Abstract
Protein binding and folding stability are important properties, not only for fundamental reasons but also for biological functions. Mutations on proteins, even at a single residue level, can significantly affect binding specificity and folding stability. A major goal of this dissertation work is to model contributions, especially those from electrostatic interactions, to binding and folding stability. The Poisson-Boltzmann (PB) equation has long been held as the benchmark for modeling electrostatic interactions in biomolecules. However, the computational cost of solving the PB equation has restricted its applications largely to single-conformation calculations. We have developed a highly efficient method based on the generalized Born (GB) model that is able to reproduce PB results with higher accuracy than current GB methods. However, even our GB method is inadequate to model small effects due to point mutations on binding and folding stability. To meet this challenge, we have designed a simple scaling scheme based on the PB results for a small number of conformations out of a large ensemble. We show that the scaled GB method is able to reproduce PB results extremely well for the whole ensemble in a variety of applications, including binding, folding, and transfer free energy. The scaling method makes it possible to thoroughly sample the transient-complex ensemble in predicting protein binding rate constants and to incorporate conformational sampling in electrostatic modeling without loss of accuracy. The body of work presented in this dissertation may serve as both tools and inspirations for research in protein design and drug design.
|
