Structural analysis of fertilinβ cyclic peptide mimics that are ligands for α₆β₁ integrin

Abstract

The NMR structural analysis of two fertilin(beta) mimics cyclo(EC2DC1)YNH2, 1, and cyclo(D2EC2D1C1)YNH2, 2 is described. Both of these mimics are moderate inhibitors of sperm-egg binding with IC50 values of 500 microm in a mouse in vitro fertilization assay. For peptide 1, the optimized conformations that best match the NMR data have a pseudo-type II' beta-turn with the linker and Glu at the i+1 and i+2 positions, respectively. The EC2D1C1 sequence is in a nonclassical (type IV) beta-turn. For peptide 2, the conformation that best matches the NMR data has two turns: a pseudo-type II' beta-turn in the D2EC2D1 sequence followed by a nonclassical beta-turn in the EC2D1C1 sequence. The Cbeta-Cbeta distance between E and D1 in peptide 1 is 9.1 A, in peptide 2, it is 7.7 A. Thus, one possibility for the high IC50 values of these cyclic peptides is that the acidic residues are not constrained to a sufficiently tight turn, and thus much entropy must still be lost upon binding to the alpha6beta1 integrin. This explains why the cyclic peptides are the same as linear peptides at inhibiting sperm-egg binding.

Exploring the conformational space of cyclic peptides by a stochastic search method

“…In such a search, we hope to find the global minimum and the ensemble of lowest energy structures, which is mainly composed of local energy minima or energies that are close to the global minimum. While in reality these molecules achieve their equilibrium conformational ensembles in short time spans [1][2][3][4][5], two main obstacles hinder the prediction in silico of their three-dimensional structures. First, a search for the lowest energy conformers on the potential energy surface is considered to be a non-polynomial (NP)-hard problem for proteins [6] and peptides should not be different.…”