Determination of the basis of ligand binding via engineering and crystallography

Fred Hutchinson Cancer Research Center, Seattle WA

Investigators

Linked publications & trials

Abstract

? DESCRIPTION (provided by applicant): A significant problem in molecular biology is our inability to accurately calculate, model and predict protein-ligand binding affinities, even when provided a high-resolution structure of the actual complex. This problem is made evident across a wide body of experimentation and literature, including (1) the poor performance of algorithms used to calculate binding affinities from structures, (2) disagreement on the physical basis for high affinity ligand binding for exceptionally well-studied proteins such as streptavidin, and (3) difficulties associated with engineering novel ligand-binding proteins. Attempts to understand the basis for tight, specific ligand binding by dissecting naturally evolved ligand binding protein, while informative, have not produced the ability to accurately predict binding affinities from structures, or to directly compute the structure of novel ligand binding proteins. Engineered proteins offer a possible advantage as alternative systems for the examination of ligand binding mechanisms. While the energetic and structural parameters that are used to create them may be inaccurate, those terms are nonetheless precisely defined during the engineering process and can be systematically altered during an iterative design project. In addition, protein engineering allows investigators to create large numbers of designed proteins against many precisely defined ligands, and then to identify the most interesting and informative constructs for detailed structural and physical analyses. The Specific Aims of this project are: (1) To determine common structural and mechanistic features of successfully designed ligand binding proteins, and to compare those results against designs that display unexpected patterns of ligand binding affinities and specificities. (2) To assess whether constraints that are placed on the design of ligand binding proteins behave as intended. In particular, this study will examine whether two fundamental components of ligand- protein designs (shape complementarity, which attempts to enforce specificity, and structural 'pre-ordering' of the binding site, which attempts to reduce entropic penalties) actually impair ligand-binding function. We believe that completion of these aims will provide both an immediate impact (by improving protein design methods) and a longer-term impact (by further elucidating rules that govern the behavior of ligand binding proteins in general).