The NZB X SWR model of lupus nephritis. I. Cross-reactive idiotypes of monoclonal anti-DNA antibodies in relation to antigenic specificity, charge, and allotype. Identification of interconnected idiotype families inherited from the normal SWR and the autoimmune NZB parents

The incidence of lupus nephritis is low in autoimmune NZB mice, but when they are crossed with normal SWR mice, almost 100% of the female F1 hybrids (SNF1) develop lethal glomerulonephritis. In a previous study we showed that anti-DNA auto-antibodies produced by the SNF1 mice were qualitatively different from those made by the NZB parents with respect to their isotype, charge, and antigenic specificity patterns. Here we studied idiotypic cross-reactions among the 65 monoclonal anti-DNA antibodies that were derived from four NZB and seven SNF1 mice. A library of 15 anti-idiotypic antibodies were prepared by immunizing rabbits with 15 monoclonal anti-DNA antibodies selected from the panel of 65. We identified 10 cross-reactive idiotype (CRI) families among this large collection of autoantibodies. Five of these CRI families were restricted to cationic anti-DNA antibodies that were exclusively of SNF1 origin, and the strongly cross-reacting members were predominantly IgG2b auto-antibodies with the allotype of the normal SWR parent. The cationic anti-DNA CRI families could be grouped into an interrelated cluster called the Id564 cluster. The other five anti-DNA CRI families were not restricted to any particular parental allotype or charge, although two of these CRI were shared exclusively by SNF1-derived autoantibodies and four of these CRI families could also be grouped into an idiotypically interrelated cluster called the Id512 cluster. In the case of seven out of the 10 CRI families, the idiotypic determinants detected were close to the antigen-binding site of the anti-DNA antibodies. The results indicate that the idiotypic repertoire of anti-DNA autoantibodies produced by the SNF1 mice is different from the NZB parents, and potentially pathogenic (cationic) antibodies produced by the SNF1 mice that are encoded by genes from the normal SWR parent can be identified as distinct CRI families. In the accompanying paper we demonstrate the role of these anti-DNA CRI families in the development of lupus nephritis.     

CD4+ T cell lines with selective patterns of autoreactivity as well as CD4- CD8- T helper cell lines augment the production of idiotypes shared by pathogenic anti-DNA autoantibodies in the NZB x SWR model of lupus nephritis

The (NZB x SWR)F1 hybrid mice (SNF1) uniformly develop lethal glomerulonephritis in marked contrast to their parents and produce nephritogenic autoantibodies that consist of highly cationic, IgG anti-DNA antibodies that share distinct cross-reactive idiotypes called IdLNF1 (idiotypes-lupus nephritis-SNF1). Herein we found that spleen cells of SNF1 mice at the late prenephritic stage, contained CD4+/CD8- and CD4-/CD8- Th that not only induced their B cells in vitro to produce highly cationic IgG autoantibodies to DNA but IdLNF1-positive IgG antibodies as well. The double-negative Th were unexpected in the SNF1 mice because they lack the lpr (lymphoproliferation) gene. We also derived IL-2-dependent CD4+/CD8- as well as CD4-/CD8- T cell lines from nephritic SNF1 mice, that could simultaneously induce IdLNF1-positive and cationic anti-DNA antibodies of IgG class. The CD4+ T cell lines consisted of "autoreactive" T cells, but not all of the lines were equal in autoantibody-inducing capability. Remarkably, the T cell lines that preferentially responded to F1-hybrid-MHC determinants, had a significantly greater ability to augment the production of pathogenic autoantibodies. The SNF1-Th could also augment autoantibody production by the NZB or SWR parent's B cells; however, IdLNF1-positive and cationic anti-DNA autoantibodies of IgG class were not induced, suggesting that the SNF1 mice possess a select population of inducible (susceptible) B cells that are committed to produce nephritogenic autoantibodies and the parental strains are deficient in such B cells. Thus, production of nephritogenic autoantibodies with IdLNF1 markers in the SNF1 mice could result from an interaction between a select population of B cells and CD4+ Th that preferentially recognize unique F1-hybrid-MHC determinants, as well as double-negative auxiliary Th.     

Lupus prone (SWR x NZB)F1 mice produce potentially nephritogenic autoantibodies inherited from the normal SWR parent

The incidence of nephritis in autoimmune NZB mice is low, but when they are crossed with normal SWR mice, almost 100% of the female F1 hybrids (SNF1) develop lethal glomerulonephritis. To define the contribution of the normal SWR strain to the development of nephritis, we analyzed 65 monoclonal anti-DNA autoantibodies derived from SNF1 mice and compared them with those obtained from the NZB parent. The majority of the SNF1-derived anti-DNA antibodies were IgG and cationic in charge. By contrast, 77% of the NZB-derived antibodies were IgM. Moreover, all three NZB-derived IgG anti-DNA antibodies were anionic. The cationic property of the SNF1-derived IgG autoantibodies was not restricted to any particular antigenic specificity pattern or IgG subclass, nor was there a preference for the allotype of either parent. However, we identified a set of highly cationic (pI at 8.2 to 8.8 pH) IgG2b anti-DNA antibodies from SNF1 hybrids that had the SWR allotype. Isoelectric focusing of intact antibodies and isolated heavy and light chains showed that the highly cationic charge of these antibodies was determined by the variable regions of their heavy chains. Because IgG anti-DNA antibodies with cationic charge are especially pathogenic, those antibodies bearing the allotype of the normal SWR parent may account for the high incidence of severe nephritis in the F1 hybrids. The results indicate that pathogenic autoantibodies, which are encoded by genes of the nonautoimmune SWR parent, are expressed in the SNF1 mice due to some cellular and genetic regulatory influence of the NZB parent.     

V region sequences of an idiotypically connected family of pathogenic anti-DNA autoantibodies

The H and L chain V region sequences of nine anti-DNA mAb that are representative of a pathogenic population of autoantibodies produced by the nephritis prone (SWR x NZB)F1 (SNF1) mice, were determined. These nine anti-DNA autoantibodies were idiotypically connected members of a cross-reactive Id family called the Id564 cluster. Moreover, these autoantibodies were all cationic in charge, IgG2b in isotype, and their H chain C regions had the normal SWR parent's allotype. Although derived from two different SNF1 animals, these pathogenic autoantibodies possessed highly homologous Leader-VH sequences that could account for their idiotypic cross-reactivity. Furthermore, the VH region sequences of these anti-DNA antibodies contained numerous basic residues that could impart their cationic charge. The Leader-VH sequences of these autoantibodies were also highly homologous to that of an anti-NP antibody-related germ-line gene of C57BL/6 mice, called VH-23. Among these nine pathogenic autoantibodies, three sets of clonally related anti-DNA antibodies could be identified. Thus the Id564 cluster of cationic anti-DNA autoantibodies of SNF1 mice are encoded by highly related VH genes, and this idiotypically connected population of pathogenic autoantibodies are selected to undergo an oligoclonal expansion in the lupus-prone SNF1 mice.     

Antigen-specific therapy of murine lupus nephritis using nucleosomal peptides: tolerance spreading impairs pathogenic function of autoimmune T and B cells

In the (SWR x NZB)F1 mouse model of lupus, we previously localized the critical autoepitopes for nephritogenic autoantibody-inducing Th cells in the core histones of nucleosomes at aa positions 10-33 of H2B and 16-39 and 71-94 of H4. A brief therapy with the peptides administered i.v. to 3-mo-old prenephritic (SWR x NZB)F1 mice that were already producing pathogenic autoantibodies markedly delayed the onset of severe lupus nephritis. Strikingly, chronic therapy with the peptides injected into 18-mo-old (SWR x NZB)F1 mice with established glomerulonephritis prolonged survival and even halted the progression of renal disease. Remarkably, tolerization with any one of the nucleosomal peptides impaired autoimmune T cell help, inhibiting the production of multiple pathogenic autoantibodies. However, cytokine production or proliferative responses to the peptides were not grossly changed by the therapy. Moreover, suppressor T cells were not detected in the treated mice. Most interestingly, the best therapeutic effect was obtained with nucleosomal peptide H416-39, which had a tolerogenic effect not only on autoimmune Th cells, but autoimmune B cells as well, because this peptide contained both T and B cell autoepitopes. These studies show that the pathogenic T and B cells of lupus, despite intrinsic defects in activation thresholds, are still susceptible to autoantigen-specific tolerogens.     

Genetic contributions of nonautoimmune SWR mice toward lupus nephritis.

(SWR x New Zealand Black (NZB))F(1) (or SNF(1)) mice succumb to lupus nephritis. Although several NZB lupus susceptibility loci have been identified in other crosses, the potential genetic contributions of SWR to lupus remain unknown. To ascertain this, a panel of 86 NZB x F(1) backcross mice was immunophenotyped and genome scanned. Linkage analysis revealed four dominant SWR susceptibility loci (H2, Swrl-1, Swrl-2, and Swrl-3) and a recessive NZB locus, Nba1. Early mortality was most strongly linked to the H2 locus on chromosome (Chr) 17 (log likelihood of the odds (LOD) = 4.59 - 5.38). Susceptibility to glomerulonephritis was linked to H2 (Chr 17, LOD = 2.37 - 2.70), Swrl-2 (Chr 14, 36 cM, LOD = 2.48 - 2.71), and Nba1 (Chr 4, 75 cM, LOD = 2.15 - 2.23). IgG antinuclear autoantibody development was linked to H2 (Chr 17, LOD = 4.92 - 5.48), Swrl-1 (Chr 1, 86 cM, colocalizing with Sle1 and Nba2, LOD = 2.89 - 2.91), and Swrl-3 (Chr 18, 14 cM, LOD = 2.07 - 2.13). For each phenotype, epistatic interaction of two to three susceptibility loci was required to attain the high penetrance levels seen in the SNF(1) strain. Although the SWR contributions H2, Swrl-1, and Swrl-2 map to loci previously mapped in other strains, often linked to very similar phenotypes, Swrl-3 appears to be a novel locus. In conclusion, lupus in the SNF(1) strain is truly polygenic, with at least four dominant contributions from the SWR strain. The immunological functions and molecular identities of these loci await elucidation.     

Failed self-tolerance and autoimmunity in IgG anti-DNA transgenic mice.

Transgenic mice were generated that express both the H and L chain genes derived from a hybridoma secreting an IgG2a mAb specific for ds- and ssDNA. This hybridoma is derived from a lupus mouse and can accelerate nephritis in young NZB x NZW F1 female mice and induce clinical nephritis in BALB/c mice. Some transgenic B cells did not exhibit allelic exclusion; they expressed both transgene-derived IgG and endogenous IgM intracellularly. Most of the B cells in transgenic mice expressed endogenous IgM, some of them expressed low levels of IgG on cell membranes. The transgenic mice, created in a strain not prone to SLE, expressed elevated serum IgG anti-DNA, and some developed clinical nephritis. The affinity of the spontaneously secreted IgG antibodies for dsDNA were similar in nephritic NZB x NZW F1 and transgenic mice. In contrast to the nontransgenic littermates, immunization of transgenic mice with murine DNA further enhanced serum levels of IgG anti-DNA in transgenic mice. Therefore, expression of transgene-encoded IgG anti-DNA mainly in the secreted form does not provide the signals necessary for allelic exclusion or self-tolerance. Expression of this Ig is sufficient to induce a mild form of autoimmune disease.     

Suppression of murine lupus nephritis by administration of an anti-idiotypic antibody to anti-DNA

The suppression of pathogenic antibodies to DNA in NZB/NZW f1 female mice was achieved by repeated inoculation of the mice with a monoclonal anti-idiotypic antibody (anti-Id). The anti-Id, an IgG1, kappa, was directed against a major cross-reactive idiotype (Id) on NZB/NZW IgG antibodies to DNA. One hundred micrograms of the anti-Id were inoculated i.p. every 2 wk, beginning at 6 wk of age (nondiseased mice--no circulating anti-DNA or proteinuria) or 20 wk of age (diseased mice--all with circulating anti-DNA, one-third with proteinuria). As controls, littermates received an IgG, kappa non-DNA-binding myeloma or no treatment. In the young mice, nephritis and anti-DNA antibodies appeared at the same time in all groups, and their circulating antibodies to DNA did not bear the target Id. In the older (20-wk-old) mice, survival was significantly prolonged because of delay in the onset of nephritis; the total quantities of antibodies to DNA were diminished, and the target Id, initially present on circulating IgG, was deleted. These benefits were transient; the suppression of antibodies was followed by the appearance of large quantities of anti-DNA that did not bear the major Id. Therefore, although administration of anti-Id was effective in reducing an undesirable antibody response after the target Id was present on circulating antibodies, the benefits were limited, probably by Id "switch" or by increased synthesis of pathogenic antibodies bearing a minor Id.     

Anti-RNA polymerase I antibodies: potential role in the induction and progression of murine lupus nephritis

Antibodies against RNA polymerase I were detected in plasma and kidney eluates of NZB/W mice. Plasma concentrations of the antibodies were the highest in mice with incipient nephritis and the lowest in mice with progressive nephritis. Mice with attenuated nephritis due to immunosuppressive therapy had intermediate plasma concentrations of the antibodies. The specific concentrations (ng/microgram IgG) of anti-RNA polymerase I antibodies in kidney eluates were significantly (10- to 70-fold) greater than the corresponding plasma concentrations. These results indicated that the decreased plasma concentration of the antibodies in mice with more advanced disease was at least partially due to selective concentration of anti-RNA polymerase I antibodies in the kidneys. The degree of this selective concentration was directly proportional (R2 = 0.9962) to the severity of renal disease, as reflected by the concentration (microgram/g tissue) of IgG eluted from the kidneys. The concentration (microgram/g tissue) of anti-RNA polymerase I eluted from the kidneys also was increased in mice with more severe renal disease. Further, the extent of this increase was greater than that of total IgG, again suggesting that anti-RNA polymerase I antibodies had been selectively concentrated in the kidneys. These findings are strongly suggestive of an important role for the RNA polymerase I/anti-RNA polymerase I antibody system in the pathogenesis of murine lupus nephritis.     

Clonotypes of anti-DNA antibodies expressing specific idiotypes in immune complexes of patients with active lupus nephritis

IEF, using 6 M urea, provides a unique opportunity to analyze the spectrotypes of antibodies in immune complexes (IC) in vivo. Using this technique, we have analyzed the clonotypes of anti-DNA antibodies expressing specific Id in the circulating IC of patients with active lupus nephritis. Serum anti-ssDNA and anti-dsDNA antibodies showed heterogeneous spectrotypes. The antibodies isolated from circulating IC had a restricted clonotype and a neutral charge and were directed mainly to ssDNA and, to a lesser extent, to dsDNA. These samples failed to form complexes with DNA when they were subjected to absorption to a DNA-coupled Sepharose column. Anti-DNA antibodies expressed specific Id, termed O-81 or NE-1, which were detected only in the IC of patients with active lupus nephritis. Anti-DNA clonotypes, including O-81 and NE-1 idiotypes, were also found in the eluates of renal glomeruli of lupus patients. These results indicate that subpopulations of anti-DNA antibodies in circulating IC are limited, and may play an important role in the pathogenesis of lupus nephritis.     

Mechanisms of T and B cell collaboration in the in vitro production of anti-DNA antibodies in the NZB/NZW F1 murine SLE model

B/W mice spontaneously develop IgG antibodies to DNA that cause lethal immune nephritis. T and B cell interactions in the in vitro anti-DNA antibody response of B/W mice were investigated, and two distinct families of helper T cells that drive these responses were defined. First, the anti-DNA antibody-forming cell (AFC) response was found to be increased in B/W mice with nephritis and was inhibited with the monoclonal antibody anti-L3T4, suggesting a major role for helper T cells. Purified splenic T cells from mice with nephritis were able to augment both the IgG and the IgM anti-DNA AFC response of young B/W B cells. T helper cells were cloned from spleens of NZB/W F female mice with high titer anti-DNA antibodies and nephritis. The cloned T cells augmented both IgG and IgM anti-DNA AFC responses of young B/W B cells. Four clones--27.9, 30.7, 30.8, and 30.10--were selected for further study. These cells proliferated, in the context of syngeneic (H2d/z) antigen-presenting cells (APC) but not to allogeneic APC. Analysis of the mechanism of T helper cell clone-mediated augmentation of anti-DNA AFC revealed two populations: "cognate" T helper cells, which specifically augment anti-DNA AFC (30.7 and 30.10), and non-antigen-specific T helper cells (27.9 and 30.8), which augment the response of B cells of differing specificity by a bystander mechanism, probably through increased release of B cell growth and differentiation factors.     

A public idiotypic determinant is present on spontaneous cationic IgG antibodies to DNA from mice of unrelated lupus-prone strains

A monoclonal anti-idiotypic antibody (anti-Id) to a public idiotype (Id) present on spontaneous IgG antibodies to DNA from NZB/NZW F1 mice recognized similar determinants on polyclonal and monoclonal IgG anti-DNA antibodies from mice of the unrelated MRL/lpr and BXSB strains. Incubation of the anti-Id with four of five monoclonal Id in solid phase inhibited their ability to bind DNA; however, different Id+ antibodies recognized different epitopes within the DNA molecule. Therefore, the public Id was located close to the antigen-binding regions but did not comprise all of those regions. Analysis of multiple polyclonal and monoclonal antibodies to DNA showed the Id on all subclasses of IgG. However, antibodies bearing the Id carried a neutral or cationic charge (10 of 10 monoclonals with pI greater than 7 were Id+); the presence of the Id on anionic IgG (pI less than or equal to 7) was infrequent (one of 21 serums, one of eight monoclonal antibodies). Therefore, IgG autoantibodies to DNA are constructed from closely related public idiotypes in several mouse strains that spontaneously develop lupus, and that Id is restricted to antibodies with a pI of 7 or greater.     

Idiotypic analysis of antibodies to hen egg-white lysozyme (HEL). III. Further studies on the idiotypic cross-reactivity among anti-HEL antibodies in mice

An isolated antibody preparation directed to the native hen egg-white lysozyme (HEL) from a single A/J mouse (#a-11), termed IDHELa-11, was inoculated into rabbits to produce anti-idiotypic sera (anti-Id). The antisera were extensively absorbed with normal A/J Ig to render then idiotype specific. Radioimmunoassay utilizing 125I-IdHELa-11 and the anti-ID sera (R103 and R104) was performed to examine the idiotypic cross-reactivity of the humoral immune response to HEL in various mouse strains and other animal species. Idiotypes shared by IdHELa-11 were detected in the sera of five mouse strains tested, but not in any of the examined sera of other animal species such as rats, goats, guinea pigs, or sheep, indicating the occurrence of species-specific cross-reactive idiotypes (CRI) of antibodies to HEL. Our experiments also suggested the presence of intrastrain CRI. The present experiments, along with our earlier results (15), suggest that idiotypic cross-reactivity among murine antibodies to HEL appears to be weak. Thus when R103 and R104 were the anti-Id sera used, the frequency of occurrence of CRI shared by IdHELa-11 in 5 micrograms of anti-HEL antibody in strain A mice was 74 ng and 111 ng; in other strains it was 25-44 ng and 60-98 ng, respectively.     

Allotype-linked production of an idiotype-specific factor that promotes idiotype expression in nonresponder strains

We had previously demonstrated that expression of the cross-reactive idiotypes (CRI) in the phenyltrimethylammonium (TMA) system depends on the presence of a second-order T helper (Th2) cell. Furthermore, we showed that this cell type can be replaced by an idiotype-specific helper factor derived from either a 24-hr concanavalin A supernatant (Con A) or the T cell hybridoma LOP 1.4. This factor, regardless of its source, is idiotype-specific, I-J+, and promotes in vitro expression of the cross-reactive phenyltrimethylammonium idiotype (CRI+-TMA) found on anti-trinitrophenyl antibodies. Because the expression of this idiotype in antigen-primed immune sera is linked to the Ig-1e heavy chain locus, experiments were conducted to test whether the production of this factor was also linked to the same locus. Of the strains tested, only splenocytes derived from the Ig-1e mice, irrespective of their background genetics, produced the factor upon Con A stimulation. Furthermore, the function of the factor is not major histocompatibility complex (MHC)-restricted because Con A supernatants derived from the C57.Ige (H-2b, Ig-1e), NZB (H-2d, Ig-1e), and A.SW (H-2s, Ig-1e) strains promoted CRI+ trinitrophenyl plaque-forming cells in A/J (H-2a, Ig-1e) cultures. Further experiments were carried out to determine if the idiotype-specific factor could promote CRI+ TNP plaque-forming cells in non-Ig-1e strains. To this end, A/J Con A and LOP 1.4-derived supernatants were added to primed C57Bl/6 (H-2b, Ig-1b) and DBA/2 (H-2d, Ig-1c) splenic cultures, both of which do not express serum CRI-TMA or produce the idiotype-enhancing factor. The cultures from either strain in the presence of the factor produced CRI+-TMA trinitrophenyl plaque-forming cells of comparable numbers to the A/J prototype strain. The results suggest an important regulatory role for this factor in allotype-linked expression of dominant idiotypes.     

Accelerated macrophage apoptosis induces autoantibody formation and organ damage in systemic lupus erythematosus

Increased monocyte/macrophage (Mphi) apoptosis occurs in patients with systemic lupus erythematosus (SLE) and is mediated, at least in part, by an autoreactive CD4(+) T cell subset. Furthermore, autoreactive murine CD4(+) T cells that kill syngeneic Mphi in vitro induce a lupus-like disease in vivo. However, it is unclear whether increased Mphi apoptosis in SLE per se is sufficient to accelerate/promote autoimmunity. We have investigated whether increased Mphi apoptosis in vivo, induced by the administration of clodronate liposomes, can exacerbate the autoimmune phenotype in NZB x SWR (SNF(1)) lupus-prone mice, and induce autoantibody production in haplotype-matched BALB/c x DBA1 (DBF(1)) non-lupus-prone mice. Lupus-prone mice SNF(1) mice that were treated with clodronate liposomes, but not mice treated with vehicle, developed significant increases in autoantibodies to dsDNA, nucleosomes, and the idiotypically related family of nephritic Abs Id(LN)F(1), when compared with untreated SNF(1) mice. Furthermore, clodronate treatment hastened the onset of proteinuria and worsened SNF(1) lupus nephritis. When compared with vehicle-treated controls, clodronate-treated non-lupus-prone DBF(1) mice developed significantly higher levels of anti-nucleosome and Id(LN)F(1) Abs but did not develop lupus nephritis. We propose that Mphi apoptosis contributes to the pathogenesis of autoantibody formation and organ damage through both an increase in the apoptotic load and impairment in the clearance of apoptotic material. This study suggests that mechanisms that induce scavenger cell apoptosis, such as death induced by autoreactive cytotoxic T cells observed in SLE, could play a pathogenic role and contribute to the severity of the disease.     

The role of anti-α-actinin antibodies in the pathogenesis and monitoring of lupus nephritis

Antibodies to double-stranded DNA are important in the pathogenesis of nephritis, a major clinical manifestation in lupus patients. Since earlier diagnosis of renal involvement may lead to better outcomes, identification of the nephritogenic specificity of lupus-associated autoantibodies is important in understanding the disease, while monitoring their titer clinically may serve as an improved biomarker. Based upon work in animal models and cross-sectional human studies, kidney α-actinin was thought to be a plausible cross-reactive target for pathogenic lupus antibodies. Manson and colleagues longitudinally evaluated anti-nucleosome, anti-DNA, and anti-α-actinin antibodies in 16 lupus patients with new-onset nephritis. While anti-nucleosome and anti-DNA antibody levels were significantly associated and correlated with measures of kidney disease, these were not found to be significant with anti-α-actinin antibodies. While in lupus patients the diagnostic use of serum anti-α-actinin antibodies, alone or with other novel biomarkers, is still under investigation, such studies are vital in improving our monitoring of systemic lupus erythematosus patients and in developing new treatment paradigms that meet the continuing clinical challenge of lupus nephritis.     

IgG anti-DNA autoantibodies within an individual autoimmune mouse are the products of clonal selection

Although mice from almost all inbred strains produce IgM anti-DNA antibody in response to B cell mitogens, only (NZB x NZW)F1 mice and mice from other strains that are genetically predisposed to autoimmunity spontaneously produce anti-DNA antibody of the IgG isotype. Because (NZB x NZW)F1 mice display marked B cell hyperactivity, anti-DNA antibody production in these mice has been thought to result from spontaneous, polyclonal B cell activation. Although this may be true for IgM anti-DNA antibodies, our results demonstrate that IgG anti-DNA antibodies are not polyclonal. Rather, IgG anti-DNA autoantibodies within an individual autoimmune mouse are oligoclonal and somatically mutated. These results demonstrate that IgG anti-DNA autoantibodies are the products of clonally selective B cell stimulation and exhibit the same characteristics as secondary immune antibodies to conventional immunogens: they are IgG, they are clonally restricted, and they are somatically mutated.