The existence of self-reactive antibodies is a potential problem with respect to the development of autoimmune disease. Antibodies are generated by random association of immunoglobulin variable, diversity and joining region genes, but the risk associated with generating a diverse repertoire in this manner is the development of self-reactive antibodies. It has been known for some time that there are three mechanisms by which self-reactive antibodies can be eliminated — receptor editing, deletion and anergy. But the extent of autoantibody production and the stage at which these mechanisms operate are not clear. A new study from Wardemann and colleagues now establishes the extent of autoantibody production and indicates two stages at which self reactivity is modified.

Antibodies were cloned from single B cells derived from the blood and bone marrow of healthy human donors. The B cells were phenotypically defined as pre-B cells, early immature cells, immature B cells, new emigrant B cells or mature B cells. Two features of antibody amino-acid sequences that have previously been associated with self-reactive antibodies — a long CDR3 (complementarity-determining region 3) and positively charged amino acids in CDR3 in the immunoglobulin heavy chain — were analysed in the cloned antibodies. These features were enriched in the pre-B and early immature B cells, but were progressively lost as the B cells matured.

To assess individual antibodies for self reactivity, the antibodies were expressed in vitro and tested using a standard clinical ELISA for reactivity against nuclear and cytoplasmic antigens. 76% of antibodies from early immature B cells were self reactive — this figure decreased to 43.1% in immature B cells and to ∼20% as recently emigrated B cells matured into mature naive B cells, showing that self-reactive antibodies are eliminated at two distinct stages of development.

Further analysis of the cloned antibodies showed that more than half of the antibodies produced from early B-cell compartments were polyreactive, in that they bound to more than one antigen in a defined panel of antigens tested, but most of these polyreactive antibodies were removed from the repertoire by the immature B-cell stage.

This study shows that a large percentage of developing B cells produce autoreactive antibodies. Although these cells are removed from the population at two distinct checkpoints, ∼20% of mature B cells still produce autoantibodies with the potential to contribute to autoimmune disease.