Abstract:Spectroscopy experiments are crucial to study membrane proteins for which traditional structure determination methods still prove challenging. Double electron-electron resonance (DEER) spectroscopy experiments provide protein residue-pair distance distributions that are indicative of their conformational heterogeneity. Atomistic molecular dynamics (MD) simulations are another tool that have been proven to be vital to study the structural dynamics of membrane proteins such as to identify inward-open, occluded, and outward-open conformations of transporter membrane proteins, among other partially open or closed states of the protein. Yet, studies have reported that there is no direct consensus between the distributional data from DEER experiments and MD simulations, which has challenged validation of structures obtained from long-timescale simulations and using simulations to design experiments. Current coping strategies for comparisons rely on heuristics, such as mapping the nearest matching peaks between two ensembles or biased simulations. Here we examine the differences in residue-pair distance distributions arising due to the choice of membranes around the protein and covalent modification of a pair of residues to nitroxide spin labels in DEER experiments. Through comparing MD simulations of two proteins, PepTSo and LeuT-both of which have been characterized using DEER experiments previously-we show that the proteins' dynamics are similar despite the choice of the detergent micelle as a membrane mimetic in DEER experiments. On the other hand, covalently modified residues show slight local differences in their dynamics and a huge divergence when the oxygen atom pair distances between spin labeled residues are measured rather than protein backbone distances. Given the computational expense associated with pairwise MTSSL labeled MD simulations, we examine the use of biased simulations to explore the conformational dynamics of the spin labels only to reveal that such simulations alter the underlying protein dynamics. Our study identifies the main cause for the mismatch between DEER experiments and MD simulations and will accelerate the development of potential mitigation strategies to improve the match.

NMR‐assisted protein structure prediction with MELDxMD

Structure Determination of Challenging Protein-Peptide Complexes Combining NMR Chemical Shift Data and Molecular Dynamics Simulations

Estimating protein complex model accuracy based on ultrafast shape recognition and deep learning in CASP15.

Automated Fragmentation QM/MM Calculation of NMR Chemical Shifts for Protein-Ligand Complexes.

MELD-Adapt: On-the-Fly Belief Updating in Integrative Molecular Dynamics

Reply To Jensen And Blackledge: Dual Quantifications Of Intrinsically Disordered Proteins By Nmr Ensembles And Molecular Dynamics Simulations

Scaled Alternating Steepest Descent Algorithm Applied for Protein Structure Determination from Nuclear Magnetic Resonance Data.

Protein structure prediction: Combining de novo modeling with sparse experimental data

Protein complex structure modeling by cross-modal alignment between cryo-EM maps and protein sequences

A new method on reconstructing protein structure from NOESY distances

Integrating deep learning, threading alignments, and a multi‐MSA strategy for high‐quality protein monomer and complex structure prediction in CASP15

Structure determination of a flexible cyclic peptide based on NMR and MD simulation 3J-coupling

Accurate and efficient structure elucidation from routine one-dimensional NMR spectra using multitask machine learning

Estimating Residue-Specific Accuracies of Protein Structure Models Using Molecular Dynamics Simulations

Protein secondary structure prediction using NMR chemical shift data

High‐accuracy protein structure prediction in CASP14

A protocol for the refinement of NMR structures using simultaneously pseudocontact shift restraints from multiple lanthanide ions

CS23D: a web server for rapid protein structure generation using NMR chemical shifts and sequence data

Reconciling membrane protein simulations with experimental DEER spectroscopy data

Construct optimization for protein NMR structure analysis using amide hydrogen/deuterium exchange mass spectrometry