Dynamic brain states and molecular signatures in primary angle-closure glaucoma.

Abstract

Background: Primary angle-closure glaucoma (PACG) has traditionally been regarded as an ocular disorder, but accumulating evidence suggests broader central nervous system involvement. Although previous neuroimaging studies have identified static functional abnormalities, the dynamic properties of large-scale brain networks and their associated molecular signatures in PACG remain insufficiently understood.

Methods: We applied Leading Eigenvector Dynamics Analysis to resting-state functional MRI data from 44 patients with PACG and 57 healthy controls to characterize recurrent whole-brain dynamic states. State-specific temporal metrics and spatial patterns were further evaluated using multiple machine learning models. To explore potential biological correlates, imaging-derived spatial patterns were linked to cortical gene expression profiles from the Allen Human Brain Atlas using partial least squares regression, followed by pathway enrichment, cell-type enrichment, and neurotransmitter receptor/transporter mapping analyses.

Results: Compared with healthy controls, PACG patients showed prolonged dwell time in one recurrent dynamic state, suggesting reduced flexibility of large-scale brain dynamics. Machine learning models showed promising classification performance within the current dataset, with the most informative features primarily located in default mode network regions. Transcriptomic decoding revealed enrichment of genes related to synaptic signaling, ion channel activity, neurotransmitter transport, and neuronal communication. Cell-type enrichment analyses further implicated excitatory neurons, inhibitory neurons, and astrocytes. In addition, a significant spatial association with VMAT2 suggested that monoaminergic systems may be relevant to the observed imaging phenotype.

Conclusion: PACG is associated with altered large-scale brain dynamics, particularly involving default mode network-related state instability. These imaging abnormalities show spatial associations with molecular, cellular, and neurotransmitter-related signatures.