Non-coding Genomic Determinants of Adipogenesis

Project Summary Cell identity is acquired by integrating multiple signaling inputs that establish complex gene regulatory networks over time. These signals are converted into expression profiles by the binding of transcription factors to non-coding regions of DNA known as enhancers. In human adipogenesis, multiple inputs converge onto a single gene, PPARG. PPARG responds to these inputs through two "super-enhancers", which are dense collections of individual enhancers. The myriad of input signals that feed into adipogenesis and drive PPARG activation suggests the presence of complex regulatory logic within the two super-enhancers located up- and downstream of the PPARG coding region. For instance, Creb1 is an early input in adipogenesis that appears dispensable later, C/EBPb is required later and subsequently activates C/EBPa to maintain PPARG transcription, and finally PPARG itself autoregulates to drive PPARG output even further, as adipocytes adopt their terminal fate. The large genomic distances between enhancers and promoter, and the observation that chromatin remodeling is a key bottleneck in adipogenesis suggesta hierarchical model of PPARγ activation. We will systematically determine which functional units of the PPARG super-enhancers respond to diverse inputs by silencing individual super-enhancer modules in a temporally resolved manner during adipogenesis. We will combine these findings with direct observation and manipulation of chromatin contacts in single cells, in order to test whether efficient signal transduction from super-enhancers to promoter requires 2-way enhancer-promoter chromatin loops or complex multi-way contacts between super-enhancer modules and the promoter.