A fruit fly eats a meal laced with harmful bacteria, gets sick, and never touches that food again. That much scientists already knew. What they did not know was how the fly’s body taught its brain to remember the danger. A study published in April 2026 in Neuron now traces that lesson to an unexpected teacher: fat.
Researchers at Tohoku University found that when fruit flies consume pathogen-contaminated food, immune activity in the fat body, an organ that functions like a combination of mammalian fat tissue and liver, triggers the release of octopamine, a chemical cousin of noradrenaline in vertebrates. That octopamine signal travels to the mushroom bodies, brain structures that handle associative learning, and stamps in a memory linking that specific food source with threat. The next time the fly encounters the same meal, it steers clear, even if the food is no longer contaminated.
Fat that talks back
The study, titled “A bidirectional brain-fat body axis for pathogen avoidance,” emphasizes that the conversation runs both ways. The brain sends signals that regulate metabolism in the fat body, and the fat body sends immune-derived signals back that reshape what the animal chooses to eat. This loop means a fly does not need to taste or smell a pathogen directly. Its body’s internal infection status can override sensory input and rewrite feeding preferences through memory.
In controlled experiments, flies that had previously eaten pathogen-laced food showed strong, lasting avoidance of that food source afterward. The response was not a momentary flinch. It was a learned association, the kind of memory formation that requires the mushroom bodies to encode and retrieve information about which foods are safe and which are not.
The finding reframes fat from passive energy storage into something more like a frontline scout: a tissue that detects microbial danger and relays that intelligence to the brain’s decision-making centers.
Parallel wiring in mammals
Fat-to-brain communication is not unique to insects. In 2022, researchers supported by the National Institute of Diabetes and Digestive and Kidney Diseases identified sensory nerve cells in mice that carry messages from adipose tissue directly to the brain. That work established that mammalian fat is physically wired into the central nervous system through dedicated neural pathways, not just through slow-acting hormones like leptin drifting through the bloodstream.
The mouse research focused on metabolic signaling, how the brain monitors fat stores and energy expenditure, rather than on pathogen detection or food aversion. But it confirmed the hardware: mammals possess direct neural lines from fat to brain that could, in principle, carry the kind of immune-related signals the Drosophila study describes.
A broader body of work, synthesized in a 2022 review in Physiological Reviews, has long established that peripheral tissues shape feeding behavior through hormones and neural circuits. Leptin reports energy reserves to the hypothalamus. Ghrelin signals hunger. Insulin modulates satiety. What the Tohoku University study adds is a new category of signal altogether: one driven not by calorie math but by immune detection of pathogens, pushing the brain toward avoidance rather than consumption.
What the study does not show
No one has yet demonstrated that this specific pathway operates in humans. The Drosophila fat body shares functional overlap with mammalian adipose tissue and liver, but the organs are not identical. Octopamine, the key molecule in the fly circuit, does not play the same role in mammalian nervous systems. Whether a parallel immune-to-fat-to-brain loop exists in mice or people, perhaps using noradrenaline or another signaling molecule, has not been tested.
The concept is biologically plausible. Humans form powerful, long-lasting food aversions after bouts of food poisoning, a phenomenon known as conditioned taste aversion, or the Garcia effect, which is among the most robust findings in behavioral neuroscience. And immune activation during infection is well known to suppress appetite, driven in part by inflammatory cytokines like interleukin-1β and tumor necrosis factor-α acting on the brain. But whether human fat tissue plays a direct, causal teaching role in these responses, the way the fly fat body does, or whether other immune and neural pathways dominate, remains an open question.
Within the fly itself, details still need filling in. How long does the learned avoidance last? Does it generalize to similar-smelling foods, or stay tightly linked to one specific source? Can starvation override it? These questions matter for understanding how flexible the system is in the wild, where an animal must constantly weigh infection risk against the pressure to find enough calories.
Why fat’s role keeps expanding
For decades, adipose tissue was treated as biological packing material, useful for insulation and energy storage but not much else. That view has eroded steadily. Fat is now recognized as an endocrine organ that secretes dozens of signaling molecules, collectively called adipokines, influencing inflammation, insulin sensitivity, and cardiovascular function. The NIDDK mouse work added a neural dimension, showing fat can communicate with the brain in real time through dedicated nerve fibers, not just through hormones circulating over minutes or hours.
The Drosophila study pushes the boundary further. It suggests fat tissue can participate in learning, not by thinking, but by generating signals that the brain interprets and converts into lasting behavioral change. If anything resembling this circuit exists in mammals, it could reshape how scientists think about appetite loss during infection, the stubborn food aversions that follow illness, and possibly even the metabolic disruptions seen in chronic inflammatory conditions.
That is a large “if.” The fly work is rigorous within its own system: it identifies specific molecular actors, a defined signaling pathway, and a measurable behavioral outcome. But translating insect neurobiology to human medicine requires new experiments in mammalian models, mapping whether immune activation in fat can instruct the brain to encode food-related memories. Until those studies arrive, the Tohoku University findings stand as a striking proof of concept, evidence that the body’s fat reserves are not just listening to the brain but actively shaping what it decides to do next.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
