Abstract
The ventral striatum (vStr) is anatomically interconnected with brain structures that exhibit prominent rhythmic activity, suggesting that oscillations in ventral striatal activity are potentially informative about systems-level interactions between these structures. However, rhythmic activity in ventral striatal neurons during behavior has only been characterized piecemeal, with individual studies focusing on a single cell type or frequency band. We performed a comprehensive analysis of (1) rhythmic activity in vStr neurons without reference to the local field potential, and (2) average as well as time-resolved spike-field relationships. Spike train rhythmicity tended to be limited to low frequencies such as delta and theta, whereas spike-field relationships were seen across a broad spectrum of frequencies, with about 90% of neurons showing spike-field locking to at least one rhythm. Using a novel time-resolved generalized linear model approach, we further show that the contribution of local field potential (LFP) phase to spike timing is dynamic over time, and enhanced by the inclusion of the LFP from the hippocampus – a new measure of inter-area coupling. These results provide a foundation for a more accurate interpretation of the ventral striatal LFP, suggest the possibility of an oscillatory taxonomy of ventral striatal neurons, and provide a starting point for understanding how rhythmic activity links cell-, circuit-, and systems-level phenomena in the ventral striatum.
Significance Statement Oscillations in neural activity are ubiquitous in the brain, readily accessible in the clinic and the lab, and shared by humans and animals to facilitate translational work. The ventral striatum (vStr) is a promising target structure for such a rhythmic activity perspective, not in the least because its local field potential (LFP) shows prominent task-related oscillations across a range of frequencies. However, recent work has shown that major components of the vStr LFP are in fact generated elsewhere in the brain, raising the question of how the LFP relates to local spiking activity. Unlike previous studies that focused on a specific cell type or frequency band of interest, we characterize rhythmic activity across a full range range of frequencies and cell types, and include novel analyses appropriate for a non-local LFP. Our results provide a foundation for more accurate interpretation of the vStr LFP and a starting point for an oscillatory taxonomy of vStr neurons.
