(294d) Exploring the Functional Conformational Transitions in Proteins Using Atomistic Simulations and Elastic Network Models

Proteins often undergo ligand-triggered large-scale conformational
changes that are involved in many cellular processes such as
transport, signal transduction, and catalysis, etc. High-resolution
crystallographic studies have indeed provided structural bases for the
stable open and closed conformations of various proteins, but
mechanistic details of the observed structural transitions are only
marginally understood. Detailed experimental characterization of these
functional transitions is significantly challenging because many
intermediate conformations of a protein along the
"activation-pathway'' are only transiently populated. Molecular
dynamics (MD) simulations can in principle provide
atomistically-resolved details of the conformational ensembles of
various proteins, but due to difficulties in observing large-scale
conformational changes on reasonable time-scales, new promising
simulation approaches have been proposed. These successful approaches
often exploit the dimensionality-reduction as a tool to explore the
free-energy landscape of proteins in large yet finite collective
variables (CVs). Traditionally, collective motions of proteins have
been analyzed successfully using normal mode analysis (NMA) of
equilibrium structures because normal modes solely arising from the
structure are intrinsically accessible to proteins and collective
protein dynamics can be usually described by the low-frequency end of
the mode spectrum. Particularly, normal mode analyses of
coarse-grained (usually C_α-based) elastic network models
(ENMs) of proteins with a single-parameter potential have been
remarkably successful in describing large-scale conformational
transitions in many biological complexes such as the ribosome. In this
work, we aim to understand large-scale conformational changes in a
class of integral membrane-proteins known as the ATP-binding cassette
(ABC) transporters that are responsible for nutrient uptake
(importers) or toxins export (exporters) using a combination of
atomistic simulations and elastic network normal mode
analysis. Specifically, we generate multiple independent
temperature-accelerated molecular dynamics (TAMD) simulations of the
components (maltose-binding protein, MBP, and nucleotide-binding
domains, NBD) of a maltose transporter. We find that a few
low-frequency normal modes of the open states of these proteins can
account for the functional transitions observed in atomistic TAMD
simulations. Therefore, we extend the NMA to understand the interplay
of conformational changes in the transporter components and the
transmembrane domains, a dynamic coupling of which leads to successful
transport events. We find reasonably high projections of the modes of
the intracellular ATP-domains on the functional displacement of the
entire transporter, and suggest that NBDs are likely the drivers of
this class of transporters.