Brain-reactive autoantibodies are nearly ubiquitous in human sera and may be linked to pathology in the context of blood-brain barrier breakdown

Previous studies have reported antibodies bound to cells in postmortem Alzheimer's disease (AD) brains, which are only rarely observed in the brains of healthy, age-matched controls. This implies that brain-reactive autoantibodies exist in the sera of AD individuals and can gain access to the brain interstitium. To investigate this possibility, we determined the prevalence of brain-reactive antibodies in sera from AD patients, patients with other neurodegenerative diseases, age-matched, non-demented controls and healthy younger individuals via immunohistochemistry and western blot analysis. Surprisingly, western analyses revealed that 92% of all human sera tested contain brain-reactive autoantibodies. When sera were used to probe western blots of human, pig, or rat brain membrane proteins, a number of comparably-sized protein targets were detected, suggesting cross-species reactivity. While the presence of brain-reactive autoantibodies was nearly ubiquitous in human sera, some autoantibodies appeared to be associated with age or disease. Furthermore, the intensity of antibody binding to brain tissue elements, especially the surfaces of neurons, correlated more closely to the serum's autoantibody profile than to age or the presence of neurodegenerative disease. However, while the blood brain barrier (BBB) in control brains remained intact, BBB breakdown was common in AD brains. Results suggest a high prevalence of brain-reactive antibodies in human sera which, in the common context of BBB compromise, leads us to propose that these antibodies may contribute to the initiation and/or pathogenesis of AD and other neurodegenerative diseases. (C) 2010 Elsevier B.V. All rights reserved.