Antibodies to guanosine triphosphate misidentified as anti-double-stranded DNA antibodies in a patient with antinuclear antibody-negative lupus, due to buckling of insolubilized assay DNA.

Objective: To investigate why the serum of a pediatric patient with systemic lupus erythematosus was persistently (>30 months) and strongly positive for antibodies to double-stranded DNA (dsDNA) as revealed by enzyme-linked immunosorbent assay (ELISA), but yielded negative results on the antinuclear antibody test (HEp-2 immunofluorescence [IF]).

Methods: The patient's antibodies were isolated on dsDNA and single-stranded DNA (ssDNA) supports, which were then examined by dsDNA ELISA and HEp-2 IF. Tests included the use of various inhibitors to determine the fine specificity of the antibodies. Other tests performed included immunoblotting, immunoprecipitation, Crithidia luciliae IF, and neutrophil IF.

Results: The antibodies isolated from the dsDNA and ssDNA supports were similar, in that they were of the IgG type, bound well in the dsDNA ELISA, and recognized a normally hidden nucleolar RNA antigen in HEp-2 cells. With both the dsDNA ELISA and nucleolar antigens, inhibition studies revealed that the epitope recognized was guanosine 5'-triphosphate (GTP). Binding of the antibodies was better to GTP than to guanosine 5'-monophosphate or cytidylyl (3'-5') guanosine, and, in turn, was better than to guanosine, while N7-methylated GTP was unreactive. The antibodies did not bind to dsDNA present in solution or in HEp-2 or Crithidia cells, but bound transfer RNA well and recognized a cytoplasmic RNA antigen in neutrophils.

Conclusion: A new problem in dsDNA ELISA is revealed in the occurrence of a hitherto-unknown and unusual buckling of the insolubilized DNA molecule, which, absent in dsDNA found in solution or in whole cells, presumably creates gaps of single-strandedness in the molecule. A new antibody specific for GTP is described in this patient, which may be clinically important.