Improving Martini 3 for Disordered and Multidomain Proteins

Abstract

Coarse-grained molecular dynamics simulations are a useful tool to determine conformational ensembles of proteins. Here, we show that the coarse-grained force field Martini 3 underestimates the global dimensions of intrinsically disordered proteins (IDPs) and multidomain proteins when compared with small-angle X-ray scattering (SAXS) data and that increasing the strength of protein−water interactions favors more expanded conformations. We find that increasing the strength of interactions between protein and water by ca. 10% results in improved agreement with the SAXS data for IDPs and multidomain proteins. We also show that this correction results in a more accurate description of self-association of IDPs and folded proteins and better agreement with paramagnetic relaxation enhancement data for most IDPs. While simulations with this revised force field still show deviations to experiments for some systems, our results suggest that it is overall a substantial improvement for coarsegrained simulations of soluble proteins.

Extended conformational states dominate the Hsp90 chaperone dynamics

“…We next probed how these large-scale conformational changes are reflected in the full-length Hsp90 dimer. To this end, we performed cgMD metadynamics simulations (see Experimental procedures ) on the dimeric yeast Hsp90 with and without the linker region using the MARTINI3 force field ( 30 ), with modified protein–water interaction that better reproduces experimental data ( 31 , 32 ). Moreover, we used AlphaFold2 ( 33 ) to assess the stability of secondary and tertiary structural elements ( via the per-residue confidence score, pLDDT) and the relative flexibility between different regions and domains ( via the predicted aligned error, PAE).…”

“…The simulations were performed at T = 310 K in an NPT ensemble with a 10 fs timestep using Gromacs ( 36 ) with PLUMED2 ( 38 , 39 ). The protein–water interaction was increased by 6% to provide a better agreement with experimental data ( 31 , 32 ). To conserve the secondary and tertiary protein structure, an elastic network model was introduced between residues with a high per-residue confidence score (pLDDT > 90) in AlphaFold2 ( 33 ).…”

Effective Molecular Dynamics from Neural-Network Based Structure Prediction Models

“…The latest MARTINI3 version 29 has expanded available bead types, resulting in more accurate and detailed representations of functional groups, although both former and current models still struggle to properly balance the challenging protein-protein and proteinsolvent interactions. In this regard, increased protein-water (P-W) interactions can improve the agreement with experimental data [35][36][37][38][39] . Due to the introduced approximations, the MARTINI model also encounters difficulties in preserving secondary and tertiary structure elements of proteins, which is why it is often combined with ENMs [23][24][25] or Gō-type models 40,41 in which the harmonic bonds are substituted by Lennard-Jones interactions.…”

Enhanced sampling for lipid-protein interactions during membrane dynamics

“…Recently, new efforts have been made to improve computational modeling of intrinsically disordered proteins (Thomasen et al, 2022;Klein et al, 2021;Tran and Kitao, 2020). Several reviews have also been dedicated to the role of intrinsic disorder during lipid-protein interactions highlighting the effects on membrane curvature (Has, Sivadas and Das, 2022;Cornish et al, 2020;Fakhree, Blum and Claessens, 2019;Snead and Stachowiak, 2018).…”