Resting-state EEG alpha-BOLD coupling spatially follows cortical transcriptomic gradients of cell types and receptors

Abstract

The coupling between electroencephalography (EEG) and blood-oxygen-level-dependent (BOLD) signals has been investigated across numerous studies, but its neurobiological underpinnings remain poorly understood. Resting-state EEG alpha-BOLD coupling follows a characteristic spatial pattern, shifting from negative correlations in sensory regions to positive correlations in association cortices. Here we compared the alpha-BOLD coupling map to 82 cortical feature maps, including transcriptomic markers of cell types and receptor subunits and structural MRI measures. Three maps were significantly associated (q < 0.05, FDR-corrected), namely a layer 6 VIP-expressing interneuron marker (In3), an excitatory layer-5 marker, and NMDA receptor subunit GRIN2C. Combined in a multiple linear regression model, they explained R2 = 0.312 of the rank-transformed spatial variance, of which dominance analysis attributed 56% to the In3 marker alone. Because many cortical maps vary along a shared sensory-to-association hierarchy, we tested whether these associations are specific to the markers by controlling for the T1/T2 ratio and the principal functional gradient. The In3 marker retained roughly half of its explained rank variance beyond these hierarchy proxies, more than the other two. Based on correlations between static group-level maps, this study identifies all three markers as concrete candidates for future computational and experimental studies of alpha-BOLD coupling.