Structure and function of a neuronal amyloid implicated in memory persistence during aging

How a transient experience produces a persistent memory has remained unsolved. The physical substrate of memory has been extensively investigated at the neuronal network level, with the synapse established as the basic unit of memory. The encoding and maintenance of a new memory involve persistent changes in the number and shape of synapses. However, the molecules or processes implicated in these persistent synaptic changes have been largely unknown. Among such processes, the experience-dependent aggregation of a synaptic protein, the cytoplasmic polyadenylation element-binding (CPEB) protein, to a functional aggregated state was proposed as a physical substrate of long-lasting memories across species. While monomeric CPEB acts as a translation repressor, aggregated CPEB activates the translation of synaptic mRNAs encoding proteins involved in synapse formation and function. By using electron cryo-microscopy, it was previously found that Drosophila CPEB, known as Orb2, attains a hydrophilic amyloid state that enhances the translation of synaptic mRNA targets implicated in memory. The hydrophilic nature of Orb2 amyloid makes it more susceptible to environmental factors that can alter its structure. Hence, we hypothesize that the changes in the cellular environment in the aging brain could lead to alterations in the Orb2 amyloid structure that can contribute to age-related memory impairment (AMI). Consequently, here we investigated the molecular basis of memory decline in flies, as both humans and Drosophila experience AMI. Our findings reveal that the hydrophilic Orb2 amyloid can be modified during AMI. Specifically, we isolated distinct Orb2 amyloid polymorphs from the brains of young flies producing robust memory ("young polymorph"), adult flies with mild AMI ("adult polymorph"), and aged flies with severe AMI ("aged polymorph"). Although the three major polymorphs were present in all groups of flies, the relative abundance of the aged polymorph, which exhibits a decreased capacity to activate the translation of synaptic mRNAs implicated in memory, increased during aging. Therefore, we conclude that changes in the prevalence of a translationally active Orb2 amyloid polymorph, the young polymorph, towards a translationally inactive Orb2 amyloid polymorph during ageing, the aged polymorph, could be a central event contributing to AMI. The knowledge obtained provides insight into the molecular basis of memory decline and prompts us to reconsider the connection between amyloids and human diseases. Furthermore, it can lead to the exploration of similar pathways in mammalian systems.