A cross-tissue transcriptome-wide association study identifies novel candidate genes associated with brain glymphatic system function.

The glymphatic system plays a key role in brain waste clearance, but its genetic regulation remains poorly understood. Diffusion Tensor Image Analysis along the Perivascular Space (DTI-ALPS) index is a non-invasive imaging biomarker to asses glymphatic system activity. We integrated mean DTI-ALPS genome-wide association study (GWAS) data from 31,021 individuals of European ancestry with GTEx v8 multi-tissue eQTL data to perform transcriptome-wide association studies (TWAS) using Unified Test for Molecular Signature (UTMOST) and Functional Summary-based Imputation (FUSION). Gene-level associations were further validated by Multi-marker Analysis of Genomic Annotation (MAGMA). Causal inference was conducted using cis-Mendelian randomization (cis-MR) and summary-data-based Mendelian randomization (SMR), while colocalization was applied to provide evidence of strong associations between two traits within a single genetic region, thereby ensuring the stability of the MR conclusions. TWAS identified 17 candidate genes (AGBL5-IT1, CENPA, CGREF1, DNAJC5G, EMILIN1, GCAT, KHK, MAPRE3, OTOF, PLCL1, PREB, RBM43, RFTN2, SERPIND1, SNAP29, TRIOBP, and UCN), among which six protein-coding genes (TRIOBP, MAPRE3, EMILIN1, KHK, GCAT, and CGREF1) were further validated by MAGMA. Cis-MR provided evidence for the causal effects of these six genes, while colocalization supported that the MR conclusions were stable for four of them (TRIOBP, MAPRE3, EMILIN1, and GCAT). Finally, SMR identified three genes (TRIOBP, GCAT, and MAPRE3) that showed consistent and robust associations with DTI-ALPS across multiple tissues. These findings provide statistical evidence for genetic regulation of glymphatic function.