Deciphering the landscape of lncRNA-driven ceRNA network in schizophrenia etiology

Abstract

Abstract Background The unifying hypothesis of competing endogenous RNA (ceRNA) wherein crosstalk between coding (mRNAs) and long non-coding RNAs (lncRNAs) via microRNA (miRNA) response elements, creates a pervasive regulatory network across the transcriptome, has been implicated in complex disorders including schizophrenia. Even with a wide range of high-throughput data, the etiology of schizophrenia remains elusive, necessitating a more holistic understanding of the altered genetic landscape, shifting focus from solely candidate gene studies and protein-coding variants. Objective We developed lncRNA-associated ceRNA networks to elucidate global molecular/regulatory signatures underlying schizophrenia using diverse data in the public domain. Methods Microarray dataset associated with peripheral blood mononuclear cells (PBMCs) of schizophrenia and control patients was used to identify differentially expressed mRNAs. Weighted gene co-expression network analysis (WGCNA) was used to identify highly correlated hubs, and genes from these overlapping Kyoto Encyclopedia of Genes and Genomes (KEGG) and gene ontology (GO) term genesets were considered key mRNA players. StarBase, Human MicroRNA Disease Database, and miRWalk were used to derive mRNA-miRNA and miRNA-lncRNA relationships. Finally, the key mRNAs, interacting lncRNAs and miRNAs were chosen to reconstruct sub-ceRNA networks based on network centrality scores. Results Bioinformatics analysis revealed the involvement of three differentially expressed mRNAs, namely ADRA1A, HAP1 and HOMER3 in the schizophrenia ceRNA networks with lncRNAs NEAT1, XIST, and KCNQ1OT1 modulating their activity by a suggestive sequestering of miR-3163, miR-214-3p and miR-2467-3p, respectively. Conclusions Furthermore, based on contextual evidence, we propose how ceRNAs could orchestrate crosstalk between neurostructural dynamics and immune/inflammatory processes and enable unifying these disparate models of schizophrenia etiology.