In transgenic animal models, humoral immunity directed against the beta-amyloid peptide (Abeta), which is deposited in the brains of AD patients, can reduce Abeta plaques and restore memory. However, initial clinical trials using active immunization with Abeta1-42 (plus adjuvant) had to be stopped as a subset of patients developed meningoencephalitis, likely due to cytotoxic T cell reactions against Abeta. Previously, we demonstrated that retrovirus-like particles displaying on their surface repetitive arrays of self and foreign Ags can serve as potent immunogens. In this study, we generated retrovirus-like particles that display the 15 N-terminal residues of human Abeta (lacking known T cell epitopes) fused to the transmembrane domain of platelet-derived growth factor receptor (Abeta retroparticles). Western blot analysis, ELISA, and immunogold electron microscopy revealed efficient incorporation of the fusion proteins into the particle membrane. Without the use of adjuvants, single immunization of WT mice with Abeta retroparticles evoked high and long-lived Abeta-specific IgG titers of noninflammatory Th2 isotypes (IgG1 and IgG2b) and led to restimulatable B cell memory. Likewise, immunization of transgenic APP23 model mice induced comparable Ab levels. The CNS of immunized wild-type mice revealed neither infiltrating lymphocytes nor activated microglia, and no peripheral autoreactive T cells were detectable. Importantly, vaccination not only reduced Abeta plaque load to approximately 60% of controls and lowered both insoluble Abeta40 as well as Abeta42 in APP23 brain, but also significantly reduced cerebral soluble Abeta species. In summary, Abeta retroparticle vaccination may thus hold promise as a novel efficient future candidate vaccine for active immunotherapy of Alzheimer's disease.