Abstract:
The incidence of suffering from Alzheimer's disease (AD) increases steadily with augmenting life expectancy in humans, generating an enormous burden for patients, their families and the national health system. Models used in previous publications covered only parts of the AD pathology. The work presented here combines transgenic technology and selective invasive methods to identify an improved animal model for the study of AD.
The suitability of animals to acquire task demands of behavioural models is an important issue that may bias measure of cognition. Therefore, several mouse strains were tested in different behavioural models. Especially C57BL/6 mice displayed constantly high performance in the T-Maze (T-CAT), Holeboard, Barnes Maze (BM) and in the Morris Water Maze (MWM). Insufficient performance in the original version of the object recognition task (ORT) was significantly ameliorated with modifications of set-up and test procedure. Against this background, four different C57BL/6 genotypes expressing different beta-amyloid protein (Abeta) niveau were tested in a longitudinal study in four behavioural models. Upon first testing, the nucleus basalis magnocellularis (NBM) of 50 % of the animals was selectively lesioned. All animals were again subjected to a second testing procedure. The results show Abeta-dependent impairment in the BM and MWM (spatial reference memory) in the first and second testing. Impairment was independent from NBM lesions. Learning performance in the ORT (working memory) was not influenced by Abeta (first and second testing). However, performance was impaired by NBM lesions in the second testing. In the T-CAT (spatial working memory), learning was unimpaired by Abeta when first tested. Interestingly, performance was reduced with Abeta and NBM lesions in the second testing. Additionally, standard therapy used in human patients (Donepezil) significantly improved learning performance in the T-CAT.
These results indicate that type (Abeta, NBM) and intensity (single- or double mutatnt) of pathology influences learning tasks in a different manner. However, the diverse symptomatology of human AD is predominantly reflected by NBM lesioned and Abeta overexpressing animals tested in the T-CAT. In conclusion, data presented here provide important contributions for the understanding of connections between pathology and learning, leading to an improvement of animal- and behavioural models in AD research.