Chaihu-Guizhi Decoction and Saikosaponin D alleviate anxiety-like behavior by engaging HTR2C-related neurotrophic and excitatory synaptic homeostasis.

BACKGROUND: Chronic stress precipitates anxiety disorders by disrupting hippocampal synaptic integrity and brain-derived neurotrophic factor (BDNF)-TrkB signalling. Chaihu-Guizhi Decoction (CGD), a classical herbal formula, is clinically prescribed for stress-related emotional disturbances. However, its bioactive constituents and underlying neurobiological mechanisms remain poorly defined. METHODS: We combined an exploratory real-world clinical observation with mechanistic investigations to explore the pharmacological basis of CGD. In a clinical cohort, 4-week modified CGD treatment was assessed in relation to anxiety severity and sleep quality. In parallel, we used a chronic restraint stress (CRS) mouse model combined with behavioural tests, electrophysiological recordings, transcriptomics, and untargeted metabolomics to delineate hippocampal mechanisms engaged by CGD. RESULTS: In an exploratory, uncontrolled clinical cohort, mCGD treatment was associated with reduced anxiety severity and improved sleep quality. In CRS mice, CGD reversed anxiety-like behaviours and restored hippocampal excitatory synaptic integrity, as evidenced by recovery of dendritic spine density, excitatory synaptic protein expression, and spontaneous excitatory postsynaptic transmission in CA1 neurons. Integrative network pharmacology and transcriptomic analyses identified the serotonin 2C receptor (Htr2c) as a stress-responsive molecular node that was downregulated by CRS and restored by CGD. Pharmacological activation of HTR2C partially recapitulated the behavioural and neurotrophic effects of CGD, supporting a role for HTR2C-related regulation of BDNF-TrkB signalling in stress resilience. Untargeted metabolomics further revealed that CGD reversed a stress-induced hippocampal metabolic signature, with pathway enrichment implicating glycerophospholipid and tryptophan-related metabolism. Chemical profiling and molecular modelling identified saikosaponin D (SSD) as a representative bioactive constituent with supportive in silico evidence for potential relevance to HTR2C-associated signalling, and SSD monotherapy reproduced key behavioural and synaptic features associated with CGD treatment. CONCLUSION: These findings support an HTR2C-related pharmacological framework involving BDNF-TrkB neurotrophic signalling, through which CGD and representative constituent SSD may restore excitatory synaptic homeostasis within a stress-responsive metabolic context. This work provides a mechanistic working model linking a classical herbal formula to pathways regulating synaptic resilience and emotional regulation under chronic stress.