ABSTRACT
Mammalian hippocampal circuits undergo extensive remodeling through adult neurogenesis. While this process has been widely studied, the specific contribution of adult-born granule cells (aGCs) to spatial operations in the hippocampus remains unknown. Here, we show that optogenetic activation of 4-week-old (young) aGCs in free-foraging mice produces a non-reversible reconfiguration of spatial maps in proximal CA3 while rarely evoking neural activity. Stimulation of the same neuronal cohort on subsequent days recruits CA3 neurons with increased efficacy but fails to induce further remapping. In contrast, stimulation of 8-week-old (mature) aGCs can reliably activate CA3 cells but produces no alterations in spatial maps. Our results reveal a unique role of young aGCs in remodeling CA3 representations, a potential that can be depleted and is lost with maturation. This ability could contribute to generate orthogonalized downstream codes supporting pattern separation.
Subject(s)
Neural Stem Cells , Humans , Mice , Animals , Hippocampus/physiology , Neurons/physiology , Brain , Neurogenesis/physiology , Dentate Gyrus/physiology , MammalsABSTRACT
Spatial navigation relies on visual landmarks as well as on self-motion information. In familiar environments, both place and grid cells maintain their firing fields in darkness, suggesting that they continuously receive information about locomotion speed required for path integration. Consistently, "speed cells" have been previously identified in the hippocampal formation and characterized in detail in the medial entorhinal cortex. Here we investigated speed-correlated firing in the hippocampus. We show that CA1 has speed cells that are stable across contexts, position in space, and time. Moreover, their speed-correlated firing occurs within theta cycles, independently of theta frequency. Interestingly, a physiological classification of cell types reveals that all CA1 speed cells are inhibitory. In fact, while speed modulates pyramidal cell activity, only the firing rate of interneurons can accurately predict locomotion speed on a sub-second timescale. These findings shed light on network models of navigation.