Gene expression meta-analysis in the prefrontal cortex: unraveling biological underpinnings of suicidal risk

AI Summary

Suicide associated PFC transcriptomic alterations are not uniform and appear diagnosis dependent, with a clearer detectable signal in cases with major depressive disorder.
MDD stratified meta analysis identified four FDR significant genes HRH3, PDE2A, NET1 and RHBDF2 linked to stress pathways, cyclic nucleotide signalling and synaptic organisation.
Overall comparison yielded ten nominally differentially expressed genes only, none surviving multiple testing; findings are hypothesis generating and require replication and functional validation.

BMC Psychiatry. 2026 May 20. doi: 10.1186/s12888-026-08170-2. Online ahead of print.

ABSTRACT

BACKGROUND: Suicide is a complex public health challenge. Although psychosocial factors are important, molecular dysregulations in the prefrontal cortex (PFC) have been implicated in suicidal behavior.

METHODS: Six post-mortem PFC transcriptomic datasets from Gene Expression Omnibus (GEO) were analyzed to investigate the molecular basis of suicide, a major public health problem whose underlying mechanisms remain incompletely understood. After harmonizing microarray and sequencing data from 248 individuals younger than 60 years, both global and diagnosis-stratified meta-analyses were performed, focusing on major depressive disorder (MDD) and bipolar disorder (BPD).

RESULTS: In the overall comparison between suicide cases and controls, 10 nominally differentially expressed genes were identified, although none remained significant after multiple-testing correction. In the MDD-stratified analysis, four genes survived false discovery rate (FDR) correction (HRH3, PDE2A, NET1, and RHBDF2) and were associated with stress-related pathways, cyclic nucleotide signaling, and synaptic organization. No significant genes were identified in the bipolar disorder subgroup.

CONCLUSION: These findings suggest that suicide-related transcriptomic alterations in the PFC are not uniform or clearly transdiagnostic, but may be partly shaped by the underlying psychiatric diagnosis, with a more clearly detectable signal in cases with MDD. This study provides a harmonized, bias-aware framework for integrating heterogeneous post-mortem transcriptomic datasets; however, the results should be interpreted as hypothesis-generating candidate signals rather than as definitive biomarkers, and require independent replication and functional validation.

PMID:42163165 | DOI:10.1186/s12888-026-08170-2

Document this CPD