Depression involves dysregulation across distributed cortico-limbic circuits, and ketamine is notable for its rapid antidepressant effects. Although depression and ketamine treatment have been linked to altered brain network topology, how within-frequency and cross-frequency coupling are jointly reorganized at the brain-wide level remains unclear. Here, we developed a frequency-varying multilayer brain functional network (FMBFN) framework to analyze local field potential recordings from eight brain regions in male C57BL/6 mice. This framework integrates within- and cross-frequency coupling and extracts multi-scale network features to characterize brain network structure. Applying this approach in the chronic social defeat stress (CSDS) model, we found that CSDS was associated with frequency-specific hyperconnectivity and selective alterations in network integration during social interaction. Ketamine reversed social avoidance and induced the distinct reorganization of multilayer network topology, including region-specific nodal changes. Notably, the lateral habenula showed the response pattern opposite to that of the other recorded regions. As an exploratory cross-modal extension, we further examined gut microbial features and found that specific ketamine-associated microbial changes were linked to global network topology, suggesting candidate gut-brain association patterns. Together, these findings establish the FMBFN framework as a systems-level tool for characterizing brain-wide neural dynamics in psychiatric disorders and for linking network-level alterations to biological contexts.
