Back

Assessing molecular, cellular and transcriptomic bases of laminar perfusion and cytoarchitecture coupling in the human cortex

Guo, F.; Zhao, C.; Bhatt, R. R.; Liu, Z.; Kim, A. J.; Yang, Z.; Xu, S.; Jann, K.; Mather, M.; Jahanshad, N.; Wang, D. J.

2025-11-05 neuroscience
10.1101/2025.11.03.686411 bioRxiv
Show abstract

Understanding how cellular architecture organizes cortical function requires mesoscopic approaches that resolve structure-function coupling in vivo. Here we introduce cerebral blood flow (CBF) and cell-body staining intensity (CSI) similarity index (CCSI), a localized similarity index between the CBF estimates from ultra-high-field 7T arterial spin labeling images and CSI from the BigBrain histology images to serve as a quantitative marker of laminar perfusion-cytoarchitecture coupling. CCSI revealed a reproducible, region- specific alignment between laminar vascular and cellular profiles across the cortex. Going beyond CBF, CCSI selectively tracked mitochondrial respiratory capacity and colocalized with capillary endothelial and mature non-myelination oligodendrocyte populations forming neurovascular interfaces. Transcriptomic enrichment highlighted pathways related to vascular remodeling, oxidative metabolism, and lipid-myelin homeostasis, indicating that CCSI reflects integrated metabolic-structural specialization. At the systems level, CCSI strengthened structure-function gradient correspondence in transmodal cortices, such as the default mode network. Together, these findings establish CCSI as a physiologically grounded, non-invasive marker of perfusion-cytoarchitecture alignment, providing a cross-scale framework linking cortical microstructure, metabolism, and functional organization.

Published in Nature Communications (predicted rank #1) · training set

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.