hide
Free keywords:
-
Abstract:
Various states of wakefulness, perceptual ability, and locomotor activity are associated with different cortical states defined by local field potential (LFP) oscillatory content. The brainstem nucleus, Locus Coeruleus (LC), contributes to cortical state via noradrenergic projections to nearly the entire central nervous system. Electrical or optogenetic LC stimulation evokes increased high-frequency (HF, >20 Hz) cortical LFP oscillations and decreased low-frequency (LF, <12 Hz) oscillations in anesthetized and non-anesthetized experiments and evokes awakening in sleeping mice. The LC stimulation-evoked cortical state is due to highly synchronous whole-LC neuronal activation because about 1600 neurons are densely packed into only about 200 × 500 × 1000μm (in rats) and stimulation even at the lower end of the current range used in these studies (30–50μA) evokes spikes 400um from the stimulation site. These studies established the conceptual model that LC generates a single aroused cortical state. Recently, however, recordings of spontaneous LC single unit spiking demonstrated that pairs of LC neurons have sparse, yet structured time-averaged cross-correlations that are uncharacteristic of the en masse population event elicited by LC stimulation. It remains unknown whether spontaneous LC population activity consists of multi-cell ensembles or how LC ensemble activity evolves over time. Here, we used non-negative matrix factorization (NMF) to analyze large populations of simultaneously recorded LC single units in the urethane anesthetized rat. NMF, unlike traditional time-averaged pairwise correlations, detects the precise neuronal composition of LC ensembles and the evolution of their activity over time. We found that LC population dynamics consists of ensembles of co-active neurons with largely non-overlapping activation dynamics. We then characterized the relationship between LC ensemble activation dynamics and cortical state. We calculated cortical LFP (area 24a) band-limited power and spectrograms aligned to spontaneous activations of LC ensembles. Spontaneous activation of distinct LC ensembles was associated with a diverse pool of cortical states. Depending on which LC ensemble fired, we observed a diverse state space of increased HF and LF, decreased HF and LF, and opposing HF and LF power. Thus, LC is not simply a switch controlling a single arousal-associated cortical state.
