Figure 1

Waxing-and-waning optogenetic stimulation of thalamic reticular nucleus (TRN) parvalbumin (PV) neurons elicits cortical electroencephalographic (EEG) events indistinguishable from spontaneously occurring sleep spindles. (A) TRN-PV neurons were bilaterally transduced with AAV-ChR2-EYFP and optogenetically stimulated. Coronal schematic representation of TRN image credit: Allen Institute, adapted from The Allen Mouse Brain Atlas (Available from: http://mouse.brain-map.org)⁷¹. (B) Representative cresyl violet staining confirmation of optogenetic fiber placement (red) in TRN. Staining did not show cellular damage due to laser application, which was consistent across all cases evaluated. Abbreviations: AV, anteroventral thalamic nucleus; AM, anteromedial thalamic nucleus; fi, fimbria of hippocampus; ic, internal capsule; TRN, thalamic reticular nucleus. Scale bar: 250 µm. (C,D) Confirmation of viral transduction efficiency and selectivity. AAV-GFP (green) was bilaterally injected into TRN of 2 PV-tdTomato (red) mice. Three coronal sections per animal representing rostral, medial, and caudal TRN were bilaterally analyzed for AAV-GFP/PV-tdTomato co-localization. 83 ± 3% of AAV-GFP transduced neurons co-expressed td-Tomato, suggesting high selectivity. 62 ± 2% of TRN-PV-tdTomato neurons were transduced by AAV-GFP, indicating high efficiency. Scale bars: 500 µm (top), 25 µm (bottom). (E–H) Spindles were detected as brief elevations in 10–15 Hz oscillatory activity in cortical EEG using a custom-built MATLAB script (see Methods). Events were counted as spindles if the root mean square (RMS) transform of band pass filtered EEG data exceeded an upper threshold (top red line, 3.5x mean RMS power) and the duration was >0.5 s as determined by crossing of a lower threshold set at 1.2x mean RMS power (bottom red line). Panel E shows an example of a spontaneously occurring (natural) spindle, while the panels F, G, & H show detected events elicited by optogenetic stimulation of TRN-PV neurons during non-rapid-eye-movement (NREM) sleep. We utilized a single pulse paradigm (F,G) and a more complex waxing-and-waning stimulation protocol (H). While both paradigms resulted in an elevation of detected spindle-like events, the events generated by the waxing-and-waning protocol were more morphologically similar to physiological spindles. The single pulse paradigm at times generated sharp EEG spike-like events (denoted by red arrows). (I,J) To characterize the single pulse laser paradigm vs. the waxing-and-waning paradigm, laser was repeatedly applied every 10 s regardless of animal’s state (without NREM gate) during the 3-h recordings. The proportion of optogenetically induced spindles in each state, including both spindle-like and spike-like spindles, was calculated by dividing the number of induced spindles in a specific state by a total number of induced spindles from all states (the inner circles of the charts). A higher percentage of induced spindles in wakefulness was observed with the single pulse laser paradigm compared to the waxing-and-waning paradigm (I,J, orange). The waxing-and-waning paradigm showed higher efficiency in eliciting spindles during NREM sleep compared to wakefulness (J, blue and orange). The proportion of spindle-like spindles and spike-like spindles was also calculated for each state and shown in the outer circles of the charts. The waxing-and-waning paradigm reduced the sharp spike-like EEG events compared to the single pulse paradigm.