A neuron-distributed DEK integrates ac4C modification and neuroinflammation in pathogenesis of Parkinson disease: evidence from Mendelian randomization, multi-omics and in vitro validation
TL;DR
This study establishes a novel AN-associated molecular patterns in PD, identifying DEK as a computationally identified neuron-specific factor associated with ac4C modification and neuroinflammatory cascades for PD patients.
Abstract
Background Epi-transcriptomic modifications, particularly N4-acetylcytidine (ac4C), and chronic neuroinflammation have emerged as pivotal players in the pathogenesis of Parkinson’s disease (PD). However, the specific molecular co-expression patterns linking ac4C RNA modification to neuronal inflammatory responses remains largely uncharted. Objective This study aimed to decode the ac4C-neuroinflammation (AN)-associated molecular patterns in PD and to identify a neuron-specific central pathogenic and therapeutic factor. Methods We integrated multi-omics analyses by using peripheral blood bulk transcriptomes (GSE18838, GSE49126, GSE22491, GSE6613, and GSE57475) and GWAS data from PD patients for identification of AN-related risk genes. Next, consensus clustering and 3 machine learning algorithms (LASSO, RF, and SVM-RFE) were applied for patient stratification, hub gene identification, and diagnostic modeling. Single-cell transcriptomic profiling of PD patients (GSE140231) was leveraged to map the cellular distribution and mechanistic roles of the hub gene within the substantia nigra (SN). An AI-driven active learning framework and the CTD database were utilized to screen therapeutic candidates targeting the hub gene, with binding affinities validated via molecular docking. In vitro experiments finally validated the expression of hub gene. Results We pinpointed 7 AN-associated risk DEGs for PD patients, including HSP90AA1, DEK, LEF1, IRF2, BCL2L1, CFL1, and BCR, which effectively stratified PD patients into 2 distinct immune-molecular subgroups. DEK can be considered as neuron-distributed and up-regulated AN-associated central pathogenic factor for PD patients. The DrugReflector active learning framework identified BRD-K57589644 as a computationally prioritized compound warranting further investigation. Conclusion This study establishes a novel AN-associated molecular patterns in PD, identifying DEK as a computationally identified neuron-specific factor associated with ac4C modification and neuroinflammatory cascades for PD patients.