Detection of anti-DNA antibody using synthetic antigens. Characterization and clinical significance of binding of poly (deoxyadenylate-deoxythymidylate) by serum.

Virtually all preparations of DNA used to detect antibody to native DNA (nDNA) by binding assays have been found to be subtly contaminated by single stranded DNA. Because recent DNA binding data have directly challenged the unique role previously attributed to these antibodies in systemic lupus erythematosus (SLE), resolution of the consequent ambiguity is of theoretical and practical importance. It is proposed that a synthetic nDNA molecule (dAT) might circumvent this difficulty by being antigenically equivalent to nDNA while, on theoretical grounds, lacking significant contamination with single stranded DNA or other cellular antigens. These expectations were generally confirmed by biochemical and immunological analyses. In clinical studies, sera from 124 pateints with SLE and from controls were examined for their ability to bind dAT. In contrast to results with KB binding, patients with non-SLE rheumatologic disorders were indistinguishable from normals by dAT binding. dAT binding was elevated in 85% of sera from SLE patients with clinically-judged active nephritis but in only 9% of those with inactive renal disease. Active non-renal disease, including cerebritis, was not associated with increased dAT binding. Individual non-lupus sera which bound increased amounts of KB DNA, failed to bind dAT. It is suggested that such binding resulted from contaminating non-nDNA antigens. When elevated, dAT binding, like KB binding, varied with disease activity and might thus be useful as a parameter thereof. In several patients elevated dAT binding led to the finding, on biopsy, of clinically silent, active, diffuse proliferative nephritis. It is concluded that use of synthetic nDNA antigens such as dAT may offer theoretical and practical advantages over naturally-derived preparations in detecting anti-nDNA, both clinically and for investigational purposes. Also, caution is urged in interpreting DNA binding data derived from incompletely characterized systems, particularly with regard to the occurrence of anti-nDNA antibodies in serum.