A discrete neuronal circuit induces a hibernation-like state in rodents


Hibernating mammals actively lower their body temperature to reduce energy expenditure when facing food scarcity1. This ability to induce a hypometabolic state has evoked great interest owing to its potential medical benefits2,3. Here we show that a hypothalamic neuronal circuit in rodents induces a long-lasting hypothermic and hypometabolic state similar to hibernation. In this state, although body temperature and levels of oxygen consumption are kept very low, the ability to regulate metabolism still remains functional, as in hibernation4. There was no obvious damage to tissues and organs or abnormalities in behaviour after recovery from this state. Our findings could enable the development of a method to induce a hibernation-like state, which would have potential applications in non-hibernating mammalian species including humans.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: Activating QrfpiCre neurons in the hypothalamus lowers body temperature and energy expenditure.
Fig. 2: Histological and functional analyses of Q neuron projections.
Fig. 3: Q-neuron-induced hypometabolism is accompanied by a lowered set-point of body temperature.
Fig. 4: Glutamatergic and GABAergic neurotransmission of Q neurons are both involved in inducing QIH.

Data availability

Data used for Bayesian estimation are included with the source code. Other data are available from the corresponding authors on request.

Code availability

All of the source code of the models used for Bayesian estimation is available at https://briefcase.riken.jp/public/JjtgwAnqQ8lAgyI. See ‘Statistical analysis’ in Methods for details.


  1. 1.

    Geiser, F. Hibernation. Curr. Biol. 23, R188–R193 (2013).

    CAS  PubMed  Google Scholar 

  2. 2.

    Melvin, R. G. & Andrews, M. T. Torpor induction in mammals: recent discoveries fueling new ideas. Trends Endocrinol. Metab. 20, 490–498 (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  3. 3.

    Bouma, H. R. et al. Induction of torpor: mimicking natural metabolic suppression for biomedical applications. J. Cell. Physiol. 227, 1285–1290 (2012).

    ADS  CAS  PubMed  Google Scholar 

  4. 4.

    Jastroch, M. et al. Seasonal control of mammalian energy balance: recent advances in the understanding of daily torpor and hibernation. J. Neuroendocrinol. 28, 12347 (2016).

    Google Scholar 

  5. 5.

    Sunagawa, G. A. & Takahashi, M. Hypometabolism during daily torpor in mice is dominated by reduction in the sensitivity of the thermoregulatory system. Sci. Rep. 6, 37011 (2016).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  6. 6.

    Vicent, M. A., Borre, E. D. & Swoap, S. J. Central activation of the A1 adenosine receptor in fed mice recapitulates only some of the attributes of daily torpor. J. Comp. Physiol. B 187, 835–845 (2017).

    CAS  PubMed  PubMed Central  Google Scholar 

  7. 7.

    Hitrec, T. et al. Neural control of fasting-induced torpor in mice. Sci. Rep. 9, 15462 (2019).

    ADS  PubMed  PubMed Central  Google Scholar 

  8. 8.

    Griko, Y. & Regan, M. D. Synthetic torpor: a method for safely and practically transporting experimental animals aboard spaceflight missions to deep space. Life Sci. Space Res. 16, 101–107 (2018).

    ADS  Google Scholar 

  9. 9.

    Choukèr, A., Bereiter-Hahn, J., Singer, D. & Heldmaier, G. Hibernating astronauts—science or fiction? Pflugers Arch. 471, 819–828 (2019).

    PubMed  Google Scholar 

  10. 10.

    Fukusumi, S. et al. A new peptidic ligand and its receptor regulating adrenal function in rats. J. Biol. Chem. 278, 46387–46395 (2003).

    CAS  PubMed  Google Scholar 

  11. 11.

    Chartrel, N. et al. Identification of 26RFa, a hypothalamic neuropeptide of the RFamide peptide family with orexigenic activity. Proc. Natl Acad. Sci. USA 100, 15247–15252 (2003).

    ADS  CAS  PubMed  Google Scholar 

  12. 12.

    Takayasu, S. et al. A neuropeptide ligand of the G protein-coupled receptor GPR103 regulates feeding, behavioral arousal, and blood pressure in mice. Proc. Natl Acad. Sci. USA 103, 7438–7443 (2006).

    ADS  CAS  PubMed  Google Scholar 

  13. 13.

    Okamoto, K. et al. QRFP-deficient mice are hypophagic, lean, hypoactive and exhibit increased anxiety-like behavior. PLoS ONE 11, e0164716 (2016).

    PubMed  PubMed Central  Google Scholar 

  14. 14.

    Roth, B. L. DREADDs for neuroscientists. Neuron 89, 683–694 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  15. 15.

    Atasoy, D., Aponte, Y., Su, H. H. & Sternson, S. M. A FLEX switch targets channelrhodopsin-2 to multiple cell types for imaging and long-range circuit mapping. J. Neurosci. 28, 7025–7030 (2008).

    CAS  PubMed  PubMed Central  Google Scholar 

  16. 16.

    Nakamura, K. Central circuitries for body temperature regulation and fever. Am. J. Physiol. Regul. Integr. Comp. Physiol. 301, R1207–R1228 (2011).

    CAS  PubMed  Google Scholar 

  17. 17.

    Zhao, Z.-D. et al. A hypothalamic circuit that controls body temperature. Proc. Natl Acad. Sci. USA 114, 2042–2047 (2017).

    CAS  PubMed  Google Scholar 

  18. 18.

    Machado, N. L. S. et al. A glutamatergic hypothalamomedullary circuit mediates thermogenesis, but not heat conservation, during stress-induced hyperthermia. Curr. Biol. 28, 2291–2301 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  19. 19.

    Piñol, R. A. et al. Brs3 neurons in the mouse dorsomedial hypothalamus regulate body temperature, energy expenditure, and heart rate, but not food intake. Nat. Neurosci. 21, 1530–1540 (2018).

    PubMed  PubMed Central  Google Scholar 

  20. 20.

    Yizhar, O. et al. Neocortical excitation/inhibition balance in information processing and social dysfunction. Nature 477, 171–178 (2011).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  21. 21.

    Morrison, S. F. Central control of body temperature. F1000Res. 5, 880 (2016).

    Google Scholar 

  22. 22.

    Ortmann, S. & Heldmaier, G. Regulation of body temperature and energy requirements of hibernating alpine marmots (Marmota marmota). Am. J. Physiol. Regul. Integr. Comp. Physiol. 278, R698–R704 (2000).

    CAS  PubMed  Google Scholar 

  23. 23.

    Snapp, B. D. & Heller, H. C. Suppression of metabolism during hibernation in ground squirrels (Citellus lateralis). Physiol. Zool. 54, 297–307 (1981).

    Google Scholar 

  24. 24.

    Moffitt, J. R. et al. Molecular, spatial, and functional single-cell profiling of the hypothalamic preoptic region. Science 362, eaau5324 (2018).

    ADS  PubMed  PubMed Central  Google Scholar 

  25. 25.

    Chen, R., Wu, X., Jiang, L. & Zhang, Y. Single-cell RNA-seq reveals hypothalamic cell diversity. Cell Rep. 18, 3227–3241 (2017).

    CAS  PubMed  PubMed Central  Google Scholar 

  26. 26.

    Walker, J. M., Glotzbach, S. F., Berger, R. J. & Heller, H. C. Sleep and hibernation in ground squirrels (Citellus spp): electrophysiological observations. Am. J. Physiol. Integr. Comp. Physiol. 233, R213–R221 (1977).

    CAS  Google Scholar 

  27. 27.

    Bratincsák, A. et al. Spatial and temporal activation of brain regions in hibernation: c-fos expression during the hibernation bout in thirteen-lined ground squirrel. J. Comp. Neurol. 505, 443–458 (2007).

    PubMed  PubMed Central  Google Scholar 

  28. 28.

    Dausmann, K. H., Glos, J., Ganzhorn, J. U. & Heldmaier, G. Physiology: hibernation in a tropical primate. Nature 429, 825–826 (2004).

    ADS  CAS  PubMed  Google Scholar 

  29. 29.

    Cerri, M. et al. The inhibition of neurons in the central nervous pathways for thermoregulatory cold defense induces a suspended animation state in the rat. J. Neurosci. 33, 2984–2993 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  30. 30.

    Heldmaier, G., Ortmann, S. & Elvert, R. Natural hypometabolism during hibernation and daily torpor in mammals. Respir. Physiol. Neurobiol. 141, 317–329 (2004).

    PubMed  Google Scholar 

  31. 31.

    Oomura, Y. et al. A new brain glucosensor and its physiological significance. Am. J. Clin. Nutr. 55, 278S–282S (1992).

    CAS  PubMed  Google Scholar 

  32. 32.

    Tan, C. L. et al. Warm-sensitive neurons that control body temperature. Cell 167, 47–59 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  33. 33.

    Mieda, M. et al. Cellular clocks in AVP neurons of the SCN are critical for interneuronal coupling regulating circadian behavior rhythm. Neuron 85, 1103–1116 (2015).

    CAS  PubMed  Google Scholar 

  34. 34.

    Osakada, F. & Callaway, E. M. Design and generation of recombinant rabies virus vectors. Nat. Protoc. 8, 1583–1601 (2013).

    PubMed  PubMed Central  Google Scholar 

  35. 35.

    Saito, Y. C. et al. Monoamines inhibit GABAergic neurons in ventrolateral preoptic area that make direct synaptic connections to hypothalamic arousal neurons. J. Neurosci. 38, 6366–6378 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  36. 36.

    Gehrmann, J. et al. Phenotypic screening for heart rate variability in the mouse. Am. J. Physiol. Heart Circ. Physiol. 279, H733–H740 (2000).

    CAS  PubMed  Google Scholar 

  37. 37.

    Sunagawa, G. A. et al. Mammalian reverse genetics without crossing reveals Nr3a as a short-sleeper gene. Cell Rep. 14, 662–677 (2016).

    CAS  PubMed  Google Scholar 

  38. 38.

    Franklin, K. B. J. & Paxinos, G. The Mouse Brain in Stereotaxic Coordinates 3rd edn (Academic Press, 2007).

  39. 39.

    Paxinos, G. & Watson, C. The Rat Brain in Stereotaxic Coordinates 7th edn (Academic Press, 2013).

  40. 40.

    Hama, H. et al. ScaleS: an optical clearing palette for biological imaging. Nat. Neurosci. 18, 1518–1529 (2015).

    CAS  PubMed  Google Scholar 

  41. 41.

    Stan Development Team. RStan: the R interface to Stan. R package v.2.19.3 (2020).

  42. 42.

    R Core Team. R: A language and environment for statistical computing, v.2.18.0 (R Foundation for Statistical Computing, 2018).

  43. 43.

    McElreath, R. Statistical Rethinking: a Bayesian Course with Examples in R and Stan 1st edn (CRC Press, 2016).

  44. 44.

    Romanov, R. A. et al. Molecular interrogation of hypothalamic organization reveals distinct dopamine neuronal subtypes. Nat. Neurosci. 20, 176–188 (2017).

    CAS  PubMed  Google Scholar 

  45. 45.

    Wickersham, I. R. et al. Monosynaptic restriction of transsynaptic tracing from single, genetically targeted neurons. Neuron 53, 639–647 (2007).

    CAS  PubMed  PubMed Central  Google Scholar 

  46. 46.

    Wang, T. A. et al. Thermoregulation via temperature-dependent PGD2 production in mouse preoptic area. Neuron 103, 309–322 (2019).

    CAS  PubMed  Google Scholar 

  47. 47.

    Tan, C. L. & Knight, Z. A. Regulation of body temperature by the nervous system. Neuron 98, 31–48 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  48. 48.

    Lazarus, M. et al. EP3 prostaglandin receptors in the median preoptic nucleus are critical for fever responses. Nat. Neurosci. 10, 1131–1133 (2007).

    CAS  PubMed  Google Scholar 

Download references


This study was supported by a JSPS KAKENHI Grant-in-Aid for Scientific Research (B) (JP 18H02595) (T.S.); a JSPS KAKENHI Grant-in-Aid for Scientific Research on Innovative Areas ‘Willdynamics’ (16H06401); JSPS KAKENHI grant number JP19K22465 (T.S.); JST CREST grant number JPMJCR1655 Japan (T.S.); the RIKEN Special Postdoctoral Researcher program (G.A.S.); a JSPS KAKENHI Grant-in-Aid for Scientific Research on Innovative Areas ‘Thermal Biology’ (18H04706) (G.A.S.); a JSPS KAKENHI Grant-in-Aid for Scientific Research (A) (19H01066) (G.A.S.); a research grant from the Astellas Foundation for Research on Metabolic Disorders (G.A.S.); and JSPS KAKENHI grant number 19J20876 (T.M.T.). We thank the animal resource centre at the University of Tsukuba and LARGE, RIKEN BDR for housing the mice; Y. Cherasse for preparing virus vectors; A. Miyasaka and Y. Niwa for discussion; and W. Gray for proofreading the manuscript. All of the brain diagrams used in figures were made based on the illustrations in Paxinos and Franklin’s atlas of the mouse brain38 and Paxinos and Watson’s atlas of the rat brain39.

Author information




T.M.T., G.A.S. and T.S. conceived the project and designed the experiments. T.M.T. performed all stereotaxic surgeries and tissue samplings and conducted experiments except as noted. T.M.T., K.I., M.T. and G.A.S. performed metabolism and biological signal-recording experiments. T.M.T. and S.S. performed histology analyses. M.A., M.Y., T.S. and K. Sakimura designed and generated genetically modified mice. T.S. and K. Sakurai. prepared AAV vectors. T.S. prepared rabies virus vectors. H.H. and A.M. performed three-dimensional imaging of transparent brains. E.H. performed electrophysiology experiments. G.A.S. performed the statistical analyses. T.M.T., G.A.S. and T.S. wrote the manuscript, with input from all co-authors.

Corresponding authors

Correspondence to Genshiro A. Sunagawa or Takeshi Sakurai.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks Rob Henning, Shaun Morrison, Richard Palmiter and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

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

Extended data figures and tables

Extended Data Fig. 1 Generation of QrfpiCre mice.

To examine the role of QRFP-producing neurons, we engineered mice in which codon-improved Cre recombinase (iCre) is inserted in the Qrfp allele. a, Targeting vector and structure of the targeted allele of QrfpiCre mice. We mated mice with the targeted genome with FLP66 mice to delete the pgk-Neo cassette and create the QrfpiCre mice used in this study. b, Distribution of Cre-positive neurons in coronal sections of brain prepared from QrfpiCre;Ai9 mice. Scale bars, 200 μm. c, Immunostaining of hypothalamic slices prepared from a QrfpiCre;Ai9 mouse with anti-mCherry, anti-orexin and anti-melanin-concentrating hormone (MCH) antibodies. Along the wall of the third ventricle, we found extensive expression of mCherry, presumably derived from tanycytes and ependymal cells. However, we could not express exogenous genes by injecting Cre-dependent AAV vectors into this region in adult mice, suggesting transient expression of Cre in these cells during the developmental stage. In addition, we observed the existence of iCre-positive neurons in the LHA in reporter mice crossed with QrfpiCre mice that were also positive for orexin-like immunoreactivity—although a previous study did not find orexin and QRFP double-positive cells in adult mice13. This suggests that a low level of iCre is expressed in some orexin neurons, and that orexin neurons and QRFP neurons might be derived from the same cell lineage. Single-cell RNA-sequencing analysis of the hypothalamus showed colocalization of Qrfp and Orexin (also known as Hcrt), and hierarchical clustering defined by molecular fingerprints showed that orexin- and QRFP-expressing neurons have a close neuronal lineage44. The middle and right images are magnifications of the boxed areas. QRFP-expressing neurons in the LHA were positive for mCherry (arrows) but negative for MCH. Scale bars, 500 μm (left); 100 μm (middle, right). d, Growth curve of QrfpiCre mice (n = 9 wild type (WT), n = 9 QrfpiCre heterozygous and n = 10 QrfpiCre homozygous). Lines show median and shaded areas denote the estimated 89% HPDI of the body weight of each group at a given age. e, Posterior distribution of estimated difference in body weight between two groups. The dotted line shows median and solid lines denote 89% HPDI of differences. Homozygous QrfpiCre mice are smaller than wild-type mice, consistent with a previous observation13.

Extended Data Fig. 2 Expression of DREADD receptors in QrfpiCre neurons.

We generated Rosa26dreaddm3 mice and crossed them with QrfpiCre mice (Extended Data Fig. 1) to obtain mice that express hM3Dq-mCherry exclusively in iCre-expressing cells (QrfpiCre;Rosa26dreaddm3 mice). a, Generation of mice that express hM3Dq and hM4Di in Cre-expressing neurons. Targeting vectors and structures of the targeted alleles of Rosa26dreaddm3 and Rosa26dreaddm4 mice. We mated these mice with FLP66 mice to delete the pgk-Neo cassette. Orange boxes indicate hM3Dq-mCherry or hM4Di-mCherry. Because the CAG promoter drives expression of hM3Dq-mCherry or hM4Dq-mCherry only after Cre-mediated excision of the floxed stopper element, this allowed us to express hM3Dq or hM4Di specifically in Cre-expressing neurons. b, Horizontal and coronal sections of brain prepared from a QrfpiCre;Rosa26dreaddm3 mouse, showing the distribution of mCherry-positive neurons in the hypothalamus. c, Top, strategy for chemogenetic excitation or inhibition of whole iCre-positive neuronal populations in QrfpiCre mice. Bottom, chemogenetic excitation of iCre-positive cells in QrfpiCre mice induced hypothermia. Heterozygous (Q-het) or homozygous (Q-homo) QrfpiCre mice with heterozygous Rosa26dreaddm3 (M3) and/or Rosa26dreaddm4 (M4) alleles were subjected to experiments. CNO was administered at ZT12 (start of the dark period). The ambient temperature was 23 °C. We found that excitatory manipulation of QrfpiCre neurons in mice resulted in severe immobility. As the posture of these mice was similar to that observed during daily torpor, we initially postulated that activation of iCre-positive cells induced a daily torpor-like state. To evaluate this hypothesis, we measured body temperature and found that the induced state of immobility was accompanied by marked, long-lasting hypothermia. TBAT decreased beginning about 5 min after CNO administration and lasted 12 h. Mice spontaneously recovered without external warming. By contrast, inhibitory DREADD manipulation of iCre-positive neurons did not have any effect on TBAT. Notably, hM3Dq-mediated activation of iCre-positive neurons in QrfpiCre;Rosa26dreaddm3 mice induced robust hypothermia, even in homozygous QrfpiCre mice in which Qrfp sequences are completely replaced by iCre in both alleles. This suggests that QRFP itself does not have a role in inducing hypothermia. The degree of hypothermia was greater in QRFP-deficient mice, which indicates that endogenous QRFP itself counteracts the hypothermia. d, Excitatory manipulation of Q neurons in QrfpiCre;Rosa26dreaddm3 mice in the light period (at ZT1) also induced a long-lasting hypothermic state (n = 4 mice for each condition). Line and shading in c, d denote mean and s.d. of each group. AHA, anterior hypothalamus; ARC, arcuate nucleus; LPO, lateral preoptic area; MM, medial mammillary nucleus; SON, supraoptic nucleus; TMN, tuberomammillary nucleus; VMH, ventromedial hypothalamus.

Extended Data Fig. 3 DREADD-mediated excitation of Q neurons.

a, Qrfp mRNA is expressed in mCherry-positive neurons in Q-hM3D mice. Dual-colour in situ hybridization for Qrfp and mCherry mRNA in brain slices prepared from Q-hM3D mice. We confirmed that CNO administration induced QIH, and subjected the mice to histological analysis. All mCherry-positive neurons were positive for Qrfp expression. Scale bars, 100 μm (left); 10 μm (middle, right). b, Representative trace of current-clamp recording from mCherry-positive Q neurons in a slice prepared from Q-hM3D mice. We performed the experiments nine times and obtained the same results. c, Comparison of spike frequency at baseline and after treatment with CNO (n = 9). d, Estimated distribution of spike frequency in baseline and CNO-treated slices. e, The estimated difference in spike frequency between CNO-treated and baseline slices was [1.44, 2.80] Hz. Because the 89% HPDI of the estimated difference is positive, the spike frequency in CNO-treated slices may be larger than baseline by more than 89%.

Extended Data Fig. 4 TBAT decreases concomitantly with body temperature during QIH.

Representative traces of TBAT examined by thermographic camera (orange) and body temperature measured by telemetry sensor (red) before and after induction of QIH in a Q-hM3D mouse, simultaneously. Grey bars indicate locomotor activity. Note that TBAT and body temperature show almost the same values both before and after induction of QIH. A.U. arbitrary units.

Extended Data Fig. 5 QIH is accompanied by low heart rate, low EEG amplitude and weak respiration.

ECG, EEG, VO2 and respiratory flow were recorded during normal, FIT and QIH states (n = 5) in Q-hM3D mice. a, The one-hour median of the heart rate (HR) at minimum VO2 during FIT was compared to that at minimum VO2 on the day before fasting. Both VO2 and heart rate showed marked decreases. Comparing two hours before and two hours after intraperitoneal injection of CNO, both VO2 and heart rate were lower during QIH. The respiratory rate (RR) was undetectable in both FIT and QIH states owing to low respiratory flow. During QIH, heart rate was markedly decreased (572 and 202 beats per min, two hours before and two hours after injection of CNO, respectively). The respiratory rate of mice was reduced from 333 breaths per min to a level undetectable by the method used, suggesting that their breathing was shallow. LF and HF represent high-frequency and low-frequency power (ms2) of HRV. b, Representative recordings of ECG, EEG and respiratory flow of recorded mice. Both FIT and QIH showed clear suppression of EEG amplitude. Even though movement of the chest wall was confirmed by visual inspection, respiratory flow became too low to measure the precise respiratory rate. c, C57BL/6J mice were fasted for 22 h from ZT0 to induce FIT (n = 4), followed by blood sampling at ZT22. The control group C1 (n = 3) was not fasted. The QIH group (n = 6; Q-hM3D mice) was given CNO at ZT12. Two other control groups, C2 (n = 4; QrfpiCre mice injected with AAV10-DIO-mCherry into the AVPe/MPA) and C3 (n = 4; Q-hM3D), were injected with saline at ZT12, followed by blood sampling at ZT22. Blood glucose levels decreased during QIH, and the QIH group of mice showed hypoglycaemia and hyponatraemia compared to control groups. Both FIT and QIH groups showed high levels of ketone bodies than control groups, although the QIH group exhibited a milder phenotype than the FIT group. Levels of aspartate aminotransferase (AST), creatine kinase (CK) and potassium were lower in QIH than in FIT. ALT, alanine transaminase; GLU, glucose; LDH, lactic acid dehydrogenase; T-KB, total ketone bodies. In the box plots, the lower and upper limits of the box correspond to the first and third quartiles; the centre line denotes the median; the upper whisker extends to the largest value that is no further than 1.5 times the interquartile range (IQR); the lower whisker extends to the smallest value that is no further than 1.5 × IQR; and the dots denote observed values that are larger or smaller than the whiskers.

Extended Data Fig. 6 Mice behave normally after recovery from QIH.

a, Food intake, body weight and activity of 6 mice were examined for 24 days before and after QIH. The first and second dashed vertical lines denote intraperitoneal injection of saline and CNO, respectively. Orange bars show the average daily food intake, and black dots represent the observed intake for each individual mouse. The bottom two panels show body weight (measured daily) and locomotor activity (measured hourly). Black lines are the average of six mice, and grey lines represent individual mice. b, Schematic schedule of behavioural tests. Q-hM3D mice (n = 5) and controls (n = 5; QrfpiCre mice with injection of AAV10-EF1α-DIO-mCherry into the AVPe/MPA) were compared. OFT, open-field test; NOR, novel object recognition test; EPM, elevated plus maze test; RR, rotarod; TST, tail suspension test. No apparent differences were observed in any behavioural tests. c, Results of the rotarod test. d, Results of the other tests. Box plots show the distribution of each group in specific tests; all elements of the box plots are as defined in Extended Data Fig. 5. e, Histology of tissues before and after QIH. We histologically examined whole regions in the brain, heart, kidney, liver and soleus muscles prepared from mice that did or did not experience QIH. Tissue sections were stained with haematoxylin and eosin. No gross pathophysiological changes were apparent in any of the tissues examined. Scale bars, 200 μm (brain and kidney); 100 μm (heart and soleus muscle); 400 μm (liver). f, Representative traces of body temperature and VO2 during QIH, which lasts for several days and can be re-induced by another injection of CNO. Line and shading denote mean and s.d. of each group.

Extended Data Fig. 7 The DMH and RPa are major target regions for the induction of QIH.

a, Strategy for delineating the axonal projection patterns of Q neurons. The neurons were visualized by injecting AAV9-hSYN-DIO-GFP into an anteromedial hypothalamic region of QrfpiCre mice to express GFP in Q neurons. b, Distribution of GFP-positive cell bodies of Q neurons in the AVPe/MPA and periventricular nucleus. Scale bars, 100 μm. c, Distribution of axons arising from Q neurons. We observed GFP-positive fibres in brain regions that are implicated in the regulation of body temperature and in sympathetic regulation. Among these regions, the DMH received especially abundant projections. Aq, aqueduct; LC, locus coeruleus; LPB, lateral parabrachial nucleus; PAG, periaqueductal grey; PVN, paraventricular hypothalamic nucleus; RVLM, rostral ventrolateral medulla; VLPO, ventrolateral preoptic area; VOLT, vascular organ of the lamina terminalis; 4V, fourth ventricle. Scale bars, 100 μm. d, Left, temporal changes in tail temperature of Q-SSFO mice (same mice as Fig. 2c, d) after optogenetic excitation. Right, representative images of thermographs. Optogenetic focal stimulation of Q neuron axons in the RPa also induced tail vasodilation. e, We implanted optic fibres in the DMH of Q-SSFO mice, applied a blue laser (1-s duration) to induce QIH and then deactivated SSFO using a 589-nm yellow laser (5-s duration) to see the effect on TBAT. The first shot of blue laser in DMH fibres rapidly triggers hypothermia. A sequential shot of yellow laser 3 min after the second shot of blue laser rapidly reverses the effect of the blue laser. Because deactivation of SSFO is not propagated along axons, this further supports the importance of the DMH projections of Q neurons in the induction of QIH. Lines and shading in d, e denote mean and s.d. of each group.

Extended Data Fig. 8 Dynamics of set-point temperature in QIH.

a, Transitions of metabolism when the ambient temperature was changed during QIH. See Fig. 3l for details. During QIH, when the ambient temperature was lowered from 28 °C to 20 °C, all mice showed decreased VO2 and body temperature. By contrast, when the ambient temperature was lowered from 20 °C to 12 °C during QIH, three out of four mice showed increased VO2 with a relatively stable body temperature. One mouse did not show an increase in VO2, which indicates individual variance in the reduction of TR. We confirmed that all mice spontaneously recovered from QIH. b, The relationship between body temperature and VO2 during QIH with changing ambient temperature. The last 48 hours of data from Fig. 3l and a were merged. The colours of the dots correspond to different ambient temperatures. c, The relationship between the minimum body temperature and VO2 during normal and QIH states. Data from Fig. 3b are summarized. Numbers in the dots denote the ambient temperature (°C) and the bars denote the distribution. d, To evaluate metabolic regulation in a normal state, wild-type C57BL/6J mice were subjected to changes in the ambient temperature. Left, the relationship between body temperature and VO2 in all mice. Of note, body temperature is tightly controlled within a narrow range—in contrast to during QIH (b). Right, change in body temperature (purple), VO2 (black) and respiratory quotient (blue) for each mouse throughout the experiment. Starting from 28 °C, the ambient temperature was lowered to 12 °C and returned to 28 °C, as shown at the top.

Extended Data Fig. 9 Hypometabolism that is induced by general anaesthesia is not regulated.

a, To evaluate how metabolic regulation during general anaesthesia was affected by ambient temperature, C57BL/6J mice (n = 4) were anaesthetized with 1% isoflurane at different ambient temperatures. Left (top row), the transition in ambient temperature. Starting from 28 °C, the set-point temperature of the chamber was lowered to 12 °C after 30 min. Because the anaesthetic machine was outside the experimental chamber and therefore the temperature of the anaesthetic gas was independent of that of the chamber, there was a delay in reaching the chamber set-point temperature. Left (middle and bottom rows), the transition in body temperature and VO2. Both decrease along with the decrease in ambient temperature. Line and shading denote mean and s.d. Right, the relationship between body temperature and VO2 in all mice. VO2 did not increase in anaesthetized mice even at low body temperature, in contrast to in QIH (compare to Extended Data Fig. 8b). b, Representative postures of mice during anaesthesia. Left, the start of isoflurane inhalation. Middle, the start of the lowering of ambient temperature. Right, 90 minutes after the set-point temperature was lowered from 28 °C to 12 °C. No change in posture was seen even at extremely low body temperature.

Extended Data Fig. 10 Blocking SNARE-complex-mediated neurotransmission in Q neurons impairs daily torpor and QIH.

a, CNO had almost no effect on body temperature and VO2 in a QrfpiCre mouse that was co-injected with AAV9-DIO-hSYN-TeTxLC-eYFP and AAV10-DIO-hM3Dq-mCherry into the AVPe/MPA (n = 1). This suggests that SNARE-mediated neurotransmission in Q neurons is indispensable for inducing QIH. b, Expression of TeTxLC–GFP in mCherry-positive neurons (Q neurons), shown by immunostaining 90 min after administration of CNO. Scale bar, 100 μm. c, Strategy for suppressing the function of Q neurons. Images show expression of TeTxLC–eYFP in the AVPe/MPA and periventricular nucleus. Scale bar, 100 μm. d, Schematic of FIT experiment schedule. e, FIT was disrupted by expressing TeTxLC in Q neurons (n = 6 mice for control and n = 5 mice for TeTxLC). The normal architecture of FIT was disrupted when neurotransmission of Q neurons was blocked in Q-TeTxLC mice. Rapid oscillatory fluctuations in metabolism were never seen in these mice. Notably, the gradual decrease in body temperature observed in these mice implies the existence of a Q-neuron-independent mechanism of metabolism reduction during FIT. f, The moving standard deviation (MSD; mean ± s.d.) was visualized for body temperature and VO2 (from e). The low MSDs that are seen in the TeTxLC group during the fasting periods demonstrate the smaller fluctuation in this group. g, FIT was induced in control, QrfpiCre heterozygous and QrfpiCre homozygous mice, showing that the lack of QRFP peptide did not affect FIT. These observations suggest that Q neurons—but not QRFP—are an indispensable component in the induction of daily torpor, and have an important role in rapidly shifting body temperature during daily torpor. The open and closed triangles denote food removal and return, respectively. h, Silencing of Q neuron neurotransmission resulted in decreased circadian fluctuations in both body temperature and VO2. The data from the first 24 h in panel e were divided into light (L) phase and dark (D) phase. Estimated differences in light phase and dark phase for both body temperature and VO2 are shown as histograms of posterior distributions. Both body temperature and VO2 showed higher values in the dark phase than in the light phase because posterior distributions are mostly positive. Although the TeTxLC group showed positive posterior distributions as well, the differences between dark phase and light phase were smaller than those in the control group. This suggests that the TeTxLC group had smaller circadian fluctuations in metabolism.

Extended Data Fig. 11 Characteristics of Q neurons.

To elucidate the possible neuronal mechanism that regulates the activity of Q neurons, we identified upstream neuronal populations that make direct synaptic contact with Q neurons by recombinant pseudotyped rabies virus vector (SADΔG(EnvA))-mediated labelling45. a, Procedure for visualizing input neurons that make mono-synaptic contact with Q neurons, using a rabies virus vector. After expressing TVA–mCherry and rabies glycoprotein (RG) in Q neurons using Cre-activatable AAV vectors35 in QrfpiCre mice, we injected SADΔG-GFP(EnvA) into the AVPe/MPA. b, Distribution of input neurons of Q neurons. Arrows show starter cells. c, Brain regions that contain input neurons. Input neurons were also observed in regions in and around the AVPe and periventricular nucleus, suggesting that local interneurons exist that regulate the function of Q neurons, and also indicating that Q neurons might form microcircuitry with interneurons within the AVPe/MPA and periventricular nucleus. Our results suggest that Q neurons receive relatively sparse direct inputs from intra-hypothalamic regions. As the MPA is implicated in the regulation of body temperature46,47, reciprocal interaction between Q neurons and the MPA might have a key role in thermoregulation. d, In situ hybridization in neurons immunostained for mCherry in the AVPe/MPA of Q-hM3D mice. Left, expression of Adcyap and Bdnf in Q neurons; right, expression of Ptger3 in Q neurons. e‚ Proportions of Adcyap-, Bdnf- and Ptger3-positive cells in Q neurons, indicating the extent to which the Q neurons overlap with genetic markers associated with thermoregulation. Numbers show the cell counts with positive signals (two mice; three slices per mouse). In the AVPe/MPA, mCherry-negative (non-Q) Adcyap1- and Bdnf-positive neurons were intermingled with Q neurons. Almost a quarter of Adcyap1- and Bdnf-positive neurons were Q neurons. We also found a small number of Q neurons that were negative for Adcyap1 and Bdnf. These observations suggest that many Q neurons constitute a subpopulation of BDNF/PACAP neurons. Although a lot of Q neurons express Adcyap1 and Bdnf, a previous report suggested that the warmth-sensing BDNF/PACAP neurons in the ventromedial preoptic area that project to the DMH are GABAergic32. As we found that excitatory Q neurons have a major role in inducing QIH (Fig. 4), Q neurons apparently constitute a unique, previously unidentified population among the group of preoptic-area neurons that are involved in thermoregulation. Notably, we found that many Q neurons express both Vgat and Vglut2 (QH neurons) (Fig. 4a). This is consistent with a previous study reporting that many BDNF/PACAP neurons in the preoptic area express both Vgat and Vglut224, because Q neurons are a subset of BDNF/PACAP neurons. Prostaglandin EP3 receptor (Ptegr3), which is implicated in causing fever16,48, is expressed in Q neurons. Again, the number of Ptegr3-positive neurons was larger than that of Q neurons, but three quarters of Q neurons expressed Ptegr3. This suggests that PGE2 inhibits Q neurons through acting on EP3 in Q neurons, although our inhibitory DREADD experiments did not show any effects on TBAT (Extended Data Fig. 2c). ac, anterior commissure; f, fornix; MnPO, median preoptic area; opt, optic tract; VLPO, ventrolateral preoptic area; VMPO, ventromedial preoptic area.

Extended Data Fig. 12 Induction of a QIH-like state in rats.

a, Procedure for the metabolic analysis with chemogenetic activation of AVPe/MPA neurons in rats. Saline and CNO were administered just before the beginning of the dark phase. Recordings were taken until the metabolism recovered to baseline levels. b, Activating AVPe/MPA neurons, including Q neurons in rats, induced a QIH-like state of hypothermia and hypometabolism (n = 7). The lines and shadings denote mean and s.d. Body temperature, VO2 and the respiratory quotient remained low for more than 24 h after intraperitoneal injection of CNO, as in mice during QIH, and then spontaneously returned to normal states. c, Representative images showing the typical posture of rats during a QIH-like state compared with during sleep. d, Schematic drawing of virus (AAV10-CaMKIIα-hM3Dq-mCherry) injections into the AVPe/MPA of the rat brain. Stereotaxic brain maps are based on Paxinos and Watson’s atlas39. The grey rectangular region in the right panel shows the area in which the following histological evaluations are focused. e, Distribution of hM3Dq-mCherry-expressing neurons in the AVPe/MPA. Arrowheads indicate hM3Dq-expressing neurons that are positive for FOS immunofluorescence 90 min after intraperitoneal injection with CNO or saline. Scale bars, 200 μm (left), 50 μm (right). f, Qrfp and mCherry transcripts detected in the AVPe/MPA of rats. Arrows denote co-expression of Qrfp and mCherry mRNAs. Scale bars, 10 μm. g, Body temperature, VO2 and respiratory quotient before and after CNO injection in rat no. 014, which did not show a QIH-like state. h, Expression of hM3Dq-mCherry in the AVPe/MPA region of rat no. 014. We observed unilateral expression of hM3Dq-mCherry in the MPA region. This suggests that bilateral proper expression of hM3Dq in the AVPe/MPA is necessary to evoke the QIH-like state. Collectively, seven out of eight rats showed a QIH-like state, characterized by a prominent decrease in body temperature. In these rats, the reduction in body temperature was accompanied by a decrease in VO2, a lowered respiratory quotient and an extended posture, showing further similarity with the QIH state in mice. The efficiency of induction of a QIH-like state in these rats is likely to be lower than that in Q-hM3D mice, owing to ectopic expression of hM3Dq in non-Q neurons within and around the AVPe/MPA.

Supplementary information


Typical immobile behaviour seen in Q-hM3D mice. The video starts at 11 hours after administration of CNO.


Cleared brain obtained by ScaleS treatment showing cell bodies and projections of Q neurons expressing GCaMP6s. The 3D image of Q neuron structure was reconstituted from images obtained by two-photon microscopy of the cleared brain of a Qrfp-iCre mouse injected with AAV9-hSYN-DIO-GCaMP6s.


Thermography of Q-SSFO mice during optogenetic induction of QIH. SSFO-expressing Q neurons in AVPe/MPA were stimulated four times by one-second light pulse every 30 minutes. The orange arrow indicates rise in tail temperature.


Representative clips of Q-hM3D mouse before and after QIH induction. Note that the mouse posture changes according to the TA even during QIH. The video is constituted of six episodes. Episode 1, normal, homeothermic state before CNO injection at TA = 28 °C. Episode 2, immobility and extended posture during QIH at TA = 28 °C. Episode 3, similar posture at TA = 20 °C. Episode 4, response to cold exposure, TA = 12 °C. Episode 5, shivering at TA = 12 °C. Episode 6, Immobility and extended posture observed 25 hours after QIH induction at TA = 28 °C.

Reporting Summary

Video 1

Typical immobile behaviour seen in Q-hM3D mice. The video starts at 11 hours after administration of CNO.

Video 2

Cleared brain obtained by ScaleS treatment showing cell bodies and projections of Q neurons expressing GCaMP6s. The 3D image of Q neuron structure was reconstituted from images obtained by two-photon microscopy of the cleared brain of a Qrfp-iCre mouse injected with AAV9-hSYN-DIO-GCaMP6s.

Video 3

Thermography of Q-SSFO mice during optogenetic induction of QIH. SSFO-expressing Q neurons in AVPe/MPA were stimulated four times by one-second light pulse every 30 minutes. The orange arrow indicates rise in tail temperature.

Video 4

Representative clips of Q-hM3D mouse before and after QIH induction. Note that the mouse posture changes according to the TA even during QIH. The video is constituted of six episodes. Episode 1, normal, homeothermic state before CNO injection at TA = 28 °C. Episode 2, immobility and extended posture during QIH at TA = 28 °C. Episode 3, similar posture at TA = 20 °C. Episode 4, response to cold exposure, TA = 12 °C. Episode 5, shivering at TA = 12 °C. Episode 6, Immobility and extended posture observed 25 hours after QIH induction at TA = 28 °C.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Takahashi, T.M., Sunagawa, G.A., Soya, S. et al. A discrete neuronal circuit induces a hibernation-like state in rodents. Nature 583, 109–114 (2020). https://doi.org/10.1038/s41586-020-2163-6

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.