Multimodal analysis of gene expression from postmortem brains and blood identifies synaptic vesicle trafficking genes to be associated with Parkinson's disease

Xiaoya Gao, Zifeng Huang, Cailing Feng, Chaohao Guan, Ruidong Li, Haiting Xie, Jian Chen, Mingchun Li, Rongfang Que, Bin Deng, Peihua Cao, Mengyan Li, Jianjun Lu, Yihong Huang, Minzi Li, Weihong Yang, Xiaohua Yang, Chunyan Wen, Xiaomei Liang, Qin YangYin Xia Chao, Ling Ling Chan, Midori A. Yenari, Kunlin Jin, K. Ray Chaudhuri, Jing Zhang, Eng King Tan, Qing Wang

Research output: Contribution to journalArticlepeer-review

Abstract

Objective: We aimed to identify key susceptibility gene targets in multiple datasets generated from postmortem brains and blood of Parkinson's disease (PD) patients and healthy controls (HC). Methods: We performed a multitiered analysis to integrate the gene expression data using multiple-gene chips from 244 human postmortem tissues. We identified hub node genes in the highly PD-related consensus module by constructing protein-protein interaction (PPI) networks. Next, we validated the top four interacting genes in 238 subjects (90 sporadic PD, 125 HC and 23 Parkinson's Plus Syndrome (PPS)). Utilizing multinomial logistic regression analysis (MLRA) and receiver operating characteristic (ROC), we analyzed the risk factors and diagnostic power for discriminating PD from HC and PPS. Results: We identified 1333 genes that were significantly different between PD and HCs based on seven microarray datasets. The identified MEturquoise module is related to synaptic vesicle trafficking (SVT) dysfunction in PD (P < 0.05), and PPI analysis revealed that SVT genes PPP2CA, SYNJ1, NSF and PPP3CB were the top four hub node genes in MEturquoise (P < 0.001). The levels of these four genes in PD postmortem brains were lower than those in HC brains. We found lower blood levels of PPP2CA, SYNJ1 and NSF in PD compared with HC, and lower SYNJ1 in PD compared with PPS (P < 0.05). SYNJ1, negatively correlated to PD severity, displayed an excellent power to discriminating PD from HC and PPS. Conclusions: This study highlights that SVT genes, especially SYNJ1, may be promising markers in discriminating PD from HCs and PPS.

Original languageEnglish
Article numberbbaa244
JournalBriefings in bioinformatics
Volume22
Issue number4
DOIs
StatePublished - 1 Jul 2021

Keywords

  • PPP2CA
  • Parkinson's disease
  • Parkinson-plus syndromes
  • SYNJ1
  • synaptic vesicle trafficking
  • weighted correlation network analysis

Fingerprint

Dive into the research topics of 'Multimodal analysis of gene expression from postmortem brains and blood identifies synaptic vesicle trafficking genes to be associated with Parkinson's disease'. Together they form a unique fingerprint.

Cite this