Systems biology approach identifies key regulators and the interplay between miRNAs and transcription factors for pathological cardiac hypertrophy

Pathological cardiac hypertrophy (CH) is associated with increased heart failure risk and sudden cardiac death. Several transcription factors (TFs) and miRNAs were implicated in CH, and their high combinatorial action in gene expression regulation is becoming more clear. We adopted a systems biology approach to construct a comprehensive TF-miRNA co-regulatory network in CH and systematically characterize its structure, from node- to systems-level properties. Parallel expression profiles for miRNAs and messenger RNA (mRNA) from an in vitro model of CH were integrated with experimentally validated interactions from seven curated databases to build the CH-related TF-miRNA regulatory network. To leverage this network, we proposed a completely unsupervised approach to identify core regulatory elements, based on Borda count aggregation of distinct network-based properties. By combining node scores derived from motif-based centrality, active pathways, and k-shell index, our approach was able to prioritize biologically meaningful TFs (e.g., MYC, SP1, AKR1B1, EGR1, NFKB1, and ETS1) and miRNAs (e.g., hsa-let-7i-5p, hsa-let-7e-5p, hsa-miR-21a-5p, and hsa-miR-27b-5p) as central nodes in the network and point potential active regulatory pathways in CH. Our findings suggest distinct roles of TFs and miRNAs, which tend to act mostly as network bottlenecks and hubs, respectively. Moreover, we identified feed-forward loop motifs and recurrent associations in the crosstalk between TFs and miRNAs, observing extensive synergistic regulation of common targets and cascade signaling among regulators. Interestingly, most TFs and miRNAs were engaged in specific regulatory roles or interconnection patterns (i.e., master regulator, intermediate regulator, co-regulation of a common target gene, reciprocal regulation, or downstream target) within this network, while few had multiplicity observed in their network function. The constructed CH-related regulatory network has the potential to provide new insights about key regulators, molecular mechanisms, and the interplay between miRNAs and TFs in the pathogenesis of CH, which after proper experimental validation may contribute to the search for new therapeutic approaches.