Improved Protein Model in SPICA Force Field

Yamada, Teppei; Miyazaki , Yusuke; Harada, Shogo; Kumar, Ashutosh; Vanni , Stefano; Shinoda, Wataru

doi:10.1021/acs.jctc.3c01016

Back

Journal article

+ 1 other files

Improved Protein Model in SPICA Force Field

DOKPE

Yamada, Teppei Okayama University, Japan
Miyazaki , Yusuke Okayama University, Japan
Harada, Shogo Nagoya University, Japan
Kumar, Ashutosh University of Fribourg
Vanni , Stefano Department of Biology
Shinoda, Wataru Okayama University, Japan

2023

Published in:

Journal of Chemical Theory and Computation. - American Chemical Society (ACS). - 2023, vol. 19, no. 23, p. 8967-8977

Intrinsically disordered protein

English The previous version of a SPICA coarse-grained (CG) force field (FF) protein model focused primarily on membrane proteins and successfully reproduced the dimerization free energies of several transmembrane helices and stable structures of various membrane protein assemblies. However, that model had limited accuracy when applied to other proteins, such as intrinsically disordered proteins (IDPs) and peripheral proteins, because the dimensions of the IDPs in an aqueous solution were too compact, and protein binding on the lipid membrane surface was over-stabilized. To improve the accuracy of the SPICA FF model for the simulation of such systems, in this study, we introduced protein secondary structure-dependent nonbonded interaction parameters to the backbone segments and re-optimized almost all nonbonded parameters for amino acids. The improved FF proposed here successfully reproduced the radius of gyration of various IDPs, the binding sensitivity of several peripheral membrane proteins, and the dimerization free energies of several transmembrane helices, including the alanine-rich peptide WALP23. The new model also showed improved agreement with the experiments on the free energy of peptide association in water. In addition, an extensive library of nonbonded interactions between proteins and lipids, including various glycerophospholipids, sphingolipids, and cholesterol, allowed the study of specific interactions between lipids and peripheral and transmembrane proteins. Hence, the new SPICA FF (version 2) proposed herein is applicable with high accuracy for simulating a wide range of protein systems.

Faculty

Faculté des sciences et de médecine

Department

Département de Biologie

Language