A couple of papers came out in Science 2 weeks ago on the
behavior of mice toward their unconscious or dead peers. They have resuscitation-like behavior that
consists of biting the face and tongue, clearing the airway, elevating the
tongue of the unconscious mouse to revive them. Mice who were the recipients of
this behavior recovered sooner than mice who did not receive these
efforts. The authors went on to see if
they could localize the behavior in the brain and using modern neuroscience techniques,
they able to show that they had localized the behavior to a brain
substrate. These papers have
implications for psychiatry both at a theoretical and practical level.
In this experiment researchers used genetically identical
mice (cross bred for 20 generations).
They followed their reactions closely across time epochs as they
encountered a familiar partner anesthetized and unresponsive or euthanized
(dead) mice. When compared with the active partner, mice spend 47.4% of the
time interacting with the unresponsive partner compared to 5.8% of the time
with the active partner. They also
displayed a specific pattern of behaviors directed facial area, trunk, limbs
and head of the unresponsive partner. 31.8% of the time was directed at the
orofacial area. The overall duration of
contact was extended with the unresponsive partner and it was more focused on
the orofacial area.
To confirm the sequence of events, short acting (isoflurane)
anesthesia was used to observe the partner response as the anesthetized mouse
became less and then more responsive.
Time spent interacting with the anesthetized partner and the topology of
interactions (more targeted orofacial activity correlated with the time
anesthetized. Deceased cagemates caused
the same pattern of interaction but not sleeping cagemates because during sleep
the directed behaviors led to movements of the sleeping mouse. This suggests
that mice can differentiate responsive and unresponsive states in their
partners.
The orofacial focused behaviors were studied under a
high-speed camera. It was discovered
that in nonresponsive mice the tongue was bitten and pulled out of the mouth by
casemates. A similar tongue protrusion
did not occur in unresponsive mice.
Pulling the tongue out resulted in a larger airway and the removal of
foreign objects in the mouth. Placing the foreign object in a non-oral orifice
did not result in removal. Stimulating the oral mucosa was also a strong
stimulus for the righting response and arousal.
Familiarity was a stronger stimulus for eliciting the
grooming and resuscitation like behaviors than sex differences or
similarities. When given a choice
between familiar partners – one anesthetized and one not – there was a
preference for the anesthetized partner.
That did not persist when both target mice were unfamiliar. The authors also demonstrated that the time
spent with the unresponsive mouse did not decrease over a period of days – it
did not show habituation.
After characterizing the unique behavioral aspects of the
resuscitation like behavior the authors looked at the neural substrate. One
hour after exposure to the unresponsive mice the cagemates were given
4-hydroxytamoxifen to label activated neurons.
Following 2 weeks of exposure several brain structures previously
implicated in the observed behaviors including the medial amygdalar nucleus
(MEA), paraventricular nucleus of the hypothalamus (PVH), basolateral amygdalar
nucleus (BLA), hippocampus, and ventromedial hypothalamic nucleus (VMH) were
examined. The number of tdTomato+ cells were noted in mice that had responded
to unresponsive mice compared with controls (active partners). This marker is a sign of the transcription
factor c-fos and that neurons were recently active. Using a second set of
probes roughly twice as many PVH oxytocin neurons were c-fos + and oxytocin +
in the mice exposed to an unresponsive partner.
An additional optogenetics experiment was done to look at
the oxytocin PVH neurons. Optogenetically
silenced PVH neuron had the expected effect of reducing interaction time with
the unresponsive partner. Optogenetically activating PVH neurons had the effect
of increasing interaction time with an unresponsive stranger mouse. An oxytocin receptor antagonist has the expected
effect of reducing interaction time with the unresponsive mouse.
The authors conclude that they have demonstrated the necessity
of both an oxytocin neuronal substrate and oxytocin signaling for very specific
interactions and behaviors toward unresponsive mice. Further, encountering a dead or anesthetized
partner is required to elicit this response but it does not occur with a sleeping
partner. They point out that rapid responses to an unresponsive partner can
reduce the time it takes for recovery and decreases the risk of predation. They suggest further work needs to be done on
defining the neural network and sensory cues necessary for these behaviors.
As I read this paper I had a few thoughts:
1: The hypothalamus
is underemphasized in psychiatry. We
spent a couple of decades studying neuroendocrinology that was primarily focused
on the pituitary gland. In clinical work, this is also an important system to
be able to assess. But in much of the work about the theoretical basis of
behavior – not much is said about the hypothalamus.
2: Social behavior
versus the neurobiological substrate – psychosocial reductionists frequently
treat the brain like a tableau rasa - not much is there until some kind of
social learning or adverse event occurs.
In this case, adaptive complicated behavior is observed without any
training. The experiment illustrates
both an anatomical and neurochemical basis for the behavior. Why would we
expect anything to be different in humans?
3: The measurements
of activity in both the resuscitation-like and non-resuscitation behaviors were
interesting. Even though the overt
behaviors were easily observed as occurring or not occurring – the neurons
involved always had some level of activity. In other words – no target behavior
did not equate to no neuronal activity.
It is not simply a case of networks being on or off. This has
implications for how we attempt to correlate networks with behavior and the meaning
of networks having a certain level of baseline activity.
All things considered; I thought this was a great paper. It
reminded me of my biochemistry and pharmacology seminars in medical school where
we would have spent a lot more time on the experimental methodology in this paper. I did check my latest copy of The Molecular
Biology of the Cell (7th edition) and found that the discussion
of optogenetics relative to this paper was very brief, but I suppose that makes
sense. From the standpoint of animal
behavior this also recalls so-called altruistic behavior of some animals. I have a file of that behavior observed in Humpback
whales (Megaptera novaeangliae) interfering with killer whales (Orcinus
orca) and hope to post something about that as well.
George Dawson, MD, DFAPA
References:
1: Sun W, Zhang GW,
Huang JJ, Tao C, Seo MB, Tao HW, Zhang LI. Reviving-like prosocial behavior in
response to unconscious or dead conspecifics in rodents. Science. 2025 Feb
21;387(6736):eadq2677. doi: 10.1126/science.adq2677. Epub 2025 Feb 21. PMID:
39977514.
2: DeNardo L, Luo L.
Genetic strategies to access activated neurons. Curr Opin Neurobiol. 2017
Aug;45:121-129. doi: 10.1016/j.conb.2017.05.014. Epub 2017 May 31. PMID:
28577429; PMCID: PMC5810937.
Image Credit:
Generated by ChatGPT based on my parameters. The original paper has a great image of the specific behaviors mentioned in this post.
