UC Riverside

Molecular recognition processes guide the preferential association of molecular species in chemical and biological systems through physical differentiation and complementarity. Recognition exists at many scales, including the long-range interactions dictating diffusional association, as well as the short-range intermolecular interactions that discriminate the selectivity of a substrate to a host molecule and dictate the configurational transition to the bound state. In this work, methods are developed and employed to study three situations in molecular association processes.

Diffusional association was investigated in the intermediate transfer between enzymes in engineered spatially organized nanostructures. In particular, the glucose oxidase-horseradish peroxidase enzyme pair was investigated in several geometric arrangements of DNA-origami scaffolds. A computationally efficient Brownian dynamics software package was developed for the determination of substrate association probability over microsecond to millisecond time scales. We found that arrangements of enzymes on a planar scaffold primarily gain efficiency from induced enzyme colocalization with moderate enhancement due to the scaffold acting as a diffusive barrier. Confinement of the enzyme system within a nanotube scaffold greatly enhanced substrate transfer, up to ten fold relative to colocalization, and up to 150 fold relative to a disorganized solution of the enzymes and substrate over the same time period. Our results have implications for the efficient engineering of synthetic enzyme cascades.

A computational method was developed for the determination of final non-covalent binding pathways for molecular complexes. Normal mode calculations were used to model coordinated natural motions of a host-guest complex, which were then utilized to connect conformational minima to form non-covalent binding pathways. Our results demonstrate that conformational transitions can be modeled and extended to construct coordinated final binding events. This approach has advantages over simulation-based methods for studying systems with slow binding processes and can help design molecules with preferred binding kinetics.

A procedure for determining receptor subtype specificity for inhibitors of the human proteasome was performed. A natural product derivative inhibitor was simulated with molecular dynamics, and molecular recognition was assessed through structural stability and energetic affinity with each receptor subtype. The determined energetic and structural preference for two of the three catalytic sites suggests potential for desirable activity as a human proteasome inhibitor.