- Email: [email protected]
Inward open characterization of EmrD transporter with molecular dynamics simulation
Accepted Manuscript Inward open characterization of EmrD transporter with molecular dynamics simulation Xianwei Tan, Boxiong Wang PII:
S0006-291X(16)30501-0
DOI:
10.1016/j.bbrc.2016.04.006
Reference:
YBBRC 35601
To appear in:
Biochemical and Biophysical Research Communications
Received Date: 17 March 2016 Accepted Date: 3 April 2016
Please cite this article as: X. Tan, B. Wang, Inward open characterization of EmrD transporter with molecular dynamics simulation, Biochemical and Biophysical Research Communications (2016), doi: 10.1016/j.bbrc.2016.04.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Inward Open Characterization of EmrD Transporter
RI PT
with Molecular Dynamics Simulation Short title: EmrD Transporter with Molecular Dynamics Simulation Xianwei Tan1, Boxiong Wang2 1
SC
School of Life Sciences, Tsinghua University, Beijing 100084, China Department of Precision Instrument, Tsinghua University, Beijing 100084, China
M AN U
2
AC C
EP
TE D
Coresponding Author: Boxiong Wang Department of Precision Instrument, Tsinghua University, Beijing 100084, China Email: [email protected]
1
RI PT
ACCEPTED MANUSCRIPT
Abstract
M AN U
SC
EmrD is a member of the multidrug resistance exporter family. Up to now, little is known about the structural dynamics that underline the function of the EmrD protein in inward-facing open state and how the EmrD transits from an occluded state to an inward open state. For the first time the article applied the AT simulation to investigate the membrane transporter protein EmrD, and described the dynamic features of the whole protein, the domain, the helices, and the amino acid residues during an inward-open process from its occluded state. The gradual inward-open process is different from the current model of rigid-body domain motion in alternating-access mechanism. Simulation results show that the EmrD inward-open conformational fluctuation propagates from a C-terminal domain to an N-terminal domain via the linker region during the transition from its occluded state. The conformational fluctuation of the C-terminal domain is larger than that of the N-terminal domain. In addition, it is observed that the helices exposed to the surrounding membrane show a higher level of flexibility than the other regions, and the protonated E227 plays a key role in the transition from the occluded to the open state.
AC C
EP
bond.
TE D
Keyword: molecular dynamics simulation; EmrD; adaptive tempering; open inward; hydrogen
2
ACCEPTED MANUSCRIPT Introduction
AC C
EP
TE D
M AN U
SC
RI PT
EmrD is a member of the multidrug resistance exporter family which belongs to the major facilitator superfamily (MFS) (1,2). It has 12 TM helices which fold into two functional domains, a C-domain and an N-domain. Each domain is a 6-helix bundle containing both buried and expose-to-membrane helices. Like other MFS transporters, EmrD operates via an alternating-access mechanism in which MFS proteins switch between two major states, an inward and an outward state (3-7). The structure of EmrD in an occluded state was determined by Yin et al in 2006 (8). Becktein O and Sansom MS described the dynamic characteristics of Lac Y and the glycerol-3-phosphate transporter of EmrD belonging to the same family (9)(10). Mchaourab HS investigated the protein helices induced by PH value using a biochemical method, but he didn’t explain the opening way of EmrD and the dynamic features of the whole protein, the structural region, the helixes and the amino acid residues in the process of opening (11).Tama F used a conventional molecular simulation (equilibrium simulation) combined with a stretching molecular simulation to investigate how a substrate passed through the EmrD. However, this investigation was conducted when the EmrD didn’t in the typical state of opening inward to the cell, and therefore he didn’t explain how the EmrD opened inwardly to the cell and the related features in the opening process (12). For the EmrD was in a closed conformation, they still didn’t presented a report on how the EmrD changed to an open-inward state from a closed state and its dynamic characteristics in the opening process. Up to now, how EmrD in closed state opens inwardly and the inward-open characteristics still remains unknown, but this is an important issue in understanding bacteria drug resistance. This paper describes a method combining the conventional molecular simulation with the adaptive tempering (AT) to make an EmrD in a closed state open inwardly to cells (13, 14). Although the AT is an effective accelerating temperature simulation method, higher temperatures are often used during its simulation for acceleration. But the drastic temperature-accelerating way can cause proteins to deform seriously, losing thus the natural structure of the protein. Therefore, AT has been only used successfully in simple systems such as the folding of simple proteins (15), simple simulations of RNA conformation transition (16), etc. All these simulations were conducted in relatively simple environments such as in water (17) or in vacuum, and the investigated objects usually have small number of molecular weights, which made the simulations easier to succeed. However, the membrane situation for the membrane protein is very complex with respect to the pure water environment. During the simulation, the instability of the membrane, the higher molecular weights, and the complex structure of membrane protein will make the simulation difficult to perform, causing easily a membrane deformation or a protein denaturation for higher temperatures in particular. Up to now, there are no successful examples of simulation applied in complex membrane transporter proteins. The AT simulation method presented in this paper selects a suitable tempering range, pressurizes skillfully the simulated system, so as not to cause a severe deformation to the protein during the drastic conformation changes, but generate a conformation change accorded with the alternated accessing mechanism(18). This article attempted to apply the AT simulation to the investigation of membrane protein EmrD, and obtained a process opening inwardly to cell from its closed state and the dynamic features of the whole protein, the domain, the helixes, and the amino acid residues in the opening 3
ACCEPTED MANUSCRIPT process.
Experiments Systems for Simulation
RI PT
Two all-atom simulation systems were constructed. The first one had a protonated E227, while the other one had a normal-state E227 (apo). The crystal structure (PDB ID: 2GFP) of protein was used as the initial structure in the two systems. All simulations were performed using NAMD2.8(19) with CHARMM27 force field for proteins, lipids, ions and water. The membrane had a POPC lipid bilayer (108×108Å2) and the system was solvated with water. The lipids that overlapped with the protein were removed. To make the system become neutral, the system was ionized with Na+ and Cl− ions.
SC
Molecular dynamics simulation
Adaptive tempering method
M AN U
A conventional simulation was carried out. Full electrostatics in the periodic system was calculated using the Particle Mesh Ewald (PME) method with a grid size of 108 Å for the x and y dimensions, and 96 Å for the z dimension. Hydrogen bonds were kept rigid, with nonbonded interactions within a 12 Å cutoff distance. Table 1 shows the flow chart of energy minimization and equilibrium simulation results (9).
Results
EP
TE D
To increase sampling efficiency, a simulation method incorporating the AT was used. The AT program was integrated into NAMD (19, 20). The adaptive tempering parameters (listed in Table 2) were set up in a NAMD configure file. To overcome energy barrier, temperatures were selected randomly between predetermined minimal and maximal values. The temperature was an essential parameter to the adaptive tempering simulation. We tried a series of minimal and maximal temperatures to run the AT simulation. The final optimal temperature ranges from 297K (minimal temperature) to 450K (maximal temperature).
AC C
A PDB structure of EmrD in occluded state was used as the initial confirmation. After a 100ns conventional MD simulation, the structure was relaxed and sustained in a stable state. No transitions from the occluded state to the inward open state were observed in both the system with protonated E227 and the control system, just the same as in previous study (9). We repeated the cMD simulation for several times and got the same result through the MD simulation combined with the AT which was carried out each for about 21ns. At the end of 11ns in the AT simulation, the classical IOF state protein appeared, showing a V-shape (8, 21-23). Then the structure lost the IOF state gradually. In this study, we focused on the 11ns process (Fig.1). We repeated the AT simulation for several times and got the same result. Changes in the distances between pairs of atoms. The distances between two pairs of the atoms, V175Cα-L378Cα and M158Cα-A355Cα, are shown in Fig.2A. The atom pairs are located at the external helix (Fig.2A). They characterize the largest distance between cytoplasmic side and 4
ACCEPTED MANUSCRIPT extracellular side of EmrD. The distance between atoms V175Cα-L378Cα located at the cytoplasmic side increases from 50 Å to 66 Å. The distance between M158Cα-A355Cα at the extracellular side decreases from 33Å to 25Å (Fig.2B). Changes in the distances between pairs of atoms indicated that the protein opened inwardly during simulation.
TE D
M AN U
SC
RI PT
Sequential deformations of amino acid. The sequential movements of protein component, such as the helix and the amino acid, are important for protein conformation change (24-26).The C-terminal region and the linker region between two domains show the largest fluctuation peaks than the other parts. This leaves a question of whether the active parts are fluctuating synchronously. The fluctuation associated with the center cavity IFO state is also not well understood (27). To answer these questions, the deformation of the amino acid is analyzed (28). The overall protein deformation is divided into several levels, at which the values of the deformation are set to 7.0 Å, 6.0 Å, 5.0 Å, 4.0 Å, 3.0 Å and 2.0 Å, respectively(29)(Fig.2C). Then the regions at different deformation levels are shown, and amino acids of protein are found to deform sequentially. The largest amount of the deformation appears in H12 of the C-domain. The first amino acid to deform is P382, whose movement is about 7.0 Å. Once the protein bends and opens inwardly, this amino acid begins to deform. The next larger deformation is the linker region, starting at P188 followed by the residue N39. The movements of residues A90, G43, G141 and A263 are at the third level of deformation. These residues, A228, I279, Q129 and N138, are at the forth level of deformation. At the final level of deformation 1.0Å, most of the amino acid starts to deform. Then the protein starts to open inwardly. It is concluded that the regions located at H12 and the linker loop between two domains are the first dynamic parts in the EmrD structure, which lead to a transition from the occluded state to an IFO state.
AC C
EP
Larger conformation fluctuation in C-domain and in helices exposed to the surrounding environment. Although the motions of the MFS protein functional domains are regarded as rigid body movements (30-31), the N-domain and the C-domain fluctuations can be observed during the AT simulation (Fig.2D). In the transition from the occluded state to the IFO state, the C-domain in the protonated E227 system shows an RMSD of 4.0 Å, which is larger than that of the N-domain (about 3.5 Å). Even if the difference is not dramatic, a higher flexibility in C-domain was still observed. After a 6 ns AT simulation (Fig.2D), the RMSD of the N-domain becomes flat. But the C-domain RMSD increases sharply in the apo system. As expected, the larger C-domain flexibility is confirmed by the RMSD curve. In addition, the C-domain and the N-domain in the protonated E227 system have shown more synchronous fluctuations than those in the apo system. To determine which region (helix) has more fluctuations, we analyzed the free energy change of the dihedral. The free energy change of the dihedral in a certain region (helix) indicates the fluctuation in related region (helix). The free energy is found as ∆G = k B T lnP/Pmax , where k B is the Boltzmann’s constant and T is the temperature in Kelvin(32). The probabilities P and Pmax are obtained from the distribution of the principal components for each structure from the trajectory (33). Helices exposed to the surrounding membrane or water, such as H3, H6, H9 and H12 (Fig.3), have shown larger dihedral angle fluctuations, while small fluctuations emerged in inner helices. The inter-domain cavity opens inward with a radius over 3Å, producing a notch in the free energy map (Fig.3). On the other hand, the root mean square fluctuation (RMSF) of non-hydrogen atoms 5
ACCEPTED MANUSCRIPT was calculated, showing a notable peak in the linking loop region between the residue A180 and the residue Y200. The RMSF of the C-terminal is larger than that of the N-terminal (Fig.4A). The reason that the C-terminal shows a larger fluctuation is that some helices, such as H12, in the C-terminal is more flexible than the other helices in the N-terminal.
Discussion
AC C
EP
TE D
M AN U
SC
RI PT
This article applied the AT simulation to investigate the membrane protein EmrD. When in a high temperature environment, membrane protein tends to denature easily. However, a higher temperature than the usual temperature is often used to accelerate the molecule simulation. In this paper, we used a range of temperatures from 297K to 450K to accelerate molecular dynamics simulation. To avoid the protein’s denaturing, pressures were increased to the simulation system. However, the reason why the increase in pressure to the protein didn’t make denaturing is still not clear. It is another question why a protein with protonated E227 undergoes such a drastic change from occlusion to the IFO conformation. According to the former MD simulation of EmrD(16), we know that the E227 is related to proton translocation in EmrD(15) and the pKa of the unprotonated E227 shifts dramatically. In our work, we have observed that some special hydrogen bonds have been formed between the protonated E227 located in H7 and other amino acids, such as the Q362 located in H12. The hydrogen bond between the E227 and the Q362 acts as a bridge connecting the helices H12 and H7 (Fig.4B). So the highly flexible H12 pulls other parts of the protein through hydrogen bond between the protonated E227 and the other amino acids, causing finally the protein to conform and open inwardly. In addition, the protonated E227 is involved in local hydrogen bond networks and interacts with the nearby residues. This local hydrogen network is a part of a larger one, which helps to maintain the stability of the whole protein structure (34,35). In our simulation experiment, EmrD opens inwardly gradually. Our results argue against a rigid-body domain motion as implied by a strict rocker-switch mechanism (36). In such a model, the protein can open or close in accompaniment with the rigid-body domain (37), but open gradually with the amino acid, helix domains. When EmrD opens gradually, we can investigate its dynamic features from the amino acid residues, the helices, the domain, and the whole protein. This investigation method can help to understand the mechanism of conformational changes in MFS proteins in an alternating-access process. The results of the simulation can be used for a further study of the EmrD dynamics related to physiological and biochemical functions. This method can be extended to study the MFS proteins with common features. Dynamics of the amino acid residues, the helices, the domain, and the whole protein plays a key role in conformation transition, which is essential to the design of antibacterial drugs. These helices or amino acids can be regarded as potential drug targets that hinder the transporter to open and make it lose the antidrug functions.
Author Contributions XW TAN designed research, performed research, analyzed data and wrote the paper; BX WANG wrote the paper.
6
ACCEPTED MANUSCRIPT Acknowledgements The authors are very grateful to Weitao Sun, Benzhuo Lu for their useful suggestions on the experiments. The authors also thank Haipeng Gong for his technical advice on the experiments.
AC C
EP
TE D
M AN U
SC
RI PT
Declarations of interest The authors declared that there is no conflict of interest.
7
ACCEPTED MANUSCRIPT
Reference 1.
Nikaido H (2009) Multidrug resistance in bacteria. Annu Rev Biochem 78:119-146 (in eng).
2.
Schweizer HP (2003) Efflux as a mechanism of resistance to antimicrobials in Pseudomonas
3.
RI PT
aeruginosa and related bacteria: unanswered questions. Genet Mol Res 2(1):48-62 (in eng). Smirnova I, Kasho V, Sugihara J, & Kaback HR (2009) Probing of the rates of alternating access in LacY with Trp fluorescence. Proc Natl Acad Sci U S A 106(51):21561-21566 . 4.
Radestock S & Forrest LR (2011) The alternating-access mechanism of MFS transporters arises
5.
SC
from inverted-topology repeats. J Mol Biol 407(5):698-715.
Kaback HR, Smirnova I, Kasho V, Nie Y, & Zhou Y (2011) The alternating access transport
M AN U
mechanism in LacY. J Membr Biol 239(1-2):85-93 . 6.
Huang Y, Lemieux MJ, Song J, Auer M, & Wang DN (2003) Structure and mechanism of the glycerol-3-phosphate transporter from Escherichia coli. Science 301(5633):616-620.
7.
Sobczak I & Lolkema JS (2004) Alternating access and a pore-loop structure in the Na+-citrate transporter CitS of Klebsiella pneumoniae. J Biol Chem 279(30):31113-31120.
8.
Yin Y, He X, Szewczyk P, Nguyen T, & Chang G (2006) Structure of the multidrug transporter
9.
TE D
EmrD from Escherichia coli.Science 312(5774):741-744. Stelzl LS, Fowler PW, Sansom MS, Beckstein O (2014) Flexible gates generate occluded intermediates in the transport cycle of LacY.J Mol Biol.426(3):735-51. D'Rozario RS1, Sansom MS (2013) Helix – dynamics – in a membrane transport protein:
EP
10.
comparative simulations of the glycerol-3-phosphate transporter and its constituent
AC C
helices.Biochemistry.52(45):7964-74. 11.
Steed PR, Zou P, Trone KE, Mchaourab HS (2013) Structure and pH-induced structural rearrangements of the putative multidrug efflux pump EmrD in liposomes probed by
site-directed spin labeling.Biochemisry. 52(45):7964-74.
12
.Baker J, Wright SH, Tama F (2012) Simulations of substrate transport in the multidrug transporter– EmrD.Proteins. 80(6):1620-32.
13.
Shim Y & Amar JG (2011) Adaptive temperature-accelerated dynamics.J Chem Phys 134(5):054127.
8
ACCEPTED MANUSCRIPT 14.
Lee MS & Olson MA (2011) Comparison of two adaptive temperature-based replica exchange methods applied to a sharp phase transition of protein unfolding-folding.J Chem Phys 134(24):244111.
15.
Schug A1, Herges T, Wenzel W.proteins (2004) All-atom folding of the three-helix HIV accessory protein with an adaptive parallel tempering method.Proteins 57(4):792-8. Gil-Ley A,Bussi G (2015) Enhanced Conformational Sampling Using Replica Exchange with
RI PT
16.
Collective-Variable Tempering.J Chem Theory Comput. 11(3): 1077 1085. 17.
Zhang C & Ma J (2010) Enhanced sampling and applications in protein folding in explicit solvent. J Chem Phys 132(24):244101.
Kaback HR, Smirnova I, Kasho V, Nie Y, & Zhou Y (2011) The alternating access transport mechanism in LacY. J Membr Biol 239(1-2):85-93 .
Phillips JC, et al (2005) Scalable molecular dynamics with NAMD. Journal of Computational Chemistry 26(16):1781-1802.
20.
Bhandarkar M, et al (2012) NAMD User's Guide -Version 2.9.(The Board of Trustees of the University of Illinois).
21.
M AN U
19.
SC
18.
Madej MG, Soro SN, & Kaback HR (2012) Apo-intermediate in the transport cycle of lactose
22.
TE D
permease (LacY). Proc Natl Acad Sci U S A 109(44):E2970-2978.
Abramson J, et al (2003) The lactose permease of Escherichia coli: overall structure, the sugar-binding site and the alternating access model for transport. FEBS Lett 555(1):96-101.
23.
Su CC, et al (2006)
Conformation of the AcrB multidrug efflux pump in mutants of the
24.
EP
putative proton relay pathway. J Bacteriol 188(20):7290-7296 . Singh-Zocchi M, Hanne J, & Zocchi G (2002) Plastic deformation of protein monolayers.
AC C
Biophys J 83(4):2211-2218. 25.
Dietz H, Berkemeier F, Bertz M, & Rief M (2006) Anisotropic deformation response of single protein molecules. Proc Natl Acad Sci U S A 103(34):12724-12728.
26.
Bartsch TF, et al (2009) Detecting sequential bond formation using three-dimensional thermal fluctuation analysis. Chemphyschem 10(9-10):1541-1547.
27.
Steed PR, Zou P, Trone KE, & McHaourab HS.2013. Structure and pH-induced structural rearrangements of the putative multidrug efflux pump EmrD in liposomes probed by site-directed spin labeling. Biochemistry 52(45):7964-7974.
28.
Velazquez-Muriel JA, et al. (2009) Comparison of molecular dynamics and superfamily spaces of protein domain deformation.BMC Struct Biol 9:6.
9
ACCEPTED MANUSCRIPT 29.
Karcher H, Lee SE, Kaazempur-Mofrad MR, & Kamm RD (2006) A coarse-grained model for force-induced protein deformation and kinetics.Biophys J 90(8):2686-2697.
30.
Khare D, Oldham ML, Orelle C, Davidson AL, & Chen J (2009) Alternating access in maltose transporter mediated by rigid-body rotations. Mol Cell 33(4):528-536 .
31.
Murakami S, Nakashima R, Yamashita E, Matsumoto T, & Yamaguchi A (2006) Crystal
RI PT
structures of a multidrug transporter reveal a functionally rotating mechanism.Nature 443(7108):173-179 . 32.
Glykos NM (2006) Software news and updates. Carma: a molecular dynamics analysis program.J Comput Chem 27(14):1765-1768.
David CC & Jacobs DJ (2014) Principal component analysis: a method for determining the
SC
33.
essential dynamics of proteins. Methods Mol Biol 1084:193-226 .
Pauling L, Corey RB, & Branson HR (1951) The structure of proteins; two hydrogen-bonded
M AN U
34.
helical configurations of the polypeptide chain.Proc Natl Acad Sci U S A 37(4):205-211. 35.
Hooft RW, Sander C, & Vriend G (1996) Positioning hydrogen atoms by optimizing hydrogen-bond networks in protein structures. in Proteins, pp 363-376.
37.
Yan H,et al (2013) Structure and mechanism of a nitrate transporter.Cell Rep. 3(3):716-23.
TE D
36.
Law CJ,et al (2007) Kinetic evidence is consistent with the rocker-switch mechanism of
AC C
EP
membrane transport by GlpT.Biochemistry.46(43):12190-7.
.
10
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
11
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
12
ACCEPTED MANUSCRIPT
Figure Legends: Figure.1 EmrD transits from an occluded state to an inward-open state. The process (A-C) shows how EmrD becomes inward-open state from the occluded state. The 11ns process has been
RI PT
snapshot for about every 4ns. Each step from A-C shows the conformation at the end of 4ns.
M AN U
SC
Figure.2 Changes in the distances between pairs of atoms. A. Locations of V175Cα-L378Cα and M158Cα-A355Cα. B. Variations of V175Cα-L378Cα, M158Cα-A355Cα distances in the AT simulation. C. Sequential deformations of amino acids. A–F show that the degree of deformation of different amino acids of protein decreased successively. The Cα atoms of the deformed amino acid are showed in red and the arrows indicate the direction in which amino acids are beginning to deform. The first deformed amino acid in different levels are shown by its name and residue number. D. RMSD evolution of C-domain and N-domain during the AT simulation. The RMSD plots of C-domain in apo protein and protonated E227 protein are represented by blue and red line, respectively. The RMSD plots of N-domain in apo protein and protonated E227 protein are represented by green and black line, respectively. The C-domain shows a larger RMSD than the N-domain in both apo system and protonated E227 system.
EP
TE D
Figure.3 Larger conformation fluctuation in C-domain and in helices exposed to the surrounding environment. The dihedral change of protein is projected to three planes. The free energy map of dihedral change of protein in each plane is divided into 7×7 sub-blocks and each block is named as a bin. A. The free energy map of dihedral change is projected to y-z plane. B. The free energy map is projected to x-z plane. C. The free energy map is projected to x-y plane. D. View of EmrD from extracellular side. H3, H6, H9 and H12 are exposed to the surrounding membrane. The energy value ranges from 0 kcal/mol to 2.5 kcal/mol presented by light blue to scarlet color.
AC C
Figure.4 A. Large fluctuations in linker region and in the C-terminal. RMSF of protein indicates fluctuation of different regions. B. Hydrogen bond formed between E227 and other amino acid. The red column represents the oxygen atom, while the white represents the hydrogen atom. The red lines represent hydrogen bonds.
13
ACCEPTED MANUSCRIPT
M AN U
SC
Extracellular end
RI PT
Figure 1
A
B
AC C
EP
TE D
Intracellular end
14
C
ACCEPTED MANUSCRIPT
EP
TE D
M AN U
SC
RI PT
Figure 2
AC C
C
D
15
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 3
16
ACCEPTED MANUSCRIPT
SC
RI PT
Figure 4
TE D
M AN U
A
AC C
EP
B
17
ACCEPTED MANUSCRIPT
Highlights: 1. This study described the dynamic features of the whole EmrD protein, the domain, the helices, and the amino acid residues during an inward-open process from its occluded state.
RI PT
2. Simulation results show that the EmrD inward-open conformational fluctuation propagates from a C-terminal domain to an N-terminal domain via the linker region during the transition from its occluded state.
SC
3. The conformational fluctuation of the C-terminal domain is larger than that of the N-terminal domain.
M AN U
4. The helices exposed to the surrounding membrane show a higher level of flexibility than the other regions, and the protonated E227 plays a key role in the transition from the occluded to the
AC C
EP
TE D
open state.
S0006-291X(16)30501-0
DOI:
10.1016/j.bbrc.2016.04.006
Reference:
YBBRC 35601
To appear in:
Biochemical and Biophysical Research Communications
Received Date: 17 March 2016 Accepted Date: 3 April 2016
Please cite this article as: X. Tan, B. Wang, Inward open characterization of EmrD transporter with molecular dynamics simulation, Biochemical and Biophysical Research Communications (2016), doi: 10.1016/j.bbrc.2016.04.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Inward Open Characterization of EmrD Transporter
RI PT
with Molecular Dynamics Simulation Short title: EmrD Transporter with Molecular Dynamics Simulation Xianwei Tan1, Boxiong Wang2 1
SC
School of Life Sciences, Tsinghua University, Beijing 100084, China Department of Precision Instrument, Tsinghua University, Beijing 100084, China
M AN U
2
AC C
EP
TE D
Coresponding Author: Boxiong Wang Department of Precision Instrument, Tsinghua University, Beijing 100084, China Email: [email protected]
1
RI PT
ACCEPTED MANUSCRIPT
Abstract
M AN U
SC
EmrD is a member of the multidrug resistance exporter family. Up to now, little is known about the structural dynamics that underline the function of the EmrD protein in inward-facing open state and how the EmrD transits from an occluded state to an inward open state. For the first time the article applied the AT simulation to investigate the membrane transporter protein EmrD, and described the dynamic features of the whole protein, the domain, the helices, and the amino acid residues during an inward-open process from its occluded state. The gradual inward-open process is different from the current model of rigid-body domain motion in alternating-access mechanism. Simulation results show that the EmrD inward-open conformational fluctuation propagates from a C-terminal domain to an N-terminal domain via the linker region during the transition from its occluded state. The conformational fluctuation of the C-terminal domain is larger than that of the N-terminal domain. In addition, it is observed that the helices exposed to the surrounding membrane show a higher level of flexibility than the other regions, and the protonated E227 plays a key role in the transition from the occluded to the open state.
AC C
EP
bond.
TE D
Keyword: molecular dynamics simulation; EmrD; adaptive tempering; open inward; hydrogen
2
ACCEPTED MANUSCRIPT Introduction
AC C
EP
TE D
M AN U
SC
RI PT
EmrD is a member of the multidrug resistance exporter family which belongs to the major facilitator superfamily (MFS) (1,2). It has 12 TM helices which fold into two functional domains, a C-domain and an N-domain. Each domain is a 6-helix bundle containing both buried and expose-to-membrane helices. Like other MFS transporters, EmrD operates via an alternating-access mechanism in which MFS proteins switch between two major states, an inward and an outward state (3-7). The structure of EmrD in an occluded state was determined by Yin et al in 2006 (8). Becktein O and Sansom MS described the dynamic characteristics of Lac Y and the glycerol-3-phosphate transporter of EmrD belonging to the same family (9)(10). Mchaourab HS investigated the protein helices induced by PH value using a biochemical method, but he didn’t explain the opening way of EmrD and the dynamic features of the whole protein, the structural region, the helixes and the amino acid residues in the process of opening (11).Tama F used a conventional molecular simulation (equilibrium simulation) combined with a stretching molecular simulation to investigate how a substrate passed through the EmrD. However, this investigation was conducted when the EmrD didn’t in the typical state of opening inward to the cell, and therefore he didn’t explain how the EmrD opened inwardly to the cell and the related features in the opening process (12). For the EmrD was in a closed conformation, they still didn’t presented a report on how the EmrD changed to an open-inward state from a closed state and its dynamic characteristics in the opening process. Up to now, how EmrD in closed state opens inwardly and the inward-open characteristics still remains unknown, but this is an important issue in understanding bacteria drug resistance. This paper describes a method combining the conventional molecular simulation with the adaptive tempering (AT) to make an EmrD in a closed state open inwardly to cells (13, 14). Although the AT is an effective accelerating temperature simulation method, higher temperatures are often used during its simulation for acceleration. But the drastic temperature-accelerating way can cause proteins to deform seriously, losing thus the natural structure of the protein. Therefore, AT has been only used successfully in simple systems such as the folding of simple proteins (15), simple simulations of RNA conformation transition (16), etc. All these simulations were conducted in relatively simple environments such as in water (17) or in vacuum, and the investigated objects usually have small number of molecular weights, which made the simulations easier to succeed. However, the membrane situation for the membrane protein is very complex with respect to the pure water environment. During the simulation, the instability of the membrane, the higher molecular weights, and the complex structure of membrane protein will make the simulation difficult to perform, causing easily a membrane deformation or a protein denaturation for higher temperatures in particular. Up to now, there are no successful examples of simulation applied in complex membrane transporter proteins. The AT simulation method presented in this paper selects a suitable tempering range, pressurizes skillfully the simulated system, so as not to cause a severe deformation to the protein during the drastic conformation changes, but generate a conformation change accorded with the alternated accessing mechanism(18). This article attempted to apply the AT simulation to the investigation of membrane protein EmrD, and obtained a process opening inwardly to cell from its closed state and the dynamic features of the whole protein, the domain, the helixes, and the amino acid residues in the opening 3
ACCEPTED MANUSCRIPT process.
Experiments Systems for Simulation
RI PT
Two all-atom simulation systems were constructed. The first one had a protonated E227, while the other one had a normal-state E227 (apo). The crystal structure (PDB ID: 2GFP) of protein was used as the initial structure in the two systems. All simulations were performed using NAMD2.8(19) with CHARMM27 force field for proteins, lipids, ions and water. The membrane had a POPC lipid bilayer (108×108Å2) and the system was solvated with water. The lipids that overlapped with the protein were removed. To make the system become neutral, the system was ionized with Na+ and Cl− ions.
SC
Molecular dynamics simulation
Adaptive tempering method
M AN U
A conventional simulation was carried out. Full electrostatics in the periodic system was calculated using the Particle Mesh Ewald (PME) method with a grid size of 108 Å for the x and y dimensions, and 96 Å for the z dimension. Hydrogen bonds were kept rigid, with nonbonded interactions within a 12 Å cutoff distance. Table 1 shows the flow chart of energy minimization and equilibrium simulation results (9).
Results
EP
TE D
To increase sampling efficiency, a simulation method incorporating the AT was used. The AT program was integrated into NAMD (19, 20). The adaptive tempering parameters (listed in Table 2) were set up in a NAMD configure file. To overcome energy barrier, temperatures were selected randomly between predetermined minimal and maximal values. The temperature was an essential parameter to the adaptive tempering simulation. We tried a series of minimal and maximal temperatures to run the AT simulation. The final optimal temperature ranges from 297K (minimal temperature) to 450K (maximal temperature).
AC C
A PDB structure of EmrD in occluded state was used as the initial confirmation. After a 100ns conventional MD simulation, the structure was relaxed and sustained in a stable state. No transitions from the occluded state to the inward open state were observed in both the system with protonated E227 and the control system, just the same as in previous study (9). We repeated the cMD simulation for several times and got the same result through the MD simulation combined with the AT which was carried out each for about 21ns. At the end of 11ns in the AT simulation, the classical IOF state protein appeared, showing a V-shape (8, 21-23). Then the structure lost the IOF state gradually. In this study, we focused on the 11ns process (Fig.1). We repeated the AT simulation for several times and got the same result. Changes in the distances between pairs of atoms. The distances between two pairs of the atoms, V175Cα-L378Cα and M158Cα-A355Cα, are shown in Fig.2A. The atom pairs are located at the external helix (Fig.2A). They characterize the largest distance between cytoplasmic side and 4
ACCEPTED MANUSCRIPT extracellular side of EmrD. The distance between atoms V175Cα-L378Cα located at the cytoplasmic side increases from 50 Å to 66 Å. The distance between M158Cα-A355Cα at the extracellular side decreases from 33Å to 25Å (Fig.2B). Changes in the distances between pairs of atoms indicated that the protein opened inwardly during simulation.
TE D
M AN U
SC
RI PT
Sequential deformations of amino acid. The sequential movements of protein component, such as the helix and the amino acid, are important for protein conformation change (24-26).The C-terminal region and the linker region between two domains show the largest fluctuation peaks than the other parts. This leaves a question of whether the active parts are fluctuating synchronously. The fluctuation associated with the center cavity IFO state is also not well understood (27). To answer these questions, the deformation of the amino acid is analyzed (28). The overall protein deformation is divided into several levels, at which the values of the deformation are set to 7.0 Å, 6.0 Å, 5.0 Å, 4.0 Å, 3.0 Å and 2.0 Å, respectively(29)(Fig.2C). Then the regions at different deformation levels are shown, and amino acids of protein are found to deform sequentially. The largest amount of the deformation appears in H12 of the C-domain. The first amino acid to deform is P382, whose movement is about 7.0 Å. Once the protein bends and opens inwardly, this amino acid begins to deform. The next larger deformation is the linker region, starting at P188 followed by the residue N39. The movements of residues A90, G43, G141 and A263 are at the third level of deformation. These residues, A228, I279, Q129 and N138, are at the forth level of deformation. At the final level of deformation 1.0Å, most of the amino acid starts to deform. Then the protein starts to open inwardly. It is concluded that the regions located at H12 and the linker loop between two domains are the first dynamic parts in the EmrD structure, which lead to a transition from the occluded state to an IFO state.
AC C
EP
Larger conformation fluctuation in C-domain and in helices exposed to the surrounding environment. Although the motions of the MFS protein functional domains are regarded as rigid body movements (30-31), the N-domain and the C-domain fluctuations can be observed during the AT simulation (Fig.2D). In the transition from the occluded state to the IFO state, the C-domain in the protonated E227 system shows an RMSD of 4.0 Å, which is larger than that of the N-domain (about 3.5 Å). Even if the difference is not dramatic, a higher flexibility in C-domain was still observed. After a 6 ns AT simulation (Fig.2D), the RMSD of the N-domain becomes flat. But the C-domain RMSD increases sharply in the apo system. As expected, the larger C-domain flexibility is confirmed by the RMSD curve. In addition, the C-domain and the N-domain in the protonated E227 system have shown more synchronous fluctuations than those in the apo system. To determine which region (helix) has more fluctuations, we analyzed the free energy change of the dihedral. The free energy change of the dihedral in a certain region (helix) indicates the fluctuation in related region (helix). The free energy is found as ∆G = k B T lnP/Pmax , where k B is the Boltzmann’s constant and T is the temperature in Kelvin(32). The probabilities P and Pmax are obtained from the distribution of the principal components for each structure from the trajectory (33). Helices exposed to the surrounding membrane or water, such as H3, H6, H9 and H12 (Fig.3), have shown larger dihedral angle fluctuations, while small fluctuations emerged in inner helices. The inter-domain cavity opens inward with a radius over 3Å, producing a notch in the free energy map (Fig.3). On the other hand, the root mean square fluctuation (RMSF) of non-hydrogen atoms 5
ACCEPTED MANUSCRIPT was calculated, showing a notable peak in the linking loop region between the residue A180 and the residue Y200. The RMSF of the C-terminal is larger than that of the N-terminal (Fig.4A). The reason that the C-terminal shows a larger fluctuation is that some helices, such as H12, in the C-terminal is more flexible than the other helices in the N-terminal.
Discussion
AC C
EP
TE D
M AN U
SC
RI PT
This article applied the AT simulation to investigate the membrane protein EmrD. When in a high temperature environment, membrane protein tends to denature easily. However, a higher temperature than the usual temperature is often used to accelerate the molecule simulation. In this paper, we used a range of temperatures from 297K to 450K to accelerate molecular dynamics simulation. To avoid the protein’s denaturing, pressures were increased to the simulation system. However, the reason why the increase in pressure to the protein didn’t make denaturing is still not clear. It is another question why a protein with protonated E227 undergoes such a drastic change from occlusion to the IFO conformation. According to the former MD simulation of EmrD(16), we know that the E227 is related to proton translocation in EmrD(15) and the pKa of the unprotonated E227 shifts dramatically. In our work, we have observed that some special hydrogen bonds have been formed between the protonated E227 located in H7 and other amino acids, such as the Q362 located in H12. The hydrogen bond between the E227 and the Q362 acts as a bridge connecting the helices H12 and H7 (Fig.4B). So the highly flexible H12 pulls other parts of the protein through hydrogen bond between the protonated E227 and the other amino acids, causing finally the protein to conform and open inwardly. In addition, the protonated E227 is involved in local hydrogen bond networks and interacts with the nearby residues. This local hydrogen network is a part of a larger one, which helps to maintain the stability of the whole protein structure (34,35). In our simulation experiment, EmrD opens inwardly gradually. Our results argue against a rigid-body domain motion as implied by a strict rocker-switch mechanism (36). In such a model, the protein can open or close in accompaniment with the rigid-body domain (37), but open gradually with the amino acid, helix domains. When EmrD opens gradually, we can investigate its dynamic features from the amino acid residues, the helices, the domain, and the whole protein. This investigation method can help to understand the mechanism of conformational changes in MFS proteins in an alternating-access process. The results of the simulation can be used for a further study of the EmrD dynamics related to physiological and biochemical functions. This method can be extended to study the MFS proteins with common features. Dynamics of the amino acid residues, the helices, the domain, and the whole protein plays a key role in conformation transition, which is essential to the design of antibacterial drugs. These helices or amino acids can be regarded as potential drug targets that hinder the transporter to open and make it lose the antidrug functions.
Author Contributions XW TAN designed research, performed research, analyzed data and wrote the paper; BX WANG wrote the paper.
6
ACCEPTED MANUSCRIPT Acknowledgements The authors are very grateful to Weitao Sun, Benzhuo Lu for their useful suggestions on the experiments. The authors also thank Haipeng Gong for his technical advice on the experiments.
AC C
EP
TE D
M AN U
SC
RI PT
Declarations of interest The authors declared that there is no conflict of interest.
7
ACCEPTED MANUSCRIPT
Reference 1.
Nikaido H (2009) Multidrug resistance in bacteria. Annu Rev Biochem 78:119-146 (in eng).
2.
Schweizer HP (2003) Efflux as a mechanism of resistance to antimicrobials in Pseudomonas
3.
RI PT
aeruginosa and related bacteria: unanswered questions. Genet Mol Res 2(1):48-62 (in eng). Smirnova I, Kasho V, Sugihara J, & Kaback HR (2009) Probing of the rates of alternating access in LacY with Trp fluorescence. Proc Natl Acad Sci U S A 106(51):21561-21566 . 4.
Radestock S & Forrest LR (2011) The alternating-access mechanism of MFS transporters arises
5.
SC
from inverted-topology repeats. J Mol Biol 407(5):698-715.
Kaback HR, Smirnova I, Kasho V, Nie Y, & Zhou Y (2011) The alternating access transport
M AN U
mechanism in LacY. J Membr Biol 239(1-2):85-93 . 6.
Huang Y, Lemieux MJ, Song J, Auer M, & Wang DN (2003) Structure and mechanism of the glycerol-3-phosphate transporter from Escherichia coli. Science 301(5633):616-620.
7.
Sobczak I & Lolkema JS (2004) Alternating access and a pore-loop structure in the Na+-citrate transporter CitS of Klebsiella pneumoniae. J Biol Chem 279(30):31113-31120.
8.
Yin Y, He X, Szewczyk P, Nguyen T, & Chang G (2006) Structure of the multidrug transporter
9.
TE D
EmrD from Escherichia coli.Science 312(5774):741-744. Stelzl LS, Fowler PW, Sansom MS, Beckstein O (2014) Flexible gates generate occluded intermediates in the transport cycle of LacY.J Mol Biol.426(3):735-51. D'Rozario RS1, Sansom MS (2013) Helix – dynamics – in a membrane transport protein:
EP
10.
comparative simulations of the glycerol-3-phosphate transporter and its constituent
AC C
helices.Biochemistry.52(45):7964-74. 11.
Steed PR, Zou P, Trone KE, Mchaourab HS (2013) Structure and pH-induced structural rearrangements of the putative multidrug efflux pump EmrD in liposomes probed by
site-directed spin labeling.Biochemisry. 52(45):7964-74.
12
.Baker J, Wright SH, Tama F (2012) Simulations of substrate transport in the multidrug transporter– EmrD.Proteins. 80(6):1620-32.
13.
Shim Y & Amar JG (2011) Adaptive temperature-accelerated dynamics.J Chem Phys 134(5):054127.
8
ACCEPTED MANUSCRIPT 14.
Lee MS & Olson MA (2011) Comparison of two adaptive temperature-based replica exchange methods applied to a sharp phase transition of protein unfolding-folding.J Chem Phys 134(24):244111.
15.
Schug A1, Herges T, Wenzel W.proteins (2004) All-atom folding of the three-helix HIV accessory protein with an adaptive parallel tempering method.Proteins 57(4):792-8. Gil-Ley A,Bussi G (2015) Enhanced Conformational Sampling Using Replica Exchange with
RI PT
16.
Collective-Variable Tempering.J Chem Theory Comput. 11(3): 1077 1085. 17.
Zhang C & Ma J (2010) Enhanced sampling and applications in protein folding in explicit solvent. J Chem Phys 132(24):244101.
Kaback HR, Smirnova I, Kasho V, Nie Y, & Zhou Y (2011) The alternating access transport mechanism in LacY. J Membr Biol 239(1-2):85-93 .
Phillips JC, et al (2005) Scalable molecular dynamics with NAMD. Journal of Computational Chemistry 26(16):1781-1802.
20.
Bhandarkar M, et al (2012) NAMD User's Guide -Version 2.9.(The Board of Trustees of the University of Illinois).
21.
M AN U
19.
SC
18.
Madej MG, Soro SN, & Kaback HR (2012) Apo-intermediate in the transport cycle of lactose
22.
TE D
permease (LacY). Proc Natl Acad Sci U S A 109(44):E2970-2978.
Abramson J, et al (2003) The lactose permease of Escherichia coli: overall structure, the sugar-binding site and the alternating access model for transport. FEBS Lett 555(1):96-101.
23.
Su CC, et al (2006)
Conformation of the AcrB multidrug efflux pump in mutants of the
24.
EP
putative proton relay pathway. J Bacteriol 188(20):7290-7296 . Singh-Zocchi M, Hanne J, & Zocchi G (2002) Plastic deformation of protein monolayers.
AC C
Biophys J 83(4):2211-2218. 25.
Dietz H, Berkemeier F, Bertz M, & Rief M (2006) Anisotropic deformation response of single protein molecules. Proc Natl Acad Sci U S A 103(34):12724-12728.
26.
Bartsch TF, et al (2009) Detecting sequential bond formation using three-dimensional thermal fluctuation analysis. Chemphyschem 10(9-10):1541-1547.
27.
Steed PR, Zou P, Trone KE, & McHaourab HS.2013. Structure and pH-induced structural rearrangements of the putative multidrug efflux pump EmrD in liposomes probed by site-directed spin labeling. Biochemistry 52(45):7964-7974.
28.
Velazquez-Muriel JA, et al. (2009) Comparison of molecular dynamics and superfamily spaces of protein domain deformation.BMC Struct Biol 9:6.
9
ACCEPTED MANUSCRIPT 29.
Karcher H, Lee SE, Kaazempur-Mofrad MR, & Kamm RD (2006) A coarse-grained model for force-induced protein deformation and kinetics.Biophys J 90(8):2686-2697.
30.
Khare D, Oldham ML, Orelle C, Davidson AL, & Chen J (2009) Alternating access in maltose transporter mediated by rigid-body rotations. Mol Cell 33(4):528-536 .
31.
Murakami S, Nakashima R, Yamashita E, Matsumoto T, & Yamaguchi A (2006) Crystal
RI PT
structures of a multidrug transporter reveal a functionally rotating mechanism.Nature 443(7108):173-179 . 32.
Glykos NM (2006) Software news and updates. Carma: a molecular dynamics analysis program.J Comput Chem 27(14):1765-1768.
David CC & Jacobs DJ (2014) Principal component analysis: a method for determining the
SC
33.
essential dynamics of proteins. Methods Mol Biol 1084:193-226 .
Pauling L, Corey RB, & Branson HR (1951) The structure of proteins; two hydrogen-bonded
M AN U
34.
helical configurations of the polypeptide chain.Proc Natl Acad Sci U S A 37(4):205-211. 35.
Hooft RW, Sander C, & Vriend G (1996) Positioning hydrogen atoms by optimizing hydrogen-bond networks in protein structures. in Proteins, pp 363-376.
37.
Yan H,et al (2013) Structure and mechanism of a nitrate transporter.Cell Rep. 3(3):716-23.
TE D
36.
Law CJ,et al (2007) Kinetic evidence is consistent with the rocker-switch mechanism of
AC C
EP
membrane transport by GlpT.Biochemistry.46(43):12190-7.
.
10
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
11
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
12
ACCEPTED MANUSCRIPT
Figure Legends: Figure.1 EmrD transits from an occluded state to an inward-open state. The process (A-C) shows how EmrD becomes inward-open state from the occluded state. The 11ns process has been
RI PT
snapshot for about every 4ns. Each step from A-C shows the conformation at the end of 4ns.
M AN U
SC
Figure.2 Changes in the distances between pairs of atoms. A. Locations of V175Cα-L378Cα and M158Cα-A355Cα. B. Variations of V175Cα-L378Cα, M158Cα-A355Cα distances in the AT simulation. C. Sequential deformations of amino acids. A–F show that the degree of deformation of different amino acids of protein decreased successively. The Cα atoms of the deformed amino acid are showed in red and the arrows indicate the direction in which amino acids are beginning to deform. The first deformed amino acid in different levels are shown by its name and residue number. D. RMSD evolution of C-domain and N-domain during the AT simulation. The RMSD plots of C-domain in apo protein and protonated E227 protein are represented by blue and red line, respectively. The RMSD plots of N-domain in apo protein and protonated E227 protein are represented by green and black line, respectively. The C-domain shows a larger RMSD than the N-domain in both apo system and protonated E227 system.
EP
TE D
Figure.3 Larger conformation fluctuation in C-domain and in helices exposed to the surrounding environment. The dihedral change of protein is projected to three planes. The free energy map of dihedral change of protein in each plane is divided into 7×7 sub-blocks and each block is named as a bin. A. The free energy map of dihedral change is projected to y-z plane. B. The free energy map is projected to x-z plane. C. The free energy map is projected to x-y plane. D. View of EmrD from extracellular side. H3, H6, H9 and H12 are exposed to the surrounding membrane. The energy value ranges from 0 kcal/mol to 2.5 kcal/mol presented by light blue to scarlet color.
AC C
Figure.4 A. Large fluctuations in linker region and in the C-terminal. RMSF of protein indicates fluctuation of different regions. B. Hydrogen bond formed between E227 and other amino acid. The red column represents the oxygen atom, while the white represents the hydrogen atom. The red lines represent hydrogen bonds.
13
ACCEPTED MANUSCRIPT
M AN U
SC
Extracellular end
RI PT
Figure 1
A
B
AC C
EP
TE D
Intracellular end
14
C
ACCEPTED MANUSCRIPT
EP
TE D
M AN U
SC
RI PT
Figure 2
AC C
C
D
15
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 3
16
ACCEPTED MANUSCRIPT
SC
RI PT
Figure 4
TE D
M AN U
A
AC C
EP
B
17
ACCEPTED MANUSCRIPT
Highlights: 1. This study described the dynamic features of the whole EmrD protein, the domain, the helices, and the amino acid residues during an inward-open process from its occluded state.
RI PT
2. Simulation results show that the EmrD inward-open conformational fluctuation propagates from a C-terminal domain to an N-terminal domain via the linker region during the transition from its occluded state.
SC
3. The conformational fluctuation of the C-terminal domain is larger than that of the N-terminal domain.
M AN U
4. The helices exposed to the surrounding membrane show a higher level of flexibility than the other regions, and the protonated E227 plays a key role in the transition from the occluded to the
AC C
EP
TE D
open state.