Changes in transcriptional regulation in the temporal lobe in patients with Alzheimer's disease

Background Alzheimer's disease (AD) is a complex neurodegenerative disorder with intricate pathophysiological mechanisms. Transcriptome analysis has been used to investigate the pathogenesis of AD from the perspectives of mRNA expression, alternative splicing, and alternative polyadenylation. However, these 3 transcriptomic regulatory layers have not been comprehensively explored, limiting our understanding of the transcriptomic landscapes of AD pathogenesis. Objective We aimed to describe the transcriptomic landscapes of AD pathogenesis, detect the contributions of different regulatory layers to the total transcriptional variance, and identify diagnostic candidates for AD prediction. Methods We collected RNA sequencing data derived from the temporal lobes of 257 patients with AD and 97 controls, performed joint transcriptional analysis with multi-omics factor analysis (MOFA2) and weighted gene co-expression network analysis (WGCNA), and evaluated the signals with regression models. Results We found that increasing Braak stage is associated with progressive downregulation of SYT1, CHN1, SNAP25, VSNL1, and ENC1 as well as upregulation of TNS1 , SGK1 , CPM , PPFIBP , and CLMN . Subsequent MOFA2 revealed that alternative splicing contributes most ( R 2 = 0.558) to the transcriptional variance between patients with AD and controls followed by alternative polyadenylation ( R 2 = 0.449) and mRNA expression ( R 2 = 0.438). In addition, the regression model constructed with SNAP25 , VSNL1 , and ENC1 expression could distinguish between patients with AD and controls (AUC = 0.752). Conclusions We systematically detailed the transcriptional landscapes in patients with AD and report mRNA signals associated with AD, offering novel insights into AD pathogenesis and therapeutic development.