Synaptic transmission is impaired prior to plaque formation in amyloid precursor protein-overexpressing mice without altering behaviorally-correlated sharp wave-ripple complexes.Neuroscience. 2009 Sep 15; 162(4):1081-90.N
One of the hallmarks of Alzheimer's disease is the accumulation of amyloid plaques in brains of affected patients. Several recent studies provided evidence that soluble oligomer forms of amyloid-beta (Abeta) rather than plaques determine cognitive decline. In vitro studies using artificial Abeta oligomer preparations suggest that such pathophysiology is caused by a specific impairment of synaptic function. We examined whether synaptic deficits occur before deposition of insoluble fibrillar Abeta by analyzing brain slices taken from young Tg2576 mice overexpressing mutant amyloid precursor protein. Excitatory synaptic transmission in the hippocampal CA1 region was strongly impaired before plaque development, suggesting a dissociation of an early synaptic impairment, probably caused by soluble oligomeric amyloid-beta, from subsequent plaque formation. At higher age neurotransmission was also decreased in wild type mice, paralleling a cognitive decline of normal aged animals. Memory formation in rats is accompanied by distinct hippocampal network oscillations. It has recently been shown that hippocampal gamma oscillations, a network correlate of exploratory behavior, are impaired in amyloid precursor protein (APP)-overexpressing mice. We determined whether sharp wave-ripple complexes, which contribute to memory consolidation during slow wave-sleep, are modified in Tg2576 mice. Interestingly, neither sharp waves nor superimposed ripples were changed at pre-plaque or plaque stages. During aging, however, there was a strong reduction of sharp wave frequency and ripple energy in wild type and APP-overexpressing animals. This indicates that the reported changes in network oscillations following APP-overexpression are specific for gamma oscillations, whereas aging has a more general effect on network properties. Taken together our data suggest that non-fibrillar forms of Abeta--possibly Abeta oligomers--specifically interfere with synaptic function in Tg2576, but do not globally alter memory-related network properties. We propose that mechanisms leading to Abeta-related cognitive decline are different from those related to aging.