ABOVE: A mouse brain shows activated neurons (white) in the brainstem (zoom on R) 3 hours after LPS injection. Scale bars: 1 mm (L), 0.25 mm (R) FIG. 2D FROM NATURE, 609:761–771, 2022. CC BY 4.0

Infections are often associated with symptoms that aren’t directly tied to the pathogen, such as lethargy and loss of appetite. Scientists have long been interested in understanding where these so-called ‘sickness behaviors’ are ultimately controlled, as that information could shed light on the brain’s influence on the immune system and potentially lead to new treatments to speed recovery from myriad illnesses. Now, research in mice published earlier this month in Nature has tracked much of that control to a set of neurons deep in the brainstem.

“I think it’s really a significant advance,” says Keith Kelley, a professor emeritus of immunophysiology at the University of Illinois and former long-time editor-in-chief of the journal Brain, Behavior, and Immunity, who was not involved in the work. “It actually shows a population of cells in the brainstem that are responsible for linking what happens in the body to what goes on in the brain.”

Zeroing in on the neurons that make us feel sick

Our bodies are constantly trying to maintain a kind of equilibrium, controlling things like our body temperature, how often we feel hungry, and how much we sleep. This careful balance, known as homeostasis, is how we manage to stay alive and healthy in the world. “Usually, these things are really well controlled, and body really prioritizes that,” says study coauthor Anoj Ilanges, a biologist at the Janelia Research Campus in Ashburn, Virginia who conducted the study while at The Rockefeller University in New York.

This balance changes when we get sick, triggering a constellation of symptoms and physiological changes collectively referred to as sickness behaviors that can help us recover.

Previous research had suggested at least some of the signals that lead to sickness behaviors originate in the brainstem but didn’t pinpoint exactly where within the structure. So, Ilanges and his colleagues decided to investigate. First, they exposed laboratory mice to lipopolysaccharide (LPS), a toxin that consists of pieces of dead bacteria and is known to elicit an immune reaction similar to that triggered by live bacteria. As expected, LPS caused the animals to act sick: They became lethargic and lost their appetites, even though they weren’t infected by a pathogen. And this effect was strong, explains study coauthor and Rockefeller University molecular biologist Jeff Friedman: LPS-exposed mice even refused food after being subjected to a long fast that normally would have pushed them to eat.

Then the scientists examined neuronal activity by looking for a protein called FOS in the brains of mice euthanized after LPS injection. FOS is involved in long-term changes in the brain and is often expressed after neurons fire, and therefore can act as a proxy for neuronal activity. Higher concentrations of FOS indicated a burst of activity in two areas: the nucleus of the solitary tract (NTS) and the area postrema (AP), which sit side-by-side in the brainstem.

But to determine if neurons in these areas are truly responsible for sickness behaviors, they needed to activate them without using LPS, as the toxin is known to cause other changes in the body and brain.

To do this, they injected a virus that delivers a molecular switch sensitive to the antipsychotic drug clozapine directly into the NTS-AP region of the brainstems of special mice. These mice had been genetically engineered such that, when exposed to the anticancer drug tamoxifen, actively-firing neurons—and only actively-firing neurons—would integrate this switch into the FOS-encoding gene. This meant that if the mice were later exposed to clozapine, the neurons that happened to be firing in the NTS-AP region where the virus was injected while the mice received a priming dose of tamoxifen would once again become active. This, in essence, gave the researchers a way to take a snapshot of neural activity as well as a way to recreate that snapshot later.

The researchers then injected the engineered mice with LPS as well as with the switch-priming, snapshot-taking tamoxifen. After a few weeks of recovery, the researchers gave the mice an injection of clozapine and, once again, NTS-AP neurons produced FOS and the mice displayed sickness behaviors, even without any LPS in their system. To the team, this confirmed that neurons in the NTS-AP region contribute to feeling ill. Further experiments using single-nucleus RNA sequencing narrowed the specificity of LPS-activated neurons down even further to ones in those regions that also express a protein called ADCYAP1.

There’s a lot going on, in terms of the immune system communicating with the brain and the brain controlling our physiology during infection. And I think this is only the beginning of really exploring this axis.

—Anoj Ilanges, Janelia Research Campus

Ilanges’ team also found that inhibiting ADCYAP1-expressing neurons reduced sickness behaviors in response to LPS injection, though it didn’t completely eliminate them.

Kelley noted that he thought the mouse model the team developed to reactivate a specific population of neurons “was really clever.” He also said he’d be interested in seeing further work on some of the sickness behaviors not included in Ilanges’ work, such as sleep disruption or the assortment of general aches and pains collectively known as myalgia.

Patricia C. Lopes, a biologist at Chapman University in California who studies sickness behaviors but didn’t work on the study, points out that NTS-AP neurons may not be the only neurons in the brain that contribute to sickness behaviors. In June, a different group of scientists, also publishing in Nature, identified neurons located in the hypothalamus that act as a kind of control hub to coordinate fever, loss of appetite, and warmth-seeking behavior. To see the two papers come out so close to each other—and in the same journal—“was exciting, but also surprising,” says Lopes. Both the brainstem and the hypothalamus had been previously identified as important to sickness behaviors but being able to identify the cell populations is remarkable, she says. “The specificity to which they’re getting is unprecedented.”

Lopes did note an interesting wrinkle in both papers: All the animals used were male. This is not uncommon in mouse studies, as female mice show large fluctuations in body temperature related to estrus (a potentially confounding factor scientists may want to avoid), but it means that any potential differences due to sex are unknown.

See “Not-so-Mellow Yellow: Pregnant Mice’s Urine Stresses Out Males”

Ilanges’ team also wasn’t able to investigate what specific bodily signals these neurons were responding to, though they note that the NTS is known to relay signals from the vagus nerve—an important communication line between the brain and internal organs—while the AP is known to sense humoral signals, such as proteins released into the blood stream. They were also not able to investigate whether the neurons were active during viral or other non-bacterial infections.

Nevertheless, they hope that others can use the data and methods they’ve set out to continue to explore how the brain and immune system interact, and Ilanges plans to continue this line of inquiry at Janelia.

Ilanges says that figuring out how the brain controls sickness behaviors could also open the door to potential methods for tweaking these mechanisms. For example, one could imagine a drug designed to help chronically ill people regain their appetites.

More broadly, this work shows that the brain plays a critical role and is an active participant in fighting infections, says Friedman. Ilanges expresses similar sentiments. “There’s a lot going on, in terms of the immune system communicating with the brain and the brain controlling our physiology during infection. And I think this is only the beginning of really exploring this axis.”