Hippocampal information topology breaks down in a mouse model of Alzheimer's disease

Hippocampal spatial coding depends on coordination among neuronal assemblies, yet how network topology organises information processing across these assemblies, and how this is disrupted in disease, remain unknown. We apply Partial Information Decomposition to CA1 calcium imaging from young and aged wild-type and 5xFAD mice, quantifying redundant and synergistic information sharing within and between assemblies. In healthy CA1, between-assembly pairs carried more joint spatial information than within-assembly pairs, and this surplus was synergistic, establishing network topology as an organising principle of spatial coding. In aged 5xFAD CA1 this topological organisation broke down through two distinct routes: redundancy lost its topology dependence as modular assembly boundaries dissolved, and synergy lost context sensitivity during novel exploration, with the breakdown greatest where ageing and the 5xFAD genotype coincided. This functional decline was also accompanied by topological effects in the functional connectivity, where the genotype-age interaction resulted in reduced modularity, weighted clustering and small-worldness. Community-level emergence revealed a complementary cross-scale shift toward higher-order integration during ageing, which was reversed by the genotype-age interaction. We isolate the compounding effect of ageing in Alzheimer's disease as the driver of disruption in information processing and functional connectivity across neuronal assemblies in the mouse hippocampus.