Parallel Arousal Pathways in the Lateral Hypothalamus

Expression of hM3Dq in the LH. A, Schematic of the AAV used. B, AAV-injected Orexin-tTA mice exhibited widespread neuronal expression of hM3Dq-mCherry despite the presence of DOX in the chow. C, Despite the absence of tTA, Gad2-IRES-Cre;R26R-EYFP mice injected with AAV-TetO-hM3Dq-mCherry exhibited a pattern of hM3Dq-mCherry expression that was similar to DOX-fed Orexin-tTA mice. Scale bar in B, C = 250 µm. Numbers in the upper left in C denote approximate AP distance from bregma (mm). D, Photomicrographs showing GAD2-EYFP⁺ neurons lateral to the fornix (Fx, left), hM3Dq-mCherry transfected neurons (center) and double-labeled neurons in yellow after merging both images (right). Scale bar = 100 µm. M and L denote medial and lateral. E, Photomicrographs showing ADA⁺ neurons in TMN (left), hM3Dq-mCherry transfected neurons (center), and merged image (right) showing the absence of double-labeled neurons. Scale bar = 50 µm.

LH activation induces continuous wakefulness for at least 4 h. Hourly percentage of time spent in W (A), NREM (B), and REM sleep (C) and hourly bout duration for W (A’) and NREM (B’) after injection of either CNO (red) or SAL (black). Asterisks denote significant difference between CNO and SAL treatment for the respective hour when a significant treatment × time interaction was found (p ≤ 0.047, n = 9; two-way repeated measures ANOVA, Bonferroni post hoc t test). B’’, No significant change in NREM δ power was observed after the prolonged W induced by CNO injection (n.s., n = 6, two-way repeated measures ANOVA). Arrow at bottom left of abscissa indicates time of dosing at ZT12.

Characterization of CNO-induced wakefulness. A, Time-weighted frequency histograms showing the proportion of wake bouts of each duration (WBD) relative to the total amount of wakefulness from ZT12 to ZT17 after CNO or SAL treatment in Gad2-IRES-Cre;R26R-EYFP mice. Asterisks denote significant differences (p < 0.04, n = 9, t test with Bonferroni correction for multiple comparisons). B, Hourly percentage of time spent in W for naïve WT mice injected with CNO or SAL during the inactive phase. C, HθWP in the dark period after CNO or SAL dosing of Gad2-IRES-Cre;R26R-EYFP mice at ZT12. Asterisk denotes a significant difference for the respective hour when a significant treatment × time interaction was found (p = 0.046, n = 9; two-way repeated measures ANOVA, Bonferroni post hoc t test). D, Normalized EEG power spectrum during W averaged from ZT13 to ZT15 after dosing of Gad2-IRES-Cre;R26R-EYFP mice with CNO (red) or SAL (black) at ZT12. E, Mean power for EEG spectral bands of Gad2-IRES-Cre;R26R-EYFP mice after dosing with either CNO or SAL, normalized by baseline. Asterisk denotes a significant difference for the respective band when ANOVA revealed a significant treatment × power band (p = 0.002, n = 9; two-way repeated measures ANOVA, Bonferroni post hoc t test). F, Percentage of time mice spent moving at different speeds from ZT12 to ZT14 during W. G, Example of simultaneous recordings of mean speed per epoch (black trace) and scale-adjusted HθWP (green trace).

Hcrt signaling is not required for LH activation-induced arousal. Upper panels show hourly percentage of time spent in W (A), NREM (B), and REM sleep (C) and lower panels (A’–C’) show corresponding mean hourly bout duration for each state after injection of either VEH or ALM during the inactive phase at ZT4 followed by either SAL or CNO at ZT5. Arrows at the bottom left of the abscissa in A’–C’ indicate the times of dosing. Symbols above each time point denote significant hourly differences between the indicated treatments according to the legends in panels A, A’ when a significant treatment × time interaction was found (p < 0.001, n = 10, two-way repeated measures ANOVA, Bonferroni post hoc t test).

LH activation at ZT5 increases spectral power in the Hθ and γ bands of the EEG during W. A, Normalized EEG power spectrum during W averaged from ZT6 to ZT8 for the different experimental treatments. B, Mean power in the spectral bands of the EEG after treatment, normalized to baseline. C, Time-weighted frequency histograms showing the proportion of wake bouts of each duration (WBD) relative to the total amount of wakefulness from ZT5 to ZT10 for the different experimental treatments. D, HθWP for the different experimental treatments. E, NREM δ power for all treatments. The discontinuity in the VEH-CNO line between ZT5 and ZT8 is because there were no sleep bouts during ZT6–ZT7. Horizontal bar denotes a significant effect of drug treatment and symbol above it indicates that the ALM-CNO versus VEH-CNO comparison was significantly different over this period (p = 0.024, n = 10, two-way repeated measures ANOVA, Bonferroni post hoc t test). Arrows at the bottom left of the abscissa in D, E indicate the times of dosing for VEH or ALM at ZT4 and for SAL or CNO at ZT5. Symbols in B–D denote significant differences between the indicated treatments (p < 0.05) according to the legends in C, E when ANOVA revealed significant interaction between drug treatment and power band.

Sparse activation of Hcrt and inhibitory neurons after CNO injection in Gad2-IRES-Cre;R26R-EYFP mice. A, Microphotographs showing hM3Dq-mCherry labeling in red (left), Hcrt neurons in green (center), and the merged images (right) in Gad2-IRES-Cre;R26R-EYFP mice transfected with AAV-TetO-hM3Dq-mCherry. White arrowheads denote Hcrt neurons coexpressing hM3Dq-mCherry. Scale bar = 100 µm. B, Microphotographs showing Hcrt immunostaining (left), c-FOS staining (center), and the merged images (right) in AAV-injected Gad2-IRES-Cre;R26R-EYFP mice. White arrowheads denote Hcrt neurons that do not express c-FOS. Scale bar = 50 µm. C, Photomicrographs of DAB immunohistochemistry of transfected neurons in LH (brown) and c-FOS expression (black nuclei) after SAL (upper panel) or CNO (lower panel) injection. D, Photomicrographs of GFP-labeled LH inhibitory neurons (green cells) and c-FOS expression (black nuclei) in LH after SAL injection (upper panel) or CNO injection (lower panel). All hypothalamic cells lateral to the dotted white line were included in the Gad2/c-FOS counts. Scale bar = 50 μm. E, GFP-labeled inhibitory neurons in the nRT (green cells) and c-FOS expression (sparse black nuclei) after SAL injection (upper panel) or CNO injection (lower panel). Scale bar = 100 µm (same scale as C).

c-FOS activity in the SuM does not increase after CNO treatment. A, c-FOS expression in the mammillary region after treatment with SAL followed by 90 min of SD. B, c-FOS expression in the SuM after CNO treatment showing similar c-FOS expression than SAL treatment. C, mCherry immunohistochemistry for the region depicted in B showing hM3Dq expression only in the lateral regions and almost no labeling in the medial aspect of SuM. Black dotted lines indicate the approximate region used for counting c-FOS in the SuM. LM, lateral mammillary nucleus; MM, medial mammillary nucleus. Scale bar = 250 μm.