Impact of the Conformational Variability of Oligopeptides on the Computational Prediction of Their CD Spectra

Although successful in the structural determination of ordered biomolecules, the spectroscopic investigation of oligopeptides in solution is hindered by their complex and rapidly changing conformational ensemble. The measured circular dichroism (CD) spectrum of an oligopeptide is an ensemble average over all microstates, severely limiting its interpretation, in contrast to ordered biomolecules. Spectral deconvolution methods to estimate the secondary structure contributions in the ensemble are still mostly based on databases of larger ordered proteins. Here, we establish how the interpretation of CD spectra of oligopeptides can be enhanced by the ability to compute the same observable from a set of atomic coordinates. Focusing on two representative oligopeptides featuring a known propensity toward an α-helical and β-hairpin motif, respectively, we compare and cross-validate the structural information coming from deconvolution of the experimental CD spectra, sequence-based de novo structure prediction, and molecular dynamics simulations based on enhanced sampling methods. We find that small conformational variations can give rise to significant changes in the CD signals. While for the simpler conformational landscape of the α-helical peptide de novo structure prediction can already give reasonable agreement with the experiment, an extended ensemble of conformers needs to be considered for the β-hairpin sequence.