Background and objectives Immune complexes (IC) and autoantibodies against nuclear antigens are associated with systemic lupus erythematosus (SLE). Autoantibodies are suggested to have a pathogenetic impact by participating in IC formation. Few studies have however focused on studying autoantibody content in SLE IC, and the relative appearance of autoantibodies in IC versus in (IC-depleted) sera. By using polyethylene glycol (PEG) precipitation, we have studied how individual autoantibodies accumulate to different extents in SLE IC. PEG precipitation is however a crude technique by also precipitating other high molecular weight proteins. Here we have developed a technique to isolate pure IC from serum to enable detailed characterisation of autoantibody content individually in pure IC and IC-depleted serum.

Material and methods IC in SLE sera were adsorbed to C1q-coated magnetic beads and thereafter eluted. Autoantibodies against dsDNA, Ro52, Ro60, SSB, Sm, Sm/RNP, ribosomal P antigen, histones and PCNA were determined by addressable laser bead immunoassay in serum, IC-depleted serum and purified IC. Circulating IC levels were determined by C1q binding. The levels of autoantibodies were normalised in relation to IgG levels. The C1q binding system was verified in real-time on a fully automated Quartz Crystal Microbalance with Dissipation Monitoring (QCM-D) system. Levels of autoantibodies in IC, IC-depleted serum and full serum were followed in five SLE patients during one year after treatment with anti-CD20 therapy.

Results Adsorption of sera resulted in a 97.2%–99.5% reduction in circulating IC. QCM-D analyses showed the importance of each step in the adsorption/elution process. Serum samples and eluted IC showed different temporal patterns, and there were also strikingly different patterns between total amounts and concentrations related to total IgG levels in IC eluates. In some patients, anti-dsDNA levels and concentrations in IC eluates dropped sharply after anti-CD20 therapy whereas the corresponding levels for anti-Ro60/Ro52/SSB increased. All patients regained circulating B cells during the follow-up, and in at least 3/5 patients a transient or remaining increase in autoantibody concentration in eluted IC heralded the re-appearance of B cells.

Conclusions We have developed a method to carefully quantify autoantibody content in circulating IC. Measurement of autoantibody amount and concentration in circulation SLE IC show individual temporal patterns that do not parallel serum autoantibody levels. Changes in autoantibody IC content/concentration might precede the re-occurrence of B-cells in rituximab-treated SLE patients.