Dynamic Gene Expression Signatures Predict Psychotherapy Response in PTSD

Static blood-based biomarkers have provided limited insight into psychiatric illness and recovery, in part because they capture only a single, unperturbed snapshot of inherently dynamic biological systems. We previously developed a functional transcriptomic approach that profiles living peripheral blood mononuclear cells (PBMCs) under controlled glucocorticoid stimulation, demonstrating that stress-responsive transcriptional programs converge across immune cells and human induced neurons. Here, we apply this platform longitudinally to characterize biological response trajectories associated with psychotherapy outcomes in post-traumatic stress disorder (PTSD). PBMCs were collected from veterans with PTSD and trauma-exposed controls before, after, and 12 weeks following cognitive processing therapy (CPT), and profiled at baseline and following in vitro dexamethasone (DEX) stimulation. PTSD participants were classified as responders or non-responders based on Clinician-Administered PTSD Scale (CAPS-5) trajectories. At baseline, responders and non-responders exhibited pronounced transcriptional differences exceeding those observed between PTSD and controls. Longitudinally, responders showed a marked shift toward control-like profiles, whereas non-responders retained persistent divergence. Pathway analyses identified a prominent temporal signal within Class C/3 G protein–coupled receptor pathways driven largely by bitter taste receptors (TAS2Rs), along with subgroup-specific regulation of GABAB, opioid, and NMDA signaling. DEX-induced transcriptional responses further differentiated subgroups and showed opposing associations with recovery trajectories, implicating glucocorticoid sensitivity as a potential predictor of treatment outcomes. Together, these findings indicate that biological responsiveness itself constitutes a measurable phenotype of recovery, and that dynamic functional profiling of living immune cells can reveal clinically relevant trajectories and candidate mechanisms not accessible through static biomarkers.