Efficient Characterization of Protein Cavities within Molecular Simulation Trajectories:trj_cavity

Abstract

Protein cavities and tunnels are critical in determining phenomena such as ligand binding, molecular transport, and enzyme catalysis. Molecular dynamics (MD) simulations enable the exploration of the flexibility and conformational plasticity of protein cavities, extending the information available from static experimental structures relevant to, for example, drug design. Here, we present a new tool (trj_cavity) implemented within the GROMACS ( www.gromacs.org ) framework for the rapid identification and characterization of cavities detected within MD trajectories. trj_cavity is optimized for usability and computational efficiency and is applicable to the time-dependent analysis of any cavity topology, and optional specialized descriptors can be used to characterize, for example, protein channels. Its novel grid-based algorithm performs an efficient neighbor search whose calculation time is linear with system size, and a comparison of performance with other widely used cavity analysis programs reveals an orders-of-magnitude improvement in the computational cost. To demonstrate its potential for revealing novel mechanistic insights, trj_cavity has been used to analyze long-time scale simulation trajectories for three diverse protein cavity systems. This has helped to reveal, respectively, the lipid binding mechanism in the deep hydrophobic cavity of a soluble mite-allergen protein, Der p 2; a means for shuttling carbohydrates between the surface-exposed substrate-binding and catalytic pockets of a multidomain, membrane-proximal pullulanase, PulA; and the structural basis for selectivity in the transmembrane pore of a voltage-gated sodium channel (NavMs), embedded within a lipid bilayer environment. trj_cavity is available for download under an open-source license ( http://sourceforge.net/projects/trjcavity ). A simplified, GROMACS-independent version may also be compiled.

Overcoming the structural versus energy dissipation trade-off in highly crosslinked polymer networks: Ultrahigh strain rate response in polydicyclopentadiene

“…Furthermore, we monitored bond energies during deformation and found the maximum values ($100 kJ/mol) were much lower than typical covalent bond dissociation energies ($350-600 kJ/mol), indicating that the simulated networks are not unrealistically stressed. We quantified nanovoids in the simulated systems using the program trj_cavity [24] with a grid spacing of 1.4 Å and a requirement that voids are surrounded on all sides. Fig.…”

Molecular Properties of Globin Channels and Pores: Role of Cholesterol in Ligand Binding and Movement

“…Globin-like proteins have been identified in bacteria, plants, fungi, and animals (Hardison, 1996, 1998) and contain internal cavities and packing anomalies that appear to reduce thermodynamic stability but may actually provide interior pathways for the diffusion of ligands (Brunori and Gibson, 2001; Teeter, 2004; Brunori et al, 2005; Olson et al, 2007). The pathways and mechanism of movement of ligands within protein cavities and tunnels have been studied by spectroscopy, by crystallography, and by stimulation as well as by mutagenesis mapping experiments (e.g., Tomita et al, 2009; Salter et al, 2012) and have been characterized using molecular dynamics simulations (e.g., Paramo et al, 2014). Studies of myoglobin suggest that thermally or photo-dissociated ligands first migrate into open spaces within the globin interior and then diffuse back to the distal pocket, where they either rebind to the iron or escape from the protein through a gate regulated by motions of His(E7)64 (reviewed in Tomita et al, 2009).…”

Selective inhibitors of trypanosomal uridylyl transferase RET1 establish druggability of RNA post-transcriptional modifications

“…The central carbonyl oxygen may form H-bonds with nearby catalytic triad aspartates Asp473, Asp310, Asp312, and sulfonyl oxygen atoms adjacent to the thiophene ring would interact with Lys505/Lys509/Ser523. Either binding mode predicts a similar interaction, leading to a snug fit of the molecule into the elongated RET1 cavity 29 . The principal reason for this is the fact that the binding site is closed off by the conformation of an N-terminal domain β-loop which protrudes into the active site, stabilized by a salt bridge between Arg358 and Glu657.…”