Nos1 neurons in the paraventricular hypothalamic area modulate lipid metabolism via the sympathetic nervous system in male mice

Kunio Kondoh; Emi Hashiuchi; Yuka Inaba; Hiroshi Inoue; Serika Yamada; Kazunari Miyamichi; Yuki Yoshimura; Ken-ichiro Nakajima; Takeshi Y Hiyama; Yasuhiko Minokoshi

doi:10.1038/s41467-026-74362-9

. 2026 Jun 29;17:8073. doi: 10.1038/s41467-026-74362-9

Nos1 neurons in the paraventricular hypothalamic area modulate lipid metabolism via the sympathetic nervous system in male mice

Kunio Kondoh ^1,^2,^✉, Emi Hashiuchi ^3,⁴, Yuka Inaba ^3,⁴, Hiroshi Inoue ^3,⁴, Serika Yamada ⁵, Kazunari Miyamichi ⁵, Yuki Yoshimura ², Ken-ichiro Nakajima ^1,⁶, Takeshi Y Hiyama ², Yasuhiko Minokoshi ^1,^7,^✉

PMCID: PMC13454591 PMID: 42373618

Abstract

The selection of the appropriate energy substrate under different physiological conditions is a key aspect of the energy metabolism homeostasis. We here show that Nos1 (nitric oxide synthase 1)-expressing cells in the paraventricular hypothalamic nucleus (PVH) serve as a pivotal node for controlling whole-body fat consumption in male mice. Nos1 neurons account for ~30% of PVH neurons that convey signals via polysynaptic pathways to individual peripheral tissues, including skeletal muscle and brown (BAT) and white (WAT) adipose tissues. Activation of these Nos1 neurons in the PVH area induces WAT lipolysis and fat oxidation in other peripheral tissues via the sympathetic nervous system, thereby increasing whole-body fat consumption. Inhibition of these neurons abolishes the increase in fat consumption during the light period, whereas long-term silencing lead to obesity independent of energy intake. These neurons are also necessary for cold-induced thermogenesis in BAT and the rapid increase in fat consumption elicited by food deprivation or other stressors. Nos1 neurons in the PVH area are therefore essential for controlling fat consumption and energy homeostasis.

Subject terms: Fat metabolism, Neurophysiology, Autonomic nervous system, Cellular neuroscience

The proper regulation of energy substrate consumption is essential for energy homeostasis and animal health. Here, the authors show that Nos1-expressing neurons in the PVH area control whole-body fat consumption and cold-induced thermogenesis under various physiological conditions.

Introduction

Most organisms rely on maintaining energy metabolism within a specific range for survival. Disruption of this energy balance can lead to disease conditions like obesity and diabetes mellitus, both of which are major health issues worldwide. The proper regulation of peripheral metabolism, along with energy intake, is essential for achieving energy homeostasis¹. This regulation involves the selective use of macronutrients, such as glucose and fat, in response to various environmental or physiological changes. Dysregulation of energy substrate selection has been linked to several diseases, including diabetes, heart failure, and cancer^2–4.

The direct action of humoral factors such as insulin on peripheral tissues greatly influences energy substrate selection⁵. However, acting via the autonomic nervous system, the brain also plays a vital role in energy metabolism and substrate selection^6,7. The hypothalamus—in particular, the paraventricular nucleus of the hypothalamus (PVH)—contains neurons that control the autonomic nervous system and glucose and fat metabolism and thereby contribute to energy homeostasis and substrate selection. Retrograde transsynaptic tracing studies using pseudorabies viruses (PRVs) have demonstrated that the PVH is highly enriched with neurons positioned upstream of the peripheral nervous system innervating multiple tissues^8–10. PVH neurons that express the melanocortin 4 receptor (MC4R) and receive inputs from both agouti-related peptide (AgRP)- and proopiomelanocortin (POMC)- expressing neurons in the arcuate nucleus of the hypothalamus have been shown to regulate glucose and lipid metabolism and are required for the proper control of peripheral metabolism^11,12. In addition, PVH neurons expressing nitric oxide synthase 1 (Nos1) and insulin receptor substrate 4 (Irs4) have recently been shown to regulate food intake and energy expenditure^13,14. Corticotropin-releasing hormone (CRH)- and oxytocin-positive neurons of the PVH also regulate the autonomic nervous system^15–17.

To identify PVH neurons that regulate peripheral tissues via the autonomic nervous system without bias, we performed retrograde transsynaptic tracing using PRVs in mice. We found that cells positive for Nos1 comprise about 30% of PRV-infected PVH neurons that connect to peripheral tissues via the peripheral nervous system, including the soleus (red skeletal muscle), BAT, and inguinal white adipose tissue (ingWAT). When combined with all individual peripheral tissues, Nos1-expressing neurons probably account for a large proportion of PRV-infected PVH neurons that connect to peripheral tissues via the peripheral nervous system. These Nos1-positive neurons were distinct from PVH neurons previously shown to regulate peripheral metabolism, such as those positive for MC4R, CRH, or oxytocin. Activating the Nos1-expressing neurons in and around the PVH area enhanced lipolysis in WAT and fat oxidation in various peripheral tissues via the sympathetic nervous system, leading to body fat consumption. The activity of Nos1-positive neurons was crucial for fat consumption during the light (resting) period, cold-induced thermogenesis, and the rapid increase in fat use caused by food deprivation and other stressors. Our findings thus demonstrate that fat consumption driven by Nos1-positive neurons in the PVH area is vital for controlling substrate utilization and maintaining energy balance.

Results

Nos1 cells constitute a substantial population of PVH neurons that convey signals via neural pathway to peripheral tissues

We used a PRV-based retrograde transsynaptic tracer system to identify hypothalamic neurons that transmit signals via wired connections to skeletal muscle and BAT, two peripheral tissues that significantly contribute to energy use. PRVs encoding green (GFP) or red (RFP) fluorescent proteins were injected into the both sides of the soleus or BAT, respectively, and the location of the virus-infected neurons in the hypothalamus was examined (Supplementary Fig. 1a). Immunohistofluorescence analysis showed that PRV-infected neurons expressing the fluorescent proteins were mostly found in the PVH and lateral hypothalamus (LH) at 3 days after virus injection (Supplementary Fig. 1b). Although they appeared in many hypothalamic regions at 4 days (Supplementary Fig. 1b–d), the PVH still contained the highest number of PRV-infected cells among hypothalamic regions after virus injection into the soleus or BAT. In each hypothalamic area, PRV-infected neurons were observed on both sides without clear bias. Some PRV neurons expressed both GFP and RFP, while others expressed only one of the fluorescent proteins. These findings suggest that PVH neurons are involved in regulating the soleus and BAT via the autonomic nervous system. We therefore focused on PRV-infected PVH neurons that convey signals via neural pathways to peripheral tissues, including the soleus and BAT.

With the use of the Allen Brain Atlas (https://mouse.brain-map.org), we examined the expression of molecular marker genes in the PRV-infected cells of the PVH. We injected a PRV encoding GFP into the soleus and BAT (Fig. 1a). GFP expression was detected in neurons of the dorsal, periventricular, and ventral portions of the PVH (Fig. 1b, d, Supplementary Fig. 1e). Some of these PVH areas have been shown to project to the intermediolateral column (IML) of the spinal cord¹⁸. The combination of immunohistofluorescence staining and in situ hybridization revealed that most PRV-infected PVH neurons are glutamatergic (67%), while a smaller portion are GABAergic (5%) (Fig. 1b, c). We note that approximately ~28% of PRV-infected neurons did not exhibit detectable expression of either excitatory or inhibitory marker genes (Fig. 1c). A similar proportion of marker-negative PRV-infected neurons was observed in our previous study¹⁹. One plausible explanation is that viral infection may reduce host gene transcription in a subset of cells²⁰, thereby limiting mRNA detectability at the time of analysis.

We found that about 26% of the PRV-infected cells in the PVH expressed Nos1 (Fig. 1d, e). In contrast, approximately 12% of PRV-infected neurons express Mc4r, consistent with previous reports suggesting MC4R-expressing PVH neurons play an important role in energy metabolism (Fig. 1d, e)^11,12,21. Cells expressing other marker genes—including those for glucagon-like peptide-1 receptor (Glp1r), oxytocin (Oxt), CRH (Crh), cocaine- and amphetamine-regulated transcript (Cartpt), proenkephalin (Penk), arginine vasopressin (Avp), thyrotropin-releasing hormone (Trh), and Irs4—showed little overlap with the PRV-infected cells (Fig. 1e, Supplementary Fig. 1e). Furthermore, the distribution of Nos1-expressing neurons (Nos1 neurons) in the PVH differed from that of neurons expressing CRH or MC4R (Supplementary Fig. 1f, g). Indeed, Mc4r expression was observed in only 1.7% of Nos1 neurons, and Nos1 expression was observed in only 2.4% of Mc4r-expressing neurons (Supplementary Fig. 1f). These findings suggest that Nos1 neurons form a distinct population separate from MC4R- and CRH-expressing neurons known to regulate energy metabolism^21,22.

We next examined whether Nos1 neurons in the PVH project to other peripheral tissues, including the liver, pancreas, and subcutaneous WAT (ingWAT). Injection of a PRV encoding GFP into these tissues revealed that ~30% of the infected cells in the PVH were Nos1 positive in each tissue (Fig. 1f, g).

Previous studies have reported that a population of PVH Nos1 neurons projects directly to the IML of the spinal cord, where they innervate presympathetic neurons, and other brain regions¹³. We mapped Nos1 neuron projections in detail by injecting an adeno-associated virus (AAV) that expresses membrane-bound EGFP and synaptophysin-tagged mRuby in a Cre-dependent manner into the PVH of Nos1-Ires-Cre (Nos1-Cre) mice (Fig. 1h, i). Nos1 neurons exhibited robust projections to the spinal IML, with preferential innervation at T1–T3 and T7–T12 spinal levels (Fig. 1j), which contain sympathetic preganglionic neurons innervating adipose tissues^23,24. In addition, Nos1 neurons projected to several brain regions, with particularly dense innervation of the nucleus of the solitary tract (NTS) and area postrema (AP), as well as projections to the median eminence (Supplementary Fig. 2a). These observations suggest that Nos1 neurons in the PVH area project to the central nervous and endocrine systems, as well as the sympathetic nervous system.

The above results suggest that Nos1 neurons in the PVH area form complex networks that regulate peripheral tissues. The results show that approximately 30% of the infected cells in the PVH are Nos1-positive in each tissue (Fig. 1f, g), suggesting that a large proportion of PRV-infected cells are Nos1-positive and project to multiple peripheral tissues. Although it is technically challenging to precisely estimate the percentage of Nos1 neurons among all PVH neurons projecting to peripheral tissues, these results suggest that Nos1 neurons in the PVH are among the key neurons regulating metabolic tissues via the autonomic nervous system, along with previously identified neurons.

To determine whether individual PVH Nos1 neurons project to one or multiple tissues, we examined GFP- and RFP-positive cells in the PVH after injecting PRV into the soleus and BAT, respectively (Supplementary Fig. 2b). Approximately 15% of PRV-infected Nos1 cells expressed both GFP and RFP, while the rest expressed only GFP or RFP (Supplementary Fig. 2c, d). Similar results were observed when PRVs were injected into ingWAT and BAT (Supplementary Fig. 2e, f). These findings suggest that Nos1 neurons are heterogeneous, with some targeting multiple peripheral tissues and others being more tissue-specific. Targeting of multiple peripheral tissues by Nos1 neurons in the PVH area may be mediated at the spinal cord and/or other brain regions connected to the Nos1 neurons.

Activation of Nos1 neurons in the PVH area increases whole-body fat consumption

To explore the potential functions of PVH Nos1 neurons in energy metabolism, we used a chemogenetic method to artificially activate them. An AAV that expresses hM3Dq-mCherry—a fusion of the fluorescent protein mCherry and an excitatory DREADD (designer receptor exclusively activated by designer drug)—in a Cre-dependent manner was injected into the PVH of Nos1-Cre mice (Fig. 2a). Most mCherry-expressing cells were Nos1-positive (Fig. 2b, c; Supplementary Fig. 3a), and no mCherry expression was found when the same AAV was injected into the PVH of C57BL/6 J wild-type mice (Supplementary Fig. 3b), confirming the specificity of hM3Dq expression in Nos1 neurons. Since mCherry was present in both the PVH and peri-PVH areas (Fig. 2b and Supplementary Fig. 3a), we defined hM3Dq-mCherry-expressing Nos1 neurons as “Nos1 neurons in the PVH area”.

The resulting Nos1-hM3Dq mice were injected intraperitoneally (ip) with the hM3Dq ligand clozapine-N-oxide (CNO) or saline (control) during the light period (zeitgeber time [ZT] 3). Administration of CNO increased the expression of c-Fos, a neuronal activation marker, in and around hM3Dq-expressing neurons in the PVH area (Supplementary Fig. 3c). Chemogenetic activation of the Nos1 neurons significantly decreased the respiratory quotient (RQ) for up to 6 h (Fig. 2d). Consistent with this effect on the RQ, the rate of fat consumption, calculated from O₂ consumption (VO₂) and CO₂ production (VCO₂), was increased after Nos1 neuron activation (Fig. 2e), whereas the rate of CHO consumption was decreased (Fig. 2f). Activation of Nos1 neurons in the PVH area also raised energy expenditure and physical activity (Fig. 2g, h), as previously described¹³. Additionally, resting energy expenditure, but not the slope, derived from plotting energy expenditure against physical activity for individual mice, was significantly increased after Nos1 neuron activation (Supplementary Fig. 3d, e) These results suggest that activating these neurons increases energy expenditure through both physical activity-dependent and independent mechanisms. Notably, although marked changes in fat consumption occurred during the 2–6 h period, CHO consumption tended to increase during the 0–2 h period and then decreased during the 2–6 h period. Energy expenditure and physical activity also only increased during the 0–2 h period. These findings suggest that Nos1 neurons in the PVH area regulate fat consumption through a mechanism distinct from those controlling CHO consumption and energy expenditure.

As reported previously¹³, Nos1 neuron activation decreased food intake during the light period (Fig. 2i). Because this decrease in food intake during the light period is small, it is unlikely to be the main factor in increasing fat consumption by activating Nos1 neurons in the PVH. Furthermore, in Fig. 5 and Supplementary Fig. 7, we examined the effects of the food-deprived condition on fat consumption and other metabolic parameters under Nos1 neuron inhibition in the PVH during the light period. The results indicated that changes in food intake primarily do not drive changes in fat consumption by modulating Nos1 neuron activity in the PVH area.

Fig. 5 — a Expression of hM4Di-mcherry in Nos1 neurons in the PVH area. b Representative PVH area images showing immunofluorescence for mCherry-fused hM4Di (magenta) and Nos1 (green) in a Nos1-Cre mouse injected with AAV-FLEX-hM4Di-mCherry (Nos1-hM4Di mouse). Dotted lines indicate the approximate PVH location. Scale bar, 200 µm. c The proportion of mCherry-positive cells in the PVH area that expressed Nos1 (n = 4 mice). **d–i** Nos1-hM4Di mice were injected with CNO or saline (t = 0) at ZT3 (**d–f**) or ZT12 (**g–i**), and the RQ (d, g), fat consumption (e, h), and food intake (f, i) were measured. Each left panel shows values for consecutive 5-min windows, and each right panel shows the average for the 6-h intervals after such administration (**d, e, g, h**) (n = 14 mice). The graphs show the cumulative food intake at the indicated time (f, i) (n = 6 mice). j, k, Nos1-hM4Di mice were injected with CNO or saline (t = 0) at ZT3 in the absence of food, and RQ (j) and fat consumption (k) were measured. Each left panel shows the values for consecutive 5-min windows, and each right panel shows the average values for consecutive 2-h intervals (n = 8 mice). **l–o** RQ (l), fat consumption (m), CHO consumption (n), and energy expenditure (o) of C57BL/6 J wild-type mice injected with vehicle (control), a mixture of phentolamine, propranolol, and L748,337 (Phe + Prop + L), or propranolol (Prop) alone. Each left panel shows values for consecutive 5-min windows with antagonists being injected at ZT4. Each right panel shows the average for ZT4-6 (n = 7 mice). Data are means ± s.d. Statistical significance for comparisons between saline and CNO administration, or saline and drug administration was determined by two-way ANOVA and Sidak’s post hoc test (f, j, k), one-way ANOVA and Sidak’s post hoc test (**l-o**), or the two-sided paired Student’s t test (d, e). Source data are provided as a Source Data file.

During the dark period (ZT12), activation of Nos1 neurons in the PVH area by injecting Nos1-hM3Dq mice with CNO produced effects similar to RQ and fat and CHO consumptions seen during the light period (Supplementary Fig. 3f–h). However, activation of Nos1 neurons in the PVH area only slightly increased energy expenditure during the initial period 0–2 h, and there was no significant impact on physical activity (Supplementary Fig. 3i, j). These results also indicate that Nos1 neurons in the PVH area regulate fat consumption independently of physical activity.

In addition, although activating Nos1 neurons mostly reduced food intake (Supplementary Fig. 3k), the decrease in food intake is unlikely to be the main reason for increased fat consumption. Fat consumption returned to control levels after 4 h, even though food intake remained much lower than in control mice at that time. Overall, these findings suggest that Nos1 neurons in the PVH area regulate fat consumption independently of physical activity or food intake.

Activation of Nos1 neurons in the PVH area during the light period significantly increased plasma levels of nonesterified fatty acids (NEFAs) and noradrenaline (NA) at 3 h after CNO administration (Supplementary Fig. 4a). Adrenaline also tended to increase. The rise in plasma NEFA levels lasted at least the following 2 h and was not accompanied by changes in plasma triacylglycerol levels (Supplementary Fig. 4b, c). Blood glucose levels increased up to 2 h after CNO administration (Supplementary Fig. 4d), likely due to increased physical activity and energy expenditure (Fig. 2g, h). It is also likely related to the tendency to increase in CHO consumption during the first 2 h (Fig. 2f). Overall, these findings support the idea that Nos1 neurons in the PVH promote fat consumption and activate the sympathetic nervous system, independent of physical activity or energy expenditure. CHO consumption may be influenced by several factors, including fat consumption, physical activity, and regulation of energy expenditure. While Nos1 neuron-induced fat consumption may partly be associated with the suppression of food intake, changes in food intake would not primarily drive the increased fat consumption caused by Nos1 neuron activation. These results suggest that Nos1 neurons in the PVH area regulate both fat consumption and food intake.

Nos1 neurons in the PVH area increase lipid turnover

To investigate how Nos1 neurons in the PVH area regulate whole-body fat intake, we studied the effects of their activation on fatty acid oxidation in specific peripheral tissues in vivo by intravenously infusing two radioisotope-labeled fatty acid tracers, [³H]-(R)−2-bromopalmitate and [¹⁴C]palmitate^25,26 (Fig. 2j). CNO administration in Nos1-hM3Dq mice (between ZT3 and ZT5) led to a significant increase in the rate of fatty acid oxidation in most examined peripheral tissues, including the heart, liver, BAT, red skeletal muscle, and WAT (Fig. 2k, l), but not in white skeletal muscle, indicating that activation of Nos1 neurons in the PVH area enhances fat oxidation in peripheral tissues. Consistent with this, activating these neurons also increased the mRNA levels of PGC1 (peroxisome proliferator–activated receptor–γ coactivator 1α) (Ppargc1a) and PPAR (peroxisome proliferator–activated receptor–α) (Ppara), both of which are key regulators of fat oxidation, in the soleus, although similar effects in BAT did not reach statistical significance (Supplementary Fig. 4e, f). Conversely, CNO administration did not influence the expression of genes involved in de novo lipogenesis in the liver (Supplementary Fig. 4g).

We next investigated whether Nos1 neuron activation affects lipolytic activity in ingWAT, which has a more enriched sympathetic innervation than epididymal white adipose tissue (epiWAT). Immunoblot analysis showed that activation of Nos1 neurons in the PVH area significantly increased the amount of phosphorylated hormone-sensitive lipase (HSL) in ingWAT (Fig. 2m), suggesting that these neurons increase lipolytic activity in this tissue. Together, these results indicated that Nos1 neurons in the PVH area stimulate lipolysis in ingWAT and fat oxidation in various peripheral tissues. Activation of these neurons thus promotes whole-body lipid turnover in mice.

Nos1 neurons in the PVH area increase fat consumption via the sympathetic nervous system

Because PVH Nos1 neurons were labeled by PRVs transported transsynaptically from multiple peripheral tissues, it is likely that these neurons regulate peripheral metabolism via specific neural circuits. Adipose tissues are mainly regulated by the sympathetic nervous system. Therefore, we examined whether Nos1 neurons in the PVH promote fat utilization via this system.

First, we evaluated the effects of combined adrenergic receptor blockade using a mixture of α1-, β1-, β2-, and β3-adrenergic antagonists (phentolamine [α1], propranolol [β1/β2], and SR59230A [β3]) on Nos1 neuron-induced increases in fat consumption. Nos1-hM3Dq mice received these blockers before CNO or saline administration during the light period. Adrenergic blockade significantly reduced the CNO-induced decrease in RQ and the rise in fat consumption (Fig. 3a, b). It also suppressed the increase in energy expenditure, and CHO consumption returned to control levels during 4–6 h (Supplementary Fig. 5a, b). These results suggest that Nos1 neuron-driven increase in fat consumption is mainly mediated by the sympathetic nervous system.

Fig. 3 — a, b Nos1-hM3Dq mice were injected with a mixture of phentolamine, propranolol, and SR59230A (Blocker) or saline 30 minutes before CNO or saline administration during the light period (ZT3, t = 0), and RQ (a), fat consumption (b) were measured. Each left panel shows the consecutive 5-min window values before and after CNO or saline administration, and each right panel shows the average values for consecutive 2-h intervals (n = 6 mice). c Activation of spinal cord-projecting Nos1 neurons in the PVH area. Two AAVs, AAVrg-FLEX-FLPo and AAV-fDIO-hM3Dq-mCherry, were injected into the T8–T12 spinal cord and the PVH area of Nos1-Cre mice, respectively. This intersectional strategy enables specific expression of hM3Dq-mCherry in spinal cord-projecting Nos1 neurons within the PVH area. d, e Representative image of the PVH area (d) and the spinal cord (e) showing mCherry expression (red) in a mouse shown in (c). Sections were counterstained with Hoechst 33342 (blue) to detect nuclei. Scale bars, 200 µm. Data are representative of 7 mice with similar results. f, g Mice expressing hM3Dq in Nos1 neurons in the PVH area projecting to the spinal cord were injected with CNO or saline at ZT11 (t = 0), and RQ (f), fat consumption (g) were measured. Each left panel shows the consecutive 5-min window values before and after CNO or saline administration, and each right panel shows the average for consecutive 1-h intervals (n = 7 mice). h Nos1-hM3Dq mice were injected with the tyrosine hydroxylase inhibitor as well as with saline or CNO at ZT2 to ZT6, and the changes in NA content of peripheral tissues were determined from measurements made at 0 and 4 h after drug administration (n = 8 mice). Data are means ± s.d. Statistical significance for comparisons between saline and CNO administration was determined by two-way ANOVA and Sidak’s post hoc test (a, b, f, g) or the two-sided unpaired Student’s t test (h). Source data are provided as a Source Data file.

To investigate the role of Nos1 neurons projecting to the spinal cord in fat consumption, we selectively manipulated these neurons in the PVH area using an intersectional viral strategy. A retrograde AAV expressing FLPo recombinase in a Cre-dependent manner was injected into the spinal cord at T8–T12 levels, followed by delivery of an AAV expressing FLPo-dependent hM3Dq-mCherry into the PVH area (Fig. 3c). The hM3Dq-mCherry expression was restricted to PVH neurons, and their axons were observed in the IML of the spinal cord (Fig. 3d, e). Even though Nos1 neurons projecting to the spinal cord at T8–T12 levels are only selectively activated, activation of these neurons quickly decreased RQ and increased fat consumption, with a shorter duration than activation of the entire Nos1 neuron group in the PVH area (Fig. 3f, g). This manipulation led to a modest yet significant increase in energy expenditure (Supplementary Fig. 5d). CHO consumption and physical activity increased after fat consumption returned to the control level (Supplementary Fig. 5e, f). These findings show that activating spinal cord-projecting Nos1 neurons in the PVH is sufficient to enhance fat consumption. Given that Nos1 neurons in the PVH area project to other brain areas and the spinal cord, activating the Nos1 neurons may directly and indirectly stimulate fat consumption via the sympathetic nervous system.

To identify peripheral tissues whose sympathetic innervation is influenced by Nos1 neurons, we employed a noradrenaline (NA) turnover assay in freely moving mice. Nos1-hM3Dq mice were treated with α-methyl-p-tyrosine, a tyrosine hydroxylase inhibitor, combined with CNO or saline during the light period. Activation of Nos1 neurons in the PVH area significantly increased the reduction in NA content over the next 4 h in multiple peripheral tissues, including skeletal muscle (soleus and gastrocnemius), BAT, ingWAT, and pancreas, with minimal effects in the heart, liver, and epiWAT, following CNO treatment (Fig. 3h, Supplementary Fig. 5g). These results suggest that chemogenetic activation of Nos1 neurons in the PVH area activates sympathetic nerves innervating peripheral tissues in a tissue-specific manner.

Nos1 neurons in the PVH area and the sympathetic nervous system are required for the change in fat consumption linked to the diurnal rhythm

To investigate the physiological role of Nos1 neurons in regulating fat metabolism, we silenced their activity and observed the effects on whole-body energy metabolism. An AAV encoding the light chain of tetanus toxin (TeNT) and EGFP in Cre-expressing cells of doxycycline-treated mice was injected into the PVH of Nos1-Cre mice (Fig. 4a). Nos1 neuron activity was then silenced by feeding the mice food containing doxycycline (hereafter referred to as Nos1-TeNT mice). Most EGFP-positive cells were Nos1-positive (Fig. 4b, c, Supplementary Fig. 6a), located in the PVH area, and no EGFP-expressing cells were found when the same AAV was injected into the PVH of C57BL/6 J wild-type mice (Supplementary Fig. 6b). Control mice showed a circadian rhythm in both RQ and fat consumption, with lower RQ and higher fat consumption during the light (resting) period, and higher RQ and lower fat consumption during the dark (active) period (Fig. 4d, e). Conversely, silencing Nos1 neurons raised RQ (Fig. 4d) and nearly eliminated fat consumption (Fig. 4e). CHO consumption was consistently higher in these mice (Fig. 4f). In contrast, silencing Nos1 neurons had little effect on energy expenditure or physical activity (Fig. 4g, h). Additionally, silencing Nos1 neuron activity increased food intake, maintaining the light-dark cycle. This is a contrast: ob/ob, KK-Ay, MC4R knockout, and AgRP knockout mice clearly retain the large RQ transition during the light-dark cycle, despite being hyperphagic^27–30. These findings suggest that a significant reduction in fat consumption throughout the day is not merely a secondary effect of changes in physical activity or energy expenditure, although increased food intake may partly contribute to reduced fat consumption and other metabolic changes. Together, these data show that silencing Nos1 neuron activity in the PVH causes a consistent decrease in fat consumption and an increase in CHO consumption throughout the day.

Fig. 4 — a Chronic silencing of Nos1 neuron in the PVH area. Two AAVs were injected into the PVH of Nos1-Cre mice. AAV-FLEX-TeNT-EGFP confers expression of TeNT and enhanced GFP (EGFP) in Cre-expressing cells in the presence of the transactivator rtTA, whose expression is conferred by AAV-tTS/rtTA in the presence of doxycycline. Control mice were injected with AAV-tTS/rtTA alone. b Representative images of the PVH area showing immunofluorescence for Nos1 (magenta) and EGFP (green) in a Nos1-Cre mouse injected with the two AAVs and provided with doxycycline (Nos1-TeNT mouse) as indicated in (a). Dotted lines indicate approximate location of the PVH. Scale bar, 200 µm. c The proportion of EGFP-positive cells in the PVH area that expressed Nos1 (n = 5 mice). **d–h** The RQ (d), fat consumption (e), CHO consumption (f), energy expenditure (g), and physical activity (h) for Nos1-TeNT and control mice. Each left panel shows the values for consecutive 5-min windows over 48 h (including two light and two dark periods), and each right panel shows the average values for the light or dark periods (each of 12 h) (n = 7 mice). i Total food intake for Nos1-TeNT and control mice during 12-h light or dark periods. Data are the average values for two light and two dark periods (n = 7 and 8 for control and Nos1-TeNT mice, respectively). j NA content in the indicated tissues of Nos1-TeNT or control mice. Tissues were collected at ZT10–11 (n = 5 mice). k Immunoblot analysis of phosphorylated (Phospho-) and total forms of HSL in ingWAT of Nos1-TeNT or control mice (left). Tissues were collected at ZT10–11. The relative phospho-/total HSL band intensity ratio was determined (right) (n = 5 mice). Data are means ± s.d. Statistical significance for comparisons between Nos1-TeNT and control mice was determined by two-way ANOVA and Sidak’s post hoc test (**d–i**) or the two-sided unpaired Student’s t test (j, k). Source data are provided as a Source Data file.

NA content in peripheral tissues correlates with NA turnover rate. Specifically, under chronic steady-state conditions, high tissue NA content associates with increased NA release and synthesis, while low tissue NA content correlates with the opposite. Therefore, we examined NA contents in peripheral tissues of Nos1 neuron activity-silenced mice. Silencing Nos1 neurons in the PVH area with TeNT reduced NA levels in skeletal muscle, BAT, and ingWAT, but not in the heart, liver, and epiWAT (Fig. 4j). In ingWAT, the density of NA fibers remained unchanged (Supplementary Fig. 6c, d). These findings suggest impaired sympathetic activity in these tissues, although sympathetic innervation in ingWAT likely remains intact. Additionally, the levels of mRNAs for genes involved in fat oxidation (Ppargc1a and Cpt1b) decreased in BAT but not in the soleus (Supplementary Fig. 6e, f). Furthermore, phosphorylated HSL levels in ingWAT were lowered after silencing these neurons (Fig. 4k, l). Overall, these results indicate that silencing Nos1 neuron activity reduces sympathetic nerve activity and associated metabolic functions in specific tissues, including skeletal muscle, BAT, and ingWAT. Of note, modulation of Nos1 neurons does not affect NA turnover and NA content in the heart, liver, and epiWAT.

Our findings from TeNT-mediated Nos1 neuron-silenced mice indicate that Nos1 neurons are essential for fat consumption during both light and dark periods. However, TeNT-mediated suppression might lead to metabolic changes that carry over from one period to another. To address this, we transiently suppressed these neurons by targeted expression of the inhibitory DREADD hM4Di (Nos1-hM4Di mice) and CNO treatment during either the light or dark period (Fig. 5a). Most of the hM4Di mCherry-expressing cells were Nos1-positive (Fig. 5b, c, Supplementary Fig. 7a) and located around the PVH area. No hM4Di mCherry-expressing cells were detected when the same AAV was injected into the PVH of C57BL/6 J wild-type mice (Supplementary Fig. 7b). CNO administration during the light period (ZT3) significantly increased the RQ of Nos1-hM4Di mice (Fig. 5d), while it reduced fat consumption (Fig. 5e) and increased both CHO consumption and energy expenditure (Supplementary Fig. 7c, d). Physical activity did not alter (Supplementary Fig. 7e). Silencing Nos1 neurons in the PVH area at ZT3 promoted food intake for 4 hours (Fig. 5f). Conversely, CNO treatment during the dark period (ZT12) had no effect on any aspects of energy metabolism or food intake (Fig. 5g–i, Supplementary Fig. 7f–h). These results suggest that Nos1 neuron activity is necessary for fat consumption during the light period but appears largely unnecessary for energy metabolism during the dark period. Additionally, these metabolic changes are independent of physical activity.

The increase in food intake observed after suppressing Nos1 neuron activity during the light period (Fig. 5f) might explain the decrease in fat consumption. To test this, we inhibited Nos1 neurons in the PVH area under food-deprived conditions. Silencing Nos1 neurons without food produced effects similar to those seen with food present (Fig. 5j, k, Supplementary Fig. 7i–k), while it increased the RQ, CHO consumption, and energy expenditure to a slightly lesser extent than when food was present. These results suggest that the reduction in fat consumption caused by Nos1 neuron silencing during the light period is independent of food intake.

To study the role of endocrine hormones in fat consumption, we measured plasma levels of corticosterone, thyroid-stimulating hormone (TSH), and insulin in food-deprived animals during the light period. Silencing Nos1 neurons in the PVH area did not affect these plasma levels (Supplementary Fig. 7l). Combined with the observed decrease in sympathetic nerve activity after chronic suppression of Nos1 neurons, these findings suggest that silencing Nos1 neuron activity mainly influences fat consumption through the sympathetic nervous system, not through endocrine pathways.

The sympathetic nervous system is less active during rest than during active periods^31,32. We next examined whether sympathetic nerve activity is required for fat consumption during the light period. Subcutaneous administration of a mixture of α1-, β1-, β2-, and β3-adrenergic receptor blockers (phentolamine [α1], propranolol [β1 and β2], and L748,337 [β3]) during the light period (ZT4) caused a significant increase in the RQ and decrease in fat consumption in C57BL/6 J wild-type mice (Fig. 5l, m). In contrast, the adrenergic antagonists did not affect CHO consumption or physical activity during the light period (Fig. 5n; Supplementary Fig. 7m), but they did lower energy expenditure (Fig. 5o). Propranolol alone showed a similar but shorter effect. These results suggest that the sympathetic nervous system promotes fat consumption, but not CHO consumption or physical activity, during the light period.

NOS1 is not required for fat consumption during the light period

NOS1 is an enzyme that catalyzes nitric oxide synthesis from arginine, with NO in turn acting as an intracellular signaling molecule that affects neuronal excitability and gene transcription as well as a retrograde extracellular signaling molecule that regulates presynaptic neuronal excitability^33,34. To examine whether the NOS1 protein expressed in PVH neurons is required for fat consumption, we created Nos1-floxed mice and injected them with an AAV encoding Cre recombinase (Supplementary Fig. 8a) to delete Nos1 in the PVH (Supplementary Fig. 8b, c). The metabolic parameters of these mice did not significantly differ from those of wild-type control mice (Supplementary Fig. 8d–i), indicating that NOS1 protein expressed in Nos1 neurons in the PVH area is dispensable for the induction of fat consumption by these neurons.

Nos1 neurons are active during the light period

The prominent role of Nos1 neuron activity in regulating fat consumption during the light period indicates that these neurons are more active during the light period than during the dark period. To investigate this, we recorded the activity of these neurons in freely moving mice over 36 h using the calcium sensor GCaMP6s and fiber photometry (Fig. 6a, b). The Nos1 neurons in the PVH area frequently showed clustered calcium activity (Fig. 6c). This clustered activity was comparable to what was seen during chemogenetic activation of Nos1 neurons in the PVH area (Supplementary Fig. 9a–d). Such activity patterns occurred more often during the light period than the dark period, although the neurons remained active during the dark (Fig. 6d, e). Consistent with this, the number of PVH Nos1 neurons expressing c-Fos protein was significantly higher during the light period (ZT4–5) than during the dark period (ZT16–17) (Fig. 6f, g). These results indicate that Nos1 neurons in the PVH area are more active during the light period than during the dark period.

Fig. 6 — a Assessment of Nos1 neuron activity in the PVH area in freely moving mice under ad libitum feeding conditions and a 12-h light/12-h dark cycle. b Representative PVH area image showing immunofluorescence for GCaMP6s (green) in a Nos1-Cre mouse injected with AAV-FLEX-GCaMP6s into the PVH. Hoechst 33342-stained nuclei (blue). Scale bar, 200 µm. Data are representative of 8 mice with similar results. c Representative GCaMP6s signal intensity in Nos1 neurons in the PVH area of a mouse treated as in (a). The arrowheads in the right inset indicate clustered activity over 1 h, and the blue bars above indicate the clustered activity. d Clustered activity distribution for Nos1 neurons for eight individual mice. e Number of clustered activities of Nos1 neurons during light (ZT0–12) and dark (ZT12–24) periods. Bars indicate means (n = 8 mice). f Representative PVH images showing immunofluorescence for Nos1 (magenta) and c-Fos (green) in C57BL/6 J wild-type mice. Scale bars, 100 µm. g The proportion of Nos1-positive cells in the PVH that expressed c-Fos. Data are means ± s.d. (n = 6 mice). h Effects of the ambient light on Nos1 neuron activity. i Clustered activity distribution for Nos1 neurons for six individual mice housed as in (h). j, k Number of clustered activities of Nos1 neurons (j) and food intake (k) during intervals corresponding to the light (RZT0–12) and dark (RZT12–24) periods as in (h). Bars indicate means (n = 6 mice (j) or n = 5 mice (k)). l Effects of the light period-restricted feeding on Nos1 neuron activity. m Clustered activity distribution for Nos1 neurons for five individual mice fed as shown in (l). **n, o** Number of clustered activities of Nos1 neurons (n) and food intake (o) for mice fed as shown in (l). Bars indicate means (n = 5 mice (n) or n = 3 mice (o)). Statistical significance for comparisons between the two periods was determined by the two-sided paired Student’s t test. Source data are provided as a Source Data file.

Light exposure regulates many circadian rhythms in the body^35–37. We therefore investigated whether environmental light affects the activity of Nos1 neurons in the PVH area. Mice housed under constant light conditions no longer showed a difference in Nos1 neuron activity between the 12-hour periods that correspond to the light period (relative ZT [RZT] 0 to 12) and the dark period (RZT12–24) (Fig. 6h–k). As shown in Fig. 4i, mice exhibit nocturnal feeding under normal environmental conditions; however, in constant light, they lost their feeding rhythm without a change in daily total calorie intake (Fig. 6k). Therefore, we tested whether this change in feeding timing might influence Nos1 neuron activity. Feeding only during the light period for more than 10 days did not alter the circadian pattern of Nos1 neuron activity (Fig. 6l–o), which remained higher during the light period than during the dark period. These findings suggest that feeding does not affect the activity of Nos1 neurons in the PVH area.

The suprachiasmatic nucleus (SCN) generates circadian rhythms in response to ambient light signals transmitted by retinal ganglion cells^37–39. We investigated whether Nos1 neurons in the PVH area receive inputs from SCN neurons. First, we used rabies virus-based monosynaptic tracing to identify neurons directly upstream of Nos1 neurons in the PVH area (Supplementary Fig. 9e), showing that starter cells in Nos1 neurons were observed in the PVH area (Supplementary Fig. 9f). Although virus-infected cells were observed in several brain areas, including multiple hypothalamic regions, we did not find rabies virus-infected cells in the SCN (Supplementary Fig. 9g–i). This suggests that SCN neurons do not project directly to Nos1 neurons in the PVH area. Next, we examined polysynaptic neural connections using a Cre-dependent PRV (PRVB180), which can infect neurons upstream of Cre-expressing neurons both directly and indirectly⁴⁰ (Supplementary Fig. 9j, k). With this approach, we detected many infected cells in the SCN and other brain regions (Supplementary Fig. 9i, l, m). These results indicate that the activity of Nos1 neurons in the PVH area may be influenced by ambient light signals transmitted indirectly via SCN neuron inputs.

Long-term silencing of Nos1 neurons in the PVH area results in obesity independent of energy intake

To examine the role of Nos1 neuron-mediated regulation of fat metabolism in whole-body energy balance, we studied the effects of long-term silencing of Nos1 neurons in the PVH area on body and tissue weights. Nos1-TeNT mice fed a regular chow diet (CD) containing doxycycline ad libitum for 8 weeks after AAV injection, gained more weight and consumed more food compared to control mice (Fig. 7a, b). They also had higher plasma glucose concentration, whereas their plasma NEFA and triacylglycerol levels remained similar to those of control mice (Supplementary Fig. 10a). The weights of all peripheral tissues examined, except skeletal muscle, were also higher in Nos1-TeNT mice 8 weeks after AAV injection, with WAT and the liver showing the largest increases (Supplementary Fig. 10b). Histological analysis showed more extensive lipid accumulation in the liver, BAT, and ingWAT of Nos1-TeNT mice than in control mice (Fig. 7c–e). These results indicate that long-term silencing of Nos1 neurons in the PVH area results in fat accumulation in peripheral tissues and increased body weight during unrestricted feeding.

Fig. 7 — a Body weight for control or Nos1-TeNT mice during ad libitum feeding with a regular chow diet (CD). Data are means ± s.d. (n = 6 mice). b Average daily food intake during ad libitum CD feeding for 8 weeks as in (a) (n = 6 mice). **c–e** Hematoxylin-eosin staining sections for the liver (c), BAT (d), and ingWAT (e) of control or Nos1-TeNT mice after ad libitum CD feeding for 8 weeks. Scale bars, 200 µm. Data are representative of 4 mice with similar results. **f–h** Cumulative food intake (f), body weight (g), and body weight change (h) for control and Nos1-TeNT mice that were pair-fed with the same amount of an HFD as consumed by control mice. Pair-feeding was initiated at 10 days after AAV injection and was continued to 70 days after AAV injection (n = 8 mice for control, n = 9 mice for Nos1-TeNT). **i–k** Hematoxylin-eosin staining sections for the liver (i), BAT (j), and ingWAT (k) of control and Nos1-TeNT mice pair-fed with HFD for 67 days. Data are representative of 4 mice with similar results. Scale bars, 200 µm. **l–p** RQ (l), fat consumption per body weight (BW) (m), CHO consumption per BW (n), energy expenditure per BW (o), and physical activity (p) for control and Nos1-TeNT mice pair-fed with the HFD for 60 days. Data are means ± s.d. (n = 8 mice for control, n = 9 mice for Nos1-TeNT). **q-s** Relations between body weight and energy expenditure (q), fat consumption (r), or CHO consumption (s) during light and dark periods for mice as in (**l–p**). Each dot represents one mouse. The ANCOVA results are shown in each panel. Statistical significance for comparison was determined by two-way ANOVA and Sidak’s post hoc test (a, **f–h**, **l–p**), ANCOVA and Sidak’s adjustment (**q–s**), or by the two-sided unpaired Student’s t test (b). All animals were maintained at 24 ± 2 °C and a 12-h light/12-h dark cycle. Source data are provided as a Source Data file.

Nos1-TeNT mice were hyperphagic (Figs. 4i, 7b). To directly assess the role of Nos1 neuron-mediated regulation of fat metabolism independent of energy intake, Nos1-TeNT mice and control mice were pair-fed a high-fat diet (HFD) containing doxycycline for 60 days (Fig. 7f). The food intake of control mice was measured every 2 to 3 days, and the same amount was given to the Nos1 neuron-silenced group over the next 2 to 3 days. Despite consuming the same amount of food, Nos1-TeNT mice gained significantly more weight than control mice (Fig. 7g, h). Plasma concentrations of NEFAs, glucose, triacylglycerol, glycerol, and insulin did not differ notably between the two groups (Supplementary Fig. 10c). However, the liver, BAT, and ingWAT weights were considerably higher in Nos1-TeNT mice than in controls (Supplementary Fig. 10d). Histological analysis showed greater lipid accumulation in these tissues of Nos1-TeNT mice (Fig. 7i–k).

HFD feeding lowered the RQ throughout the day compared with CD-fed mice, with minor fluctuations that were unaffected by the light-dark cycle (Supplementary Fig. 10e). However, even under HFD pair-fed conditions, which often induce a starvation state especially in hyperphagic animals and decrease the RQ (see Fig. 8), Nos1-TeNT mice increased the RQ throughout the day (Fig. 7l). They also significantly reduced fat consumption during the dark period but not during the light period (Fig. 7m), and CHO consumption increased across the day (Fig. 7n). Consequently, the difference in energy expenditure between the light and dark periods was eliminated in Nos1-TeNT mice (Fig. 7o). Physical activity decreased during the dark period (Fig. 7p).

Fig. 8 — a Rectal temperature for control and Nos1-TeNT mice during acute cold exposure (4 °C) (n = 4 for control, n = 5 for Nos1-TeNT mice). **b–d** Relative mRNA abundance for *Ucp1* (b), *Cpt1b* (c), and *Ppargc1a* (d) in BAT of control or Nos1-TeNT mice subjected to cold exposure for 8 h (4 °C) or maintained at room temperature (25 °C). Data are means ± s.d. (n = 4 mice (control at 25 °C), and n = 5 mice (other groups)). e Representative PVH area images showing immunofluorescence for Nos1 (magenta) and c-Fos (green) in C57BL/6 J wild-type mice housed at 4 °C or 25 °C for 8 h. The boxed regions are shown at higher magnification to the right. Scale bars, 100 µm. The bar graph shows the proportion of Nos1-positive cells in the PVH that expressed c-Fos. Data are means ± s.d. (n = 4 mice for 4 °C, n = 5 mice for 25 °C). f, g The RQ (f) and fat consumption (g) for Nos1-TeNT and control mice after food removal at ZT12. Data are means ± s.d. Each left panel shows values for consecutive 5-min windows from 1 h before to 17 h after food removal, and each right panel shows the average for the indicated periods after food removal (n = 7 for control, n = 8 for Nos1-TeNT mice). Statistical significance for comparisons between Nos1-TeNT and control mice, or mice housed at 4 °C and 25 °C, respectively, was determined by two-way ANOVA and Sidak’s post hoc test (a, f, g), one-way ANOVA with Sidak’s post hoc test (**b–d**), or the two-sided unpaired Student’s t test (e). All animals were maintained at 24 °C ± 2 °C and a 12-h light/12-h dark cycle. Source data are provided as a Source Data file.

Fat and carbohydrate (CHO) consumption, along with energy expenditure, are linked to body weight. Therefore, we examined how silencing Nos1 neurons in the PVH area affects fat and CHO intake, as well as energy expenditure, using body weight as a covariate. Although energy expenditure during both the light and dark periods was similar between Nos1-TeNT and control mice pair-fed with the HFD for 60 days (Fig. 7q), analysis of covariance (ANCOVA) showed that fat consumption significantly decreased while CHO consumption increased in Nos1-TeNT mice during the dark period (Fig. 7r, s). During the light period, the effects of chronic Nos1 neuron silencing on fat and CHO consumption varied depending on body weight. Fat consumption tended to decline in Nos1-TeNT mice weighing 35–45 g but increased at weights of 50–60 g (Fig. 7r). This suggests that the lack of differences in fat consumption between groups during the light period is due to its greater dependence on body weight in Nos1-TeNT mice, which exhibit the larger weight gain. Additionally, CHO intake was notably higher in Nos1-TeNT mice compared to controls, especially at weights of 35–50 g (Fig. 7s). These findings indicate that long-term silencing of Nos1 neurons in the PVH area influences fat and CHO consumption during both light and dark periods.

The decrease in physical activity is unlikely to be the primary reason for changes in fat and CHO consumption during the light and dark period, because it would mainly reduce CHO consumption or both fat and CHO consumption, thereby affecting energy expenditure. Together with the lack of change in physical activity during acute inhibition of Nos1 neurons in the PVH area, the reduction in physical activity may result from chronic metabolic changes induced by long-term neuronal inhibition.

We also examined whether silencing Nos1 neurons in the PVH area affects energy metabolism when mice are pair-fed with CD and doxycycline for 74 days (Supplementary Fig. 11a). Although the differences were less noticeable than in mice pair-fed with HFD, Nos1-TeNT mice on CD showed greater body weight gain and increased liver, BAT, and ingWAT weights compared to control mice (Supplementary Fig. 11b–d). Overall, these findings suggest that Nos1 neuron activity in the PVH area is necessary for regulating fat consumption, which, in turn, affects CHO consumption and energy balance. The neurons also exert a strong inhibitory effect on food intake.

Nos1 neuron activity in the PVH area is required for metabolic adaptation to cold exposure, food deprivation, and other stressors

Fat consumption increases not only during the light period but also in response to stimuli such as cold exposure and food deprivation. We investigated whether Nos1 neurons in the PVH area also regulate this additional physiological control of fat consumption. We first examined whether Nos1 neurons in the PVH area are necessary for thermogenesis during acute cold exposure. Nos1-TeNT and control mice fed a doxycycline-containing diet were kept at 4 °C for 8 h. While control mice maintained their rectal temperature, Nos1-TeNT mice did not, with their rectal temperature dropping by over 4 °C (Fig. 8a). Since BAT is the main thermogenic tissue activated by cold, we analyzed thermogenesis-related gene expression in BAT. The cold-induced increase in uncoupling protein 1 (Ucp1) mRNA in BAT of control mice was significantly reduced in Nos1-TeNT mice (Fig. 8b). The amount of carnitine palmitoyltransferase 1b (Cpt1b) mRNA in BAT was also lower in Nos1-TeNT mice than for control mice under cold conditions (Fig. 8c). Conversely, both groups showed a similar increase in PGC1α (Ppargc1a) mRNA (Fig. 8d). The abundance of PPARα (Ppara) mRNA did not change in BAT of either Nos1-TeNT or control mice with or without cold exposure (Supplementary Fig. 12a, b). The percentage of Nos1 neurons in the PVH area expressing c-Fos in C57BL/6 J wild-type mice significantly increased after 8 hours of cold exposure (Fig. 8e), indicating neuronal activation by acute cold exposure. These findings suggest that acute silencing of Nos1 neurons in the PVH reduces thermogenesis and fat oxidation in BAT, highlighting their crucial role in the acute thermogenic response to cold exposure.

Next, we examined whether Nos1 neurons in the PVH area are necessary for fasting-induced fat intake. In control mice, food (CD) deprivation starting at ZT12 caused a rapid decrease in the RQ (Fig. 8f) and CHO consumption (Supplementary Fig. 12c), with a similar change in energy expenditure (Supplementary Fig. 12d). Fat consumption increased rapidly within about 2 hours (Fig. 8g). In contrast, Nos1-TeNT mice exhibited a higher RQ and CHO consumption, and the increase in fat consumption was delayed during the first 0–4 hours of food deprivation. These measures later reached levels comparable to those of control mice. These findings suggest that Nos1 neurons in the PVH area are essential for the rapid rise in fat consumption following food deprivation.

We examined how other stressors affect Nos1 neuron activity using fiber photometry. Several stressors—including spraying water on the face, hanging by the tail (tail suspension), and exposure to a large open space (open field)—activated Nos1 neurons in the PVH area (Fig. 9a, b, Supplementary Fig. 12e, f). To determine if Nos1 neurons are necessary for water-spray-induced metabolic changes, we inhibited Nos1 neuron activity using the hM4Di receptor. Stress exposure lowered the RQ and increased both fat consumption and energy expenditure during the first 60 min in saline-treated Nos1-hM4Di mice, indicating that the stressor promotes fat consumption and energy expenditure (Fig. 9c, d, Supplementary Fig. 12g, h). Conversely, blocking Nos1 neuron activity before stress reduced the decrease in the RQ and the rise in fat consumption (Fig. 9c, d). It also increased CHO consumption (0–60 min), thereby maintaining energy expenditure in stressed mice (Supplementary Fig. 12g, h). Physical activity did not change by inhibition of Nos1 neuron activity (Supplementary Fig. 12i). These results suggest that Nos1 neuron activity in the PVH area is required for the stress-induced increase in fat consumption but not for changes in energy expenditure. Collectively, this set of experiments demonstrates that Nos1 neurons in the PVH area play a critical role in metabolic adaptations to various environmental stimuli, including cold exposure, food deprivation, and other stressors that increase whole-body fat consumption.

Fig. 9 — a Changes in GCaMP6s signal intensity in Nos1 neurons in the PVH area of Nos1-Cre mice subjected to water spray stress at t = 0. Data are means ± s.d. (n = 6 mice). b Average GCaMP6s signal intensity in Nos1 neurons in the PVH area at 300 s after water spray stress, 30 s after the onset of tail suspension, or 1200 s after open-field exposure. Data are means ± s.d. (n = 6 mice). c, d RQ (c) and fat consumption (d) for Nos1-hM4Di mice treated with saline or CNO and subjected (or not) to water spray stress. Saline or CNO was administered 60 min before stress exposure at t = 0 (ZT12). Each left panel shows values for consecutive 5-min windows (Data are means ± s.d.), and each right panel shows averages for the indicated periods after stress (Bars indicate means) (n = 5 mice). Statistical significance for comparisons between saline and CNO administration with or without stress was determined by two-way ANOVA and Sidak’s post hoc test (c, d), or the two-sided one-sample t test (b). Source data are provided as a Source Data file.

Discussion

Selection of the appropriate energy substrate under different pathophysiological conditions is a key aspect of the homeostatic regulation of energy metabolism. We have now shown that Nos1 neurons in the PVH area function as a node for fat consumption and thereby affect energy substrate selection in male mice. Activation of these neurons promotes fat consumption by increasing fatty acid oxidation in various peripheral tissues and lipolysis in WAT, in a manner dependent on the sympathetic nervous system. Conversely, we found that inhibition of these neurons markedly reduced fat consumption during the light period. The Nos1 neurons are also required for rapid adaptation to cold exposure, food deprivation, and various other stressors. Monitoring their activity supported the notion that Nos1 neurons in the PVH area regulate fat consumption in response to these physiological conditions. Of note, we found that modulation of Nos1 neuron activity in the PVH area alters fat consumption, independent of total energy expenditure, physical activity, and food intake. Long-term inhibition of the neurons induced obesity under pair-fed conditions with HFD and CD. Therefore, our results reveal a previously unrecognized mechanism by which Nos1 neurons in the PVH area regulate fat consumption and energy balance.

The hypothalamus is a crucial brain region involved in energy homeostasis. It contains neurons that control intermediate metabolism in peripheral tissues through the autonomic nervous system or hormonal signals. Our PRV tracing analysis showed that the PVH is rich in neurons connecting to peripheral tissues via polysynaptic circuits through the autonomic nervous system, consistent with previous studies using PRV injections into peripheral tissues^8–10. We found that Nos1 neurons account for about 30% of PRV-infected cells in the PVH area that convey signals via neural pathways to tissues such as the soleus, BAT, and ingWAT. Although it is technically challenging to precisely estimate the percentage of Nos1 neurons among all PVH neurons projecting to peripheral tissues, these findings indicate that over 30% of PRV-infected neurons are Nos1-positive across all peripheral tissues, and that some Nos1 neurons in the PVH area connect to multiple tissues via neural pathways. CRH and MC4R neurons, which are key in regulating peripheral metabolism^15,37,38, were found to be distinct from Nos1 neurons in the PVH area. Nos1 neurons in the PVH represented a larger portion of PRV-infected cells than other PVH neurons, such as those expressing MC4R and CRH. These results suggest that Nos1 neurons in the PVH are among the key neurons regulating metabolic tissues via the autonomic nervous system, along with previously identified neurons.

Whereas some Nos1 neurons in the PVH were previously shown to express oxytocin¹³, PRV-infected oxytocin neurons in the PVH were much less abundant than PRV-infected Nos1 neurons, suggesting that oxytocin-negative cells may be predominant among presympathetic neurons that regulate metabolism in the soleus and BAT. Consistent with this, the previous study reported that selective activation of oxytocin neurons had minimal effects on oxygen consumption or food intake¹³. However, neurons positive for both oxytocin and Nos1 in the PVH may regulate other peripheral tissues such as the pancreas, liver, and ingWAT via the autonomic nervous system^41,42.

Activation of Nos1 neurons increased sympathetic nerve activity in skeletal muscle, ingWAT, and BAT, as well as the phosphorylation level of HSL in ingWAT and fatty acid oxidation in various peripheral tissues. Silencing of Nos1 neurons in the PVH area had the opposite effect, decreasing NA content in skeletal muscle, BAT, and ingWAT in Nos1-TeNT mice. The increase in fat consumption caused by Nos1 neuron activation was significantly reduced by the blockade of adrenergic signaling. In addition, silencing of Nos1 neuron activity did not affect corticosterone, TSH, or insulin levels. These results suggest that Nos1 neurons in the PVH area coordinately regulate fat mobilization and consumption in some peripheral tissues via the sympathetic nervous system.

As reported previously¹³, Nos1 neurons in the PVH area send dense projections to the spinal cord. Furthermore, we found that these Nos1 neurons projecting directly to the spinal cord are functionally capable of driving fat consumption. Therefore, our results suggest that Nos1 neurons in the PVH area directly project to the sympathetic nervous system and stimulate fat consumption. However, as reported previously¹³, Nos1 neurons in the PVH area also project to other brain areas, including the brainstem. Therefore, Nos1 neurons probably stimulate fat consumption through the sympathetic nervous system, directly via the IML, and indirectly via other brain areas, including the brainstem.

We found that Nos1 neurons in the PVH area play a more significant role during the light period than during the dark period. Sympathetic nerve activity decreases in peripheral tissues such as the heart during rest^31,32. Therefore, activation of the sympathetic nervous system during the light period may be limited to specific tissues such as skeletal muscle, BAT, and ingWAT, coordinately stimulating both fat mobilization and consumption in these tissues. BAT thermogenesis appears necessary to maintain body temperature during rest when physical activity is reduced. Furthermore, modulation of Nos1 neuron activity in the PVH area did not alter NA turnover and NA content in the heart or liver, suggesting that these Nos1 neurons are more influential on tissues regulating fat metabolism during the light period. Understanding how Nos1 neurons in the PVH area control specific tissues is a key question for future research.

In addition to fat consumption, modulation of Nos1 neurons in the PVH area alters CHO consumption, energy expenditure, physical activity, and food intake. Activating Nos1 neurons in the PVH area increased fat consumption and decreased CHO consumption, whereas both acute and chronic suppression of these neurons produced opposite effects. Notably, energy expenditure and physical activity levels briefly rise during the first few hours after Nos1 neuron activation, whereas the rise in fat consumption lasts up to 6 h. Furthermore, acute silencing of Nos1 neuron activity in the PVH does not affect energy expenditure or physical activity but significantly reduces fat consumption. These results suggest that Nos1 neurons in the PVH regulate fat consumption independently of energy expenditure and physical activity.

Compared to the significant change in fat consumption after modulation of Nos1 neurons, the changes in CHO consumption and energy expenditure are relatively small and have no effect in some experiments. Therefore, the effect of Nos1 neurons in the PVH area on CHO consumption is likely mediated by a compensatory mechanism that maintains total energy expenditure constant. The idea is further supported by differential effects of adrenergic receptor blockers on CHO consumption during the light period after acute neuronal suppression, when no significant changes in CHO consumption were observed. The mechanism remains unknown. However, whereas CHO consumption may be autonomously regulated to maintain total energy expenditure, other mechanisms are also involved. Leptin has been shown to promote both fat and CHO utilization through the sympathetic nervous system^25,43,44, which are regulated differently⁴⁵. Orexin neurons in the hypothalamus also increase glucose utilization in skeletal muscle⁴⁶. Further studies are necessary to explore this possibility.

As reported previously¹³, modulation of Nos1 neurons in the PVH area alters food intake. Since Nos1 neuron activity in the PVH area is higher during the light period than the dark, the inhibitory effect of Nos1 neurons in this area is likely consistent with the reduced food intake during this time. Although changes in food intake partially contribute to changes in fat consumption, they are unlikely to be the primary factor driving fat consumption in response to Nos1 neuron activation. In fact, although activation of Nos1 neurons during the dark period consistently suppressed food intake for up to 4 h, fat consumption returned to baseline levels by then. Additionally, even under food-deprived or pair-feeding conditions, inhibiting Nos1 neurons in the PVH increased fat consumption. These findings suggest that Nos1 neurons in the PVH area stimulate fat consumption independently of their effects on food intake. The projection of Nos1 neurons to brain regions other than the spinal cord may play a role in regulating food intake. Identifying subclasses of Nos1 neurons in the PVH through gene expression or single-cell calcium imaging will help clarify the mechanisms underlying their different roles in metabolic and food intake regulation. Such techniques will also be needed to uncover the upstream and downstream neural pathways that control PVH Nos1 neurons and the physiological processes they influence.

An important question is whether different subpopulations of PVH Nos1 neurons in the PVH area control fat metabolism in individual tissues in a “one-to-one” manner or if subsets of these neurons regulate fat metabolism across multiple tissues in a “one-to-all” pattern. We found that a significant portion of PVH Nos1 neurons project to only one of the two organs (BAT and either soleus or ingWAT) injected with PRVs, supporting the one-to-one scenario and indicating that individual PVH Nos1 neurons coordinately regulate fat metabolism in specific peripheral tissues via sympathetic nerves innervating those tissues. We also identified a subset of neurons that project to BAT and either soleus or ingWAT, suggesting the existence of another subset that influences multiple peripheral tissues involved in fat metabolism. Thus, these findings imply that PVH Nos1 neurons may have complex regulatory roles in fat metabolism, acting through both neurons dedicated to specific tissues and those that influence multiple tissues. These multiple connections to peripheral tissues could be mediated at the spinal cord level or through other brain areas, such as the brainstem. We have previously shown that surgical sympathetic denervation, but not denervation of motor nerves, inhibits leptin-induced fatty acid oxidation and AMPK activation in skeletal muscle²⁵, suggesting that the brain directly promotes fat consumption in certain tissues as well as lipolysis in WAT through the sympathetic nervous system. Future research focusing on subgroups of Nos1 neurons within the PVH area will help uncover the detailed mechanisms by which these neurons control the function of peripheral tissues.

Switching of energy substrate utilization occurs under various physiological conditions, including the daily light-dark transitions^47–50. Our results now indicate that Nos1 neurons in the PVH area play a crucial role in fat and CHO consumption linked to the diurnal rhythm. In contrast, ob/ob, KK-Ay, MC4R knockout, and AgRP knockout mice clearly retain the large transition of the RQ during the light-dark cycle^27–30. Thus, the role of Nos1 neurons in the PVH area in fat metabolism might differ from that of neurons, such as MC4R neurons and other leptin-regulated neurons.

We found that the daily light-dark cycle of Nos1 neuron activity in the PVH area was disrupted in mice housed in a constant-light environment. Nos1 neurons in the PVH area receive indirect neuronal inputs from the SCN. SCN neurons, most of which express γ-aminobutyric acid (GABA), were previously shown to project to PVH neurons^51,52. Since SCN neurons are activated by light, they may activate Nos1 neurons in the PVH area via disinhibition of inhibitory neurons innervating the Nos1 neurons. The polysynaptic connection from the SCN to Nos1 neurons in the PVH area might underlie the differences in the daily pattern of fat consumption between humans and rodents.

Nos1 neurons in the PVH area are also required for cold-induced thermogenesis and the fasting-induced transition from CHO to fat consumption. Mice with silenced Nos1 neurons in the PVH area failed to maintain their body temperature and to increase the expression of genes related to thermogenesis and fatty acid oxidation in BAT. Silencing of Nos1 neurons in the PVH area also impaired the switch to fat consumption immediately after food deprivation. The rapid shift from CHO to fat consumption within ~2 h of food deprivation likely reflects an anticipatory response to protect against depletion of CHO stores. Nos1 neurons in the PVH area are therefore essential for acute metabolic changes induced by cold exposure or food deprivation. Furthermore, our results indicate that these neurons contribute to the increase in fat consumption, but not that in energy expenditure, induced by other acute stressful stimuli. Thus, Nos1 neurons in the PVH play an essential role in fat consumption under stress conditions.

Long-term silencing of Nos1 neurons in the PVH area led to significant body weight gain, along with increased fat storage in adipose tissue and the liver. Mice with silenced Nos1 neurons in this area also gained more weight than control mice despite consuming the same amount of food. Interestingly, long-term silencing of Nos1 neuron activity disrupted the regulation of CHO and fat consumption during both dark and light periods, indicating that proper selection of energy substrates and nutrient partitioning are crucial for maintaining energy balance. Consistent with this, humans with lower fat metabolic rates have a higher risk of developing obesity^53,54. Nos1-TeNT mice gained more weight on the same calorie intake as controls, even though their energy expenditure was similar. This may be because changes in energy expenditure induced by silencing Nos1 neurons in the PVH area were below the detection limit of indirect calorimetry. Alternatively, the pair-feeding approach used might have caused a starvation-like state in Nos1-TeNT mice, thereby affecting body weight regulation, though starvation decreased CHO consumption in our study. Direct calorimetry might be necessary to clarify this.

Collectively, our findings suggest that Nos1 neurons constitute a substantial population in the PVH that play a key role in orchestrating fat consumption in peripheral tissues by activating the sympathetic nervous system.

Methods

Animals

Male C57BL/6 J mice were obtained from Nihon SLC, and transgenic mice expressing Cre recombinase in Nos1 neurons (Nos1-Ires-Cre, or Nos1-Cre, mice: B6.129-Nos1^tm1(cre)Mgmj/J) were obtained from The Jackson Laboratory (Strain#017526). Both heterozygous and homozygous Nos1-Cre mice were used for the study.

Nos1-floxed (Nos1^fl/fl) mice were generated by insertion of a loxP site on each side of exon 3 (Supplementary Fig. 8a)⁵⁵. The insertion of loxP sequences was achieved in a single step with the use of the CRISPR/Cas9 system as previously described⁵⁶. The CRISPR RNAs (crRNAs) targeting intron 2 (5′-CAGGATTGCAACCTTGATCT-3′) and intron 3 (5′-GGTCCCACACTTCCTCGGGT-3′) of mouse Nos1 were designed with the use of an online single guide RNA (sgRNA) design tool (https://sg.idtdna.com/site/order/designtool/index/CRISPR_CUSTOM) and were obtained from Integrated DNA Technologies (IDT). Donor single-stranded DNA containing the loxP sites, which result in excision of exon 3 and formation of a premature stop codon as a result of a frameshift, was produced with the use of the Guide-it Long ssDNA Production System v2 (Takara Bio) and with gBlocks Gene Fragments synthesized by IDT as templates. Single-stranded oligodeoxynucleotide (40 ng/μl), crRNA (0.31 μM), trans-activating crRNA (tracrRNA, 0.31 μM), and Cas9 protein (0.18 μM), all obtained from IDT and dissolved in Opti-MEM (Thermo Fisher Scientific), were introduced into fertilized C57BL/6 J mouse embryos at the pronuclear stage by microinjection. The embryos were subsequently washed and cultured overnight in potassium simplex optimization medium (KSOM, ARK Resource). A total of 123 two-cell embryos was transferred to pseudopregnant recipient ICR mice (CLEA Japan). Pups derived from the transplanted embryos were considered the F0 generation.

All mice were fed with laboratory chow (CE-2, CLEA Japan) and were maintained at a constant temperature of 24 °C ± 2 °C, at humidity of 50 ± 20% and with a 12-h-light, 12-h-dark cycle unless indicated otherwise. Eight- to 24-week-old male mice housed individually were used for all experiments and were randomly assigned to experimental groups unless indicated otherwise. Animals of similar age ranges were adopted for each experiment. All experimental procedures complied with institutional guidelines for the care and handling of experimental animals and were authorized by the Institutional Animal Care and Use Committees of the National Institutes of Natural Sciences and Tottori University.

Cells

PK15 cells were obtained from American Type Culture Collection (cat. no. CCL-33) for PRV production. B7GG cells, BHK-EnvA cells, and HEK293-TVA cells used for rabies virus production were kindly provided by Edward Callaway (Salk Institute). All cells were cultured under 5% CO₂ and at 37 °C in Dulbecco’s modified Eagle’s medium (DMEM) containing glucose at 4500 mg/l (Fujifilm Wako) and supplemented with 10% fetal bovine serum (Sigma-Aldrich).

PRV preparation

PRV152 and PRV614 were kindly provided by Lynn Enquist (Princeton University). PRV152, PRV614, and PRVB180 were propagated as described previously¹⁹. PK15 cells were infected with the parental viruses at a multiplicity of infection of 0.01 to 0.1, and the cells were harvested when a prominent cytopathic effect was apparent. They were disrupted by ultrasonic treatment, and cell debris was removed by centrifugation twice at 1000 × g for 5 min. The cleared supernatant was used for experiments. The titer of viral stocks was determined by standard plaque assays with PK15 cells⁵⁷.

Rabies virus preparation

Rabies viruses were obtained as described previously⁵⁸. For preparation of a GFP-expressing rabies virus with G protein deletion, B7GG cells were transfected with pSADΔG-GFP-F2, pcDNA-SADB19N, pcDNA-SADB19P, pcDNA-SADB19L, and pcDNA-SADB19G (Addgene, cat. no. 32630, 32631, 32632, 32633, 32635). The SADΔG rabies viruses released into the culture supernatant were concentrated by ultracentrifugation and then used to infect BHK-EnvA cells for the production of EnvA-pseudotyped SADΔG viruses. One day after virus infection, the cells were washed with phosphate-buffered saline (PBS), treated with 0.25% trypsin-EDTA, and replated on new dishes. The culture supernatant containing viruses was collected after 3 days, passed through a 0.45-μm filter, and concentrated by two rounds of ultracentrifugation. The EnvA-pseudotyped rabies viruses were titrated with HEK293-TVA cells.

Virus injection into peripheral tissues

Male C57BL/6 J mice were anesthetized by ip injection of a mixture of ketamine (100 mg/kg) (Daiichi Sankyo) and xylazine (10 mg/kg) (Sigma aldrich), or a mixture of medetomidine hydrochloride (0.75 mg/kg) (Zenoaq), midazolam (4 mg/kg) (Maruishi Pharma), and butorphanol tartrate (5 mg/kg) (Meiji Seika Pharma). Tissues were injected with a total of 2 µl (soleus), 2.5 µl (BAT), or 4 µl (other tissues) of PRV suspension containing 1.0 × 10¹⁰ plaque-forming units (pfu)/ml with the use of a Hamilton 701RN Neuros syringe equipped with a 33 G needle.

For injection of viruses into BAT, the ventral side of inter scapular brown fat pads was exposed by making an incision at the caudal end and PRVs were injected at a total of four sites on both sides of each pad. For infection of the soleus, the skin overlying the gastrocnemius muscle was incised and the soleus located underneath was exposed with forceps. PRVs were then injected at four sites along the soleus. For infection of ingWAT, inguinal fat pads were exposed by incisions of the inguinal area and PRVs were injected at a total of eight sites on both sides of each fat pad. For injection of viruses into the liver or pancreas, a transverse cut was made under the ribcage to expose the liver and PRVs were injected at eight sites in the right lobe of the liver or in the widest portion of the pancreas.

Skin incisions were sutured after injections, and the health of each mouse was monitored daily. Three to four days after surgery, the brain of each animal was analyzed after administration of deep anesthesia with a mixture of ketamine and xylazine and transcardiac perfusion with 4% paraformaldehyde.

Stereotaxic injection and optic fiber implantation in the brain

Mice were anesthetized by ip injection of a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg), or a mixture of medetomidine hydrochloride (0.75 mg/kg), midazolam (4 mg/kg), and butorphanol tartrate (5 mg/kg), the skull was fixed to a stereotaxic frame, and AAVs were injected into the PVH with a glass capillary connected to a 5-µl syringe (Hamilton). The stereotaxic coordinates for PVH injection relative to the bregma were obtained from the atlas of Paxinos and Franklin⁵⁹: −0.70 mm anteroposterior, ± 0.25 mm mediolateral, and −5.00 mm dorsoventral.

For stimulation or transient inhibition of Nos1 neurons, 250 nl of AAV8-hSyn-DIO-hM3Dq-mCherry (1.2 × 10¹³ genome copies/ml; AAV-FLEX-hM3Dq-mCherry, Addgene, cat. no. 44361) or 250 nl of AAV9-hSyn-DIO-hM4Di-mCherry (2.5 × 10¹³ genome copies/ml; AAV-FLEX-hM4Di-mCherry, Addgene, cat. no. 44362), respectively, were injected into each side of the PVH in Nos1-Cre mice. For silencing of Nos1 neurons by TeNT, 150 nl of AAV8-TRE-DIO-TeNT-t2a-EGFP (1.1 × 10¹³ genome copies/ml; AAV-FLEX-TeNT-EGFP, custom-made by Vector Builder Japan) and 150 nl of AAV8-Syn1-tTS/rtTA (4.2 × 10¹² genome copies/ml; AAV-tTS/rtTA, custom-made by Vector Builder Japan) were injected into each side of the PVH in Nos1-Cre mice; control mice were injected with 150 nl of AAV-tTS/rtTA only. For ablation of Nos1, 300 nl of AAV8-hSyn-GFP-Cre (6.5 × 10¹² genome copies/ml; AAV-Cre-GFP, University of North Carolina [UNC] Vector Core) were injected into each side of the PVH in Nos1^fl/fl mice; wild-type littermates were similarly injected as a control. For measurements of calcium signal of Nos1 neurons, 250 nl of AAV9-hSyn-DIO-GCaMP6s (2.1 × 10¹³ genome copies/ml; AAV-FLEX-GCaMP6s, Addgene, cat. no. 100843) were injected into one side of the PVH in Nos1-Cre mice, and a 250-µm-diameter mono–fiber optic cannula (NA0.50, RWD Life Science) was implanted above the injected area. For mapping the projection of Nos1 neurons, 200 nl of AAV1-hSyn-FLEX-mGFP-2A-Synaptophysin-mRuby (1.1 × 10¹³ genome copies/ml; AAV-hSyn-FLEx-mGFP-2A-Synaptophysin-mRuby, Addgene, cat. no. 71760) were injected into each side of the PVH in Nos1-Cre mice. For rabies virus-mediated monosynaptic tracing, 250 nl of AAV8-CAG-FLEX-RVG (2.5 × 10¹² genome copies/ml; AAV-FLEX-RVG, UNC Vector Core) and 250 nl of AAV8-EF1a-FLEX-TVA-mCherry (5.4 × 10¹² genome copies/ml; AAV-FLEX-TVA-mCherry, UNC Vector Core) were injected into each side of the PVH in Nos1-Cre mice, and 500 nl of EnvA-pseudotyped SADΔG rabies viruses (2.3 × 10⁸ ifu/ml; RVΔG-GFP:EnvA) were injected at the exact same locations 3 weeks after AAV injection. For polysynaptic tracing, 200 nl of PRVB180 (6.6 × 10⁹ pfu/ml) were injected into each side of the PVH in Nos1-Cre mice.

After virus injection, mice were injected subcutaneously with the analgesic meloxicam (5 mg/kg, Nippon Boehringer Ingelheim), a drop of the analgesic lidocaine (Aspen) was applied to the wound, and the skin was sutured. Antibiotic ointment was also applied around the wound to prevent infection, and the mice were monitored for several days.

Virus injection in the spinal cord

Virus injections into the lower thoracic spinal cord were performed as described previously⁶⁰ using 4-5-weeks-old Nos1-Cre male mice. Mice were anesthetized by ip injection of a mixture of medetomidine hydrochloride (0.75 mg/kg), midazolam (4 mg/kg), and butorphanol tartrate (5 mg/kg). An incision was made over the thoracic region of the spinal cord, and the paraspinal muscles and dorsal vertebral elements, including the dura mater, were carefully removed to expose the spinal cord. A total of 1000 nl of AAVrg-EF1a-DIO-FLPo (8.0 × 10¹² genome copies/ml; AAVrg-FLEX-FLPo, Addgene, cat. no. 87306) with 0.1% Fast Green (Sigma-Aldrich, cat. no.F7252) was injected bilaterally at 10 distinct sites across the T8–T12 spinal segments using a glass capillary connected to a 5-µl Hamilton syringe. In the same animals, 300 nl of AAV8-hSyn-fDIO-hM3Dq-mCherry (1.4 × 10¹³ genome copies/ml; AAV-fDIO-hM3Dq-mCherry, Addgene, cat. no. 154868) was simultaneously injected into the PVH as described above.

Stimulation of Nos1 neurons by hM3Dq

Nos1-Cre mice were allowed ~3 weeks to recover after injection of AAV-FLEX-hM3Dq-mCherry into the PVH. They were then subjected to indirect calorimetry with ip injection of CNO (2.5 mg/kg, Cayman Chemical, cat. no. 16882) or 0.9% NaCl (saline) at ZT3 or ZT12. Mice were returned to their regular cages and maintained under the ad libitum feeding condition for several days before tissue collection (as described below) at 3 h after injection of CNO or saline at ZT2–5. Some of the animals received CNO or saline at ZT4 after food deprivation for 16 h for analysis of blood metabolites and subsequent tissue collection at 3 h after CNO or saline injection. Some mice were directly subjected to measurement of fatty acid oxidation or NA turnover.

Transient inhibition of Nos1 neurons by hM4Di

Nos1-Cre mice were allowed ~3 weeks to recover after injection of AAV-FLEX-hM4Di-mCherry into the PVH. They were then subjected to indirect calorimetry with or without food and with or without stress exposure as described below. Some mice were subjected to stress experiments and hormonal measurement as described below.

Silencing of Nos1 neurons by TeNT

Nos1-Cre mice were allowed 1 week to recover after injection of AAV-tTS/rtTA with or without (control) AAV-FLEX-TeNT-EGFP into the PVH. They were then fed with the AIN-93G diet (Oriental Yeast) supplemented with doxycycline (100 mg/kg) for 4 days before being subjected to indirect calorimetry. Some mice were subsequently again subjected to indirect calorimetry after food deprivation, with food being removed at ZT12 and energy metabolism being monitored for the subsequent 17 h. Some mice were subjected to cold exposure. After these experiments, mice were euthanized, and tissues were collected for NA content analysis and histochemical examination of ingWAT, as described below. All these experiments were performed within 3 weeks after AAV injection, before the TeNT-expressing animals began to manifest significant weight gain (Fig. 7a).

Deletion of Nos1 in the PVH

Nos1^fl/fl mice or their wild-type littermates were allowed to recover for ~3 weeks after injection of AAV-Cre-GFP before being subjected to indirect calorimetry followed by transcardiac perfusion of 4% paraformaldehyde.

Pair-feeding

Eight-week-old Nos1-Cre mice were injected with AAV-tTS/rtTA with or without (control) AAV-FLEX-TeNT-EGFP and were subjected to pair-feeding experiments 10 days later. The food intake of the control group of mice was measured every 2 to 3 days, and the amount of food consumed by these mice was provided to the Nos1 neuron–silenced group at ZT11 over the next 2 to 3 days. Mice were fed with AIN-93G (Oriental Yeast) supplemented with doxycycline (100 mg/kg) for the chow diet (CD) experiment, or with an HFD (60% kcal from fat; Research Diets, cat. no. D12492) supplemented with doxycycline (100 mg/kg). The mice were pair-fed with the HFD for 60 days and then subjected to indirect calorimetry. Pair-feeding was then continued for an additional week before tissue collection as described below. The mice were pair-fed with CD for 74 days in the CD experiment.

Indirect calorimetry

Oxygen consumption (VO₂) and carbon dioxide production (VCO₂) for each mouse were measured with a mass spectrometric analyzer (Arco-2000, Arco System). For measurement of locomotor activity, a force plate system (Actracer-2000, Arco System) was mounted beneath the cage to detect low-intensity movements. VO₂, VCO₂, and locomotor activity were simultaneously measured every 5 min after 2 days of acclimation. Total energy expenditure and the amounts of utilized fat and carbohydrate were calculated from VO₂ and VCO₂ as described previously⁶¹: energy expenditure (cal/min) = (3.816 × VO₂) + (1.231 × VCO₂); fat oxidation (cal/min) = [1.67 × (VO₂ − VCO₂)] × 9.3; carbohydrate oxidation (cal/min) = [(4.51 × VCO₂) – (3.18 × VO₂)] × 4.1.

For analysis of the effects of Nos1 neuron activation or transient inhibition, Nos1-Cre mice expressing hM3Dq or hM4Di, respectively, were injected ip with CNO (2.5 mg/kg) or saline at ZT3 (light period) or ZT12 (dark period). For analysis of the effects of Nos1 neuron silencing, Nos1-Cre mice expressing TeNT were subjected to indirect calorimetry for two consecutive days, and the averages of values for two light and two dark periods were determined.

Administration of adrenergic receptor antagonists

C57BL/6 J mice or Nos1-hM3Dq mice were allowed to acclimate to the cages for indirect calorimetry for 2 days and were then subcutaneously injected with adrenergic receptor antagonists: propranolol (20 mg/kg, Sigma-Aldrich) alone, a mixture of phentolamine (10 mg/kg, Sigma-Aldrich), propranolol (20 mg/kg), and L748,337 (10 mg/kg, Cayman Chemical), or a mixture of phentolamine (10 mg/kg, Sigma-Aldrich), propranolol (20 mg/kg), and SR59230A (5 mg/kg, Selleck). Control mice were injected with saline containing the same amount of dimethyl sulfoxide vehicle. The injection into Nos1-hM3Dq mice was performed at ZT2.5, 30 min before CNO injection. The injection into C57BL/6 J mice was performed at ZT4, and energy metabolism was analyzed for 4 h.

Cold exposure, stress treatment, and fasting

For analysis of the effects of cold exposure on Nos1 neuron-silenced mice, Nos1-Cre mice expressing TeNT or control mice were placed in a cold (4 °C) chamber at ZT3, and their rectal temperature was measured at 0, 2, 5, and 8 h thereafter with a probe thermometer (BAT7001H, Physitemp). The brain and BAT were then isolated for analysis.

To assess the effects of fasting in Nos1 neuron-silenced mice, Nos1-Cre mice expressing TeNT or control mice were placed in cages for indirect calorimetry and food was removed from the cage at ZT12 for 17 h.

For analysis of the effects of stress exposure on Nos1 neuron activity, mice were allowed to acclimate to their home cage for 5 min after attachment of the optic fiber, and baseline activity was recorded for 5 min. For the induction of water spray stress, water was sprayed three times on the face of the mouse, and Nos1 neuron activity was measured for 5 min. For the induction of tail suspension stress, each mouse was suspended by its tail three times for 10 s, with an interval of 50 s between suspensions, and Nos1 neuron activity was measured for 5 min beginning with the initial tail suspension. For application of open-field stress, mice were introduced to a novel large metal cage (width and depth of 50 cm, height of 30 cm), and Nos1 neuron activity was measured for 20 min.

Fiber photometry

Changes in the cytosolic Ca²⁺ concentration of Nos1 neurons of freely moving mice were recorded with the use of a fiber photometry system (Doric Lenses) equipped with a 465-nm LED (light-emitting diode) light source for excitation of GCaMP6s and a 405-nm LED light source for excitation at the isosbestic point. The excitation light power at the fiber tip was adjusted to 25 µW. Fluorescence signals detected by photoreceiver were amplified 10 times, filtered by a 12-Hz low-pass filter at a sampling rate of 1 Hz (light-dark phase analysis) or 12 Hz (stress response analysis) with Doric Neuroscience Studio software (Doric Lenses) before the data analysis. The detection of two different wavelength signals were achieved with “the lock-in mode” where the input will only record signals at a given reference frequency.

Mice were allowed by acclimate to handling and optic fiber attachment for 5 days as well as to a custom photometry cage (30 cm wide by 30 cm deep by 50 cm high) for 1 week. For analysis of the effects of ambient brightness on Nos1 neuron activity, mice were kept in the photometry cage in a room with the light on constantly for >10 days before and during the experiments. For analysis of the effects of feeding time on Nos1 neuron activity, the mice were maintained in the photometry cage with access to food only during the light period (ZT0–12) for >10 days before and during the experiments.

On the day of the experiment, measurements were started at the beginning of the dark period (ZT12), 1 h after connection of the optical fiber. Data for the light and dark periods were collected from 12 (ZT0) to 24 (ZT12) hours and from 24 (ZT12) to 36 (ZT0) hours, respectively.

Analysis of fiber photometry data

Relative changes in fluorescence intensity were calculated with R software. Fluorescence changes at individual wavelengths were first calculated as ΔF/F₀ = (F_t – F₀)/F₀, where F₀ is the mean of the entire signal (long-term recording) or that for the 300-s period before stress treatment (stress response recording), and the relative change in fluorescence intensity (ΔF/F₀) was obtained by subtraction of fluorescence changes at 405 nm (ΔF⁴⁰⁵/F₀⁴⁰⁵) from those at 465 nm (ΔF⁴⁶⁵/F₀⁴⁶⁵) to compensate photobleaching. The ΔF/F₀ data were then analyzed using Doric Neuroscience Studio software to identify peaks of sustained synchronized calcium activity. The peaks were detected using the “spike finding” function within the electrophysiology analysis module. To adapt this algorithm to the slower temporal dynamics of fiber photometry signals, one hour of recording time was scaled to one second within the software. Peaks were defined using the following criteria: (1) maximum signal of >2.5 standard deviation above the average; (2) the signal spanning 3.6 s before to 10.8 s after this maximum signal included as a single peak; and (3) minimum time between two peaks of 108 s. Raster plots were automatically generated by Doric Neuroscience Studio software based on the detected peaks using the same thresholding parameters described above.

Measurement of in vivo fatty acid oxidation

The rate of fatty acid oxidation was measured as described previously²⁵. Three or four days before the experiment, a catheter was implanted into the right jugular vein of Nos1-Cre mice expressing hM3Dq in the PVH. On the day of the experiment, the mice were injected ip with CNO (2.5 mg/kg) or saline, and food was removed from the cage. After 3 h, a mixture of 8 µCi of [9,10-³H]-(R)−2-bromopalmitic acid ([³H]bromopalmitate, American Radiolabeled Chemicals, cat. no. ART1270) and 8 µCi of [1-¹⁴C]palmitic acid ([¹⁴C]palmitate, Revvity, cat. no. NEC075H) was injected through the jugular vein catheter over 5 min. Blood samples were collected from the tail vein at 0, 2, 5, 10, and 15 min after the onset of tracer injection, and the amounts of ³H and ¹⁴C radioactivity in the blood were measured. Additional blood samples were collected immediately before tracer administration and before tissue collection (15 min) for measurement of plasma fatty acid concentrations. All blood samples were immediately centrifuged in the presence of EDTA, and the plasma supernatant was rapidly frozen. Mice were killed by cervical dislocation 15 min after the start (10 min after the end) of tracer injection, and peripheral tissues—including ingWAT, epiWAT, interscapular BAT, heart, liver, soleus, and red and white portions of the gastrocnemius—were rapidly dissected and frozen in liquid nitrogen. Tissue samples were solubilized overnight at 55 °C with SOLVABLE (Revvity), and the amounts of ³H and ¹⁴C radioactivity were measured with a scintillation counter (Tri-Carb 3100TR, Revvity). For analysis of [³H]bromopalmitate and [¹⁴C]palmitate radioactivity in plasma samples, fatty acids were extracted with Dole’s reagent⁶². Plasma fatty acid concentrations were measured with a colorimetric assay (LabAssay NEFA kit, Fujifilm). For determination of the lumped constant ( ${L C}^{*}, see below$ ), some mice were injected with etomoxir (30 µmol/kg), an inhibitor of fatty acid oxidation, via the catheter 30 min before tracer injection.

The fatty acid oxidation rate in individual tissues in vivo was calculated as described previously²⁶. The fatty acid utilization rate ( $R_{f}$ ) in individual tissues was calculated by the following equation:

R_{f} = C_{P} m_{B} / {L C}^{*} \int_{0}^{T} c_{B} (t) d t

where $C_{P}$ is the steady-state venous plasma free fatty acid concentration, $c_{B}$ is the venous plasma concentration of [³H]bromopalmitate as a function of time (t), $m_{B}$ is the tissue content of ³H associated with metabolic products of [³H]bromopalmitate (at t = T), and T is the time of tissue collection (15 min). ${L C}^{*}$ is the lumped constant for each tissue, representing the proportion between the uptake of [³H]bromopalmitate and that of [¹⁴C]palmitate after inhibition of fatty acid oxidation by etomoxir injection.

The fatty acid storage rate ( $R_{f s}'$ ) was calculated with the following equation:

R_{f s}' = C_{P} m_{P} / \int_{0}^{T} c_{P} (t) d t

where $c_{P}$ is the venous plasma concentration of [¹⁴C]palmitate as a function of time (t) and $m_{P}$ is the tissue content of ¹⁴C associated with metabolic products of [¹⁴C]palmitate (at t = T).

The fatty acid oxidation rate in individual tissues was then calculated by subtraction of $R_{f s}'$ from $R_{f}$ .

NA turnover assay and measurement of tissue NA content

NA turnover was measured on the basis of the decline in tissue NA content after the inhibition of catecholamine biosynthesis with α-MT as described previously⁶³. Nos1-Cre mice expressing hM3Dq were injected ip with α-MT (200 mg/kg, Sigma-Aldrich) together with CNO (2.5 mg/kg) or saline. At 0, 2, and 4 h after the injection, the animals were killed by cervical dislocation and tissues were rapidly removed and weighed. For assay of tissue NA and adrenaline content, ~30 mg of each tissue were homogenized in 0.2 M perchloric acid containing 0.1 mM EDTA, and the homogenate was centrifuged at 11,100 × g for 20 min at 4 °C. 3,4-Dihydroxybenzylamine was added to the resulting supernatant as an internal control, and NA was then isolated from the supernatant with activated alumina and assayed by high-performance liquid chromatography (EP-300 system, Eicom) with a reversed-phase column (inner diameter of 2.1 by 150 mm; CA-5ODS, Eicom) and electrochemical detector (ECD-300, Eicom). Data are presented as nanograms of NA per gram of tissue weight.

Tissue collection

Mice were deprived of food for 3 h and then killed by cervical dislocation. Blood was collected from the heart via an incision and was centrifuged at 1200 × g for 15 min in EDTA-containing tubes for isolation of plasma. The brain and peripheral tissues were collected, and the weight of the latter was measured. The brain and portions of ingWAT, BAT, and liver were fixed with 4% paraformaldehyde, and all other tissues and plasma samples were rapidly frozen in liquid nitrogen. Frozen specimens were preserved at –80 °C until analysis.

Immunohistofluorescence staining

The brain was fixed with 4% paraformaldehyde, immersed in 30% sucrose until it sank, and frozen in optimal cutting temperature (OCT) compound (Sakura Finetek) mixed with 30% sucrose at a volume ratio of 2:1. Tissue blocks were sectioned at a thickness of 20 µm with a cryostat (Frigocut 2800, Leica Biosystems; or Cryostar NX-70, Epredia), and the sections were mounted on glass slides and stored at –20 °C. The frozen slides were returned to room temperature and washed twice with PBS. The sections were incubated for 5 min first with 0.5% Triton X-100 in PBS and then with TNT (0.1 M Tris-HCl [pH 7.5], 150 mM NaCl, 0.05% Tween 20). They were then incubated overnight at 4 °C (or at room temperature in the case of guinea pig anti–c-Fos) with primary antibodies—including rabbit anti-Nos1 (1:500 dilution; Immunostar, cat. no. 24287 and 24431), guinea pig anti-c-Fos (1:1000 dilution;Synaptic Systems, cat. no. 226-004), guinea pig anti-CRH (1:1000 dilution; BMA Biomedicals, cat. no. T-5007), goat anti-GFP (1:1000 dilution; Rockland, cat. no. 600-101-215), chicken anti-GFP (1:1000 dilution; Aves Labs, cat. no. GFP-1010), rabbit anti-RFP (1:1000 dilution; Rockland, cat. no. 600-401-379), and goat anti-RFP (1:1000 dilution; Rockland, cat. no. 200-101-379)—all of which were diluted in DSB (1% bovine serum albumin and 5% normal donkey serum in TNT). After three washes with TNT, the slides were incubated for 1 h at room temperature with Alexa Fluor 488-, Alexa Fluor 568-, or Alexa Fluor 647-labeled donkey antibodies to rabbit, goat, or guinea pig immunoglobulin G or chicken immunoglobulin Y (1:1000 dilution, Thermo Fisher, cat. no. A21206, A10042, A31573, A11057, A11055, A78948, or Jackson Laboratories, cat. no. 706-545-148) and with Hoechst 33342 (0.5 μg/ml, Thermo Fisher). Coverslips were then applied with Fluoromount-G (Southern Biotech). For immunostaining of CRH, colchicine (200 ng/g) was injected intracerebroventricularly at 1 day before tissue isolation.

For immunostaining of white adipose tissue, ingWAT samples were delipidated prior to staining as previously described⁶⁴. Tissues were washed twice with PBS, dehydrated through an ascending methanol series (20, 40, 60, 80, and 100%) in B1n buffer (0.3 M glycine and 0.1% Triton X-100), and immersed in 100% dichloromethane with three solvent changes until the samples sank. Tissues were then washed twice with 100% methanol, rehydrated through a descending methanol series (80, 60, 40, 20, and 0%) in B1n buffer, washed with PBS, and incubated overnight in 30% sucrose. Samples were embedded in OCT compound, cryosectioned at 25 µm thickness, mounted on glass slides, and stored at −20 °C until immunostaining. Immunostaining was performed as described above using chicken anti-tyrosine hydroxylase (1:1000 dilution; Abcam, cat. no. ab76442), rabbit anti-perilipin 1 (1:1000 dilution; Cell Signaling Technology, cat. no. 9349), with Hoechst 33342 used for nuclear counterstaining.

In situ hybridization combined with immunohistofluorescence staining

Complementary RNA probes (riboprobes) for Vglut1, Vglut2, Gad1, Gad2, Nos1, Mc4r, Glp1r, Oxt, Crh, Irs4, Cartpt1, Penk, Avp and Trh were adopted from the Allen Brain Atlas (https://mouse.brain-map.org). The probe sequence for each gene flanked by Sp6 and T7 promoter sequences was synthesized and cloned into the pUCF vector by Fasmac. The plasmids were then linearized, and digoxigenin (DIG)–labeled riboprobes were prepared with the use of a DIG RNA Labeling Kit (Roche).

Mice were anesthetized with a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg), or a mixture of medetomidine hydrochloride (0.75 mg/kg), midazolam (4 mg/kg), and butorphanol tartrate (5 mg/kg) and transcardially perfused with 4% paraformaldehyde in PBS. The brain was removed, placed consecutively in 4% paraformaldehyde for 4–6 h and in 30% sucrose for 24–48 h, frozen in OCT compound mixed with 30% sucrose at a volume ratio of 2:1, and cut into 14-μm coronal sections with a cryostat. Brain sections were incubated with 0.3% H₂O₂ in PBS for 30 min, permeabilized with proteinase K (6 μg/ml, Sigma-Aldrich) for 10 min, and then subjected to acetylation with 0.1 M triethanolamine and 25 mM acetic anhydride for 15 min. The sections were incubated at 56 °C for 1 h with hybridization buffer (yeast tRNA [250 μg/ml], heparin [300 U/ml], 1 mM dithiothreitol, herring sperm DNA [100 μg/ml], 5× Denhardt’s solution, 5× saline sodium citrate [SSC], and 50% formamide) and then subjected to hybridization overnight at 56 °C with the DIG-labeled riboprobes (0.2–0.5 ng/μl) in hybridization buffer. They were washed twice for 30 min at 68 °C with 0.2× SSC, incubated for 2 h at room temperature with POD-conjugated antibodies to DIG (1:500 dilution; Sigma-Aldrich, cat. no. 11207733910) and goat antibodies to GFP (1:500 dilution; Rockland, cat. no. 600-101-215) or rabbit antibodies to Nos1 (1:500 dilution; Immunostar, cat. no. 24287 and 24431) diluted in TNB (0.5% blocking reagent [Akoya Bioscience] in TNT), and then processed with a TSA-plus Cy3 Kit (Akoya Bioscience, cat. no. NEL744001KT). The sections were finally incubated for 1 h at room temperature with Hoechst 33342 (1 μg/ml) and Alexa Fluor 488-conjugated donkey antibodies to goat or rabbit immunoglobulin G (1:500 dilution) before coverslip application with Fluoromount-G.

Hematoxylin-eosin staining

The liver, BAT and ingWAT were fixed in 4% paraformaldehyde, washed with PBS, dehydrated with an ascending gradient of ethanol solutions (50, 60, 70, 80, 90, and 100%) and xylene, and then immersed and embedded in paraffin. The samples were sectioned at a thickness of 8 µm and subjected to hematoxylin-eosin staining.

RNA extraction and RT-qPCR analysis

Total RNA was extracted from frozen tissues with the Trizol reagent (Thermo Fisher) and was purified with the use of a NucleoSpin RNA Kit (Takara). Portions (1 µg) of the RNA were subjected to RT with an oligo(dT) primer and ReverTra Ace reverse transcriptase (Toyobo), and the resulting cDNA was purified with the use of a QIAquick PCR Purification Kit (Qiagen) and subjected to real-time PCR analysis with Power SYBR Green Master Mix (Thermo Fisher) and custom-made PCR primers (Fasmac) in a StepOnePlus real-time PCR system (Thermo Fisher). The abundance of target mRNAs was normalized by that of Rplp0 (36B4) mRNA. PCR primer sequences are provided in Supplementary Table 1.

Immunoblot analysis

Tissue samples were homogenized in RIPA buffer (50 mM Tris-HCl [pH 7.6], 150 mM NaCl, 1% IGEPAL CA-630 [Sigma-Aldrich], 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA) supplemented with a protease inhibitor cocktail (Sigma-Aldrich, cat. no. P8340) and phosphatase inhibitor cocktail (Merck Millipore, cat. no. 52462). The homogenate was incubated for 1 h at 4 °C before centrifugation at 17,400 × g for 15 min at 4 °C, and the protein concentration of the resulting supernatant was measured with the use of a Pierce BCA Protein Assay Kit (Thermo Fisher). Portions (5 μg of protein) of the supernatant were fractionated by SDS-polyacrylamide gel electrophoresis, the separated proteins were transferred to a polyvinylidene difluoride membrane, and the membrane was then probed with primary antibodies to Ser⁶⁶⁰-phosphorylated HSL (1:500 dilution; Cell Signaling Technology, cat. no. 45804) and to HSL (1:2000 dilution; Cell Signaling Technology, cat. no. 4107). Immune complexes were visualized with horseradish peroxidase-conjugated secondary antibodies (1:1000 dilution; Cell Signaling Technology, cat. no. 7074) and Clarity Western ECL Substrate (Bio-Rad).

Measurement of plasma glucose, insulin, NEFA, glycerol, triacylglycerol, corticosterone, and TSH concentrations

Plasma glucose, NEFA, glycerol, and triacylglycerol concentrations were measured with a LabAssay Glucose Kit (Fujifilm Wako), LabAssay NEFA Kit (Fujifilm Wako), Glycerol Assay Kit (Sigma-Aldrich) and LabAssay Triglyceride Kit (Fujifilm Wako), respectively. The plasma concentration of insulin, corticosterone and TSH was measured with an enzyme-linked immunosorbent assay kit for insulin (Fujifilm Wako), corticosterone (Enzo Life Sciences) and TSH (Cloud Clone), respectively.

Image acquisition and analysis

Immunofluorescence images were acquired with a confocal microscope (FV3000, Olympus) with the use of FV31S-SW software (Olympus), or a confocal microscope (LSM900, Zeiss) with the use of ZEN3 software (Zeiss), or with an all-in-one microscope (BZ-X800, Keyence). The brightness and contrast of images were adjusted with FIJI software (National Institutes of Health) as needed for visibility. Every fifth section was analyzed with the use of FIJI software to count the number of PRV-infected neurons or c-Fos-expressing neurons. TH-immunoreactive fiber density in ingWAT was quantified with the use of FIJI software. Regions of interest from eight different locations were randomly selected, and background fluorescence was reduced by rolling-ball background subtraction, followed by median filtering. Images were then automatically thresholded using the MaxEntropy algorithm and converted to binary masks. TH fiber density was quantified from the resulting masks using FIJI’s measurement functions. Immunoblot images were acquired with an EzCapture MG system (ATTO), and the density of protein bands was analyzed with FIJI software.

Statistical analysis

Quantitative data are presented as means ± s.d. and were analyzed with GraphPad Prism software (version 10.2.3) unless indicated otherwise. Comparisons among multiple groups were performed with one-way or two-way analysis of variance (ANOVA) and Sidak’s post hoc test, or with the one-sample t test, whereas those between two groups were performed with the unpaired or paired Student’s t test. Energy metabolism of Nos1 neuron–silenced mice and control mice under the HFD pair-fed condition was compared by ANCOVA with SPSS software (IBM). A p value of <0.05 was considered statistically significant.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Supplementary Information^{(29.8MB, pdf)}

Reporting Summary^{(2.8MB, pdf)}

Transparent Peer Review file^{(867KB, pdf)}

Source data

Source Data^{(12.3MB, xlsx)}

Acknowledgements

We thank Megumi Hayashi and Kaori Sato for technical assistance as well as members of the Division of Endocrinology and Metabolism, National Institute for Physiological Sciences, and members of Division of Integrative Physiology, Faculty of Medicine, Tottori University for discussion and comments. This research was partly performed at the Tottori Bio Frontier managed by Tottori prefecture, Japan.

Author contributions

K.K. and Y.M. conceived and designed the project, interpreted data, and wrote the manuscript. K.K. performed experiments and analyzed the data. E.H., Y.I., and H.I. generated Nos1-floxed mice. S.Y. and K.M. provided technical advice for virus injection to the spinal cord. Y.Y. supported animal experiments. K.-i.N. and T.Y.H. provided intellectual input and advice. All authors approved the manuscript.

Peer review

Peer review information

Nature Communications thanks Yi Zhou who co-reviewed with Shaowen Qian; Daisuke Ino, Fei Yin and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Fundings

This work was supported by KAKENHI grants from the Japan Society for the Promotion of Science (20H03736 and 23H02965 to Y.M., and 18K08494, 21H03387 and 25K03065 to K.K.), grants from the Japan Science and Technology Agency PRESTO (JPMJPR1784 and JPMJPR21S5 to K.K.), a joint grant from the Japan Society for the Study of Obesity and Novo Nordisk Japan (to K.K.), and grants from the Novartis Foundation (Japan) for the Promotion of Science, the Ono Medical Research Foundation, the Takeda Science Foundation, The Hori Sciences and Arts Foundation, and the Inamori Foundation (to K.K.).

Data availability

The data generated in this study are provided in the Source Data file. All microscopy data are provided upon request to the corresponding authors. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Kunio Kondoh, Email: kkondoh@tottori-u.ac.jp.

Yasuhiko Minokoshi, Email: minokosh@sugiyama-u.ac.jp.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-74362-9.

References

1.World Health Organization. Obesity and overweight. https://www.who.int/en/news-room/fact-sheets/detail/obesity-and-overweight.
2.Manceau, R., Majeur, D. & Alquier, T. Neuronal control of peripheral nutrient partitioning. Diabetologia63, 673–682 (2020). [DOI] [PubMed] [Google Scholar]
3.Goodpaster, B. H. & Sparks, L. M. Metabolic flexibility in health and disease. Cell Metab.25, 1027–1036 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Smith, R. L., Soeters, M. R., Wüst, R. C. I. & Houtkooper, R. H. Metabolic flexibility as an adaptation to energy resources and requirements in health and disease. Endocr. Rev.39, 489–517 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Randle, P. J., Garland, P. B., Hales, C. N. & Newsholme, E. A. The glucose fatty-acid cycle. Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet1, 785–789 (1963). [DOI] [PubMed] [Google Scholar]
6.Myers, M. G., Jr & Olson, D. P. Central nervous system control of metabolism. Nature491, 357–363 (2012). [DOI] [PubMed]
7.Brüning, J. C. & Fenselau, H. Integrative neurocircuits that control metabolism and food intake. Science381, eabl7398 (2023). [DOI] [PubMed] [Google Scholar]
8.Oldfield, B. J. et al. The neurochemical characterisation of hypothalamic pathways projecting polysynaptically to brown adipose tissue in the rat. Neuroscience110, 515–526 (2002). [DOI] [PubMed] [Google Scholar]
9.Bamshad, M., Song, C. K. & Bartness, T. J. CNS origins of the sympathetic nervous system outflow to brown adipose tissue. Am. J. Physiol. Regulatory Integrative Comparative Physiol.276, 1569 (1999). [DOI] [PubMed]
10.Kerman, I. A., Akil, H. & Watson, S. J. Rostral elements of sympatho-motor circuitry: a virally mediated transsynaptic tracing study. J. Neurosci.26, 3423–3433 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
11.de Souza Cordeiro, L. M. et al. Hypothalamic MC4R regulates glucose homeostasis through adrenaline-mediated control of glucose reabsorption via renal GLUT2 in mice. Diabetologia64, 181–194 (2021). [DOI] [PMC free article] [PubMed]
12.Toda, C. et al. Distinct effects of leptin and a melanocortin receptor agonist injected into medial hypothalamic nuclei on glucose uptake in peripheral tissues. Diabetes58, 2757–2765 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Sutton, A. K. et al. Control of food intake and energy expenditure by nos1 neurons of the paraventricular hypothalamus. J. Neurosci.34, 15306–15318 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Sutton, A. K. et al. Paraventricular, subparaventricular and periventricular hypothalamic IRS4-expressing neurons are required for normal energy balance. Sci. Rep.10, 5546 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Sawchenko, P. E. Evidence for differential regulation of corticotropin-releasing factor and vasopressin immunoreactivities in parvocellular neurosecretory and autonomic-related projections of the paraventricular nucleus. Brain Res.437, 253–263 (1987). [DOI] [PubMed] [Google Scholar]
16.Hallbeck, M., Larhammar, D. & Blomqvist, A. Neuropeptide expression in rat paraventricular hypothalamic neurons that project to the spinal cord. J. Comp. Neurol.433, 222–238 (2001). [DOI] [PubMed] [Google Scholar]
17.Zhang, X. et al. Brain control of humoral immune responses amenable to behavioural modulation. Nature581, 204–208 (2020). [DOI] [PubMed] [Google Scholar]
18.Swanson, L. W. & Kuypers, H. G. The paraventricular nucleus of the hypothalamus: cytoarchitectonic subdivisions and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J. Comp. Neurol.194, 555–570 (1980). [DOI] [PubMed] [Google Scholar]
19.Kondoh, K. et al. A specific area of olfactory cortex involved in stress hormone responses to predator odours. Nature532, 103–106 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Lin, H.-W. et al. Role of the UL41 protein of pseudorabies virus in host shutoff, pathogenesis and induction of TNF-α expression. J. Vet. Med. Sci.72, 1179–1187 (2010). [DOI] [PubMed] [Google Scholar]
21.Nogueiras, R. et al. The central melanocortin system directly controls peripheral lipid metabolism. J. Clin. Invest.117, 3475–3488 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Zhu, C. et al. Disrupted hypothalamic CRH neuron responsiveness contributes to diet-induced obesity. EMBO Rep. 21, EMBR201949210 (2020). [DOI] [PMC free article] [PubMed]
23.Huesing, C. et al. Sympathetic innervation of inguinal white adipose tissue in the mouse. J. Comp. Neurol. 529, 1465–1485 (2021). [DOI] [PMC free article] [PubMed]
24.François, M. et al. Sympathetic innervation of the interscapular brown adipose tissue in mouse. Ann. N. Y. Acad. Sci.1454, 3–13 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Minokoshi, Y. et al. Leptin stimulates fatty-acid oxidation by activating AMP-activated protein kinase. Nature415, 339–343 (2002). [DOI] [PubMed] [Google Scholar]
26.Oakes, N. D. et al. Development and initial evaluation of a novel method for assessing tissue-specific plasma free fatty acid utilization in vivo using (R)−2-bromopalmitate tracer. J. Lipid Res.40, 1155–1169 (1999). [PubMed] [Google Scholar]
27.Butler, A. A. et al. Melanocortin-4 receptor is required for acute homeostatic responses to increased dietary fat. Nat. Neurosci.4, 605–611 (2001). [DOI] [PubMed] [Google Scholar]
28.Ishihara, K. et al. Amino acid composition of soybean protein increased postprandial carbohydrate oxidation in diabetic mice. Biosci. Biotechnol. Biochem.67, 2505–2511 (2003). [DOI] [PubMed] [Google Scholar]
29.Albarado, D. C. et al. Impaired coordination of nutrient intake and substrate oxidation in melanocortin-4 receptor knockout mice. Endocrinology145, 243–252 (2004). [DOI] [PubMed] [Google Scholar]
30.Joly-Amado, A. et al. Hypothalamic AgRP-neurons control peripheral substrate utilization and nutrient partitioning. EMBO J.31, 4276–4288 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Grassi, G., Bombelli, M., Seravalle, G., Dell’Oro, R. & Quarti-Trevano, F. Diurnal blood pressure variation and sympathetic activity. Hypertens. Res.33, 381–385 (2010). [DOI] [PubMed] [Google Scholar]
32.Anderson, C. et al. Non-canonical role of the sympathetic nervous system in the day-night rhythm in heart rate. Philos. Trans. R. Soc. Lond. B Biol. Sci.378, 20220179 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Solanki, K. et al. The expanding roles of neuronal nitric oxide synthase (NOS1). PeerJ10, e13651 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Zhou, Q.-G., Zhu, X.-H., Nemes, A. D. & Zhu, D.-Y. Neuronal nitric oxide synthase and affective disorders. IBRO Rep.5, 116–132 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Foster, R. G., Hughes, S. & Peirson, S. N. Circadian photoentrainment in mice and humans. Biology9, 180 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Duffy, J. F. & Wright, K. P. Jr. Entrainment of the human circadian system by light. J. Biol. Rhythms20, 326–338 (2005). [DOI] [PubMed] [Google Scholar]
37.LeGates, T. A., Fernandez, D. C. & Hattar, S. Light as a central modulator of circadian rhythms, sleep and affect. Nat. Rev. Neurosci.15, 443–454 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Hastings, M. H., Maywood, E. S. & Brancaccio, M. Generation of circadian rhythms in the suprachiasmatic nucleus. Nat. Rev. Neurosci.19, 453–469 (2018). [DOI] [PubMed] [Google Scholar]
39.Berson, D. M., Dunn, F. A. & Takao, M. Phototransduction by retinal ganglion cells that set the circadian clock. Science295, 1070–1073 (2002). [DOI] [PubMed] [Google Scholar]
40.Hanchate, N. K. et al. Connect-seq to superimpose molecular on anatomical neural circuit maps. Proc. Natl. Acad. Sci. USA117, 4375–4384 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Papazoglou, I. et al. A distinct hypothalamus-to-β cell circuit modulates insulin secretion. Cell Metab.34, 285–298.e7 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Stanley, S. et al. Identification of neuronal subpopulations that project from hypothalamus to both liver and adipose tissue polysynaptically. Proc. Natl. Acad. Sci. USA107, 7024–7029 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Haque, M. S. et al. Role of the sympathetic nervous system and insulin in enhancing glucose uptake in peripheral tissues after intrahypothalamic injection of leptin in rats. Diabetes48, 1702–1706 (1999). [DOI] [PubMed]
44.Minokoshi, Y., Haque, M. S. & Shimazu, T. Microinjection of leptin into the ventromedial hypothalamus increases glucose uptake in peripheral tissues in rats. Diabetes48, 287–291 (1999). [DOI] [PubMed] [Google Scholar]
45.Shiuchi, T. et al. Induction of glucose uptake in skeletal muscle by central leptin is mediated by muscle β2-adrenergic receptor but not by AMPK. Sci. Rep.7, 15141 (2017). [DOI] [PMC free article] [PubMed]
46.Shiuchi, T. et al. Hypothalamic orexin stimulates feeding-associated glucose utilization in skeletal muscle via sympathetic nervous system. Cell Metab.10, 466–480 (2009). [DOI] [PubMed] [Google Scholar]
47.Randle, P. J. Regulatory interactions between lipids and carbohydrates: the glucose fatty acid cycle after 35 years. Diab. Metab. Rev.14, 263–283 (1998). [DOI] [PubMed] [Google Scholar]
48.Satoh, Y., Kawai, H., Kudo, N., Kawashima, Y. & Mitsumoto, A. Time-restricted feeding entrains daily rhythms of energy metabolism in mice. Am. J. Physiol. Regul. Integr. Comp. Physiol.290, R1276–R1283 (2006). [DOI] [PubMed] [Google Scholar]
49.Vollmers, C. et al. Time of feeding and the intrinsic circadian clock drive rhythms in hepatic gene expression. Proc. Natl. Acad. Sci. USA106, 21453–21458 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Zitting, K.-M. et al. Human resting energy expenditure varies with circadian phase. Curr. Biol.28, 3685–3690.e3 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Kriegsfeld, L. J., Leak, R. K., Yackulic, C. B., LeSauter, J. & Silver, R. Organization of suprachiasmatic nucleus projections in Syrian hamsters (Mesocricetus auratus): an anterograde and retrograde analysis. J. Comp. Neurol.468, 361–379 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Ono, D., Honma, K.-I., Yanagawa, Y., Yamanaka, A. & Honma, S. Role of GABA in the regulation of the central circadian clock of the suprachiasmatic nucleus. J. Physiol. Sci.68, 333–343 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Schutz, Y. Abnormalities of fuel utilization as predisposing to the development of obesity in humans. Obes. Res3, 173S–178S (1995). [DOI] [PubMed] [Google Scholar]
54.Zurlo, F. et al. Low ratio of fat to carbohydrate oxidation as predictor of weight gain: study of 24-h RQ. Am. J. Physiol.259, E650–E657 (1990). [DOI] [PubMed] [Google Scholar]
55.Balke, J. E., Zhang, L. & Percival, J. M. Neuronal nitric oxide synthase (nNOS) splice variant function: insights into nitric oxide signaling from skeletal muscle. Nitric Oxide82, 35–47 (2019). [DOI] [PubMed] [Google Scholar]
56.Yang, H. et al. One-step generation of mice carrying reporter and conditional alleles by CRISPR/Cas-mediated genome engineering. Cell154, 1370–1379 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Engel, E. A., Card, J. P. & Enquist, L. W. Transneuronal circuit analysis with pseudorabies viruses. Curr. Protoc.3, e841 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Osakada, F. & Callaway, E. M. Design and generation of recombinant rabies virus vectors. Nat. Protoc.8, 1583–1601 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Paxinos, G. & Franklin, K. B. J. Paxinos and Franklin’s the Mouse Brain in Stereotaxic Coordinates. (Academic Press, 2019).
60.Harima, Y. et al. Parallel labeled-line organization of sympathetic outflow for selective organ regulation in mice. Nat. Commun.15, 10478 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Ishihara, K., Oyaizu, S., Onuki, K., Lim, K. & Fushiki, T. Chronic (-)-hydroxycitrate administration spares carbohydrate utilization and promotes lipid oxidation during exercise in mice. J. Nutr.130, 2990–2995 (2000). [DOI] [PubMed] [Google Scholar]
62.Hagenfeldt, L. A gas chromatographic method for the determination of individual free fatty acids in plasma. Clin. Chim. Acta13, 266–268 (1966). [DOI] [PubMed] [Google Scholar]
63.Saito, M., Minokoshi, Y. & Shimazu, T. Accelerated norepinephrine turnover in peripheral tissues after ventromedial hypothalamic stimulation in rats. Brain Res.481, 298–303 (1989). [DOI] [PubMed] [Google Scholar]
64.Chi, J., Crane, A., Wu, Z. & Cohen, P. Adipo-clear: a tissue clearing method for three-dimensional imaging of adipose tissue. J. Vis. Exp.137, 58271 (2018). [DOI] [PMC free article] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information^{(29.8MB, pdf)}

Reporting Summary^{(2.8MB, pdf)}

Transparent Peer Review file^{(867KB, pdf)}

Source Data^{(12.3MB, xlsx)}

Data Availability Statement

The data generated in this study are provided in the Source Data file. All microscopy data are provided upon request to the corresponding authors. Source data are provided with this paper.

[CR1] 1.World Health Organization. Obesity and overweight. https://www.who.int/en/news-room/fact-sheets/detail/obesity-and-overweight.

[CR2] 2.Manceau, R., Majeur, D. & Alquier, T. Neuronal control of peripheral nutrient partitioning. Diabetologia63, 673–682 (2020). [DOI] [PubMed] [Google Scholar]

[CR3] 3.Goodpaster, B. H. & Sparks, L. M. Metabolic flexibility in health and disease. Cell Metab.25, 1027–1036 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Smith, R. L., Soeters, M. R., Wüst, R. C. I. & Houtkooper, R. H. Metabolic flexibility as an adaptation to energy resources and requirements in health and disease. Endocr. Rev.39, 489–517 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Randle, P. J., Garland, P. B., Hales, C. N. & Newsholme, E. A. The glucose fatty-acid cycle. Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet1, 785–789 (1963). [DOI] [PubMed] [Google Scholar]

[CR6] 6.Myers, M. G., Jr & Olson, D. P. Central nervous system control of metabolism. Nature491, 357–363 (2012). [DOI] [PubMed]

[CR7] 7.Brüning, J. C. & Fenselau, H. Integrative neurocircuits that control metabolism and food intake. Science381, eabl7398 (2023). [DOI] [PubMed] [Google Scholar]

[CR8] 8.Oldfield, B. J. et al. The neurochemical characterisation of hypothalamic pathways projecting polysynaptically to brown adipose tissue in the rat. Neuroscience110, 515–526 (2002). [DOI] [PubMed] [Google Scholar]

[CR9] 9.Bamshad, M., Song, C. K. & Bartness, T. J. CNS origins of the sympathetic nervous system outflow to brown adipose tissue. Am. J. Physiol. Regulatory Integrative Comparative Physiol.276, 1569 (1999). [DOI] [PubMed]

[CR10] 10.Kerman, I. A., Akil, H. & Watson, S. J. Rostral elements of sympatho-motor circuitry: a virally mediated transsynaptic tracing study. J. Neurosci.26, 3423–3433 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.de Souza Cordeiro, L. M. et al. Hypothalamic MC4R regulates glucose homeostasis through adrenaline-mediated control of glucose reabsorption via renal GLUT2 in mice. Diabetologia64, 181–194 (2021). [DOI] [PMC free article] [PubMed]

[CR12] 12.Toda, C. et al. Distinct effects of leptin and a melanocortin receptor agonist injected into medial hypothalamic nuclei on glucose uptake in peripheral tissues. Diabetes58, 2757–2765 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Sutton, A. K. et al. Control of food intake and energy expenditure by nos1 neurons of the paraventricular hypothalamus. J. Neurosci.34, 15306–15318 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Sutton, A. K. et al. Paraventricular, subparaventricular and periventricular hypothalamic IRS4-expressing neurons are required for normal energy balance. Sci. Rep.10, 5546 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Sawchenko, P. E. Evidence for differential regulation of corticotropin-releasing factor and vasopressin immunoreactivities in parvocellular neurosecretory and autonomic-related projections of the paraventricular nucleus. Brain Res.437, 253–263 (1987). [DOI] [PubMed] [Google Scholar]

[CR16] 16.Hallbeck, M., Larhammar, D. & Blomqvist, A. Neuropeptide expression in rat paraventricular hypothalamic neurons that project to the spinal cord. J. Comp. Neurol.433, 222–238 (2001). [DOI] [PubMed] [Google Scholar]

[CR17] 17.Zhang, X. et al. Brain control of humoral immune responses amenable to behavioural modulation. Nature581, 204–208 (2020). [DOI] [PubMed] [Google Scholar]

[CR18] 18.Swanson, L. W. & Kuypers, H. G. The paraventricular nucleus of the hypothalamus: cytoarchitectonic subdivisions and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J. Comp. Neurol.194, 555–570 (1980). [DOI] [PubMed] [Google Scholar]

[CR19] 19.Kondoh, K. et al. A specific area of olfactory cortex involved in stress hormone responses to predator odours. Nature532, 103–106 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Lin, H.-W. et al. Role of the UL41 protein of pseudorabies virus in host shutoff, pathogenesis and induction of TNF-α expression. J. Vet. Med. Sci.72, 1179–1187 (2010). [DOI] [PubMed] [Google Scholar]

[CR21] 21.Nogueiras, R. et al. The central melanocortin system directly controls peripheral lipid metabolism. J. Clin. Invest.117, 3475–3488 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Zhu, C. et al. Disrupted hypothalamic CRH neuron responsiveness contributes to diet-induced obesity. EMBO Rep. 21, EMBR201949210 (2020). [DOI] [PMC free article] [PubMed]

[CR23] 23.Huesing, C. et al. Sympathetic innervation of inguinal white adipose tissue in the mouse. J. Comp. Neurol. 529, 1465–1485 (2021). [DOI] [PMC free article] [PubMed]

[CR24] 24.François, M. et al. Sympathetic innervation of the interscapular brown adipose tissue in mouse. Ann. N. Y. Acad. Sci.1454, 3–13 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Minokoshi, Y. et al. Leptin stimulates fatty-acid oxidation by activating AMP-activated protein kinase. Nature415, 339–343 (2002). [DOI] [PubMed] [Google Scholar]

[CR26] 26.Oakes, N. D. et al. Development and initial evaluation of a novel method for assessing tissue-specific plasma free fatty acid utilization in vivo using (R)−2-bromopalmitate tracer. J. Lipid Res.40, 1155–1169 (1999). [PubMed] [Google Scholar]

[CR27] 27.Butler, A. A. et al. Melanocortin-4 receptor is required for acute homeostatic responses to increased dietary fat. Nat. Neurosci.4, 605–611 (2001). [DOI] [PubMed] [Google Scholar]

[CR28] 28.Ishihara, K. et al. Amino acid composition of soybean protein increased postprandial carbohydrate oxidation in diabetic mice. Biosci. Biotechnol. Biochem.67, 2505–2511 (2003). [DOI] [PubMed] [Google Scholar]

[CR29] 29.Albarado, D. C. et al. Impaired coordination of nutrient intake and substrate oxidation in melanocortin-4 receptor knockout mice. Endocrinology145, 243–252 (2004). [DOI] [PubMed] [Google Scholar]

[CR30] 30.Joly-Amado, A. et al. Hypothalamic AgRP-neurons control peripheral substrate utilization and nutrient partitioning. EMBO J.31, 4276–4288 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Grassi, G., Bombelli, M., Seravalle, G., Dell’Oro, R. & Quarti-Trevano, F. Diurnal blood pressure variation and sympathetic activity. Hypertens. Res.33, 381–385 (2010). [DOI] [PubMed] [Google Scholar]

[CR32] 32.Anderson, C. et al. Non-canonical role of the sympathetic nervous system in the day-night rhythm in heart rate. Philos. Trans. R. Soc. Lond. B Biol. Sci.378, 20220179 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.Solanki, K. et al. The expanding roles of neuronal nitric oxide synthase (NOS1). PeerJ10, e13651 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Zhou, Q.-G., Zhu, X.-H., Nemes, A. D. & Zhu, D.-Y. Neuronal nitric oxide synthase and affective disorders. IBRO Rep.5, 116–132 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Foster, R. G., Hughes, S. & Peirson, S. N. Circadian photoentrainment in mice and humans. Biology9, 180 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Duffy, J. F. & Wright, K. P. Jr. Entrainment of the human circadian system by light. J. Biol. Rhythms20, 326–338 (2005). [DOI] [PubMed] [Google Scholar]

[CR37] 37.LeGates, T. A., Fernandez, D. C. & Hattar, S. Light as a central modulator of circadian rhythms, sleep and affect. Nat. Rev. Neurosci.15, 443–454 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Hastings, M. H., Maywood, E. S. & Brancaccio, M. Generation of circadian rhythms in the suprachiasmatic nucleus. Nat. Rev. Neurosci.19, 453–469 (2018). [DOI] [PubMed] [Google Scholar]

[CR39] 39.Berson, D. M., Dunn, F. A. & Takao, M. Phototransduction by retinal ganglion cells that set the circadian clock. Science295, 1070–1073 (2002). [DOI] [PubMed] [Google Scholar]

[CR40] 40.Hanchate, N. K. et al. Connect-seq to superimpose molecular on anatomical neural circuit maps. Proc. Natl. Acad. Sci. USA117, 4375–4384 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR41] 41.Papazoglou, I. et al. A distinct hypothalamus-to-β cell circuit modulates insulin secretion. Cell Metab.34, 285–298.e7 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Stanley, S. et al. Identification of neuronal subpopulations that project from hypothalamus to both liver and adipose tissue polysynaptically. Proc. Natl. Acad. Sci. USA107, 7024–7029 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Haque, M. S. et al. Role of the sympathetic nervous system and insulin in enhancing glucose uptake in peripheral tissues after intrahypothalamic injection of leptin in rats. Diabetes48, 1702–1706 (1999). [DOI] [PubMed]

[CR44] 44.Minokoshi, Y., Haque, M. S. & Shimazu, T. Microinjection of leptin into the ventromedial hypothalamus increases glucose uptake in peripheral tissues in rats. Diabetes48, 287–291 (1999). [DOI] [PubMed] [Google Scholar]

[CR45] 45.Shiuchi, T. et al. Induction of glucose uptake in skeletal muscle by central leptin is mediated by muscle β2-adrenergic receptor but not by AMPK. Sci. Rep.7, 15141 (2017). [DOI] [PMC free article] [PubMed]

[CR46] 46.Shiuchi, T. et al. Hypothalamic orexin stimulates feeding-associated glucose utilization in skeletal muscle via sympathetic nervous system. Cell Metab.10, 466–480 (2009). [DOI] [PubMed] [Google Scholar]

[CR47] 47.Randle, P. J. Regulatory interactions between lipids and carbohydrates: the glucose fatty acid cycle after 35 years. Diab. Metab. Rev.14, 263–283 (1998). [DOI] [PubMed] [Google Scholar]

[CR48] 48.Satoh, Y., Kawai, H., Kudo, N., Kawashima, Y. & Mitsumoto, A. Time-restricted feeding entrains daily rhythms of energy metabolism in mice. Am. J. Physiol. Regul. Integr. Comp. Physiol.290, R1276–R1283 (2006). [DOI] [PubMed] [Google Scholar]

[CR49] 49.Vollmers, C. et al. Time of feeding and the intrinsic circadian clock drive rhythms in hepatic gene expression. Proc. Natl. Acad. Sci. USA106, 21453–21458 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR50] 50.Zitting, K.-M. et al. Human resting energy expenditure varies with circadian phase. Curr. Biol.28, 3685–3690.e3 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Kriegsfeld, L. J., Leak, R. K., Yackulic, C. B., LeSauter, J. & Silver, R. Organization of suprachiasmatic nucleus projections in Syrian hamsters (Mesocricetus auratus): an anterograde and retrograde analysis. J. Comp. Neurol.468, 361–379 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Ono, D., Honma, K.-I., Yanagawa, Y., Yamanaka, A. & Honma, S. Role of GABA in the regulation of the central circadian clock of the suprachiasmatic nucleus. J. Physiol. Sci.68, 333–343 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR53] 53.Schutz, Y. Abnormalities of fuel utilization as predisposing to the development of obesity in humans. Obes. Res3, 173S–178S (1995). [DOI] [PubMed] [Google Scholar]

[CR54] 54.Zurlo, F. et al. Low ratio of fat to carbohydrate oxidation as predictor of weight gain: study of 24-h RQ. Am. J. Physiol.259, E650–E657 (1990). [DOI] [PubMed] [Google Scholar]

[CR55] 55.Balke, J. E., Zhang, L. & Percival, J. M. Neuronal nitric oxide synthase (nNOS) splice variant function: insights into nitric oxide signaling from skeletal muscle. Nitric Oxide82, 35–47 (2019). [DOI] [PubMed] [Google Scholar]

[CR56] 56.Yang, H. et al. One-step generation of mice carrying reporter and conditional alleles by CRISPR/Cas-mediated genome engineering. Cell154, 1370–1379 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Engel, E. A., Card, J. P. & Enquist, L. W. Transneuronal circuit analysis with pseudorabies viruses. Curr. Protoc.3, e841 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Osakada, F. & Callaway, E. M. Design and generation of recombinant rabies virus vectors. Nat. Protoc.8, 1583–1601 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR59] 59.Paxinos, G. & Franklin, K. B. J. Paxinos and Franklin’s the Mouse Brain in Stereotaxic Coordinates. (Academic Press, 2019).

[CR60] 60.Harima, Y. et al. Parallel labeled-line organization of sympathetic outflow for selective organ regulation in mice. Nat. Commun.15, 10478 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Ishihara, K., Oyaizu, S., Onuki, K., Lim, K. & Fushiki, T. Chronic (-)-hydroxycitrate administration spares carbohydrate utilization and promotes lipid oxidation during exercise in mice. J. Nutr.130, 2990–2995 (2000). [DOI] [PubMed] [Google Scholar]

[CR62] 62.Hagenfeldt, L. A gas chromatographic method for the determination of individual free fatty acids in plasma. Clin. Chim. Acta13, 266–268 (1966). [DOI] [PubMed] [Google Scholar]

[CR63] 63.Saito, M., Minokoshi, Y. & Shimazu, T. Accelerated norepinephrine turnover in peripheral tissues after ventromedial hypothalamic stimulation in rats. Brain Res.481, 298–303 (1989). [DOI] [PubMed] [Google Scholar]

[CR64] 64.Chi, J., Crane, A., Wu, Z. & Cohen, P. Adipo-clear: a tissue clearing method for three-dimensional imaging of adipose tissue. J. Vis. Exp.137, 58271 (2018). [DOI] [PMC free article] [PubMed]

PERMALINK

Nos1 neurons in the paraventricular hypothalamic area modulate lipid metabolism via the sympathetic nervous system in male mice

Kunio Kondoh

Emi Hashiuchi

Yuka Inaba

Hiroshi Inoue

Serika Yamada

Kazunari Miyamichi

Yuki Yoshimura

Ken-ichiro Nakajima

Takeshi Y Hiyama

Yasuhiko Minokoshi

Abstract

Introduction

Results

Nos1 cells constitute a substantial population of PVH neurons that convey signals via neural pathway to peripheral tissues

Fig. 1. Nos1 neurons of the PVH project to peripheral metabolic organs.

Activation of Nos1 neurons in the PVH area increases whole-body fat consumption

Fig. 2. Activation of Nos1 neurons in the PVH area increases fat consumption.

Fig. 5. Activity of Nos1 neurons in the PVH area and the sympathetic nervous system is required for fat consumption during the light period.

Nos1 neurons in the PVH area increase lipid turnover

Nos1 neurons in the PVH area increase fat consumption via the sympathetic nervous system

Fig. 3. Nos1 neurons in the PVH area control fat consumption via the sympathetic nervous system.

Nos1 neurons in the PVH area and the sympathetic nervous system are required for the change in fat consumption linked to the diurnal rhythm

Fig. 4. Chronic silencing of Nos1 neuron activity suppressed the fat consumption during the light period.

NOS1 is not required for fat consumption during the light period

Nos1 neurons are active during the light period

Fig. 6. Activity of Nos1 neurons in the PVH area is increased during the light period.

Long-term silencing of Nos1 neurons in the PVH area results in obesity independent of energy intake

Fig. 7. Chronic silencing of Nos1 neurons in the PVH area induces obesity independent of energy intake.

Fig. 8. Silencing of Nos1 neuron activity in the PVH area disrupts metabolic adaptation to cold exposure and food deprivation.

Nos1 neuron activity in the PVH area is required for metabolic adaptation to cold exposure, food deprivation, and other stressors

Fig. 9. Silencing of Nos1 neuron activity in the PVH area disrupts metabolic adaptation to other stressors.

Discussion

Methods

Animals

Cells

PRV preparation

Rabies virus preparation

Virus injection into peripheral tissues

Stereotaxic injection and optic fiber implantation in the brain

Virus injection in the spinal cord

Stimulation of Nos1 neurons by hM3Dq

Transient inhibition of Nos1 neurons by hM4Di

Silencing of Nos1 neurons by TeNT

Deletion of Nos1 in the PVH

Pair-feeding

Indirect calorimetry

Administration of adrenergic receptor antagonists

Cold exposure, stress treatment, and fasting

Fiber photometry

Analysis of fiber photometry data

Measurement of in vivo fatty acid oxidation

NA turnover assay and measurement of tissue NA content

Tissue collection

Immunohistofluorescence staining

In situ hybridization combined with immunohistofluorescence staining

Hematoxylin-eosin staining

RNA extraction and RT-qPCR analysis

Immunoblot analysis

Measurement of plasma glucose, insulin, NEFA, glycerol, triacylglycerol, corticosterone, and TSH concentrations

Image acquisition and analysis

Statistical analysis

Reporting summary

Supplementary information

Source data

Acknowledgements

Author contributions

Peer review

Peer review information

Fundings

Data availability

Competing interests

Footnotes

Contributor Information

Supplementary information

References

Associated Data

Supplementary Materials

Data Availability Statement