Limiting-dilution analysis of T cell reactivity to mycobacterial antigens in peripheral blood and synovium from rheumatoid arthritis patients.

Limiting-dilution analysis (LDA) was used to quantify the frequency of Mycobacterium bovis BCG- and 65-kD-reactive T cells in paired samples of peripheral blood and synovial tissue from patients with rheumatoid arthritis. The frequency of BCG-reactive T cells detected in the peripheral blood of patients ranged from 1/585 to 1/7639 versus a control frequency range of 1/480 to 1/6773. The frequency of such cells in the synovium was found to be much lower than it was in peripheral blood; in fact, in 80% of patients synovial BCG-reactive T cells were not detected. The frequency of 65-kD-reactive cells in the peripheral blood of each individual was lower than the frequency of BCG-reactive cells (range 1/3738 to 1/55,324), as would be expected. However, no synovial 65-kD-reactive cells were detected from any of the patients studied. The LDA assay for the 65-kD antigen was consistent with the single hit model, that for BCG was not. The relatively high proportion of mycobacterial-reactive precursors seen in the peripheral blood of non-vaccinated individuals may reflect a population of cells induced either by natural environmental exposure to mycobacteria or, given the highly conserved nature of heat shock proteins across phylogeny, by some other infection. The results also suggest that the frequent finding of reactivity to proteins such as the 65-kD heat shock protein contained within BCG may not be a generalized phenomenon in rheumatoid synovium.