 30.07 25.64 23.18 18.69 17.77 16.52 15.46 1s.12 14.48 13.65 13.36 12.83 12.00 11.16 10.44 8.86 8.53 8.34 7.98 7.85 7.58 7.39 7.35 6.85 6.50 5.19 4.92 3.74 3.64 2.46 2.44 1.62 0.69 0.19
Reactivity with antigensb) Act Tub Mg Spe DNATNP Alb PO Myo Tf
+ +
+ +
+
-
+ + -
+ + +
-
Tg Ker
+ + - + + + + + + - + + + + +
-
-
-
-
-
+
Mean antigen binding activity) (A495 P d 132.91 148.58 15.40 0.00 7.92 1.77 0.00 21.57 16.01 3.49 10.84 13.71 22.91 0.00 45.18 34.78 4.10 12.19 1.22 111.74 11.12 9.29 101.72 63.13 30.49 35.04 0.00 0.00 1.67 12.36 1.05
0.00 4.12 11.45 0.00 0.00 0.00 1.08 0.00
a) A4y5/pgof IgM obtained in a binding assay to F(ab‘)2 IVIg using a 1 :5 dilution of culture SN b) + depict reactivity with an antigen. - indicate lack of reactivity.The threshold of reactivity in the ELISAwas defined as the h Y 5 value obtained with a 1 :225 dilution of a normal human serum. c) Sum of &95 obtained in binding assays of a 1 : 5 dilution of culture SN to the 12 antigens divided by 12 and by IgM concentration.
The binding to TNP of polyreactive IgM with anti-TNP specificity was inhibited by the antigen in a dose-dependent manner. The binding to TNP of one of the IgM antibodies was inhibited by F(ab’)2 fragments from IVIg, indicating the presence in IVIg of anti-Id that may interfere with the antibody-combining site of polyreactive IgM antibodies. IgM antibodies secreted by EBV-transformed normal B cells are multireactive as shown by their ability to bind to a broad panel of self-related antigens, antigens and haptens, both in mice [20] and man [21].We examined whether there was a relationship between polyreactivity of natural IgM antibodies and their ability to bind to F(ab’)2 of IVIg. A relatively weak correlation was found in the case of IgM from B lymphoid cell lines; no correlation was found in the
case of IgM mAb.When taking all the IgM which bound to IVIg into consideration, no difference in the mean binding activity of IgM to F(ab’)z from IVIg was found between monospecific IgM and those IgM that were highly polyreactive. IgM mAb bound with higher affinity to IVIg than IgM antibodies secreted by B lymphoid cell lines. However, a higher proportion of IgM from B lymphoid cell lines recognized IVIg. The latter observations may relate to the fact that the cell lines were derived from B cells of elderly donors. Our observations indicate that the ability of IgM to recognize an idiotype on IVIg or to be recognized by anti-Id in IVIg does not increase with increased polyreactivity of natural antibodies. This reinforces the concept of specific interactions between some of these IgM antibodies and IVIg.
Idiotypic interactions between human IgG and natural IgM antibodies
Eur. J. Immunol. 1990. 20: 2089-2094
2093
Table 2. Summary of binding to IVIg and polyreactivity of IgM antibodies
B cell line
Anti-IVIgF( ab’)z activity) (‘4495 CLg)
Reactivity with antigensb) Act Tub Myo DNA TNP Mg
Alb
Tg 14.81 9.64 486.51 4.31 0.00 5.Y7 8.12 0.00 0.00 0.00 0.0 15.82 0.00 184.70 0.00 22.00 57.57 0.00 0.00 0.00 0.00 3.05 100.30 0.00 236.58
672.93 444.49 402.36 316.37 314.50 302.73 298.87 285.33 270.54 251.62 201.53 170.58 152.73 117.20 101.08 84.03 79.08 71.57 51.79 26.46 26.18 25.57 13.07 10.52 4.97
8-sc-111 7-SC-137 8-sc-57 7-SC-136 7-SC-127 7-SC-106 7-SC-105 7-SC-107 7-SC-108 7-sc-66 8-SC-107 2-SP-90 7-SP-72 7-SC-134 2-SP-44 2-SP-41 7-SC-120 8-sc-106 7-SP-46 8-sc-138 2-SP-86 2-SP-54 2-SP-80 8-SC-1.10 8-sc-74
Tf
Mean antigen binding activity)
a) A4&pg of IgM obtained in a binding assay to F(ab’)z using a 1 :5 dilution of culture SN b) + depict reactivity with an antigen. - indicate lack of reactivity.The threshold of reactivity was defined as the A value obtained with a 1 : 225 dilution of a normal human serum. c) Calculated as described in Sect. 2.7
IVIg are pools of IgG prepared from the plasma of a large number of donors, usually between 6000 and 15000. IVIg display a large number of antibody specificities against external antigens [22] and highly conserved mammal autoantigens [9]. The present study demonstrates that IVIg also react with various haptens. Previous studies have indicated that IVIg contain anti-Id against autoantibodies in the serum of patients with various autoimmune diseases [9, 23, 241 and against autoantibodies present in IVIg [9]. As the number of donors to the pool of IVIg increases, the ability of IVIg to express anti-Id activity against antibodies with the preparation is increased [25,26]. The present finding of Id/anti-Id interactions between IVIg and natural
IgM antibodies further documents that IVIg contain anti.body specificities against Ig from normal individuals. These interactions are indicative of an Id recognition between IgG and IgM in the serum of normal individuals who contribute to the pool of IVIg. Our observations also suggest that antibodies of the IgG isotype belonging to the actual B cell repertoire may participate in network interactions with natural polyreactive IgM antibodies. Such interactions may possibly influence the expression of the potential B cell repertoire which is analyzed through the study of EBVtransformed B cells. Intravenous administration of Ig into mice modifies the expressed B cell repertoire as assessed by a change in the preferential use of specificVH families [27].
200 500
-
150
0 %
100
E d C
T
3 50
O f 0
I
200
800 lVlQ
400
(PQW
800
1000
0
100
200
300 lvlg (pglm 1)
400
500
Figure 3. Binding of IVIg to a panel of antigens and haptens. Left panel: binding t o Alb (A),Act (OhDNA (O), Mg (I)My0 , (Oh microsomes (M),Tg (*),Tf (x) and Tub (A). Right panel: binding to TNP-OVA ( O ) , PCOVA (A),ARS-OVA (I) and Phox-OVA (*). IVIg had been preadsorbed with OYA .
2094
F. Rossi, B. Guilbert, C. Tonnelle et al.
Id interactions between IVIg and natural IgM antibodies could be relevant to the suppressive effect of IVIg in several autoimmune diseases in which IVIg may have modified the expressed repertoire by perturbing the Id network. Received April 26, 1990.
5 References 1 Guilbert, B., Dighiero, G. and Avrameas, S., J. Immunol. 1982. 128: 2779. 2 Dighiero, G., Lymberi, F?, Mazie, J. C., Rouyre, S., ButlerBrowne, G. S.,Whalen, R. G. and Avrameas, s.,J. lmmunol. 1983. 131: 2267. 3 Guilbert, B., Mahana, W., Gilbert, M., Mazie, J. C. and Avrameas, S., Immunology 1985. 56: 401. 4 Dighiero, G., Lymberi, F?, Holmberg, D., Lundquist, I., Coutinho, A. and Avrameas, S., J. Immunol. 1985. 134: 765. 5 Holmberg, D., Forsgren, S., Ivars, F. and Coutinho, A., Eur. J. Immunol. 1984. 14: 435. 6 Sultan,Y., Kazatchkine, M. D., Maisonneuve, F? and Nydegger, U. E.. Lancet 1984. ii: 765. 7 Vermeulen, M.,Van der Mesche, F. G. A., Speelman, J. D., Weber, A. and Busch, H. F. M., J. Neurol. Sci. 1985. 70: 317. 8 Carreas, L. O., Pkrez, G. N.,Vega, H. R. and Cazavilla, F., Lancet 1988. ii: 393. 9 Rossi, F. and Kazatchkine, M. D., J. Immunol. 1989. 143: 4104. 10 Rossi, F., Dietrich, G. and Kazatchkine, M. D., Immunol. Rev. 1989. 110: 135. 11 Spudich, J. A. and Watt, S., J. Biol. Chem. 1971. 246: 4866.
Eur. J. Immunol. 1990.20: 2089-2094
12 Whalen, R. G., Butler-Browne, G. S. and Gros, F., J. Mol. Biol. 1978. 126: 415. 13 Shelansky, M. L.,Gaskin, F. and Cantor, C., Proc. Natl. Acad. Sci. USA 1973. 70: 765. 14 Little, J. R., Eisen, H. N., Biochemistry 1966. 5: 3385. 15 Nisonoff, A., Methods Immunol. Immunochem. 1967. 1: 120. 16 Pery, F?, Luffau, G., Charley, J., Petit, A., Rouze, P. and Bernard, S., Annu. Immunol. (Inst. Pasteur) 1979. 130C: 517. 17 Makela, O., Kaartinen, M., Pelkonen, J. L.T. andKarjalainen, K., J. Exp. Med. 1978.148: 1644. 18 Seigneurin, M. J., Guilbert, B., Bourgeat, M. J. and Avrameas, S., Blood 1988. 3: 581. 19 Guigou,V., Guilbert, B., Moinier, D.,Tonnelle, C., Bloubly, C., Avrameas, S., Fougereau, M. and Fumoux, F., J. lmmunol. 1990, in press. 20 Lymberi, F?, Dighiero, G. ,Ternynck,T. and Avrameas, S., Eur. J. Immunol. 1985. 15: 702. 21 Garzelli, C., Taub, F. E., Scharff, J. E., Probhakar, B. S., Ginsberg-Fellner, F., Notkins, A. L., J. Virol. 1984. 52: 722. 22 Morel], A. and Nydegger, U. E. (Eds.), Clinical Use of Intravenous Immunoglobulins, Academic Press, London 1986. 23 Rossi, F., Sultan, Y. and Kazatchkine, M. D., Clin. Exp. Immunol. 1988. 74: 311. 24 Van Doom, F! A., Rossi, F., Brand, A., Van Lint, M., Vermeulen, M. and Kazatchkine, M. D., J. Neuroimmunol. 1990, in press. 25 Tankersley, D. L., Preston, M. S. and Finlay, J. S., Mol. Immunol. 1988. 25: 41. 26 Gronsky, F?, Barrer, R., Bodenbender, L., Kanzy, E. J., Schmidt, K. H., Zilag, H. and Seiler, E R., Behring lnst. Mitt. 1988. 82: 127. 27 Sundblat, A., Hauser, S., Holmberg, D., Cazenave, F? A. and Coutinho, A., Eur. J. lmmunol. 1989. 19: 1425.
