High-Resolution Adsorbed and Solution-State Ensembles of a Naturally Evolved Biomineralization Protein: Blind and NMR-Guided Predictions

A high-resolution multi-scale modeling technique, RosettaSurface, is developed to study the solution and adsorbed state ensembles of a naturally evolved biomineralization protein. Metropolis Monte-Carlo simulations accounting for all torsional and rigid-body protein degrees of freedom, starting from a fully extended chain in solution, are applied to the statherin hydroxyapatite (HAp) system. Results show minimal structural deviations between solution and adsorbed states of statherin, and that HAp consistently selects for lattice matching conformers of statherin that are similar to its solution state energy minima. HAp shows specificity for the N-terminal binding domain of statherin; however, when all three equilibrium crystal faces of HAp (001, 010, and 100) are present, statherin exhibits promiscuous binding behavior that may be indicative of the geometric and chemical similarities of the three faces. Results agree well with 15 high-resolution NMR distance and angle measurements, although two long-range intra-molecular distance measurements consistently disagree with simulation. Adding all NMR distance and angle constraints to the simulation in the form of a harmonic potential produce results that suggest these two long-range distance measurements may in fact be inter-molecular, rather than intra-molecular as initially interpreted. Results also show promise for a combined NMR RosettaSurface protocol that may prove effective where protein structure prediction at interfaces is concerned.