Among the 35 million people who received the AstraZeneca COVID-19 vaccine in the UK, approximately 1,050 developed vaccine-induced immune thrombocytopenia and thrombosis — a life-threatening blood-clotting disorder. Published in the New England Journal of Medicine and covered in Nature in February 2026, researchers traced the disorder to a single amino acid substitution in the antibody-producing B cells of affected patients. A positively charged lysine was swapped for a negatively charged glutamic acid. One charge flip in one amino acid position reshaped the antibody so that it bound platelet factor 4 instead of the viral protein it was supposed to target.
All 21 VITT patients studied carried this mutation. The researchers described the result as “the first time we've been able to trace an autoimmune disorder back to the original triggering event” — from the initial antigen exposure through the somatic hypermutation that created the pathogenic antibody to the clinical syndrome.
The structural insight is about the thinness of the boundary between function and catastrophe at the molecular level. The immune system generates antibody diversity through deliberate mutation — somatic hypermutation randomly alters antibody genes to produce variants that might bind targets better. Most variants are neutral or harmful and get eliminated. The VITT mutation wasn't eliminated because it bound something strongly — it just bound the wrong thing. The selection mechanism that keeps the immune system functional is imperfect, and the cost of imperfection is not gradual degradation but a discontinuous jump from protection to self-destruction.
One proton's worth of charge difference. Not a broken gene, not a missing protein, not a toxin — a charge reversal in a single molecular position. The mutation is within the normal range of somatic hypermutation; the pathology is catastrophic. The distance between a working immune system and a lethal one, measured in angstroms, is essentially zero.