APPDutch transgenic (tg) mice develop cerebral amyloid angiopathy (CAA) that consists mainly of AbetaDutch40, with virtually no parenchymal amyloid plaques. To modulate cerebral amyloidosis, we crossbred APPDutch mice with either BACE1 tg mice to increase total AbetaDutch, or with G384A-mutated PS1 tg mice to elevate the ratio of AbetaDutch42 to AbetaDutch40. We analyzed all mice at 22 months of age. Compared to APPDutch mice, double-tg APPDutch/BACE1 mice revealed increased CAA mainly due to extensive vascular amyloid accumulation in the thalamus. In addition, they developed parenchymal amyloid in cortex and subiculum. In contrast, APPDutch/G384A-PS1 mice showed extensive, predominantly parenchymal amyloid throughout the entire brain, interestingly, even in the thalamus. The amyloid, composed largely of AbetaDutch42, was different compared to that in APPDutch/BACE1 mice which was composed mainly of AbetaDutch40. In summary, these mouse models reveal a broad variety and region-specificity of parenchymal versus vascular cerebral amyloid. This is partially explained by the absolute amount of neuronally produced AbetaDutch42 and AbetaDutch40 and ratio between the two. We conclude that the absolute levels of Abeta in combination with the ratio of Abeta42 to Abeta40 play a key role in determining the cerebral compartment and brain region in which Abeta is deposited.
2007 S. Karger AG, Basel