Examining simultaneous encoding of local and remote space across distinct hippocampal subnetworks

点击翻译按钮获取中文摘要

Project Summary/Abstract Current experiences, memories, and imagined futures all affect behavior. Evidence demonstrates that the hippocampus (HPC) can represent past locations, present position, and possible future trajectories. The HPC, however, is often referred to as a single structure capable of only one type of representation at a time, but the HPC can be divided into subregions along its dorsoventral axis. There is evidence for different functional roles of the dorsal and ventral HPC in spatial navigation, but recent gene expression studies suggest the dorsoventral axis can be divided into at least three subregions: dorsal, intermediate, and ventral. In contrast to these partitions, electrophysiology signatures evolve gradually along the HPC longitudinal axis, suggesting a gradual evolution of function. Whether the dorsal HPC (dHPC) and intermediate HPC (iHPC) behave as distinct networks as suggested by molecular subdivisions or as parts of a larger functional gradient is unknown. Both subregions are known to represent space, and studies in the dHPC have shown that representations of local and nonlocal space are temporally patterned around a regular oscillation in the field potential. This oscillation is present but phase-shifted in the iHPC, so the temporal patterning of local and nonlocal representations may differ from the patterning in the dHPC. This means that the HPC may simultaneously represent past, present, and future locations, and if so, the two subregions should be viewed as distinct networks that both contribute to encoding space. Through a combination of experimental and analytic methods, this project tests the hypothesis that dHPC and iHPC are best understood as distinct subregions that represent space noncoherently. We will test these hypotheses by recording simultaneously in the rat dHPC and iHPC during a spatial navigation task and studying the structure and temporal organization of spatial representations (Aim 1) as well as the characteristics of correlated activity across the two subregions using a model of network activity (Aim 2). This computational model serves as a complementary approach to understanding the nature of correlated activity in the HPC; it is completely agnostic to what neurons may encode and has no free parameters to adjust. By combining both experimental methods with cutting edge analytic techniques, the proposed aims have the potential to uncover previously unknown richness of representations in the HPC that will give further insight into how memories and simulations of the future are coordinated with present experiences. The simultaneous existence of these representations may serve as a neural substrate for relating past, present, and future experiences to one another across time.