Memory loss in young APPswe/PS1dE9 mice and associated changes in brain metabolism analyzed using a 3-D voxel-based approach

Background: APPswe/PS1dE9 is a mouse model of Alzheimer's disease that starts to develop amyloid plaques by the age of 4 months. This study evaluated early alterations of spatial memory and brain glucose metabolism in this mouse strain. The latter biomarker is of great interest in translational studies, as it can be evaluated both in animals and humans. Methods: APPswe/PS1dE9-transgenic (Tg) and C57BL/6 J amyloid free control (WT) mice were studied at 4.5 months, i.e. at the very beginning of amyloid deposition. Spatial memory was evaluated using the Morris water maze (nTg = 14; nWT = 14). During the training (4 trials/day for 4 days) the latency and distance to reach platform (PF) were evaluated. For the memory retention test, the PF was removed and mice were allowed to navigate for 60 seconds. The time spent in the target quadrant, the distance moved, the number of crossing into PF zone, and the swim speed were evaluated (Ethovision videotracking system, Noldus). Glucose uptake was quantified in awake mice (nTg = 6; nWT = 7) by [14C]-2-deoxyglucose autoradiography of the whole brain (20-μm serial sections). 3D autoradiographic brain volumes were reconstructed using BrainRAT (freely available software, http://www.brainvisa.info) and statistical analyses were performed without any a priory hypothesis using SPM5. Cerebral amyloid deposits were examined in the same animals (BAM10 immunostaining and Congo red). Results: APPswe/PS1dE9 mice displayed significant spatial memory alterations that could be detected in training (increased distance and latencies to reach the PF) and retention sessions (reduced time spent in the quadrant). Despite these alterations, SPM analysis of glucose metabolism revealed no statistically significant hypometabolic voxels in Tg as compared to WT animals. On the contrary, hypermetabolic voxels were identified bilaterally. They were located in lateral olfactory tracts and in the somatosensory and motor parietal cortex (increases of 12.0%/13.7% and 13.9%/13.0% for the left and right sides of these structures). Conclusions: Although young Tg mice displayed spatial memory impairment, no obvious hypometabolic brain regions was detected. Strikingly, Tg mice even present with discrete areas of hypermetabolism in the somatosensory and motor parietal cortex and in olfactory areas. The fine relationship between these behavioral and metabolic changes remains to be elucidated.