Abstract
Suicidality is the most severe sign of major depressive disorder (MDD). However, network-wide structural dynamics alterations and transcriptional patterns in MDD patients with suicidality remain unclear. A total of 124 participants were recruited, including 31 MDD patients without suicidal ideation (MDDNSI), 34 MDD patients with suicidal ideation (MDDSI), 22 MDD patients with suicidal behavior (MDDSB), and 37 healthy controls (HCs). All participants completed diffusion tensor imaging scans and suicidal assessment. Network-based statistics (NBSs) were used to compare differences in structural connectivity, while average network controllability (ANC) reflecting brain state dynamics was measured based on suicidality-based circuits. Weighted gene coexpression network analysis (WGCNA) identified gene modules associated with neuroimaging phenotypes. Gene functions and cellular types in MDD patients were determined with enrichment analysis and cell-type-specific expression analysis. The results showed that in the suicidal ideation (SI) circuit, MDDSI patients represented decreased ANC compared to MDDNSI patients, with upregulated genes related to neuronal bundles and downregulated genes related to ribosomes. In the suicidal behavior (SB) circuit, MDDSB patients exhibited reduced ANC compared to MDDNSB patients, with upregulated genes associated with cytoplasmic pathways and positive regulation of signaling and downregulated genes related to synaptic signal transmission. Importantly, genes in both the SI circuit and SB circuit were enriched in gamma-aminobutyric acid (GABA) neurons. The study highlights that the differential ANC of the SI network and the SB network exerts a distinctive influence on brain state trajectories in MDD patients and identifies the specific transcriptional pattern and cell types associated with brain regions related to SI or SB in depression.