'Mathematical microscope' reveals novel, energy-efficient mechanism of working memory that works even during sleep

Working memory can be maintained with lower metabolic cost even during sleep, according to UCLA Health researchers studying cortico-entorhinal circuitry. The team proposes a “mathematical microscope” based on a simplified model of interactions between neocortex input and entorhinal cortex recurrent connections. Using in vivo measurements with identified entorhinal neurons and whole-cell patch clamp recordings, the researchers test quantitative predictions of how the system behaves when cortical activity oscillates between on and off states. The model predicts and the experiments support two memory states: spontaneous persistent activity and a newly highlighted spontaneous persistent inactivity state. Together they more than double memory capacity while cutting metabolic energy cost by about half, offering a mechanistic target in brain regions where Alzheimer’s disease begins.