Engram cell vulnerability in neurodegeneration: from mechanistic insights to therapeutic horizons.

Neurodegenerative diseases such as Alzheimer's disease (AD), Parkinson's disease (PD), and Huntington's disease (HD) share progressive memory impairment, yet their upstream pathologies differ. This review synthesizes evidence across these disorders to frame memory decline through engram ensembles and proposes an excitability-vulnerability tradeoff, in which the physiological requirements for engram function, including elevated excitability, synaptic plasticity, and coordinated network activity, also increase sensitivity to synaptic stress, network disruption, and inflammatory perturbations. We summarize disease-related pathological cascades. To date, AD has the most robust body of direct experimental evidence linking pathology to impaired engram function. Solid experimental findings demonstrate that amyloid-β (Aβ)/tau causes hippocampal engram inaccessibility in AD. For PD and HD, inferences drawn from synaptic and circuit dysfunction suggest that α-synuclein (α-syn) pathology and dopaminergic loss may disrupt striatal ensembles, and mutant huntingtin (mHTT) presumably impairs striatocortical coordination. We further integrate a shared amplification route in which microglial activation and complement-linked synaptic remodeling weaken memory-relevant connectivity and reduce cue-driven ensemble reinstatement. This mechanism is well validated in AD, while supporting evidence for PD and HD remains largely indirect. Finally, we outline a translational roadmap that integrates circuit-targeted delivery, engram-relevant clinical endpoints, and scalable neuromodulation, supported by human neuronal-glial molecular profiling to guide mechanism-informed combination strategies. Importantly, these multi-tiered therapeutic approaches act on the full spectrum of neural cells rather than only memory engram ensembles, and cognitive rescue arises from a multifactorial regulatory process involving both global neuroprotection and ensemble stabilization. We also systematically discuss key limitations of current engram research, including methodological heterogeneity in engram labelling tools, interspecies translational barriers, and substantial clinical obstacles for optogenetic and circuit-based therapeutic applications.