We used male transgenic orexin-Cre C57BL/6 mice (8∼12 weeks old, weight 25–30 g) with Cre recombinase and EGFP under the prepro-orexin promoter, a gift from Dr. Akihiro Yamanaka, Nagoya University, Japan (Inutsuka et al., 2014). All experimental procedures involving animals were approved by the Institutional Animal Care and Use Committee, University of California, San Francisco. Mice were maintained in a strictly controlled environment with ad libitum access to food and water. The light cycle starts at 6:00 a.m. and ends at 6:00 p.m. and the dark cycle runs from 6:00 p.m. to 6:00 a.m. The temperature was controlled between 20 and 22^°C. Altogether 72 mice were used in the experiments for imaging, electrophysiology, and optogenetic behavior assays.

Three different Cre-dependent viruses, AAV-Syn-DIO-ChrimsonR-tdTomato (62723-AAV5; Addgene), AAV-EF1α-DIO-hChR2(H134R)-mCherry-WPRE-HGHpA (20297-AAVretro; Addgene), and AAV-hSyn-DIO-mCherry (50459-AAV8; Addgene) were used in this study. Hereafter we refer to them as AAV5-DIO-ChrimsonR-tdT, AAVretro-DIO-ChR2-mCherry, and AAV8-DIO-mCherry. The titers of virus were 5 × 10¹² vg/mL (AAV5-DIO-ChrimsonR-tdT), 1 × 10¹³ vg/mL (AAVretro-DIO-ChR2-mCherry and AAV8-DIO-mCherry). The viruses were aliquoted and stored at –80^°C.

Intracranial injections were performed under balanced anesthesia using a robot stereotaxic instrument (Neurostar, Germany). Ketamine (80 mg/kg, i.p. injection), meloxifen (5 mg/kg, s.c. injection), ampicillin (10–20 mg/kg, s.c. injection), buprenorphine (0.1 mg/kg, s.c. injection), 0.25% bupivacaine (0.025 ml, s.c. injection) and isoflurane (Henry Schein Animal Health) were used. A 5 μl glass syringe with a 33 gauge needle (syringe: 87,943, needle: 7762–06; Hamilton Company) was used for the injections, and 0.5 μl virus was delivered over a 10 min (0.05 μl/min) period. The needle was left in situ for 10 min and slowly removed. The skin was stapled after surgery.

The injection coordinates were as follows: lateral hypothalamus (AP, –1.46 mm, ML, ± 0.95 mm, and DV, 5.1–5.0 mm), medial amygdala nucleus (AP, –1.94 mm, ML, ± 2.3 mm, and DV, 4.76–4.66 mm), medial preoptic area (AP, –0.1 mm, ML, ± 0.5 mm, and DV, 5.0–4.9 mm), lateral habenula (AP, –1.7 mm, ML, ± 0.44 mm, and DV, 3.2–3.1 mm), periaqueductal gray (AP, –4 mm, ML, ± 0.3 mm, and DV, 3.2–3.1 mm), and locus coeruleus (AP, –5.34 mm, ML, ± 0.85 mm, and DV, 4.5–4.4 mm).

For optogenetic stimulation experiments, all mice were implanted with bilateral optical fibers (400 μm Core, 0.39 N.A.; R-FOC-BL400C39NA; RWD Life Science) into the lateral hypothalamus (AP, –1.46 mm, ML, ± 0.95 mm, and DV, 5.0 mm).

One week after receiving bilateral stereotaxic viral injections, the animals were subjected to optical fiber implantation and electroencephalogram (EEG)/electromyography (EMG) device implantation (Pinnacle Technologies). A 23-gauge surgical needle was used to drill four guide holes, including two frontal cortical area (AP, 1 mm, ML, ± 1.25 mm) and two parietal area (AP, –3 mm, ML, ± 2.5 mm). Four screws with wires were placed into the skull through the holes. The wires were then soldered onto a six-pin connector EEG/EMG headstage. EMG leads were placed into the neck muscle. The headstage was secured with dental cement (C&B Metabond^® Quick Adhesive Cement System, Parkell) onto the skull. Behavioral experiments were conducted 2 weeks later after surgery to allow for sufficient recovery. Each time before the behavioral tests, mice were anesthetized with 3% isoflurane for 1 min in an induction chamber followed by connecting with the EEG device.

The mice were transcardially perfused 3–4 weeks after virus injection with 1X cold PBS followed by a cold solution of 4% paraformaldehyde (PFA, Electron Microscopy Services) in PBS. Brains were removed and post-fixed in 4% PFA for 6 h at 4°C before being transferred to 30% sucrose for at least 2 days of dehydration until the tissue stayed afloat. A series of 50 μm coronal/sagittal brain slices were sliced using a cryostat (Leica CM3050S). Fixed brain slices were used for immunohistochemistry staining.

Slices were blocked with blocking solution (5% donkey serum, 3% BSA, and 0.3% Triton-X100 in PBS) for 1 h at room temperature, followed by overnight incubation with primary antibodies at 4^°C for 18 h. After the brain slices were washed three times with PBS, 10 min each time, the slices were incubated with secondary antibodies for 2 h at room temperature followed by 3 washes. The slices were covered with DAPI Fluoromount-G mounting medium (0100-20, SouthernBiotech).

The antibodies were diluted in blocking solution as follows.

mCherry. Primary, chicken anti-mCherry (1:400, NBP2-25158, Novus Biologicals); rabbit anti-mCherry (1:100, ab167453, Abcam). Secondary, Cy3-conjugated AffiniPure goat anti-chicken IgY++ (1:400, 103-165-155, Jackson Labs); goat anti-rabbit Cy3 (1:100, 111-165-003, Jackson Labs).

cFos. Primary, rabbit anti-cFos (1:100, 2250S, cell signaling). Secondary, Alexa Fluor 488-conjugated donkey anti-rabbit IgG (H + L) (1:400, 711-546-152, Jackson Labs).

Orexin. Primary, mouse anti-Orexin-A (1:100, sc-80263, Santa Cruz). Secondary, Alexa Fluor 488-conjugated donkey anti-mouse IgG (H + L) (1:400, 715-545-151, Jackson Labs).

Confocal images were taken with the Leica TCS SP8. Light sheet images were taken with ZEISS Lightsheet Z.1. Lightsheet images were collected using EC Plan-Neofluar 5x NA 0.16 objective. The pixels are 1,920 × 1,920, and the image size is 4867.7 μm × 4867.7 μm. Tile scans are assisted with the setup of a tiling experiment with Zen (ZEISS) for Lightsheet Z.1 software. Tiles were stitched and 3D video was generated by Imaris software (OXFORD INSTRUMENTS). ImageJ was used to visualize the 3D images.

Animal was transcardially perfused using 15 ml ice-cold 1X PBS (3 ml/min for 5 min), followed by 20 ml of PFA-Monomer hydrogel solution (1.3% acrylamide, 0.0125% Bis-acrylamide, 4% PFA, 0.25% VA-044 initiator (w/v) in 1X PBS solution (capture-clarity.org/optimized-clarity). The brain was cut into two hemispheres and placed into hydrogel solution for 3 days for monomer diffusion at 4^°C in the cold room. Afterward, the hemispheres were transferred to a 50 ml conical tube, which was placed in a vacuum chamber. The tube was shaken gently at 37^°C for 24 h for polymerization. The hemispheres were transferred into 200 mM NaOH-Boric buffer (pH 8.5), 8% SDS for 6–12 h with gentle shaking to remove residual PFA. Next, they were cleared in 100 mM Tris-Boric buffer (pH 8.5), 8% SDS with gentle stirring at 37^°C over the course of 21 days to remove the phospholipids. The cleared hemispheres were then washed in 1X PBST (0.1% Triton-X100) for 24 h to remove SDS. Thereafter, one hemisphere was stained with the primary antibody for 6 days at 37°C (2 mL of PBST + rabbit anti-mCherry, 1:100 dilution). After washing in 1X PBST, the hemisphere was stained in 1:100 dilution of goat anti-rabbit Cy3 at 37^°C for 5 days. And then washed the hemispheres extensively with 1X PBST at 37^°C for 3 days with solution change twice a day. Before imaging, the hemispheres were placed in the EasyIndex solution (Lifecanvas, refractive index—1.46) for 3 days at room temperature.

AAVretro-DIO-ChR2-mCherry was injected into the lateral hypothalamus of orexin-Cre mice 3–4 weeks prior to electrophysiological experiments. For brain slice preparation, mice were anesthetized with isoflurane and transcardially perfused with NMDG-HEPES: 93 mM NMDG, 93 mM HCl, 2.5 mM KCl, 1.2 mM NaH₂PO₄, 30 mM NaHCO₃, 20 mM HEPES, 25 mM Glucose, 5 mM sodium ascorbate, 2 mM Thiourea, 3 mM sodium pyruvate, 10 mM MgSO4.7 H₂O, 0.5 mM CaCl2.2H₂O (pH 7.4 by HCl, 300–310 mOsm), bubbled with carbogen (95% O2/5% CO2). The coronal sections were cut into 300 μm thickness sections by a microtome (Leica VT1000). The brain slices were recovered in the cutting solution for 15 min at 34^°C and in the artificial cerebrospinal fluid (aCSF) recovery solution for 45 min at room temperature before recording. The aCSF solution contained: 125 mM NaCl, 26 mM NaHCO3, 1.25 mM NaH₂PO₄, 2.5 mM KCl, 1 mM MgCl2, 2 mM CaCl₂, 12.4 mM glucose, 300–310 mOsm, and bubbled with carbogen. The final pH was 7.4.

Loose patch recordings were made on mCherry-positive cell bodies identified under the fluorescence microscope. The internal solution within whole-cell recording pipettes (3–5 MΩ) contained: 139.95 mM KGluc, 8.5 mM KCl, 10 mM HEPES, 0.1 mM EGTA, 4 mM MgATP2, 0.3 mM NaGTP (pH 7.3–7.4, 300–310 mOsm). During the recordings, the neurons were perfused with the carbogen-bubbled aCSF solution with a peristaltic pump. The recordings were performed using a Multiclamp 700B amplifier (Molecular Devices). The data were acquired and analyzed using pClamp and ClampEx software (Molecular Devices). For light stimulation, 470 nm LED (M470F3, Thorlabs) was used to deliver light pulses (1, 5, 10, 20 Hz, 10 ms pulse width). The optical fiber was directed right above the patched cell with a micromanipulator (Sutter Instrument) under the recording microscope.

The animals were anesthetized with 3% isoflurane for 1 min in an induction chamber and then quickly attached with optical fibers. Then the mouse was placed in a clear isoflurane chamber with optical fibers exiting through a sealed port. The chamber was then equilibrated with 0.75% isoflurane for 10 min, before light pulse (473 nm, 20 Hz, 20 ms pulse width, 30 s, Aurora-200, Newdoon) was applied to the animals. After the 30 s stimulation, isoflurane was continued for another 1 min before it was switched to 100% O₂ to let the mouse recover. The time from starting the light pulse to the point when the mouse started to wake up (moving, kicking, or tail rising) was recorded as the latency to wake. If animals had no response to the light stimulations, the time from laser on to when the animals woke up after the termination of isoflurane was recorded as latency to wake. All anesthesia chambers were placed on a temperature-controlled heating pad to automatically regulate the temperature between 35 and 37^°C. The trials were repeated three times in the same day from 9:00 to 15:00, and again repeated two more times with 3-day intervals. Mice with AAV8-DIO-mCherry injections were used as control and received the same treatment of optical stimulation. All trials were age-matched with similar animals.

The experimental animals received 2% isoflurane for 30 min in the induction chamber. Afterward, the animals were quickly attached with optical fibers and then were treated with or without optical stimulation (20 Hz, 20 ms pulse width, 1 s on, 1 s off) in a clear acrylic chamber open to the room air while placed on their backs to test how quickly the righting reflex can return (RoRR). The RoRR, when the animal turned around and stood on all four paws, was used as the indicator for the emergence from anesthesia. The emergence time from turning off the isoflurane to the RoRR was recorded for analysis. The trials were performed from 9:00 to 15:00 and repeated three times at 3-day intervals. The isoflurane induction chamber was placed on a temperature-controlled heating pad and a temperature probe connected to a temperature controller was placed underneath the animal body to automatically regulate the temperature between 35 and 37^°C. Mice with AAV8-DIO-mCherry injections were used as control and received the same treatments. All trials were age-matched with similar animals.

The animals were anesthetized in the induction chamber with 2% isoflurane for 1 min to facilitate optical fiber attachment. Then the animals went through 5 min of optical stimulation with or without the laser (20 Hz, 20 ms pulse width, 1 s on, 1 s off), followed by the hot plate test. An acrylic container was used as an enclosure to prevent the animal from escaping. The temperature used in the hot plate test was 55^°C. The animals were placed onto the hot plate while continuing the optical stimulation until the first time when animals showed licking, fanning, or jumping, at which point the animal was immediately removed from the hot plate and the paw withdrawal latency was recorded. The animal was removed if there was no response within 45 s. The trials were repeated three times in the same day from 9:00 to 15:00, with 60 min intervals, and again repeated two more times with 3-day intervals. Mice with AAV8-DIO-mCherry injections were used as control and received the same treatments. All trials were age matched with similar animals.

Data analysis was done by experimenters blinded to experimental conditions. Paired t-test was used to analyze the significance between the same group of mice with and without optogenetic stimulation. Unpaired t-test was used to analyze the significance of the data from two groups. Data are shown as mean ± s.e. of the mean (s.e.m.). Statistical significance was set at P < 0.05. All data were analyzed using Prism 7.0 (GraphPad Software Inc. San Diego, CA, United States).

To express opsins in orexin neurons specifically, AAV5-DIO-ChrimsonR-tdT or AAVretro-DIO-ChR2-mCherry was injected into orexin-Cre mouse brains (Figure 1A). We stained the brain slices with antibodies recognizing orexin and Cre, which showed that Cre is colocalized with orexin in the lateral hypothalamus in orexin-Cre mice but no Cre was found in wildtype (WT) mice, suggesting that Cre was exclusively expressed in orexin neurons with colocalization of 92.87 ± 0.6974% (n = 5) (Figure 1B). Next, we injected AAV5-DIO-ChrimsonR-tdT into the lateral hypothalamus and AAVretro-DIO-ChR2-mCherry into the locus coeruleus. The brain slice imaging result showed 93.7 ± 2.456% (n = 5) colocalization of tdTomato with orexin in AAV5 group (Figures 1C,E), and 90.07 ± 1.593% (n = 5) colocalization of mCherry with orexin in AAVretro group in the lateral hypothalamus (Figures 1D,E). Interestingly, not all orexin neurons were mCherry positive, but all mCherry positive neurons were also orexin positive (Figure 1D), suggesting that not all of the orexin neurons in the lateral hypothalamus project to locus coeruleus. It’s noticeable that the processes are clearly visible with mCherry staining, but barely with orexin staining (Figure 1D), suggesting AAV expression of opsin tagged with fluorescence marker provided an excellent tool to label the neuronal projections. Lateral hypothalamus injections of both AAVretro-DIO-ChR2-mCherry and AAV5-DIO-ChrimsonR-tdT can produce similarly vigorous expression throughout the mouse brain (Figure 1F). In contrast, there was little signal in WT mice that received either AAV5 or AAVretro injections, similar to orexin-Cre mice with saline injections (Figure 1F).

AAV5-DIO-ChrimsonR-tdT was injected into the lateral hypothalamus of orexin-Cre mice, and sagittal (Figures 2A–D) and coronal (Figure 3) slices were imaged. Overall, strong tdTomato signals were observed in many subcortical regions, including the septum, basal forebrain, amygdala, hypothalamus, the superficial regions of thalamus, midbrain, pons, and medulla (Figures 2, 3, and Supplementary Table 1). Only sparse signals can be found in the olfactory area, caudoputamen, and cerebral cortex (Figures 2A–D, 3B–I). In contrast, several structures had little or no signals, including hippocampal formation, the core of thalamus, and cerebellum (Figures 2A–D, 3F–I). Interestingly, we found dense orexin fibers projecting to many sleep-wake and pain control nuclei (Supplementary Figure 1).

To visualize the three-dimensional (3D) orexin projections, we applied light sheet technology to image ChR2-mCherry in the clarified mouse brain (Figures 2G–K). Certain anatomical features, which were hard to recognize in the two-dimensional (2D) images by confocal, were more easily identifiable in the 3D light sheet rendering (Supplementary Figure 2 and Supplementary Video 1).

To confirm the projection patterns above, AAVretro-DIO-ChR2-mCherry was injected into selected brain targets of orexin-Cre mice to see whether the retrograde expression can trace back to the cell bodies.

Five target sites were chosen for the retrograde virus injection. All of them produced robust expressions of ChR2-mCherry. Medial amygdala nucleus (Figures 4A,B), and PAG (Figure 4E) injections produced signals in the thalamic regions but much less than the injections from medial preoptic area (Figure 4C) or lateral habenula (Figure 4D). On the slices from lateral habenula injections, the mCherry signals were found not only in the thalamus, but also in anterior hypothalamus and midbrain (Figure 4D). Locus coeruleus injections showed the soma staining was concentrated at dorsolateral hypothalamus, and no expression was noted in the thalamus region including the lateral habenula (Figure 4F). Furthermore, we observed the rostral projection beyond the cell bodies to the anterior hypothalamus and BST regions. These differences implied that different orexin neurons have specific projection preferences but all orexin neurons seem to project both rostrally and caudally.

After the successful viral expression, we validated the function of the optogenetic stimulation via the loose patch recording and cFos staining (Figure 5A). The orexin neurons expressing ChR2-mCherry in the acute brain slices showed reliable firing upon blue light (473 nm) photoactivation targeting the orexin cell bodies. We also confirmed that optogenetic stimulations excite orexin neurons reliably at different frequencies (1 Hz, 5 Hz, 10 Hz, and 20 Hz, 10 ms pulse width) (Figure 5B).

Compared with the mCherry group, optical stimulation remarkably elevated the percentage of double positive cell (cFos⁺+orexin⁺) among orexin neurons (orexin⁺) in the ChR2-mCherry group (mCherry group: 7.9 ± 2.072% vs. ChR2-mCherry group: 62.86 ± 3.923%; n = 5; P < 0.0001; Figures 5C,D). Our cFos staining confirmed that optogenetic stimulation can activate orexin neurons faithfully.

To investigate the role of the orexin neurons in arousal and pain behavior in live animals, we first explored the arousal effect by optogenetically activating orexin neurons while the animals were under a light isoflurane anesthesia (Figure 6A). The results showed that optogenetic stimulation of orexin neurons can robustly induce arousal while the animals were under 0.75% isoflurane, whereas the control group completely stayed sedated (ChR2-mCherry group: 9.429 ± 3.804 s; n = 7 vs. mCherry group: 238.2 ± 17.42 s; n = 6; P < 0.0001; Figure 6B). The EEG/EMG recording showed the reversible changes of EEG spectrum upon the optogenetic stimulation (Figure 6C). When the mice were exposed to isoflurane at 1% or greater, neither group can be aroused upon stimulation.

We further asked whether the orexin activation can facilitate the emergence speed after 30 min of surgical level of isoflurane (2%) (Figure 6D). The paired t-test comparing the same animals showed the optogenetic activation significantly shortened the emergence time of the ChR2-mCherry group (no light: 237.2 ± 19.26 s vs. 20 Hz: 109.5 ± 13.59 s; n = 6; P = 0.005) but not the mCherry group (no light: 219.2 ± 24.15 s vs. 20 Hz: 213.8 ± 21.77 s; n = 6; P = 0.483; Figure 6E). Unpaired t-test showed the optogenetic activation significantly shortened the emergence time of the ChR2-mCherry group (mCherry group: 213.8 ± 21.77 s vs. ChR2-mCherry group: 109.5 ± 13.59 s; n = 6; P = 0.0023; Figure 6E).

Next, we tested the analgesic effect by activating orexin neurons through a 55^°C hot plate test (Figure 6F). The paired t-test comparing the same animals showed that the optogenetic activation can lengthen the latency to withdraw significantly for the ChR2-mCherry group (no light: 8.133 ± 0.5846 s vs. 20 Hz: 11.45 ± 1.185 s; n = 6; P = 0.0159) but not the mCherry group (no light: 9.117 ± 0.663 s vs. 20 Hz: 8.767 ± 0.7775 s, n = 6; P = 0.5952; Figure 6G). Unpaired t-test displayed significantly longer latencies in the ChR2-mCherry group (mCherry group: 8.767 ± 0.7775 s vs. ChR2-mCherry group: 11.45 ± 1.185 s, n = 6; P = 0.0317; Figure 6G). It illustrated that optogenetic stimulation of the orexin neurons in the lateral hypothalamus increased the animal’s thermal pain tolerance.