Identification of Functionally Key Residues in AMPA Receptor with a

Jul 3, 2013 - ABSTRACT: AMPA receptor mediates the fast excitatory synaptic transmission in the central nervous system, and it is activated by the bin...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/JPCB

Identification of Functionally Key Residues in AMPA Receptor with a Thermodynamic Method Ji Guo Su,† Hui Jing Du,† Rui Hao,† Xian Jin Xu,‡ Chun Hua Li,‡ Wei Zu Chen,‡ and Cun Xin Wang*,‡ †

College of Science, Yanshan University, Qinhuangdao, China College of Life Science and Bioengineering, Beijing University of Technology, Beijing, China



ABSTRACT: AMPA receptor mediates the fast excitatory synaptic transmission in the central nervous system, and it is activated by the binding of glutamate that results in the opening of the transmembrane ion channel. In the present work, the thermodynamic method developed by our group was improved and then applied to identify the functionally key residues that regulate the glutamate-binding affinity of AMPA receptor. In our method, the key residues are identified as those whose perturbation largely changes the ligand binding free energy of the protein. It is found that besides the ligand binding sites, other residues distant from the binding cleft can also influence the glutamate binding affinity through a long-range allosteric regulation. These allosteric sites include the hinge region of the ligand binding cleft, the dimer interface of the ligand binding domain, the linkers between the ligand binding domain and the transmembrane domain, and the interface between the N-terminal domain and the ligand binding domain. Our calculation results are consistent with the available experimental data. The results are helpful for our understanding of the mechanism of long-range allosteric communication in the AMPA receptor and the mechanism of channel opening triggered by glutamate binding.

I. INTRODUCTION AMPA receptor is an ionotropic glutamate receptor that mediates the fast excitatory synaptic transmissions in the central nervous system, which plays a central role in human learning and memory.1−4 It has been proved to be involved in many neurodegenerative disorders, such as amyotrophic lateral sclerosis and Parkinson’s and Alzheimer’s diseases, and therefore it is considered as a potential target for drug design.5−7 The activation of AMPA receptor requires the binding of glutamate to the ligand binding domain of the protein, which results in the opening of the channel and influx of cations.2−4,8,9 Experimental studies indicated that the binding affinity of AMPA receptor for their agonists is correlated with its functional responses.10,11 Therefore, identification of key residues that regulate the glutamate binding affinity is important for our understanding of the mechanism for the function of AMPA receptor. Obviously, the residues in the binding site, which directly interact with the ligand, play an important role for the association of the receptor with its ligand. Besides that, experimental studies have found that there exist other functional sites that are distant from the binding pocket can also regulate the binding affinity of AMPA receptor with its ligand through a long-range allosteric communication.4,8,9,12,13 Identification of these key residues involved in allosteric communication is critically useful for our understanding of the biological function of AMPA receptor. Although experimental methods, such as NMR,14,15 timeresolved X-ray crystallography,16 time-resolved spectroscopy,17 and biochemical methods,18 can effectively explore the © 2013 American Chemical Society

allosteric communication pathway and hot spots within proteins, the procedures are very expensive and timeconsuming. Therefore, effective computational approaches need to be developed to identify functionally key residues that are involved in allosteric communication. It has been widely accepted that protein structural topology and its intrinsic dynamics play critical roles in determining its functions. On the basis of this point, several computational methods have been developed to recognize functionally important sites through the analysis of protein topological properties. In these methods, the protein structures are represented as residue interaction networks, and several topology-related parameters of the networks were proposed to identify the allosteric communication pathways and functionally key residues.19−22 Besides the static characteristics of protein structure, it is increasingly clear that protein structure-encoded dynamics also play an important role in protein allostery and the performance of protein functions.23−25 The elastic network model (ENM), is a simple yet effective tool for exploring protein topology-encoded dynamics.26−29 An advantage of ENM is that the vibrational partition function of the system can be computed analytically, and thus the thermodynamic quantities, such as vibrational entropy and free energy, can be calculated easily.30,31 Based on ENM, several theoretical approaches have been developed to identify Received: March 6, 2013 Revised: June 3, 2013 Published: July 3, 2013 8689

dx.doi.org/10.1021/jp402290t | J. Phys. Chem. B 2013, 117, 8689−8696

The Journal of Physical Chemistry B

Article

free energy with a residue perturbed in the receptor. In this figure, the apostrophe represents that a residue perturbation is introduced into the system. We desire to calculate the quantity ΔΔG = ΔG′ − ΔG. The larger the ΔΔG value, the more important the residue for protein−ligand binding. The functionally key sites are identified as the residues whose perturbation induces relative large ΔΔG value. However, the interactions between the receptor and its ligand are complex, and the direct calculation of the binding free energy ΔG and ΔG′ is difficult. In order to facilitate the free energy calculation, a thermodynamic cycle was constructed, as shown in Figure 1. We designed two nonphysical processes: receptor → receptor′ and complex → complex′, and the free energy changes corresponding to these two processes are represented as ΔGrec and ΔGcomp, respectively. Considering that free energy is a state property, the change of its value depends only on the initial and final states instead of the path between them. Thus,

functionally key sites in proteins. The Bahar group applied ENM to obtain the low frequency modes of protein motions and they found that the functionally important residues tend to be located at the hinge sites of the softest modes.32 Atilgan et al. proposed a perturbation-response scanning technique based on ENM to determine the controlling sites that can be manipulated to achieve the functionally relevant conformational transition of proteins.33,34 Haliloglu and Erman developed an ENM-based statistical thermodynamics approach to identify the functionally important residues involved in energy exchange with the ligand and other key residues along the interaction pathway in the protein.35,36 Erman proposed a method to identify the allosteric signal transduction pathway and the involved key residues through analyzing the pairwise correlation of distance fluctuations of different residue pairs.37 Zheng and co-workers proposed a structural perturbation method based on ENM to determine the key residues that regulate the allosteric conformational transitions of proteins.38 Moreover, they also developed a dynamical correlation analysis approach to identify the hot-spot residues that display allosteric coupling with the active sites in proteins.39 Ming and Wall proposed an ENM-based method called Dynamics Perturbation Analysis (DPA) for investigating the allosteric influence of molecular interactions on protein conformational distributions, and the functional residues are identified as the sites where interactions cause a large change in the protein conformational distributions.40,41 Our group proposed a thermodynamic method based on ENM to predict the functionally key residues involved in protein conformational transitions.42 In the present work, our proposed ENM-based thermodynamic method was improved and then applied to determine the functional residues that play critical roles for the binding of AMPA receptor with its ligand glutamate. In this method, the functionally important residues are identified as those whose perturbation largely changes the ligand binding free energy of the protein. The residue perturbation is introduced by changing the force constant of the springs connecting to this residue. Then, the change of the ligand binding free energy in response to the residue perturbation is calculated. To facilitate the calculation, a thermodynamic cycle was constructed, and the change of free energy was computed by using ENM. The present study is helpful for our understanding the mechanism of long-range allosteric communication in the AMPA receptor and the mechanism of channel opening triggered by glutamate binding.

ΔΔG = ΔG′ − ΔG = ΔGcomp − ΔGrec

(1)

In this work, the values of ΔGcomp and ΔGrec were calculated based on the theory of Gaussian network model, one of the ENM. B. Gaussian Network Model. The Gaussian network model (GNM)26,29 describes the tertiary structure of a protein as an elastic network of Cα atoms, in which harmonic springs with an uniform constant are used to represent the interactions of the reside pairs within a cutoff distance Rc (8 Å is adopted in the present work). Given all contacting residues, the internal energy of the protein can be written as H=

1 γ(ΔRT ΓΔR ) 2

(2)

where γ is the force constant of the springs; {ΔR} represents the column vector of fluctuations of the residue Cα atoms, where N is the number of residues in the protein; the superscript T denotes the transpose; and Γ is the N × N symmetric Kirchhoff matrix whose elements are written as ⎧ −1 if i ≠ j and R ij ≤ Rc ⎪ ⎪ 0 if i ≠ j and R ij > Rc Γ=⎨ ⎪ ⎪− ∑ Γij if i = j ⎩ i ,i≠j

(3)

where Rij is the distance between the ith and jth Cα atoms and Rc is the cutoff distance. The mean-square fluctuation of the ith residue and the crosscorrelation fluctuation between the ith and jth residues are respectively given by

II. THEORY AND METHODS A. Thermodynamic Cycle Method. In our method, the key residues thermodynamically coupled with ligand binding were identified as those whose perturbations significantly change the binding free energy of the receptor with its ligand. As illustrated in Figure 1, ΔG represents the ligand-binding free energy of the wild-type receptor, and ΔG′ denotes the binding

⟨(ΔR i)2 ⟩ =

3kBT −1 [Γ ]ii γ

⟨ΔR i·ΔR j⟩ =

3kBT −1 [Γ ]ij γ

(4)

(5)

where kB is the Boltzmann constant; T is absolute temperature; and the meaning of γ and Γ are the same as eq 2. The inverse of the Kirchhoff matrix can be decomposed as

Γ−1 = U Λ−1UT

Figure 1. Schematic illustration of the thermodynamic cycle. 8690

(6)

dx.doi.org/10.1021/jp402290t | J. Phys. Chem. B 2013, 117, 8689−8696

The Journal of Physical Chemistry B

Article

mutation is the sum of ΔΔG values for all the perturbed springs. It should be noted that our method only applies to the springs that are present both in the receptor and the complex structures. However, in the complex structure, there exist springs formed by protein residues with the ligand, which are absent in the receptor structure. In this work, the ΔΔG value for this kind of spring was approximately evaluated by using the largest ΔΔG value of all the springs formed by the corresponding residue.

where U is an orthogonal matrix whose columns ui (1 ≤ i ≤ N) are the eigenvectors of Γ, and Λ is diagonal matrix of eigenvalues λi of Γ. The vibrational entropy of the protein can be expressed as43 S=

⟨H ⟩ − F 3 = (N − 1)kB + kB ln Z T 2

(7)

where ⟨H⟩ is the average vibrational energy. According to the law of equipartition of energy, the average internal energy for each mode of motion is 3/2kBT, and thus for all the N − 1 internal motion modes, the average vibrational energy ⟨H⟩ = 3/ 2(N − 1)kBT; F = −kBT ln Z is the vibarational Helmholtz free energy; Z is the configurational integral part of the vibrational partition function given by Z = ∫ exp(−H/kBT)d{ΔR}; kB is the Boltzmann constant; T is the absolute temperature; and N is the number of residues in the protein. When a perturbation was introduced into the protein, such as when the force constant of a spring is reduced, the vibrational entropy of the system will change correspondingly. In the linear approximation, the change of the entropy in response to the perturbation of the spring connecting the ith and jth residues can be written as42 ΔS =

III. RESULTS AND DISCUSSION AMPA receptor is a tetrameric protein assembled as dimer of dimer,2−4,9,44 as shown in Figure 2. Each subunit consists of

∂S Δγ ∂γij ij

=−

1 (⟨(ΔR i)2 ⟩ + ⟨(ΔR j)2 ⟩ − 2⟨ΔR iΔR j⟩)Δγij 2T (8) Figure 2. Left: The crystal structure of full-length tetrameric AMPA receptor (PDB code: 3KG2) in the resting state, in which the four subunits A (red), B (yellow), C (green) and D (blue), are assembled as dimer of dimer. Each subunit consists of four domains: the NTD, the LBD, the TMD, and the CTD. The CTD was not resolved in the crystal structure. Right: The constructed full-length AMPA receptor dimer complexed with glutamate based on the crystal structure of 1FTJ and 3KG2. Subunit A and D are also shown in red and blue colors, respectively. The ligand glutamate is displayed in space-filling model with gray color.

here, Δγij is the change of the force constant of the spring when a perturbation is introduced. Δγij is a negative value representing the decrease of the force constant. Based on the theory of GNM, we now calculate the value of eq 1. According to thermodynamic theory, free energy is composed of potential energy and entropy. Eq 1 can be written as ΔΔG = ΔGcomp − ΔGrec = (ΔUcomp − ΔUrec) − T (ΔScomp − ΔSrec)

(9)

four domains: the N-terminal domain (NTD), the ligandbinding domain (LBD), the transmembrane domain (TMD) and the C-terminal domain (CTD). The extracellular NTD contributes to receptor assembly, trafficking, and modulation. Recently, it has been suggested that the NTD can also allosterically control the channel function of the receptor, in which the conformational change of NTD transmits through the NTD−LBD interface to LBD and eventually down to the channel region.45,46 The extracellular LBD is composed of two lobes, D1 and D2, that form a clamshell for the binding of glutamate. Glutamate binding triggers large-scale domain movements that results in the closure of the clamshell. The conformational motion of LBD then propagates to TMD via the LBD−TMD linkers and induces the ion channel gating. TMD is the ion channel pore-forming domain that consists of three transmembrane helices, i.e., M1, M3 and M4, and a short re-entrant loop M2. Experimental studies have indicated that the linkers between LBD and TMD play a crucial role in mediating the signal transduction from LBD to TMD, which control the gating of the ion channel.47 The intracellular CTD mainly regulates the anchoring and trafficking of the receptor. According to the above discussion, the binding of glutamate to LBD is crucial for the function of AMPA receptor, which triggers the activation of the receptor. Experimental studies also

where, ΔU and ΔS denote the potential energy and entropy changes of the systems, respectively. In our method, the perturbation is introduced by reducing the force constant of the spring. For pairwise residues i and j that form a contact both in the receptor and the complex structures, the reduction of the force constant will result in the same change of the potential energy for the receptor and the complex states, i.e., ΔUcomp = ΔUrec. Thus, ΔΔG = −T(ΔScomp − ΔSrec) = −T ΔΔS

(10)

Combing eq 8 and eq 10, the ΔΔG value for the perturbation of each spring can be calculated. The springs with relative large ΔΔG value are considered to be important for receptor−ligand binding. In our calculation, the first 30 slowest normal modes of the receptor are taken into account, and for the complex structure, all the motion modes are projected onto these considered 30 modes of the receptor. The above method only evaluates the effects of a perturbation of a single spring on protein−ligand binding. In order to identify the key residues in the protein, each residue should be perturbed, such as mutation of the residue. When a residue was mutated, all the springs connecting to the residue are perturbed. Thus, the ΔΔG value caused by the point 8691

dx.doi.org/10.1021/jp402290t | J. Phys. Chem. B 2013, 117, 8689−8696

The Journal of Physical Chemistry B

Article

on our proposed method. The residues with relatively large ΔΔG value were identified as the thermodynamically key sites that are important for the binding of glutamate to AMPA receptor. The calculation results are displayed in Figure 3. As

indicated that the binding affinity of AMPA receptor for their agonists is correlated with its functional responses.10,11 Therefore, identification of key residues that regulate the agonists binding affinity is important for our understanding of the mechanism for the function of AMPA receptor. Intuitively, the residues in the binding cleft may be important for the binding of the ligand. Many experiments have successfully determined the key sites in the binding cleft that are crucial for the binding affinity of glutamate and then modulated the function of the receptor through mutation of the corresponding key residues.48,49 Besides that, more and more evidence has indicated that there exist other functional sites, which are distant from the binding cleft, can also modulate the binding and dissociation of the agonist and then regulate the function of the receptor.4,8,9,12,13 Many positive and negative allosteric modulators binding at the LBD dimer interface have been found or designed to effectively modulate the binding affinity of glutamate and affect the activation/deactivation of the receptor.8,9,12 Several experimental studies also shown that there are allosteric interactions among the four domains of the receptor.4,13,45 The conformational change in ATD can transmits to LBD and then to TMD, which can regulate the binding affinity of glutamate in LBD and ion channel gating in TMD. The CTD was also proved by experiments to be important for the regulation of membrane trafficking and receptor function. However, the CTD was not resolved in the crystal structure of AMPA receptor, therefore, this domain will not be discussed in this study. In the present work, the functionally key residues that regulate the binding of glutamate to LBD are identified using our proposed method, which is helpful for understanding the allosteric communication within the AMPA receptor and the mechanism of protein function. The crystal structure of the full-length AMPA receptor (the CTD was not resolved) in the resting state, where an antagonist bound to the LBD, has been reported (PDB code: 3KG2).44 However, for the glutamate-activated state, only the isolated LBD dimer bound with glutamate has been determined (PDB code: 1FTJ),50 but the NTD, TMD, and CTD were not be resolved. At first, the full-length AMPA receptor dimer complexed with glutamate was constructed based on these two structures. The following strategy was adopted: (1) D1 domain of the LBD-glutamate complex (PDB code: 1FTJ) was superimposed with the D1 domain of the full-length AMPA receptor (PDB code: 3KG2), and then the NTD was grafted from the latter to the former; (2) the TMD was grafted from the full-length AMPA receptor to the LBD−glutamate complex after superimposition of the D2 domain. It should be noted that the structure in 3KG2 includes four subunits assembled as dimer of dimer, and only subunits A and D were used to construct the full-length dimeric complex structure of AMPA receptor with its ligand glutamate. The constructed complex structure is shown in Figure 2. It is found that the structure of subunit A and D is not symmetrical. There is a substantial interface between ATD and LBD in subunit A, whereas in subunit D, these two domains are distant. In the present work, we focus on the binding of glutamate to subunit A. In our calculation, this constructed structure was used as the complex structure, and the structure after removing glutamate from subunit A was used as the receptor structure. Based on the constructed structures of the AMPA receptor and the receptor−glutamate complex, the corresponding elastic networks were set up. Then each residue in the structures was perturbed, and the ΔΔG value in response was calculated based

Figure 3. ΔΔG value for each residue perturbation. There are 16 clusters of residues with relative large ΔΔG values, which are marked by the numbers 1−16 in the figure.

shown in this figure, there are 16 peaks in the curve. Each peak corresponds to a cluster of residues with relatively large ΔΔG values, and the residue with the maximum ΔΔG value in the peak is defined as the central residue of the corresponding cluster. These 16 residue clusters are centered at Lys187, Gly217, Glu402, Tyr424, Lys449, Thr480, Leu498, Ser635, Gly653, Thr686, Tyr711, Gly731, Ala749, and Trp767 in subunit A, and Ser497, and Gln756 in subunit D. These clusters of residues are denoted by the numbers 1−16 in Figure 3. In order to display these key residues more intuitively, the central residues of these clusters are also marked on protein tertiary structure, as shown in Figure 4. According to their location in the protein structure, these residue clusters can be classified into four groups. (1). The Glutamate Binding Pocket in LBD. The residues of clusters 3, 4, 5, 6, 9, 10, 11 are located in the ligand binding pocket. The crystal structure and the GNM show that residues Tyr 450 in cluster 5, Pro478, and Thr480 in cluster 6, Ser654 in cluster 9, and Glu705 in cluster 11 directly interact with the ligand, which are responsible for the tightly binding of glutamate with the receptor. The crystal structures of the receptor complexed with various agonists have shown that the interactions between these residues and the ligand are highly conserved. Holm et al. have reported that the substitution of Tyr450 with Ala results in a dramatic reduction in the potency of glutamate to AMPA receptor.48 Experimental studies have indicated that when Glu705 was mutated, the binding of agonist was abolished. The experimental studies of Armstrong et al. have showed that the mutation of Leu650 in cluster 9 to Thr decreases the potency of glutamate to AMPA receptor by 8.5 fold.51 The residues in the clusters 3, 4, 10 are not directly interacting with the ligand, but they are located on the opposite sides of the binding cleft and form cross-cleft interactions. These interdomain interactions stabilize the closure of the cleft and have significant impacts on the binding affinity of AMPA receptor with its agonists. This view has been supported by the experimental observations. Several pieces of experimental evidence have indicated that the degree of binding cleft closure is correlated with agonist efficacy.52−54 For example, Residue Glu402 in cluster 3 and Thr686 in cluster 10 are located at the mouth of the binding cleft. These two residues form a crosscleft interaction that plays an important role in stabilization of 8692

dx.doi.org/10.1021/jp402290t | J. Phys. Chem. B 2013, 117, 8689−8696

The Journal of Physical Chemistry B

Article

Figure 4. The central residues of the observed 16 clusters are mapped on the crystal structure of AMPA receptor. The interface between NTD and LBD, the ligand binding pocket and the dimer interface of LBD, and the interface between LBD and TMD are highlighted in the left.

the cleft-closure conformation. The mutations of Glu402 or Thr686 that disrupt such interdomain interaction decrease both binding affinity and efficacy of the agonist.49,55−57 (2). The Hinge Region between D1 and D2 of LBD. The two lobes of LBD, D1 and D2, are connected by a hinge region composed of two polypeptide strands (residues 494− 502 and 726−734). Upon the binding of glutamate, the two lobes undergo a large-scale open-to-closed conformational transition axed at the hinge region. The residue clusters 7 and 12 are located at the hinge region. These residues mediate the domain motion of LBD and control the opening and closing of the ligand-binding cleft. The crystal structures showed that the backbone torsion angles between Ser497 and Ile500 in cluster 7, and between Leu727 and Gly731 in cluster 12 undergo 20° shifts during the cleft conformational transition from the open to the closed state. Therefore, these residues have significant impact on the binding of glutamate to AMPA receptor. Jin et al. have showed that aniracetam or CX614 bound at the site adjacent to the hinge of LBD can slow AMPA receptor deactivation.12 They also found that the mutations of Ser497 in cluster 7 and Ser 729 in cluster 12 will impair the effects of CX614.12,53 (3). The LBD Dimer Interface. The LBD dimer interface serves as a critical structural element in the gating process of AMPA receptor, which mediates the transmission of the binding energy from the glutamate binding cleft to the ion channel region in TMD. It was proposed that the cleft closure triggered by ligand binding puts strains on the dimer interface. The strains are relieved by the opening of the ion channel, or

by rearrangement of the dimer interface that results in the desensitization of the receptor. Therefore, the LBD dimer interface plays a crucial role in the activation, deactivation, and desensitization of AMPA receptor. Experimental evidence has indicated that the rates of desensitization and deactivation are largely determined by the stability of the dimer interface, where strengthening the dimer interface can reduce desensitization and deactivation.4,8,9,58−61 Our predicted key residue clusters 6, 7, 12, 13, 15, and 16 are located at the LBD dimer interface. The residue clusters 6 and 13 are located at helix D (residues 480−486) and J (residues 743−756) in subunit A, and cluster 16 is located at helix J in the adjacent subunit D. Crystallographic studies have revealed that helix D of one subunit and helix J of the adjacent subunit form significant intermolecular contacts to stabilize the dimer interface. The experiments have indicated that the mutation of Leu483 to Tyr in cluster 6 stabilizes the dimer interface and thus reduces desensitization of AMPA receptor,58,62 whereas, the Ser754Asp mutation in cluster 16 destabilizes the dimer interface and promotes desensitization.58,63 Horning and Mayer have designed several residue substitutions that disrupt the intermolecular contacts between helices D and J in adjacent subunits and greatly accelerate desensitization of AMPA receptor.59 The other predicted residue clusters 7, 12, and 15 are located at the center of the dimer interface. Many allosteric modulators for AMPA receptor have been found to bind at this region, which can regulate the deactivation and desensitization of the receptor. For example, aniracetam and CX614 that bind to the center of the dimer interface can stabilize the closed-cleft 8693

dx.doi.org/10.1021/jp402290t | J. Phys. Chem. B 2013, 117, 8689−8696

The Journal of Physical Chemistry B

Article

(2) The hinge region of the ligand binding cleft. These sites mediate the opening-closure motion of the cleft and play an important role for the binding and dissociation of the ligand. (3) The LBD dimer interface. This region is subjected to strains caused by the conformational change of the cleft and in turn influences the opening-closure motion of the cleft. (4) The linkers between LBD and TMD. These linkers mediate the transmission of the conformational motion between ligandbinding cleft and the ion channel. (5) The interface between NTD and LBD. These residues mediate the transmission of domain motion in NTD to LBD and eventually down to TMD. For these key residues, except for the first group, the other four groups do not directly interact with the ligand and even are distant from the ligand binding cleft. The results imply that these distant key residues can regulate the glutamate binding affinity through allosteric communications. Our studies are helpful for our understanding the mechanism of the allosteric regulation in AMPA receptor and exploring the mechanism of channel opening triggered by glutamate binding. It should be noted that our calculation results are supported by the available experimental observations; however, at present, the amount of available experimental data is limited. The validity of our method should be verified by more experimental results in the future.

conformation of the LBD, and thus reduce the deactivation of the receptor.4,12 (4). The Linkers between LBD and TMD. The ion channel-forming domain TMD is composed of four segments (M1−M4), where M3 is the major structural element involved in channel opening and closure. LBD connects with TMD through three interdomain linkers, i.e., LBD−M1, LBD−M3, and LBD−M4 linkers. The cleft-closure motion of LBD triggered by ligand binding can transmit to TMD through these linkers and then induce the opening of the channel pore. Among these three linkers, the linker between LBD and M3 mainly contribute to the transfer of the conformational changes in LBD to TMD. Our predicted key residue clusters 8 and 14 are located at the LBD−TMD linkers. Residue cluster 8 (centered at Ser635) is located at the LBD−M3 linker and cluster 14 (centered at Trp767) at the LBD−M4 linker. The experiments with mutations have indicated that the LBD−M3 linker greatly affects the desensitization and channel gating of AMPA receptor.64 (5). The Interface between NTD and LBD. Based on normal-mode analysis (NMA) and molecular dynamics (MD) simulation results, Sukumaran et al.45 and Dutta et al.46 proposed that the NTD can modulate the channel function of AMPA receptor. They suggested that NTD is capable of undergoing clamshell motion, and the domain motion in NTD transmits through the NTD−LBD interface to LBD and eventually down to the channel region. Our predicted key residue clusters 1 and 2 are located at the interface between NTD and LBD. NMA has revealed that for the clamshell motion of NTD, the αF (residues 174−183), αG (residues 198−211) and αI (residues 233−238) in the lower lobe of NTD exhibit largest fluctuations. The predicted residue cluster 1 (centered at Lys187) is located at the contacting region between αF and LBD, and the cluster 2 (centered at Gly217) at the region between αG and LBD. These key residues transmit the large motion of NTD down to LBD and then influence the movement of LBD.



AUTHOR INFORMATION

Corresponding Author

*Address: College of Life Science and Bioengineering, Beijing University of Technology, Beijing 100024, China; Tel: +86-1067392724; E-mail: [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.



IV. CONCLUSION AMPA receptor is one subtype of the ionotropic glutamate receptors that plays a critical role for the excitatory synaptic transmission in the central nervous system. Upon the binding of glutamate, the LBD of AMPA receptor undergoes cleftclosure motion, which then induces the opening of the ion channel and activates the receptor. Experimental evidence has suggested that the binding affinity of glutamate significantly affects the functional response of AMPA receptor. Therefore, identification of the functionally key residues that influence the glutamate binding affinity is important for our understanding of the mechanism of glutamate-triggered channel opening. In the present work, the ENM-based thermodynamic method developed by our group was applied to identify the key residues that regulate the glutamate binding affinity. In our method, the functionally important sites were determined as the residues whose perturbation significantly alters the ligand binding free energy. In order to facilitate the calculation, a thermodynamic cycle was constructed and the change of free energy was computed by using ENM. Our calculation results showed that the functionally key residues can be classified into five groups according to their locations in the protein structure: (1) The glutamate binding sites. These residues directly interact with the ligand glutamate, which are critically important for the glutamate binding affinity.

ACKNOWLEDGMENTS C.X.W. gratefully acknowledges the grant from the International S&T Cooperation Program of China (No. 2010DFA31710). J.G.S. acknowledges support from the National Natural Science Foundation of China (11204267) and the Science and Technology Research and Development Program of Hebei Province (11215170). C.H.L. is thankful for the support from the National Natural Science Foundation of China (31171267).



REFERENCES

(1) Nakanishi, S. Molecular Diversity of Glutamate Receptors and Implications for Brain Function. Science 1992, 258, 597−603. (2) Madden, D. R. The Structure and Function of Glutamate Receptor Ion Channels. Nat. Rev. Neurosci. 2002, 3, 91−101. (3) Gouaux, E. Structure and Function of AMPA Receptors. J. Physiol. 2004, 554, 249−253. (4) Traynelis, S. F.; Wollmuth, L. P.; McBain, C. J.; Menniti, F. S.; Vance, K. M.; Ogden, K. K.; Hansen, K. B.; Yuan, H.; Myers, S. J.; Dingledine, R. Glutamate Receptor Ion Channels: Structure, Regulation, and Function. Pharmacol. Rev. 2010, 62, 405−496. (5) O’Neill, M. J.; Witkin, J. M. AMPA Receptor Potentiators: Application for Depression and Parkinson’s Disease. Curr. Drug Targets 2007, 8, 603−620. (6) Bowie, D. Ionotropic Glutamate Receptors and CNS Disorders. CNS Neurol. Disord.: Drug Targets 2008, 7, 129−143.

8694

dx.doi.org/10.1021/jp402290t | J. Phys. Chem. B 2013, 117, 8689−8696

The Journal of Physical Chemistry B

Article

Dynamics Using the Gaussian Network Model. Nucleic Acids Res. 2006, 34, W24−W31. (30) Zimmermann, M. T.; Leelananda, S. P.; Kloczkowski, A.; Jernigan, R. L. Combining Statistical Potentials with Dynamics-Based Entropies Improves Selection from Protein Decoys and Docking Poses. J. Phys. Chem. B 2012, 116, 6725−6731. (31) Zimmermann, M. T.; Leelananda, S. P.; Gniewek, P.; Feng, Y.; Jernigan, R. L.; Kloczkowski, A. Free Energies for Coarse-Grained Proteins by Integrating Multibody Statistical Contact Potentials with Entropies from Elastic Network Models. J. Struct. Funct. Genomics 2011, 12, 137−147. (32) Dutta, A.; Bahar, I. Metal-Binding Sites Are Designed to Achieve Optimal Mechanical and Signaling Properties. Structure 2010, 18, 1140−1148. (33) Atilgan, C.; Gerek, Z. N.; Ozkan, S. B.; Atilgan, A. R. Manipulation of Conformational Change in Proteins by SingleResidue Perturbations. Biophys. J. 2010, 99, 933−943. (34) Atilgan, C.; Atilgan, A. R. Perturbation-Response Scanning Reveals Ligand Entry−Exit Mechanisms of Ferric Binding Protein. PLoS Comput. Biol. 2009, 5, e1000544. (35) Haliloglu, T.; Erman, B. Analysis of Correlations between Energy and Residue Fluctuations in Native Proteins and Determination of Specific Sites for Binding. Phys. Rev. Lett. 2009, 102, 088103. (36) Haliloglu, T.; Gul, A.; Erman, B. Predicting Important Residues and Interaction Pathways in Proteins Using Gaussian Network Model: Binding and Stability of HLA Proteins. PLoS Comput. Biol. 2010, 6, e1000845. (37) Erman, B. A Fast Approximate Method of Identifying Paths of Allosteric Communication in Proteins. Proteins 2013, DOI: 10.1002/ prot.24284. (38) Zheng, W.; Brooks, B. R.; Thirumalai, D. Low-frequency Normal Modes That Describe Allosteric Transtitons in Biological Nanomachines Are Robust to Sequence Variations. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 7664−7669. (39) Zheng, W.; Tekpinar, M. Large-Scale Evaluation of Dynamically Important Residues in Proteins Predicted by the Perturbation Analysis of a Coarse-Grained Elastic Model. BMC Struct. Biol. 2009, 9, 45. (40) Ming, D.; Wall, M. E. Allostery in a Coarse-Grained Model of Protein Dynamics. Phys. Rev. Lett. 2005, 95, 198103. (41) Ming, D.; Wall, M. E. Quantifying Allosteric Effects in Proteins. Proteins 2005, 59, 697−707. (42) Su, J. G.; Xu, X. J.; Li, C. H.; Chen, W. Z.; Wang, C. X. Identification of Key Residues for Protein Conformational Transition Using Elastic Network Model. J. Chem. Phys. 2011, 135, 174101. (43) Bahar, I.; Atilgan, A. R.; Demirel, M. C.; Erman, B. Vibrational Dynamics of Folded Proteins: Significance of Slow and Fast Motions in Relation to Function and Stability. Phys. Rev. Lett. 1998, 80, 2733− 2736. (44) Sobolevsky, A. I.; Rosconi, M. P.; Gouaux, E. X-ray Structure, Symmetry and Mechanism of an AMPA-Subtype Glutamate Receptor. Nature 2009, 462, 745−756. (45) Sukumaran, M.; Rossmann, M.; Shrivastava, I.; Dutta, A.; Bahar, I.; Greger, I. H. Dynamics and Allosteric Potential of the AMPA Receptor N-Terminal Domain. EMBO J. 2011, 30, 972−982. (46) Dutta, A.; Shrivastava, I. H.; Sukumaran, M.; Greger, I. H.; Bahar, I. Comparative Dynamics of NMDA- and AMPA-Glutamate Receptor N-Terminal Domains. Structure 2012, 20, 1−12. (47) Balannik, V.; Menniti, F. S.; Paternain, A. V.; Lerma, J.; SternBach, Y. Molecular Mechanism of AMPA Receptor Noncompetitive Antagonism. Neuron 2005, 48, 279−288. (48) Holm, M. M.; Naur, P.; Vestergaard, B.; Geballe, M. T.; Gajhede, M.; Kastrup, J. S.; Traynelis, S. F.; Egebjerg, J. A Binding Site Tyrosine Shapes Desensitization Kinetics and Agonist Potency at GluR2. A Mutagenic, Kinetic, and Crystallographic Study. J. Biol. Chem. 2005, 280, 35469−35476. (49) Robert, A.; Armstrong, N.; Gouaux, J. E.; Howe, J. R. AMPA Receptor Binding Cleft Mutations That Alter Affinity, Efficacy, and Recovery from Desensitization. J. Neurosci. 2005, 25, 3752−3762.

(7) Johnson, K. A.; Conn, P. J.; Niswender, C. M. Glutamate Receptors as Therapeutic Targets for Parkinson’s Disease. CNS Neurol. Disord.: Drug Targets 2009, 8, 475−491. (8) Hansen, K. B.; Yuan, H.; Traynelis, S. F. Structural Aspects of AMPA Receptor Activation, Desensitization and Deactivation. Curr. Opin. Neurobiol. 2007, 17, 281−288. (9) Kumar, J.; Mayer, M. L. Functional Insights from Glutamate Receptor Ion Channel Structures. Annu. Rev. Physiol. 2013, 75, 313− 337. (10) Shahi, K.; Baudry, M. Increasing Binding Affinity of Agonists to Glutamate Receptors Increases Synaptic Responses at Glutamatergic Synapses. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 6881−6885. (11) Robert, A.; Armstrong, N.; Gouaux, J. E.; Howe, J. R. AMPA Receptor Binding Cleft Mutations That Alter Affinity, Efficacy, and Recovery from Desensitization. J. Neurosci. 2005, 25, 3752−3762. (12) Jin, R.; Clark, S.; Weeks, A. M.; Dudman, J. T.; Gouaux, E.; Partin, K. M. Mechanism of Positive Allosteric Modulators Acting on AMPA Receptors. J. Neurosci. 2005, 25, 9027−9036. (13) Zheng, F.; Erreger, K.; Low, C. M.; Banke, T.; Lee, C. J.; Conn, P. J.; Traynelis, S. F. Allosteric Interaction between the Amino Terminal Domain and the Ligand Binding Domain of NR2A. Nat. Neurosci. 2001, 4, 894−901. (14) Selvaratnam, R.; Chowdhury, S.; VanSchouwen, B.; Melacini, G. Mapping Allostery through the Covariance Analysis of NMR Chemical Shifts. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 6133−6138. (15) Manley, G.; Loria, J. P. NMR Insights into Protein Allostery. Arch. Biochem. Biophys. 2012, 519, 223−231. (16) Knapp, J. E.; Pahl, R.; Šrajer, V.; Royer, W. E., Jr. Allosteric Action in Real Time: Time-Resolved Crystallographic Studies of a Cooperative Dimeric Hemoglobin. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 7649−7654. (17) Sato, A.; Gao, Y.; Kitagawa, T.; Mizutani, Y. Primary Protein Response after Ligand Photodissociation in Carbonmonoxy Myoglobin. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 9627−9632. (18) Jin, M.; Song, G.; Carman, C. V.; Kim, Y.; Astrof, N. S.; Shimaoka, M.; Wittrup, D. K.; Springer, T. A. Directed Evolution to Probe Protein Allostery and Integrin I Domains of 200,000-fold Higher Affinity. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 5758−5763. (19) Chennubhotla, C.; Bahar, I. Markov Propagation of Allosteric Effects in Biomolecular Systems: Application to GroEL-GroES. Mol. Syst. Biol. 2006, 2, 36. (20) Vijayabaskar, M. S.; Vishveshwara, S. Interaction Energy Based Protein Structure Networks. Biophys. J. 2010, 99, 3704−3715. (21) Hansia, P.; Ghosh, A.; Vishveshwara, S. Ligand Dependent Intra and Inter Subunit Communication in Human Tryptophanyl tRNA Synthetase as Deduced from the Dynamics of Structure Networks. Mol. Biosyst. 2009, 5, 1860−1872. (22) Vishveshwara, S.; Ghosh, A.; Hansia, P. Intra and Intermolecular Communications through Protein Structure Network. Curr. Protein Pept. Sci. 2009, 10, 146−160. (23) Böde, C.; Kovács, I. A.; Szalay, M. S.; Palotai, R.; Korcsmáros, T.; Csermely, P. Network Analysis of Protein Dynamics. FEBS Lett. 2007, 581, 2776−2782. (24) Bahar, I.; Lezon, T. R.; Yang, L. W.; Eyal, E. Global Dynamics of Proteins: Bridging between Structure and Function. Annu. Rev. Biophys. 2010, 39, 23−42. (25) Kamerlin, S. C.; Warshel, A. At the Dawn of the 21st Century: Is Dynamics the Missing Link for Understanding Enzyme Catalysis? Proteins 2010, 78, 1339−1375. (26) Haliloglu, T.; Bahar, I.; Erman, B. Gaussian Dynamics of Folded Proteins. Phys. Rev. Lett. 1997, 79, 3090−3093. (27) Hinsen, K. Analysis of Domain Motions by Approximate Normal Mode Calculations. Proteins 1998, 33, 417−429. (28) Atilgan, A. R.; Durell, S. R.; Jernigan, R. L.; Demirel, M. C.; Keskin, O.; Bahar, I. Anisotropy of Fluctuation Dynamics of Proteins with an Elastic Network Model. Biophys. J. 2001, 80, 505−515. (29) Yang, L. W.; Rader, A. J.; Liu, X.; Jursa, C. J.; Chen, S. C.; Karimi, H. A.; Bahar, I. oGNM: Online Computation of Structural 8695

dx.doi.org/10.1021/jp402290t | J. Phys. Chem. B 2013, 117, 8689−8696

The Journal of Physical Chemistry B

Article

(50) Armstrong, N.; Gouaux, E. Mechanisms for Activation and Antagonism of an AMPA-Sensitive Glutamate Receptor: Crystal Structures of the GluR2 Ligand Binding Core. Neuron 2000, 28, 165− 181. (51) Armstrong, N.; Mayer, M.; Gouaux, E. Tuning Activation of the AMPA-Sensitive GluR2 Ion Channel by Genetic Adjustment of Agonist-Induced Conformational Changes. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 5736−5741. (52) Mayer, M. L.; Armstrong, N. Structure and Function of Glutamate Receptor Ion Channels. Annu. Rev. Physiol. 2004, 66, 161− 181. (53) Zhang, W.; Cho, Y.; Lolis, E.; Howe, J. R. Structure and SingleChannel Results Indicate That the Rates of Ligand Binding Domain Closing and Opening Directly Impact AMPA Receptor Gating. J. Neurosci. 2008, 28, 932−943. (54) Jin, R.; Banke, T. G.; Mayer, M. L.; Traynelis, S. F.; Gouaux, E. Structural Basis for Partial Agonist Action at Ionotropic Glutamate Receptors. Nat. Neurosci. 2003, 6, 803−810. (55) Uchino, S.; Sakimura, K.; Nagahari, K.; Mishina, M. Mutations in a Putative Agonist Binding Region of the AMPA-Selective Glutamate Receptor Channel. FEBS Lett. 1992, 308, 253−257. (56) Mano, I.; Lamed, Y.; Teichberg, V. I. A Venus Flytrap Mechanism for Activation and Desensitization of α-Amino-3-hydroxy5-methyl-4-isoxazole Propionic Acid Receptors. J. Biol. Chem. 1996, 271, 15299−15302. (57) Abele, R.; Keinanen, K.; Madden, D. R. Agonist-induced Isomerization in a Glutamate Receptor Ligand-binding Domain. A Kinetic and Mutagenetic Analysis. J. Biol. Chem. 2000, 275, 21355− 21363. (58) Sun, Y.; Olson, R.; Horning, M.; Armstrong, N.; Mayer, M.; Gouaux, E. Mechanism of Glutamate Receptor Desensitization. Nature 2002, 417, 245−253. (59) Horning, M. S.; Mayer, M. L. Regulation of AMPA Receptor Gating by Ligand Binding Core Dimers. Neuron 2004, 41, 379−388. (60) Timm, D. E.; Benveniste, M.; Weeks, A. M.; Nisenbaum, E. S.; Partin, K. M. Structural and Functional Analysis of Two New Positive Allosteric Modulators of GluA2 Desensitization and Deactivation. Mol. Pharmacol. 2011, 80, 267−280. (61) Ward, S. E.; Bax, B. D.; Harries, M. Challenges for and Current Status of Research into Positive Modulators of AMPA Receptors. Br. J. Pharmacol. 2010, 160, 181−190. (62) Stern-Bach, Y.; Bettler, B.; Hartley, M.; Sheppard, P. O.; O’Hara, P. J.; Heinemann, S. F. Agonist-selectivity of Glutamate Receptors Is Specified by Two Domains Structurally Related to Bacterial Amino Acid Binding Proteins. Neuron 1994, 13, 1345−1357. (63) Partin, K. M.; Fleck, M. W.; Mayer, M. L. AMPA Receptor Flip/ flop Mutants Affecting Deactivation, Desensitization, and Modulation by Cyclothiazide, Aniracetam, and Thiocyanate. J. Neurosci. 1996, 16, 6634−6647. (64) Yelshansky, M. V.; Sobolevsky, A. I.; Jatzke, C.; Wollmuth, L. P. Block of AMPA Receptor Desensitization by a Point Mutation Outside the Ligand-Binding Domain. J. Neurosci. 2004, 24, 4728−4736.

8696

dx.doi.org/10.1021/jp402290t | J. Phys. Chem. B 2013, 117, 8689−8696