Computational Simulations Of Conformational Changes And Transport Mechanism Of AcrAB-TolC Multidrug Efflux System

Proteins are the main carriers of the biological phenomena, and the most abundant biological macromolecules. They have specific 3-dimensional structures which are folded from peptides with certain amino acid sequences under physiological conditions. In the early 1940s,3-dimensional structures of some proteins have been obtained from X-ray crystallography. Due to the static crystal structure, rigidity made the first impression of proteins. As the development of protein structure determination methods and more and more input of human and material resources, it was found that the structures of one protein, obtained from different crystallization environments, were in different conformations, among which collective conformational changes were observed. So the significance of conformational flexibility of proteins for their biological functions was gradually realized. In the view of physicists, proteins are soft materials that sample a large ensemble of conformations around the average structure as a result of thermal energy. A complete description of proteins requires a multidimensional energy landscape that defines the relative probabilities of the conformational states (thermodynamics) and the energy barriers between them. A crystal structure is only a single point on the energy landscape. To have a comprehensive understanding of protein function and reveal the molecular mechanism of biological process, we should not only determine the static structures, but also try to reveal the dynamic personalities, including the conformational flexibility and the transitions between different conformational states. So it becomes very necessary to introduce time dimension to conventional structural biological studies and extend the structure-function relationship to the structure-dynamics-function relationship.Protein dynamics covers a very broad time and amplitude scale, ranging from fast and local conformational changes, such as atomic fluctuations, side-chain rotations, to slow and global motions, such as collective domain movements, protein folding, allosteric effect, and protein denaturation. Many experimental techniques can be used to explore the conformational changes of proteins, such as X-ray crystallography, Nuclear Magnetic Resonance (NMR), Fluorescence Resonance Energy Transfer (FRET), Circular Dichroism (CD), Infrared spectrum (IR), Raman spectrum, and Electron Paramagnetic Resonance (EPR). Computational simulation has its own advantages over experiments in protein dynamics. It can describe protein dynamics on the atomic level, trace the evolving process of every atom to predict the property of protein, and give an atomistic description of protein dynamics to infer the structure-dynamics-function relationship, which is very difficult if not impossible to achieve by experimental techniques. At present, molecular dynamics are limited in time scale from picoseconds to nanoseconds, due to the computing power, parallel efficiency and calculation precision. But there are different methods and strategies to overcome this difficulty, such as replacement of all-atom force field with a coarse-grained one or introduction of external force to accelerate the conformational transition.The system we studied is AcrAB-TolC multidrug efflux system. It is hoped that computational simulations of this system describe its conformational dynamics, could enhance the understanding of multidrug resistance, reveal the molecular mechanism of the functional process, and provide more information and theoretical foundation for drug design and disease treatment.Multidrug resistance, the simultaneous acquisition of resistance to many chemically, structurally and functionally unrelated drugs or secretion of cell toxins to which the cell may have never been exposed, is the essential reason of the failure of bacterial infection treatment and human tumors chemotherapy. Multidrug resistance is, in large part, the story of multidrug efflux pump located on the membrane of bacterial or tumor cell. Multidrug efflux pumps exist widely in nature, and have been found in gram-negative bacteria, gram-positive bacteria and eukaryotic cell. Structures have now been obtained for multidrug transporters from five distinct superfamilies:the ABC (ATP binding cassette) superfamily, the RND (resistance nodulation division) superfamily, the MFS (major facilitator superfamily) superfamily, the MATE (the multidrug and toxic-compound extrusion) superfamily, the SMR (small multidrug resistance) superfamily. RND proteins are found in both prokaryotic and eukaryotic cells. They are active secondary transporters, energized by proton movement down the transmembrane electrochemical gradient. By far the best characterized RND protein is AcrB, the most important protein for multidrug resistance in E. coli.Two helper proteins, AcrA and TolC, are required for AcrB to pump antibiotics out of the cell. These three proteins make up of the AcrAB-TolC multidrug efflux system. In this system, AcrB is a main component located on the inner membrane that provides energy for substrate export by proton transmembrane transport, determines the substrate specificity, and translocates substrates into the TolC pore;. outer membrane protein TolC transfers substrates across both the periplasm and outer membrane; AcrA locates in the periplasm, interacts with both TolC and AcrB. AcrA is essential to the transport function of the system, but the detailed mechanism is poorly understood at present.The crystal structure revealed that AcrB is a homo-trimer. The three monomers are in different conformational states:access (A), binding (B), and extrusion (E), representing three consecutive functional states of a substrate transport cycle, and suggesting a three-step functionally rotating mechanism in which monomers undergo sequential conformational changes:A→B→>E→A. This mechanism proposed based on the crystal structure was supported by recent mutagenesis and cross-linking experiments, but can not be directly confirmed. Many questions remain unanswered about the working mechanism. Conventional molecular dynamics simulations of AcrB were first carried out, revealing that PD domains of AcrB trimer, especially the PC2 subdomains, have large conformational movement. The crystal structure with substrate binding in the binding pocket of B monomer and the protonation state of D407HD408H(E) is stable in the 50ns simulations, but it is difficult to simulate the functional conformational transition of AcrB by conventional MD simulations. So we resorted to targeted molecular dynamics simulation and normal mode analysis to study the large-scale conformational changes of this system. It is found that the modes with low frequency are dominated mainly the conformational movement of PD domain, and the low frequency modes have large contribution in the functional conformational transition. The normal modes analysis of AcrB trimer revealed that the conformational coupling between intra-monomer PD and TMD domains are very weak, but much more significant between inter-monomer PD and TMD domains, and between inter-monomer PN domains. The whole conformational transition (ABE→BEA) revealed by targeted MD simulations started from the conformational motion of PN2 subdomain of A monomer, transferring to B and E monomer by the conformational coupling between the PN subdomains. The intrinsic flexibility of PD domain, especially the PC2 subdomains, could provide an entropic driving force for the subsequent conformational changes required to complete the transport cycle. The transport of proton and substrate were after the collective conformational movement of PD domain, that is to say, the conformational changes of the side-chains of local residues in proton transport pathway may be the rate determining step of the substrate transport process.The substrate gets into the TolC pore after leaving AcrB transpoter. Outer membrane protein TolC is a channel-tunnel. The crystal structure of TolC is supposed to be a close state. From the variance of the radius along the pore, two gates are at each end of TolC:the extracellular gate (EG) and the periplasmic gate (PG). The extracellular gate is formed by the three large extracellular loops, and the periplasmic gate results from the inclined a-helices towards the molecular threefold axis, locked by a circular network of intra-and inter-monomer hydrogen bonds and salt bridges. To understand the free energy pathway of drugs along TolC pore, we have investigated the substrate permeability and gating mechanism of TolC by calculating the potential of mean forces (PMFs) for transporting sodium ion and doxorubicin through the TolC pore using the adaptive biased force (ABF) method. The transport mechanism is turned out to be substrate dependent. Gate opening was found to occur at the periplasmic entrance for the passage of both Na+ and doxorubicin, but the conformational gating does not lead to permeation barrier for Na+ at this region. The extracellular loops and K283 residues cause permeation barriers for Na+ at the extracellular entrance, but not for doxorubicin due to the extensive interactions between the drug molecule and the protein. TolC exhibited high conformational flexibility during the transport of Na+, while doxorubicin seems to be able to stabilize TolC in the resting state with closed periplasmic gate. The association of the TolC docking domain of AcrB does not help to open the periplasmic gate, while the gate opening induces the dissociation of the TolC-AcrB complex. To sample the conformation in open state in the conformational space of TolC, conventional molecular dynamics simulations of various mutants were carried out. We found that the totally open conformations are symmetric, but the intermediate conformations between close state and totally open state are asymmetric. And the targeted molecular dynamics simulations of the transition between close and totally open state revealed that the transition process is in asymmetric manner.The other component of AcrAB-TolC system is membrane fusion protein (MFP) AcrA, an elongated, sickle-shaped molecule, containing four domains:α-hairpin domain, lipoyl domain, β-barrel domain, membrane proximal (MP) domain. The role of AcrA in the tripartite AcrAB-TolC complex is poorly understood at present. The crystal structure only contains the core part without N- and C-terminal, introducing four mutations. Structural and EPR studies showed that AcrA has high conformational flexibility and exhibited pH-induced conformational change. We built the complete structure of AcrA (Model1 and Model2) through homology modeling and performed atomistic simulations of AcrA at different pH values. It was shown that the individual domain is very stable and the conformational flexibility of AcrA originates from the motions of α-hairpin and MP domains. The conformational dynamics of AcrA is sensitive to specific point mutations and pH values. In agreement with the EPR experiments, the inter-domain motions were restrained upon lowering pH from 7.0 to 5.0 in the simulations. It was found that the protonation/deprotonation of His285 underlies the pH-regulated conformational dynamics of AcrA by disturbing the local hydrogen bond interactions. Based on the crystal structure of AcrB and TolC, and the complete structure model of AcrA, we obtained the tripartite AcrAB-TolC assembly through molecular docking. From the normal modes analysis of the assembly, concerted motions were observed not only at the direct contact interfaces between various components, but also between distant parts of the whole complex. The presence of AcrA was shown to significantly strengthen the motional couplings between AcrB and TolC.We have performed a systemic and comprehensive research on the conformational flexibility and the transport mechanism of AcrAB-TolC multidrug efflux system, characterized the substrate transport in AcrAB-TolC system at the atomistic view, which may provide ideas and theoretical foundation for experimental design and disease treatment. On the other side, our work could enrich the methods and means of computational simulations of large protein system, and provide more information for the understanding of the structure-dynamics-function relationship. We anticipate that with the development of cpu speed and the optimization algorithms computational simulation will be more powerful in the exploration of conformational changes of proteins, and combining with experiments, the mysteries of life will be understood more deeply.