Construction and Design of ‚-Sheets

Introduction De novo protein design and architecture both focus on the construction and design of three-dimensional structures. Although these disciplines work on vastly different scales, they nevertheless share two requirements: a structural design and an understanding of the physical properties which govern the stability of that structure. Our knowledge of physics and engineering has allowed architects to devise magnificent buildings that can be stably constructed to serve the intended purpose. Protein designers, on the other hand, have a greater challenge realizing their intended structures because accurately predicting a protein’s stability is not yet possible. Although it is well-documented that a protein’s folded threedimensional structure is encoded by its amino acid sequence, currently that folded structure cannot be predicted from sequence information alone. Therefore, the studies of protein stability, protein secondary structure, and de novo protein design are intimately interconnected. Stability studies provide insight for the design of proteins that will fold into predetermined structures and perform specified functions. Protein design, on the other hand, provides an opportunity to test our grasp of the rules that underlie protein structure and stability. Understanding â-sheet formation is the key to a host of problems and applications involving protein folding and design. For example, the formation of a â-hairpin has a profound effect on reducing the conformational space and defining the long-range interactions for a folding protein. Although the characterization and de novo design of R-helical structures have dominated the field in the past, interest in â-sheet stability and design has intensified for several reasons. Recent studies have emphasized that there are many proteins in which â-sheets play functionally important roles. â-Sheets can provide the key element in protein-DNA,1 protein-RNA,2 and protein-protein recognition.3 Several of these interactions are based upon direct, edge-on â-sheet contacts, which can often be mimicked by peptides, for example, the dimerization of HIV protease4 and P pilin binding to the PapD chaperone.5 Even the behavior of the hormone erythropoetin can be mimicked by disulfide-linked â-hairpin peptides.6 Aggregated protein fibrils exhibiting predominantly â-structure have been implicated in amyloid diseases.7 Recently, several groups have begun to quantify the energetics of the interactions that stabilize â-structure in simple model systems and to formulate guidelines which will allow the structure and stability of â-sheets to be manipulated in a rational fashion.