Regional structural-functional connectivity coupling in major depressive disorder is associated with neurotransmitter and genetic profiles

Abstract

BackgroundAbnormalities in structural–functional connectivity (SC–FC) coupling have been identified globally in patients with major depressive disorder (MDD). However, investigations have neglected the variability and hierarchical distribution of these abnormalities across different brain regions. Furthermore, the biological mechanisms underlying regional SC–FC coupling patterns are not well understood. MethodsWe enrolled 182 patients with MDD and 157 healthy control (HC) subjects, quantifying the intergroup differences in regional SC–FC coupling. The extreme gradient boosting (XGBoost), support vector machines (SVM) and random forest (RF) models were constructed to assess the potential of SC–FC coupling as biomarkers for MDD diagnosis and symptom prediction. Then, we examined the link between changes in regional SC–FC coupling in patients with MDD, neurotransmitter distributions, and gene expression. ResultsWe observed increased regional SC–FC coupling in default mode network (T = 3.233) and decreased coupling in frontoparietal network (T = −3.471) in MDD relative to HC. XGBoost (AUC = 0.853), SVM (AUC = 0.832) and RF (p < 0.05) models exhibited good prediction performance. The alterations in regional SC–FC coupling in patients with MDD were correlated with the distributions of four neurotransmitters (p < 0.05) and expression maps of specific genes. These genes were strongly enriched in genes implicated in excitatory neurons, inhibitory neurons, cellular metabolism, synapse function, and immune signaling. These findings were replicated on two brain atlases. ConclusionsThis work enhances our understanding of MDD and pave the way for the development of additional targeted therapeutic interventions.