Abstract Details

Poster 26: Molecular Dynamics Simulation of Bicyclic Peptides

Xian Jin1, Julian Schwartz1, Prof. Dr. Jean-Louis Reymond1
1Department of Chemistry and Biochemistry, University of Bern
In rational drug design, the application of peptides that perturb protein-ligand interactions is an attractive avenue to target disease[1]. Peptides are particularly well suited for this task since they are readily accessible by solid-phase peptide synthesis (SPPS)[2]. Despite the activity and receptor selectivity of naturally occurring bioactive peptides as drug, they often show different
disadvantages for practical application in medicine, such as short half-life in vivo and lack of bioavailability[3]. These limitations can be solved by rigidifying the active sequence of their structure in a defined conformation. Cyclisation is one way to reduce the conformational space to achieve high structure rigidity.

Herein we focused on the almost pristine cyclic peptide topological area, bicyclic topologies. In the previous work, we reported the synthesis of bicyclic ‘‘non-zero bridged’’ homodetic (in which amino acids are connected by amide bonds only) peptides prepared by SPPS using an orthogonal protection scheme[2]. At the same time, 3D structures of the bicyclic peptides were investigated by molecular dynamics (MD) simulations with NMR data restraint using AMBER 10 package[2]. The analysis of MD simulation trajectory indicated the conformationally rigidity of the molecular, especially the backbone. On the other hand, from the analysis of molecular shapes[4], these “non-zero bridged” bicyclic peptides have globular shapes that similar to short segments of folded alpha-helices. Therefore, bicyclic peptides with such conformation properties can be well achieved the desired activity and selectivity in molecular interactions with protein targets. Bicyclic peptides can define a very large structural family. For instance, there are 25 possible bicyclic peptide scaffolds from 5 to 13 amino acids and with 20 different amino acids considered at the non-bridgehead positions, it can form 3.3 trillion possible products. It can widely expend the chemical space for drug-like small molecules. Thus, based on the previous work, we extended the way of structure modeling of bicyclic peptides for establishing the bicyclic peptide 3D structure library to investigate this novel area of peptide topological space. This library will be used for virtual screening of the search of bioactive conformers in further research.

As the main idea of building bicyclic peptide conformation library, the automatic generation of reasonable 3D structure without previous information of the peptide, like NMR data, is the crucial task. MD simulation with optimized forcefield parameters was the well suited tool to predict these “unknown” 3D structures. For obtaining the most reasonable conformations, we adapted several different MD strategies to simulate the peptide structures and validated these selected conformations by comparing to the NMR data through calculating root-mean-square deviation (RMSD). The comparison result demonstrated that the MD simulation in the explicit solvent offered the most reasonable conformation and characterization of hydrogen bonds, relaxation behavior and intramolecular distances of bicyclic peptides.

[1] K. H. Bleicher, H. J. Bohm, K. Muller and A. I. Alanine, Nat. Rev. Drug Discovery, 2003, 2, 369–378.
[2] M. Bartoloni, RU. U. Kadam, J. Schwartz, J. Furrer, T. Darbre, J.-L. Reymond, Chem. Commun., 2011, 47, 12634-12636.
[3] V. Marx, Chem. Eng. News, 2005, 83, 17-24.
[4] W. H. Sauer and M. K. Schwarz, J. Chem. Inf. Comput. Sci.,2003, 43, 987–1003.

Return to Programme