Structural basis for autoantibody recognition of phosphatidylserine-beta 2 glycoprotein I and apoptotic cells.

Apoptotic cells contain nuclear autoantigens that may initiate a systemic autoimmune response. To explore the mechanism of antibody binding to apoptotic cells, 3H9, a murine autoantibody with dual specificity for phospholipids and DNA, was used. H chain mutants of 3H9 were constructed, expressed as single-chain Fv (scFv) in Escherichia coli, and assessed for binding to phosphatidylserine, an antigen expressed on apoptotic cells. Both 3H9 and its germline revertant bound to dioleoyl phosphatidylserine in ELISA, and binding was enhanced by beta 2 glycoprotein I (beta 2GPI), a plasma protein that selectively binds to apoptotic cells. Higher relative affinity for DOPS-beta 2GPI was achieved by the introduction of Arg residues into the 3H9 H chain variable region at positions previously shown to mediate DNA binding. Specificity of the two structurally most diverse scFv for apoptotic cells was shown by flow cytometry, and two populations of scFv-bound cells were identified by differences in propidium iodide staining. The results suggest that, in autoimmunity, B cells with Ig receptors for apoptotic cells and DNA are positively selected, and that the antibodies they produce have the potential to affect the clearance and processing of apoptotic cells.

Genes coding evolutionary novel anti-carbohydrate antibodies: studies on anti-Gal production in alpha 1,3galactosyltransferase knock out mice.

This study analyzes the gene repertoire coding for antibodies to an evolutionary novel immunogenic carbohydrate antigen in mice. The alpha-gal epitope (Gal alpha 1-3Gal beta 1-4GlcNAc-R) is an autoantigen, abundantly expressed in wild type mice, but absent in alpha 1,3galactosyltransferase knock-out (KO) mice, where it can induce the production of the anti-Gal antibody. Hybridoma clones secreting anti-Gal were isolated from different mice and their immunoglobulin genes were analyzed. All anti-Gal clones were found to be encoded by the heavy chain gene VH22.1 and light chain gene VK5.1. Moreover, one 'forbidden' anti-Gal clone, produced in a wild type mouse, was also encoded by VH 22.1 and VK 5.1. The genes coding for the different anti-Gal clones were found to contain somatic mutations and different CDR3 domains. These data imply that a highly restricted gene usage combined with junctional diversity and somatic mutations can generate new antibodies that have not been produced in the course of the evolution of a species.

Diverse roles for the third complementarity determining region of the heavy chain (H3) in the binding of immunoglobulin Fv fragments to DNA, nucleosomes and cardiolipin.

Autoantibodies to DNA and chromatin employ junctional diversity and somatic mutations to generate or enhance antigen recognition. To define the role of diversity generating mechanisms in the etiology of autoantibodies to nuclear antigens, the heavy (H) chain of a murine autoantibody, 3H9, was used in its somatically mutated or germ-line form in conjunction with its own or with heterologous CDR3 (H3) domains. The resulting H chains were expressed together with the 3H9 light (L) chain as single-chain Fv (scFv) in Escherichia coli and assayed for binding to DNA, nucleosomes, or cardiolipin by enzyme-linked immunosorbent assay. All recombinant scFv exhibited nearly identical binding to cardiolipin. In contrast, the binding to nuclear antigens was drastically reduced by the reversion of mutations in 3H9 or the exchange of H3, such that only 3H9 itself bound strongly to single-stranded DNA, double-stranded DNA and nucleosomes. The results illustrate diverse interactions between a single combining site and different autoantigens. The analysis of these interactions suggests that the 3H9 VH domain, as encoded by the germ line, directs binding to cardiolipin, whereas structural determinants of H3, in concert with the remainder of the combining site, guide the maturation of antibody binding toward nuclear autoantigens.

Analysis of autoimmune bone marrow by antibody-phage display: somatic mutations and third complementarity-determining region arginines in anti-DNA gamma and kappa V genes.

OBJECTIVE: To examine anti-double-stranded DNA (anti-dsDNA) IgG autoantibodies from the bone marrow of individuals with systemic lupus erythematosus (SLE). METHODS: A library of single-chain variable fragments (scFv) was constructed from SLE bone marrow complementary DNA of gamma, kappa, and lambda isotype by cloning into the pHENIX phagemid vector. The library was screened with dsDNA in solution, and 2 anti-DNA phage, DNA1 and DNA4, were isolated and their Ig V genes sequenced. Soluble scFv corresponding to DNA1 and DNA4, and their heavy (H)- and light (L)-chain recombinants, were prepared, purified, and analyzed for binding to DNA by enzyme-linked immunosorbent assay. RESULTS: DNA1 and DNA4 used different Ig H-chain (3-30 and 5-51, respectively) and L-chain (DPK15 and DPK22, respectively) V genes. The ratios of replacement mutations to silent mutations in DNA1 and DNA4 suggest that their V genes were selected for improved antigen binding in vivo. The recombinant between DNA4VH and DNA1VL showed the highest relative affinity for both single-stranded DNA and dsDNA. These 2 Ig subunits contained third complementarity-determining region arginines and had acquired the majority of replacement mutations. CONCLUSION: Anti-dsDNA IgG autoantibodies from the bone marrow of SLE patients exploit diverse V genes and cationic V-D-J and V-J junctions for DNA binding, and accumulate replacement mutations that enhance binding.

Tandem affinity tags for the purification of bivalent anti-DNA single-chain Fv expressed in Escherichia coli.

Antibodies to DNA define an important autospecificity that arises in systemic lupus erythematosus (SLE). To elucidate the molecular features that may explain the pathogenesis of SLE, a heterologous system for expression of cloned V genes is often desirable. Here, a single-chain Fv coding domain was constructed by using the heavy- and light-chain V genes of a high-affinity site-directed mutant of the murine anti-dsDNA autoantibody, 3H9. This scFv was joined in frame to the c-jun leucine zipper for dimerization, and to two affinity tags, domain B of the staphylococcal protein A and a pentahistidine peptide, for purification. Dimerization of the scFv was determined by size-exclusion chromatography. The yields of the scFv following affinity purification on IgG agarose or Ni-NTA agarose were compared, and the activities of the resulting protein fractions were determined. A two-step purification of periplasmic extracts on Ni-NTA agarose and IgG agarose, followed by elution with 3.5 M MgCl(2), yielded scFv with the highest specific activity. The final purified material bound DNA by ELISA, electrophoretic mobility shift assay, and immunofluorescence of fixed Hep-2 cells. Antibodies purified in this fashion should have applications in structure/function studies in which it is essential to generate highly purified antigen-combining sites.

Initiation of systemic autoimmunity and sequence specific anti-DNA autoantibodies.

Antibodies to double-stranded DNA (dsDNA) are a defining feature of Systemic Lupus Erythematosus (SLE). The molecular characterization of anti-dsDNA autoantibodies reveals that they are actively selected for binding to antigen. Evidence for antigen selection includes the use of suitable rearrangement products, the switching of IgM isotype to IgG, and the acquisition of somatic mutations that raise the affinity for dsDNA. Through a process of specificity maturation, anti-dsDNA antibodies can arise from anti-single stranded DNA (ssDNA) antibodies that also occur in nonautoimmune individuals. To clarify circumstances leading to the initiation of systemic autoimmunity, we compare features of immune responses to nucleic acids that operate before and after disease develops. Evidence indicating that anti-dsDNA antibodies bind with DNA sequence preference is highlighted to propose that sequence-specific anti-dsDNA antibodies may be induced by an infectious agent and in turn may extend the response to endogenous nuclear antigens. Thus, sequence-specific anti-dsDNA B cells may provide an important stimulus to break the tolerance to self.

A sensitive assay for measuring alpha-Gal epitope expression on cells by a monoclonal anti-Gal antibody.

BACKGROUND: The assessment of a-gal epitope (Galalpha1-3Galbeta1-4GlcNAc-R) expression on various cells and tissues is important for the prediction of anti-Gal-mediated immune rejection of xenografts. This study describes an enzyme-linked immunosorbent assay (ELISA inhibition assay) developed for this purpose, which uses the monoclonal anti-Gal antibody M86. METHODS: Cells at various concentrations were incubated overnight with M86 at 1/100 dilution. The cells and bound antibody were removed, and the residual antibody in the supernatant was measured in an ELISA assay with a-gal-bovine serum albumin as a solid phase antigen. The extent of a-gal epitope expression on cells correlates with the subsequent inhibition of M86 binding in ELISA. The inhibition binding curves at various cell concentrations were compared with those of a standard cell line with a known number of epitopes per cell. RESULTS AND CONCLUSIONS: The mouse IgM M86 monoclonal antibody was highly specific for a-gal epitopes. Using this antibody in an ELISA inhibition assay with cells at a wide range of concentrations enables the detection of at least 5 x 10(4) and up to more than 5 x 10(7) a-gal epitopes per cell. This assay can be used also for the detection of a-gal epitopes on membranes from tissue homogenates, and thus it enables the determination of the extent of decrease in a-gal epitope expression in animals that are genetically manipulated to alter their carbohydrate make-up.

Cloning of anti-Gal Fabs from combinatorial phage display libraries: structural analysis and comparison of Fab expression in pComb3H and pComb8 phage.

Anti-Gal is the most abundant natural antibody in humans. It interacts specifically with the carbohydrate epitope Gal alpha1-3Galbeta1-4GlcNAc-R (termed the alpha-galactosyl epitope). In an attempt to characterize the Ig genes encoding anti-Gal, two combinatorial phage display libraries in phagemid pComb3H were screened for anti-Gal Fabs. For this purpose, phages were incubated with biotinylated BSA coupled with alpha-galactosyl epitopes (designated alpha-Gal-BSA). Subsequently, phages complexed with alpha-Gal-BSA were isolated by streptavidin-coupled magnetic beads. Because of the low affinity of this antibody, a characteristic shared with other anti-carbohydrate antibodies, only two clones displaying anti-Gal activity were isolated. Clone G9 contained the VH gene V3-43 and VL gene DPK15, whereas clone P19 contained the VH gene V3-15 and VL gene DPL16. Both clones contained between five and 14 mutations in their H and L chain V genes. The affinity of clone G9 was found to be higher than that of clone P19, as only the former could bind to solid-phase alpha-galactosyl epitopes in an enzyme-linked immunosorbent assay. This interaction could be increased by expressing Fabs in phagemid pComb8 grown in the presence of IPTG. Under such conditions, the IPTG-activated Lac-Z promoter induces an increased expression of Fabs that are linked to phage envelope protein VIII, resulting in multiple Fab display on the phage. The data suggest that screening combinatorial phage display libraries for anti-carbohydrate antibodies may be more effective with pComb8 phage grown in the presence of IPTG.

Selection of recurrent V genes and somatic mutations in autoantibodies to DNA.

Antigen selection of autoantibodies to DNA results in the use of limited sets of immunoglobulin heavy and light chains, characteristic VDJ and VJ joining regions, and recurrent patterns of somatic mutations. In the past, we have used site-directed mutagenesis to examine the roles of two recurrent features of anti-DNA antibodies: VHCDR3 arginine and somatic mutations to arginine. We observed that in one prototypic anti-DNA antibody, 3H9, a suitable CDR3 conformation is essential for DNA binding and depends on arginine. In addition, arginines at any of five positions in CDR1, CDR2, or FWR3 of the heavy chain contribute contacts with the antigen. Here, we extend these studies and report that arginines at positions 52 and 58 improve relative DNA binding but that binding critically depends on the germline-encoded arginine at position 50. These observations provide a more detailed view of the anti-DNA combining site and suggest that structural features account, at least in part, for the recurrence of heavy chain variable regions in anti-DNA antibodies.

Monoclonal antibodies from NZW x BXSB F1 mice to beta2 glycoprotein I and cardiolipin. Species specificity and charge-dependent binding.

NZW x BXSB F1 mice develop a systemic autoimmune syndrome with various lupus-like manifestations. Male animals develop a degenerative coronary disease with myocardial infarction, resulting in death before 6 mo of age. The presence in these mice of anti-phospholipid Abs reacting with beta2-glycoprotein I may contribute to the pathogenesis of the cardiovascular lesions. beta2-glycoprotein I, a plasma protein implicated in various aspects of the coagulation pathway, is also the target of autoantibodies in humans with the anti-phospholipid syndrome. We obtained several mAbs from NZW x BXSB F1 mice that were selected for binding to cardiolipin. Two mAbs are specific for beta2-glycoprotein I and display a species-dependent pattern with preferential reactivity to mouse beta2-glycoprotein I. The other mAbs display charge-mediated interactions with anionic phospholipids in the absence of beta2-glycoprotein I. The analysis of the V region sequences of the mAbs suggests that cationic residues in the H chain complementarity-determining region 3 are important for their phospholipid reactivity. The structural features of the V(H)-D-J(H) junctions of these mAbs further support the view that an increased frequency of unusual V(D)J rearrangements directly contributes to the development of murine autoimmunity.

Constitutive secretion of transgene-encoded IgG2b autoantibodies leads to symptoms of autoimmune disease.

Anti-DNA Ab are strictly regulated, except in autoimmunity, where they are expressed and may contribute to pathogenicity. To study constitutive anti-DNA Ab secretion in nonautoimmune mice, two anti-dsDNA H/L chain transgene combinations were constructed using an IgG2b C region with secretory but no transmembrane domain exons. One H/L combination, consisting of the VH3H9 H and V kappa 4 L chain transgenes, was chosen to recreate 3H9, an autoantibody that originally arose in an autoimmune MRL/lpr mouse; the other paired a higher affinity variant of VH3H9, 56R, with the same V kappa 4 L chain. Elevated titers of IgG2b along with normal levels of other isotypes were observed in transgene-positive mice, indicating that constitutive transgene-directed Ab secretion was achieved. Sera and hybridoma supernatants from VH3H9 gamma transgene-positive animals exhibited binding to dsDNA, ssDNA, and cardiolipin. Mice expressing the 56R gamma H chain and the V kappa 4 L chain showed enhanced binding. Expression of the transgenes correlated with signs of autoimmune disease, including prolonged plasma clotting in vitro, and reduced litter size. The results suggest that even a single autoreactive H chain that escapes tolerance may suffice to induce features of autoimmune disease.

Light chain contribution to specificity in anti-DNA antibodies.

We studied mice expressing one of two H chain transgenes. Both transgenes expressed the same 3H9 anti-DNA VDJ, but differed in their constant domains. The IgM transgene efficiently induced tolerance and selected for a subset of endogenous L chains that prevented dsDNA binding. In contrast, the IgG2b secreted-only H chain allowed expression of a broad range of L chains, most of which yielded anti-dsDNA Ab. To deduce the features of L chains that affect DNA binding, we derived hybridomas from LPS-stimulated splenic B cells from the two transgene lines and compared the V kappa sequences of Ab they secreted. Identification of L chains with related sequences but different binding to ssDNA, dsDNA, and cardiolipin allowed us to pinpoint L chain residues that correlate with enhanced or reduced binding. Arginines at the junction of V kappa 1 or V kappa 8 regions and J kappa 1, and arginines or asparagines in CDR1 or CDR2 enhanced DNA binding. Negatively charged residues at the same positions were found to interfere with binding. Thus, we predict that appropriate amino acids at these positions may form contacts with DNA. The likely locations of contact residues in the combining site were evaluated by inspection of previously determined Ab structures. Our results indicate that L chains in anti-DNA Ab are able to modulate DNA binding and contribute contact sites for additional determinants on a complex autoantigen.

Human anti-Gal heavy chain genes. Preferential use of VH3 and the presence of somatic mutations.

Anti-Gal is the most abundant natural Ab known in humans. It interacts specifically with the carbohydrate structure Gal alpha 1-3 Gal beta 1-4GlcNAc-R (termed the alpha-galactosyl epitope), constitutes approximately 1% of circulating Ig, and is found in all three isotypes in the serum. Anti-Gal is produced in Old World monkeys, apes, and humans, and in no other mammalian species. The objective of this study was to determine the VH genes involved in the synthesis of anti-Gal. B lymphocytes from various individuals were transformed by EBV, the clones producing anti-Gal were isolated, the specificity and affinity of the Abs were determined, and the VH genes were sequenced. The affinity of anti-Gal clones for the free radiolabeled alpha-galactosyl epitope ranged between 1.1 x 10(6) M-1 and 5 x 10(8) M-1. Eight of the nine clones studied used VH3 family genes and one clone used a VH1 family gene. Four of the five VH3 genes used were found to form a cluster of related sequences, suggesting that functional constraints may lead to the use of VH3 genes with structural motifs that are suited for specific interactions with the alpha-galactosyl epitope. Comparison with known germline sequences for all of the clones studied and analysis of autologous germ-line genes in two of the clones indicate that anti-Gal VH genes undergo somatic mutations. These somatic mutations may provide a pool of variants that are available for affinity maturation.

The site and stage of anti-DNA B-cell deletion.

Antibodies to DNA and nucleoproteins are found in sera of individuals with systemic autoimmune disease. In the population (and in the autoimmune mouse strain MRL/lpr) there is a great variety of such antinuclear antibodies, but individuals with systemic lupus erythematosus or single MRL mice express a subset only of the antinuclear specificities found in the population. These observations have been interpreted to mean that these antibodies arise by immunization. The oligoclonal nature of the autoantibody response and the evidence of selection acting on somatically mutated autoantibodies favour this interpretation. Specific activation of autoantibodies in disease implies either that autoantibodies are regulated in non-diseased individuals or that autoantigen availability is variable. The former has been demonstrated in anti-DNA transgenic mice. In normal mice, transgene-encoded antibodies against double-stranded (ds) DNA are not expressed in serum or on B cells. Here we describe modified anti-dsDNA transgenic mice which allow us to study the site and developmental stage at which such B-cell regulation occurs. This model shows that in normal mice B cells expressing anti-DNA specificity are deleted in the bone marrow at a pre-B to immature B transitional stage.

Deletion and editing of B cells that express antibodies to DNA.

We previously demonstrated that in mice transgenic for genes coding for an anti-ssDNA autoantibody B cells were functionally inactivated but not physically deleted. We have now extended this model by introducing an arginine into the CDR2 of the heavy chain transgene. This change alters the specificity of the Ab from anti-ssDNA to anti-dsDNA and increases the affinity for ssDNA. Mice carrying this transgene displayed a significant reduction of peripheral B cells and anti-dsDNA B cells were not recovered from the spleens. The remaining B cells escape deletion by revising their Ag receptors in several ways: 1) elimination of the transgenic heavy chain gene via intrachromosomal recombination, followed by rearrangement and expression of endogenous VH genes; 2) ongoing rearrangement of endogenous kappa light chain genes to generate a non-dsDNA-binding Ab; and 3) expression of a rare V lambda gene, V lambda x, to generate a non-DNA-binding Ab.

Genetic and structural evidence for antigen selection of anti-DNA antibodies.

The primary structure of anti-DNA antibodies is highly diverse, a result of different germline variable (V) gene use, different combinations of immunoglobulin gene segments, peculiar heavy chain complementarity determining region 3 (H-CDR3) segments, and somatic mutations. Nevertheless, tertiary structure predictions reveal common features that yield information about likely contact sites in the anti-DNA combining site. That these contacts are involved with DNA binding is supported by recurrent features of a newly compiled set of homology groups of 13 variable regions of heavy chains (VH) and 11 variable regions of light chains (VL), characteristic pattern of somatic mutations, and the results of site-directed mutagenesis. The role of antigen in the etiology of the autoimmune response is viewed in light of recent data on overlaps between anti-DNA and anti-nucleic acid binding protein specificities.

Residues that mediate DNA binding of autoimmune antibodies.

Somatic mutations to arginine (R) are a common feature of a subset of J558 H chain genes that code for the majority of high-affinity, anti-dsDNA antibodies in autoimmune MRL/lpr mice. To examine the consequences of such amino acid substitutions on DNA binding, we reverted three somatic mutations of a prototypic anti-dsDNA H chain gene, VH3H9, and assayed the effect of those reversions by expression in a V lambda 1 L chain-only plasmacytoma line. Reversion of R53 eliminated virtually all dsDNA binding and sharply reduced ssDNA affinity. While the complete germ-line revertant of VH3H9 retained a low level of DNA binding, the substitution of R96, a product of N base addition in the third complementarity determining region (CDR3), with glycine (G) was sufficient to abolish measureable DNA specificity. Antibodies with higher affinity for DNA were generated by introducing arginines into VH3H9 at any one of four positions where somatic mutations to arginine had been identified by sequencing other anti-dsDNA J558 H chain genes. All four arginine mutants showed affinity increments consistent with their direct involvement in DNA binding, although one such mutant, K64R, required the simultaneous reversion of an adjacent aspartic acid (D) to the germ-line glycine. Two variants with three nongerm-line arginines showed further improvements in DNA affinity suggesting that their contributions to DNA binding may be additive. Molecular modeling of antibody and mutant F(ab) structures and calculations of their electrostatic potentials were used as an aid in interpreting the results and in predicting the location and size of possible combining sites.