%0 Generic %A Alvarez Canales, Gilberto Alejandro %C Heidelberg %D 2025 %F heidok:36747 %R 10.11588/heidok.00036747 %T Developmental constraints and the regulatory logic of Drosophila enhancers %U https://archiv.ub.uni-heidelberg.de/volltextserver/36747/ %X Summary Enhancers can generate specific patterns of gene expression, which are essential for organismal development. Despite decades of research, how enhancers encode the information necessary to drive these precise patterns remains unclear. To address this question, in this project, I focused on the interplay of two different approaches: 1) Using well-known endogenous enhancers, I searched for putative missing regulatory elements by integrating information from different mutational screenings, and 2) Using the combination of synthetic enhancers expression profiles and thermodynamic models, I tested the possibility of several mechanisms of their associated Transcription Factors (TFs). In the first approach, I focused on two of the best-characterized animal enhancers, the minimal stripe 2 enhancer (MS2E) and the E3N enhancer. The MS2E enhancer can generate the second of the seven stripes of the Even-skipped gene expression pattern. The E3N enhancer regulates the expression of the Shavenbaby gene, which has an eight-striped pattern. For the MS2E enhancer, a systematic and targeted mutagenesis screening was performed and analyzed, while for the E3N enhancer, I analyzed data from a randomized mutagenesis approach. For both enhancers, I estimated and associated affinity changes of multiple TF binding sites (TFBSs) with the observed expression phe- notypes. A dense encoded architecture was observed for the two enhancers where almost each mutated section generates a phenotype that differs from the wild-type. Affinity changes and motif turnover analyses of known TFBSs could explain only a small fraction of these phenotypes. Finally, I explored experimental alterna- tives for finding additional associated TFs and the evolutionary implications of a dense encoded enhancer architecture. These results suggest the need to find ad- ditional regulators and improve current versions of the binding profiles of already known TFs. In the second approach, I evaluated the roles of early patterning TFs through synthetic enhancers and mechanistic modeling. I observed that early embryonic enhancers are associated with higher information at the sequence level composition than later and synthetic random enhancers. Additionally, the syn- thetic system could encode different sharp gene expression patterns with different combinations of known TFBSs. The results from implementing generalized ther- modynamic models suggested that these TFs are highly context-dependent. This analysis suggested multiple equally performing mechanisms for different possibili- ties for TF-TF interactions, TF function, and modes of regulation. These models could predict observed expression patterns, such as a broad stripe in the center of the embryo, but other expression patterns, such as the presence of additional anterior expression, could not be explained. Finally, I propose alternative mecha- nisms by which known TFs work in this context and new directions for enhancer design to expand the understanding of encoding developmental patterns.