In infections associated with immune complex disease, microbial antigens have rarely been found in the complexes. Using an enzyme-linked immunoassay, we studied the immune complexes of a patient who had hematogenous Pseudomonas aeruginosa osteomyelitis associated with palpable purpura, arthritis, and microscopic hematuria. After separation of the complexes into high and low molecular weight fractions, a 6-fold selective concentration of P aeruginosa antibody was found in the low molecular weight fraction compared with the concentration in the serum. FolIowing disruption of immunoglobulin, the high molecular weight fraction competed with solid-phase P aeruginosa antigen for P aeruginosa antibody from another source. After successful treatment of the infection, the patient's symptoms resolved, and the complexes disappeared. These findings strongly suggest that immune complexes in this patient contained P aeruginosa antigen and antibody that may have been pathogenetic in his disease.
A. HARDINAlthough circulating immune complexes are thought to be present in many diseases, it has often been difficult to identify antigen in the complexes (1). Even when immune complexes occur in infections, microbial antigens have rarely been found in the complexes (2). Furthermore, in many diseases, it remains unclear whether circulating immune complexes are pathogenetic or simply an epiphenomenon (3). We studied the immune complexes of a patient who had Pseudornonas aeruginosa vertebral osteomyelitis associated with palpable purpura, arthritis, and microscopic hematuria. We found evidence of both P aeruginosa antigen and antibody in the complexes.
MATERIALS AND METHODSThe patient's blood was drawn weekly during his acute illness and less frequently (approximately monthly) during convalescence. Serum from each sample was analyzed for putative circulating immune complexes by the 1251-Clq binding assay ( 4 3 , C3 levels by radial immunodiffusion, and P aeruginosa antibody titers by indirect immunofluorescence. The isolate of P aeruginosa from the patient's blood was maintained in continuous culture on cysteine-trypticase agar.To separate the immune complexes into antigen and antibody components, 60 ml of the patient's serum, pooled from blood samples drawn early in his illness, was precipitated with polyethylene glycol as previously described (6). The precipitate (total protein 49 mg) was then dissolved in 0.lM glycine buffer, pH 3.0, and chromatographed (column dimensions: 0.9 x 60 cm) in the same buffer at 4°C on Sephacryl 5-300 (superfine) (Pharmacia Fine Chemicals, Piscataway, NJ), which had not been used in previous experiments. The eluate (total protein 37 mg) contained 2 protein peaks, one near the 19s marker (the high molecular weight fraction) and the other near the 7s marker (the low molecular weight fraction). The protein in each peak was