Transcriptional Regulation Analysis in a Neurotoxin-induced Apoptosis of Human Neuroblastoma SH-EP Cells with a State Space Model

Understanding the molecular mechanisms of Parkinson's disease (PD) is essential to development of therapeutic strategies. Despite many studies of the pathogenesis of PD, its exact mechanism is still unknown. Information on gene regulation in PD might provide an insight to PD pathogenesis. Time course gene expression data have been used to predict gene regulatory mechanisms in dynamic biological processes, such as development, drug response and the cell cycle. 1-Methyl-4-phenylpyridinium (MPP+), a dopaminergic neurotoxin, produces in vivo and in vitro cellular changes characteristic of PD that include cytotoxicity, which result in apoptosis. In this study, a time series microarray experiment was performed for MPP+ treated human neuroblastoma SH-EP cells. Prior to estimation of regulation structure, 997 MPP+ response genes were identified by the Extraction of Differential Gene Expression program. These MPP+ response genes were assigned to eight transcriptional modules including M-1 +/-, M-2 +/-, M-3 +/-, and M-4 +/- by a state space model and gene ontology analysis identified significantly enrich terms related to apoptosis signaling pathway in the three modules including M1+, M1- and M3+. The regulation networks of MPP+ response genes were estimated using the auto-regressive form of the state space model. In the networks, four hub genes including CHAC1, MTHFD2, SH2D5 and LOC100134537 were identified. These hub genes showed direct or indirect positive feedback connection with genes, such as to AEN encoding apoptosis enhancing nuclease and ATF6 encoding a transcription factor that activates target genes for the unfolded protein response during ER stress. This network might provide an insight for interactions of mitochondrial dysfunction, endoplasmic reticulum stress and apoptosis in MPP+-induced model of PD.