A molecular signature for social isolation identified in the brain

Extended social isolation causes debilitating effects in social mammals such as humans. A study in mice shows that the gene Tac2 is upregulated throughout the brains of socially isolated animals, driving massive behavioural changes.
Noga Zilkha is in the Department of Neurobiology, Weizmann Institute of Science, Rehovot 76100, Israel.

Search for this author in:

Tali Kimchi is in the Department of Neurobiology, Weizmann Institute of Science, Rehovot 76100, Israel.

Search for this author in:

Even the toughest prisoners fear solitary confinement. There is a growing awareness across the globe that we are facing an epidemic of loneliness. Prolonged social isolation and loneliness can lead to many profound physiological and neuropsychiatric conditions, including depression and heart disease, and to increased mortality rates1. In the United States, more than 50% of people over the age of 60 experience loneliness2, and the United Kingdom has appointed a government minister to tackle the issue of loneliness. But the biological mechanisms underlying the effects of social isolation are poorly understood. Writing in Cell, Zelikowsky et al.3 reveal a signalling mechanism that acts in several brain regions in mice to drive some of the harmful effects of the stress caused by chronic social isolation.

The authors examined the effects of two weeks of social isolation on the brains and behaviour of male mice (equivalent to more than a year in these conditions for humans4). First, the researchers used an array of behavioural tests to compare mice kept in isolation with control mice that had been housed in groups. These assays revealed widespread effects. Compared to control animals, isolated mice showed enhanced aggression and hypersensitivity to diverse stressful stimuli. For example, the socially isolated mice responded more aggressively to an unfamiliar mouse placed in their cage. In another assay, the researchers presented mice with a dark circle that loomed overhead, simulating an approaching predator. Control animals froze in response to the threat, but moved normally after the stressful stimulus was removed, whereas isolated mice remained frozen long after the apparent threat was removed.

Next, Zelikowsky et al. investigated the brain mechanisms underlying this behaviour. In a previous study of fruit flies, the same group had identified the gene Tac as essential for the regulation of aggression induced by social isolation5. Rodents have two versions of Tac, which are expressed in various brain regions, including regions associated with social behaviour, anxiety and emotions. Using several independent methods, Zelikowsky and colleagues now found a massive increase in the expression of Tac2 throughout the brain following social isolation.

The gene Tac2 encodes a protein called neurokinin B (NkB), which binds specifically to the receptor Nk3R. The researchers performed a series of experiments to alter NkB signalling in the brain. First, they systemically inhibited NkB signalling in isolated male mice using a drug called osanetant, which inhibits the activity of Nk3R. Admini-stration of osanetant, either throughout the social-isolation period or 20 minutes before behavioural testing, substantially reduced the effects of social isolation on behaviour. Next, the authors genetically upregulated Tac2 expression and simultaneously activated Tac2-expressing neurons in group-housed animals, using specially designed viruses that were injected intravenously but could cross the blood–brain barrier to reach the brain. They found that this genetic manipulation led to group-housed mice behaving in a similar way to those that had been isolated.

Finally, Zelikowsky et al. locally manipulated Tac2 expression and NkB signalling, by injecting either osanetant or viruses to downregulate Tac2 expression or inhibit the activity of Tac2-expressing neurons, into particular locations in the brain. These experiments enabled the authors to attribute specific behaviours to regulation of Tac2 in specific brain regions. The main social effect of isolation — enhanced aggression towards an intruder — was controlled by Tac2 in the dorsomedial hypothalamus. By contrast, acute and persistent stress responses were regulated primarily by Tac2 in the central amygdala (Fig. 1).

Diagram of experimental set-up and of relevant brain regions

Figure 1 | The gene Tac2 mediates various effects of social isolation in mice. Zelikowsky et al.3 investigated how two weeks of isolation affected the brains and behaviour of male mice. They found that Tac2 expression is upregulated throughout the brain, and that the gene’s upregulation in particular areas — including the central amygdala and dorsomedial hypothalamus — led to specific changes in the animals’ social behaviour and in their response to various stressful stimuli.

This work opens a gateway to much future research. First and foremost, it will be interesting to determine whether TAC3, the human equivalent of Tac2, is involved in mediating the effects of loneliness and social isolation in people. To our knowledge, TAC3 has not yet been directly associated with sociality or social behaviour of any kind in humans. However, it is expressed in the human brain and has shown abnormal gene-expression levels in children with autism-spectrum disorder6, which profoundly affects social interaction. The systemic manipulations presented in Zelikowsky and colleagues’ paper could be rapidly applied to humans, because osanetant and other NkB inhibitors have already been tested in clinical trials. These drugs could potentially treat anti-social disorders induced by isolation, as well as mood and anxiety disorders.

Although most of their experiments focused on male mice, Zelikowsky et al. found upregulation of Tac2 in response to social isolation in both males and females. Sex differences in response to stress and isolation are well documented, and are usually conserved across species7. It will therefore be interesting to test whether the roles of Tac2 in mediating the effects of social isolation in females are similar to or different from those in males.

The need for social interactions and the response to social isolation can differ enormously between and within species. Mice and humans, for example, are typically considered to be highly social creatures8. When their social needs are not filled, they can experience debilitating outcomes1,9. Some species (and individuals within a species), however, are more solitary, or even avoid social inter-actions10. Such species or individuals might harbour neuronal mechanisms that are adapted to the lack of social inter-action. Whether or not members of the Tac gene family act differently in solitary individuals or species compared to how they do in more-social individuals or species remains to be determined.

Finally, one has to wonder: to what extent can we rely on a mouse model of social isolation to truly examine the underlying mechanisms of human loneliness? After all, loneliness and mental isolation are subjective, and a person might feel alone even when surrounded by other people. The traits exhibited by mice under prolonged social isolation greatly resemble those found in humans experiencing solitary confinement, so these animals do provide a good model for studying this process. What we currently lack are relevant animal models for other forms of human loneliness, such as social withdrawal or antisocial personality disorder. Expanding our research toolbox — for example, by studying various species, including non-social and community-living animals, as well as humans — might bring us closer to understanding the biology of human loneliness.

Nature 559, 38-40 (2018)

doi: 10.1038/d41586-018-05447-9
Nature Briefing

Sign up for the daily Nature Briefing email newsletter

Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.

Sign Up


  1. 1.

    Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T. & Stephenson, D. Perspect. Psychol. Sci. 10, 227–237 (2015).

  2. 2.

    Gerst-Emerson, K. & Jayawardhana, J. Am. J. Public Health 105, 1013–1019 (2015).

  3. 3.

    Zelikowsky, M. et al. Cell 173, 1265–1279 (2018).

  4. 4.

    Dutta, S. & Sengupta, P. Life Sci. 152, 244–248 (2016).

  5. 5.

    Asahina, K. et al. Cell 156, 221–235 (2014).

  6. 6.

    Martin, J. et al. J. Am. Acad. Child Adolescent Psychiatry 53, 761–770 (2014).

  7. 7.

    Bale, T. L. & Epperson, C. N. Nature Neurosci. 18, 1413 (2015).

  8. 8.

    Insel, T. R. & Fernald, R. D. Annu. Rev. Neurosci. 27, 697–722 (2004).

  9. 9.

    House, J., Landis, K. & Umberson, D. Science 241, 540–545 (1988).

  10. 10.

    Rehan, S. M. & Toth, A. L. Trends Ecol. Evol. 30, 426–433 (2015).

Download references