Figshare · Apr 19, 2026
Symptom-Specific Senescence Pathways in Major Depressive Disorder: Multi-Tissue Transcriptome-Wide Association Evidence for Fatigue and Sleep Disturbance Endotypes
By Cheung N
Major depressive disorder is clinically heterogeneous, yet most biological work still treats it as a single condition. That has limited progress toward more individualized treatment. Cellular senescence, a state marked by stable cell-cycle arrest, mitochondrial dysfunction, resistance to apoptosis, and secretion of pro-inflammatory factors known as the senescence-associated secretory phenotype, has been linked to chronic low-grade inflammation and metabolic dysregulation in age-related conditions, including depression (Franceschi et al., 2018; Seitz-Holland et al., 2023; Lorenzo et al., 2024). What remains unclear is whether senescence-related pathways track particular depressive symptoms rather than major depressive disorder as a whole.To address that question, we conducted a multi-tissue transcriptome-wide association study using summary statistics from the 2024 Psychiatric Genomics Consortium symptom-structured genome-wide association study of depression symptoms (Adams et al., 2024). Predicted gene expression was examined across subcutaneous adipose tissue, cerebral cortex, hypothalamus, liver, and pancreas, and gene-set enrichment analyses focused on mitochondrial function, oxidative phosphorylation, NAD metabolism, intrinsic apoptosis, and senescence-associated secretory phenotype signatures.The clearest signal involved fatigue. Mitochondrial and oxidative phosphorylation gene sets showed significant positive enrichment for fatigue, with Stouffer’s Z values above 4.1 and permutation p values below .005. Several mitochondrial genes, including LARS2 and components of respiratory chain complexes I and IV, showed especially strong positive associations. A different pattern emerged for sleep symptoms. Intrinsic apoptosis pathways and the SenMayo senescence-associated secretory phenotype signature showed opposing directional patterns for insomnia and hypersomnia, with some of the largest differences involving MMP9 and members of the BCL2 family. NAD and sirtuin-related genes contributed weaker but directionally consistent support for linked metabolic regulation.These findings suggest that senescence-related biology may contribute to depressive heterogeneity in symptom-specific ways. In this dataset, fatigue aligned most strongly with mitochondrial dysfunction, whereas sleep-disturbance subtypes appeared to differ in apoptotic and secretory pathway activity. The results are descriptive and need replication, colocalization, and functional follow-up, but they offer a biologically grounded rationale for testing senolytic or NAD-restoring strategies in symptom-stratified subgroups rather than across depression as a single undifferentiated diagnosis.