Opposing responses of hippocampal theta oscillations to running and a forelimb-dominated sensorimotor behavior

Abstract

Hippocampal theta oscillations regulate the timing of neurons to support navigation, memory formation, and sensorimotor integration. Theta is modulated by running speed, breathing, whisking, and jumping and increases in tasks involving memory encoding or retrieval. The positive relationship between theta frequency and running speed is believed stabilize hippocampal representations of space amid movement variability. Here, we incorporated a novel string-pulling task to determine if established relationships between movement and theta hold when progress to a reward is determined by the length of string pulled. This task eliminates many speed-associated inputs, such vestibular, visual, and hindlimb information, and allows an unprecedented level of precision in the analysis of individual paw movements. Given that animals move the string a fixed length to acquire a reward, we predicted that the positive relationship between theta frequency and speed would hold. ApproachFive Sprague Dawley rats (4 mo.) were trained to continuously pull a string a fixed distance of 208 cm using an automated string-pulling system and run on a track for food reward. Local-field data was acquired from electrodes in dorsal CA1. ResultsRelationships between theta and movement speed were distinct during string pulling and running. While theta was robust in both conditions, frequency was significantly reduced during string-pulling and showed no speed-frequency coupling, unlike running. This difference could result from the conflict between hindlimb and forelimb signals, with only forelimb movement signaling advancement. Fine-grained analysis of paw movements during string-pulling (lift, advance, grasp, pull, push) revealed that theta power and frequency peaked during the contralateral paws downward push despite paw speed being low during this action. This suggests that theta frequency and power could respond to effort rather than purely kinematic information. Notably, running-associated theta may similarly reflect both speed and effort as most locomotor tasks conflate these variables. Finally, theta phase aligned from one reach-pull cycle to the next during the downward pull motion - the first action that directly advances the string forward. Since phase-locking has been associated with sensorimotor gating, synchrony at this point could reflect the gating of inputs that are the most causally relevant for reaching the reward, potentially facilitating integration of action-outcome signals for memory encoding and navigation. Taken together, these data support a dual-scale view of hippocampal processing and theta-band activity where macroscale theta activity requires suprathreshold sensory, vestibular, and proprioceptive drive and microscale theta remains sensitive to subsecond limb movements.