There are considerable data on synaptic dysfunction in Alzheimer's disease (AD). However, the precise molecular basis for synaptotoxicity in AD is not known. We tested the hypothesis that amyloid beta (Abeta), as produced in Tg2576 mice overexpressing a mutant form of amyloid precursor protein, leads to changes in SNAP-25, a molecule required for Ca-sensitive neurotransmitter vesicle exocytosis. Anti-Abeta antibody was injected into the third ventricle (icv) of 10-month-old Tg2576 mice, preceding formation of plaques. Immunodensity of glial fibrillary acidic protein (GFAP) and SNAP-25 were quantitated in the hippocampus 1 month later. SNAP-25 was reduced by 96% in the inner molecular layer (SMi) of dentate gyrus, by 95% in the hilum, and by 75-76% in stratum lucidum (SL), stratum oriens (SO), and stratum radiatum (SR) of CA1-CA3 of the Tg2576 mice. GFAP was increased by more than 50-fold, specifically within the neuropil of CA1-CA3, and by twofold in portions of fimbria. One injection of 10 microg of anti-Abeta antibody into the third ventricle at 10 months completely prevented or restored changes in GFAP at 11 months of age. The restoration of SNAP-25 by anti-Abeta antibody compared with wild type was 69% in CA1-SO, 93% in CA1-SR, 85% in CA3-SL, 77% in SMi, and 60-73% in hilum. In addition, whereas control injections of saline or IgG produced greatly increased GFAP diffusely in the hippocampus of Tg2576 animals, there was no increase in GFAP after anti-Abeta injection, suggesting a synergistic interaction of nonspecific trauma with Abeta in the transgenic mice. This is the first report of depleted SNAP-25 immunoreactivity in Tg models and the first report of icv injection of anti-Abeta antibody in this model of AD. The largest reductions of the SNAP-25 are in hilum and SMi, so either reduction in the septal-hilum-SMi path is primary or reduction in this path begins at an earlier age than in CA3-CA1 fields. A single icv injection of anti-Abeta antibody is potent in reversing Abeta effects and, therefore, represents a suitable model for investigating early Abeta toxicity. In addition, intrathecal or icv antibody may be an efficient means of treating or preventing toxicity in AD, particularly under conditions of immune hyporesponsivity.