The conformational change in amyloid beta (Abeta) peptide from its monomeric form to aggregates is crucial in the pathogenesis of Alzheimer's disease (AD). In the healthy brain, some unidentified chaperones appear to prevent the aggregation of Abeta. Here we reported that lipocalin-type prostaglandin D synthase (L-PGDS)/beta-trace, the most abundant cerebrospinal fluid (CSF) protein produced in the brain, was localized in amyloid plaques in both AD patients and AD-model Tg2576 mice. Surface plasmon resonance analysis revealed that L-PGDS/beta-trace tightly bound to Abeta monomers and fibrils with high affinity (K(D) = 18-50 nM) and that L-PGDS/beta-trace recognized residues 25-28 in Abeta, which is the key region for its conformational change to a beta-sheet structure. The results of a thioflavin T fluorescence assay to monitor Abeta aggregation disclosed that L-PGDS/beta-trace inhibited the spontaneous aggregation of Abeta (1-40) and Abeta (1-42) within its physiological range (1-5 microM) in CSF. L-PGDS/beta-trace also prevented the seed-dependent aggregation of 50 microM Abeta with K(i) of 0.75 microM. Moreover, the inhibitory activity toward Abeta (1-40) aggregation in human CSF was decreased by 60% when L-PGDS/beta-trace was removed from the CSF by immunoaffinity chromatography. The deposition of Abeta after intraventricular infusion of Abeta (1-42) was 3.5-fold higher in L-PGDS-deficient mice and reduced to 23% in L-PGDS-overexpressing mice as compared with their wild-type levels. These data indicate that L-PGDS/beta-trace is a major endogenous Abeta-chaperone in the brain and suggest that the disturbance of this function may be involved in the onset and progression of AD. Our findings may provide a diagnostic and therapeutic approach for AD.