Antibody binding to total protein extracted from a mammalian source (HeLa cells) and from a prokaryotic source (Escherichiu coli) was compared in sera from patients with systemic lupus erythematosus (SLE) and sera from normal subjects. When the average numbers of peptides or proteins recognized by IgG antibodies were compared on immunoblots, SLE sera bound to a significantly greater number of proteins from the HeLa cell extract than did sera from normal individuals (P < 0.001). In contrast, SLE sera actually bound to fewer E coli proteins than did the sera obtained from normal controls, although the difference was not statistically significant. There was no correlation between the number of E coli proteins and HeLa proteins recognized by individual SLE sera, and there was no trend toward reactivity with a larger number of antigens in sera containing higher levels of IgG. IgG from SLE sera did not bind to 6 purified eukaryotic protein standards (selected solely on the basis of differences in size and charge) either in their denatured state or in their native state. These findings indicate that the high levels of IgG antibodies against selected eukaryotic intracellular pro- May 23, 1988. teins in patients with SLE cannot be explained by a random polyclonal B cell activation.The etiology of autoantibody production in systemic lupus erythematosus (SLE) is uncertain. Some investigators have attempted to explain autoantibody production by intrinsic B cell hyperactivity (for review, see ref. 1) or by T lymphocyte suppressor cell deficiency (2,3). Such mechanisms would be expected to result in a broad (polyclonal) B cell expansion (4), with a presumably random profile of abnormal antibody reactivity. Recent analyses of autoantibody specificities in SLE, however, suggest that of the approximately 2,000 mammalian intracellular proteins detectable by current methods (3, only relatively few are targeted by lupus autoantibodies (6).To further determine whether sera from patients with SLE are broadly reactive to the universe of potential antigens, we compared the antibody reactivity of lupus sera with that of sera from normal individuals. Purified eukaryotic protein standards and total Escherichia coli cell extracts were used as antigen sources. Our findings indicate that lupus sera do not show increased or unusual binding to randomly selected eukaryotic proteins nor to the many hundreds or thousands of potential E coli protein epitopes (7) displayed on an immunoblot .