Network analysis of transcriptomic data uncovers molecular signatures and the interplay of mRNAs, lncRNAs, and miRNAs in human embryonic stem cells

Arindam Ghosh,Anup Som
DOI: https://doi.org/10.1016/j.diff.2023.11.001
IF: 3.533
Differentiation
Abstract:Growing evidence has shown that besides the protein coding genes, the non-coding elements of the genome are indispensable for maintaining the property of self-renewal in human embryonic stem cells and in cell fate determination. However, the regulatory mechanisms and the landscape of interactions between the coding and non-coding elements is poorly understood. In this work, we used weighted gene co-expression network analysis (WGCNA) on transcriptomic data retrieved from RNA-seq and small RNA-seq experiments and reconstructed the core human pluripotency network (called PluriMLMiNet) consisting of 375 mRNA, 57 lncRNA and 207 miRNAs. Furthermore, we derived networks specific to the naïve and primed states of human pluripotency (called NaiveMLMiNet and PrimedMLMiNet respectively) that revealed a set of molecular markers (RPS6KA1, ZYG11A, ZNF695, ZNF273, and NLRP2 for naive state, and RAB34, TMEM178B, PTPRZ1, USP44, KIF1A and LRRN1 for primed state) which can be used to distinguish the pluripotent state from the non-pluripotent state and also to identify the intra-pluripotency states (i.e., naïve and primed state). The lncRNA DANT1 was found to be a crucial as it formed a bridge between the naive and primed state-specific networks. Analysis of the genes neighbouring DANT1 suggested its possible role as a competing endogenous RNA (ceRNA) for the induction and maintenance of human pluripotency. This was computationally validated by predicting the missing DANT1-miRNA interactions to complete the ceRNA circuit. Here we first report that DANT1 might harbour binding sites for miRNAs hsa-miR-30c-2-3p, hsa-miR-210-3p and hsa-let-7b-5p which may influence pluripotency.
What problem does this paper attempt to address?