Inferring the Hoxa1 Gene Regulatory Network in Mouse Embryonic Stem Cells: Time-Series RNA-seq Data and Computational Modeling Approach

Abstract

AbstractThe homeotic geneHoxa1plays a pivotal role in regulating embryonic pattern formation and morphogenesis during mouse embryogenesis. However, despite the identification of a number of putative Hoxa1 target genes, the intricate regulatory relationships between these targets remain largely elusive. Leveraging the advancements in high-throughput technologies and sophisticated computational methods, we aimed to infer the Gene Regulatory Networks (GRNs) governed by Hoxa1 that direct cellular function, morphology, and/or differentiation. To achieve this, we generated time-series RNA-seq data from Retinoic Acid (RA)-treated Wild Type versusHoxa1-null mouse ES cells, enabling the construction of the Hoxa1 GRN. To create this GRN, we employed NARROMI, a published technique known for its noise reduction capabilities and improved accuracy in gene-regulation inference. Using this technique, we identified putative direct and indirect connections between Hoxa1 and a set of genes with known relevance in embryonic development. Validation through qPCR confirmed the Hoxa1-dependence on mRNA expression for selected genes, both within the immediate vicinity (direct) and in secondary interactions (indirect). Furthermore, by mapping the candidate genes to relevant Gene Ontology (GO) networks, we verified their involvement in processes likely regulated by Hoxa1. Our findings provide compelling evidence supporting the accuracy of the NARROMI analysis in generating a hierarchical network of genes under the transcriptional control of Hoxa1 Transcription Factor (TF), specifically in mouse ES cells. This network reveals a pool of promising candidate genes that may function as direct targets of Hoxa1. However, further investigations, including the characterization of Hoxa1 protein interactions with target loci DNA, are necessary to confirm their direct regulatory relationship with this TF. Moreover, the time-series RNA-seq data from Wild Type ES cells, coupled with the methodology employed in this study, hold the potential for constructing GRNs for additional TFs activated by RA. This comprehensive approach can shed further light on the intricate regulatory networks governing cellular function, morphology, and differentiation, advancing our understanding of embryonic development and gene regulation processes.