(427c) Computational Design of High Affinity Monomeric Streptavidin

Streptavidin binds biotin with Kd ~ 10-14 M, but the binding affinity depends critically on forming the tetrameric structure. As such, mutations that disrupt the native quaternary structure significantly reduce the binding affinity. Although the role of subunit cooperativity has been suggested, the mechanism by which the tetrameric structure contributes to the high binding affinity is not well understood. Similarly, the analysis of high resolution apo and ligand-bound structures of streptavidin does not elucidate the fundamental coupling between its quaternary structure and ligand binding. As a result, it has been difficult to design monomeric streptavidin that binds biotin with a comparable affinity as the tetramer. Engineering high affinity monomeric streptavidin would in the least require that i) we identify the structural elements in the tetramer that are most relevant to biotin binding and ii) introduce mutations within a monomeric streptavidin to reconstruct these properties. To address the first question we performed MD simulations of ligand bound streptavidin tetramer as well as ligand bound streptavidin monomer, and compared the stability of bound biotin in the two structures. This led to a hypothesis that the most significant difference between the monomeric and the tetrameric structures is that the ligand binding pocket is solvent exposed in the monomeric structure. If the functional differences between the monomer and the tetramer are due to differences in the solvent accessibility of the ligand binding pocket in the two conformations, then we should be able to regain much of the binding affinity in monomeric streptavidin by blocking the solvent access to the binding pocket. To test this idea, we designed a monomeric streptavidin in which the first two b strands are elongated by eleven amino acids. The residues on the extended strands were selected primarily to stabilize the new secondary structure and to create a hydrophobic cover over the biotin binding space. Additional mutations were introduced at other locations to stabilize the structure, thus bringing the total number of mutations to 27 out of 121 streptavidin amino acids. The designed sequence was stable by MD and biotin remained bound in the ligand binding pocket significantly longer than for wild type streptavidin. A monomeric streptavidin with high affinity toward biotin can be useful in numerous scientific and medical applications as a molecular biology tool to purify rare proteins from the cell and label targets without causing aggregation, as well as a novel cancer drug delivery vehicle.