Transnasal delivery of human A-beta peptides elicits impaired learning and memory performance in wild type mice

Background Murine models of Alzheimer’s disease (AD) are mainly based on overexpression of pathologic amyloid precursor protein and/or presenilins. Those genes resemble underlying cause of early onset type of AD while about 99 % of all human cases are to be characterized as sporadic, late onset. Appropriate animal models for this type of AD are still missing. We here investigated, if transnasal delivery of A-beta 42 peptides might serve to mimic pathological effects in mice. Results A-beta 42 peptides, used for the behavioral study, showed the expected dose-dependent toxicity in neur oblastoma cell line SH-SY5Y and were able to form higher molecular weight species in vitro. Upon delivery into nostrils of wild type mice, protein bands that might represent aggregation products of the exogenously applied human A-beta 42 were only observed in total brain homogenates from mice pre-treated with mannitol. By using TAMRA-labeled A-beta 42 peptides we demonstrated, that transport throughout the brain was achieved already 1 h after administration. FVB/N mice treated with A-beta 42 for 3 days were significantly impaired in the cue-retention condition of the fear conditioning task as compared to controls whereas A-beta-treated C57B6/J mice were impaired in the context condition. In the Morris water maze test, these mice also displayed a delayed learning performance, indicated by significantly longer time to find the platform. Those deficits were also seen for memory performance in the probe trial as measured by number of crossings of the former platform position and time spent in the goal quadrant. Conclusions Existing AD mouse models are of genetic origin and need prolonged housing time before onset of pathology. Our short-term treatment induced learning and memory deficits via exogenous application of A-beta peptides comparable to those observed for the transgenic animals. With the transnasal A-beta 42 treatment we present an approach to investigate purely A-beta related changes suitable as a model for symptoms of Alzheimer’s dementia (AD). Resulting behavioral deficits were indicative for familial type of Alzheimer’s disease as well as for the late onset variant.