Molecular basis underlying functional connectivity of premotor cortex subregions

2403 

Abstract Submission 

Qian Wang¹, Zhang Yu², Jinfeng Hou¹, Shengfeng Liu³, Tianzi Jiang³, Lingzhong Fan³, Luqi Cheng^1,2

¹School of Life and Environmental Sciences, Guilin University of Electronic Technology, Guilin, China, ²Research Center for Augmented Intelligence, Zhejiang Lab, Hangzhou, China, ³Brainnetome Center, Institute of Automation, Chinese Academy of Sciences, Beijing, China

Qian Wang  
School of Life and Environmental Sciences, Guilin University of Electronic Technology
Guilin, China

Zhang Yu  
Research Center for Augmented Intelligence, Zhejiang Lab
Hangzhou, China

Jinfeng Hou  
School of Life and Environmental Sciences, Guilin University of Electronic Technology
Guilin, China

Shengfeng Liu  
Brainnetome Center, Institute of Automation, Chinese Academy of Sciences
Beijing, China

Tianzi Jiang  
Brainnetome Center, Institute of Automation, Chinese Academy of Sciences
Beijing, China

Lingzhong Fan  
Brainnetome Center, Institute of Automation, Chinese Academy of Sciences
Beijing, China

Luqi Cheng  
School of Life and Environmental Sciences, Guilin University of Electronic Technology|Research Center for Augmented Intelligence, Zhejiang Lab
Guilin, China|Hangzhou, China

The premotor cortex (PM) is a cortical region within the frontal agranular cortex[1] that regulates motor and cognitive functions[2]. The functional diversity of the PM implies that it has functionally distinct subregions. Here, we identified the subdivisions of the PM based on structural connectivity and explored the corresponding behavioral domains and genetic association.

40 right-handed healthy adults, including T1w images, diffusion-weighted images and resting-state functional MRI images were randomly selected from the Human Connectome Project (HCP)[3] database. The images had been preprocessed following the HCP's minimal preprocessing pipeline[4]. The ROI of the PM was derived from HCP-MMP1.0[5] including regions of FEF, PEF, 55b, 6d, 6v, 6r, and 6a. Connectivity-based parcellation scheme[6] was used to identify the subregions of the PM. The appropriate cluster number was determined by evaluating the consistency of parcellation between subjects using the silhouette index. Functional decoding was conducted based on the Neurosynth database. Brain gene expression data were obtained from the AHBA dataset[7]. We obtained normalized expression data of 5013 genes for 1298 tissue samples. Cross-sample Pearson correlations between gene expression and rsFC were performed in a gene-wise manner, yielding 5013 correlation coefficients for each subject. Next, genes with significant correlations (using FDR-BH) in more than 90% of subjects were used for functional enrichment analyses with the Metascape portal[8] to characterize the biological functions and diseases of the rsFC-related gene sets.

The PM could be subdivided into six subregions as the silhouette index showed the highest values in the six-cluster solution with the exception of the two-cluster solution (Fig1B). Three clusters were located in the dorsal part and three clusters in the ventral part(Fig1A). Meta-analysis showed that dorsal regions were mainly associated with movement and ventral regions were mainly associated with visual attention and linguistics. Particularly, the C1 of dorsal part was more associated with working memory, the C2 of dorsal part was more associated with visual and spatial motion while the C6 in left ventral part was more associated with language-related functions(Fig1C). For the functional enrichment, expression measures of 618, 2565, 2605, 2455, 1783 genes were associated with rsFC of the C2-C6, respectively, whereas no significant rsFC-related genes were found in the C1. In dorsal part, the rsFC-related genes of the C2 were especially enriched in adrenaline, and the C3 showed extra enrichment in Neuroinflammation and glutamatergic signaling pathway. RsFC-related genes of the ventral C4-C6 were primarily enriched for neurotransmitter pathways, with the C5 more related to hormone response and the C6 more related to muscle contraction(Fig2A). As for diseases, Absence Seizures was significantly enriched in the C2 of the dorsal part while Hypoxia, Subarachnoid Hemorrhage and Febrile Convulsions were enriched in the three ventral regions(Fig2B).

We demonstrated that the human PM can be subdivided into six distinct subregions based on their structural connectivity. Functional enrichment analyses suggested that rsFC of different PM subregions were modulated by their specific associated gene expression profiles. These findings may offer new insights into the molecular basis underlying the functional heterogeneity of the PM.

Genetic Association Studies ²

Multi-Modal Imaging ¹

Other - parcellation, premotor cortex, diffusion weighted imaging, functional connectivity, gene expression, Allen Human Brain Atlas