Dysfunction of synaptic communication in cortical and hippocampal networks has been suggested as one of the neuropathological hallmarks of the early stages of Alzheimer's disease (AD). Also, several lines of evidence have linked disrupted levels of activity-regulated cytoskeletal associated protein (Arc), an immediate early gene product that plays a central role in synaptic plasticity, with AD "synaptopathy". The mapping of Arc expression patterns in brain networks has been extensively used as a marker of memory-relevant neuronal activity history. Here we evaluated basal and behavior-induced Arc expression in hippocampal networks of the 3xTg-AD mouse model of AD. The basal percentage of Arc-expressing cells in 10-month-old 3xTg-AD mice was higher than wild type in CA3 (4.88% versus 1.77% , respectively) but similar in CA1 (1.75% versus 2.75% ). Noteworthy, this difference was not observed at 3 months of age. Furthermore, although a Morris water maze test probe induced a steep (∼4-fold) increment in the percentage of Arc+ cells in the CA3 region of the 10-month-old wild-type group, no such increment was observed in age-matched 3xTg-AD, whereas the amount of Arc+ cells in CA1 increased in both groups. Further, we detected that CA3 neurons with amyloid-β were much more likely to express Arc protein under basal conditions. We propose that in 3xTg-AD mice, intraneuronal amyloid-β expression in CA3 could increase unspecific neuronal activation and subsequent Arc protein expression, which might impair further memory-stabilizing processes.

Spatial water maze (WM) overtraining induces hippocampal mossy fiber (MF) expansion, and it has been suggested that spatial pattern separation depends on the MF pathway. We hypothesized that WM experience inducing MF expansion in rats would improve spatial pattern separation in the hippocampal network. We first tested this by using the the delayed non-matching to place task (DNMP), in animals that had been previously trained on the water maze (WM) and found that these animals, as well as animals treated as swim controls (SC), performed better than home cage control animals the DNMP task. The “catFISH” imaging method provided neurophysiological evidence that hippocampal pattern separation improved in animals treated as SC, and this improvement was even clearer in animals that experienced the WM training. Moreover, these behavioral treatments also enhance network reliability and improve partial pattern separation in CA1 and pattern completion in CA3. By measuring the area occupied by synaptophysin staining in both the stratum oriens and the stratun lucidum of the distal CA3, we found evidence of structural synaptic plasticity that likely includes MF expansion. Finally, the measures of hippocampal network coding obtained with catFISH correlate significantly with the increased density of synaptophysin staining, strongly suggesting that structural synaptic plasticity in the hippocampus induced by the WM and SC experience is related to the improvement of spatial information processing in the hippocampus.