Multiscale design of coarse-grained elastic network-based potentials for the μ opioid receptor
- 359 Downloads
Despite progress in computer modeling, most biological processes are still out of reach when using all-atom (AA) models. Coarse-grained (CG) models allow classical molecular dynamics (MD) simulations to be accelerated. Although simplification of spatial resolution at different levels is often investigated, simplification of the CG potential in itself has been less common. CG potentials are often similar to AA potentials. In this work, we consider the design and reliability of purely mechanical CG models of the μ opioid receptor (μOR), a G protein-coupled receptor (GPCR). In this sense, CG force fields (FF) consist of a set of holonomic constraints guided by an elastic network model (ENM). Even though ENMs are used widely to perform normal mode analysis (NMA), they are not often implemented as a single FF in the context of MD simulations. In this work, various ENM-like potentials were investigated by varying their force constant schemes and connectivity patterns. A method was established to systematically parameterize ENM-like potentials at different spatial resolutions by using AA data. To do so, new descriptors were introduced. The choice of conformation descriptors that also include flexibility information is important for a reliable parameterization of ENMs with different degrees of sensitivity. Hence, ENM-like potentials, with specific parameters, can be sufficient to accurately reproduce AA MD simulations of μOR at highly coarse-grained resolutions. Therefore, the essence of the flexibility properties of μOR can be captured with simple models at different CG spatial resolutions, opening the way to mechanical approaches to understanding GPCR functions.
KeywordsGPCR Molecular dynamics Coarse-graining Multiscale modeling Elastic network models Graph theory
M.F., L.L., and D.P.V. acknowledge Doctor François Meurant, anesthetist at the Centre Hospitalier Tubize-Nivelles (Nivelles, Belgium) for fruitful discussions. All authors acknowledge the Swedish National Infrastructure for Computing (SNIC, Sweden), the Consortium des Equipements de Calcul Intensif (C.E.C.I., supported by the F.R.S.-FNRS, Belgium), and the Plateforme Technologique de Calcul Intensif (P.T.C.I., supported by the F.R.S.-FNRS, Belgium) for computing resources. M.F. thanks the Belgian National Fund for Research (F.R.S.-FNRS) for his F.R.I.A. doctoral scholarship. M.F., L.L., and D.P.V. also thank the Interuniversity Attraction Poles Programmes n° 7/05: ‘Functional supramolecular systems’ initiated by the Belgian Science Policy Office for partial financial support. A.L. wishes to thank the Swedish Science Council (VR) for financial support.
- 9.Dror RO, Young C, Shaw DE (2011) Anton: a special-purpose molecular simulation machine. In: Padua D (ed) Encyclopedia of parallel computing. Springer, New York, pp 60–71Google Scholar
- 11.Dror RO, Green HF, Valant C, Borhani DW, Valcourt JR, Pan AC, Arlow DH, Canals M, Lane JR, Rahmani R, Baell JB, Sexton PM, Christopoulos A, Shaw DE (2013) Structural basis for modulation of a G-protein couples receptor by allosteric drugs. Nature 503:295–299Google Scholar
- 37.Cutolo P, Basdevant N, Bernadat G, Bachelerie F, Ha-Duong T (2016) Interaction of chemokine receptor CXCR4 in monomeric and dimeric state with its endogenous ligands CXCL12: coarse-grained simulations identify differences. J Biomol Struct Dyn (in press). doi: 10.1080/07391102.2016.1145142
- 44.Gorecki A, Szypowski M, Dlugosz M, Trylska J (2009) RedMD—reduced molecular dynamics package. J Comput Chem 30:2364–2373Google Scholar
- 50.Kmiecik S, Gront D, Kolinski M, Wieteska L, Dawid AE, Kolinski A (2016) Coarse-grained protein models and their applications. Chem Rev 116:7898–7936. doi: 10.1021/acs.chemrev.6b00163
- 57.Development Core Team R (2008) R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna. ISBN 3-900051-07-0Google Scholar