Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

A hypothalamic novelty signal modulates hippocampal memory


The ability to recognize information that is incongruous with previous experience is critical for survival. Novelty signals have therefore evolved in the mammalian brain to enhance attention, perception and memory1,2. Although the importance of regions such as the ventral tegmental area3,4 and locus coeruleus5 in broadly signalling novelty is well-established, these diffuse monoaminergic transmitters have yet to be shown to convey specific information on the type of stimuli that drive them. Whether distinct types of novelty, such as contextual and social novelty, are differently processed and routed in the brain is unknown. Here we identify the supramammillary nucleus (SuM) as a novelty hub in the hypothalamus6. The SuM region is unique in that it not only responds broadly to novel stimuli, but also segregates and selectively routes different types of information to discrete cortical targets—the dentate gyrus and CA2 fields of the hippocampus—for the modulation of mnemonic processing. Using a new transgenic mouse line, SuM-Cre, we found that SuM neurons that project to the dentate gyrus are activated by contextual novelty, whereas the SuM–CA2 circuit is preferentially activated by novel social encounters. Circuit-based manipulation showed that divergent novelty channelling in these projections modifies hippocampal contextual or social memory. This content-specific routing of novelty signals represents a previously unknown mechanism that enables the hypothalamus to flexibly modulate select components of cognition.

This is a preview of subscription content, access via your institution

Access options

Buy this article

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: The SuM responds to contextual and social novelty.
Fig. 2: SuM-Cre mice exhibit robust SuM–hippocampus projections.
Fig. 3: DG-projecting and CA2-projecting SuM neurons respond to contextual and social novelty, respectively.
Fig. 4: SuM–hippocampus circuits selectively modulate contextual and social memory.

Similar content being viewed by others

Data availability

Data supporting the findings of this study are available from the corresponding authors upon reasonable request.

Code availability

Custom code used in this study is available from the corresponding authors upon reasonable request.


  1. Ranganath, C. & Rainer, G. Neural mechanisms for detecting and remembering novel events. Nat. Rev. Neurosci. 4, 193–202 (2003).

    Article  CAS  Google Scholar 

  2. van Kesteren, M. T., Ruiter, D. J., Fernández, G. & Henson, R. N. How schema and novelty augment memory formation. Trends Neurosci. 35, 211–219 (2012).

    Article  Google Scholar 

  3. Lisman, J. E. & Grace, A. A. The hippocampal–VTA loop: controlling the entry of information into long-term memory. Neuron 46, 703–713 (2005).

    Article  CAS  Google Scholar 

  4. McNamara, C. G., Tejero-Cantero, Á., Trouche, S., Campo-Urriza, N. & Dupret, D. Dopaminergic neurons promote hippocampal reactivation and spatial memory persistence. Nat. Neurosci. 17, 1658–1660 (2014).

    Article  CAS  Google Scholar 

  5. Takeuchi, T. et al. Locus coeruleus and dopaminergic consolidation of everyday memory. Nature 537, 357–362 (2016).

    Article  ADS  CAS  Google Scholar 

  6. Pan, W. X. & McNaughton, N. The supramammillary area: its organization, functions and relationship to the hippocampus. Prog. Neurobiol. 74, 127–166 (2004).

    Article  Google Scholar 

  7. Saper, C. B. & Lowell, B. B. The hypothalamus. Curr. Biol. 24, R1111–R1116 (2014).

    Article  CAS  Google Scholar 

  8. Wirtshafter, D., Stratford, T. R. & Shim, I. Placement in a novel environment induces fos-like immunoreactivity in supramammillary cells projecting to the hippocampus and midbrain. Brain Res. 789, 331–334 (1998).

    Article  CAS  Google Scholar 

  9. Ito, M., Shirao, T., Doya, K. & Sekino, Y. Three-dimensional distribution of Fos-positive neurons in the supramammillary nucleus of the rat exposed to novel environment. Neurosci. Res. 64, 397–402 (2009).

    Article  Google Scholar 

  10. Kobayashi, Y. et al. Genetic dissection of medial habenula-interpeduncular nucleus pathway function in mice. Front. Behav. Neurosci. 7, 17 (2013).

    Article  CAS  Google Scholar 

  11. Allen Institute for Brain Science. Allen Mouse Brain Atlas (2006).

  12. Franklin, K. & Paxinos, G. The Mouse Brain in Stereotaxic Coordinates (Academic, 2007).

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

    Article  CAS  Google Scholar 

  14. Soussi, R., Zhang, N., Tahtakran, S., Houser, C. R. & Esclapez, M. Heterogeneity of the supramammillary-hippocampal pathways: evidence for a unique GABAergic neurotransmitter phenotype and regional differences. Eur. J. Neurosci. 32, 771–785 (2010).

    Article  Google Scholar 

  15. Reijmers, L. G., Perkins, B. L., Matsuo, N. & Mayford, M. Localization of a stable neural correlate of associative memory. Science 317, 1230–1233 (2007).

    Article  ADS  CAS  Google Scholar 

  16. Pedersen, N. P. et al. Supramammillary glutamate neurons are a key node of the arousal system. Nat. Commun. 8, 1405 (2017).

    Article  ADS  Google Scholar 

  17. Hashimotodani, Y., Karube, F., Yanagawa, Y., Fujiyama, F. & Kano, M. Supramammillary nucleus afferents to the dentate gyrus co-release glutamate and GABA and potentiate granule cell output. Cell Rep. 25, 2704–2715 (2018).

    Article  CAS  Google Scholar 

  18. Tritsch, N. X., Granger, A. J. & Sabatini, B. L. Mechanisms and functions of GABA co-release. Nat. Rev. Neurosci. 17, 139–145 (2016).

    Article  CAS  Google Scholar 

  19. Boehringer, R. et al. Chronic loss of CA2 transmission leads to hippocampal hyperexcitability. Neuron 94, 642–655 (2017).

    Article  CAS  Google Scholar 

  20. Resendez, S. L. et al. Social stimuli induce activation of oxytocin neurons within the paraventricular nucleus of the hypothalamus to promote social behavior in male mice. J. Neurosci. 40, 2282–2295 (2020).

    Article  CAS  Google Scholar 

  21. Wu, Z., Autry, A. E., Bergan, J. F., Watabe-Uchida, M. & Dulac, C. G. Galanin neurons in the medial preoptic area govern parental behaviour. Nature 509, 325–330 (2014).

    Article  ADS  CAS  Google Scholar 

  22. Strange, B. A., Witter, M. P., Lein, E. S. & Moser, E. I. Functional organization of the hippocampal longitudinal axis. Nat. Rev. Neurosci. 15, 655–669 (2014).

    Article  CAS  Google Scholar 

  23. Leutgeb, J. K., Leutgeb, S., Moser, M. B. & Moser, E. I. Pattern separation in the dentate gyrus and CA3 of the hippocampus. Science 315, 961–966 (2007).

    Article  ADS  CAS  Google Scholar 

  24. McHugh, T. J. et al. Dentate gyrus NMDA receptors mediate rapid pattern separation in the hippocampal network. Science 317, 94–99 (2007).

    Article  ADS  CAS  Google Scholar 

  25. Wintzer, M. E., Boehringer, R., Polygalov, D. & McHugh, T. J. The hippocampal CA2 ensemble is sensitive to contextual change. J. Neurosci. 34, 3056–3066 (2014).

    Article  CAS  Google Scholar 

  26. Chiang, M. C., Huang, A. J. Y., Wintzer, M. E., Ohshima, T. & McHugh, T. J. A role for CA3 in social recognition memory. Behav. Brain Res. 354, 22–30 (2018).

    Article  CAS  Google Scholar 

  27. Hitti, F. L. & Siegelbaum, S. A. The hippocampal CA2 region is essential for social memory. Nature 508, 88–92 (2014).

    Article  ADS  CAS  Google Scholar 

  28. Piskorowski, R. A. et al. Age-dependent specific changes in area CA2 of the hippocampus and social memory deficit in a mouse model of the 22q11.2 deletion syndrome. Neuron 89, 163–176 (2016).

    Article  CAS  Google Scholar 

  29. Alexander, G. M. et al. Social and novel contexts modify hippocampal CA2 representations of space. Nat. Commun. 7, 10300 (2016).

    Article  ADS  CAS  Google Scholar 

  30. Smith, A. S., Williams Avram, S. K., Cymerblit-Sabba, A., Song, J. & Young, W. S. Targeted activation of the hippocampal CA2 area strongly enhances social memory. Mol. Psychiatry 21, 1137–1144 (2016).

    Article  CAS  Google Scholar 

  31. Meira, T. et al. A hippocampal circuit linking dorsal CA2 to ventral CA1 critical for social memory dynamics. Nat. Commun. 9, 4163 (2018).

    Article  ADS  Google Scholar 

  32. Chen, S, et al. Near-infrared deep brain stimulation via upconversion nanoparticle-mediated optogenetics. Science 359, 679–684 (2018).

    Article  ADS  CAS  Google Scholar 

Download references


We thank C. Yokoyama and Y. Mu for comments on the manuscript; J. Shi and X. Wang for experimental support; H. Kurokawa for help with preparing image data; M. Fujisawa and Y. Goto for daily assistance; the Advanced Manufacturing Support Team at RIKEN Center for Advanced Photonics for their assistance in microdrive production; all the members of the Laboratory for Circuit and Behavioral Physiology for advice; J. Johansen for advice and reagents for rabies tracing; and S. Itohara for supplying the Rosa-NLSLacZ Cre reporter mouse. This work was supported by a Japan Society for the Promotion of Science (JSPS) Postdoctoral Fellowship (16F16386 to S.C.), RIKEN Special Postdoctoral Researchers Program (S.C.), a Human Frontier Science Program Postdoctoral Fellowship (LT000579/2018 to S.C.), Grant-in-Aid for Young Scientists from MEXT (the Ministry of Education, Culture, Sports, Science and Technology of Japan) (16K18373 and 18K14857 to S.C.), RIKEN Incentive Research Project Grant for Individual Germinating Research (S.C.), Narishige Neuroscience Research Foundation Grant (S.C.), Nakatani Foundation Grant Program (S.C.), Grant-in-Aid for Scientific Research from MEXT (19H05646 to T.J.M.; 16H04663 to H. Hioki), Grant-in-Aid for Challenging Exploratory Research from MEXT (15K14357 to T.J.M.; 17K19451 to H. Hioki), Grant-in-Aid for Scientific Research on Innovative Areas from MEXT (17H05591, 17H05986 and 19H05233 to T.J.M.; 15H05948 to A.M.; 15H01430 and 18H04743 to H. Hioki), Brain/MINDS from the Japan Agency of Medical Research and Development (AMED) (JP18dm0207064 to H. Hioki), the NeurImag facility at the Institute of Psychiatry and Neurosciences of Paris (IPNP), the Foundation Recherche Médicale (FRM; FTD20170437387 to V.R.), a NARSAD independent investigator grant from the Brain and Behavior Research Foundation (R.A.P.), Ville de Paris Programme Emergences (R.A.P.), Agence Nationale de la Recherche (ANR-13-JSV4-0002-01 and ANR-18-CE37-0020-01 to R.A.P.) and RIKEN Brain Science Institute and Center for Brain Science (T.J.M.).

Author information

Authors and Affiliations



S.C. conceived the study, conducted experiments and analysed data under the guidance and supervision of T.J.M. S.C., L.H., Y.T., M.G., A.J.Y.H. and T.J.M. performed tracing and histology. S.C., L.H., M.E.W., A.Z.W. and M.G. performed behavioural experiments. S.C. performed in vivo electrophysiology and analysed data. R.B. and L.H. contributed to data collection. D.P. and S.J.M. contributed to data analysis. T.J.M. and A.J.Y.H. generated the mouse lines. A.J.Y.H. produced all AAVs. V.R., L.T., V.C. and R.A.P. performed in vitro electrophysiology, immunohistochemistry and analysis. K.N., H. Hama and A.M. performed Scale. H. Hioki provided AAV vectors for Scale. All figures were prepared by S.C. with inputs from all authors. S.C. and T.J.M. wrote the manuscript. All authors discussed the manuscript.

Corresponding authors

Correspondence to Shuo Chen or Thomas J. McHugh.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks Masanobu Kano 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 Identification of hypothalamic regions activated by contextual and social novelty.

a, Counts of FOS+ cells in hypothalamic regions that show response to contextual and social novelty. *P = 0.0140, **P = 0.0023, ***P = 0.0009, ****P < 0.0001. n.s., not significant. Two-way ANOVA with Tukey’s post-hoc test, n = 4 mice for all groups. Error bars indicate mean ± s.e.m. Open circles are values from individual mice. b, Example FOS immunostaining (red) in the hypothalamus of one mouse from a total of 4 replicates at various distances from bregma along the rostrocaudal axis of mice 1.5 h after exposure to a familiar context, a novel context or a novel mouse. AHN, anterior hypothalamic nucleus; DMH, dorsal medial hypothalamic nucleus; LHA, lateral hypothalamic area; LPO, lateral preoptic area; MM, medial mammillary nucleus; MPN, medial preoptic nucleus; MPO, medial preoptic area; PH, posterior hypothalamic nucleus; PVH, paraventricular hypothalamic nucleus; SuM, supramammillary nucleus; VMH, ventromedial hypothalamic nucleus. Scale bars, 500 μm.

Extended Data Fig. 2 Histology of recorded mice and firing rate analysis of interneurons.

a, Tetrode recordings in the SuM. b, An example coronal section of the SuM from one of 17 mice used for recording with tetrode locations indicated by arrows. Scale bar, 400 μm. c, Reconstructed tetrode tip locations (71 in target and 14 off target) from 17 mice overlaid onto schematic mouse brain slices. Numbers indicate posterior distances from bregma. d, Behavioural protocol for a recording session, including sequential exposure to a familiar context, a novel context and a novel mouse. e, Mean firing rate of all putative interneuron (pIN; n = 22 from 17 mice) units during different exposures. One-way repeated measures ANOVA. n.s., not significant. f, g, Scatter plots of the firing rates of all interneuron units during novel contextual (f) and novel social (g) exposures, plotted against firing rates during familiar contextual exposure. Coloured dots indicate interneurons that showed a significant increase in firing rate. Horizontal and vertical green bars indicate 10 Hz, the low threshold of firing rate to classify interneurons. Pie graphs show the percentage of putative interneuron units that exhibited significant increase, significant decrease, or no change in firing rate during the respective novelty exposure. h, Venn diagram of overlap between subpopulations with significantly increased activity during novel contextual and novel social exposures.

Extended Data Fig. 3 Generation and characterization of the SuM-Cre mouse line.

a, Recombination strategy for SuM-specific expression of Cre recombinase driven by the Csf2rb2 promoter. Diagram shows the schematic of the Csf2rb2 promoter-Cre-FRT-AmpR-FRT construct. b, Example anti-Cre immunostaining in the SuM of the SuM-Cre mouse (one of three replicates). Dotted white line indicates the anatomical boundary of the SuM. Scale bar, 500 μm. c, d, Coronal (c) and sagittal (d) sections (scale bars, 2 mm) prepared from an adult progeny (age 12 weeks) of SuM-Cre and Rosa-NLSLacZ crosses and immunostained with anti-β-galactosidase (β-gal, recombination marker). Representative images of from all brain regions in which recombination is identified are shown from one mouse of four replicates of each. Dotted white lines indicate the boundaries of these brain regions. On the right of each image in c is a magnified view of the area within the white box (scale bars, 1 mm). BLA, basolateral amygdala; LHA, lateral hypothalamic area; PAG, periaqueductal grey; PF, parafascicular thalamic nucleus; PH, posterior hypothalamic nucleus; RL, rostral linear nucleus raphe; SuM, supramammillary nucleus. Scale bars, 2 mm.

Extended Data Fig. 4 Brain-wide mapping of projections from SuM neurons.

a, Coronal sections prepared from a SuM-Cre mouse injected with AAV-DIO-eYFP into the SuM. AcbSh, accumbens nucleus, shell; AHN, anterior hypothalamic nucleus; CM, central medial thalamic nucleus; DG, dentate gyrus; DP, dorsal peduncular cortex; fi, fimbria of hippocampus; IAM, interanteromedial thalamic nucleus; IG, induseum griseum; LHA, lateral hypothalamic area; LHb, lateral habenular nucleus; LPO, lateral preoptic area; LS, lateral septal nucleus; MD, mediodorsal thalamic nucleus; MDl, mediodorsal thalamic nucleus, lateral part; MPO, medial preoptic area; MS, medial septal nucleus; PH, posterior hypothalamic nucleus; RE, nucleus of reuniens; SFi, septofimbrial nucleus; SHi, septohippocampal nucleus; SuM, supramammillary nucleus; vhc, ventral hippocampal commissure. Example is one mouse of four replicates. b, Sagittal sections prepared from a SuM-Cre mouse injected with AAV-DIO-eYFP into the SuM. Example shown is one mouse of 3 replicates. Scale bars, 500 μm. c, Whole-brain examination of SuM projections by clearing and rapid three-dimensional (3D) imaging with light sheet microscopy using the ScaleS method. AAV-DIO-eYFP was injected into the SuM of SuM-Cre mice to label SuM projections by Cre-dependent eYFP expression. Acb, accumbens nucleus; DG, dentate gyrus; EC, entorhinal cortex; SuM, supramammillary nucleus. Scales: 14.6 mm (x) × 17.1 mm (y) × 7.5 mm (z) for horizontal views and 11.7 mm (x) × 17.6 mm (y) × 9.0 mm (z) for the sagittal view. One example of 3 replicates is shown. d, AbScale shows eYFP+/FOS+ neurons in the SuM after exposure to contextual (left, mouse #252-1263) or social novelty (right, mouse #252-1265). Coronal slices (800-μm thickness) were cleared and FOS was stained by AbScale, followed by rapid 3D imaging with laser-scanning confocal microscopy. One example of 3 replicates is shown. Scale bars, 1 cm (top panel); 500 μm (other panels).

Extended Data Fig. 5 Retrograde tracing of SuM neurons projecting to the DG and CA2.

a, Collateral projections of DG- and CA2-projecting SuM neurons. Dual-colour Cre-dependent retrograde AAVs were used to label the cell bodies and axons of SuM neurons projecting to the DG (mCherry, red) and CA2 (eYFP, green) (see also Fig. 2d, e). Weak collateral projections from DG- and CA2-projecting SuM neurons could be seen in the medial (MS) and lateral septal nucleus (LS) and lateral hypothalamic area (LHA). One example of 6 replicates is shown. Scale bars, 500 μm. b, Dual-colour CTB labelling was used to trace the cell bodies of SuM neurons projecting to DG (CTB 488, green) and CA2 (CTB 594, red). The majority of the identified SuM cell bodies were found to project solely either to the DG or CA2 with only 2.6 ± 0.5% (n = 3 mice, n = 708 neurons) projecting to both. Scale bars, 250 μm. c, d, Strategy for multiplex fluorescent in situ hybridization (RNAscope) to identify Vglut2 and Vgat mRNAs present in the cell bodies of DG- and CA2-projecting SuM neurons. Cre-dependent retrograde AAV was used to label the cell bodies and axons of SuM neurons projecting to the DG (c) and CA2 (d). e, f, DG-projecting (e) and CA2-projecting (f) SuM neurons were stained by antisense probes directed against Vglut2 (570 nm), Vgat (690 nm) and eYFP (520 nm) mRNAs. Scale bars, 500 μm. g, h, Quantification of eYFP+Vglut2+Vgat+ and eYFP+Vglut2+Vgat neurons in the DG-projecting (g) and CA2-projecting (h) SuM neurons. n = 2 mice for each group. All the DG- and CA2-projecting neurons are Vglut2+. The majority (average: 71.3%, mouse 1: 72.2%, mouse 2: 70.3%) of DG-projecting neurons are Vgat+, whereas only a small portion (average: 13.1%, mouse 1: 12.3%, mouse 2: 13.8%) of CA2-projecting neurons are Vgat+.

Extended Data Fig. 6 Validation of the functional segregation of SuM–DG and SuM–CA2 circuits.

a, b, Strategies of whole-cell voltage clamp recording experiments to validate the physiological segregation of SuM–DG and SuM–CA2 circuits. Retrograde AAV-DIO-ChR2-eYFP was bilaterally injected into the DG (a) or CA2 (b) of SuM-Cre mice to infect DG-projecting (a) or CA2-projecting (b) SuM cells. Two weeks later, acute transverse hippocampal slices were prepared and whole-cell voltage clamp recordings were performed in CA2 pyramidal neurons (PNs, a) or DG granule cells (GCs, b) to determine whether the DG-projecting SuM neurons also innervated CA2 cells (a), and vice versa (b). All recorded cells were filled with biocytin for post-hoc identification. Bottom panels are confocal images of transverse hippocampal slices recorded with Nissl (blue), ChR2–eYFP expressing SuM fibres (green), and biocytin-filled (red) CA2 PNs (a) or DG GCs (b). Scale bars, 250 μm. c, d, Example light-evoked IPSCs (10 traces, light red; average trace, red) and light-evoked EPSCs (10 traces, grey; average trace, black) recorded with 5 light pulses in CA2 PNs (c) and DG GCs (d). No inhibitory or excitatory light-evoked current was detected for any of the 17 CA2 PNs from 6 mice and 19 DG GCs from 7 mice, suggesting that DG-projecting SuM cells do not innervate CA2, and vice versa. e, f, Strategies of behavioural experiments to examine the effect from light bleed-through on the optogenetic dissection of SuM–DG and SuM–CA2 circuits. RetroAAV-DIO-ChR2 was bilaterally injected into the DG (e) or CA2 (f) of SuM-Cre mice to infect DG-projecting (e) or CA2-projecting (f) SuM cells. Optic fibres were bilaterally implanted in the uninjected region of each mouse: DG–ChR2 with fibres in CA2 (e) or CA2–ChR2 with fibres in DG (f). gj, Contextual (g, h) and social (i, j) memory tests. Light was delivered to the uninfected region of the hippocampus: DG–ChR2 with light delivery to CA2 (g, i) or CA2–ChR2 with light delivery to DG (h, j). Two-way ANOVA with Bonferroni post-hoc test, n = 8 mice (DG–ChR2) versus n = 8 (DG–eYFP) in g, n = 8 mice (CA2–ChR2) versus n = 8 (CA2–eYFP) in h. Two-tailed unpaired t-tests, n = 8 mice (DG–ChR2) versus n = 8 (DG–eYFP) in i, n = 8 mice (CA2–ChR2) versus n = 8 (CA2–eYFP) in j. N.s., not significant. Error bars show mean ± s.e.m. Boxes denote median (centre line), 25th and 75th percentiles (box edges) and full range of values (whiskers).

Extended Data Fig. 7 Circuit-specific FOS mapping and optogenetic identification and recording.

a, Experimental design to examine the projection specificity of SuM populations that respond to novel contextual or novel social exposure. b, d, Strategy to label DG-projecting (b) or CA2-projecting (d) SuM cells using retroAAV-DIO-eYFP with the SuM-Cre mice. c, e, Images of overlapping populations in the SuM that had contextual or social novelty-induced FOS expression and projected to the DG or CA2. On the right of each image is a magnified view of the area within the white box (200 × 200 μm). One example of 3 replicates is shown. f, eYFP+/FOS+ overlap SuM cell counts for the conditions shown in c and e. DG-projecting SuM neurons were preferentially activated by contextual novelty (DG-projecting: 53.3 ± 6.1% versus CA2-projecting: 37.1 ± 0.4%, *P = 0.0429, one-way ANOVA with Tukey’s post-hoc test, n = 3 mice for both groups), whereas CA2-projecting neurons responded more to social novelty (CA2-projecting: 50.2 ± 1.8% versus DG-projecting: 25.4 ± 2.7%, **P = 0.0042, one-way ANOVA with Tukey’s post-hoc test, n = 3 mice for both groups). All error bars show mean ± s.e.m. Open circles are values from individual mice. Scale bars, 200 μm. g, j, Strategies for optogenetic identification and recording DG-projecting (g) and CA2-projecting (j) neurons in the SuM. h, k, Example coronal sections of the SuM with tetrode locations indicated by arrows. Scale bars, 200 μm. i, l, Histology of recorded mice for circuit-specific optogenetic identification and recording. Reconstructed tetrode tip locations overlaid onto schematic mouse brain slices are shown. Numbers indicate posterior distances from bregma. Thirty-five tetrodes were in target and 8 out of target from 8 mice for recording DG-projecting neurons, and 49 tetrodes were in target and 9 out of target from 11 mice for recording CA2-projecting neurons.

Extended Data Fig. 8 Behavioural tests for contextual memory.

a, Familiar (A1–A4 in b) and novel contexts (B1 and B2 in b) used for contextual memory tests. b, Behavioural protocol. ce, Example travelling trajectories of mice during each contextual exposure described in b under different optogenetic manipulations of the SuM–DG circuit. fi, Percentage of distances travelled during each contextual exposure under various optogenetic manipulations. Light was delivered in sessions A4 and B1 to the DG or CA2. The no-light data were used as the control, supplementary to the eYFP controls in Fig. 4b–e. Two-way ANOVA with Bonferroni post-hoc test, n = 11 mice (DG–ChR2–light) versus n = 11 (DG–ChR2–no light) in f (**P = 0.0013, F (20, 100) = 2.882), n = 11 (DG–eNpHR–light) versus n = 11 (DG–eNpHR–no light) in g (****P < 0.0001, F (20, 100) = 2.402), n = 11 (CA2–ChR2–light) versus n = 11 (CA2–ChR2–no light) in h, n = 10 (CA2–eNpHR–light) versus n = 10 (CA2–eNpHR–no light) in i. All error bars show mean ± s.e.m.

Extended Data Fig. 9 Synaptic transmission from SuM terminals to DG granule cells and CA2 pyramidal neurons.

a, Identification of direct glutamatergic transmission in CA2 from SuM inputs. Left, diagram illustrating the whole-cell recordings of hippocampal pyramidal neurons (PNs) and SuM fibre stimulation in acute slice preparation. Caesium-based intracellular solution was used in the recording pipette and all cells were recorded in voltage-clamp mode with a membrane potential held at -70 mV to record excitatory glutamatergic transmission. Right, diagram mapping the location of all PNs recorded in the hippocampus with closed circles indicating connected cells and open circles indicating unconnected cells with no detected transmission. Cells were classified as CA1 (n = 11, all unconnected), CA2 (n = 46 connected, n = 30 unconnected) and CA3 (n = 19 connected, n = 29 unconnected) based on morphology (the presence of thorny excrecsences for CA3) and location. Circle colour indicates cell type with CA2 PNs black, CA3 PNs brown and CA1 PNs orange. b, Identification of feed-forward inhibitory transmission in CA2 from SuM inputs. Left, diagram illustrating the local circuitry involved in the feed-forward inhibition recruited by SuM inputs in the hippocampus. The same recording configuration was used as shown in a, except that cells were held at +10 mV to measure feed-forward inhibitory transmission. Right, diagram mapping PNs in the hippocampus that receive feed-forward transmission from light-evoked SuM inputs with unconnected cells shown as open circles and connected cells as filled circles. No CA1 PNs received feed-forward inhibition (n = 11), whereas over half of CA2 PNs received detectable feed-forward inhibition (n = 30 connected, n = 18 unconnected) as well as a fraction of CA3 PNs (n = 20 connected, n = 18 unconnected). c, Identification of excitatory and inhibitory transmission to DG granule cells (GCs) from SuM inputs. Left, diagram illustrating the experimental setup to measure local excitatory and inhibitory circuitry recruited by SuM inputs in the DG. The same recording configuration was used as shown in a and b. Cells were held at -70 mV to measure excitatory transmission and then at +10 mV. Inhibitory transmission was measured before and after the addition of 10 μM NBQX and 50 μM APV to distinguish feed-forward and direct inhibitory transmission. Right, diagram mapping GCs in the DG that receive excitatory and inhibitory (both feed-forward and direct) transmission from light-evoked SuM inputs. Blue filled circles mark the location of cells that received both excitatory and inhibitory transmission (n = 21). All cells received both direct and feed-forward inhibition. Yellow-filled circles mark cells that did not receive detectable excitatory transmission, but both direct and feed-forward inhibitory transmission (n = 3). Data for ac were collected from 156 slices from 92 mice. Abbreviations: so, stratum oriens; sp., stratum pyramidale; sr, stratum radiatum; mf, mossy fibre; ml, molecular layer; gcl, granule cell layer. d, Inverted epifluorescent image of a hippocampal slice with a DG GC and a CA2 PN. Each neuron had been recorded and filled with biocytin which was stained with streptavidin-conjugated fluorophore. e, Dendritic (blue) and axonal (red) reconstruction of the cells in d with the hippocampal regions demarcated. f, Diagram illustrating the local circuitry and whole-cell recording configuration of DG GCs in acute brain slices prepared from SuM-Cre mice injected with AAV-DIO-ChR2-eYFP. SuM axonal terminals were illuminated by a 0.5-ms 488-nm light stimulation. IN, interneuron. g, Sample traces of light-evoked EPSCs recorded at −70 mV (individual traces in grey, average trace in black) and IPSCs recorded at +10 mV (individual traces in light red, average trace in red) in a same DG GC under voltage clamp. The blue line denotes when the light stimulus was applied. h, Normalized cumulative distribution of latencies for DG GC EPSCs (black) and IPSCs (red), both displaying response latencies consistent with direct monosynaptic transmission (4.1 ± 0.3 ms for EPSCs, n = 20 GCs; 4.16 ± 0.3 ms for IPSCs, n = 22 GCs; Student’s t-test, P = 0.94; Kolmogorov–Smirnov test, P = 0.97). i, j, Sample traces (i) and time course of amplitudes (j, IPSCs only, n = 22 GCs from 8 mice) of light-evoked EPSCs (black) and IPSCs (red) recorded in DG GCs before and after application of 10 μM NBQX and 50 μM APV (grey), and further application of 1 μM SR95531 and 2 μM CGP55845A (green). AMPA and NMDA receptor blockers completely blocked EPSCs (13 ± 3.1 pA, n = 20 GCs), but only partially blocked IPSCs (by 36%, 86 ± 21 pA before, 31 ± 7.2 pA following NBQX and APV addition, **P = 0.0037, Wilcoxon signed-rank test, n = 22 GCs from 8 mice), indicating that there is feed-forward inhibition recruited by SuM inputs, accompanied by a larger amount of direct inhibitory transmission. The remaining light-evoked IPSCs were entirely blocked by the subsequent addition of GABAA and GABAB receptor blockers, suggesting that DG-projecting neurons in the SuM are capable of simultaneously releasing both glutamate and GABA. k, Diagram similar to f, showing the local circuitry and whole-cell recording configuration of CA2 PNs. l, Sample traces of EPSCs recorded at −70 mV (individual traces in grey, average trace in black) and IPSCs recorded at +10 mV (individual traces in light red, average trace in red) in the same CA2 PN under voltage clamp. Note the increased latency of IPSCs onset compared to EPSCs. m, Normalized cumulative distribution of latencies for CA2 PN EPSCs (black) and IPSCs (red). IPSCs displayed significantly different response latencies from the EPSCs, with a longer IPSC response latency consistent with bi-synaptic feed-forward inhibition (2.9 ± 0.1 ms for EPSCs, n = 166 PNs; 6.2 ± 0.4 ms for IPSCs, n = 69 PNs; ****P < 0.0001, Mann–Whitney U test; Kolmogorov–Smirnov test, ****P < 0.0001, data from 92 mice). n, o, Sample traces (n) and time course of amplitudes (o, IPSCs only, n = 7 from 4 mice) of light-evoked EPSCs (black) and IPSCs (red) recorded in CA2 PNs before and after application of 10 μM NBQX and 50 μM APV (grey). Both the EPSCs (16 ± 4.8 pA, n = 6 PNs from 4 mice) and IPSCs (167 ± 40 pA, n = 7 PNs from 4 mice) were completely blocked by the application of AMPA and NMDA receptor blockers, indicating that the synaptic transmission from the SuM is entirely glutamatergic in CA2. Notably, both SuM–DG and SuM–CA2 transmissions recruit a robust feed-forward inhibition. p, Confirmation of SuM–CA2 transmission by the addition of TTX and 4-aminopyridine (4-AP) to isolate transmitter release resulting from direct optical terminal depolarization. Application of TTX and 4-AP abolishes IPSCs and spares EPSCs in CA2 pyramidal neurons (PNs), consistent with mono-synaptic excitation and di-synaptic inhibition. Control sample traces are shown in black (EPSC) and red (IPSC), and traces following application of TTX and 4-AP are shown in grey. q, Time course of light-evoked EPSC and IPSC amplitudes after application of 0.2 μM TTX and 100 μM 4-AP in which a longer light stimulus was used to directly depolarize the SuM axonal terminal. Initial amplitudes were 20.2 ± 6.1 pA for EPSCs and 110 ± 50 pA for IPSCs, and 9.3 ± 7.7 pA for EPSCs and 4.8 ± 0.7 pA for IPSCs following TTX and 4-AP, indicating a 61 ± 24% block of EPSCs and 88 ± 4.2% block of IPSCs by TTX and 4-AP. P = 0.22 for EPSCs, P = 0.016 for IPSCs, Wilcoxon signed-rank test, n = 7 PNs from 3 mice. All error bars show mean ± s.e.m.

Extended Data Fig. 10 Retrograde tracing of upstream inputs to the DG-projecting and CA2-projecting SuM neurons.

a, b, Strategies of projection-specific retrograde tracing to map the upstream inputs to the DG-projecting (a) and CA2-projecting (b) neurons. Retrograde AAV helper viruses expressing the rabies G protein and the TVA receptor are injected into DG (a) or into CA2 (b), and three weeks later EnvA pseudotyped rabies virus expressing mCherry is injected into the SuM. This allows for the retrograde labelling of upstream neurons that send efferent axons to DG- or CA2-projecting SuM cells by the rabies virus. c, d, Coronal sections showing DG-projecting (c) and CA2-projecting (d) starter cells in the SuM. Cells in red (mCherry) are TVA+ cells that express rabies virus (one example of four mice for DG injections, one example of five mice for CA2 injections). Cells in green (GFP) express oG (optimized glycoprotein). The mCherry+/GFP+ overlapping cells (yellow) are starter cells in which the trans-complementation of RVdG with oG results in the production of oG+RVdG that spreads trans-synaptically to input cells. Scale bar, 1 mm. e, f, Coronal sections with trans-synaptically labelled input cells upstream to the DG-projecting (e) and CA2-projecting (f) SuM neurons. Scale bars, 1 mm. g, h, Quantification of inputs from various brain regions to the DG-projecting (g, n = 4 mice) and CA2-projecting (h, n = 5 mice) SuM neurons. Both populations received extensive inputs from subcortical regions including the hypothalamus, brainstem, septum and nucleus accumbens. However, the inputs to the DG-projecting population were comparatively biased to brain regions in the reward and motor systems, such as the VTA, SI, AcbSh, LS and MS, whereas the CA2 projectors received proportionally greater inputs from neurons in socially engaged regions, particularly the PVH (*P = 0.0159, Mann–Whitney U test) and MPO. All error bars show mean ± s.e.m. LS, lateral septal nucleus; MS, medial septal nucleus; SI, substantia innominata; AcbSh, accumbens nucleus, shell; ZI, zona incerta; LHA, lateral hypothalamic area; MPO, medial preoptic area; LPO, lateral preoptic area; PVH, paraventricular hypothalamic nucleus; PH, posterior hypothalamic nucleus; PAG, periaqueductal grey; MRN, midbrain reticular nucleus; VTA, ventral tegmental area; Raphe: DR, dorsal raphe nucleus and MnR, median raphe nucleus.

Supplementary information

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chen, S., He, L., Huang, A.J.Y. et al. A hypothalamic novelty signal modulates hippocampal memory. Nature 586, 270–274 (2020).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


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.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing