He and others have shown since then that this protective aggregation response occurs in other neurodegenerative diseases as well. It may explain the repeated failure of experimental trials for treating Alzheimer’s disease by targeting plaques, he said: If the amyloid plaques characteristic of the disease form to protectively bind up the defective protein, then breaking up the plaques might do more harm than good.

“It is a hard concept for humans to grasp, since it seems intuitive that things that look abnormal should be ‘bad’ and pathogenic,” Finkbeiner wrote. “But biology is complex, full of many feedback loops, so it is important that people don’t get fooled jumping to conclusions.”