The anterior cingulate cortex (ACC), a fold of brain tissue tucked just atop the corpus callosum at the center of the brain, has become known as a hotbed of emotional processing, including the affective components of pain. But could neurons of the ACC respond to another person’s pain as well as our own, contributing to our sense of empathy?
New work from Christian Keysers and colleagues at the Netherlands Institute for Neuroscience, Amsterdam, reveals the presence of emotional “mirror neurons” in the rat ACC that selectively encode pain (but not fear) in the self and in others. Silencing the ACC with muscimol, a GABAA receptor agonist, rendered rats insensitive to social contagion of pain, suggesting this brain region produces the emotion associated with observing pain rather than simply reacting to it.
“The science is very elegant, and I think it’s a very relevant finding in pain and social learning that will help us understand the role of mirror neurons in encoding actions, emotions, and pain,” said Luana Colloca, a pain researcher at the University of Maryland, Baltimore, US, who was not involved in the new study. Most exciting, she said, is that the authors “identify a subpopulation of neurons that respond to shared pain, when it’s observed in another.”
The work was published April 22, 2019, in Current Biology.
Mirror in the brain
Mirror neurons were first identified in the 1990s, when researchers recorded from neurons in the motor system of a monkey as it observed a researcher picking up an apple from a table.
“These neurons activate both when you do something, and when you watch someone else do it,” Keysers said.
That discovery changed the way neuroscientists conceived of how we process the actions of others, Keysers said.
“At that point, people still thought that observing other people would primarily activate visual-spatial processing in the brain. There was a belief that somehow we just make sense of it like a game of chess, in an intellectual way.” The existence of mirror neurons “was the first evidence that we don't just engage visual and intellectual brain regions, but also those associated with our own body and actions,” he said.
Now, Keysers and colleagues have shown that mirror neurons—those activated by direct and by vicarious experience—also encode pain. The authors stipulated that, in order to qualify as mirror neurons, cells would need to respond selectively to one emotion but not another, and would use a common neural code to encode that emotion in the self and in others.
To find mirror neurons, the researchers made electrophysiological recordings from the ACC of rats while the animals underwent pain or observed another rat in pain. Pain in the former condition was evoked by a CO2 laser, while in the observed pain condition, the animals saw another rat undergoing painful foot shocks.
To determine whether the ACC responses were selective for pain, the researchers compared pain with another salient, aversive emotion: fear. To establish a conditioned stimulus (CS) that would evoke fear, rats were exposed to painful foot shocks paired with a 20-second tone. The tone later served as the CS.
“It was very clever to add the fear condition along with the painful stimulus. That allowed them to separate what belongs to pain versus fear circuits” in the ACC, Colloca commented.
Co-first authors Maria Carrillo and Yinging Han, also of the Netherlands Institute for Neuroscience, implanted rats with electrodes to record multiunit activity (MUA), the pooled spiking activity of thousands of nearby neurons. They recorded from 425 channels in 17 rats. Three quarters (313) of the channels responded with increased firing to at least one of these conditions: pain, observed pain, or fear evoked by the CS. Of those 313 channels, 62 percent (193) responded more robustly to observation of another rat in pain than to a control condition that replicated the sound of the foot shock machinery. Among those 193 channels, 71 percent also responded to direct experience of either pain or fear.
To get a better look at the activity of individual neurons, the authors used a data analysis program to separate spiking patterns of single cells from the MUA channels. They identified 84 cells in 13 rats for analysis. Seventy-three neurons responded to at least one stimulus condition. Of the responders, 81 percent responded to observation of pain in another rat, and two-thirds of those also responded to a “self” experience of pain or fear, qualifying them as potential mirror neurons. Only three of these neurons responded to both the painful laser and the fear CS, whereas 25 were selective for pain and 11 for fear.
“Here we observed that 60 percent of the neurons that responded to pain experience [of the self] also responded to pain of the other,” Keysers said. “That was really a strikingly high number, and that shows how prevalent the representation of another animal’s pain is in a system we associate mainly with personal pain.” In previous studies, by contrast, “in the motor system, only about 10 percent of neurons were active both when one is doing something and when observing.” (See Gallese et al., 1996Kohler et al., 2002).
After recording, the authors made an electrical lesion to mark the location of the electrodes, and the brains were sectioned and stained. Histological reconstruction of the recorded cells, mainly in area 24 of the ACC, revealed that the mirror neurons were not clustered but spatially distributed and interspersed with non-mirror neurons.
A common neural code
The researchers next wanted to determine whether neurons used a common mechanism to encode distress in the self and in others, so they applied a decoding algorithm to the spike counts of 69 neurons to determine what firing patterns the cells were using to encode relevant information. They first trained the algorithm to decode spikes during observation of pain. The same algorithm was able to decode activity during laser-evoked pain in the self, indicating the neurons indeed used a common neural code.
“It’s been shown in humans that there is a similar, shared mechanism for ‘self’ pain and our ability to ‘feel’ the pain of others,” Colloca told PRF (see Zaki et al., 2016). But although data from functional magnetic resonance imaging (fMRI) suggests that the same brain regions are activated by direct and vicarious pain, the technique cannot show that the same neurons are activated by the two phenomena. “This is really the first time we see direct and observed pain represented in individual neurons in an animal study, and we see that the neural encoding is actually similar,” Colloca said.
“Before this, there was no evidence for single neurons that map others’ pain onto your pain,” Keysers added.
Next, the team wondered what specifically about observing a peer in pain triggered activity in the ACC mirror neurons. They found that observed (demonstrator) rats in pain jumped and squeaked dramatically during the one-second period following the shock stimulus, when the ACC response in observer rats was maximal.
“As we record from the rats, the demonstrator rat emits a pain squeak, and 100 milliseconds after that, the observer has activity in the pain system. That’s really quite rapid. It tells us it’s a bit of a reflex-like reaction, this mapping of pain onto the observer’s own pain,” said Keysers.
Pain-evoked nocifensive behaviors in response to the laser included paw licking but not jumping or squeaking, indicating that the same neurons responded to both visual-auditory stimuli from others and to nociceptive inputs from the self. That showed that while those ACC neurons were tuned to respond to pain, they were also multimodal, rather than responding to only one type of sensory stimulus.
In a final experiment, the researchers wanted to determine whether ACC mirror neuron activity was responsible for the vicarious emotion, or simply a reflection of neural activity elsewhere. They made microinjections into the ACC using muscimol, a potent GABAA receptor agonist found in the psychoactive mushroom Amanita muscaria, to deactivate the area. The investigators then compared rats’ freezing behavior in two conditions: in response to the fear CS, and in response to observing another rat in pain, indicating social transmission of pain. Following muscimol injection, freezing was reduced in the observation condition but not in response to the CS, suggesting that social contagion of pain (but not freezing itself) depended on ACC activity.
The ability of humans to learn from one another is critical to our survival. And the finding of the mirror neurons in rats, Colloca said, “suggests it’s a very old mechanism of learning. No matter what species, animals observe one another and experience a great learning opportunity to optimize behaviors.” That learning, she added, is an extremely complex phenomenon that occurs across multiple brain regions and involves many different types of mechanisms.
The new work also shows that those learning opportunities extend beyond simple behaviors.
“When we observe something more complex, something more emotionally associated, that can help us make inferences about the thoughts and actions of others,” Colloca said.
Colloca says the new findings might even translate to better pain treatments.
“They complement very well what we’re learning about social modification of pain in people, and the implications are significant,” she said. “Once we learn the mechanisms behind social learning, we can actually use that to improve therapeutic benefits.”
For example, patients might watch a video of another person undergoing successful treatment for pain. “If they can be exposed to the experience of others and learn from it, that’s much more effective than merely telling a patient, ‘we observed this benefit.’ This can amplify our ability to manage pain.”
Stephani Sutherland, PhD, is a neuroscientist and freelance journalist in Southern California. Follow her on Twitter @SutherlandPhD.
Image credit: Alexander Kharchenko/123RF Stock Photo.