Somatic mutation and light chain rearrangement generate autoimmunity in anti-single-stranded DNA transgenic MRL/lpr mice.

Antibodies to single-stranded (ss)DNA are expressed in patients with systemic lupus erythematosus and in lupus-prone mouse models such as the MRL/Mp-lpr/lpr (MRL/lpr) strain. In nonautoimmune mice, B cells bearing immunoglobulin site-directed transgenes (sd-tgs) that code for anti-ssDNA are functionally silenced. In MRL/lpr autoimmune mice, the same sd-tgs are expressed in peripheral B cells and these autoantibodies gain the ability to bind other autoantigens such as double-stranded DNA and cell nuclei. These new specificities arise by somatic mutation of the anti-ssDNA sd-tgs and by secondary light chain rearrangement. Thus, B cells that in normal mice are anergic can be activated in MRL/lpr mice, which can lead to the generation of pathologic autoantibodies. In this paper, we provide the first direct evidence for peripheral rearrangement in vivo.

An idiotype D23-bearing polyspecific, murine anti-DNA monoclonal antibody forms glomerular immune deposits. Pathogenic role of natural autoantibodies?

Identification of the immunochemical and structural properties of pathogenic anti-DNA antibodies is a major goal for understanding their origins and the mechanisms whereby they induce tissue lesions. Herein, we report on the production of an IgG2a,k anti-DNA monoclonal antibody (4B1), derived from a 12-month-old (NZB x NZW)F1 lupus mouse, able to form glomerular immune deposits. mAb 4B1 is a polyspecific antibody able to bind to ssDNA, actin, tubulin, cardiolipin and to laminin as shown by solid phase ELISAs. Indirect immunofluorescence labeling of HEp-2 cells gave a cytoplasmic staining pattern similar to that obtained with anti-cytoskeleton antibodies. Western blot analysis demonstrated that mAb 4B1 bore idiotype D23, previously shown to be characteristic of natural antibodies derived from normal mice. After injecting the 4B1-secreting hybridoma intraperitoneally into normal (NZW x BALB/c)F1 mice, glomerular immune deposits were observed along the capillary wall. These deposits contained mainly IgM, IgG2a and mAb 4B1, as demonstrated by direct immunofluorescence using a biotinylated-rat anti-4B1 idiotype mAb and kidney eluate analysis. Nucleotide sequence analysis of the VH and VL genes showed that mAb 4B1 is encoded by VH Q52, DSP2.9 and JH2 genes with minimal mutations and by VK8 very similar to the canonic D23 light chain, and JK1 germline genes. No arginine residues were observed in the VH CDR and both chains lacked N-segment addition. Thus, no structural characteristics deduced from the primary structure of mAb 4B1 could explain its pathogenic potential. However, the immunochemical and structural properties suggest that autoantibodies closely related to natural autoantibodies may be pathogenic.

Modeling of anti-nucleosome immunoglobulin Fv domains: analysis of electrostatic interactions.

Three-dimensional structural models of six murine anti-(H2A-H2B) monoclonal autoantibody variable fragments were built by comparative molecular modeling using the COMPOSER software. Analysis of the antibody combining sites is based on the hypothesis that ionic and/or electrostatic interactions predominate in antigen antibody binding, as suggested by the cationic nature of histones and the amino acid sequences of the antibody hypervariable regions. The study of the electrostatic potentials of their combining site surfaces, computed with the MOLCAD software, and the comparison with the electrostatic potentials of 13 selected control mAbs show the lack of a unique electrostatic pattern. One group of three mAbs expresses a strong and large electronegative area, supporting the hypothesis that ionic interactions predominate in antigen recognition. The second group, containing the other three mAbs, exhibits an alternation of electropositive and electronegative areas. All, however, present a localized electronegative area in the vicinity of H-CDR1 and H-CDR2 loops that is generated by the presence of at least one acidic residue. The model suggesting that the binding activity may depend on charged residues at the same site is reminiscent of what was previously reported in anti-DNA mAbs. In addition, the alternation of electropositive areas and electronegative areas in second group mAbs is also frequently observed in certain anti-DNA mAbs. These data argue for the existence of relationships between these two autoantibody populations and suggest that they share a common immunogenic particle formed by anionic and cationic components, such as a nucleosome.

[Lupus diseases]. / Les maladies lupiques.

Systemic lupus erythematosus is a multifactorial, non organ-specific autoimmune disease. Studies of the disease in humans, of spontaneous murine models and of induced experimental models have made it possible to identify the various factors involved. These factors act on the immune system, the abnormalities of which include polyclonal B cell activation, selection of autoreactive B clones by autoantigen, and intervention of helper T cells. Other abnormalities, such as disturbance of the idiotypic network and alteration of suppressor T cell function, are not excluded.

Structural characterization of an (NZB x NZW)F1 mouse-derived IgM monoclonal antibody that binds through V region-dependent interactions to murine IgG anti-DNA antibodies.

Among B/W-derived IgM mAb, 12.5H was selected because of its ability to recognize, in solid phase ELISAs, IgG mAb with DNA-binding activity. mAb 12.5H bound preferentially to IgG2a mAb and did not react with B/W- or BALB/c-derived IgG2a mAb with no DNA-binding activity, suggesting V region-mediated interactions. mAb 12.5H also bound to IgG2a isolated from the sera of old B/W mice but did not react with Ig present in the sera of young B/W (< 6 months) or BALB/c mice. Further analysis of IgM/IgG interaction showed that the binding of 12.5H to IgG polyclonal anti-DNA antibodies could be inhibited by DNA and that mAb 12.5H bound to the F(ab')2 fragments of B/W-derived IgG2a anti-DNA mAb, demonstrating that the interaction occurred through the variable regions of the molecules. When the antigen-binding capacity of mAb 12.5H was evaluated, it was demonstrated to bind to self-antigens such as myosin, actin, tubulin and histones, to bind poorly to ssDNA and not to react with dsDNA. mAb 12.5H gave a cytoplasmic staining pattern by indirect immunofluorescence on HEp-2 cells. Nucleotide sequence analysis of the H and L V genes showed features previously demonstrated to be recurrent in two categories of SLE autoantibodies, i.e. arginine residues in the VHCDR3 and at position 96 on the light chain, both considered to be characteristic of antibodies reacting with dsDNA and acidic amino-acid residues in VHCDR2 reported to be frequent on antihistone antibodies. Taken together, these results show that the B/W mouse repertoire contains IgM autoantibodies: (i) that react in an idiotypic manner with DNA binding antibodies and (ii) that, because of structural characteristics, may constitute the common precursor of different categories of SLE autoantibodies, and the prototype of the idiotypically connected SLE autoantibodies accounting for the production of autoantibodies upon immunization with cognate idiotype and the experimental model of cross-idiotype-induced lupus.

Do naturally occurring autoantibodies participate in the constitution of the pathological B-cell repertoire in systemic lupus erythematosus?

Systemic lupus erythematosus (SLE) is the prototype of systemic autoimmune diseases. In both human and mouse SLE diseases, the autoimmune response targets a restricted set of autoantigens. Many of them are nucleic acids and proteins involved in the synthesis and processing of DNA or RNA, a characteristic which should be taken into consideration to elucidate the origins of non organ-specific autoantibodies. Several observations, in particular those obtained from experimental models of SLE induced in normal mice, suggest that the breakdown of B-cell tolerance occurs in the periphery. Herewith, we present data further supporting the proposition that SLE-associated autoantibodies originate from natural autoantibody-secreting B cells activated in the internal environment of lupus mice. Thus, one may hypothesize that certain clones of the expanded primary B-cell repertoire are selected to differentiate into harmful IgG autoantibody-secreting clones, thereby raising the question of the nature of immunogenic structures involved in SLE. Our analysis of the immunochemical and structural properties of anti-nucleosome and anti-myeoloperoxidase monoclonal antibodies derived from (NZB x NZW)F1 mice leads us to propose that complexes formed by the association of DNA and DNA-binding proteins and, more generally, by anionic molecules associated with proteins, possess a selective advantage over other autoantigens to induce the differentiation of certain B-cell clones and the very special profile of the SLE-autoimmune response. These DNA/DNA-protein complexes could also play a role in the activation of the T-cell compartment in SLE.

Idiotypic analysis of anti-nucleosome monoclonal antibodies derived from lupus mice.

3F6 and 2E1 are anti-(H2A-H2B) monoclonal antibodies derived from a 12-month-old (NZBxNZW)F1 mouse that diverged from the same clonal precursor by somatic mutations. Rabbit anti-idiotypic antisera were prepared against these two monoclonal antibodies and used as probes to analyse the properties and expression of 3F6 and 2E1 idiotypes. Both idiotypes were conformational, distant from the antigen binding site and did not correlate with a VH- or VL-chain usage. 3F6 was preferentially bound by anti-3F6 idiotype but was weakly recognized by anti-2E1 idiotype suggesting, since 3F6 derives from 2E1, that 3F6 bore idiotypic determinants generated by somatic mutations. While none of the murine anti-DNA monoclonal antibodies tested expressed 2E1 or 3F6 idiotypes, 3F6 idiotype could be detected on approximately one-third of anti-(H2A-H2B) monoclonal antibodies derived from other lupus strains of mice, demonstrating the presence of cross-reactive idiotypes on autoantibodies directed against a nucleosome that could result from somatic mutations.

Structural models of antibody variable fragments: a method for investigating binding mechanisms.

The value of comparative molecular modeling for elucidating structure-function relationships was demonstrated by analyzing six anti-nucleosome autoantibody variable fragments. Structural models were built using the automated procedure developed in the COMPOSER software, subsequently minimized with the AMBER force field, and validated according to several standard geometric and chemical criteria. Canonical class assignment from Chothia and Lesk's [Chottin and Lesk, J. Mol. Biol., 196 (1987) 901; Chothia et al., Nature, 342 (1989) 877] work was used as a supplementary validation tool for five of the six hypervariable loops. The analysis, based on the hypothesis that antigen binding could occur through electrostatic interactions, reveals a diversity of possible binding mechanisms of anti-nucleosome or anti-histone antibodies to their cognate antigen. These results lead us to postulate that antinucleosome autoantibodies could have different origins. Since both anti-DNA and anti-nucleosome autoantibodies are produced during the course of systemic lupus erythematosus, a non-organ specific autoimmune disease, a comparative structural and electrostatic analysis of the two populations of autoantibodies may constitute a way to elucidate their origin and the role of the antigen in tolerance breakdown. The present study illustrates some interests, advantages and limits of a methodology based on the use of comparative modeling and analysis of molecular surface properties.

Structural properties and mutation patterns of anti-nucleosome monoclonal antibodies are similar to those of anti-DNA antibodies.

Four monoclonal antibodies (mAb) derived from an (NZB x NZW)F1 mouse bound to nucleosomes, total histones and to the H2A-H2B dimers but not to individual histones or DNA. Sequencing of their heavy (H)- and light (L)-chain variable region genes showed that they derived by somatic mutations from the same B cell precursor. The distribution of negatively and positively charged amino acids in the H-chain complementarity-determining regions was very similar to that observed not only in anti-H2A-H2B mAb derived from different lupus-prone mouse strains but also in anti-DNA mAb. Combined analysis of the mAb structures and their interactions with immobilized H2A-H2B dimer or total histones by plasmon resonance allowed us to assign the H-chain mutations a major role in the binding profiles of these anti-nucleosome mAb. Interestingly, four of the five H-chain mutations that distinguished mAb 3F6 from 2E1 generated negatively or positively charged amino acid residues, and two of them occurred at positions 56 and 76, which are frequently involved in the maturation process of anti-DNA antibodies. A modeling study of the 3F6 variable fragment (Fv) predicted that acidic residues occupy the cleft of the Ab combining site and have the potential to participate in electrostatic interactions. Thus, the demonstration that (NZB x NZW)F1-derived anti-H2A-H2B antibodies share certain structural features and mutation patterns with anti-DNA mAb suggest that common selection and maturation processes account for the production of these lupus-related autoantibodies.

IgM anti-myeloperoxidase antibody-secreting lymphocytes are present in the peripheral repertoire of lupus mice but rarely differentiate into IgG-producing cells.

Two IgM, kappa anti-myeloperoxidase (MPO) monoclonal antibodies, 6D6 and 9B5, bound to MPO in a solid-phase enzyme-linked immunosorbent assay were derived from the splenocytes of (NZB x NZW) F1 and MRL/lpr-lpr mice, respectively. 6D6 gave a characteristic perinuclear immunofluorescence staining pattern on ethanol-fixed human neutrophils, bound to the native form of MPO by immunoblotting and had a high constant affinity for MPO as demonstrated by real-time specific interaction. 9B5 produced a cytoplasmic immunofluorescence staining pattern, reacted with the heavy chain of MPO and had a low constant affinity for MPO. The heavy-and light-chain variable region genes of monoclonal antibodies (mAb) 6D6 and 9B5 were sequenced and found to be highly homologous to germline genes and to contain negatively charged amino acids in the complementarity determining regions. IgM MPO-binding activity was observed in most BW and MRL/lpr-lpr mouse sera, which may correspond to polyclonal activation of B cells, whereas IgG anti-MPO antibodies could be rarely detected. Thus, this study indicates that (i) BW and MRL/lpr-lpr mice do not delete IgM anti-MPO secreting B cells, do not maintain these B cells in a state of anergy, but most individuals are not able to spontaneously induce the class-switching of this autoantibody population; (ii) IgM anti-MPO antibodies can recognize different epitopes on MPO and produce different immunofluorescence staining pattern on ethanol-fixed human neutrophils, as demonstrated by the immunochemical properties of the two lupus-mouse derived mAb.

A complementarity-determining region peptide of an anti-desmosome autoantibody may interact with the desmosomal plaque through molecular mimicry with a cytoplasmic desmoglein 1 sequence.

The sera of patients with pemphigus, a group of autoimmune blistering skin diseases, contain autoantibodies directed against components of adhering junctions termed desmosomes. F12, a human monoclonal antibody derived from a pemphigus patient, recognizes an unknown polypeptide of the desmosomal and hemidesmosomal plaques. The third complementarity-determining region of the F12 heavy chain (VH-CDR3) was shown to share a four-amino-acid sequence (GSSG) with the intracellular domains of desmoglein 1 and bullous pemphigoid antigen 2 which interact with components of, respectively, the desmosomal and hemidesmosomal plaques. Computer modeling of F12 showed that the GSSG sequence protudes inside the antigen-combining site and thus might be involved in antigen interactions. The GSSG sequence is essential to F12 function, since a peptide containing the VH-CDR3 inhibited its binding to target antigens while VH-CDR3 peptides with specific modifications of the GSSG sequence did not. These data allow us to hypothesize that certain autoantibodies produced during the course of an autoimmune disease can behave as adhesion molecules, through the molecular mimicry of the motif involved in protein/protein adhesion, and to propose a new self-antigen binding mechanism for some autoantibodies.