Abstract
Orexins, which are produced within neurons of the lateral hypothalamic area, play a pivotal role in the regulation of various behaviors, including sleep/wakefulness, reward behavior, and energy metabolism, via orexin receptor type 1 (OX1R) and type 2 (OX2R). Despite the advanced understanding of orexinergic regulation of behavior at the circuit level, the precise distribution of orexin receptors in the brain remains unknown. Here, we develop a new branched in situ hybridization chain reaction (bHCR) technique to visualize multiple target mRNAs in a semiquantitative manner, combined with immunohistochemistry, which provided comprehensive distribution of orexin receptor mRNA and neuron subtypes expressing orexin receptors in mouse brains. Only a limited number of cells expressing both Ox1r and Ox2r were observed in specific brain regions, such as the dorsal raphe nucleus and ventromedial hypothalamic nucleus. In many brain regions, Ox1r-expressing cells and Ox2r-expressing cells belong to different cell types, such as glutamatergic and GABAergic neurons. Moreover, our findings demonstrated considerable heterogeneity in Ox1r- or Ox2r-expressing populations of serotonergic, dopaminergic, noradrenergic, cholinergic, and histaminergic neurons. The majority of orexin neurons did not express orexin receptors. This study provides valuable insights into the mechanism underlying the physiological and behavioral regulation mediated by the orexin system, as well as the development of therapeutic agents targeting orexin receptors.
Significance Statement
The neuropeptide orexin regulates sleep and other behaviors through its receptors, OX1R and OX2R, which are targets for the development of therapeutic agents for sleep and related disorders. However, the cellular distribution of orexin receptors in the brain is only partially known. We applied a newly developed branched in situ hybridization chain reaction (bHCR) technique and conducted a whole-brain mapping of orexin receptor mRNA expression in the brain with neuron subtype markers. Few cells expressed both OX1R and OX2R, which were expressed in the different neuronal subtypes in many brain regions. This study fills an important gap in understanding and modulating the orexin system.
Introduction
The orexin system, which consists of the neuropeptides, orexin A and orexin B, and their receptors, orexin receptor type 1 (OX1R) and type 2 (OX2R), serves as a core neural circuit for sleep and wakefulness (Sakurai, 2007; Liu and Dan, 2019). In addition, orexinergic neurons are involved in multiple aspects of behaviors and metabolism, such as motivation, addiction, stress response, food intake, energy metabolism, glucose metabolism, sympathetic nervous regulation, and pain sensation (Tsujino and Sakurai, 2009; Funato, 2015; Sargin, 2019; Kang et al., 2021; James and Aston-Jones, 2022; Mavanji et al., 2022).
Orexins, also known as hypocretins, are highly conserved neuropeptides among vertebrates and are exclusively produced in specific neurons of the lateral hypothalamic area (LHA; de Lecea et al., 1998; Sakurai et al., 1998). Orexins A and B are processed from the common precursor peptide, prepro-orexin. Orexin A is a 33-amino acid peptide with 2 intrachain disulfide bonds formed between cysteine residues. Orexin B is a 28-amino acid peptide with an amidated C terminus. OX1R preferentially binds to orexin A, whereas OX2R binds to orexins A and B with similar high affinity (Sakurai et al., 1998). OX1R and OX2R are G-protein–coupled receptors (GPCRs) and have distinct roles in sleep/wake regulation and energy metabolism (Willie et al., 2003; Funato et al., 2009; Kakizaki et al., 2019; Xiao et al., 2021). Orexin receptors were thought to be coupled primarily to Gq and partially to Gi. However, recent comprehensive analysis indicates that both OX1R and OX2R couple promiscuously to Gq, Gs, and Gi (Inoue et al., 2019; Hauser et al., 2022; Kukkonen, 2023).
Orexinergic neurons send their axons throughout the central nervous system. Optogenetic and chemical genetic analysis identified target regions of orexinergic neuron projections that regulate sleep/wakefulness (Saito et al., 2018; Feng et al., 2020; De Luca et al., 2022), aggression (Flanigan et al., 2020), fear (Soya et al., 2017), and glucose metabolism (Xiao et al., 2021). However, in contrast to the advanced circuit-level understanding of the orexin system, the detailed distribution of OX1R and OX2R in the brain has not been shown. The rough distribution of orexin receptors was reported in rat brains by radiolabeled probes approximately 20 years ago and is still the most comprehensive information available (Trivedi et al., 1998; Marcus et al., 2001). Cell resolution information of OX1R and OX2R has been available only for brain regions that abundantly express orexin receptors, such as the amygdala, tuberomammillary nucleus, raphe nuclei, locus ceruleus, and laterodorsal tegmental nucleus (Mieda et al., 2011; Ikeno and Yan, 2018; Xiao et al., 2021; Schneeberger et al., 2022; Yaeger et al., 2022). At this time, the Allen Brain Atlas fails to show positive signals for orexin receptors, and reliable antibodies for OX1R and OX2R that work for immunohistochemistry are not available.
To obtain whole-brain information on cells expressing orexin receptors in a semiquantitative manner, we developed a new branched in situ hybridization chain reaction (HCR) technique (Fig. 1A, Extended Data Fig. 1-1), by combining short hairpin HCR with split-initiator probe (Tsuneoka and Funato, 2020) and hyperbranched HCR systems (Bi et al., 2015). This branched HCR (bHCR) can visualize not only multiple target mRNAs but also proteins combined with immunohistochemistry because bHCR does not require proteinase K treatment, which reduces the immunogenicity of proteins. Using bHCR combined with immunohistochemistry, we characterized the distribution and transmitters of cells expressing orexin receptors throughout the brain. This study also provides information on orexin receptor expression in monoaminergic and cholinergic neurons. We discuss orexin receptor expression in terms of different behavioral modalities such as sleep/wakefulness, reward behavior, feeding, energy metabolism, and the autonomic nervous system.
Outline of branched HCR (bHCR) amplification and its specificity. A, A pair of split-initiator probes formed an initiator sequence to start the 1st HCR when the probes hybridized with the target mRNAs in the correct position. In the first HCR, the long hairpin DNA and the short hairpin DNA formed a “trunk” on the mRNA. Subsequently, the remaining hairpin domains of the trunk were hybridized with linear ssDNA to expose initiator sequences to start the polymerization of fluorophore-labeled hairpin DNAs (second HCR). For details, please see Extended Data Figure 1-1. B–D, Reconstructed distribution of Ox1r- and Ox2r-expressing cells approximately bregma 0 level. The receptor-expressing cells are indicated by green (Ox1r) or red (Ox2r) circles. The large circles denote the cells with high expression. The rectangles in B and C indicate the region shown in D, and the rectangle in D indicates the magnified view in E. E–G, Representative microphotographs of Ox1r- and Ox2r-stained sections of the MPOA of wild-type, Ox1r KO, and Ox2r KO mice. Cell nuclei are shown in blue. The arrows indicate the receptor-expressing cells. The outlined arrowheads in F and G show a single granule of noise. Scale bars: 1 mm (B), 200 µm (D), and 50 µm (E). H,I, Reconstructed three-dimensional images showing receptor expression of Ox1r (H) and Ox2r (I). The regions showing notable expressions are indicated by arrows. A scheme showing detailed reaction steps is provided in Extended Data Figure 1-1.
Figure 1-1
Principle of branched in situ hybridization chain reaction (HCR) Split-initiator probes hybridize with the target mRNA to form an initiator sequence (bc) for hairpin DNA H0. In the 1st HCR, H0 hybridizes with the pair of split-initiator probes, followed by exposure of initiator sequence (a'b') for double stem-hairpin DH. DH hybridizes with the 3' end of H0, and the initiator sequence for H0 is exposed from DH. As a consequence of HCR, the DNA trunk with the repeated hairpin-containing domain is formed on the target mRNA. In the 2nd HCR, the assist oligo (AS) hybridizes with the hairpin-containing domain to expose the initiator sequence (xy') of fluorescent hairpin DNA FH1. FH1 hybridizes with the single-strand region of the trunk to form branched DNA with the initiator sequence (zy) of FH2. HCR of FH1 and FH2 occur in multiple regions of the DNA trunk, enabling higher polymerization of fluorescent hairpins per single target site than non-branched HCR. Download Figure 1-1, TIF file.
Materials and Methods
Animals
All animal procedures were conducted in accordance with the Guidelines for Animal Experiments of Toho University and were approved by the Institutional Animal Care and Use Committee of Toho University (Approval Protocol ID #21-53-405). Breeding pairs of C57BL/6J mice were obtained from Japan SLC and CLEA Japan. Mice were raised in our breeding colony under controlled conditions (12 h light/dark cycle, lights on at 8:00 A.M., 23 ± 2°C, 55 ± 5% humidity, and ad libitum access to water and food). We also used male Ox1r-deficient mice (Kakizaki et al., 2019) and male Ox2r-deficient mice (Willie et al., 2003).
Twelve- to 15-week-old male mice were anesthetized with sodium pentobarbital (50 mg/kg, i.p.) and then transcardially perfused with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) for 2–4 h after the start of the light phase. The brains were postfixed in 4% PFA at 4°C overnight, followed by cryoprotection in 30% sucrose in PBS for 2 d, embedded in Surgipath (FSC22, Leica Biosystems), and stored at −80°C. The frozen brains except olfactory bulbs were cryosectioned coronally at a thickness of 40 μm. The sections were stored in an antifreeze solution (0.05 M phosphate buffer, 30% glycerol, 30% ethylene glycol) at −25°C until use. Every third section from the serial sections was processed for the same staining procedures. To confirm the reproducibility of the staining, sections from two or three mice were used for each staining combination.
Branched HCR with IHC
Preparation of probes and hairpin DNAs
The probes for ISH-HCR were designed to minimize off-target complementarity using a homology search by NCBI Blastn (https://blast.ncbi.nlm.nih.gov), and they were designed to have split-initiator sequences with an mRNA binding site. For Ox1r probes, five target sequences overlapped with exons 5–6, which are absent in Ox1r KO mice. For Ox2r probes, 10 target sequences were selected from the complete cDNA sequence because mRNA expression was completely abolished in Ox2r KO mice. Other probes were designed according to our published protocol (Table 1). The DNA probes were synthesized as standard desalted oligos (Integrated DNA Technologies) and purified by denaturing polyacrylamide gel electrophoresis (PAGE) using 20% polyacrylamide gels. The fluorescent hairpin DNAs were prepared by conjugation of succinimidyl ester of fluorophores (Sarafluore488, ATTO550, or ATTO647N) and synthesized with a C12 amino linker at the 5′ end (Integrated DNA Technologies or Tsukuba Oligo Services). The DNA probes and short hairpin DNAs were purified by PAGE using 20% polyacrylamide gels. The double-stem hairpin DNAs were designed as a combination of our established hairpin DNA sets according to a previous study for branched HCR (Bi et al., 2015). The double-stem hairpins were purified by PAGE using 15% polyacrylamide gels. The detailed procedures for DNA purification were described in our previous study (Tsuneoka and Funato, 2020).
Split-initiator probe sequences
Probe hybridization
ISH procedures were similar to those previously described with some modifications (Tsuneoka and Funato, 2020; Katayama et al., 2022). The free-floating sections were washed with PBS containing 0.2% Triton (PBST) and immersed in methanol for 10 min, followed by PBST washing for 5 min twice. After washing, the sections were prehybridized for 5 min at 37°C in a hybridization buffer containing 10% dextran sulfate, 0.5× SSC, 0.1% Tween 20, 50 μg/ml heparin, and 1× Denhardt's solution. The sections were treated with another hybridization solution containing a mixture of 10 nM probes and incubated overnight at 37°C. One to three probe sets were selected from the probe sets for Ox1r, Ox2r, Vgat, Vglut1, Vglut2, and Chat mRNAs. The control experiment was performed without probes or using sections of Ox1r- or Ox2r-deficient mice. After hybridization, the sections were washed three times for 10 min in 0.5× SSC containing 0.1% Tween 20 at 37°C. Then, the sections were bleached by an LED illuminator for 60 min in PBST to quench autofluorescence (Tsuneoka et al., 2022).
Branched HCR amplification
For the first HCR amplification, the DH and H0 hairpin DNAs were used to form a DNA trunk (Extended Data Fig. 1-1, Table 2). These hairpin DNA solutions were separately snap-cooled (heated to 95°C for 1 min and then gradually cooled to 65°C for 15 min and 25°C for 40 min) before use. The sections were incubated in amplification buffer (10% dextran sulfate in 8× SSC, 0.2% Triton X-100, 100 mM MgCl2) for 5 min and then immersed in another amplification buffer containing 150 nM DH and 60 nM H0 hairpin DNAs for 45 min at 25°C. The samples were washed three times with PBST at 37°C for 15 min followed by second HCR amplification. The sections were incubated in the amplification buffer described above for 5 min and then immersed in another amplification buffer containing 60 nM snap-cooled fluorescent hairpin DNA and 120 nM assist oligo DNA for 2 h at 25°C.
Hairpin sequence and conjugated dye
In the case of combined ISH and immunohistochemistry, the sections were blocked using 0.8% Block Ace/PBST (Dainihon Seiyaku), which was followed by overnight incubation with one or two antibodies from rabbit anti-TH (Millipore, AB152, 1:4,000, RRID: AB_390204), mouse anti-calbindin D28k (Sigma-Aldrich, C9848, 1:2,000, RRID: AB_476894), guinea pig HDC (American Research Product, 03-16046, 1:2,000, RRID: AB_1541511), goat 5-HT (Immunostar, 20079, 1:2,000, RRID: AB_572262), and rabbit orexin A (Millipore AB3704, 1:2,000, RRID: AB_91545) antibodies in 0.4% Block Ace/PBST at 4°C. After washing three times with PBST, sections were incubated with a secondary antibody mixture with Hoechst 33342 (1 μg/ml) for an hour at room temperature. The sections were mounted on the slide glass and coverslipped with mounting media containing antifade (1% n-propyl gallate and 10% Mowiol 4-88 in PBS).
Histological analysis
Fluorescence photomicrographs were obtained using a Nikon Eclipse Ni microscope equipped with the A1R confocal detection system under 20×/0.75 NA objective lenses at 4 μs/pixel speed at 0.628 μm/pixel resolution using autofocus (Nikon Instruments). Tiled images were captured and automatically stitched by NIS-Elements C software (Nikon Instruments).
Images were analyzed using ImageJ software (version 1.50i, NIH). Quantification of the fluorescent photographs was performed at the same threshold and adjustment of contrast. All of the mapping procedures of receptor-expressing cells were performed manually under blinded conditions. If more than one granule contacted each cell nucleus, the cell was judged as mRNA-positive. If more than five puncta were observed, the expression level of each cell was regarded as high expression. The distributions of mRNA-positive cells were reconstructed by drawing circles on the original photographs. The three-dimensional images of receptor-expressing cells were reconstructed from serial sections of which positions were adjusted manually according to the published brain atlas. The area location was basically decided according to the brain atlases (Watson and Paxinos, 2010; Paxinos and Franklin, 2019), while the area boundary shown in Figures 2⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓–17 was adjusted from the expression of Vgat, Vglut1, Vglut2, and Chat mRNAs, calbindin, and TH immunoreactivity. In each brain region, whether orexin receptor–positive cells were glutamatergic or GABAergic was examined using bHCR sections for Ox1r, Vgat, and Vglut2 and bHCR sections for Ox2r, Vgat, and Vglut2. For orexinergic neurons, orexin receptor expression was examined using bHCR sections for Ox1r and Ox2r with immunostaining of anti-orexin A antibody. For monoaminergic neurons, orexin receptor expression was examined using bHCR sections for Ox1r and Ox2r with immunostaining of anti-TH, HDC, or 5-HT antibodies. For cholinergic neurons, orexin receptor expression was examined using bHCR sections for Ox1r, Ox2r, and Chat.
Distribution of orexin receptor–expressing cells and regional markers (1 of 16). Panels 1–12. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. A, Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. B, Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. C, Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. D, Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (2 of 16). Panels 13–24. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (3 of 16). Panels 25–36. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (4 of 16). Panels 37–48. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (5 of 16). Panels 49–60. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (6 of 16). Panels 61–72. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (7 of 16). Panels 73–84. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (8 of 16). Panels 85–96. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (9 of 16). Panels 97–108. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (10 of 16). Panels 109–120. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (11 of 16). Panels 121–138. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (12 of 16). Panels 139–153. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (13 of 16). Panels 154–168. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (14 of 16). Panels 169–183. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (15 of 16). Panels 184–192. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Distribution of orexin receptor–expressing cells and regional markers (16 of 16). Panels 193–222. Numbers show the anterior–posterior position of the complete serial sections with 40 μm thickness from a single mouse brain. Dashed-line 1 mm grids overlapped. (Leftmost column) Reconstructed distribution of the orexin receptor–expressing cells. The green and red dots indicate Ox1r- and Ox2r-expressing cells, respectively. (Second column from left) Photographs of Vgat (green), Vglut1 (blue), and Vglut2 (red) mRNA staining. (Third column from left) Photographs of calbindin (green) and TH (blue) immunostaining and Chat (red) mRNA staining. (Rightmost column) Atlas of brain regions examined.
Results
To obtain whole-brain information on orexin receptor expression at a cellular resolution in a semiquantitative manner, we developed bHCR, which consists of two steps of HCRs (Fig. 1A, Extended Data Fig. 1-1). In the first step, the hybridization of a pair of split-initiator probes to target mRNAs triggers HCRs of double-stem hairpin and short hairpin, which forms a “hairpin-containing DNA trunk” (Choi et al., 2018; Tsuneoka and Funato, 2020). In the second step, assist oligo DNAs to bind to double-stem hairpins of the DNA trunk to stretch self-hybridization of double-stem hairpins, which allows double-stem hairpins to hybridize to fluorophore-labeled hairpin DNAs. Then, HCRs further occur on the trunk to form fluorescently tagged branches.
bHCR visualized Ox1r and Ox2r signals throughout the brain in a region-specific manner (Fig. 1B–D,H,I). Since our HCR system can visualize a single mRNA molecule (Tsuneoka and Funato, 2020), the number of fluorescent granules produced by HCRs is thought to roughly correlate with the expression level of the target mRNAs. The locus ceruleus (LC) and ventral part of the tuberomammillary nucleus (TMV) contain cells expressing a large number of fluorescent granules for Ox1r and Ox2r, respectively (Fig. 1H,I, Fig. 23E,H) as previously reported (Marcus et al., 2001; Mieda et al., 2011). However, the majority of orexin receptor–positive cells exhibited fewer than five fluorescent granules around the nucleus, reflecting their low expression (Fig. 1E). The specificity of the fluorescent granules for Ox1r and Ox2r was verified using Ox1r- and Ox2r-deficient brains (Fig. 1E–G). A few fluorescent granules were sparsely and evenly distributed in all brain regions, which may be artifacts or background signals (Fig. 1E–G). Therefore, we considered a cell with two or more fluorescent granules to be a positive cell. Depending on the number of fluorescence granules, positive cells were further classified into low or high expression. We drew small circles for cells with low expression and large circles for cells with high expression on original microscopic images (Figs. 1B–D, 2⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑⇑–17). The reconstructed 3-D images of cells expressing Ox1r and Ox2r showed that the density of orexin receptor–positive cells largely varied among brain regions (Fig. 1H,I). Importantly, many brain regions had Ox1r-positive cells and Ox2r-positive cells, but Ox1r- and Ox2r–double-positive cells were rarely observed (Fig. 1B–D), except for some brain regions, such as the paraventricular thalamus (PVT), ventromedial nucleus (VMH), and dorsal raphe nucleus (DRN), where a large number of double-positive cells were found (see below).
To make a whole-brain map of Ox1r- and Ox2r-expressing cells, we used a complete series of 40-μm-thick coronal sections from a single male mouse. For every three sections, bHCR for Ox1r and Ox2r; bHCR for Vgat, Vglut1, and Vglut2; and bHCR for Chat with immunostaining for calbindin and TH were performed (Figs. 2–17). In addition, to characterize Ox1r- or Ox2r-expressing cells, bHCR for Ox1r or Ox1r was combined with bHCR for Vgat, Vglut2, or Chat, followed by immunostaining of either tyrosine hydroxylase (TH), histidine decarboxylase (HDC), 5-HT or orexin. Regional expression of the orexin receptor, coexpression of glutamatergic or GABAergic markers (Vglut2 or Vgat), and distribution of area markers (Vglut1, Vglut2, Vgat, Chat, TH, and calbindin) are summarized in Tables 3 and 4.
The cellular amount of Ox1r or Ox2r mRNA for certain cell types and the number of orexin receptor–expressing cells
Examined brain areas with regional marker information
Cerebrum
In the isocortex, a large number of Ox1r-positive cells and Ox2r-positive cells were observed in layers V and VI (Figs. 2–10, 18A–D). Ox1r- or Ox2r-positive cells included glutamatergic cells and GABAergic cells. A small number of Ox1r- and Ox2r–double-positive cells were found in layer VI (Figs. 2–10, 18A–D). In layers II/III, a moderate number of Ox2r-positive cells were observed, which were Vgat-positive cells, while a small number of Ox1r-positive cells were observed, which were Vgat-negative cells (Table 3). Few orexin receptor–expressing cells were found in layers I and IV.
In the anterior olfactory area, such as the anterior olfactory nucleus (AON), dorsal taenia tecta (DTT), ventral taenia tecta (VTT), and olfactory tubercle (OT), Ox2r-positive cells were abundant, and Ox1r-positive cells were scarce (Figs. 2, 3). The majority of Ox2r-positive cells were Vgat-positive. In the piriform area, anterior to the bregma level, abundant Ox2r-positive cells and few Ox1r-positive cells were observed, but in the piriform area posterior to the bregma level, the number of Ox2r-positive cells decreased, and the number of Ox1r-positive cells increased (Figs. 5, 18S–Z). In the anterior cortical amygdala (ACo) and nucleus of the lateral olfactory tract (LOT), Ox1r was expressed in Vglut2-positive neurons. In the posterolateral cortical amygdala (PLCo), Ox1r-positive cells were rare. Similar to the anterior olfactory area, more than half of the Ox2r-positive cells were Vgat-positive in the cortical amygdala and adjacent nuclei.
In general, the striatum and related regions contained many Ox2r-positive cells with strong expression and fewer Ox1r-positive cells. A small population of GABAergic neurons in the caudoputamen (CPu) expressed Ox2r. In the CPu, there were scattered Vgat- and Chat–double-positive large cells, one-third of which strongly expressed Ox2r. The nucleus accumbens expressed Ox2r but not Ox1r. The nucleus accumbens shell (Nash) was one of the most densely populated regions in the brain with cells expressing Ox2r (Fig. 3). The diagonal band (NDB), magnocellular nucleus (MN), and globus pallidus (GP) were enriched in Vgat-positive cells highly expressing Ox2r (Fig. 18E–H). Regarding the septal area, Ox2r expression was found mainly in Vgat-positive cells in the medial septum (MS), while Ox2r expression was found in Vglu2-positive cells in the triangular septal nucleus (TRS), septofimbrial nucleus (SF), and lateral septum (LS).
Representative expression of Ox1r and Ox2r in the cerebrum. (A–D) Ox1r and Ox2r expression in isocortex layer VI. E–H, Ox2r expression in the cell region of the nucleus accumbens (F), the nucleus of the diagonal band (G), and the substantia innominata (H). I–L, Ox1r and Ox2r expression in the BNST principal nucleus. M–P, Ox1r expression in the medial amygdala, posterodorsal (N), basolateral amygdala (O), and basomedial amygdala (P). Q,R, Ox2r expression in the CA2 region of the hippocampus. The green, red, and yellow arrows denote Ox1r-, Ox2r-, and both receptor–expressing cells. S–Z, Ox1r, Ox2r, Vglut2, and Vgat expression in the piriform cortex (PIR). Scale bars: 50 μm (A–R), 25 μm (V,Z).
For the extended amygdala, the principal nucleus of the BNST (BNSTpr) contained Ox1r-expressing cells, which are Vgat-positive neurons, while cells expressing Ox2r were mainly positive for Vglut2 (Fig. 18I–L). In the BNST nuclei, except for BNSTpr, Ox1r- and Ox2r-positive cells were Vgat-positive. The posterodorsal and posteroventral medial amygdala (MePD, MePV) were rich in Ox1r-positive cells, and they were mainly Vglut2-positive (Fig. 18M–P). The central amygdala (CEA) expressed Ox2r but not Ox1r, and Ox2r-positive cells were Vgat-positive. The anterodorsal medial amygdala (MeAD) and anteroventral medial amygdala (MeAV) contained Ox1r-positive cells and Ox2r-positive cells, which were mainly Vgat-positive.
In the endopiriform nucleus (Epi) and claustrum (CLA), the expression levels of Ox1r and Ox2r were relatively low. Among the lateral, basolateral, and basomedial amygdala, area boundaries can be easily divided by Vglut1 and Vglut2 expression (Fig. 8, Panels 74, 77, and 80). In the basolateral amygdala (BLA) and basomedial amygdala (BMA), a dense distribution of Ox1r-positive glutamatergic cells was observed (Fig. 18M–P), while a relatively small number of Ox2r- and Vgat-positive cells were distributed in the posterior part of the basolateral amygdala (BLP), the posterior part of the basomedial amygdala (BMP), lateral amygdala (LA), and amygdala hippocampal area (Ahi).
In the hippocampus, there was a high density of Ox2r-positive cells in the CA2 region that were Vgat-negative and appeared to be pyramidal cells and a moderate density of Ox2r-positive cells in the dentate gyrus, CA1, CA3, and subiculum, which were Vgat-positive (Fig. 18Q,R, Table 3). Ox1r was expressed in Vgat-negative cells, which were localized in the polymorphic cell layer of the dentate gyrus and CA3 region. In the posterior part of the CA1, CA3, and subiculum, Ox2r-positive cells were Vgat-negative.
Thalamus
In the thalamus, Ox1r and Ox2r were moderately expressed in the midline regions, such as the paraventricular nucleus of the thalamus (PVT) and centromedian nucleus (CM) but low in other thalamic regions. Since the thalamus is composed of excitatory neurons, Ox1r expression was also found in Vglut2-positive cells of the anteromedial nucleus (AM), paraxiphoid nucleus (PaXi), and subparafascicular nucleus (SPF). Ox2r-positive cells were more frequently found around midline cells, especially in the parataenial nucleus (PT) and xiphoid nucleus (Xi). In the posterior group, Ox2r-positive cells were found in the subparafascicular area (SPA), SPF, and perireunensis nucleus (PR). The anterior part of the PVT, which is one of the major target areas of orexin neurons to promote wakefulness (Ren et al., 2018), moderately expressed Ox1r and Ox2r and contained many Ox1r- and Ox2r–double-positive cells (Fig. 19A–D). Regarding GABAergic areas surrounding the thalamus, Ox1r and Ox2r were not expressed in the reticular thalamus (RT) but were expressed in Vgat-positive cells of the intergeniculate leaflet (IGL) and pregeniculate nucleus (PrG; Fig. 19E,F). The medial and lateral habenula expressed Ox2r but not Ox1r.
Representative expression of Ox1r and Ox2r in the diencephalon. A–D, Ox1r and Ox2r expression in the periventricular thalamus. E,F, Ox1r expression in the intergeniculate leaflet. G,H, Ox2r expression in the paraventricular hypothalamic nucleus. I–L, Ox1r and Ox2r expression in the VMH, dorsomedial. M–P, Ox1r and Ox2r expression in the lateral mammillary nucleus. The green, red, and yellow arrows denote Ox1r-, Ox2r-, and both receptor–expressing cells. Scale bars: 50 μm.
Hypothalamus
Consistent with the vicinity of orexin neurons localized in the LHA and dense orexin fiber projections in the hypothalamus, almost all hypothalamic areas expressed either Ox1r or Ox2r, except for the suprachiasmatic nucleus (SCN) and vascular organ of the lamina terminalis (VOLT).
In the preoptic region, the MPOA was one of the densest areas of receptor-expressing cells in the brain, although the expression intensity per cell was not high (Fig. 1B–D). Ox1r-positive cells were found in Vgat-positive cells but not in Vglut2-positive cells. Ox1r-positive cells were distributed in the relatively posterior part of the MPOA. In contrast, Ox2r-positive cells were found in both Vgat- and Vglut2-positive cells and distributed in the relatively anterior part of the MPOA. In the median preoptic nucleus (MnPO) and lateral preoptic areas (LPOA), the expression was relatively low.
In the periventricular area, notable expression of Ox2r was found, although the expression intensity per cell was not high. In the paraventricular nucleus (PVN), Ox2r was expressed in many Vglut2-positive cells but not in Vgat-positive cells (Fig. 19G,H). The arcuate nucleus (ARC), which is composed of GABAergic neurons, expressed Ox2r but not Ox1r. In the dorsomedial nucleus (DMH), most of the Ox2r-positive cells were Vgat-positive.
Ox1r- and Ox2r-positive cells were abundant in the ventromedial nucleus (VMH), which is composed of excitatory neurons. In the dorsomedial part of the VMH, positive cell density was highest, followed by the ventrolateral part. Double-positive cells for Ox1r and Ox2r were frequently found in the dorsomedial part (Fig. 19I–L). In the LHA, a moderate number of receptor-expressing cells were observed in Vgat-positive and Vglut2-positive cells. In the posterior hypothalamus (PH), subthalamic nucleus (STN), and parasubthalamic nucleus (PST), many Vglut2-positive cells expressed Ox2r, but Vgat-positive cells did not express either Ox1r or Ox2r.
In the hypothalamus, cells strongly expressing Ox2r were most abundant in the premammillary and mammillary areas. In the dorsal premammillary nucleus (PMD) and ventral premammillary nucleus (PMV), strong and dense expression of Ox2r-positive cells was observed, whereas weakly expressed Ox1r mRNAs were observed in a moderate number of cells. The dorsal tuberomammillary nucleus (TMD) and ventral tuberomammillary nucleus (TMV) were highly dense in strongly Ox2r-expressing cells but did not express Ox1r. The lateral mammillary nucleus (LM), the lateral cluster of Vglut2-positive neurons in the mammillary body, was abundant in Ox1r- and Ox2r–double-positive cells, in addition to single-positive cells (Fig. 19M–P). In the supramammillary nucleus (SuM), which contains cells expressing both Vglut2 and Vgat (Root et al., 2018), a moderate number of Ox2r-positive cells were observed.
Midbrain
In the pretectal region, the anterior pretectal nucleus (APN) and medial pretectal nucleus (MPT) express the orexin receptor. In the APN, a large number of Vglut2- and Vgat-positive neurons expressed Ox2r, and a few of them expressed Ox1r.
In the ventral dopaminergic region, a moderate number of Ox1r- or Ox2r-expressing cells were found, but the expression per cell was low for both Ox1r and Ox2r. Among these regions, Ox1r expression was greater. In the retrorubral area (RR), substantia nigra pars compacta (SNc), and substantia nigra pars reticulata (SNr), approximately half of the receptor-expressing neurons were Vgat-positive, and the remaining neurons were Vgat- and Vglut2-negative and were presumed to be dopaminergic cells (Fig. 11, also see Monoamine neurons). In the ventral tegmental area (VTA), cells expressing either Ox1r or Ox2r were Vgat-positive cells, Vglut2-positive cells, or Vgat- and Vglut2-double-negative cells.
The periaqueductal gray (PAG) and adjacent nuclei were densely populated with Ox1r-positive cells and sparsely populated with Ox2r-positive cells. In the PAG, especially the lateral and ventrolateral subdivisions, approximately half of the neurons were Ox1r-positive. Most of the Ox1r-positive neurons were Vgat-positive and a few were Vglut2-positive. Conversely, Ox2r-positive neurons in the PAG were predominantly Vglut2-positive neurons. The highest expression of Ox2r per cell was observed in Vglut2-positive neurons of the medial accessory oculomotor nucleus (MA3), or Edinger–Westphal nucleus.
In the midbrain raphe nuclei, Ox1r and Ox2r were highly expressed. Whereas most Ox1r-positive neurons in the central linear nucleus (CLi) were Vgat-positive cells, most Ox1r-positive neurons in the rostral linear nucleus (RLi) were Vglut2-positive cells. The DRN is one of the regions with the highest per-cell expression of both Ox1r and Ox2r. The vast majority of receptor-expressing cells were negative for both Vgat and Vglut2 and were probably serotonergic cells (also see Monoamine neurons).
In the other midbrain nuclei, notable expression of Ox1r was found in the nucleus sagulum (SAG) and pedunculopontine tegmental nucleus (PPT). The SAG was a small cluster of Vglut2-positive cells, and almost all cells in the SAG showed low expression of Ox1r. In the PPT, most Vglut2-negative neurons, presumed to be cholinergic neurons, expressed Ox1r at the highest intensity in the brain. In contrast, Ox2r-positive cells in the PPT were Vglut2-positive. In the nucleus of the brachium of the inferior colliculus (NB) and parabigeminal nucleus (PBG), which are composed of Vglut2-positive neurons, cells expressing a high level of Ox2r were densely localized. Moderate expression of Ox2r was found in the superior colliculus (SC) and red nucleus (RN). In the motor-related SC (SCm), Ox2r expression was found in both Vgat-positive and Vglut2-positive cells, whereas Ox2r expression was found only in Vgat-positive cells in the sensory-related SC (SCs).
Pons
In the locus ceruleus (LC) and laterodorsal tegmental nucleus (LDT), there were numerous Ox1r-expressing cells with high Ox1r expression and no Ox2r-expressing cells. Nearly all Ox1r-positive cells in the LC and LDT were negative for both Vgat2 and Vgat, which are assumed to be adrenergic and cholinergic neurons, respectively. In the sublaterodorsal tegmental area (SLD), a moderate number of Vglut2-positive cells were Ox1r- or Ox2r-positive, and a small number of Vglut2-positive cells expressed both Ox1r and Ox2r. In Barrington's nucleus (BAR), which is composed of mostly Vglut2-positive neurons, Ox1r-positive cells, Ox2r-positive cells, and Ox1r- and Ox2r–double-positive cells were observed (Fig. 20A–D). The pontine gray (PG) and tegmental reticular nucleus (TRN) were abundant in Ox2r-positive neurons, which were Vglut2-positive (Fig. 20E–H). In addition, the PG and TRN also intensely expressed Vglut1 (Figs. 12, 13). The principal sensory nucleus of the trigeminal nerve (PSV) expressed only Ox2r. In the ventrolateral PSV (PSVvl), Ox2r-positive neurons were Vglut2-positive, while in the dorsomedial PSV (PSVdm), Ox2r-positive neurons were mainly Vgat-positive.
Representative expression of Ox1r and Ox2r in the brainstem. A–D, Ox1r and Ox2r expression in Barrington's nucleus. E,F, Ox2r expression in the pontine gray. G,H, Ox2r expression in the tegmental reticular nucleus. I–K, Ox2r expression in the external cuneate nucleus (J) and the lateral reticular nucleus (K). The green, red, and yellow arrows denote Ox1r-, Ox2r-, and both receptor–expressing cells. Scale bars: 50 μm. All sections every 120 μm from a single mouse were quantified.
Medulla and cerebellum
In the medulla, Ox1r expression was generally low and sparse. A relatively dense population of Ox1r-positive cells was observed in the dorsal motor nucleus of the vagus nerve (10N), compact part of the nucleus ambiguous (AMBc), medial vestibular nucleus (MVE), lateral part of the paragigantocellular reticular nucleus (PGRNl), nucleus raphe obscurus (RO), nucleus raphe pallidus (Rpa), and spinal vestibular nucleus (SPIV), although the expression per cell was low. In contrast, a dense population of Ox2r-expressing cells with a high expression level was observed in many nuclei in the medulla. In the cochlear nuclei, almost all neurons in the granular region (GRC) expressed Ox2r, while the ventral cochlear nucleus (VCO) did not express Ox2r and the dorsal cochlear nucleus (DCO) moderately expressed Ox2r. Cholinergic motor nuclei also showed moderate expression of Ox2r (also see Cholinergic neuron). Moderate Ox2r expression was found in Vgat-positive cells and Vglut2-positive cells in the cuneate nucleus (CU), intermediate reticular nucleus (IRN), magnocellular reticular nucleus (MARN), dorsal part of the medullary reticular nucleus (MdV), nucleus of the solitary tract (NTS), parvocellular reticular nucleus (PARN), dorsomedial part of the spinal nucleus of the trigeminal nerve (SPVd), and oral part of the SPV (SPVo). The external cuneate nucleus (ECU) contained numerous cells that were negative for both Vgat and Vglut2 and highly expressed Ox2r. In the linear nucleus (LIN), lateral reticular nucleus (LRN), ventral part of the nucleus prepositus (PRPv), and nucleus X, a dense population of highly Ox2r-expressing cells was observed (Fig. 20I–K), and the Ox2r-expressing cells were Vglut2-positive (Fig. 17).
Orexin neurons
In the lateral hypothalamic area, both Ox1r and Ox2r expression was observed. To evaluate the receptor expression in the orexin neurons, we conducted orexin A immunohistochemistry combined with Ox1r and Ox2r bHCR. Only a few orexin A–positive neurons were positive for either Ox1r or Ox2r (5 of 101 cells, Fig. 21).
Orexin receptor expression of orexin neurons. A, Immunofluorescence of orexin A in the hypothalamus. The rectangle indicates the region shown in (B–D). Scale bar: 500 μm. B–D, Ox1r and Ox2r expression and immunofluorescence of orexin A. The cyan and green arrows denote Ox1r- and Ox2r-expressing cells, respectively. Scale bar: 50 μm.
Monoamine neurons
In serotonergic neurons, high expression levels of Ox1r and/or Ox2r were observed, but the proportions of Ox1r and Ox2r expression differed among cell groups (Fig. 22, Extended Data Figs. 22-1, 22-2). In the anterior DRN, referred to as the B7 group, where relatively large 5-HT-positive cells were distributed, numerous neurons expressed either Ox1r or Ox2r, or both (Fig. 22A–D). The dominant cell types in the B7 group were Ox1r-positive (Ox1r, 36.6%; Ox2r, 18.8%; Ox1r + Ox2r, 22.6%; Fig. 22K). In the posterior DRN, referred to as B4/6 groups, where relatively small cells were clustered in the midline, the expression level per cell was weaker than in other serotonergic nuclei, and the proportion of orexin receptor–expressing cells was the smallest among serotonergic cell groups (Ox1r, 11.0%; Ox2r, 20.1%; Ox1r + Ox2r, 9.1%; Fig. 22E–H,K). In the MRN, called B5/8/9 cell groups, there was a large number of Ox1r-positive neurons, whereas the number of Ox2r-positive neurons was relatively small (Ox1r, 45.8%; Ox2r, 6.3%; Ox1r + Ox2r, 10.4%; Fig. 22K). The ventral raphe nucleus (VRN), called B1/2/3 cell groups, was a relatively small cluster of serotonergic neurons located in the ventral medulla, and the majority of receptor-expressing cells was Ox1r-positive (Ox1r, 62.5%; Ox2r, 1.3%; Ox1r + Ox2r, 2.5%; Fig. 22I–K).
Orexin receptor expression of serotonergic neurons. A,E,I, A reconstructed image showing Ox1r- and Ox2r-positive serotonergic neurons. The open, green, red, and yellow circles denote no receptor, Ox1r-, Ox2r-, and both receptor–positive cells, respectively. Scale bars: 500 μm. B–D, Ox1r and Ox2r expression and immunofluorescence of 5-HT in the B7 region of the dorsal raphe nucleus. F–H, Immunofluorescence of 5-HT in the B6 region of the dorsal raphe nucleus. J, Ox1r expression and immunofluorescence of 5-HT in the ventral raphe nucleus. 5-HT immunofluorescence is shown in blue. The green, red, and yellow arrows denote Ox1r-, Ox2r-, and both receptor–expressing cells, respectively. Outlined arrowheads denote the no-receptor–expressing 5-HT-ir neurons. Scale bars: 50 μm. K, Regional differences in the expression of receptor subtypes in serotonergic neurons. The numbers above the bar plot indicate the total number of cells. All sections every 120 μm from a single mouse were quantified. Data of all sections are provided in Extended Data Figure 22-1 and 22-2.
Figure 22-1
Distribution of orexin receptor-expressing serotonergic neurons (1/2). White, green, red, and yellow circles indicate receptor-negative, Ox1r-positive, Ox2r-positive, and both Ox1r and Ox2r-positive serotonergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 22-1, TIF file.
Figure 22-2
Distribution of orexin receptor-expressing serotonergic neurons (2/2). White, green, red, and yellow circles indicate receptor-negative, Ox1r-positive, Ox2r-positive, and both Ox1r and Ox2r-positive serotonergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 22-2, TIF file.
Dopaminergic cell groups A11–15, in addition to periventricular nuclei, PVD, and PVH, were distributed in the hypothalamus. In these cell groups except A11, a small proportion of dopaminergic cells were Ox2r-positive, and very few cells were Ox1r-positive (Ox1r, 1.0%; Ox2r, 8.3% in total; Fig. 23). A11 dopaminergic cells were scattered around the dorsal border of the posterior hypothalamus, and approximately half of these cells were Ox2r-positive (51.7%, Fig. 23A,B). In contrast to the hypothalamic dopaminergic neurons, midbrain dopaminergic neurons called A8–10 groups contained a relatively high proportion of Ox1r-positive cells. Although the distributions of Ox1r- or Ox2r-positive cells were not uniform within each nucleus (Extended Data Fig. 23-3), similar proportions of Ox1r-positive cells were found in A8 RR, A9 SNc, and A10 VTA neurons (RR, 13.8%; SNc, 14.8%; VTA, 16.6%; Fig. 23C,D,G). Dopaminergic neurons in the midbrain were also distributed in the DRN, anterior to the B7 serotonergic cell clusters. In these DRN dopaminergic cells, receptor-positive neurons were never found.
Orexin receptor expression of dopaminergic, noradrenergic, and histaminergic neurons. A,C,E, A reconstructed image showing Ox1r- and Ox2r-positive cells in TH-positive neurons. The open, green, and red circles denote no receptor, Ox1r-, and Ox2r-positive cells, respectively. Scale bars: 500 μm. B, Ox2r expression and immunofluorescence of TH in the A11 region of the hypothalamus. D, Ox1r expression and immunofluorescence of TH in the ventral tegmental area. F, Ox1r expression and immunofluorescence of TH in the locus ceruleus. TH immunofluorescence is shown in blue. The green and red arrows denote the Ox1r- and Ox2r-expressing cells, respectively. Outlined arrowheads denote the no-receptor–expressing TH-ir neurons. Scale bars: 50 μm. G, Regional differences in the expression of receptor subtypes in TH-ir neurons. The numbers above the bar plot indicate the total number of cells. For details, please see Extended Data Figures 23-1, 23-2, 23-3, 23-4, 23-5, and 23-6. H, A reconstructed image showing Ox2r-positive histaminergic neurons. A total of 97.4% of histidine decarboxylase (HDC)-ir neurons were positive for Ox2r. Scale bar: 500 μm. I, Immunofluorescence of histidine decarboxylase (green). J, mRNA of Ox2r (red). K, Merged image of (I) and (J). Cell nuclei are shown in blue. The green, red, and yellow arrows denote the histidine decarboxylase-, Ox2r-, and both-expressing cells, respectively. Scale bar: 50 μm. All sections every 120 μm from a single mouse were quantified. Data of all sections are provided in Extended Data Figures 23-1, 23-2, 23-4, 23-5, 23-6, and 23-7.
Figure 23-1
Distribution of orexin receptor-expressing dopaminergic/adrenergic neurons (1/6). White, green and red circles indicate receptor-negative, Ox1r-positive, and Ox2r-positive dopaminergic/adrenergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 23-1, TIF file.
Figure 23-2
Distribution of orexin receptor-expressing dopaminergic/adrenergic neurons (2/6). White, green and red circles indicate receptor-negative, Ox1r-positive, and Ox2r-positive dopaminergic/adrenergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 23-2, TIF file.
Figure 23-3
Distribution of orexin receptor-expressing dopaminergic/adrenergic neurons (3/6). White, green and red circles indicate receptor-negative, Ox1r-positive, and Ox2r-positive dopaminergic/adrenergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 23-3, TIF file.
Figure 23-4
Distribution of orexin receptor-expressing dopaminergic/adrenergic neurons (4/6). White, green and red circles indicate receptor-negative, Ox1r-positive, and Ox2r-positive dopaminergic/adrenergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 23-4, TIF file.
Figure 23-5
Distribution of orexin receptor-expressing dopaminergic/adrenergic neurons (5/6). White, green and red circles indicate receptor-negative, Ox1r-positive, and Ox2r-positive dopaminergic/adrenergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 23-5, TIF file.
Figure 23-6
Distribution of orexin receptor-expressing dopaminergic/adrenergic neurons (6/6). White, green and red circles indicate receptor-negative, Ox1r-positive, and Ox2r-positive dopaminergic/adrenergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 23-6, TIF file.
Figure 23-7
Distribution of Ox2r-expressing histaminergic neurons White and red circles indicate Ox2r-negative and Ox2r-positive histaminergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 23-7, TIF file.
Noradrenergic neurons in the pons, called A4/A6 (LC) and A7 groups, showed strong and dense expression of Ox1r (Fig. 23E,F) but no expression of Ox2r (LC Ox1r, 67.3%; A7 Ox1r, 57.1%). These features were well contrasted with the lack of expression of receptors of TH-ir neurons in the medulla, which are presumed to be adrenergic and/or noradrenergic neurons (Fig. 23G, Extended Data Fig. 23-6).
Histaminergic neurons showed a scattered distribution throughout the hypothalamus but clustered in the TMV (Fig. 23H). Strong expression of Ox2r was observed in almost all histaminergic neurons (97.3%, Fig. 23H–K, Extended Data Fig. 23-7), while no expression of Ox1r was found.
Cholinergic neurons
In the cerebral nuclei, a moderate proportion of cholinergic neurons expressed either Ox1r or Ox2r (Fig. 24). A total of 15.5% of Chat-positive neurons in the MS and NDB were Ox1r-positive, whereas 3.4% of those were Ox2r-positive. In the Cpu, there were scattered large Chat-positive neurons, approximately one-third of which were Ox2r-positive (34.9%). Although Chat-positive neurons were also found in layers II/III of the isocortex and MHb, almost all of them were receptor-negative (layers II/III, 99.3%; MHb, 99.7%).
Orexin receptor expression in cholinergic neurons. A,C,E,G,I,M, A reconstructed image showing Ox1r- and Ox2r-positive cholinergic neurons. The open, green, red, and yellow circles denote no receptor-, Ox1r-, Ox2r-, and both receptor–positive cells, respectively. Scale bars: 500 μm. B, Ox1r and Chat expression in the pedunculopontine nucleus. D, Ox2r and Chat expression in the parabigeminal nucleus. F, Ox1r and Chat expression in the laterodorsal tegmental nucleus. H, Ox2r and Chat expression in the lateral part of the facial motor nucleus (7N). J–L, Ox1r, Ox2r and Chat expression in the ambiguus nucleus, compact. N–P, Ox1r, Ox2r, and Chat expression in the dorsal motor nucleus of the vagus nerve (10N). Chat mRNA is shown in blue. The green, red, and yellow arrows denote Ox1r-, Ox2r-, and both receptor–expressing cells, respectively. The outlined arrowheads denote the no-receptor–expressing Chat-positive neurons. Scale bars: 50 μm. Q, Regional differences in the expression of receptor subtypes in cholinergic neurons. The numbers above the bar plot indicate the total number of cells. All sections every 120 μm from a single mouse were quantified. Data of all sections are provided in Extended Data Figures 24-1 and 24-2.
Figure 24-1
Distribution of orexin receptor-expressing cholinergic neurons in the brainstem (1/2). White, green, red, and yellow circles indicate receptor-negative, Ox1r-positive, Ox2r-positive, and both Ox1r- and Ox2r-positive cholinergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 24-1, TIF file.
Figure 24-2
Distribution of orexin receptor-expressing cholinergic neurons in the brainstem (2/2). White, green, red, and yellow circles indicate receptor-negative, Ox1r-positive, Ox2r-positive, and both Ox1r- and Ox2r-positive cholinergic neurons, respectively. Panels are arranged in anterior-posterior order. Scale bar: 500 μm. Download Figure 24-2, TIF file.
In the brainstem, very high expression of Ox1r and/or Ox2r was found. The 3N has been known to be a small cluster of Chat-positive neurons, and Ox2r-positive neurons were dominant in the 3N (Ox1r, 4.7%; Ox2r, 33.7%). In the PBG, there was a dense cluster of small neurons that expressed both Chat and Vglut2 mRNAs. These neurons also expressed Ox2r in a high proportion (52.6%), but the expression level per cell was relatively low (Fig. 24C,D). In contrast to these Ox2r-positive nuclei in the midbrain, almost all Chat-positive neurons in the PPT showed highly intense expression of Ox1r (Fig. 24A,B). Similar to PPT Chat neurons, Chat-positive neurons in the LDT also show strong expression of Ox1r with a high proportion (PPT, 85.8%; LDT, 76.9%; Fig. 24E,F). The 5N and PBN in the pons showed a small percentage of receptor-positive cells (Ox1r, 8.1%; Ox2r, 9.6% in 5N; Ox1r, 10.9% in PBN; Extended Data Fig. 24-1). In the 6N and 7N, there was dominant expression of Ox2r with high intensity, and a subregional difference in Ox2r expression was found between the medial and lateral 7N (Ox2r in 6N, 33.3%; 7N medial, 12.6%; 7N lateral, 56.6%; Fig. 24G,H). Subregional differences in receptor expression were also found in the AMB. The AMBc was a relatively large cluster of Chat-positive cells located in the anterior portion of the AMB, and almost all neurons expressed Ox1r and/or Ox2r (Ox1r, 12.0%; Ox2r, 56.0%; Ox1r and Ox2r, 24.0%; Fig. 24I–L). In contrast, the Chat-positive neurons in the AMBsc, the posterior small cell cluster of AMB, showed no expression of the receptor mRNAs. Similar to AMBc, the proportion of Ox1r- and Ox2r–double-positive cells was high in the 10N (Ox1r, 20.5%; Ox2r, 23.6%; Ox1r and Ox2r, 16.4%; Fig. 24M–P). The expression profile in the 12N was different from that in the 10N and 18.1% of the Chat-positive neurons in the 12N expressed Ox2r.
Discussion
In the present study, the bHCR system was used to visualize Ox1r and Ox2r mRNA, combined with other neuron type markers, which provides comprehensive information on the distribution of orexin receptor mRNA and neuron types expressing orexin receptors in mouse brains. We showed that many brain regions express both Ox1r and Ox2r mRNAs, but only a few cells express both Ox1r and Ox2r mRNAs. Double receptor-positive cells are observed in certain brain regions, such as the DRN, LM, BAR AMBc, VMH, and 10N. Since orexin receptors are promiscuous GPCRs (Inoue et al., 2019), how downstream signaling of double receptor–positive cells differs from that of single receptor–positive cells needs to be determined individually. The subcellular localization of OX1R and OX2R may be different.
We also showed that Ox1r-expressing cells and Ox2r-expressing cells in the same region tend to be classified into different subpopulations, such as glutamatergic and GABAergic neurons. For example, Ox1r and Ox2r were expressed predominantly in Vgat-negative cells and Vgat-positive cells, respectively, in isocortex layer II/III, olfactory areas, and the cortical subplate in the cerebrum, as shown in the BLA (Yaeger et al., 2022). In the cerebrum, although Vglut2 was not expressed in almost all of the excitatory neurons, almost all of the Vgat-negative neurons expressed Vglut1. Therefore, we assumed that Vgat-negative neurons were excitatory neurons in the cerebrum.
In the thalamus and more caudal brain regions, the majority of excitatory neurons express Vglut2. Vglut1 is also expressed in certain neural groups along with Vglut2, such as PG, TRN, PRPv, nucleus X, LRN, and LIN. The ECU and GRC express only Vglut1. Interestingly, these Vglut1-positive regions strongly expressed Ox2r and are Barhl1-lineage cells (Rose et al., 2009). Barhl1 is a homeobox transcription factor and is expressed in a subset of developing and mature glutamatergic cells in the hindbrain and the cerebellum. Exceptionally, the VCO, which expresses Vglut1 but is not the Barhl1-lineage (Rose et al., 2009), did not express Ox2r. These findings imply that Barhl1-lineage cells constitute a certain subpopulation of hindbrain excitatory neurons that are characterized by Vglut1 and strong Ox2r expression.
Regarding the wake-promoting action of the orexin system (Liu and Dan, 2019), each orexin receptor exhibits a distinct distribution in many wake-active or wake-promoting neurons. Ox1r is abundantly expressed in the DRN, LC, PPT, and LDT, mildly expressed in the NDB, VTA, and parabrachial nucleus, and not expressed in the histaminergic neurons of the TMV. Ox2r is abundantly expressed in the NDB, histaminergic TMV neurons, DRN, and PPT, mildly expressed in the VTA and parabrachial nucleus, and not expressed in the LC or LDT. These observations are generally consistent with those of previous studies (Lu et al., 2000; Mieda et al., 2011; Ch’ng and Lawrence, 2015; Ikeno and Yan, 2018; Xiao et al., 2021; Schneeberger et al., 2022; Yaeger et al., 2022). Furthermore, the cellular resolution analysis showed that the DRN expresses both Ox1r and Ox2r but serotonergic cells expressing both receptors are a minor population of DRN neurons. Consistently, single-cell RNA-seq of Pet1-lineage DR neurons showed that cell clusters expressing Ox1r are largely different from cell clusters expressing Ox2r (Okaty et al., 2020). Ox1r-expressing clusters express Gad2, whereas Ox2r-expressing clusters express Vglut3.
Electrophysiological studies showed that the application of orexin led to the activation of orexinergic neurons via orexin receptors expressed on local glutamatergic neurons or orexinergic neurons (Li et al., 2002; Yamanaka et al., 2010). In the current study, orexin neurons themselves rarely expressed orexin receptors (Fig. 21), which suggests that orexin neurons do not augment their activity via autoreceptor feedback.
The expression of the orexin receptor in wake-promoting neurons supports the role of orexins in wakefulness, but orexin receptors are also expressed in NREM sleep-promoting neurons (Liu and Dan, 2019). Ox2r is abundantly expressed in the MPOA, including the ventrolateral POA, and in the ventrolateral periaqueductal gray (vlPAG), subthalamic nucleus, and nucleus of the solitary tract, caudoputamen, and nucleus accumbens. Given that orexin neurons are generally inactive during NREM sleep (Mileykovskiy et al., 2005; Hassani et al., 2009), orexin receptor signaling in these sleep-promoting neurons may be involved more in other behaviors during wakefulness than in sleep regulation.
The identification of neural populations involved in REM sleep and its abnormalities, such as SLD, PPT, and LDT, continues, and orexin receptors are expressed in many of these neural populations (Dauvilliers et al., 2018; Iranzo, 2018; Liu and Dan, 2019). Orexinergic neurons project to SLD glutamatergic neurons, which express Ox1r and/or Ox2r. Orexin-excited SLD neurons and activation of these neurons promoted a REM sleep-related symptom, cataplexy, in the orexin-deficient mice (Torontali et al., 2019; Feng et al., 2020). SLD glutamatergic neurons project to glycinergic/GABAergic neurons in the ventromedial medulla including the PGRNI (Weber et al., 2015; Uchida et al., 2021), which express moderately Ox1r and mildly Ox2R. Glycinergic/GABAergic neurons in the ventromedial medulla project to spinal motor neurons and are involved in the generation of muscle atonia during REM sleep (Uchida et al., 2021). In the PPT and LDT, a vast majority of cholinergic neurons only express Ox1r. Whereas the LDT did not express Ox2r, the PPT contained abundant Ox2r-expressing cells that were glutamatergic, not cholinergic. Furthermore, the basolateral amygdala, which is involved in the initiation of REM sleep (Hasegawa et al., 2022), has a dense distribution of Ox1r-positive glutamatergic cells. Thus, orexin may regulate REM sleep through its receptors at various locations in the neural circuits that regulate REM sleep, and loss of orexin may therefore be associated with a variety of abnormalities related to REM sleep.
Regarding the metabolic effects of the orexin system, orexin receptors are abundantly expressed in brain regions regulating energy metabolism, such as the medullary raphe, DMH, and ARC. More than 50% of serotonergic neurons in the medullary raphe, including the raphe pallidus nucleus (B1), raphe obscurus nucleus (B2), and raphe magnus nucleus (B3), express Ox1r but not Ox2r, which serve as sympathetic premotor neurons to enhance thermogenesis by brown adipose tissue (BAT; Tupone et al., 2011; Nakamura et al., 2021). The DMH is one of the major upstream regions of medullary raphe neurons, and the excitatory projection from the DMH to the medullary raphe enhances sympathetic outflows. However, Ox2r-expressing DMH cells were mainly GABAergic. The role of DMH GABAergic neurons in the regulation of sympathetic and other functions remains unclear. In the ARC, GABAergic neurons express Ox2r but not Ox1r. Orexin activates ARC GABAergic neurons (Burdakov et al., 2003) and suppresses POMC neurons, a subpopulation of ARC GABAergic neurons (Ma et al., 2007). Since the loss of either Ox1r or Ox2r only partially mimics the obesity-prone phenotype of orexin-deficient mice (Kakizaki et al., 2019), it is suggested that both Ox1r and Ox2r may function independently to render mice less prone to obesity.
The orexin system is also involved in reward behavior and addiction (Mehr et al., 2021; James and Aston-Jones, 2022). Pharmacological studies using orexin and orexin receptor antagonists have repeatedly reported the role of OX1R in the VTA in reward-related behavior (James et al., 2011; Zarepour et al., 2014; Bentzley and Aston-Jones, 2015; Terrill et al., 2016; Pantazis et al., 2022). However, the expression level of Ox1r per cell was very low, and <20% of dopaminergic cells expressed Ox1r. It may be possible that orexin receptors are functional even if the number of mRNA copies per cell is very low (Leonard and Kukkonen, 2014). In addition, orexin may change the presynaptic function of VTA dopaminergic cells (Baimel and Borgland, 2015). The nucleus accumbens contained a dense population of Ox2r-expressing cells. Orexin modulates risk-avoidance and reward behaviors through nucleus accumbens neurons (Thorpe and Kotz, 2005; Blomeley et al., 2018). OX2R in the nucleus accumbens modulates anxiety (Li et al., 2021) and nociception (Yazdi et al., 2016).
High expression of Ox2r with variable Ox1r expression is observed in cholinergic motor neurons in the brainstem, such as the lateral facial nucleus (7N) that regulates muscles for facial expression, and the ambiguus nucleus compact part (AMBc) that innervates the pharyngeal constrictor and the cervical esophageal muscles (McGovern and Mazzone, 2010; Takezawa et al., 2018; Coverdell et al., 2022). In contrast, the medial 7N and ambiguus nucleus semicompact part (AMBsc), which are not associated with motor regulation (Martin and Lodge, 1977; Klein and Rhoades, 1985; McGovern and Mazzone, 2010; Coverdell et al., 2022), did not show Ox2r expression. Ox1r and Ox2r are expressed in cholinergic neurons of the dorsal motor nucleus of the vagus nerve (10N), which serves parasympathetic functions in many thoracic and abdominal organs. These findings suggest that orexin receptor signaling, mainly via OX2R, works to enhance motor action and parasympathetic activity.
In this study, we visualized orexin receptor mRNA in the brains of male mice. However, female mice could exhibit a different distribution of orexin receptors than males, especially in the hypothalamus. In female rats, Ox1r expression in the hypothalamus changed with the estrous cycle, but Ox2r expression was stable (Wang et al., 2003). Consistently, the regions expressing sex steroid hormone receptors, such as the MPOA, BNSTpr, VMH, MeA, and PAG, also abundantly expressed Ox1r. Sexual differences in orexin receptor expression will be the focus of future studies.
The bHCR system allows us to detect target molecules with much higher sensitivity (Xu and Zheng, 2016; Bi et al., 2017, 2015; Liu et al., 2018; Xu et al., 2019), and this is the first report to apply branched HCR to the in situ detection of multiple targets. The time required for obtaining confocal images was reduced due to the ability to perform highly sensitive detection by branched in situ HCR, allowing us to inspect many brain sections within a shorter time. Split-initiator probes also suppressed noise signals in the branched in situ HCR, as confirmed by staining sections of Ox1r and Ox2r-deficient mice. Very few cells expressing both receptors indicate that the signals detected in this study are highly specific.
In summary, the detailed distribution of orexin receptors helps interpret the consequences of optogenetic manipulation of orexin circuits and provides valuable insight into the development of orexin receptor agonists and antagonists that are therapeutically useful not only for human sleep-related disorders but also for other conditions, including disorders of consciousness following traumatic brain injury.
Footnotes
The authors declare no competing financial interests.
We thank N. Ohno, H. Arai and A. Iijima for animal care. This work was supported by the Precise Measurement Technology Promotion Foundation (to Y.T.), JSPS Kakenhi (21K06414 to Y.T.), a research grant from Uehara Memorial Foundation (to H.F.), and Toho University Grant for Research Initiative Program (TUGRIP).
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license, which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.
References
In this issue
