Unraveling Solvent-Driven Equilibria between a- and 310-Helices through an Integrated Spin Labeling and Computational Approach

In this work we present an effective and flexible computational approach, which is the result of an ongoing development in our groups, allowing the complete a priori simulation of the ESR spectra of complex systems in solution. The usefulness and reliability of the method are demonstrated on the very demanding playground represented by the tuning of the equilibrium between 3 10- and α-helices of polypeptides by different solvents. The starting point is the good agreement between computed and X-ray diffraction structures for the 310-helix adopted by the double spin-labelled heptapeptide Fmoc-(Aib-Aib-TOAC)2-Aib-OMe. Next, density functional computations, including dispersion interactions and bulk solvent effects, suggest another energy minimum corresponding to an α-helix in polar solvents, which, eventually, becomes the most stable structure. Computation of magnetic and diffusion tensors provides the basic ingredients for the building of complete spectra by methods rooted in the Stochastic Liouville Equation (SLE). The remarkable agreement between computed and experimental spectra at different temperatures allowed us to identify helical structures in the various solvents. The generality of the computational strategy and its implementation in effective and user-friendly computer codes pave the route toward systematic applications in the field of biomolecules and other complex systems. © 2007 American Chemical Society.