Back

Sexual Dimorphism in c-Fos Networks Governing Aggression

Aubry, A. V.; Li, L.; Durand, R.; Karpman, E.; Oh, H.; Tavallaei, A. A.; Russo, S.

2025-10-07 neuroscience
10.1101/2025.10.06.680658 bioRxiv
Show abstract

Circuit based studies of aggression often focus on the output of a small group of regions referred to as the "core aggression circuit," yet a whole-brain map of activity has yet to be produced. Using resident-intruder assays in male and female Swiss Webster mice, we combined iDISCO+ c-Fos imaging with a weighted co-expression network analyses to identify mesoscale co-activation modules in aggressive (AGG) vs. non-aggressive (NON) animals. We performed a module preservation analysis between each phenotype within each sex followed by differential correlation to localize edge-level changes. Behaviorally, AGG mice spent more time attacking, while females showed greater social investigation. Network analyses revealed that NON networks often preserved density of AGG modules, but connectivity reorganized with aggression. In male AGGs there was large spread reorganization of numerous modules: a large sensorimotor-subcortical "blue" module, a pallidal/hypothalamic/brainstem "yellow" module, and a brainstem-heavy "green" module. Surprisingly the "brown" module, enriched for classic social-behavior regions was moderately preserved in male NONs. In females, the turquoise module spanning somatosensory/interoceptive cortex and midbrain regions and the "brown" sensory cortex-amygdala module were the least preserved. These results indicate that aggression recruits distributed mesoscale communities via edge-specific gain, with males displaying broad strengthening along a cortex/basal ganglia/ hypothalamus/brainstem axis and females showing more targeted potentiation within sensory/arousal modules. This framework nominates candidate hubs and edges for causal manipulation.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.