首页 | 本学科首页   官方微博 | 高级检索  
     


Molecular dynamics simulations of the chemokine CCL2 in complex with pull down-derived heparan sulfate hexasaccharides
Authors:Sophie Winkler  Rupert Derler  Bernd Gesslbauer  Elmar Krieger  Andreas J. Kungl
Affiliation:1. Institute of Pharmaceutical Sciences, University of Graz, Schubertstraße 1/1, 8010 Graz, Austria;2. Antagonis Biotherapeutics GmbH, Strasserhofweg 77a, 8045 Graz, Austria;3. YASARA Biosciences GmbH & CMBI, Wagramer Strasse 25/3/45, 1220 Vienna, Austria
Abstract:

Background

Binding of chemokines to glycosaminoglycans (GAGs) is a crucial step in leukocyte recruitment to inflamed tissues.

Methods

A disaccharide compositional analysis of the HS dp6 fraction in combination with MS analysis of the CCL2-depleted dp6 fraction was the basis for target GAG ligand structure suggestions. Four experimentally-derived heparan sulfate hexasaccharides, two potentially chemokine-specific and two unspecific, have been docked to CCL2. Subsequent 300?ns molecular dynamics simulations were used to improve the docked complexes.

Results

Hexasaccharides with four sulfations and no acetylations are suggested for selective and high affinity chemokine binding. Using the Antithromin-III/heparin complex as positive control for docking, we were able to recover the correct complex structure only if the previously liganded ATIII structure was used as input. Since the liganded structure is not known for a CCL2-GAG complex, we investigated if molecular dynamics simulations could improve initial docking results. We found that all four GAG oligosaccharides ended up in close contact with the known binding residues after about 100?ns simulation time.

Conclusions

A discrimination of specific vs. unspecific CCL2 GAG ligands is not possible by this approach. Long-time molecular dynamics simulations are, however, well suited to capture the delicate enthalpy/entropy balance of GAG binding and improve results obtained from docking.

General significance

With the comparison of two methods, MS-based ligand identification and molecular modelling, we have shown the current limitations of our molecular understanding of complex ligand binding which is could be due to the numerical inaccessibility of ligand-induced protein conformational changes.
Keywords:Corresponding author at: Institute of Pharmaceutical Sciences   University of Graz   Schubertstraße 1/1   8010 Graz   Austria.
本文献已被 ScienceDirect 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号