An R Shiny App for Integrating Sequencing Datasets in the Analysis of Toxicant Defects Within Osteogenic Differentiation
Abstract
A class of small, non-coding RNAs, microRNAs (miRNAs), may play a pivotal role in the development of the neural crest (NC)- and mesoderm-derived skeleton. They are involved in a variety of biological processes that contribute to skeletal development, such as cell migration, Wnt signaling and specification of cellular fate towards osteogenesis. Identifying miRNAs involved with regulating genes that drive the two lineages that give rise to osteoblasts during development can provide distinct gene regulatory mechanisms that give rise to craniofacial deformities. Potential microRNA-to-gene target interaction (MTIs) pairs can then potentially be used as biomarkers for in utero diagnostic or treatments for skeletal deformities in the face or limbs. The aim of this study was to develop a set of miRNA biomarkers that correlate with NC and mesoderm cell differentiation into osteoblasts, and to identify miRNAs specific to these pathways. Our study was conducted in vitro using human embryonic stem cells (hESCs) that were differentiated into osteoblasts. Toxicants known to repress NC and mesoderm induction were supplied to the in vitro cultures and the RNA sequenced to obtain both small RNAs and mRNAs. To facilitate this analysis, an R Shiny web app was developed to implement the sequencing results of both microRNA and RNA. The app provides tools to separately display and visualize the results of the differential regulation of microRNAs and mRNAs. Gene ontology (GO) analysis revealed that the identified miRNA-target interactions were indeed involved in select GO pathways related to osteogenic, NC and mesoderm differentiation. The app narrowed down significantly correlated MTIs specific to each differentiation lineage. Within the NC lineage miR-30a, miR-148a and miR-195 were correlated with cadherin2 (CDH2), DNA methyltransferase 3 beta (DNMT3B) and transcription factor AP-2 alpha (TFAP2A), respectively. Within the mesoderm lineage miR-148a, miR-7 and miR-20a were correlated with DNMT3B, JARID2 and VANGL1, respectively. In the future, these miRNAs could be further validated as potential biomarkers in the context of neurocristopathies.