Transcription Factor Co-expression Networks of Adipose RNA-Seq Data Reveal Regulatory Mechanisms of Obesity

Background: Transcription Factors (TFs) control actuation of genes in the genome and are key mediators of complex processes such as obesity. Master Regulators (MRs) are the genes at the top of a regulation hierarchy which regulate other genes. Objective: To elucidate clusters of highly co-expressed TFs (modules), involved pathways, highly interconnected TFs (hub-TFs) and MRs leading to obesity and leanness, using porcine model for human obesity. Methods: We identified 817 expressed TFs in RNA-Sequencing dataset representing extreme degrees of obesity (DO; lean, obese). We built a single Weighted Transcription Factor Co-expression Network (WTFCN) and TF sub-networks (based on the DO). Hub-TFs and MRs (using iRegulon) were identified in biologically relevant WTFCNs modules. Results: Single WTFCN detected the Red module significantly associated with DO (P < 0.03). This module was enriched for regulation processes in the immune system, e.g.: Immune system process (Padj = 2.50E-06) and metabolic lifestyle disorders, e.g. Circadian rhythm - mammal pathway (Padj = 2.33E-11). Detected MR, hub-TF SPII was involved in obesity, immunity and osteoporosis. Within the obese sub-network, the Red module suggested possible associations with immunity, e.g. TGF-beta signaling pathway (Padj = 1.73E-02) and osteoporosis, e.g. Osteoclast differentiation (Padj = 1.94E02). Within the lean sub-network, the Magenta module displayed associations with type 2 diabetes, obesity and osteoporosis e.g. Notch signaling pathway (Padj = 2.40E-03), osteoporosis e.g. hub-TF VDR (a prime candidate gene for osteoporosis). Conclusion: Our results provide insights into the regulatory network of TFs and biologically relevant hub TFs in obesity.