Biochemical Pathway And Protein-protein Interaction Network Based Analyses Of Genes Associated With Parkinson's Disease

Objective:Parkinson's disease(PD), one of the most common neurodegenerative disorders, not only affects quality of life of a large proportion of the elder population and their caregivers, but also causes heavy economic burden on society. Although there have been lots of studies devoting to deciphering the pathogenesis of this disease, little has been known for this complex neurological disorder. Analyzing biological function enrichment and protein-protein interaction network(PPIN), current study performed a systematic analysis on a human gene set associated with PD, referred to as PDgset, identified by genetic studies, to identify major biological pathways in the etiology and development of Parkinson's disease, constructed and analyzed the corresponding PPIN related to PDgset, which laid great foundation for understanding the molecular mechanisms underlying Parkinson's disease. Meanwhile, since there is a typical co-morbidity relation between Parkinson's disease and Alzheimer's disease(AD), we also used biological function and network based method to compare genes genetically associated with these two neurological phenotypes, in order to explore the specific interrelation between these two disorders at the molecular level. Methods:1. Data source Querying for reports related to PD and AD with specific retrieval terms and reviewing all those related reports, Parkinson's disease related genes gene set(PDgset) and Alzheimer's disease related genes gene set(Alzgset) were curated separately.2. Functional enrichment analysis of PDgset Via WebGestalt software, GO Biological Process(BP) terms enriched in PDgset were identified and then the major biological processes involved in PD were uncovered.3. Biological pathway and pathway crosstalk analysis The significantly enriched biological pathways in PDgset were detected by ToppGene software, in turn observing the crosstalks among significant enrichment pathways, and then visualizing pathway crosstalk results using Cytoscape.4. PPIN analysis In the context of human PPIN, we calculated the global network topological metric, i.e., degree, associated with PDgset, to evaluate the corresponding topological characteristics.5. PD-specific protein network inference The distinctive protein network was inferred via node-weighted Steiner minimal tree algorithm and its unrandomness was evaluated. Also, several PD-related potential genes were predicted via the inferred network.6. Biological function and PPIN based comparison between PDgset and Alzgset Analyzing biological function and PPIN topology, we compared three gene datasets, i.e., PDgset, Alzgset and Common gene set for PDgset and Alzgset(PAgset) with 78 gene members, to unravel molecular relation of PD and AD.Results:1. Datasets source Screening all related genetic association studies, we curated PDgset with 242 genes and Alzgset consisting of 430 related genes.2. GO enrichment analysis Lots of GO biological processes related to drug response and metabolism, neurodevelopment or synaptic transmission and immune response were enriched among PDgset, such as response to nicotine, xenobiotic metabolic process, dopamine uptake involved in synaptic transmission, regulation of neuronal synaptic plasticity, positive regulation of interleukin-6 production.3. Biological pathway and pathway crosstalk analysis Those PDgset significantly enriched pathways, were mainly related to neurotransmitter, immune system and metabolism, e.g., dopaminergic synapse signaling, serotonergic synapse signaling, tyrosine metabolism pathway, IL-10 anti-inflammatory signaling and signal transduction through IL1 R. In addition, pathways related to estrogen signaling and adipocytokine signaling were also enriched in the candidate genes, which indicated the potential links between PD and both two signaling pathways. What's more, pathway crosstalk analysis could suggest that there are two main interlinked modules: one mainly consisted of neuronal signaling related pathways as well as the metabolic pathways of neurotransmitters or drugs, the other primarily dominated by immune system-related pathways.4. PPIN analysis and PD-specific protein network inference Indicated by the results, the degree distribution of PDgset was apparently right-skewed in the context of human protein network. Compared to cancer genes, degree distributions and average degrees of PDgset and nicotine addiction related genes(NAGenes) tended to be closer, obviously distinctive from those metrics of cancer genes, indicating apparent distinctions between neurological disorders and cancer, at least from the aspect of network. Through constructing PD-specific protein network, we also identified several promising PD-related genes, e.g., SLC9A3R2, YWHAB and TYROBP.5. Biological function and PPIN based comparison between PDgset and Alzgset Comparing to PDgset, We found shared GO terms and pathways of PDgset and Alzgset enriched GO BP terms and pathways ranked higher in Alzgset; via network metrics comparison, compared with PDgset, the metrics of PAgset appeared to be more closer with those of Alzgset. In sum, we could conclude that at least from molecular aspect, PAgset tended to be closer to Alzgset, suggesting shared parts of both diseases are more likely to approach AD. Conclusions:1. The molecular pathogenetic mechanisms of PD are quite complex. Current study indicated that, there have been some alterations in the biological processes related to synaptic transmission and related metabolism, immune system, resulting in degeneration and death of dopaminergic neurons.2. Compared to cancer genes, PDgset and NAGenes possessed fewer protein connections, reflecting distinctive differences between neuronal disorders and cancer.3. SLC9A3R2, YWHAB and TYROBP might act vitally in the pathogenesis and development of PD, and thus they are the potential targets for the following researches related to PD.4. The results from our work, at the molecular level, implicated shared parts of both diseases and AD were closer to each other, a valuable hint for subsequent researches, and the framework proposed in our current work can be used to infer pathological molecular mechanisms related to a complex disease.