Alzheimer's disease (AD) is a neurodegenerative disease, characterised by the
presence of amyloid plaques and neurofibrillary tangles, and is associated with a
progressive memory loss. It has been hypothesized that beta-amyloid (Aß)
accumulation and deposition play the central role in the pathogenesis of AD.
However, recent evidence suggests that the contribution of Aß plaques to cognitive
impairment of AD is very controversial. Genetic studies showed that mutations of
amyloid precursor protein (APP), presenilins (PS1 and PS2) and apolipoprotein s4
(apoE4) genes in familial forms of autosomal AD (FAD) resulted in increased
amyloid plaques in the brain. PDAPP and other APP transgenic mice have previously
been shown to exhibit age-related deposition of amyloid plaques, other AD-like
neuropathology and age-independent learning deficits prior to any plaque deposition.
A cross-sectional and a
longitudinal design were employed to address whether
PDAPP mice exhibit any plaque-related learning deficit. First, the cross-sectional
study indicated that PDAPP mice simultaneously displayed an early (plaqueindependent) and age-related learning impairment. Further analysis showed that the
age-related learning deficit was highly correlated with plaque burden in the
hippocampus of aged PDAPP mice, suggesting that amyloid plaques play a very
important role in memory loss of AD. Second, the longitudinal study showed that the
same PDAPP mice exhibited significant age-related learning deficits in trials to
criterion and learning capacity tasks when they aged. Interestingly, cued navigation
and object recognition in both cross-sectional and longitudinal studies were
unaffected, indicating normal sensorimotor function and recognition memory of
PDAPP mice. The longitudinal study further showed that the age-related learning
impairment was significantly correlated with significantly reduced size of field
potentials, suggesting important roles of amyloid plaques in disturbing synaptic
transmission and cognitive function.
Preclinical data indicated that A(3 immunotherapy is very effective to improve both
Aß-like neuropathology and cognitive impairment in APP transgenic mice. Using
active Aß immunisation, long-term (9 months: prevention study) and short-term months: treatment study) effects of Aß vaccines (AN-1792) on synaptic transmission
and spatial learning of PDAPP mice were investigated. Although neither of the two
studies showed that Aß vaccines had significant improvement on either synaptic
transmission or spatial learning of PDAPP mice, both long-term and short-term Ap
vaccinations effectively reduced the levels of total Aß. Thus, these two studies
indicated that active Aß immunotherapy might be very effective to improve AD-like
pathology, but ineffective to rescue the learning impairment in APP mice.
The γ- and ß-secretases have become the prime targets for pharmaceutical
intervention of AD. In this thesis, it was investigated, using a genetic and a
pharmacological method, whether inhibition of the γ-secretase has any effects on
synaptic transmission and spatial learning of APP transgenic mice. First, using
genetically modified APP mice, PSlcKO;APP mice, in which the PS1 gene is
conditionally inactivated in the forebrain, their spatial learning was studied.
Interestingly, young PSlcKO;APP mice exhibited normal spatial learning in the trials
to criterion and learning capacity tasks, whereas young APP transgenic mice were
impaired, suggesting a significant rescue effect of PS1 inactivation on spatial learning
at young age. However, although old PSlcKO;APP mice displayed normal learning in
the trials to criterion task, they exhibited significant impairment on a cuetask and a
learning capacity task, suggesting that long-term PS1 inactivation may only partially
rescue impaired learning in old APP animals. Thus, these findings raise the possibility
that massive accumulation of APP CTFs can cause learning deficits. Secondly, a
potent functional y-secretase inhibitor, ELN 44989 was administered to PDAPP mice
to investigate its chronic in vivo effects on AD-like pathology and spatial learning.
This study showed that ELN 44989 had no significant effect on Ap levels in the brain
and also showed no effect on impaired spatial learning of PDAPP mice. Future work
needs to examine whether more potent and selective γ-secretase inhibitor will show
any effect on improvement of neuropathology and memory loss in APP mice.
