Abstract
There is a close relationship between Parkinson's disease (PD) and circadian rhythm disturbances. To investigate genes linked to circadian rhythm disruption in PD, we analyzed the PD dataset via bioinformatic techniques. We downloaded 2 datasets from the GEO database; one dataset (GSE20163) was used as an internal training set, whereas the other dataset (GSE20164) was used as an external set. We first performed differential gene expression analysis on the GSE20163 dataset to identify differentially expressed genes (DEGs). We then conducted weighted gene co-expression network analysis on this dataset to select the most significant modules (turquoise modules) and took the intersection of the two to obtain crossover genes (85 genes), which were functionally enriched in the Disease Ontology (DO), Gene Ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG). These 85 crossover genes were again taken to intersect with circadian rhythm-related genes (2531) to form circadian rhythm-related genes (16). Box plot presentation was performed and found that all p-values were significant. Least absolute shrinkage and selection operator (LASSO) regression analysis was carried out to find 2 representative genes (CDK5, BSCL2). ROC curve analysis was conducted in both the cohort and the validation set, demonstrating statistically significant in each group. Gene set enrichment analysis (GSEA) was conducted on these 2 genes in the internal training set. Analysis of these two genes revealed that they are involved in the regulation of biological processes and synaptic signaling. The diagnostic and predictive model for Parkinson's disease (PD), which relies on circadian-related genes (BSCL2 and CDK5), has excellent diagnostic and predictive performance. This model could serve as a valuable device in the diagnosis of circadian disturbance in PD. Finally, immunohistochemistry and immunoblotting-related experiments were performed for further confirmation. Immunoblot analysis revealed decreased protein levels of BSCL2 and CDK5 in the PD group. Immunohistochemistry further demonstrated reduced TH-positive neuron counts in this experimental cohort. Taking these findings together, BSCL2 and CDK5 are important genes associated with circadian rhythm disturbance in PD.