The conformational propensity of amino acids in the unfolded state of peptides
Corinne Rancurel,Mahvash Khosravi,A. Keith Dunker,David Karlin,Xiaoying Zhang,Mingxin Tang,David Harris,Yingxin Li,Hiroyuki Nakayama,Hongyu Zhang,Remus Berretta,Andrea Eckhart,Walter Koch,Jeffrey Molkentin
2009-01-01
Abstract:and proteins is the subject of ongoing deliberation. Recent research has mostly focused on alanine, owing to its abundance in proteins and its relevance for the understanding of helix coil transitions. We have analyzed the amide I’ band profiles of the IR, isotropic and anisotropic Raman, and VCD profiles of a series of GXG peptides, X representing a subset of the naturally occurring amino acids, in terms of a conformational model which explicitly considers the entire ensemble of possible conformations rather than representative structures. The distribution function utilized for satisfactory simulations of the amide I’ band profiles was found to also reproduce a set of seven J-coupling constants reported by Graf et al. The results of our analysis reveal a PPII fraction of 0.91 for the central alanine residue in GAG, which strongly corroborates the notion that alanine has a very high PPII propensity. We performed a similar analysis for X1⁄4E, F, S, V, K, L and M. Preliminary indication is that E, F, K and L exhibit a substantial PPII propensity, whereas S, V and M exhibit a less pronounced PPII propensity with an increased propensity for b-strand. We also used distributions from coil libraries and MD simulations to model amide I band profiles and J-coupling constants for alanine and valine. We found most of them to be inconsistent with our experimental data. Thus, these results clearly demonstrate that caution must be taken in using coil libraries and MD simulations to describe the unfolded state of peptides and proteins, and that experimental data are a prerequisite for quantitative evaluation of amino acid residues.