The study was led by Alexei Morozov of the Fralin Biomedical Research Institute at VTC, who has long been interested in how the brain converts environmental warnings into behavior. Scientists have documented for decades that animals cluster together when confronted with acute threats such as predators, a phenomenon sometimes described as collective defense. Far less understood is whether a conditioned stimulus, a neutral signal that has acquired meaning through experience, can recruit the same kind of social pulling-together. The Virginia Tech team designed their experiments specifically to separate the learned nature of the threat from its immediate physical presence, asking whether memory alone could reshape social positioning.

The experimental logic was straightforward but demanding. Each mouse was first trained alone to associate a particular tone with a short, mild foot shock, a standard conditioning procedure that reliably produces a lasting memory of the sound as a predictor of harm. One to two days later, when that memory had consolidated but the shock was long past, the researchers placed the mice in pairs and played the tone again. The animals that had previously lived together drew measurably closer to one another when the tone sounded, while mice paired with unfamiliar partners showed no consistent change in the distance between them. The contrast between the two conditions was the central result: familiarity determined whether the warning signal produced social approach.

Importantly, the tendency to draw closer was not simply a byproduct of freezing, the well-known fear response in which mice become motionless when they detect danger cues. The researchers found no link between how much an animal froze and whether it moved toward its partner, indicating that companion-seeking and freezing are distinct behavioral outputs generated by the same warning signal. This dissociation matters for how scientists interpret fear conditioning experiments more broadly, because it suggests that a single conditioned cue can split into multiple parallel behavioral streams, one expressed through body posture and stillness and the other through deliberate spatial reorientation toward a trusted individual.

What the familiar pairs did was subtle rather than theatrical. The team did not observe overt comforting behaviors such as grooming or huddling of the kind sometimes described in other social species. Instead, their quantitative measure was a decrease in the distance between the animals’ snouts, showing that the familiar mice not only closed the physical gap between them but also oriented their bodies and heads toward one another. That orientation detail is significant, because it implies a directed social engagement rather than random crowding. The animals appeared to be monitoring and positioning themselves relative to a specific, recognized individual, which is precisely the kind of behavior a social-memory system would be expected to regulate.

Morozov framed the stranger result in terms of the ecology of the species. Mice are territorial animals, and an unfamiliar mouse of the same sex can read more like an intrusion than an ally, so a danger cue that would drive a resident toward a known cagemate produces no such effect with a stranger. He drew a parallel to human behavior, noting that people, too, may be less inclined to cooperate with groups they do not know, and that learning about one another can help dissolve that barrier and make collective responses to shared threats easier. The analogy is suggestive rather than direct, but it points to why the researchers believe the circuitry they identified could have relevance well beyond rodent behavior, particularly for neuropsychiatric conditions in which social approach is impaired.

The mechanistic core of the study lies in a specific connection between two of the brain’s most intensively studied structures. The basolateral amygdala is essential for processing threatening cues and attaching emotional significance to sensory signals, while the ventral hippocampus is deeply involved in social memory, the storage and retrieval of information about specific individuals. The researchers temporarily suppressed neuronal communication along the pathway running from the basolateral amygdala to the ventral hippocampus. When that projection was disrupted, the mice no longer moved closer to their familiar companions upon hearing the tone, even though their freezing responses were unchanged. The result cleanly separates the threat-detection side of the circuit from the social-approach side: the animals still recognized the tone as dangerous, but the signal no longer translated into proximity-seeking.

A second line of evidence implicated oxytocin, the neuropeptide famous for its roles in social recognition, bonding and affiliative behavior across mammals. When the team blocked receptors for oxytocin, the overall proximity response disappeared as well. Intriguingly, the researchers have not yet determined exactly where in the brain oxytocin acts to influence this particular behavior, leaving an open question that the group considers a priority for future work. Together, the two manipulations sketch a circuit-level model in which the amygdala, having identified a threat, recruits the hippocampus to coordinate a social response, while the hippocampus simultaneously acts as a gatekeeper, permitting that coordination only between animals with an established social history.

That gatekeeper role is what Morozov sees as the most conceptually interesting aspect of the findings. In his view, the amygdala recognizes the threat and the hippocampus holds the social memories, and the conjunction of the two determines whether danger produces approach or indifference. He has suggested that studying hippocampal activity during these experiments will help uncover how that gate actually operates at the cellular level, a question that touches on one of the central puzzles in social neuroscience: how the brain tags specific individuals as safe or unsafe and then uses those tags to bias behavior in real time. The current study provides the behavioral assay and the first circuit-level leverage points for answering it.

Michael Friedlander, Virginia Tech’s vice president for health sciences and technology and executive director of the Fralin Biomedical Research Institute, emphasized the translational promise of the work. In his assessment, the experiments by Morozov and his team represent far more than the exploration of a basic mechanism; they begin to deliver the kind of mechanistic understanding that will be essential for developing precise therapies for neuropsychiatric disorders in which adaptive social interactions in humans are compromised. Conditions ranging from autism spectrum disorder to social anxiety involve difficulties in deciding whom to approach and under what circumstances, and a defined amygdala-hippocampus circuit gated by oxytocin offers a concrete biological target for that broader clinical effort.

The study was conducted by Wataru Ito and Alexei Morozov of the Fralin Biomedical Research Institute, where Morozov is a faculty member of the Center for Neurobiology Research and also holds an appointment in the Department of Psychiatry and Behavioral Medicine at the Virginia Tech Carilion School of Medicine. The research was published as an experimental study in Neuropsychopharmacology under the title describing proximity in mice induced by an auditory-conditioned stimulus, with an article publication date of 15 July 2026. Funding came from the National Institutes of Health and the Seale Innovation Fund, and the authors declared no competing interests. For a field that has long treated fear conditioning as a story about individual animals and their internal states, the work adds a distinctly social dimension: the brain’s alarm system, it seems, does not merely command the body to freeze or flee, but also checks who is standing nearby before deciding whether closeness is a comfort worth seeking.

Subject of Research: Learned threat cues driving familiarity-dependent social proximity in mice through amygdala-hippocampal circuitry and oxytocin signaling

Article Title: When danger is about, mice look to their friends

Article References: When danger is about, mice look to their friends. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: mice, social behavior, fear conditioning, amygdala, hippocampus, oxytocin, neuropsychopharmacology, threat response, social memory, Virginia Tech, behavioral neuroscience, proximity

Traditional antigen discovery has relied heavily on bulk omics. RNA sequencing of tumour biopsies identifies transcripts that are elevated in tumours relative to normal tissues, and mass spectrometry of immunopeptidomes identifies the peptides actually presented on human leukocyte antigen molecules. These approaches have been enormously productive, yielding the tumour-associated antigens that underpin therapeutic cancer vaccines, bispecific antibodies, and adoptive cell therapies. Yet they share a structural blind spot: they measure abundance, not arrangement. A transcript that is highly expressed in a tumour may also be expressed in a vital healthy tissue, and a peptide that appears abundant in a dissociated tumour sample may, in the intact tissue, be presented only on stromal cells rather than on the malignant compartment itself. Dissociation destroys the spatial relationships that the immune system, operating in intact tissue, would encounter.

Proximity labelling technologies emerged precisely to address this limitation. By fusing an engineered enzyme, such as a promiscuous biotin ligase or a peroxidase, to a protein of interest, researchers can covalently tag molecules that come within a few tens of nanometres of that protein in living cells. The tagged molecules can then be enriched and identified by mass spectrometry, producing a snapshot of the local molecular neighbourhood rather than the global composition of the cell. Applied to tumour biology, proximity labelling offers a way to ask what molecules physically congregate around a given marker protein, information that bulk profiling cannot provide. The microenvironment of a membrane protein, its partners, its neighbours in the plasma membrane, and the proteins trafficked alongside it, constitutes a layer of biological organisation that is invisible to standard transcriptomics and only partially accessible to conventional proteomics.

The challenge that arises once proximity data are generated is computational. A proximity labelling experiment produces a network-like structure: bait proteins, their tagged neighbours, the abundances of those neighbours, and the connections among them across conditions and cell states. Representing this as a simple list discards most of its meaning. Graph-structured data demand graph-aware analysis, and this is where modern graph learning methods become relevant. Graph neural networks and related architectures are designed to learn representations of nodes in a network that incorporate information from their local neighbourhoods, so that the identity of a protein is encoded not only by its own properties but by the company it keeps. In biological settings, this inductive bias is often exactly right: function in cellular systems is relational, and proteins that occupy similar network positions frequently share functional roles even when their sequences are unrelated.

Applying graph learning to proximity data in the tumour context creates an opportunity to formalise an intuition that immunologists have long held informally. An antigen that is safe to target therapeutically is not simply one that is absent from healthy tissue in a bulk measurement; it is one whose presentation is unlikely to occur on healthy cells under the conditions the immune system will actually encounter. Proximity provides a proxy for this contextual specificity. A candidate antigen that is consistently found in the immediate molecular neighbourhood of validated tumour markers, and that participates in the same spatial programmes as known malignant-surface proteins, carries more evidence of tumour association than a candidate identified by expression alone. Graph learning allows this evidence to be aggregated systematically across many candidates and many data modalities, rather than assessed one protein at a time by expert curation.

The therapeutic stakes of this problem are considerable. CAR T cell therapy, which engineers a patient’s T cells to recognise a surface antigen, has produced remarkable outcomes in haematological malignancies where a truly tumour-restricted target such as CD19 exists. Solid tumours have proven far more resistant, and a major reason is the absence of comparable target antigens. The antigens that are available on solid tumours, such as HER2, EGFR, or B7-H3, are frequently shared with essential healthy tissues, and targeting them produces on-target off-tumour toxicity that can be dose-limiting or fatal. The field has responded with a variety of engineering strategies, including logic-gated CARs that require two antigens to be present simultaneously, tunable affinity receptors that respond only to high antigen density, and adaptor systems that allow dosing control. All of these strategies depend on knowing which combinations of antigens are jointly specific to tumours, and this is a question about spatial co-organisation that proximity-guided approaches are well positioned to answer.

There is also a deeper immunological rationale for thinking in terms of proximity. The immune synapse itself is a proximity phenomenon: a T cell commits to killing only after sustained engagement with a target cell, integrating signals from many receptor-ligand interactions across the contact interface. Antigens that cluster together on the tumour surface may be recognised more effectively than antigens presented in isolation, because multivalent engagement strengthens T cell receptor signalling and can overcome the inhibitory signals that tumours deploy. Conversely, an antigen that is spatially segregated from co-stimulatory context may be immunologically silent even if it is abundant. Understanding the spatial organisation of tumour antigens therefore has implications not only for target selection but for predicting the quality of the immune response that targeting will provoke.

The tumour microenvironment adds further layers of complexity that proximity-aware methods are suited to capture. Tumours are not homogeneous masses of malignant cells; they are ecosystems containing fibroblasts, endothelial cells, macrophages, T cells, and extracellular matrix, all arranged in structured architectures that vary between patients and between regions of the same tumour. A candidate antigen expressed by tumour-associated fibroblasts, for example, might be attractive for stromal targeting strategies but inappropriate for direct tumour-cell killing. Distinguishing the malignant compartment from the reactive stromal compartment requires information about which proteins co-localise with which, and dissociated single-cell methods, while powerful, lose the tissue architecture that defines these compartments. Proximity labelling performed in intact systems, combined with graph-based inference, offers a route to recovering some of this architectural information from molecular data.

From a machine learning perspective, the tumour antigen problem illustrates a broader trend in computational biology: the shift from classification of individual entities to inference over relational structures. Early bioinformatics treated each gene or protein as an independent feature, and predictive models were built on expression vectors. The realisation that biological molecules operate in networks led to a family of methods that propagate information across interaction graphs, including network propagation algorithms, random walk approaches, and eventually graph neural networks. Each generation of methods has expanded the kinds of questions that can be asked. Where earlier approaches could ask whether a protein is connected to known disease genes, graph learning can ask whether a protein occupies a network position characteristic of disease-relevant molecules, a subtler and often more robust criterion. In the antigen discovery setting, this means the model can learn what the neighbourhood of a validated tumour antigen looks like and then score uncharacterised candidates by the similarity of their neighbourhoods.

Validation remains the essential counterweight to computational prediction, and the history of antigen discovery offers sobering lessons about candidates that looked compelling in silico but failed in vivo. A predicted proximity antigen must survive several successive tests: confirmation that the protein is genuinely presented on the tumour cell surface by human leukocyte antigen molecules, demonstration that healthy tissues lack comparable presentation, evidence that T cells capable of recognising the presented peptide exist in patients, and finally evidence that engaging those T cells produces tumour killing without tissue damage. Each of these tests is demanding, and attrition between stages is high. Computational methods that incorporate proximity information improve the front end of this pipeline by enriching the candidate list for molecules likely to pass the later stages, which is a meaningful advance even though no computational prediction can substitute for experimental validation.

The timing of this work within the broader technological landscape is notable. Spatial transcriptomics and spatial proteomics have matured rapidly, multiplexed imaging now routinely profiles dozens of proteins in intact tissue sections, and proximity labelling has been adapted to an expanding range of model systems. Meanwhile, immunopeptidomics has improved in sensitivity to the point where thousands of presented peptides can be catalogued from limited clinical material. The convergence of these technologies with graph-based machine learning creates conditions in which antigen discovery can become less dependent on serendipity. Historically, many successful tumour antigens were found through patient-specific approaches, such as isolating tumour-infiltrating lymphocytes and identifying their targets, which is powerful but slow and individualised. A systematic, proximity-informed discovery framework aims to generalise this process, producing catalogues of candidate antigens that can serve the broader population of patients.

There are also implications beyond T cell therapy. Antibody-drug conjugates require surface antigens with sufficient differential expression and internalisation behaviour, and the proximity context of an antigen can inform predictions about its trafficking and its accessibility to circulating antibodies. Bispecific molecules that bridge tumour cells and T cells depend on pairs of antigens whose joint expression pattern is tumour-restricted, and proximity data speak directly to co-localisation. Even vaccine design benefits, since peptide antigens that arise in the context of a tumour-specific molecular programme are more likely to elicit responses that discriminate tumour from self. In each of these therapeutic modalities, the fundamental question is the same: which molecular features distinguish the tumour cell surface in its intact context, and how confidently can that distinction be made for a given patient population.

As the field moves forward, the integration of proximity-guided graph learning into antigen discovery pipelines will likely be judged by a practical standard: whether the candidates it surfaces translate into therapies with wider therapeutic windows than those discovered by expression-based approaches alone. The conceptual contribution, however, may prove equally durable. By treating the tumour cell surface as a structured, relational system rather than a list of abundances, this line of work aligns the computational representation of tumour biology with the way the immune system itself reads that biology, through contact, context, and the molecular neighbourhoods in which every antigen is embedded.

Subject of Research: Proximity-guided graph learning reveals tumour-associated proximity antigens

Article Title: Proximity-guided graph learning reveals tumour-associated proximity antigens

Article References: Scandore, C., Malone, C. F., May, C. K., de Regt, A. K., Guernsey, J., Ma, H., Dephoure, N., Setter, B., Howell, R. A., Johnson, K. R., Farr, C. L., Romero, S., Vignale, L., Vittum, T., Dawson, E., Habtetsion, T., Nardi, F., Woodruff, B., Mathay, M., … Fadeyi, O. O. (2026). Proximity-guided graph learning reveals tumour-associated proximity antigens. Nature. https://doi.org/10.1038/s41586-026-11003-7

Image Credits: AI Generated

DOI: 10.1038/s41586-026-11003-7

Keywords: Proximity-guided, graph, learning, reveals, tumour-associated, proximity, antigens, scientific research