A B S T R A C T To elucidate the relationship between Epstein-Barr virus (EBV) and rheumatoid arthritis (RA), we measured antibodies to RA-associated nuclear antigen (anti-RANA) and three other EBV-related antigens in the sera of RA patients and controls. Our study groups consisted of 89 patients with definite or classical RA, mean age 56, male/female ratio 47:42; and 53 normal and osteoarthritis controls, mean age 51, male/female ratio 25:28. In addition to anti-RANA, we measured antibodies to viral capsid antigen (anti-VCA), early antigen (anti-EA) and EBV-associated nuclear antigen (anti-EBNA).Anti-RANA was detected in 71% of RA patients but in only 6% of controls. Elevated anti-VCA titers (>1: 160) were more common in RA patients than controls, 31% compared with 15%. The geometric mean titer of anti-VCA was significantly higher in the RA group, 133 compared with 58. Anti-EA was present in 53% of RA patients but only 19% of controls. Anti-EA in elevated titers (>1:20) was present in 26% of RA patients but only 7% of controls. Characterization of the anti-EA antibodies revealed that the RA patients reacted primarily with the diffuse component, whereas the majority of the controls reacted with the restricted component ofthe EA complex. In contrast, the frequencies, distributions, and geometric mean titers of anti-EBNA were not significantly different between the two groups. Correlative analysis ofthese antibodies showed highly significant relationships between anti-VCA and anti-EA, and anti-RANA and anti-EBNA in the RA group. These data are compatible with the interpretation that RA patients have either more active EBV infections than controls or an altered regulation of their immune response to this infectious agent.

The immunologic specificities of antinuclear antibodies (ANA) in sera from 97 patients with rheumatoid arthritis (RA) were studied. The frequency of ANA detected by immunofluorescence with mouse kidney sections as substrate was 60% (59/97) compared to 13% (7/ 55) in controls. IgM ANA was positive in 41% (40/97) of RA sera, IgG ANA in 40% (39/97), and IgA ANA in 33% (32/97). Specificities of antibodies in the ANA positive sera were examined by radioimmunoassay for antibodies to DNA and by immunodiffusion for antibodies to Sm antigen, nuclear ribonucleoprotein, and the SS-A and SS-B nuclear antigens. Antibodies to these nuclear antigens were found infrequently (1.7% to 6.8%). Sera were further investigated for antibodies to histones by an immunofluorescent method previously described using acid-extracted and histone reconstituted tissue sections. Twenty-four percent of the ANA positive sera (14/55) reacted with a histone-dependent nuclear antigen. The relationship of rheumatoid factor to ANA was studied by isolating rheumatoid factors (RF) from aggregated IgG absorbant columns. Ten of 19 isolated RF showed ANA activity, which did not result from complexes of rheumatoid factor with IgG ANA since IgM rheumatoid factor, dissociated from IgG under acid con- ditions, continued to show ANA activity. In 8 of 16 RA sera tested, both ANA and rheumatoid factor activity could be inhibited by aggregated IgG. Five of the crossreacting rheumatoid factors reacted specifically with a histone-dependent antigen in the reconstitution assay. These findings show that approximately WO of rheumatoid factors possess cross-reactivity with nuclear components, and histones are involved in a significant number of these reactions.

To determine if rheumatoid arthritis (RA) patients have an increased frequency of Epstein-Barr virus (EBV) shedding into saliva, throat washings were collected from 59 patients with RA and 64 healthy adult controls. EBV in filtered samples was detected by transformation of human umbilical cord lymphocytes. EBV was detected in 27% (16/59) of throat washings of RA patients compared to 11% (7/64) of control samples. This difference is significant (p < 0.05). RA patients on steroids had a frequency of positivity of 43% (10/23) compared to 17% (6/36) in patients not on steroids. No correlation was demonstrated between steroid dose received by the patient and detection of EBV in throat washings. Parotid fluid samples were collected from 18 RA patients and 11 healthy controls and were tested for EBV by the transformation bioassay. EBV was detected in 18% (2/11) of samples from normal controls, but samples from RA patients were uniformly negative.