Specificity of anti-DNA antibodies in SLE--II. Relative contribution of backbone, secondary structure and nucleotide sequence to DNA binding

Hydrolysis of Oligodeoxyribonucleotides on the Microarray Surface and in Solution by Catalytic Anti-DNA Antibodies in Systemic Lupus Erythematosus

Anti-DNA antibodies are known to be classical serological hallmarks of systemic lupus erythematosus (SLE). In addition to high-affinity antibodies, the autoantibody pool also contains natural catalytic anti-DNA antibodies that recognize and hydrolyze DNA. However, the specificity of such antibodies is uncertain. In addition, DNA binding to a surface such as the cell membrane, can also affect its recognition by antibodies. Here, we analyzed the hydrolysis of short oligodeoxyribonucleotides (ODNs) immobilized on the microarray surface and in solution by catalytic anti-DNA antibodies from SLE patients. It has been shown that IgG antibodies from SLE patients hydrolyze ODNs more effectively both in solution and on the surface, compared to IgG from healthy individuals. The data obtained indicate a more efficient hydrolysis of ODNs in solution than immobilized ODNs on the surface. In addition, differences in the specificity of recognition and hydrolysis of certain ODNs by anti-DNA antibodies were revealed, indicating the formation of autoantibodies to specific DNA motifs in SLE. The data obtained expand our understanding of the role of anti-DNA antibodies in SLE. Differences in the recognition and hydrolysis of surface-tethered and dissolved ODNs need to be considered in DNA microarray applications.

The binding of sera of patients with SLE to bacterial and mammalian DNA

Systemic lupus erythematosus (SLE) is a prototypic autoimmune disease characterized by antibodies to DNA (anti-DNA). Although these antibodies have features of antigen drive, the source of this DNA is not defined. To assess the potential role of foreign and self-DNA as driving antigens, the specificity of SLE sera for bacterial and mammalian DNA was evaluated. Micrococcus lysodeikticus (MC) and calf thymus (CT) DNA were tested as antigens, with absorption on CT DNA columns used to identify antibodies to antigenic sites on the two DNA. Among 9 sets of longitudinal sera tested, all showed binding to both DNA, and none showed exclusive or predominant binding to CT DNA. With absorbed sera, antibodies could be distinguished in terms of cross-reactive or selective binding to the DNAs. These findings suggest that anti-DNA antibodies vary in specificity and are consistent with a role of both foreign and self-DNA in anti-DNA induction.

Lupus autoantibodies to native DNA preferentially bind DNA presented on PolIV

While immunoglobulin G (IgG) antibodies to double-stranded (ds)DNA are serological markers of systemic lupus erythematosus (SLE), not all antibodies to DNA (anti-DNA) are able to cause tissue damage to a similar extent. It has been proposed that anti-DNA-induced renal damage could be linked to differences in the fine specificity of the antibodies. In an attempt to gain insight into their fine binding properties, we investigated the cross-reactivity of two human lupus monoclonal IgG anti-dsDNA (B3 and RH14) to a recently described Escherichia coli PolIV (a DNA polymerase). These autoantibodies possess distinct pathogenic properties in severe combined immunodeficient (SCID) mice. Although both antibodies cause proteinuria, only RH14 induces early histological features of lupus nephritis. Both RH14 and B3 bound PolIV; however, they exhibited a marked difference in their reactivity to the PolIV-dsDNA complex. Alhough RH14 exhibited significant activity to the complex, the binding of B3 to PolIV complexed with dsDNA was almost abolished. Furthermore, there was a significant difference in the way the lupus sera recognized naked dsDNA and that presented on PolIV. Although 67% of lupus sera bound naked dsDNA, approximately 90% of these sera (93% calf thymus DNA; 90% synthetic oligonucleotide) reacted to the complex when dsDNA was presented on PolIV. Thus, the IgG anti-dsDNA likely to exist in lupus patients may be distinguished into those that recognize dsDNA in the context of PolIV and those which do not. This difference in binding ability may help to distinguish those dsDNA antibodies that are more pathogenic.

Anti-DNA antibodies exhibit different binding motif preferences for single stranded or double stranded DNA

A common feature for most anti-DNA antibodies (Abs) is their induction in an antigen (Ag)-driven specific clonal expansion pattern though crossreactivity. However, the fine sequences in DNA Ags that interact directly with immune system and the ability of DNA to induce immune responses is poorly understood. In order to define the characteristics of possible antigenic determinants in DNA Ags, we immunized mice with the pBR322 plasmid and used antisera as source of anti-DNA Abs. A systemic evolution of ligands by exponential enrichment (SELEX) procedure was performed on an oligodeoxynucleotide library either in single stranded (ss-) or double stranded (ds-) form. The SELEXed fragments were cloned and sequenced. The resulting sequences were analyzed using the Multiple Alignment Construction and Analysis Workbench program. We show that the fragments of ss- or ds- form bound by a same stock of antibodies were different in their conserved sequences. ss-DNA fragments recognized by anti-DNA Abs were rich in cacc, caccc, accc or cccc blocks, while the same stock of Abs exhibited significant preference for the (5'gcg3'/3'cgc5') motif located in ds-DNA. At the same time sera from unimmunized control mice showed no sequence preference in either ss-DNA or ds-DNA. Future improvement of this work and the potential use of SELEX for studies of DNA Ags are also discussed.

Repair of DNA damage induced in systemic lupus erythematosus skin fibroblasts by simulated sunlight

Skin fibroblasts derived from three normal individuals and three patients exhibiting the disease systemic lupus erythematosus (SLE) were exposed to the simulated sunlight produced by a solar simulator. The induction and repair of DNA damage induced by this treatment were examined. The total number of lesions repaired by excision, as well as the removal of pyrimidine dimers and E. coli endonuclease III--sensitive sites did not differ significantly in the three SLE cell strains compared with normal cells. However, abnormalities in the formation and maintenance of DNA-protein crosslinks (DPC) and DNA single-strand breaks (SSB) were found in SLE-4 and SLE-5 following simulated sunlight exposure. In contrast, SLE-3 cells exhibited responses similar to normal cells in reference to SSB and DPC formation. These findings correlate well with the previously determined UV sensitivity of these SLE cell strains.

Anti-DNA antibodies induced by BK virus inoculations. Demonstration of the specificities for eukaryotic dsDNA and synthetic polynucleotides

The ability of BK virus to induce anti-DNA antibodies in rabbits, and the ability of these antibodies to bind natural eukaryotic DNA and synthetic polynucleotides have been analysed. The specificity of the binding was assayed by inhibition of anti-dsDNA and -ssDNA ELISA tests with dsDNA, ssDNA, and synthetic single-stranded as well as double-stranded polynucleotides. The anti-dsDNA activity of two rabbit antisera was effectively inhibited by dsDNA and ssDNA and poly(dAdT)-poly(dAdT). The other nucleotide antigens produced relatively less inhibition. The anti-ssDNA binding was most efficiently inhibited by the homologous antigen, whereas inhibition by dsDNA only reached approximately 70% of the maximum as defined by ssDNA as inhibitor. This indicates the existence of a selective anti-ssDNA antibody population and a population recognizing both ssDNA and dsDNA within the sera. Cross-reaction of the induced anti-DNA antibodies with phospholipid antigens, such as cardiolipin, phosphatidylic acid, and bacterial cell surface, could not be demonstrated. We conclude that antibodies resulting from inoculation with BK virus specifically bind to dsDNA and ssDNA and possess a high affinity for the synthetic duplex poly(dAdT). In this way, they have some similarities with anti-DNA antibodies encountered in SLE (systemic lupus erythematosus) in both man and mouse.

Thymine and guanine base specificity of human myeloma proteins with anti-DNA activity

To further our understanding of the molecular basis of DNA-autoantibody interactions, we have characterized the specificities of three IgG human myeloma proteins that bind DNA. We measured their binding to synthetic single- and double-stranded homopolynucleotides, random and alternating copolymers, oligonucleotides, and nucleotides or nucleosides conjugated to non-nucleic acid carriers. All three antibodies bound single-stranded nucleic acids, including both polyribonucleotides and polydeoxyribonucleotides. They varied in relative affinities for polynucleotides of varying base composition. Polymers containing the purines guanine or hypoxanthine and/or the pyrimidine thymine were most reactive with all three proteins. A myeloma protein that reacted with poly(G), poly(I), or poly(dT) also bound to the corresponding nucleosides or nucleotides conjugated to bovine serum albumin. None of the antibodies reacted with base-paired double-helical polynucleotides (double-stranded RNA, RNA-DNA hybrid or double-stranded DNA). The results indicate that base specificity is prominent in their reactions and that the accessible epitopes in single-stranded polynucleotides become masked upon base pairing in double-stranded helices. These findings suggest a model in which positions N1 and O6 of guanine and hypoxanthine and N3 and O4 of thymine interact with amino acids of the antibody-combining site.

A recognition site on synthetic helical oligonucleotides for monoclonal anti-native DNA autoantibody

The binding site in native DNA for a murine monoclonal anti-DNA autoantibody was investigated by measurements of competitive binding of a series of synthetic helical oligonucleotides. The antibody bound to a (dG-dC)3 or (dG-dC)4 core in the center of a base-paired octadecanucleotide. Reactions of analogues containing modifications or substitutions at specific sites indicated that the antibody bound to portions of cytosine and guanine in the major groove, a limited region of the backbone, and the 2-amino group of one guanine in the minor groove. For these interactions to occur, the antibody combining site would straddle the backbone of one of the helical strands of DNA.

Antibodies to DNA

Antibodies that recognize specific conformational variations of DNA structure provide sensitive reagents for testing the extent to which such conformational heterogeneity occurs in nature. A most dramatic recent example has been the development and application of antibodies to left-handed Z-DNA. They provided the first identification of Z-DNA in fixed nuclei and chromosomes, and of DNA sequences that form Z-DNA under the influence of supercoiling. Antibodies have also been induced by chemically modified DNA and by synthetic polydeoxyribonucleotides that differ from the average B-DNA structure. These antibodies recognize only the features that differ from native DNA. In most experiments, native DNA itself is not immunogenic. Antibodies that do react with native DNA occur in sera of patients with autoimmune disease, but even monoclonal anti-DNA autoantibodies usually react with other polynucleotides as well. Anti-DNA antibodies, especially those of monoclonal origin, provide a model for the study of protein-nucleic acid recognition.

Detection of human autoantibodies specific for 5'-m7GMP and m7G(5')ppp(5')N

An enzyme-linked immunosorbent assay was utilized for the detection of spontaneously occurring antibodies with apparent specificities for m7G, 5'-m7GMP, and m7G(5')ppp(5')C. From the sera of 50 patients containing anti-nuclear antibodies, 48 (96%) possessed antibodies which bound to one or more immobilized nucleoside-BSA antigens (A-, G-, C-, U-, and T-BSA). Additionally, 8 (16%) of these sera contained immunoglobulins that reacted with m7G-BSA antigen. In these latter sera, soluble competitors such as m7G, 5'm7GMP, and m7G(5')ppp(5')C (but not 5'-AMP, -GMP, -CMP, -UMP, and -TMP or m1G and m22G) effectively inhibited antibody-binding to immobilized m7G-BSA. These results indicate the existence of spontaneously occurring anti-m7G antibodies in autoimmune diseases which are distinct from anti-G antibody populations.

A method for binding specificity analysis of anti-DNA autoantibodies in SLE

A pattern of differential binding between an NZB/NZW mouse-derived monoclonal anti-ssDNA antibody, V'D2, and restriction fragments of plasmid pBR322 DNA was shown by electrophoresis of the fragments through a denaturing agarose gel followed by their transfer onto nitrocellulose membrane and subsequent reaction of the immobilized DNA with the antibody and 125I-protein A. The antibody showed preferential binding to a 328 base pair Alu I + Hinf I fragment (denoted FD) (AT content, 60%), compared with the other fragments (AT contents, 40-56%). In dot blot assays the antibody bound only to poly(dT) and poly(dA,dT), failing to bind to other synthetic deoxyribopolynucleotides even at the highest concentration tested (300 ng). In competition experiments, the ability of unlabeled DNA to inhibit binding of V'D2 to FD increased with AT content of the DNA. It is concluded that V'D2 has preference for AT-rich DNA. In addition, poly(dA,dT) inhibited binding to a greater extent than either poly(dA) or poly(dT), indicating that base sequence may be important in defining the antigenic determinant. The method, appropriately modified, may be applicable to a wide range of natural nucleic acids and monoclonal antibodies, allowing detection and isolation of specific DNA fragments for detailed studies of antigenic determinants.