Normal mode-guided transition pathway generation in proteins

Abstract

The biological function of proteins is closely related to its structural motion. For instance, structurally misfolded proteins do not function properly. Although we are able to experimentally obtain structural information on proteins, it is still challenging to capture their dynamics, such as transition processes. Therefore, we need a simulation method to predict the transition pathways of a protein in order to understand and study large functional deformations. Here, we present a new simulation method called normal mode-guided elastic network interpolation (NGENI) that performs normal modes analysis iteratively to predict transition pathways of proteins. To be more specific, NGENI obtains displacement vectors that determine intermediate structures by interpolating the distance between two end-point conformations, similar to a morphing method called elastic network interpolation. However, the displacement vector is regarded as a linear combination of the normal mode vectors of each intermediate structure, in order to enhance the physical sense of the proposed pathways. As a result, we can generate more reasonable transition pathways geometrically and thermodynamically. By using not only all normal modes, but also in part using only the lowest normal modes, NGENI can still generate reasonable pathways for large deformations in proteins. This study shows that global protein transitions are dominated by collective motion, which means that a few lowest normal modes play an important role in this process. NGENI has considerable merit in terms of computational cost because it is possible to generate transition pathways by partial degrees of freedom, while conventional methods are not capable of this.

Unraveling the Compositional and Molecular Features Involved in Lysozyme-Benzothiazole Derivative Interactions

“…Afterward, we examine how the BTS-ligand could affect the communication efficiency in the inter-residue network of the lysozyme receptor in the unbound and bound state (i.e., in the absence and presence of the BTS-ligand). To this end, we carried out an elastic network approach (ENM models) by regarding the local perturbations induced by the BTS-ligand in the C(α)-residue network that forms the lysozyme binding site [ 44 , 45 , 46 ]. In this instance, the local perturbations are propagated like an allosteric signal pathway across the communication network composed by C(α)-atom nodes which represent the residues and the network edges representing their mutual contacts connected by the two associated C(α)-residue nodes within a distance of 5.0 Å between C-alpha atoms in the lysozyme receptor structure.…”

“…In this sense, the local perturbations are defined by the collective anisotropic fluctuations generated from the intra-segment C α -C α atomic residue distances (labelled by the color intensity in the 2D-perturbation matrix) by the presence of a given ligand (i.e., the best-crystallographic binding pose of the BTS-ligand) which could affect the function and conformational properties of the binding residues of lysozyme. In this context, both the unbound and bound state are fully defined by the local perturbation response 3 N × 3 N Hessian matrix ( H i,j ) which describes the inter-residue allosteric communication pathway in the lysozyme [ 44 , 45 , 46 ]. Its N -elements (1 ≤ i, j ≤ N ) are the second derivatives of the anisotropic network potential (V lysozyme or V lysozyme+ligand ) which describes the (3 N )- XYZ -displacements ( ) of all the ( i , j )-residue pairs sensors ( i ) and effectors ( j ) in N -blocks of three dimensions each, and where γ R = γ L = 0.5 is an interaction constant, in the absence (lysozyme unbound state), and upon interactions (or perturbations as lysozyme plus BTS in the bound state) following the Equations (8) and (9): …”

Coarse-Grained Protein Dynamics Studies Using Elastic Network Models

“…In the examples above, to reproduce transitions between two structures, we used one structure as the reference and the other as the initial condition. This is of course a crude simplification, and smoother morphing methods have also been introduced [53,54,55,56,57]. Multiscale hybrid methods (e.g., all-atom MD and ANM-NMA [102]) have been developed, and combination with ENM-MD is possible.…”

“…The nonlinear effect was formerly discussed in the context of transitions between structural states, sometimes involving partial unfolding (“proteinquake”) [52]. Attempts were made to connect (interpolate) multiple structures by iteratively applying NMA or introducing a smooth potential surface [53,54,55,56,57]. Also, ENM variants explicitly including nonlinear (higher-order) terms have been developed (e.g., nonlinear network model (NNM) [58]) and used for MD simulations (e.g., [59]).…”

Computer Simulation of Protein Materials at Multiple Length Scales: From Single Proteins to Protein Assemblies

“…Moreover, they have extended ENI coupled to a rigid cluster model [119], in which the rigid domains of a protein are modeled as a rigid body, while the interaction between the rigid domains is described by elastic springs. A recent study by Kim et al [120] has suggested a normal mode-guided elastic network interpolation (NGENI) in such a way that low-frequency normal modes are used in the linear interpolation for obtaining an intermediate conformation.…”