Analysis, design and "in silico" evaluation of e-selectin antagonists

E-selectin, is member of a family of cell-adhesion proteins, which plays a crucial role in many physiological processes and diseases [1], and in particular, in the early phases of the inflammatory response. Its role is to promote the tethering and the rolling of leukocytes along the endothelial surface [2]. These steps are then followed by integrin-mediated firm adhesion and final transendothelial migration. Therefore, control of the leukocyte-endothelial cell adhesion process may be useful in cases, where excessive recruitment of leukocytes can contribute to acute or chronic diseases such as stroke, reperfusion injury, psoriasis or rheumatoid arthritis [3]. In this work, efforts to develop in silico-based protocols to study the interaction between E-selectin and its ligands, are presented. Hence, different protocols had to be developed and validated. In particular, a new procedure for the analysis of the conformational preferences of E-selectin antagonists was established and the results compared to those obtained with the MC(JBW)/SD approach, which had already demonstrated its validity in the past [161,168]. Thus, the comparison between the two protocols permitted to recognize a different conformational preference of the two methods for the orientation of the sialic acid moiety of sLex (3) (torsions Φ3 and Ψ3, Figure A), which reflects the contrasting opinions existing for the conformation adopted by sLex (3) in solution [150–168]. A more detailed analysis revealed that probably both approaches deliver only a partially correct view and that in reality, in solution, sLex (3) exists as a mixture of low energy conformers and not as supposed to date [150–154,161–163] as a population of a single conformer.
In addition, a docking routine was established and the impact of different partialcharge
methods and of explicit solvation on the binding mode studied.
MD simulations enabled to gain an insight into the dynamical character of the
protein-ligand interactions. In particular, the observations done in an atomic-force
microscopy study [350], describing the interactions between the carboxylic group
of sLex and Arg97, and between the 3– and 4–hydroxyls of fucose and the
calcium ion, as the two main energy barriers for the dissociation process of the
protein-ligand complex, found confirmation in our MD-investigations. Thus, these
two contacts always lasted longer than any other in the MD simulation.
QSAR-models with Quasar [270–272,351] and Raptor [315,316,335] were
successfully derived and will permit a semi-quantitative in silico estimation of the
binding affinity for the ligands that will be designed in the future.
Finally, the developed protocols and models were applied for the development of
new E-selectin antagonists. Unfortunately, to date, only few biological data is
available to evaluate our design strategies. However, the impact of the ligand’s
pre-organization on the binding affinity could be established at least for the Lexcore
of sLex (3). Hence, the importance of the exo-anomeric effect, of the steric
compression, and of the hydrophobic interaction between the methyl group of
fucose and the β-face of galactose was clearly demonstrated.