| Summary: | Sleep spindle oscillations are a hallmark of non-REM (NREM) sleep in mammals. These discrete events appear as 0.5- 3 second periods of 8-15 Hz oscillations in the thalamocortical system, and have a characteristic waxing-waning shape in cortical field potentials. The spindle 'pacemaker' rhythm is generated by interactions between cells of the thalamic reticular nucleus (TRN) and thalamocortical (TC) cells, the latter of which relay the rhythm to the neocortex. The mechanisms behind the generation of the thalamic spindle rhythm have been subject to much scrutiny and are now fairly well understood. By comparison, much less is known about the nature of cortical activity during spindles. In this thesis, I set out to characterize dynamics in the thalamocortical system over the time scale of individual spindle events, using in vivo electrophysiological recordings from naturally sleeping rats. (1) Using simultaneous recordings from TRN and the medial prefrontal cortex (mPFC), I found a robust acceleration of the spindle rhythm, both in the cortical LFP and in the spike trains of single TRN units. Spindle acceleration was accompanied by a precession of TRN spike phases relative to the mPFC LFP oscillation. I identified two distinct subpopulations of putative pyramidal (Py) cells: 'Early' and 'Late' cells, which fired at opposite phases of local spindles and had markedly different overall mean firing rates. (2) I made use of a pre-existing dataset to compare cortical activity during spindles in mPFC and posterior parietal cortex (PPC), in naturally sleeping rats. In deep-layer PPC recordings, spindles were associated with a distinct 5-8 Hz oscillation. Furthermore, spindles detected in either cortical area were accompanied by the emergence of 5-8 Hz synchrony between the two areas. (3) Taking advantage of hippocampal recordings in the same dataset used in (2), I linked the spiking of single units during spindles to their coupling with two other NREM oscillations: cortical slow waves, and hippocampal ripples. In mPFC and PPC, the spiking of many cells was temporally coupled with ripples, and strong coupling with slow oscillations was associated with elevated firing rates during spindles. Finally, by exploiting recordings of awake behaviour in this dataset, I found that theta-modulated mPFC cells were preferentially synchronized with PPC in the 5-8 Hz band during spindles.
|
|---|
