BPS2026 – Deconstructing the complexity: Understanding ATP binding with de novo design and component parts

Adenosine triphosphate is the major source of energy in the cell and a common signaling and regulatory molecule. Proteins that bind this molecule in nature do so with an incredibly high level of selectivity over the four other nucleotide triphosphates that exist. How this selectivity is achieved has been the subject of several studies, and while though, successful, the inherent complexity of proteins has made it difficult to extract the core fundamental biophysics that allow for a faithful reproduction of this function in a non-natural fold or a new context. In our lab, we are using de novo design to better study these fundamental biophysics in a simplified context. In addition to that we are also reducing the complexity of the ligand. ATP has three chemically distinct domains that require different amino acids to interact with. We believe that by minimizing the complexity of both the protein and the ligand we can better extract the fundamental biophysics underlying binding and selectivity. To that end, we have established a 4-helix bundle that can selectively bind Adenine over Guanine, and a beta barrel protein that can bind ATP. Our modeling indicates that buried hydrogen bonds are the main driver of affinity, and that bulk solvent molecules disrupt these interactions. We believe that our work will lead to more selective and physics-based design rules for engineering ATP binding and catalytic proteins.