Combining computational structure prediction techniques and experimental biophysical validation, we aim to develop a tool to design proteins to recognize and bind with a particular orientation to inorganic surfaces. Detailed knowledge of interfacial biophysics is paramount to novel technologies such as biocompatible materials and molecular electronics, but progress is hindered by a lack of experimental tools able to resolve atomic detail at the protein-surface interface. Using a design approach partially circumvents this difficulty as it allows us to investigate how changes in protein composition can affect such interactions. Here we begin to develop design principles by exploring a model system, the synthetic protein TOP7 and redesigned variants adsorbing to muscovite mica. Recently, we adapted the Monte Carlo structure prediction algorithm Rosetta to model protein interactions with inorganic materials. The new algorithm combines rigid backbone docking, sequence space searches, and backbone minimization in an iterative dock-design-refine process. Design decoys are sorted by relative energies which include orientation dependent hydrogen bonding, a modified Lennard-Jones 6-12 potential, implicit solvation, rotamer probability, and Coulomb electrostatics with a distance-dependent dielectric constant. Mutants will be expressed, purified, and preliminarily screened for relative affinity using atomic force microscopy. Promising designs will then be tested with sum frequency generation spectroscopy (a nonlinear spectroscopy capable of probing molecules at interfaces) and hydrogen-exchange mass-spectrometry to probe stability, structure, and orientation on the surface.

Extended Abstract Status: Not Uploaded