Identifying the hub gene and immune infiltration of Parkinson's disease using bioinformatical methods

Parkinson's disease (PD) is a common neurodegenerative disorder that affects 1%-2% of the population over 60 years old. Immune response dysfunction in the brain contributes to the occurrence and development of PD. This study aimed to uncover the potential diagnostic genes for PD and characterize the immune cell infiltrates.

We downloaded the microarray data of patients with PD samples from the Gene Expression Omnibus (GEO) database. Weighted Gene Co-Expression Network Analysis (WGCNA) was used to identify the modules linked to PD in the GSE20163 dataset. Meanwhile, differentially expressed genes (DEGs) between the healthy control samples and PD samples were also identified. Then the PD-related genes were integrated based on the genes in the key module and DEGs. Functional enrichment analysis was used to explore the molecular mechanisms of these PD-related genes. Protein-protein interaction (PPI) network and least absolute shrinkage and selection operator (LASSO) analysis were used to further screen candidate genes for PD. Gene set enrichment analysis (GSEA) was applied to explore the biological functions of these candidate genes. The infiltration of immune cells was detected by single-sample gene set enrichment analysis (ssGSEA) algorithm in the GSE20163 dataset, and Pearson analysis was used to investigate the correlation of candidate genes with immune cells and immune checkpoint proteins. The expression of candidate genes in clinical samples was verified by qPCR.

Altogether, we found a unique gene module related to PD, where 109 DEGs were identified in the GSE20163 dataset. Following these results, we screened 68 genes associated with PD. Gene Expression Omnibus (GEO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses suggested that these genes were markedly enriched in the pathway of synthesis and transport of neurons. Three candidate genes (SLC18A2, CALB1, and SYNGR3) were further identified in PD patients through PPI network and LASSO analysis. The receiver operating characteristic (ROC) curve indicated that the three candidate genes had a good performance in distinguishing the PD samples from healthy control samples. The proportions of the aDC, DC, NK CD56dim cells, and follicular helper T cells (TFH) were obviously different between the healthy control and PD samples. Moreover, CTLA4, LAG3, CEACAM1, and CD27 were highly expressed in the PD group. GSEA analysis for candidate genes revealed that they were all closely related to the neurogenic disease. Additionally, the three candidate genes were all strongly correlated with the above immune cells and immune checkpoint proteins. The qPCR results validated the expression differences of SLC18A2 and SYNGR3 in the clinical PD and control samples.

The three candidate genes may be a useful tool for diagnosing PD patients. These findings provide a reference for exploring new therapeutic targets and strategies for PD treatment.