Skip to main content

Thank you for visiting nature.com. 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 neural network for intermale aggression to establish social hierarchy

Nature Neuroscience volume 21, pages 834–842 (2018)Cite this article

Subjects

Abstract

Intermale aggression is used to establish social rank. Several neuronal populations have been implicated in aggression, but the circuit mechanisms that shape this innate behavior and coordinate its different components (including attack execution and reward) remain elusive. We show that dopamine transporter-expressing neurons in the hypothalamic ventral premammillary nucleus (PMvDAT neurons) organize goal-oriented aggression in male mice. Activation of PMvDAT neurons triggers attack behavior; silencing these neurons interrupts attacks. Regenerative PMvDAT membrane conductances interacting with recurrent and reciprocal excitation explain how a brief trigger can elicit a long-lasting response (hysteresis). PMvDAT projections to the ventrolateral part of the ventromedial hypothalamic and the supramammillary nuclei control attack execution and aggression reward, respectively. Brief manipulation of PMvDAT activity switched the dominance relationship between males, an effect persisting for weeks. These results identify a network structure anchored in PMvDAT neurons that organizes aggressive behavior and, as a consequence, determines intermale hierarchy.

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

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

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

Fig. 1: Aggression activates PMvDAT neurons, and their electrical activity correlates with aggression phenotype.
Fig. 2: PMvDAT neuron activity is necessary and sufficient for aggression.
Fig. 3: Membrane properties and wiring of PMvDAT neurons provide a substrate for persistent activity.
Fig. 4: Monosynaptic PMvDAT projections in VMHvl and SuM drive aggression intensity and aggression reward.
Fig. 5: Brief manipulation of PMvDAT neuron activity causes a long-lasting switch of intermale hierarchy.

Similar content being viewed by others

References

  1. Moyer, K. E. Kinds of aggression and their physiological basis. Commun. Behav. Biol. 2, 65–87 (1968).

    Google Scholar 

  2. Yang, T. & Shah, N. M. Molecular and neural control of sexually dimorphic social behaviors. Curr. Opin. Neurobiol. 38, 89–95 (2016).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  3. Fernandez-Espejo, E. & Mir, D. Ethological analysis of the male rat’s socioagonistic behavior in a resident-intruder paradigm. Aggress. Behav. 16, 41–55 (1990).

    Article  Google Scholar 

  4. Natarajan, D., de Vries, H., Saaltink, D. J., de Boer, S. F. & Koolhaas, J. M. Delineation of violence from functional aggression in mice: an ethological approach. Behav. Genet. 39, 73–90 (2009).

    Article  PubMed  Google Scholar 

  5. Hess, W. R. & Brügger, M. Das subkortikale Zentrum der affektiven Abwehrreaktion. (The subcortical center for affective defense reactions.). Helv. Physiol. Pharmacol. Acta 1, 33–52 (1943).

    Google Scholar 

  6. Kruk, M. R. et al. Discriminant analysis of the localization of aggression-inducing electrode placements in the hypothalamus of male rats. Brain Res. 260, 61–79 (1983).

    Article  PubMed  CAS  Google Scholar 

  7. Lammers, J. H., Kruk, M. R., Meelis, W. & van der Poel, A. M. Hypothalamic substrates for brain stimulation-induced attack, teeth-chattering and social grooming in the rat. Brain Res. 449, 311–327 (1988).

    Article  PubMed  CAS  Google Scholar 

  8. Olivier, B. & Wiepkema, P. R. Behaviour changes in mice following electrolytic lesions in the median hypothalamus. Brain Res. 65, 521–524 (1974).

    Article  PubMed  CAS  Google Scholar 

  9. Levinson, D. M., Reeves, D. L. & Buchanan, D. R. Reductions in aggression and dominance status in guinea pigs following bilateral lesions in the basolateral amygdala or lateral septum. Physiol. Behav. 25, 963–971 (1980).

    Article  PubMed  CAS  Google Scholar 

  10. Golden, S. A. et al. Basal forebrain projections to the lateral habenula modulate aggression reward. Nature 534, 688–692 (2016).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  11. Unger, E. K. et al. Medial amygdalar aromatase neurons regulate aggression in both sexes. Cell Rep. 10, 453–462 (2015).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  12. Ulrich, R. E. & Craine, W. H. Behavior: persistence of shock-induced aggression. Science 143, 971–973 (1964).

    Article  PubMed  CAS  Google Scholar 

  13. Lee, H. et al. Scalable control of mounting and attack by Esr1+ neurons in the ventromedial hypothalamus. Nature 509, 627–632 (2014).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  14. Donato, J. Jr. et al. Leptin’s effect on puberty in mice is relayed by the ventral premammillary nucleus and does not require signaling in Kiss1 neurons. J. Clin. Invest. 121, 355–368 (2011).

    Article  PubMed  Google Scholar 

  15. Motta, S. C. et al. Ventral premammillary nucleus as a critical sensory relay to the maternal aggression network. Proc. Natl. Acad. Sci. USA 110, 14438–14443 (2013).

    Article  PubMed  PubMed Central  Google Scholar 

  16. Lin, D. et al. Functional identification of an aggression locus in the mouse hypothalamus. Nature 470, 221–226 (2011).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  17. Kollack-Walker, S. & Newman, S. W. Mating and agonistic behavior produce different patterns of Fos immunolabeling in the male Syrian hamster brain. Neuroscience 66, 721–736 (1995).

    Article  PubMed  CAS  Google Scholar 

  18. Soden, M. E. et al. Genetic isolation of hypothalamic neurons that regulate context-specific male social behavior. Cell Rep. 16, 304–313 (2016).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  19. Van den Berg, M. J., Ter Horst, G. J. & Koolhaas, J. M. The Nucleus premammillaris ventralis (Pmv) and aggressive-behavior in the rat. Aggress. Behav. 9, 41–47 (1983).

    Article  Google Scholar 

  20. Canteras, N. S., Simerly, R. B. & Swanson, L. W. Projections of the ventral premammillary nucleus. J. Comp. Neurol. 324, 195–212 (1992).

    Article  PubMed  CAS  Google Scholar 

  21. Cavalcante, J. C., Bittencourt, J. C. & Elias, C. F. Distribution of the neuronal inputs to the ventral premammillary nucleus of male and female rats. Brain Res. 1582, 77–90 (2014).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  22. Ziegler, D. R., Cullinan, W. E. & Herman, J. P. Distribution of vesicular glutamate transporter mRNA in rat hypothalamus. J. Comp. Neurol. 448, 217–229 (2002).

    Article  PubMed  CAS  Google Scholar 

  23. Meister, B. & Elde, R. Dopamine transporter mRNA in neurons of the rat hypothalamus. Neuroendocrinology 58, 388–395 (1993).

    Article  PubMed  CAS  Google Scholar 

  24. Blanchard, D. C. & Blanchard, R. J. Ethoexperimental approaches to the biology of emotion. Annu. Rev. Psychol. 39, 43–68 (1988).

    Article  PubMed  CAS  Google Scholar 

  25. Curran, T. & Morgan, J. I. Fos: an immediate-early transcription factor in neurons. J. Neurobiol. 26, 403–412 (1995).

    Google Scholar 

  26. Polston, E. K. & Erskine, M. S. Patterns of induction of the immediate-early genes c-fos and egr-1 in the female rat brain following differential amounts of mating stimulation. Neuroendocrinology 62, 370–384 (1995).

    Article  PubMed  CAS  Google Scholar 

  27. Barbosa, F. et al. Hysteresis effect in the processing of facial expressions of emotion and its neurophysiological correlates. Psychophysiology 50, S132 (2013).

    Google Scholar 

  28. McCormick, D. A. & Pape, H. C. Properties of a hyperpolarization-activated cation current and its role in rhythmic oscillation in thalamic relay neurones. J. Physiol. (Lond.) 431, 291–318 (1990).

    Article  CAS  Google Scholar 

  29. Falkner, A. L., Dollar, P., Perona, P., Anderson, D. J. & Lin, D. Decoding ventromedial hypothalamic neural activity during male mouse aggression. J. Neurosci. 34, 5971–5984 (2014).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  30. Yang, T. et al. Social control of hypothalamus-mediated male aggression. Neuron 95, 955–970.e4 (2017).

    Article  PubMed  CAS  PubMed Central  Google Scholar 

  31. Tong, Q. et al. Synaptic glutamate release by ventromedial hypothalamic neurons is part of the neurocircuitry that prevents hypoglycemia. Cell Metab. 5, 383–393 (2007).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  32. Fish, E. W., DeBold, J. F. & Miczek, K. A. Escalated aggression as a reward: corticosterone and GABA(A) receptor positive modulators in mice. Psychopharmacology (Berl.) 182, 116–127 (2005).

    Article  CAS  Google Scholar 

  33. Burgdorf, J. et al. Ultrasonic vocalizations of rats (Rattus norvegicus) during mating, play, and aggression: behavioral concomitants, relationship to reward, and self-administration of playback. J. Comp. Psychol. 122, 357–367 (2008).

    Article  PubMed  Google Scholar 

  34. Ikemoto, S., Witkin, B. M., Zangen, A. & Wise, R. A. Rewarding effects of AMPA administration into the supramammillary or posterior hypothalamic nuclei but not the ventral tegmental area. J. Neurosci. 24, 5758–5765 (2004).

    Article  PubMed  CAS  Google Scholar 

  35. Ikemoto, S. & Bonci, A. Neurocircuitry of drug reward. Neuropharmacology 76 Pt B, 329–341 (2014).

    Article  PubMed  CAS  Google Scholar 

  36. Spiteri, T. et al. The role of the estrogen receptor alpha in the medial amygdala and ventromedial nucleus of the hypothalamus in social recognition, anxiety and aggression. Behav. Brain Res. 210, 211–220 (2010).

    Article  PubMed  CAS  Google Scholar 

  37. Desjardins, C., Maruniak, J. A. & Bronson, F. H. Social rank in house mice: differentiation revealed by ultraviolet visualization of urinary marking patterns. Science 182, 939–941 (1973).

    Article  PubMed  CAS  Google Scholar 

  38. Wong, M. & Balshine, S. Fight for your breeding right: hierarchy re-establishment predicts aggression in a social queue. Biol. Lett. 7, 190–193 (2011).

    Article  PubMed  Google Scholar 

  39. Bernstein, I. S., Rose, R. M., Gordon, T. P. & Grady, C. L. Agonistic rank, aggression, social-context, and testosterone in male pigtail monkeys. Aggress. Behav. 5, 329–339 (1979).

    Article  CAS  Google Scholar 

  40. Young, C., Majolo, B., Schulke, O. & Ostner, J. Male social bonds and rank predict supporter selection in cooperative aggression in wild Barbary macaques. Anim. Behav. 95, 23–32 (2014).

    Article  Google Scholar 

  41. van den Berg, W. E., Lamballais, S. & Kushner, S. A. Sex-specific mechanism of social hierarchy in mice. Neuropsychopharmacology 40, 1364–1372 (2015).

    Article  PubMed  Google Scholar 

  42. Lindzey, G., Winston, H. & Manosevitz, M. Social dominance in inbred mouse strains. Nature 191, 474–476 (1961).

    Article  PubMed  CAS  Google Scholar 

  43. Zhou, T. et al. History of winning remodels thalamo-PFC circuit to reinforce social dominance. Science 357, 162–168 (2017).

    Article  PubMed  CAS  Google Scholar 

  44. Südhof, T. C. Reproducibility: Experimental mismatch in neural circuits. Nature 528, 338–339 (2015).

    Article  PubMed  CAS  Google Scholar 

  45. Hoebel, B. G. The neural and chemical basis of reward - new discoveries and theories in brain control of feeding, mating, aggression, self-stimulation and self-injection. J. Soc. Biol. Struct. 5, 397–408 (1982).

    Article  Google Scholar 

  46. Gil, M., Nguyen, N. T., McDonald, M. & Albers, H. E. Social reward: interactions with social status, social communication, aggression, and associated neural activation in the ventral tegmental area. Eur. J. Neurosci. 38, 2308–2318 (2013).

    Article  PubMed  Google Scholar 

  47. Anderson, P. K. Density, social structure, and nonsocial environment in house-mouse populations and the implications for regulation of numbers. Trans. N. Y. Acad. Sci. 23, 447–451 (1961).

    Article  CAS  Google Scholar 

  48. Singleton, G. R. The social and genetic-structure of a natural colony of house mice, Mus musculus, at Healesville Wildlife Sanctuary. Aust. J. Zool. 31, 155–166 (1983).

    Article  Google Scholar 

  49. Chambers, L. K., Singleton, G. R. & Krebs, C. J. Movements and social organization of wild house mice (Mus domesticus) in the wheatlands of northwestern Victoria, Australia. J. Mamm. 81, 59–69 (2000).

    Article  Google Scholar 

  50. Ekstrand, M. I. et al. Progressive Parkinsonism in mice with respiratory-chain-deficient dopamine neurons. Proc. Natl. Acad. Sci. USA 104, 1325–1330 (2007).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  51. Russo, S. J. et al. Nuclear factor kappa B signaling regulates neuronal morphology and cocaine reward. J. Neurosci. 29, 3529–3537 (2009).

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  52. Petersson, U. et al. Nucleobindin is produced by bone cells and secreted into the osteoid, with a potential role as a modulator of matrix maturation. Bone 34, 949–960 (2004).

    Article  PubMed  CAS  Google Scholar 

  53. Tulke, S. et al. Nucleobindin 1 (NUCB1) is a Golgi-resident marker of neurons. Neuroscience 314, 179–188 (2016).

    Article  PubMed  CAS  Google Scholar 

  54. Franklin, K.B.J. & Paxinos, G. The Mouse Brain in Stereotactic Coordinates, 3rd ed. (Academic Press, Cambridge, MA, 2008).

Download references

Acknowledgements

The authors thank I. Lazaridis for advice on optogenetics; C. Bellardita and V. Caggiano for advice on hardware, software coding, and experiments; and K. Ampatzis for advice on illustrations. We thank members of the Broberger laboratory for discussion during the preparation of this manuscript; N.G. Larsson and O. Kiehn for sharing the DAT-Cre and loxP-flanked tdTomato mice, respectively; and K. Fuxe for sharing reagents. We thank M. Wendel (Dept. of Dental Medicine, Karolinska Institutet) for generously sharing primary NUCB-1 antiserum. The authors acknowledge I. Lazaridis, C. Bellardita, and A. El Manira for providing insightful suggestions on the manuscript. This study was supported by a European Research Council starting grant (261286), the Swedish Research Council (2014-3906), the Strategic Research Program in Diabetes at Karolinska Institutet, Hjärnfonden (the Swedish Brain Foundation), Novo Nordisk Fonden, and internal funds from Karolinska Institutet.

Author information

Authors and Affiliations

  1. Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden

    Stefanos Stagkourakis, Giada Spigolon, Paul Williams, Jil Protzmann, Gilberto Fisone & Christian Broberger

Authors
  1. Stefanos Stagkourakis
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Giada Spigolon
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Paul Williams
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Jil Protzmann
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Gilberto Fisone
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Christian Broberger
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

S.S., G.S., and P.W. designed, performed, and analyzed experiments. J.P. performed neuron reconstructions. G.F. and C.B. designed experiments. S.S. and C.B. wrote the manuscript. All authors reviewed the manuscript.

Corresponding authors

Correspondence to Stefanos Stagkourakis or Christian Broberger.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

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

Integrated Supplementary Information

Supplementary Figure 1 PMvDAT neuron numbers in the rostrocaudal axis, dopamine transporter (DAT) coexistence and c-Fos immunoreactivity post-RI test.

a, Confocal micrograph of PMv from a DAT-tdTomato mouse section immunostained for the pan-neuronal marker NUCB1 (n = 10 medial PMv sections from 10 mice). b, Schematic representation of the rostrocaudal extension of PMv. c, Total and DAT+ neuron numbers at the different rostrocaudal PMv levels (n=10 per group, one-way ANOVA with Tukey’s test, comparisons from left to right, P < 0.0001, P < 0.0001 and P = 0.0002). d, In situ hybridization in DAT-Cre-floxed-tdTomato PMv slices, in low magnification (right—n = 4 mice) and high magnification (left). Cre-driven tdTomato mRNA (red), DAT mRNA (green) and DAPI nuclear staining (blue). e, Quantification of tdTomato and DAT mRNA suggestive of near 100% coexistence in PMv DAT-tdTomato cells [quantification performed on tissue from 4 mice, with 3 samples of 40X confocal Z-stack PMv images per mouse (n = 12)]. f, Confocal micrograph of a DAT-tdTomato mouse PMv section immunostained for DAT (n = 5 mice) and (g), high magnification inset showing the DAT immunofluorescence in PMvDAT cells. h, Experimental design utilized in the RI paradigm for the respective c-fos studies. i-q, Representative c-fos immunoreactivity from control, AGGs and NONs post-RI test as quantified in Fig. 1h (i-k, control group of AGGs alone in cage n = 9, l-n, AGGs vs intruder n = 13, o-p, NONs vs intruder n = 7). r, Linear regression of attack duration and c-fos immunoreactivity in PMv neurons (n = 13, P = 0.5321, R2=0.03). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Data expressed as mean ± s.e.m. Image scale bars, (a, d (right - low magnification image), f, i-q) 100 μm and (d (left—high magnification images of single cells) and g) 10 μm.

Supplementary Figure 2 In vitro and in vivo validation of optogenetics in PMvDAT neurons.

a, Schematic drawing of PMv viral transduction with ChR2 in the DAT-Cre mouse. b, Light stimulation -evoked action potential (AP) firing. c, Quantification of spike fidelity in relation to photo-stimulation frequency and pulse duration (n = 10 neurons from 5 mice), and d, inward current induced by ChR2 stimulation (n = 9 neurons from 4 mice). In d, the shaded region represents the standard error. e, Schematic drawing of PMv viral transduction with eNpHR3 in the DAT-Cre mouse. f, Repeated eNpHR3-mediated hyperpolarizations of similar amplitude and g, eNpHR3-mediated outward current (n = 7 neurons from 4 mice). In g, the shaded region represents the standard error. h, Schematic drawing of PMv viral transduction with ChR2 and fiber implants in DAT-Cre AGGs and NONs. i, Photostimulation paradigm to test optogenetic activation of PMvDAT neurons. j-m, Virally transduced PMv expressing eYFP or ChR2, and c-fos immunoreactivity post-photostimulation (n = 4 bilateral PMv sections from 4 mice). n, Quantification of PMvDAT/c-fos colocalizing neurons (n = 4 PMv sections from 4 mice per group, two-tailed unpaired t-test, P < 0.0001). o, Quantification of PMvDAT/DIO-eYFP and PMvDAT/DIO-ChR2-eYFP colocalizing neurons (n = 6 PMv sections from 6 ChR2 being expressed in PMv, DAT-tdTomato mice and 11 PMv sections from 11 eYFP being expressed in PMv, DAT-tdTomato mice, two-tailed unpaired t-test, P = 0.2138). p, Representative confocal micrograph Z-stacks of PMv from DAT-tdTomato mouse brain sections immunostained for the virally induced eYFP or ChR2-eYFP expression (n = 6 bilateral PMv sections from 6 ChR2 being expressed in PMv, DAT-tdTomato mice, and 11 PMv sections from 11 eYFP being expressed in PMv, DAT-tdTomato mice). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. In (c) data are expressed as mean ± s.e.m. In box-and-whisker plots, centre lines indicate medians, box edges represent the interquartile range, and whiskers extend to the minimal and maximal values. Image scale bars, (j-m, p) 50 μm.

Supplementary Figure 3 ChR2 photostimulation of PMvDAT neurons in AGGs does not lead to attack against inanimate objects, such as an inflated glove or a plastic mouse.

a-c, Snapshots of video recordings with AGG residents presented with inanimate objects. d, No attack was elicited across animals and trials, pre- or during photostimulation (glove n = 10 trials with 10 ChR2 AGGs, black plastic mouse n = 8 trials with 8 ChR2 AGGs, and white plastic mouse n = 6 trials with 6 ChR2 AGGs).

Supplementary Figure 4 Targeted genetic ablation of PMvDAT neurons reduces intermale aggression.

a, Schematic drawing illustrating viral transduction of the diphtheria toxin A (dtA) subunit gene in PMvDAT neurons to accomplish cell ablation. b, c, Examples of DAT-tdTomato fluorescence from PMvDAT neurons in control (b, n = 9 bilateral PMv sections/mice) and dtA-transduced (c, n = 9 bilateral PMv sections/mice) mice. d, Bilateral quantification of PMvDAT neurons in control and dtA-ablated PMv nuclei (n = 9 medial PMv sections from 9 mice per group, two-tailed unpaired t-test, P < 0.0001). e-j, Bilateral quantification of neighbouring Cre-expressing DAT neuron populations. g, Bilateral quantification of SNDAT and VTADAT neurons in control and dtA-ablated PMv nuclei (substantia nigra (SN) comparison, n = 9 SN sections from 9 mice per group, two-tailed unpaired t-test, P = 0.4002, and ventral tegmental area (VTA) comparison, n = 9 VTA sections from 9 mice per group, two-tailed unpaired t-test, P = 0.4002). j, Bilateral quantification of ArcuateDAT neurons in control and dtA-ablated PMv nuclei (n = 9 Arcuate sections from 9 mice per group, two-tailed unpaired t-test, P = 0.1007). k-n, Quantification of aggression parameters in control and lesioned mice. k, Attack duration (n = 9 mice per group, two-tailed unpaired t-test, P = 0.0007). l, First attack latency (n = 9 mice per group, two-tailed unpaired t-test, P = 0.0133). m, Close investigation duration (n = 9 mice per group, two-tailed unpaired t-test, P = 0.0363). o-q, Quantification of mating parameters in control and lesioned resident male mice against females in estrus. o, Mounting duration (n = 9 mice per group, two-tailed unpaired t-test, P = 0.5933). p, Latency to first mounting (n = 9 mice per group, two-tailed unpaired t-test, P = 0.9055). q, Close investigation duration (n = 9 mice per group, two-tailed unpaired t-test, P = 0.7671). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. In box-and-whisker plots, centre lines indicate medians, box edges represent the interquartile range, and whiskers extend to the minimal and maximal values. Image scale bars, (b, c, e, f, h, i) 200 μm.

Supplementary Figure 5 DAT+→DAT+ and DAT+→DAT– PMv neuron connectivity.

a, Schematic illustration of the experimental design, transducing PMvDAT neurons with ChR2 to study DAT+ to DAT+ intra-PMv neuron synaptic connectivity. b, Average responses of a neuron clamped at −70 mV and photostimulation with a single 5 ms 447 nm light pulse, evoking an AP followed by an immediate compound depolarization (cEPSP, n = 5 cells). c, cEPSP amplitude on baseline (green), in Gabazine (blue), in TTX+4AP (red), and in CNQX+AP5 (black) (n = 5 cells per group from 3 mice, one-way ANOVA with Tukey’s test, Baseline vs Gabazine P = 0.1059, Gabazine vs TTX+4AP P = 0.9688, TTX+4AP vs CNQX+AP5 P = 0.0001). d, Time constant of the cEPSP on baseline (green), in Gabazine (blue), in TTX+4AP (red), and in CNQX+AP5 (black) (n = 5 cells per group from 3 mice, one-way ANOVA with Tukey’s test, Baseline vs Gabazine P = 0.9921, Gabazine vs TTX+4AP P = 0.0222, TTX+4AP vs CNQX+AP5 P < 0.0001). e, Voltage-clamp recording example of a DAT+ neuron photostimulated for 500 ms at 20 Hz. Sample of EPSCs (EPSCs) recorded immediately prior to (100 ms, e, f, green) and 10 s post-photostimulation (e, f, blue). g, EPSCs inter-event interval, pre- (green) and post- photostimulation (blue), (two-sided Kolmogorov-Smirnov test, p < 0.01). h, EPSC frequency pre- (green) and post- (blue) photostimulation (n = 6 cells per group from 3 mice, two-tailed unpaired t-test, P = 0.0005). i, EPSCs amplitude, pre- (green) and post- photostimulation (blue), (two-sided Kolmogorov-Smirnov test, p<0.05). j, EPSC amplitude pre- (green) and post- (blue) photostimulation (n = 6 cells per group from 3 mice, two-tailed unpaired t-test, P = 0.0304). k, Schematic illustration of the experimental design, transducing PMvDAT neurons with ChR2 to study DAT+ to DAT- intra-PMv neuron chemical connectivity (n = 3 mice). l, Quantification of DAT- neurons responsive to DAT+ neuron photostimulation (n = 6 cells, two per mouse). m, Averaged lEPSC amplitude evoked on baseline (green), in Gabazine (blue), TTX+4AP (red), in CNQX and AP5 (black), and in non-responsive neurons (grey) (n = 6 cells connected, n = 1 not-connected, recordings collected from 3 mice, one-way ANOVA with Tukey’s test, Baseline vs Gabazine P = 0.9981, Gabazine vs TTX+4AP P = 0.1603, TTX+4AP vs CNQX+AP5 P = 0.0050). The shaded region represents the standard error. n, Reverberation of activity evoked by optogenetic stimulation in a PMvDAT neuron, and quantification of its duration (o) and number of action potentials (APs, p) fired during that period. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. In bar graphs data are expressed as mean ± s.e.m. In box-and-whisker plots, centre lines indicate medians, box edges represent the interquartile range, and whiskers extend to the minimal and maximal values.

Supplementary Figure 6 PMvDAT neuron conductances, morphology, and contralateral projections.

a, PMvDAT neurons exhibit a hyperpolarization-activated depolarizing sag (sag) sensitive to ZD-7288 (n = 7 per group from 3 mice, two-tailed unpaired t-test, P = 0.0002), and a post-inhibitory rebound (PIR) sensitive to ZD-7288 and ML-218 (n = 7 per group from 3 mice, two-tailed unpaired t-test, P < 0.0001). b, Confocal micrograph shows an example of a neurobiotin-filled PMvDAT neuron (green) with extensive arborisation within the PMv. c, Confocal Z-stack micrograph indicating close appositions of the neuron in b (green), onto other PMvDAT cells (red). d-i, Six examples of reconstructed neurobiotin-filled PMvDAT neurons from coronal slices after in vitro recordings. j, Sholl analysis quantification of the number of PMvDAT neuron branches within and outside PMv (n = 12). k, Unilateral (right hemisphere) injection of AAV5-DIO-ChR2-eYFP in a DAT-tdTomato mouse and l, contralateral ChR2-eYFP terminals. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Data expressed as mean ± s.e.m. Image scale bars, (b-i, k, l) 50 μm.

Supplementary Figure 7 Anterograde and retrograde tracing of PMv input neurons to VMHvl and SuM.

a, Schematic of the experimental design, transducing PMvDAT neurons with ChR2 and identification of ChR2-eYFP terminals enriched in VMHvl and SuM. b, PMvDAT ChR2-eYFP expressing neurons in a DAT-tdTomato mouse (n = 11 mice). c, ChR2-eYFP terminals in VMHvl in the same brain as (b, n = 11 mice). d, Coronal brain section showing fiber projection from PMv and dense ChR2-eYFP expressing terminals in SuM (n = 11 mice). e, Schematic illustration of the experimental design, transducing PMvDAT neurons with ChR2, and evoking post-synaptic light responses in SuM neurons in in vitro recordings. f, Neurobiotin-stained SuM neuron surrounded by ChR2 terminals and responsive to light stimulation (left) and quantification of SuM neurons with lEPSCs (right). g, Averaged lEPSCs amplitude evoked on baseline (green), in Gabazine (blue), TTX and 4AP (red) and in CNQX and AP5 (black) (n = 5 connected, n = 1 not-connected, recordings collected from 3 mice, one-way ANOVA with Tukey’s test, Baseline vs Gabazine P = 0.2707, Gabazine vs TTX+4AP P = 0.0260, TTX+4AP vs CNQX+AP5 P = 0.0154). h-k, Green retrobead injections in VMHvl in the DAT-tdTomato mouse and identification of PMvDAT (tdTomato expressing) retrobead-containing neurons. h, Schematic of the experimental design (n = 8 mice). i, Injection site in VMHvl (n = 8 mice). j,k, PMv from the same tissue as in (i), containing retrobeads (n = 8 mice). l-o, Green retrobead injection in SuM in the DAT-tdTomato mouse and identification of PMvDAT (tdTomato expressing) retrobead-containing neurons. l, Schematic of the experimental design (n = 6 mice). m, Injection site in SuM (n = 6 mice). n,o, PMv from the same tissue as in (m), containing retrobeads (n = 6 mice). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. In box-and-whisker plots, centre lines indicate medians, box edges represent the interquartile range, and whiskers extend to the minimal and maximal values. All image scale bars are 50 μm.

Supplementary Figure 8 VMHvl and SuM neurons targeted by PMvDAT cells and their membrane properties.

a, Schematic of the experimental design, transducing PMvDAT neurons with ChR2 to study PMvDAT to VMHvl neuron connectivity. b, Example of a neurobiotin-filled photostimulation-responsive VMHvl neuron post-recording (n = 6 cells, two per mouse). c-f, ChR2-eYFP terminals in close apposition to the VMHvl neuron’s dendrite. g, Neurobiotin reconstruction of the neuron in (b), with numerous ChR2-eYFP terminals in close (<10 μm) apposition indicated in green. h, Estrogen receptor α (ERα) immunostaining in the Arcuate Nucleus and in VMHvl (red, n = 5 mice). i, High magnification (63x) inset from h, shows ChR2-eYFP PMvDAT terminals in proximity to VMHvl ERα positive neurons (n = 5 mice). j, Response of a VMHvl neuron to a single light pulse, on baseline (green), in Gabazine (blue), TTX+4AP (red) and in CNQX and AP5 (black). k, Resting membrane potential (RMP) and l, Input resistance (Rinput) of VMHvl neurons. m, Current-step pulse protocol and n, quantification of the number of VMHvl neurons with post-inhibitory rebound (PIR). o, Schematic of the experimental design, transducing PMvDAT neurons with ChR2 to study PMvDAT to SuM neuron connectivity. p, Response of a SuM neuron to a single light pulse, on baseline (green), in Gabazine (blue), TTX+4AP (red) and in CNQX and AP5 (black). q, RMP and r, Rinput of recorded SuM neurons. s, Current-step pulses protocol and t, quantification of the number of SuM neurons with inward rectification (Kir) and PIR. u, Quantification of number of APs per single 5 ms photostimulation pulse (n = 5 SuM cells and 9 VMHvl cells collected from 3 mice, two-tailed unpaired t-test, P = 0.7677) and v, instantaneous firing frequency of VMHvl and SuM neurons (n = 5 SuM cells and 9 VMHvl cells collected from 3 mice, two-tailed unpaired t-test, P < 0.0001). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. In bar graphs data are expressed as mean ± s.e.m. In box-and-whisker plots, centre lines indicate medians, box edges represent the interquartile range, and whiskers extend to the minimal and maximal values. Image scale bars, (b, g-i) 50 μm, and (c-f) 10 μm.

Supplementary Figure 9 PMvDAT neuron stimulation results in real-time place preference.

a, Schematic drawing of PMv viral transduction with ChR2 in AGG DAT-Cre mice. b, Schematic illustration of the behaviour protocol utilized in ChR2 stimulation experiments to test real-time place preference, NS (No Stimulation), LS (Light Stimulation). c, Example of real-time place-preference in a control (left heatmap) and a stimulation trial (right heatmap), and quantification (n = 15 per group, one-way ANOVA with Tukey’s test, eYFP comparison P = 0.9976, ChR2 comparison P = 0.0013). d, Schematic drawing of PMv viral transduction with eNpHR3 in AGG DAT-Cre mice. e, Schematic illustration of the behaviour protocol utilized in eNpHR3 PMv silencing experiments to test real-time place aversion. f, Samples of real-time place aversion in a control (left heatmap) and an eNpHR3 stimulation trial (right heatmap), and quantification (n = 10 per group, one-way ANOVA with Tukey’s test, eYFP comparison P = 0.9261, eNpHR3 comparison P < 0.0001). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Data expressed as mean ± s.e.m.

Supplementary Figure 10 PMvDAT neuron stimulation modulates cocaine-conditioned place preference (CPP).

a, Schematic of the experimental design and cocaine CPP procedure. b, Representative CPP heatmaps for PMvDAT neuron stimulation of eYFP and ChR2 transfected animals. c, Duration of stay in the cocaine paired chamber Duration of stay in the intruder paired chamber (n = 5 trials with 5 eYFP mice, n = 9 trials with 9 ChR2 mice, two-tailed unpaired t-test, eYFP in PMv AGGs with optic fibers in PMv, pretest vs CPP test day—no stimulation P = 0.0474, CPP test day no stimulation vs CPP test day—stimulation P = 0.4560, ChR2 in PMv AGGs with optic fibers in PMv, pretest vs CPP test day—no stimulation P = 0.0317, CPP test day no stimulation vs CPP test day—stimulation P = 0.0435); for colour-coding see h. d, Normalized CPP (n = 5 trials with 5 eYFP mice, n = 9 trials with 9 ChR2 mice, two-tailed unpaired t-test comparisons, eYFP in PMv AGGs with optic fibers in PMv CPP test day—no stimulation vs CPP test day—photostimulation P = 0.7053, ChR2 in PMv AGGs with optic fibers in PMv CPP test day—no stimulation vs CPP test day—photostimulation P = 0.0454). *P < 0.05. In box-and-whisker plots, centre lines indicate medians, box edges represent the interquartile range, and whiskers extend to the minimal and maximal values.

Supplementary Figure 11 PMvDAT neuron manipulation in anxiety-type assays.

a, Schematic of the experimental design for testing anxiety-type behaviours upon PMvDAT neuron activity modulation. b, Quantification of thigmotaxis (time spent in the centre) and related parameters in the open field test. Distance moved (n = 31 eYFP, n = 15 ChR2, n = 10 eNpHR3, n = 9 DTA mice, one-way ANOVA with Bonferroni’s test, eYFP vs ChR2 P = 0.9306, eYFP vs eNpHR3 P = 0.1372, eYFP vs DTA P = 0.3694). Time spent in centre (n = 31 eYFP, n = 15 ChR2, n = 10 eNpHR3, n = 9 DTA mice, one-way ANOVA with Bonferroni’s test, eYFP vs ChR2 P = 0.7518, eYFP vs eNpHR3 P = 0.9928, eYFP vs DTA P > 0.9999). Time spent in periphery (n = 31 eYFP, n = 15 ChR2, n = 10 eNpHR3, n = 9 DTA mice, one-way ANOVA with Bonferroni’s test, eYFP vs ChR2 P = 0.9608, eYFP vs eNpHR3 P = 0.1598, eYFP vs DTA P = 0.9732). c, Quantification of parameters in the elevated plus maze. Distance moved (n = 32 eYFP, n = 14 ChR2, n = 6 eNpHR3, n = 9 DTA mice, one-way ANOVA with Bonferroni’s test, eYFP vs ChR2 P = 0.4967, eYFP vs eNpHR3 P = 0.8539, eYFP vs DTA P = 0.9716). Time spent in open arms (n = 32 eYFP, n = 14 ChR2, n = 6 eNpHR3, n = 9 DTA mice, one-way ANOVA with Bonferroni’s test, eYFP vs ChR2 P > 0.9999, eYFP vs eNpHR3 P = 0.8859, eYFP vs DTA P = 0.1640). Time spent in closed arms (n = 32 eYFP, n = 14 ChR2, n = 6 eNpHR3, n = 9 DTA mice, one-way ANOVA with Bonferroni’s test, eYFP vs ChR2 P = 0.2690, eYFP vs eNpHR3 P = 0.2206, eYFP vs DTA P = 0.1699). In box-and-whisker plots, centre lines indicate medians, box edges represent the interquartile range, and whiskers extend to the minimal and maximal values.

Supplementary Figure 12 PMvDAT neuron glutamatergic terminal stimulation in VMHvl or SuM excites local target cells.

a-b, Schematic of the experimental design transducing PMvDAT neurons with ChR2 and placement of fiberoptic implants in VMHvl (a) or SuM (b). c, Schematic illustration of the experimental protocol utilized to validate PMvDAT neuron terminal excitation. d, Representative confocal micrograph Z-stack highlighting a confined increased c-fos expression in the photostimulated sites (VMHvl or SuM, n = 4 mice per group). e,f, Quantification of c-fos cell number across nuclei in photostimulated VMHvl animals (e, n = 4 bilateral PMv sections per group from 4 mice, one-way ANOVA with Tukey’s test, VMHvl vs PMv P = 0.0066, VMHvl vs SuM P = 0.0060) and photostimulated SuM animals (f, n = 4 bilateral PMv sections per group from 4 mice, one-way ANOVA with Tukey’s test, SuM vs VMHvl P = 0.0072, SuM vs PMv P = 0.0067)). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. In box-and-whisker plots, centre lines indicate medians, box edges represent the interquartile range, and whiskers extend to the minimal and maximal values. Acronyms: PMv; ventral premammillary nucleus, VMHvl; ventrolateral division of the ventromedial hypothalamus, VMHdm; dorsomedial division of the ventromedial hypothalamus, SuM; supramammillary nucleus. Image scale bars, (j-m, p) 100 μm.

Supplementary Figure 13 Experimental design and control parameters of the hierarchy corridor test (HCT).

a, Schematic of the behavioural test (HCT) used to test intermale hierarchy in male pairs. b, To assess robustness of the test, twelve pairs performed baseline repeated sessions where in 98.75 ± 1.25 % of cases the dominant mouse displaced the subordinate from the corridor back to its home chamber (n = 80 trials with 12 mouse pairs, P < 0.0001, two-tailed unpaired t-test). c, Dominant mice were heavier than subordinate mice (dominant mouse weight average 35.86 ± 1.03 g, subordinate mouse weight average 27.86 ± 0.46 g, n = 12 mouse pairs, P = 0.0002, two-tailed paired t-test). d, Prior to the HCT test, all subjects were habituated alone to the arena. Quantification of the cumulative duration spent in all three chambers (left chamber vs corridor vs right chamber) is indicative of no place preference. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. In bar graphs data are expressed as mean ± s.e.m. In box-and-whisker plots, centre lines indicate medians, box edges represent the interquartile range, and whiskers extend to the minimal and maximal values.

Supplementary Figure 14 Monochrome snapshots during HCT at start and near end of duels.

a, Dominant mouse expressing eNpHR3 and subordinate ChR2 in PMvDAT neurons pre- photostimulation (n = 35 as quintuplicates from 7 rhodopsin mouse pairs). b, During photostimulation hierarchy reverses. c, Post- photostimulation hierarchy remains reversed. d-f, Hierarchy remains unaffected by photostimulation in eYFP-expressing control animals (n = 25 as quintuplicates from 5 eYFP mouse pairs).

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–14

Reporting Summary

Supplementary Video 1 - Photostimulation of eYFP expressing PMvDAT neurons in a male DAT-Cre mouse has no effect in intermale social behaviour.

The video shows an AGG male DAT-Cre mouse (C57BL/6J) injected in PMv with DIO-eYFP, interacting with a castrated male intruder (BALB/c). No changes in behaviour are observed upon photostimulation of the resident. Video at 3X.

Supplementary Video 2 - Photostimulation of ChR2-eYFP expressing PMvDAT neurons in a male DAT-Cre mouse elicits attacks against the intruder.

The video shows an AGG male DAT-Cre mouse (C57BL/6J) injected in PMv with DIO-ChR2-eYFP, interacting with a castrated male intruder (BALB/c). Upon photostimulation at 20 Hz with 5 msec pulse duration, the resident initiates several attack bouts. Video at 3X.

Supplementary Video 3 - Photostimulation of eNpHR3-eYFP expressing PMvDAT neurons in a male DAT-Cre mouse stops attacks against the intruder.

The video shows an AGG male DAT-Cre mouse (C57BL/6J) injected in PMv with DIO-eNpHR3-eYFP, interacting with a male intruder (C57BL/6J). Initially the resident attacks the intruder, and the attack bout ends upon photostimulation. Escalation from close investigation to attack is inhibited by photostimulation. Video at 3X (0 - 22 sec), and 0.5X (22 - 37 sec).

Supplementary Video 4 - Hierarchy corridor test and establishment of social rank between a pair of AGG male mice.

The video shows (left) a dominant AGG DAT-Cre mouse (C57BL/6J) injected in PMv with DIO-eNpHR3 in a social rank duel against (right) a subordinate AGG DAT-Cre mouse (C57BL/6J) injected in PMv with DIO-ChR2. The dominant mouse pushes the subordinate out of the corridor which marks the end of the test. Video at 1X.

Supplementary Video 5 - Hierarchy corridor test during photostimulation, and reversal of hierarchical status.

During this session both mice are photostimulated simultaneously, silencing PMvDAT neurons in the dominant (continuous light for eNpHR3 mediated silencing) and stimulating PMvDAT neurons in the subordinate (20 Hz for ChR2 stimulation). Hierarchy reverses typically on the 2nd or 3rd trial during this session. Video at 1X, and of the same pair of mice as supplementary video 4.

Supplementary Video 6 - Hierarchy corridor test post-photostimulation, illustrating a sustained effect on hierarchy.

This video shows immediately post-photostimulation, the two mice maintaining the newly acquired social status between them. This effect persisted in 6/7 pairs after 14 days. Video at 1X, and of the same pair of mice as supplementary videos 4 and 5.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Stagkourakis, S., Spigolon, G., Williams, P. et al. A neural network for intermale aggression to establish social hierarchy. Nat Neurosci 21, 834–842 (2018). https://doi.org/10.1038/s41593-018-0153-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41593-018-0153-x

This article is cited by

Search

Advanced search

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