Alzheimer's disease (AD) differs from other forms of dementia in its relation to amyloid beta peptide (Abeta). Abeta, a proteolytic product of amyloid precursor proteins (APP), has a toxic effect on neuronal cells, which involves perturbation of their Ca(2+) homeostasis. This effect implies that changes of protein expression in neuronal cells with calcium stress should provide a molecular marker for this disease. In the present study, we used the supernatant from a neuronal cell culture after incubation with or without Abeta and isolated a Ca(2+)-dependent acidic phospholipid binding fraction to perform a proteomic study. Several unique proteins were identified after incubation with Abeta. We focused on annexin A5, among these proteins, because it binds both Ca(2+) and lipids likely to be involved in calcium homeostasis. Tg2576 transgenic mice (AD model) overexpressing mutant human APP showed a significant increase of annexin A5 in the brain cortex but not in other organs, including liver, kidney, lung, and intestine. In human plasma samples, the level of annexin A5 was significantly increased in a proportion of AD patients compared with a control group (P < 0.0001 in the logistic regression analysis). From the receiver operating characteristic (ROC) curve with plasma annexin A5 concentrations, the mean area under the curve (AUC 0.898) suggests that annexin A5 is a favorable marker for AD.