Integrative transcriptomic- neuroimaging analysis reveals polygenic correlates of interhemispheric functional decoupling in ischemic stroke
TL;DR
The findings provide a molecular framework for understanding post-stroke neural network reorganization and offer a link between functional neuroimaging phenotypes and gene expression, though all associations remain correlational and require mechanistic validation.
Abstract
Background Functional magnetic resonance imaging (fMRI) has revealed abnormal brain activity patterns in stroke patients, yet the genetic correlates underlying functional homotopy – defined as synchronized spontaneous activity between bilateral homologous brain regions – remain poorly characterized. This study investigates the genetic basis of voxel-mirrored homotopic connectivity (VMHC) abnormalities in stroke patients. Methods We analyzed resting-state fMRI data from 50 stroke patients and 50 healthy controls (HC) to quantify VMHC. Spatial transcriptome-neuroimaging correlations were established using the Allen Human Brain Atlas (AHBA) to identify VMHC-associated genes. Transcriptomic analyses combined pathway-centric functional annotation (DAVID) with protein-protein interaction (PPI) network modeling (STRING v12.0). Results Stroke patients exhibited significantly reduced VMHC in the rectus gyrus, superior temporal gyrus, middle occipital gyrus, cuneus, and right calcarine/left posterior cingulate gyrus (p < 0.05, GRF-corrected). VMHC alterations correlated positively and negatively with 1,198 genes each. Transcriptomic profiling revealed significant enrichment in synaptic vesicle trafficking, mitochondrial energy metabolism, and neuroinflammation-related pathways. PPI mapping uncovered multi-tiered networks with hub genes including BRCA1, CDK9, ACTB, and ATP6V1A/F involved in transcriptional regulation, cytoskeletal dynamics, and vesicular acidification. Conclusion This multimodal integration study elucidates polygenic correlates of post-stroke VMHC abnormalities, demonstrating that interhemispheric coordination may depend on synergistic interactions among functionally diverse gene clusters. Our findings provide a molecular framework for understanding post-stroke neural network reorganization and offer a link between functional neuroimaging phenotypes and gene expression, though all associations remain correlational and require mechanistic validation.