Transcriptomic profiling of the human habenula reveals a shared molecular architecture across mood disorders

The habenula is a critical regulator of monoaminergic and reward circuitry and is increasingly implicated in the neurobiology of mood disorders. Preclinical studies demonstrate that habenula hyperactivity drives depressive-like behaviours and can be reversed by interventions such as ketamine and deep brain stimulation. However, the molecular architecture of the human habenula remains largely unexplored. Here, we applied a transdiagnostic framework to characterise shared and disorder-specific transcriptomic alterations across affective illness. Bulk RNA sequencing was performed on postmortem habenula-enriched tissue from controls (n = 6), major depressive disorder (MDD; n = 6), and bipolar disorder (BPD; n = 6) cases from the Netherlands Brain Bank. Differential gene expression and differential transcript usage (DTU) analyses identified diagnosis- and sex-associated transcriptional changes. Individual diagnostic comparisons revealed modest numbers of differentially expressed genes (MDD: 60; BPD: 66; FDR < 0.05), consistent with limited power and/or subtle disorder-specific effects. In contrast, transdiagnostic analysis combining affective cases (MDD + BPD; n = 12) identified 378 differentially expressed genes, indicating a robust shared molecular signature across mood disorders. Upregulated genes were enriched for potassium channel activity, calcium homeostasis, and Wnt signalling, consistent with altered neuronal excitability, while downregulated genes were enriched for metal ion binding. DTU analysis identified 49 isoform switches, highlighting isoform-specific regulation not captured at the gene level. Biological sex contributed substantially to transcriptomic variation, with 67 differentially expressed genes and 18 isoform switches differing between males and females, including sex-dependent regulation of GPR151, NLGN3, and KIF17, genes known to influence neuronal excitability. Together, these findings support a shared molecular architecture across mood disorders and underscore the importance of transdiagnostic and sex-informed approaches.