Clues to this self-regulatory behavior emerged from earlier immunization studies. The group succeeded in combating autoimmune diseases by injecting adjuvant plus DNA vaccines encoding pro-inflammatory mediators. Antibodies against the vaccine-encoded mediators apparently dampened both inflammation and disease. But this antibody response looked less like a de novo response and more like the amplification of an existing response.
Sure enough, when the Israeli group looked in models of autoimmunity, they found antibody responses against common pro-inflammatory mediators such as TNF-α. The anti–TNF-α response could be prevented by inducing neonatal tolerance to TNF-α; this resulted in a much more serious disease after subsequent induction of autoimmunity mimicking rheumatoid arthritis.
The natural antibody response was directed at pro-inflammatory rather than regulatory mediators, and was seen only during autoimmune rather than local inflammatory reactions. This specificity remains a mystery. “The next question is to find the difference between an immune response and an autoimmune response,” says Karin. “Nobody has really found a difference.” One possibility is that the immune system somehow reacts to any self protein that rises far above its normal level. Or there may be earlier controls on the production or regulation of the cells that make the antimediator antibodies. “We're entering an empty field here,” says Karin. “It's not an easy question to answer.” ▪