A surprisingly large number of B cells express two different Ig light chain alleles (red and green).

Transgenic B cells forced to express one self-reactive Ig allele can switch on a second nonautoreative Ig in a process called allelic inclusion. This second Ig appears to dilute the autoreactive potential of the cell. But it is unknown whether allelic inclusion, which defies the "one lymphocyte–one antibody" theory of B cell specificity, can occur under normal physiological conditions.

Using physiologically normal mice that carried one human and one endogenous allele of the Ig light chain gene (to enable identification of both), Casellas et al. discovered that approximately 10% of B cells express two Ig light chain alleles.

The coexpressing cells might arise either because allelic exclusion—the permanent switching-off of one allele—fails or because the second allele gets switched on during receptor editing. Such editing, which occurs if the first Ig produced by a B cell is discovered to be autoreactive, involves switching back on the machinery necessary for Ig gene recombination and expression. With this machinery back on, the second allele might get activated.

Consistent with this hypothesis, coexpression and autoreactivity were correlated. The team also sequenced the Ig genes in coexpressing cells and found strong evidence that editing had occurred. The developmental path of these coexpressing cells was slower than that of monoexpressing cells, consistent with a necessary pause for editing.

The coexpressing B cells, though large in number, appear to have their autoreactive tendencies tempered by the second Ig allele. However, B cells undergo somatic hypermutation at their Ig genes during an immune response, suggesting that autoimmune disease might be triggered if the innocuous allele is edited into oblivion and the wolf is let loose.