Molecular Immunology, Vol. 29, No. 3, pp. 343-351, Printed in Great Britain.

1992

ANTI-CYCLOSPORINE MONOCLONAL ANTIBODIES THEIR ANTI-IDIOTOPIC COUNTERPART: STRUCTURE AND BIOLOGICAL ACTIVITY

0161.5890192

$5.00 + 0.00

;c’ 1992 Pergamon Press plc

AND

GERHARD ZENKE,*~ GABRIELLE ZEDER,$ ULRIKE STRITTMATTER,* ELSEBETH ANDERSEN,* HANS PETER KOCHER,* VALERIE F. J. QUESNIAUX,* MAX H. SCHREIER* and MARC H. V. VAN REGENMORTEL~ *Preclinical Research, Sandoz Pharma Ltd, CH-4002 Base], Switzerland and ZLaboratoire d’Immunochimie, Institut de Biologie Mol&culaire et Cellulaire, F-67084 Strasbourg, France (First received 15 May 1991; accepted 25 June 1991)

Abstract--In order to study the structural and functional mimicry of an antigen by anti-idiotypic antibodies, we generated anti-idiotopic monoclonal antibodies (anti-Id mAbs) against a mAb (R45-45-11) with specificity for the immunosuppressive cyclic undecapeptide cyclosporine (Cs; SandimmuneR). Three out of five anti-Id mAbs inhibited the binding of Cs to the anti-Cs mAb R45-45-11. All anti-Id mAbs cross-reacted only with one (antiCs mAb V45-271-10) out of 19 anti-Cs mAbs. The anti-Cs mAb V45-271-10 recognizes an epitope on the Cs molecule which is very similar to that recognized by R45-45-11. R45-45-11 and V45-271-10 differ only by one amino acid in the variable region. The anti-Id mAbs which recognize combining site-associated idiotopes (Ids) reverse the blocking effect of the anti-Cs mAb R45-45-11 on Cs immunosuppression in vitro. The sequences of the variable regions of heavy and light chain of one anti-Id mAb were determined. X-ray analysis of the corresponding Fab fragment, either alone or complexed with the Fab fragment of the Id, is currently in progress.

INTRODUCTION

sional structure of idiotypes can be obtained by X-ray diffraction. So far, one crystallographic study of a Idiotypic-anti-idiotypic interactions play an important complex between an anti-lysozyme monoclonal antibody role in a variety of immunological processes. Of special (mAb) and an anti-idiotopic (anti-Id) mAb has been interest are anti-idiotypic antibodies which, according to reported (Bentley et al., 1989; Bentley et nl., 1990). the immune network hypothesis, represent an internal We have chosen the immunosuppressive cyclic undeimage of the antigen and can act as the antigen itself, e.g. capeptide cyclosporine (Cs; SandimmuneR) as a model mimicking biological activities of the antigen (Jerne, antigen to study the structural parameters involved in 1974). Our knowledge about the structure of idioantigen-antibody recognition as well as in idiotope type-anti-idiotype interaction and in particular about (Id)-anti-Id interaction for several reasons. Cs has a the molecular mimicry of antigen by anti-idiotypic antirigid, well-characterized three-dimensional structure bodies is still fragmentary. The serological characteriz(Petcher et al., 1976; Loosli et al., 1985). Hundreds of Cs ation of idiotypes has significantly contributed to our derivatives have been synthesized (Wenger, 1986). More understanding of the molecular and structural basis of than 180 anti-Cs mAbs have been prepared to analyze idiotypy (for review see Davie et al., 1986). A direct the immunochemical properties of Cs (Quesniaux et al., correlation of the three-dimensional structure of an 1986). On the basis of their cross-reactivity with various antibody with idiotype has been attempted by immuCs derivatives, the epitopes recognized by different noelectron microscopy and led to the localization of a mAbs were characterized at the level of individual amino combining site-associated idiotype in the distal terminus acid residues on the Cs molecule (Quesniaux et al., of Fab arms (Roux t~f al., 1982; Rous et al., 1987). The 1987a, 1990). Sequences of the variable region of seven relatively low resolution of this method did not allow a anti-Cs mAbs have been determined (Schmitter et al., more precise definition of the idiotype-anti-idiotype 1990). One of these mAbs (R450-45 I 1) has been selected interaction. Direct information about the three-dimenfor determining the three-dimensional structure of the Cs-Fab complex by crystallization and X-ray analysis (Altschuh et al., 1989). tAuthor to whom correspondence should be addressed: Here, we report the immunochemical and functional Gerhard Zenke, Preclinical Research 386/127, Sandoz characterization of five anti-Id mAbs generated against Pharma Ltd, CH-4002 Basel, Switzerland. the anti-Cs mAb R45-45 11. The amino acid sequence of Ahbwriations-BSA, bovine serum albumin; Cs, cyclosporine; the variable regions of heavy and light chain of one of Id(s), idiotopic/idiotope(s); mAb(s), monoclonal antithe anti-Id mAbs (114), which inhibits the interaction body(-ies); PBS, phosphate buffered saline; V,,V,, variable domain of heavy and light chain. between R45-45-11 and Cs, has been determined. 343

344

G.ZENKE

Table

I. Summary

ELISA

format

Sandwich

Isotype

ELISA

I

ELISA

ELISA

II

ELISA

“In all ELISAs

used for detection and characterization

of successive

I-AP”

I-BSA-O/anti-Cs-biotin/streptavidin-AP + anti-Id (competitor) I-R45-45 11/anti-Id-biotin/streptavidin-AP + anti-Id (competitor) I-anti-IgG + IgM/anti-Id/NS/R45-45-I 1-AP + anti-Q (competitor) I-anti-IgG + IgM/anti-Id~anti-isotype-biotin/ strep~~vidin-AP

the first component

is used to coat microtiter

of anti-Id mAbs

Suitabilitv

steDs of the assavY

I”-R45-45-l l/anti-Id/R45-45-1

inhibition

Competition

immunoassays

Summary

ELISA

Competition Mutual

of solid-phase

et al.

for defining

Ids (combining site- and non-combining site-associated) Combining site-associated Ids Spatial

relationship

Cross-reactive Ig isotype

wells and the following

of Ids

Ids

and subclass

reagents

are added

sequentially. ‘1: Solid-phase, ‘AP: alkaline

phosphatase.

Furthermore, X-ray analysis of the Fab fragment of the anti-Id mAb 114 either alone or as a complex with the R45-45-11 Fab is in progress (Mikol and Zurini, in preparation). These studies should provide the basis for the structural definition of an Id and its relation to the antigen combining site. In addition, the anti-Id mAbs blocking the interaction between the idiotype (R45-45- I 1) and the antigen (Cs) could, according to the internal image concept (Jcrne et al., 1974), express structures which resemble those of the epitope recognized by the idiotype on the antigen Cs. Such particular anti-Id mAbs, classified as Ab2fl (Jerne et al., 1982), could allow a direct investigation of the structural and functional mimicry of a well-defined epitope by the Ab2P. Some aspects of the mode of action of Cs including the search for potential Cs receptors could also be addressed with Ab2@ (Gaulton and Green, 1986; Linthicum and Farid, 1987). MATERIALS

AND METHODS

Antibodies The generation, purification and characterization of the anti-Cs mAbs have been described elsewhere (Quesniaux et al., 1987a). All anti-Cs mAbs are of BALB/c origin. Conjugates of alkaline phosphatase (Sigma, St Louis, MO) and the anti-es mAb R45-45-11 were obtained essentially as described by Kearny et al, (1979) using glutaraldehyde linkage between antibody and enzyme. For biotinylation the purified mAbs (1 mg/ml) were dialyzed against 0.1 M NaCO, and reacted for 1 hr at room temp at a ratio of 1: 5 (w/w) with a biotin succinimide ester with an extended spacer arm (Sulfosuccinimidyl 6-(biotin-amido) hexanoate; Pierce Chemicals, Rockford, IL). The biotinylated mAbs were dialyzed and stored in phosphate buffered saline (PBS) containing 0.02% NaN,.

~e~~r~tion of anti-Id mAhs Eight-week-old female BALB/c or C56BL/6 mice (Madiirin, Fiillingsdorf, Switzerland) were immunized

with 25 pg purified R45-45- 11, given in a 1: 1 mixture of complete Freund’s adjuvant (Difco Laboratories, Detroit, MI) and PBS, subcutaneously in each rear footpad. At 4-day intervals the mice received three additional injections of the same amount of antigen in PBS in the rear footpads. One day after the last injection the local draining lymph nodes were removed, prepared as single cell suspensions and fused with the mouse myeloma cell line PAI-O (Stocker et al., 1982) essentially as described (Fazekas de St Groth and Scheidegger, 1980). Supernatants of growth-positive cultures were screened for anti-Id mAbs in different solid-phase ELISA formats (see below). Selected hybridomas were cloned by limiting dilution and adapted to serum-free medium. Anti-Id mAbs of the IgG isotypes were purified by affinity chromatography on Protein A according to the manufacturer instructions (Pharmacia, 1986). IgM anti-Id mAbs were partially purified by ammonium sulfate precipitation (Quesniaux et al., 1987a).

Enzyme immunoassays Anti-Id mAbs were screened and characterized using different ELISA formats (Table 1). The general conditions for the different ELISA were as follows: all reagents were diluted in PBS containing 1% bovine serum albumin (BSA; Fraction V; Boehringer, Mannheim, Germany) except the reagents coupled to the ELISA plates, which were diluted in PBS. Reaction volumes were 100 ~1 per well, except for the blocking step which required 200 ,uI per well. Incubation times were 2-3 hr at 37°C or overnight at 4”C, only streptavidin alkaline phosphatase was incubated for l/2 hr at room temp. Washings between all incubation steps were done four times with PBS containing 0.05% Tween 20 (Merck-Schuchardt, Hohenbrunn, Germany). One hour after addition of the substrate p-nitrophenyl phosphate (Sigma; 1 mg/ml in 0.1 M diethanolamine, pH 9.8) the absorbance was measured at 4051492 nm. Sandwich ELBA. The ELTSA used for screening the products of the first fusion for anti-Id mAbs was a sandwich type of ELISA (Table 1). The idiotype-carrying anti-Cs mAb R45-45-11 (IgGl/K) was coupled to

Monoclonal anti-idiotopic antibodies to anti-cyclosporine antibodies 96-well microtiter plates (NuncImmuno Pfate II F, Gibco BRL AG, Basle, Switzerland) at lO~gg/ml and subsequently incubated with anti-Id mAb containing hybridoma supernatants (1: 5 or 1: 10). Bound antibodies were detected with alkaline phosphatase conjugated R45-45-11 (0.5 hdg/ml). In this assay, the anti-Id mAbs are used to link the 1abeledmAb R45-45-11 to the solid phase-coupled mAb R45-45-11. Other IgGl/K anti-0 mAbs coated on the solid phase were used as controls for Id specificity. This sandwich ELISA should detect both combining site-associated and non-combining site-associated Ids. Competition ELISA I. For detection of combining site-associated Ids, a competition ELISA was performed (competition ELISA I; Table I). A derivative of D-Lys’Cs coupled to BSA (Quesniaux et al., 1987a) was coated onto ELISA plates at IO0 ng/ml. Purified anti-Id mAbs (l-10,000 ng/ml) were used to inhibit the binding of the biotinylated anti-Cs mAbs R45-45-I 1 or V45-271-10 (40 ng/ml) to the Cs conjugate absorbed to the solid phase. Bound biotinylated R45-45-11 was detected with streptavidin-conjugated alkaline phosphatase (Jackson Tmmuno-Research Laboratories, Avondale, PA). MAbs of the same isotype but with unrelated specificities were used as competitors in control reactions. Competition ELISA II. The cross-reactivity of the anti-Id mAbs with different anti-Cs mAbs was tested in competition ELISA II (Table 1). Anti-mouse IgG and anti-mouse IgM (affinity-purified goat IgG, 0.1 pgg/ml; Sigma) were coated onto ELISA plates. After incubation with 400 ng/ml anti-Id mAbs, pooled normal serum from BALB/c mice (1: 5000) was used to saturate the remaining free anti-mouse binding sites. The R45-45-1 Ialkaline phosphatase conjugate (0.2 pg/ml) was then added in the presence of unlabeled R45-45-11 or other anti-es mAbs (8-5000 ng/ml). Only anti-Cs mAbs carrying cross-reactive Ids should inhibit the binding of labeled R45-45-11 to the anti-Id, fixed onto the solid phase. The following anti-es mAbs were used as competitors: 26-86, 26-118, 29-355, 78-299, R14-85-1, R14212-6, R45-45- 11, R45- 109-22, R45- 180-7, R45-200-7, R45-246-2, R45-269-2, R45-620-40, R14-7-20, V14-5420, Vl4-203, V14-343, V45-180-56, V45-187-15, V45271-10 (previously described in Quesniaux et al., 1987a). IC,, values (concn of competitor required for 50% inhibition) were calculated by curve fitting using RSjl (Bolt, Beranek and Newman, Boston, MA), a commercially available data analysis system. Isot);pe ELISA. The isotypes of the anti-Id mAbs were determined by coating 0.1 @g/ml of goat antimouse IgG and IgM onto ELISA plates, incubation with 1: 50 diluted hybridoma supernatants and detection of bound mAbs with biotinylated, isotype-specific goat anti-mouse Ig antibody preparations (Southern Biotechnology Associates, Inc., Birmingham, AL) followed by a streptavidin-alkaline phosphatase conjugate.

345

penicillin, 100 pg/ml streptomycin, 2 mM L-glutamine (all Gibco) and 50pM ~-mercaptoethanol (Fluka Chemie AG, Buchs, Switzerland) in flat bottom tissue culture microtiter plates (Costar, Cambridge, MA) were stimulated with 2 pgg/ml Concanavalin A (Pharmacia). CsA (Sandoz Pharma AG, Basel, Switzerland) was used at a concn of 30 nM, which causes >90% inhibition of the proliferation. The anti-Cs mAb R45-45-11 was added either without or with the anti-Id mAbs 66 and 114 and a mAb with unrelated specificity (216; all mAbs 120 nM). After incubation for 43 hr at 37°C. 1 PCi 3H-thymidine (25 Ci/mmol; Amersham, England) was added to each well and incubated for additional 5 hr. The cells were harvested on filter paper (Inotech AC, Wohlen, Switzerland), which was washed, dried and counted after addition of scintillation liquid (Beckman Instruments, Inc., Palo Alto, CA, U.S.A.). mRNA sequencing For determination of the heavy and light chain sequences of the anti-Id mAb 114 by mRNA sequencing, hybridoma cells were expanded in RPM1 1640 (Gibco, Paisley, Scotland) containing 10% heat-inactivated fetal calf serum (Boehringer), 2 mM ~-giutamine, 0.05 mM 2-mercaptoethanol~ 100 IU/ml peniciIlin and 100 p g/ml streptomycin (all Gibco). Cells were collected and immediately frozen in liquid nitrogen. Total cellular RNA was isolated by a modification of the guanidinium/ cesium chloride method (Glisin et al., 1974). Cells (3 x lo*) were homogenized in 5 volume lysis buffer containing 5 M guanidinium isothiocyanate (Sigma, St Louis, MO, U.S.A.), 5 mM sodium citrate pH 7.0,O.l M fi-mercaptoethanol and 0.5% sarkosyl. The homogenate was layered on 5.7 M CsCl in 0.1 M EDTA pH 7.5 and centrifuged at 1125OOg ~25000 rpm, Beckman rotor SW28) for 20 hr. The RNA pellet was dissolved in 0.3 to 1 ml water, and extracted twice with phenoI/~hloroform (l/l v/v) and once with ether. After ethanol precipitation the total RNA was resuspended in water and kept at - 20°C. Polyadenylated RNA was prepared by means of oligo(dT)-cellulose T7 (Pharmacia LKB Biotechnology, Uppsala, Sweden), as described by Maniatis et al. (1982). The direct sequencing of variable region mRNA was performed by a modification of the method described by Geliebter et al. (1986). Poly(A)+RNA (2-4 pg) were allowed to anneal with 20 ng of y3’P-ATP end-labeled primer. The elongation occurs in 3.3 ~1 RT buffer (24 mM Tris-HCI, pH 8.3; 16 mM MgC12, 8 mM DTT; 0.4 mM dATP; 0.4 mM dCTP; 0.8 mM dGTP; 0.4 mM dTTP) containing AMV reverse-transcriptase (5 U/ Boehringer), in the presence of one reaction; ddNTP/reaction (0.8 ~1 of 1 mM ddATP, or 0.5 mM ddCTP, or 1 mM ddGTP, or 2 mM ddTTP). After 45 min incubation at 5O”C, the cDNA fragments were analyzed on a 6% acryamide sequencing gel and detected by autoradiography.

Spleelz ceil proiiferaiion assay Spleen cells of C57BL/6 mice (Iffa Credo, Lyon, France; 2 x lo5 per well) in 200 ~1 RPM1 1640 containing 10% inactivated fetal calf serum, 100 units/ml

Amino acid sequencing The heavy and light chains of anti-Id mAb I14 were separated as described (Kocher et aE., 1980). The isolated

346

G. ZENKEef ai.

Table 2. Specificity of anti-Id mAbs in the sandwich ELISAff Anti-Id

R45-45- 11

Medium 135 G12 G9

0.002’ 1.117 1.715 0.437

Anti-Cs mAbsb 2685 26-118 0.065 0.084 0.080 0.106

0.093 0.092 0.099 0.113

78-214 0.119 0.119 0.119 0.137

“Anti-Cs mAbs were adsorbed to ELISA plates and incubated with anti-Id mAb supernatants. Bound mAbs were detected with alkaline phosphatase labeled R45-45-I 1. ‘All anti-Cs mAbs are of the IgGl/K isotype. R45-4% f I recognizes BSA-Lys”-Cs and BSA-Thr?-Cs, whereas 26-85,26-l 18 and 78-214 recognize BSATh&Cs only (Quesniaux et al., 1990). ‘BLISA absorbance values.

light chain of the anti-Id mAb 114 was applied to Polybrene (Sigma) treated glass-fiber disks and subjected to amino acid sequence determination on an Applied Biosystems 470A protein sequencer equipped with an on-line phenylthiohydanto~n derivative detection system according to standard procedures.

Gmeration of anti-Id mAbs

The immunization protocol for the generation of syngeneic or allogeneic anti-Id mAbs followed the procedure described by Kearny (in Paul, 1984), which was adapted from immunization schedules developed by Lieberman et at. fI972) for the production of conventional heterogeneous ante-idiotypi~ antisera. The immunization of RALBJc mice with the syngeneic anti-es mAb R45-45-I I should not elicit mAbs directed against the constant part of the R45-45-I I. Therefore, we used as primary screening for anti-Id mAbs a sandwich ELISA in which the idiotype-carrying anti-Cs mAb R45-45-1 I was used both to coat plates as well as in its labeled form to detect antibodies bound to coated plates (Table 1). This ELISA format should detect any mAbs directed against Ids, irrespective of whether they are associated with the combining site or not. Three mAbs, selected from a first fusion, were significantly positive in this assay (Table 2, second column). The speciftcity of these mAbs for Ids of the mAb R45-45-I 1 was subsequently confirmed, since they did not react with antiCs mAbs of the same isotype (IgG I> but of different fine specificity (Table 2; anti-Cs mAbs 26-85, 26-l 18, 78-214; Quesniaux et al., 1990). To control that equivalent amounts of these anti-Cs mAbs were coated to the ELISA plates, bound mAbs were revealed with rabbit anti-mouse IgGI and a goat anti-rabbit Ig alkaline phosphatase conjugate in parallel experiments (data not shown). Isotype determinations of the anti-Id mAbs showed that all of them were IgMlE, indicating that the sandwich ELBA used for screening might. preferentially select IgM antibodies. Because of their pentameric struc-

ture IgM anti-Id mAbs are probably able to link the labeled R45-45 I I mAb to the solid-phase coupled R4545-11 more easily than IgG mAbs. Anti-Id mAbs spec$c

-for combining site-associcrted Id.T

In order to obtain IgG anti-Id mAbs we immunized C56BL/6 mice with the BALB/c derived R4S-45 11 mAb to use the difference in IgH allotype as a carrier effect. This required a more specific screening to avoid detection of anti-allotype antibodies. In addition, we wanted to select preferentialfy anti-Id mAbs specific for combining site-associated Ids. Therefore, the competition ELISA I was devetoped in which the binding of labeled R45-45-I 1 to solid-phase coupled BSA-CsA was inhibited by the anti-Id mAbs (Table I). From a second fusion (C57BL/6) two anti-Id mAbs (66 and 114) were selected, which inhibited the binding of R45-45-11 to GSA in a dose dependent fashion (Fig. 1; 50% inhibition at 32 ng/ml of anti-Id mAb 66 and 36 ng/ml of anti-Id mAb 114). The two anti-Id mAbs 66 and 114 are of the IgG2a/x and IgGl/rc isotype, respectively. The three IgM anti-Id mAbs selected from the first fusion fBALB/c) with the sandwich ELISA were also analyzed in this competition ELISA (Fig. I). The anti-Id mAb G12 inhibited the binding of R45-45-l t to Cs although not as effectively as the anti-Id mAbs 66 and 1I4 (50% inhibition at 470 ng/ml). The anti-Id mAbs G9 and 135 showed no inhibition (50% inhibition not reached at 10,000 ng/ml). IgM/k- and IgGI/K mAbs with unrelated specificity did not show any significant inhibition in this assay (data not shown). To exclude that these results were due to differences in the affinities of the anti-Id mAbs for their Ids, relative affinities of purified anti-Id mAbs were determined in ELISA. AII the IgM anti-Id mAbs (135, G9 and GlZf showed similar affinities to solid-phase coupled R45-45 f I, whereas the relative affinities of the two IgG anti-Id mAbs 66 and 114 were fivefold higher [data not shown). It is unlikely that these relatively small difl’erences in affinity would cause the approximately 300-fold difference in inhibitory capacity of the anti-Id mAbs G9 and 135 compared to the anti-Id mAbs 66 and 114 in the competition ELISA. 7001

Antl-ld

mAbs

[lg/ml]

Fig. 1. Inhibition of R45-45-l I binding to Cs by anti-Id mAbs in competition ELISA I (Table 1). Purified anti-Id mAbs (I 14: a; 66: ,A_;G12: /‘J; G9: A; 135: 0) were used to inhibit the binding of biotinyiat~d R45-45-1 I to solid-phase coupled Cs. The TC, values (antibody conca required for 50% i~l~~bition) were 32 ng/ml for anti-Id mAb 66,36 ngjml for 114.470 ngiml for Gi2 and > fO$HOngjmf For G9 and 135.

Monoclonal Table

binding

anti-idiotopic

antibodies

Anti-Id 66

66 114 Gl2 135 G9

mAb-biotin 114

6’ II 7 I8 >33

antibodies

347

and Gl2 are spatially closely related. In addition, the non-combining site-associated Id recognized by anti-Id mAb 135 seems to be spatially related to the combining site-associated Ids although the second non-combining site-associated Id recognized by anti-Id mAb G9 is not.

3. Mutual competition of anti-Id mAb to R45-45-11 in mutual inhibition ELISA”

Inhibitor

to anti-cyclosporine

6 6 9 14 >33

Cross-reactivity of the anti-Id mAbs with d$erent antiCs mAbs Cross-reactivity of the anti-Id mAbs with different anti-Cs mAbs was analyzed in competition ELISA II, where a series of unlabeled antiCs mAbs were used to compete with the binding of labeled R45-45-11 to the anti-Id mAbs (See Table 1). As shown in Fig. 2, in addition to unlabeled R45-45-11 (homologous inhibitor) only one anti-es mAb (V45-271-10) inhibited the binding of labeled R45-45-11 to all the different anti-Id mAbs. Equivalent inhibition was obtained with V45271-10 and with the homologous inhibitor R45-45-11 in this system. These results indicate that the anti-Cs mAb V45-271-10 expresses all the Ids recognized by the five anti-R45-45 11 anti-Id mAbs. In a second series of experiments, the association of the Ids recognized by the five anti-Id mAbs with the combining site of the anti-Cs mAb V45-271-10 was investigated using competition ELISA I. The binding of labeled V45-27 1- 10 to solid-phase coupled BSA-CsA was inhibited by the anti-Id mAbs, showing very similar results to those obtained with labeled R45-45-11 (Fig. 3). From these results we conclude that the two anti-Cs mAbs R45-45- 11 and V45-27 1- 10 are idiotypically indistinguishable. This idiotypic similarity is in agreement with the high sequence homology of these two mAbs since their two light chains are identical and the heavy chains differ only in one amino acid (a phenylalanine to tyrosine substitution in CDR2; Schmitter et al., 1990). The two anti-Cs mAbs R45-45- 11 and V45-27 1- 10 which were derived from the same mouse in two different fusion experiments recognize very similar and extended epitopes on the Cs molecule in contrast to most of the anti-Cs mAbs which recognize more restricted epitopes (Quesniaux et af., 1987a; 1990). Additional 18 anti-Cs mAbs tested in the competition ELISA II did not show

“Binding of biotinylated anti-Id mAb to sohdphase coupled R45-45- 1I in the presence of unlabeled anti-Id mAb (inhibitor). “Molar ratio of inhibitor to labeled anti-Id mAb required for 50% inhibition. Three groups of anti-Id mAbs could therefore be identified: (1) The two anti-Id mAbs 66 and 114 which recognize combining site-associated Ids, (2) the anti-Id mAb G 12 which recognizes an Id only partially overlapping with the combining site, and (3) the two anti-Id mAbs 135 and G9 which seem to be specific for Ids not associated with the combining site of R45-45-1 I.

In order to analyze the spatial relationship of the Ids, competition experiments were performed with the antiId mAbs in a mutual inhibition ELISA (as described in Table 1). Binding of biotinylated anti-Id mAb 66 to the Id-carrying anti-Cs mAb R45-45-11 was inhibited equally well by unlabeled anti-Id mAb 66 as by anti-Id mAb 114 and conversely biotinylated anti-Id mAb 114 was inhibited equally well by anti-Id mAbs 114 and 66 (Table 3). Binding of the biotinylated anti-Id mAbs 66 and 114 was as effectively inhibited by the anti-Id mAb Gl2 as by the homologous anti-Id mAbs 66 and 114 (Table 3). The anti-Id mAb 135 competes with the binding of the labeled anti-Id mAbs 66 and 114, although less effectively than the anti-Id mAb G12 and no inhibition was observed with the anti-Id mAb G9. The preceding results confirm that the three combining site-associated Ids recognized by anti-Id mAbs 66, 114

,_ PC / Anti-Id mAbs

66

-7

2 z

114

loo_ r r 50 o_,;,‘lO,OOOng/ml for G9 and 135.

A. 1 D GAC

IV A’M

GTG

L CTG

T ACC

Q CAA

S TCI.

P CCA

10 A S GCX TCT

L Tl’G CDR I

A GCT

V GTG

40

S TCT 30

L CTC

G Q GGG CAG

50

GIPARFSGSGSRTDFTLTIN GGC ATC CCT GCC AGG 80 P V E A D CCT GTG GAG GCI- GAT 100 T F G G G ACG TTC GGT GGA GGC I JKl

20

R A T AGG GCC ACC

CDR 2

70 ‘ITC

AGT

GGC

AGT

GGG TLT

D GAT

V GlT

A GCA

T ACC

Y TAT

T ACC

K AAG

L CI’G

E GAA

I ATC

Y TAC 107 K AAA

AGG ACA C TGT

Q CAG

GAC 90 Q CAA

l-X S AGT

ACC D GAT

CTC CDR

3

E GAA

ACC

All-

AAT

N AAT

P CCT

R CGG t-

b

B. 20

IO

;AYLQQSGAELVRPGSSVKM CAG GCT TAT mA CAG

CAG

TCI-

GGG GCl- GAG 30

Cl-G

SCKASGYTFTSYNMHWVKQT TCC TGC AAG GCT T(TI

GGC

TAC

ACA

TIT

ACC 50

AGI- TAC

AAT

ATG

CAC

Cl-G GAA

TGG

Al-l-

GGA

GCT

Y TAT

P CCA

G GGA

K AAG

A T GCC UCTG

A’IT 70 T ACT

T ACA

S TCT

PRQGLEWIGAI CCT AGA CAG 60

80 M ATG Y TAC -

GGC

L S S AGC AGC CTG CDR 3 S G S I D TCC GGT AGT ATA GAC DFL16.1__, .

Q CAG

L CTC

Y TAC

L

E D GAA GAC 105 W G Q TGG GGC CAA

S TCT

GTA

GTG AAG

ATG 40

TGG

ACA

N AAT

GI’A AAG CAG CDR 2 G D T S GGT GAT ACT TCC

V D Gl-A GAC

S TCC

S TCC

A GCG

K AAA 90 V Y GTC TAT

Y TAC 80 Y TAC

F C ‘I-l-C TGT

A GCA

T ACT

T ACA

S TCC

G T GGC ACC IfiT

AGG CCT CDR I

L CTC

GGG TCC

V GTC

TCA

S ACi

T ACA

A GCC

AGA Rlzg GGG GAT 113 S TCA b

Fig. 4. Nucleotide and amino acid sequences of the anti-Id mAb 114 variable region. The amino acid numbers and the complementarity determining regions (CDR, in boxes) are defined as in Kabat et al. (1987). Two synthetic oligonucleotides per chain were used as primers. A. Light chain. 5’ CACTGGATGGTGGG 3’ hybridizes to the 5’ end of the constant domain of the x chain (Altenburger et al., 1981). 5’ GGTGGCCGTCCTG 3’ hybridizes to the underlined sequence of the light chain variable region. B. Heavy chain. 5’ GCAGCAGATCCAGC 3’ hybridizes to the 5’ end of the first constant domain (CHl) of the y 1 chain (Honjo et al..79). 5’ GTGTGGCCTTGCC 3’ hybridizes to the underlined sequence of the heavy chain variable region.

Monoclonal

anti-idiotopic

antibodies

resemble some structures of Cs. The V, and V, regions sequences were determined by direct mRNA sequencing using two specific primers. The first primer is complementary to a part of the constant region directly preceding the rearranged variable region. The second oligonucleotide is chosen in the variable region, and is complementary to a sequence in the framework region 2 or 3. The nucleotide and deduced amino acid sequences of the anti-Id mAb 114 light and heavy variable regions are presented in Figs 4A and 4B, respectively. The 68 first amino acids of the light chain were confirmed by amino acid sequencing. The V, sequence was found to belong to the V, group 3 family according to Kabat’s classification (Kabat et al., 1987). All the invariant amino acids characteristic of that group are present in the V, of the anti-Id mAb 114. To form the V, region, a J,, gene (Sakano et al., 1979) is recombined with a V,21C gene (Heinrich et al., 1984). The expressed V, gene shows a very high homology (97%) with the V,21C germline gene. The V, sequence could be assigned to the heavy chain subgroup 5A. The V, of the anti-Id mAb 114 displays all characteristic invariant amino acids except three out of 60: Ala 2, Arg 42 and Gln 62 instead of Val, Gly and Glu, respectively. The VH sequence is also close to the heavy chain subgroup 2b; in this case only one of the invariant amino acids is different (Ala 2 instead of Val). To form the V, region, a DLF16.1 (Tonegawa, 1983) and a J,, gene (Sakano et al., 1980) are joined to a V, gene of the 5558 family. The recombined VH gene reveals 85% homology with VH3 germline gene (Bothwell et al., 1981). The heavy and light chain sequences will be used to interprete the X-ray data obtained from the Fab fragment of this anti-Id mAb either alone or in a complex with the idiotope R45-45-11. The comparison of the sequences of the anti-Id mAb 114 and the anti-Cs mAb R45-45-11 showed that they exhibit enough sequence differences to be distinguished in the X-ray electron density map of the complex. Since the antigen Cs is of peptidic nature it would be tempting to speculate whether an ‘internal image’ of Cs could be detected in the amino acid sequence of the CDRs of anti-Id 114. However, this is rather unlikely, since Cs is made of mostly non-natural amino acids and is N-methylated in four peptide bonds. Effect of anti-G mAbs and anti-Id mAbs on Cs immunosuppression in vitro To probe for a possible structural and functional mimicry of Cs by the anti-Id mAbs, we took advantage of the functional properties of Cs (Bore1 et al., 1989) and its affinity for the cytosolic binding protein cyclophilin (Handschumacher et al., 1984). In order to test whether the anti-Id mAbs which recognize combining-site associated Ids would affect the reversal of Cs-mediated immunosuppression by the antiCS mAb R45-45-11 in vitro we analyzed the effects of combinations of these mAbs on the proliferation of Concanavalin A activated mouse spleen cells (Fig. 5).

to anti-cyclosporine

349

antibodies cpm [xl o-31

CsA

Anti-Cs

AntI-Id

+

-

+

+

+

+

+

+

66

+ --

+

216 E

0

100

200

300

-I 114

Fig. 5. Effect of anti-Cs mAbs and anti-Id mAbs on Cs immunosuppression in vitro. Concanavalin A activated mouse spleen cells were incubated with CsA, CsA and R45-45-11 or CsA and R45-45-11 in combination with the anti-Id mAbs 66 and 114 and the control mAb 216 as described in Materials and Methods. The anti-Id mAbs reverse the blocking effect of R45-45-11 on Cs immunosuppression at an equimolar ratio.

CsA at a concn of 30 nM causes a total suppression (97%) of the proliferation which can be nearly completely reversed by a 4-fold molar excess of the anti-Cs mAb R45-45- 11. The two combining site-associated anti-Id mAbs 66 and 114 reverse the blocking effect of the anti-Cs mAb R45-45-11 when added in equimolar concns. A mAb with unrelated specificity had no effect. This indicates that the anti-Id mAbs very effectively compete with Cs in the combining site of R45-45-11. The anti-Id mAbs 66 and 114 could therefore be anti-Id mAbs of the internal image type. These two anti-Id mAbs were further tested directly in the mouse spleen cell proliferation assay. They did not reproduce a Cs-like immunosuppressive effect nor did they antagonize Cs immunosuppression in this system (data not shown). In order to show structural mimicry between an anti-Id mAb and the antigen Bentley et al. (1989) generated Ab3 against the anti-Id mAb and tested them for binding to the antigen. We examined recognition of the anti-Id mAbs by the natural cytosolic Cs-binding protein cyclophilin. No significant binding of cyclophilin to the anti-Id mAbs 66 and 114 could be observed (data not shown). There could be several reasons for these findings: The anti-Id mAbs are combining site-associated but no real internal images, or they are internal images of the epitope recognized by the anti-Cs mAb R45-45-11 but the mimicry in structural terms does not completely reproduce the active region of Cs. The residues 1, 2, 3, 10 and 11 of Cs are essential for the immunosuppressive activity of Cs as shown by structure and activity studies of Cs using a great variety of synthetic and natural Cs derivatives (Wenger et al., 1986). Residues 1, 2, 10 and 11 were also shown to be involved in the interaction of Cs with cyclophilin (Quesniaux et al., 1987b). Thus, the generation of mAbs directed towards the Cs-binding site of cyclophilin seems to be particularly interesting. However, efforts to raise mAbs with that specificity have been unsuccessful so far, probably because of the high evolutionary conservation of cyclophilin (Harding et al., 1986). In order to prepare anti-Id mAbs of the internal image type, we selected the anti-Cs mAb R45-45-11 whose epitope residues (1, 2, 3, 4, 5, 9 and 11) showed maximal overlap with the region

350

G. ZENKE et al.

of Cs required for immunosuppressive activity and cyclophilin binding (Quesniaux et al., 1990). The finding that the anti-Id mAbs did not interfere with the Cs-cyclophilin interaction and did not show Cs-like activity would suggest that one prerequisite for obtaining internal image anti-Id mAbs mimicking Cs immunosuppressive activity would be to use an anti-Cs mAb that recognizes an epitope corresponding more precisely to the residues involved in cyclophilin-binding and required for immunosuppressive activity of Cs. REFERENCES

Altenburger W., Neumaier P. S., Steinmetz M. and Zachau H. G. (1981) DNA sequence of the constant region of the mouse immunoglobulin kappa chain. Nucl. Acids Res. 9, 971-981. Altschuh D., Kocher H. P., Quesniaux V. F. J., Schmitter D., Van Regenmortel M. H. V. and Thierry J. C. (1989) Crystallization and preliminary X-ray investigation of a complex between a Fab fragment and its antigen, cyclosporin. J. mol. Biol. 209, 177-178. Bentley G. A., Bhat T. N., Boulot G., Fischmann T., Navaza J., Poljak R. J. and Riottot M. M. (1989) Immunochemical and crystallographic studies of antibody D1.3 in its free, antigen-liganded, and idiotope-bound stages. Cold Spring Harbor Symposia on Quantitative Biology LIV, 239-245. Bentley G. A., Boulot G., Riottot M. M. and Poljak R. J. (1990) Three-dimensional structure of an idiotope-antiidiotope complex. Nature 348, 254-257. Bore1 J. F., Di Padova F.. Mason J., Quesniaux V., Ryffel B. and Wenger R. (1989) Pharmacology of cyclosporine (Sandimmune). Pharmac. Rev. 41, 239434. Bothwell A. M., Paskind M., Reth M., Imanishi-Kari T., Rajewsky K. and Baltimore D. (1981) Heavy chain variable region contribution to the NPb family of antibodies: Somatic mutation evident in a gamma 2a variable region. Cell 24, 625-637. Boulot G., Rojas C., Bentley G. A., Poljak R. J., Barbier E., Le Guern C. and Cazenave P. A. (1987) Preliminary crystallographic study of a complex between the Fab fragment of a monoclonal anti-lysozyme antibody (D1.3) and the Fab fragment from an anti-idiotopic antibody against D1.3. J. mol. Biol. 194, 577.-579. Davie J. M., Seiden M. V., Greenspan N. S., Lutz C. T., Bartholow T. L. and Clevinger B. L. (1986) Structural correlates of idiotopes. A. Rec. Immun. 4, 147-165. Gaulton G. N. and Greene M. I. (1986) Idiotypic mimicry of biological receptors. A. Rev. Immun. 4, 253-280. Geliebter J., Zeff R. A., Melvold R. W. and Nathenson S. G. (1986) Mitotic recombination in germ cells generated two major histocompatibility complex mutant genes shown to be identical by RNA sequence analysis: Kbm9 and Kbm6. Proc. natn. Acad. Sci. U.S.A. 83, 3371-3375. Glisin V., Crkvenjakov R. and Byus C. (1974) Ribonucleic acid isolation by cesium chloride centrifugation. Biochemistry 13, 2633-2637. Handschumacher R. E., Harding M. W., Rice J., Drugge R. J. and Speicher D. W. (1984) Cyclophilin: a specific cytosolic binding protein for cyclosporin A. Science 226, 544547. Harding M. W., Handschumacher R. E. and Speicher D. W. (1986) Isolation and amino acid sequence of cyclophilin. J. hiol. Chem. 261, 8547-8555.

Heinrich G., Traunecker A. and Tonegawa S. (1984) Somatic mutation creates diversity in the major group of mouse immunoglobulin kappa light chains. J. exp. Med. 159, 417435. Honjo T., Obata M., Yamawaki-Kataoka Y., Kataoka T., Kawakami T., Takahashi N. and Mano Y. (1979) Cloning and complete nucleotide sequence of mouse immunoglobulin gamma 1 chain gene. Cell 18, 559-568. Jerne N. K. (1974) Towards a network theory of the immune system. A. Immun. (Inst. Pasteur) 125 C, 373-389. Jerne N. K., Roland J. and Cazenave P. A. (1982) Recurrent idiotopes and internal images. EMBO J. 1, 243-247. Kabat E. A., Wu T. T., Reid-Miller M., Perry H. M. and Gottesman K. S. (1987) Sequences of Proteins of Immunological Interest. Fourth Edition. U.S. Department of Health and Human Services. Kearney J. F. (1984) Hybridomas and monoclonal antibodies. In Fundamental ImmunologJj (Edited by Paul W.) p. 755. Raven Press, New York. Kearney J. F., Radbruch A.. Liesegang B. and Rajewsky K. (1979) A new mouse myeloma cell line that has lost immunoglobulin expression but permits the construction of antibody-secreting hybrid ceil lines. J. Immun. 123, 1548-1550. Kocher H. P., Berek C., Schreier M. H., Cosenza H. and Jaton J. C. (1980) Phosphorylcholin-binding hybridoma proteins of normal and idiotypically suppressed BALB/c mice. 11. Variable region N-terminal amino acid sequences. Eur. J. Immun. 10, 264-267. Lieberman R. and Humphrey W. (1972) Association of H-2 types with genetic control of immune responsiveness to IgG (gamma 2a) allotypes in the mouse. J. e.up. Med. 136, 1222-1230. Linthicum D. S. and Farid N. (I 987) Anti-idiotypes, Receptors and Molecular Mimicry. Springer, New York. Loosli H. R., Kessler H., Oschkinat H., Weber H. P.. Petcher T. J. and Widmer A. (1985) The conformation of ‘cyclosporin A’ in the crystal and in solution. Hell.. chim. Acta 68, 682-704. Maniatis T., Fritsch E. F. and Sambrook J. (1982) Molecular Cloning, A Laboratory Manual. Cold Spring Harbor Laboratory, New York. Petcher T. J., Weber H. P. and Rueegger A. (1976) Crystal and molecular structure of an iodo-derivative of the cyclic decapeptide cyclosporin. A. Hell:. chim. Acta 59, 1480-1488. Quesniaux V., Tees R., Schreier M. H.. Wenger R. M., Donatsch P. and Van Regenmortel M. H. V. (1986) Monoclonal antibodies to cyclosporin. In Cyclosporin. Progress in Allergy (Edited by Bore1 J. F.) Vol. 38, pp. 108-122. Karger, Basel. Quesniaux V., Tees R., Schreier M. H., Wenger R. M. and Van Regenmortel M. H. V. (1987a) Fine specificity and crossreactivity of monoclonal antibodies to cyclosporine. Moles. Zmmun. 24, 1159-l 168. Quesniaux V. F. J., Schmitter D., Schreier M. H. and Van Regenmortel M. H. V. (1990) Monoclonal antibodies to cyclosporin are representative of the major antibody populations present in antisera of immunized mice. Molec. Immun. 27, 227-236. Quesniaux V. F. J., Schreier M. H., Wenger R. M.. Hiestand P. C., Harding M. W. and Van Regenmortel M. H. V. (19876) Cyclophilin binds to the region of cyclosporine involved in its immunosuppressive activity. Eur. J. fmmun. 17, 1359-1365.

Monoclonal

anti-idiotopic

antibodies

Roux K. H. and Metzger D. W. (1982) Immunoel~tron microscopic localization of idiotypes and allotypes on immunoglobulin molecules. J. Immun. 129, 780-783. Roux K. H., Monafo W. J., Davie J. M. and Greenspan N. S. (1987) Construction of an extended three-dimensional map by electron microscopic analysis of idiotope-anti-idiotope complexes. Proc. natn. Acad. Sri. U.S.A. 84, 4984-4988. Sakano H., Hueppi K., Heinrich G. and Tonegawa S. (1979) Sequences at the somatic recombination sites of immunoglobulin light-chain genes. Nature 280, 188-294. Sakano H., Maki R., Kurosawa Y., Roeder W. and Tonegawa S. (1980) Two types of somatic recombination are necessary

to anti-cyclosporine

antibodies

351

for the generation of complete immunoglobuiin heavy-chain genes. Nature 286, 676683. Schmitter D., Poch O., Zeder G., Heinrich G. F., Kocher H. P., Quesniaux V. F. J. and Van Regenmortel M. H. V. (1990) Analysis of the structural diversity of monoclonal antibodies to cyclosporine. Molec. Irnmun. 27, 1029-1038. Tonegawa S. (1983) Somatic generation of antibody diversity. Nature 302, 575-58 1. Wenger R. M. (1986) Cyclosporines and analogues. Isolation and synthesis. Mechanism of action and structural requirements for pharmacological activity. Progr. Chem. Organic Natural Products 50, 123-168.