The lag between beta-amyloid (Abeta) deposition and neurodegeneration in Alzheimer's disease (AD) suggests that age-dependent factors are involved in the pathogenesis. Racemization of Ser and Asp in Abeta is a typical age-dependent modification in AD. We have shown recently that Abeta1-40 racemized at Ser(26) ([D-Ser(26)]Abeta1-40) is soluble and non-toxic to neuronal cells, but is easily converted by brain proteases to truncated toxic fragments, [D-Ser(26)]Abeta25-35/40. Furthermore, [D-Ser(26)]Abeta1-40 in vivo, produced a drastic and synergistic neuronal loss by enhancing the excitotoxicity when co-injected into rat hippocampus with ibotenic acid, an excitatory amino acid, suggesting an in vivo conversion of nontoxic [D-Ser(26)]Abeta1-40 to toxic fragments including [D-Ser(26)]Abeta25-35/40. In this study, we further investigated the mechanism behind the in vivo neuronal loss by [D-Ser(26)]Abeta1-40 and ibotenic acid in rats, and also searched for the presence of [D-Ser(26)]Abeta25-35/40 antigens in AD brains. Quantitative analyses of the damaged area indicate clearly that non-toxic [D-Ser(26)]Abeta1-40 caused as much neurodegeneration as toxic [D-Ser(26)]Abeta25-35/40. MK-801, an NMDA receptor antagonist, completely inhibited the neurodegeneration. The immunohistochemical analyses using anti-[D-Ser(26)]Abeta25-35/40-specific antibodies demonstrated the presence of [D-Ser(26)]Abeta25-35/40 antigens in senile plaques and in degenerating hippocampal CA1 neurons in AD brains, but not in age-matched control brains. These results strengthen our hypothesis that soluble [D-Ser(26)]Abeta1-40, possibly produced during aging, is released from plaques and converted by proteolysis to toxic [D-Ser(26)]Abeta25-35/40, which damage hippocampal CA1 neurons by enhancing excitotoxicity in AD. This may account for the lag between Abeta deposition and neurodegeneration in AD. (C) 2002 Wiley-Liss, Inc.
