Oxytocin in the Brain’s Septum Lets Mice Learn Safe Foods From Friends

For a mouse in the wild, a single sniff of a companion’s breath can mean the difference between a nourishing meal and a poisonous mistake. This remarkable ability, known as the social transmission of food preference, allows animals to learn which foods are safe by interacting with others, rather than by risking a potentially lethal taste test themselves. It is one of the most elegant examples of social learning in nature, and it has long fascinated neuroscientists who study how the brain weaves together social signals and survival behaviors. Yet despite decades of research into the psychology of this behavior, the precise neural circuitry that allows a social encounter to rewrite an animal’s food preferences has remained stubbornly obscure. A new study published in PLOS Biology now identifies a critical piece of that puzzle, pointing to a small and often overlooked brain structure called the lateral septum, and to the famous social hormone oxytocin as the molecular key that unlocks it.

The research, led by Hanyu Fu and colleagues, set out to answer a deceptively simple question: when a mouse smells a food odor on the breath of a conspecific and later prefers that food, where in the brain does that social information get converted into a lasting preference? The team focused on the lateral septum, a region nestled beneath the corpus callosum that sits at a strategic crossroads between the olfactory system, the hippocampus, and the hypothalamic circuits that govern feeding and social behavior. Anatomically, the lateral septum is ideally positioned to serve as an integration hub, receiving a copy of the social and sensory information an animal encounters and relaying it to the motivational machinery that decides what is worth eating. But position alone does not prove function, and the researchers knew they would need direct evidence that this region actually participates in socially guided food learning.

To obtain that evidence, the team turned to the standard laboratory paradigm for studying this behavior. In the social transmission of food preference task, a demonstrator mouse eats a flavored diet and then interacts with an observer mouse. The observer, detecting the food’s odor clinging to the demonstrator’s breath, later shows a robust preference for that same flavor when offered a choice between it and an unfamiliar alternative. This learned preference can persist for days, overriding any innate attraction the animal might otherwise have. Using this paradigm, the researchers monitored neuronal activity in the lateral septum during social learning and made a striking observation: neurons in the region became significantly more active following the social interaction, and their firing patterns reliably distinguished between foods that had been socially cued and foods that had not. In other words, the lateral septum appeared to be encoding exactly the information that the behavior depends on.

Correlation, however, is not causation, so the team next asked whether this activity was actually necessary for the behavior. Using chemogenetic tools, a technique that allows researchers to dial neuronal activity up or down with engineered receptors activated by otherwise inert drugs, they selectively silenced neurons in different parts of the lateral septum. The results revealed a sharp subregional divide. Suppressing activity in the dorsal portion of the lateral septum completely abolished the food preference that social learning would normally produce, leaving the observer mice indifferent between the cued and uncued foods. Silencing the ventral portion, by contrast, left the behavior intact. This dissociation demonstrates that the dorsal lateral septum is not merely a passive relay but an essential node in the circuit, and that its contribution is anatomically specific rather than a diffuse property of the region as a whole.

Having established where the critical computation happens, the researchers turned to the question of how. Social behaviors across the mammalian brain are heavily modulated by oxytocin, a neuropeptide best known for its roles in childbirth, lactation, and pair bonding, but increasingly recognized as a versatile modulator of social information processing throughout the nervous system. Oxytocin receptors are expressed in many brain regions, and the team found that the dorsal lateral septum is among them. When they blocked oxytocin receptor signaling specifically within the dorsal lateral septum, the effect was dramatic: mice could no longer acquire a food preference through social interaction, even though their basic olfactory abilities and feeding motivation remained functional. The social channel of food learning had been selectively switched off.

The researchers then dug into the cellular mechanism underlying this requirement, and their findings provide a satisfying mechanistic account of what oxytocin is actually doing in this circuit. Oxytocin receptors belong to a family of G protein-coupled receptors that, upon binding their ligand, activate intracellular signaling cascades. In the dorsal lateral septum, the team found that oxytocin receptor activation engages the phospholipase C pathway, a canonical signaling route that ultimately modifies the excitability of neurons. Specifically, oxytocinergic signaling enhanced the excitability of dorsal lateral septum neurons, making them more responsive to the inputs they receive during the social encounter. When this enhancement was removed, either pharmacologically or genetically, the neurons were less responsive and the social learning failed. The picture that emerges is one in which oxytocin acts as a gain control, amplifying the neural representation of socially transmitted food information at the precise moment it needs to be encoded into memory.

This mechanistic clarity is what sets the study apart from earlier work on social food learning. Previous research had implicated a variety of structures, including the olfactory bulb, the hippocampus, and the medial prefrontal cortex, in different aspects of the behavior, and oxytocin had been shown to influence social recognition in other circuits. What the new study adds is a complete chain of evidence linking a specific subregion, a specific neuromodulator, a specific intracellular pathway, and a specific behavioral outcome. The dorsal lateral septum, according to this account, is where social information about food safety is integrated with the animal’s feeding motivation, and oxytocin is the signal that permits that integration to occur. Disrupt any link in that chain, and the animal falls back on its own unguided preferences.

The findings also carry broader implications for how scientists think about the lateral septum, a region that has historically received far less attention than its neighbors in the septohippocampal system. The demonstration of a sharp functional divide between the dorsal and ventral subdivisions suggests that the lateral septum is not a homogeneous structure but a collection of parallel circuits with distinct behavioral roles. It also reinforces a growing appreciation that oxytocin’s influence on social behavior is not confined to a single reward hub or a single receptor hotspot, but is distributed across the brain in a region-specific fashion. Understanding where and how oxytocin acts, rather than simply whether it acts, may be essential for interpreting the mixed results that have plagued attempts to translate oxytocin-based interventions to human social disorders.

For the field of social neuroscience more generally, the study offers a compelling example of how a socially acquired memory is built. The food preference that an observer mouse acquires is not a reflex; it is a durable association between an odor and the implicit endorsement of a trusted companion, and it must be encoded, consolidated, and later retrieved to guide choice. The new work suggests that the dorsal lateral septum participates in this process by tagging socially cued foods with motivational significance, a computation that depends on oxytocin receptor signaling to boost the excitability of the neurons doing the tagging. Future research will likely explore how this septal mechanism interacts with the hippocampal and cortical systems that support the longer-term storage of the food-odor memory, and whether similar oxytocin-dependent gating operates in other forms of social learning, from mate choice to predator avoidance.

There is also an evolutionary dimension worth savoring. The ability to learn what is safe to eat from the experience of others is a form of cultural inheritance, one that predates language and operates in species as distant as rodents and primates. The new findings hint at the ancient neural machinery that makes such inheritance possible: a peptide that signals social presence, a receptor that translates that signal into cellular excitability, and a septal circuit that converts a companion’s breath into a decision about dinner. For mice, and perhaps by extension for other social mammals, the lesson is that the brain treats information from trusted companions as uniquely valuable, and it has dedicated molecular hardware to make sure that valuable information is not lost. What remains to be discovered is how far this principle extends up the phylogenetic tree, and whether the same oxytocinergic logic governs the ways in which human food cultures are passed from one generation to the next.

Subject of Research: Oxytocin receptor signaling in the dorsal lateral septum enabling social transmission of food preference in mice

Article Title: Oxytocin receptor signaling in the dorsal lateral septum is essential for social transmission of food preference in mice

Article References: Fu, H., Liu, Z., Zhou, J., Zhang, L., Liu, C., Tang, D., Zhang, S., Li, A., & Wang, D. (2026). Oxytocin receptor signaling in the dorsal lateral septum is essential for social transmission of food preference in mice. PLOS Biology, 24(10), e3004044. https://doi.org/10.1371/journal.pbio.3004044

Image Credits: AI Generated

DOI: 10.1371/journal.pbio.3004044

Keywords: oxytocin, lateral septum, social learning, food preference, neuroscience, mice, chemogenetics, phospholipase C, social behavior, PLOS Biology, neural circuits, feeding behavior