A unifying model for helices in proteins.

摘要

Helices are the most common secondary structural element in proteins. This one simple structural motif has been used to explain all types of protein structural hypotheses. Yet, critical analysis of helices in protein crystallographic structures indicates that common interpretation of the helical model needs to be updated to reflect high resolution data and modern physical models. Data from 45,661 protein crystallography experiments were analyzed and found to support a "molten helix" interpretation. Thus, high resolution experimental data is unified with the concept of a statistical mechanical peptide chain that samples conformational space from a single smooth potential well located between the classical alpha-helical and 310-helical coordinates on the potential energy landscape. Shared hydrogen bonds provide a mechanism to ease the transition between classical i → i+3 and i → i+4 hydrogen bonds endpoints, and are an essential feature of the "molten helix" model.;Detailed analysis of crystallographic and theoretical methods support a unified "molten helix" model. Statistical distributions of hydrogen bonds helical conformations are presented to obviate sources of experimental error and inaccurate modeling assumptions. These observations suggest improvements in crystallographic refinement procedures, protein folding models, protein-protein interface constraints, peptidomimetic design, and more.;Crystallographic structures represent a single low-energy "snapshot" of a protein conformation, but proteins are truly dynamic. Molecular dynamics simulations employing molecular mechanics force fields are state-of-the-art technology for simulating dynamic protein behavior. Simulations are analyzed for helix and hydrogen bond content and compared to crystallographic data.;From analysis of molten helices, an empirical "molten helix" potential was derived. Weighted distributions of helical backbone conformations may be enumerated with all possible sidechain configurations for all possible amino acids. This library of helical configurations was screened, in silico, for shape similarity with a complementary protein cleft. Thus, a computational algorithm is presented for selecting an optimal sequence of helical sidechains for binding a cognate protein cleft.