Dogs that habitually hear a bell at chow time become classically conditioned to drool at the mere chime, as the physiologist Ivan Pavlov showed in the 1890s: Their brains learn to associate the bell with food and instruct the salivary glands to respond accordingly.
More than a century later, in a paper published today in Cell, the neuroimmunologist Asya Rolls has shown that a similar kind of conditioning extends to immune responses. Using state-of-the-art genetic tools in mice, her team at the Technion in Haifa, Israel, identified brain neurons that became active during experimentally induced inflammation in the abdomen. Later, the researchers showed that restimulating those neurons could trigger the same types of inflammation again.
“This is an outstanding body of work,” said Kevin Tracey, a neurosurgeon and president of the Feinstein Institutes for Medical Research in Manhasset, New York. It “establishes that the classic concept of immunological memory can be represented in neurons.” Others before Rolls have suggested that the brain could remember and retrieve immune responses, he said, but “she proved it.”
Ruslan Medzhitov, an immunologist at the Yale School of Medicine in New Haven, Connecticut, considers the new research “very provocative.” But unlike other groundbreaking studies that push boundaries and challenge conventional concepts, he said that this one also evokes “the ‘Oh, it makes sense’ type of reaction.”
Decades of research and everyday experience offer striking examples of the interplay between mind and body. Around the time Pavlov was experimenting with drooling dogs, the American physician John Mackenzie watched one of his patients develop an itchy throat and struggle to breathe upon seeing an artificial rose — suggesting that the perception that pollen was present was enough to provoke her allergy symptoms. In the 1970s, scientists discovered a similar phenomenon while conducting taste-aversion experiments on rats: They repeatedly gave the animals an immunosuppressive drug along with the artificial sweetener saccharin; eventually, they found they could quell the animals’ immune activity with saccharin alone. Many of us can recall times when the mere scent of a food that once made us sick could trigger nausea anew.
But the mechanism responsible for these psychosomatic reactions has always been shadowy. Such experiences “cannot be guided by immunological memory as we know it,” said Rolls. Rather, it seems that these immune responses start in the brain, she said. “Somehow, there are these thoughts that initiate real physiological processes.”
In recent years Rolls’ lab has begun to get a handle on how thoughts and emotions could affect physical health. In 2018, she and her co-workers reported that stimulating neurons in the brain’s pleasure centers in mice disabled a subset of immune cells that suppress the body’s defenses; tumor growth slowed in those animals. In a study published in May, her team found that activating specific nerves in the colon prevented immune cells in the blood from entering the tissue — offering a mechanism for brain control over local inflammation.
Given that these groups of neurons regulated immune activity with such precision, Rolls couldn’t imagine that the brain would control a system without knowing its status. “So we wanted to see how the brain represents the state of the immune system,” she said.
Her team focused on the insular cortex, a structure deep within the brain that processes pain, emotions and the body’s inner physical sensations. “It would make perfect sense that the immune system would be part of this interoceptive information,” Rolls said.
To find out if that was true, the researchers slipped a chemical into the drinking water of laboratory mice to give them a weeklong bout of colitis. The chemical disrupted the inner lining of the colon and triggered a rush of immune cells to the damage, which then harmfully spiraled out of control. A genetic modification in the mice enabled Rolls and her team to fluorescently label neurons active on the day the inflammation peaked, lighting up cells in the insula. They then used a second genetic tool to do something more powerful: They placed a molecular on/off switch onto the activated insula cells.
Then Rolls and her co-workers waited. Several weeks after the colitis subsided and the mice recovered, the researchers used their on/off switch to reactivate the neurons — and triggered a similar inflammatory response in the colon. They saw similar results in mice that had been induced to develop a different inflammatory disease, peritonitis, in the abdominal lining.
The immune responses sparked by neural stimulation “were reminiscent of the original” disease state, Rolls said. The similarities extended to the molecular level: In the mice with induced peritonitis, white blood cells carrying a specific receptor protein became more abundant in the abdominal lining during both the original inflammation and the inflammation evoked later.
The researchers also observed the opposite effect: When they instead inhibited the initial set of activated neurons, the animals’ disease symptoms weren’t as severe. This suggests that even during chemically induced inflammation, signals from the brain may be helping to determine its severity.
In a set of nerve-mapping experiments, the team determined that the insula neurons that kicked into action during the initial inflammation in fact “have a way to deliver a message all the way to the colon,” Rolls said.
In Tracey’s view, the new research shows “you can’t separate the state of the neuron activity from the state of the immune system activity. It’s a two-way street.”
In 2002, Tracey and his colleagues broke ground in this area with their discovery that the brain can send anti-inflammatory signals to other parts of the body through the vagus nerve. This line of research has advanced to the point where bioelectronic devices are being developed and studied to control inflammation in rheumatoid arthritis, pulmonary hypertension and other diseases.
Unlike the vagal nerve system, however, the insula neurons in Rolls’ mechanism sense the inflammation, remember that immune state and can reactivate it — a behavior that is more like Pavlovian conditioning than a negative feedback response, Medzhitov said. Tracey thinks of it this way: The vagus nerve is like a brake line in a car. Rolls’ study shows “there is a driver,” he said. “There is someone who decides whether to hit the brake or the gas pedal.”
However, as Rolls and her colleagues noted in their paper, they cannot yet say whether the insula neurons’ “memory” of the inflammation in some way describes the immune response itself, or if it’s instead a record of the sensations from the inflamed body tissues — in effect, the memory of what it felt like to be sick with that inflammation. They also can’t rule out that other parts of the brain could be involved in remembering the immune response too. What the study does show is that “this information is encoded even though it may not be consciously experienced,” said Medzhitov.
The research could have far-reaching implications. Describing an anatomical pathway that links “your emotional state all the way to the inflammation in the colon,” Medzhitov said, “that, to me, is probably the best demonstration available for psychosomatic control.”
The new findings also upend the common top-down view of the brain. “Most people tend to think, ‘We’re so smart, we decide what to do,’ and then we make our body do it,” Tracey said. “But that’s not how the nervous system works.” Instead, the brain receives and synthesizes information about changes in the body — an infection, a fever — and delivers a response.
Rolls’ work shows that “the brain is inseparable from the immune system,” said Tracey. “I think immunologists and neuroscientists both are going to be excited and surprised.”