R102Q Mutation Shifts the Salt-Bridge Network and Reduces the

Oct 24, 2014 - Analysis of the salt bridge network in both WT and the R102Q variant demonstrates that the R102Q-mutation- induced salt bridge alternat...
0 downloads 4 Views 3MB Size
Download PDF
Article pubs.acs.org/JPCB

R102Q Mutation Shifts the Salt-Bridge Network and Reduces the Structural Flexibility of Human Neuronal Calcium Sensor‑1 Protein Yuzhen Zhu,† Ying Wu,† Yin Luo,‡ Yu Zou,† Buyong Ma,§ and Qingwen Zhang*,† †

College of Physical Education & Training, Shanghai University of Sport, 399 Chang Hai Road, Shanghai, 200438, China Department of Physics, Fudan University, 220 Handan Road, Shanghai, 200433, China § Basic Science Program, Leidos Biomedical Research, Inc. Cancer and Inflammation Program, National Cancer Institute, Frederick, Maryland 21702, United States ‡

S Supporting Information *

ABSTRACT: Neuronal calcium sensor-1 (NCS-1) protein has a variety of different neuronal functions and interacts with multiple binding partners mostly through a large solvent-exposed hydrophobic crevice (HC). A single R102Q mutation in human NCS-1 protein was demonstrated to be associated with autism disease. Solution NMR study reported that this R102Q mutant had long-range chemical shift effects on the HC and the C-terminal tail (L3). To understand the influence of the R102Q mutation on the HC and L3 of NCS-1, we have investigated the conformational dynamics and the structural flexibility of wild type (WT) NCS-1 and its R102Q mutant by conducting extensive allatom molecular dynamics (MD) simulations. On the basis of six independent 450 ns MD simulations, we have found that the R102Q mutation in NCS-1 protein (1) dramatically reduces the flexibility of loops L2 and L3, (2) facilitates L3 in a more extended state to occupy the hydrophobic crevice to a larger extent, (3) significantly affects the intersegment salt bridges, and (4) changes the subspace of the free energy landscape of NCS-1 protein. Analysis of the salt bridge network in both WT and the R102Q variant demonstrates that the R102Q-mutationinduced salt bridge alternations play a critical role on the reduced flexibility of L2 and L3. These results reveal the important role of salt bridges on the structural properties of NCS-1 protein and that R102Q mutation disables the dynamic relocation of Cterminus, which may block the binding of NCS-1 protein to its receptors. This study may provide structural insights into the autistic spectrum disorder associated with R102Q mutation.

1. INTRODUCTION Physical activity and the brain have a dynamic, bidirectional relationship.1 Molecular mechanisms of exercise and brain health may involve neuronal growth factors,2 immune factors, and stress hormones,2 and many other signaling proteins like neuronal calcium sensor (NCS) proteins.3−5 Physical exercise enhances cognitive function in mice,6 and one of the important proteins linking physical activity and cognitive function is the Neuronal calcium sensor-1 (NCS-1) protein, which has been shown to render different effects of swimming training on object recognition memory tasks.3 NCS-1 protein also plays key roles in central neuronal system disorders. It was reported that NCS-1 protein is associated with various psychiatric diseases.7 For example, a rare missense in NCS-1, substituting argnine 102 for glutamine (R102Q) was identified in an autistic patient.8 NCS-1 and orthologs are expressed in all organisms from yeast to humans. A variety of different neuronal functions have been attributed to NCS-1. It has been mostly associated with central nervous system development and neurotransmission.9 NCS-1 is involved in the efficient transduction of nerve stimuli in synaptic nerve endings, where it potentiates neurotransmitter release.10 A role in development may be inferred from its © 2014 American Chemical Society

regulation of neurite outgrowth and synapse formation, which has been seen in flies, molluscs, birds and mammals et al.11,12 Studies on its other several physiological functions have also been reported, including the regulation of neurotransmission, synaptic plasticity,13 learning and memory,4,14 membrane traffic,15 and the activity of ion channels.16−18 NCS-1 protein belongs to the NCS family. NCS proteins are characterized by their structural similarities and a high (low micromolar) affinity for Ca2+,16 and are a conserved subclass of the calmodulin superfamily that triggers different biological processes to regulate signal transduction in neurons and photoreceptor cells.17−19 The structure of NCS is characterized by helix−loop−helix motifs, called EF-hand motifs, which contains four EF-hand motifs, namely EF1, EF2, EF3, and EF4. Three of the four EF hands (EF2, EF3, and EF4) bind Ca2+. EF2 and EF3 have been identified as structural sites, being occupied by Mg2+ in the absence of Ca2+, while EF4 is regarded as a sensory site, binding only Ca2+ with lower affinity.20 However, Ca2+ does not bind to EF4 in recoverin21 and Received: August 6, 2014 Revised: October 21, 2014 Published: October 24, 2014 13112

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

VILIPs.22 EF1 is not able to bind Ca2+ ions due to a conserved Cys/Pro substitution in the calcium binding motif.19,23 Four EF-hands always occur in pairs called an EF domain,24 which are defined as the N-domain and the C-domain. The N-domain consists of EF1 and EF2, and the C-domain consists of EF3 and EF4. NCS-1 has multiple binding partners25 and is involved in a range of cellular processes. Different NCS proteins can regulate distinct target proteins, and they can have highly specific, or more wide-ranging, distributions and cellular roles.16 NCS-1 is able to bind and transmit very small changes in intracellular Ca2+ concentrations by binding a large number of substrates and reshaping its structure depending on the environmental conditions.26 Once binding Ca2+, conformational changes convert the NCS-1 to its active state and thereby mediate effects of cytosolic Ca2+.27 NCS-1 contains 190 residues and is characterized by a consensus signal for N-terminal myristoylation.28−30 The N-terminal myristoylation is a lipid anchor modification, and the attachment of the lipid moiety results in an increase of hydrophobicity that anchors the protein to the membrane.31 NCS-1 interacts with a broad spectrum of interaction partners mostly through a large solvent-exposed hydrophobic crevice (HC) near the C-terminus.29,30,32 The Cterminal tail of human NCS-1 regulates the conformational stability of the activated state.33 It could initially bind to the binding site HC pocket and subsequently release the binding to the incoming NCS-1 interacting proteins. On the contrary, recent experimental study on Caenorhabditis elegans NCS-1 indicated that full length of the hydrophobic groove is required for the regulatory interactions underlying NCS-1 function, whereas the C-terminal tail of NCS-1 is not essential.34 Closely related to the above suggestions of the NCS-1 is the effect of the R102Q mutation on the structure and function of human NCS-1 protein. Previous studies9,26,33 have already demonstrated that the R102Q variant produced pronounced distant as well as local changes, particularly in helices H6 and H9 and loop L3. However, there are two conflict mechanisms of R102Q mutation. One NMR spectroscopy study suggested that the R102Q mutation affected the structure of human NCS1 particularly with an increase in the extent of conformational exchange in the C-terminus.9 However, in another study, highresolution NMR structure and chemical shift measurements indicated that R102Q mutation damped the mobility of ending loop L3 of the C-terminus.33 Delineation of the R102Q mutation and regulation of the C-terminus is of fundamental significance for allosteric proteins in general35,36 and particularly has implication for antipsychotic drug development, since human NCS-1 and its D2 dopamine receptor represent an emerging new class of molecular targets for pharmacological drug development.37 In this study, we investigate at the atomistic level the effect of R102Q mutation on the structural dynamics of human NCS-1 protein by conducting three independent 450 ns molecular dynamics (MD) simulations with explicit solvent for both WT NCS-1 and its R102Q mutant (MT), starting from three different initial structures. We have first compared the conformational dynamics and the local structural flexibility of WT/R102Q MT, then identified the salt bridges that were affected most dramatically by the R-to-Q mutation at position 102. We have found that R102Q mutation (1) dramatically reduces the flexibility of loops L2 and L3, (2) helps L3 in a more extended state, which occupies the hydrophobic crevice to a larger extent, (3) significantly alters the intersegment salt

bridges of NCS-1 protein, and (4) changes the subspace of the free energy landscape of NCS-1 protein. Detailed analysis revealed that R102Q-mutation-induced salt bridge alternation plays critical roles on the observed structural flexibility changes and the more extended state of L3. Consistent with the suggestion that R102Q mutation damped the mobility of ending loop L3 of the C-terminus,33 our study demonstrates that R102Q mutation disables the dynamics relocation of the C-terminus, which may block the binding of human NCS-1 protein to its receptors.

2. MATERIALS AND METHODS Human WT NCS-1 Protein and Its R102Q Mutant. The initial states of three MD simulations of WT NCS-1 were the first three models of 20 solution NMR structures of the unmyristoilated NCS-1 (PDB ID: 2LCP).33 The three initial states are labeled as iWT-1, iWT-2, iWT-3. The backbone RMSDs of full-length NCS-1 are 0.29 nm for the iWT-1−iWT2 pair, 0.35 nm for the iWT-1−iWT-3 pair, and 0.31 nm for the iWT-2−iWT-3 pair. The three initial conformations of the R102Q mutant were modeled by taking the three initial states (iWT-1, iWT-2, and iWT-3) of the WT species as the starting point, substituting argnine 102 for glutamine, and then energyminimizing them. We note that the X-ray crystal structure of unmyristoilated human NCS-1 protein is also available in the protein data bank (PDB ID: 2G8I).19 Here we use only the NMR structure as the starting state of our MD simulations due to the following three reasons. First, our MD simulation study is motivated by the solution NMR structure study of human NCS-1 protein, which reported that, in the absence of a binding partner, R102Q mutation affected the C-terminal tail loop L3 and the binding hydrophobic crevice (HC) in NCS-1.33 Second, in the X-ray static structure, the HC is occupied by polyethylene glycol (PEG) molecules, and this PEG binding may affect the X-ray static structure as well as the structural flexibility of NCS-1. We found that this is true for the static structure by calculating the backbone RMSD between each of the 20 NMR structure and the X-ray structure. The RMSD values are calculated for both the core structure spanning residues E11 ∼ K174 and the full-length NCS-1 protein (M1 ∼ V190) (see below for more details about its structure and amino acid sequence). It can be seen from Table S1 that the RMSD values span from 0.33 to 0.46 nm for the core structure and from 0.42 to 0.58 nm for the full-length NCS-1. Finally, the aim of our study is to investigate at the atomistic level how the R102Q mutation influences the HC and loop L3 of NCS-1 protein in aqueous solution in the absence of PEG molecules. The use of NMR structure as the starting point of our MD simulations also allows us to make direct comparison with the solution NMR study.33 Details of Molecular Dynamics Simulations. Molecular dynamics (MD) simulations have been carried out using the GROMACS 4.5.3 package38 and the CHARMM27 force field with CMAP corrections,39 in accordance with a recent MD study by Bellucci et al.26 A previous study on the evaluation of eight different force fields showed that four force fields, including CHARMM27, provide a reasonably accurate description of the native states of two small proteins with αhelix and β-sheet structures,40 close to the ensembles that were reconstructed to fit the experimental data.41 The TIP3P model of water was used to solvate protein in a PBC box of 10.07 × 6.98 × 6.70 nm3. Additional NaCl was added to the system with a concentration of 0.05 M. Three independent 450 ns MD 13113

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

Figure 1. Cartoon diagram of the NMR structure of unmyristoilated calcium-bound human NCS-1 (PDB: 2LCP) (a) and its amino acid sequence (b). For each segment, we use the same color in the structure as in the amino acid sequence. For clarity, Ca2+ ions are not shown.

Figure 2. Time evolution of three quantities of WT NCS-1 (black curve) and its R102Q MT (red curve) in three different MD trajectories (WT1, WT2, and WT3 for the WT species and MT1, MT2, and MT3 for the mutant). The three quantities include the backbone RMSDs of the core structure (E11−K174) with respect to the corresponding energy-minimized initial structures (iWT-1, iWT-2, and iWT-3) of WT NCS-1 (aligned on all backbone atoms from residues E11−K174) (a,b,c), number of backbone H-bonds (d,e,f), and percentage of helix (g,h,i).

simulations have been performed for both WT NCS-1 and the R102Q MT at 310 K in an isothermal−isobaric (NPT) ensemble. The MD simulations for WT species started from the first three conformations of the NMR structures deposited in the Protein Data Bank by Heidarsson et al.33 (PDB ID: 2LCP). The three MD simulations for R102Q MT started from the conformations mutated from the first three NMR structure models (model-1, model-2, and model-3). The solute and solvent were separately coupled to external temperature bath using the velocity rescaling method42 and pressure bath using the Parrinello−Rahman method.43 The temperature and pressure were maintained at 310 K and 1 bar using coupling constants of 0.1 and 1.0 ps, respectively. Bond lengths within protein and water molecules were respectively constrained by the LINCS 44 and SETTLE algorithms, 45 allowing an integration time step of 2 fs. The particle mesh Ewald (PME) method was used to calculate the electrostatic interaction with a real space cutoff of 1.0 nm, and the van der Waals interactions were calculated using a cutoff of 1.4 nm. Analysis Methods. Root mean square deviations (RMSDs) were calculated after the protein structure alignment. Unless

specified, the RMSD calculated in this study is the backbone RMSD of the core structure (residues E11 ∼ K174) of human NCS-1 protein. The structural stability of NCS-1 is further illustrated by the time evolution of the number of backbone Hbonds. A hydrogen bond is considered to be formed if the distance between N and O is ≤3.5 Å and the angle of N−H···O is ≥150°. The DSSP program was used to determine the secondary structure.46 Root mean square fluctuation (RMSF) was calculated for each residue with respect to the MD generated average structure in the last 250 ns. A salt bridge between a pair of oppositely charged residues is considered to be formed if the centroids of the side-chain charged groups in oppositely charged residues lie within 0.4 nm of each other.47 Distances between the centroids of the side-chain charged groups were calculated to show the change of the salt bridges. The free energy landscape (or potential mean force) was constructed using the relation −RT ln H(x,y), where H(x,y) is the histogram of two selected reaction coordinates, x and y. 13114

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

3. RESULTS AND DISCUSSION The solution structure of unmyristoilated calcium-bound human NCS-1 has been solved recently using NMR spectroscopy.33 The cartoon representation of its 3D structure (the first model of 2LCP) and its amino acid sequence are given in Figure 1. The NCS-1 structure primarily consists of nine αhelices arranged in four EF hands and three loops. According to Heidarsson et al.,33 the nine helices are helix 1 (H1) (residues E11−R18), H2 (E24−F34), H3 (A45−Q54), H4 (T62−F72), H5 (F82−S93), H6 (D98−Y108), H7 (R118−V132), H8 (E146−M155), H9 (L166−K174). The three loops are L1 (F56−P61), L2 (G133−P145), L3 (D176−V190). The eight helices H2−H9 form a hydrophobic crevice (HC), of which helices H4, H5, and H6 form the floor of the crevice, and H3 and H7 on one side and H9 on the other side form the long “walls” of the HC, and helices H2 and H8 close the HC at the opposite edges. The 3D structure shows that, in the absence of a binding partner, L3 partly occupies the HC as a ligand mimic. Prior to characterizing the structural properties of the R102Q mutant of human NCS-1 protein, we examine the conformational dynamics of WT species by performing three independent 450 ns MD simulations at 310 K (WT1, WT2, and WT3). The initial states of WT NCS-1 for the three MD simulations were the first three models of 20 NMR structures (PDB: 2LCP). Figure 2 presents the time evolution of the backbone RMSD of WT NCS-1 with respect to the corresponding minimized initial state in the three different MD trajectories. It can be seen from Figure 2a,b,c that the backbone RMSDs of the core structure (residues E11−K174) of WT NCS-1 increase quickly with time within the first 50 ns, then change gradually and stabilize respectively around 0.35, 0.40, and 0.45 nm after t = 350 ns. We note that a RMSD value of ∼0.4 nm with respect to the initial state is a bit large. To examine what happens for WT NCS-1 during the 450 ns MD simulations, we monitor the number of backbone H-bonds, the percentage of helix, and the secondary structure profile as a function of time. The number of backbone H-bonds (Figure 2d−f fluctuates around 80, and the helix percentage remains around 55% during the 450 ns duration of MD simulations. The fluctuation of the number of H-bonds corresponds to a disruption and reformation of backbone H-bonds. A similar extent of fluctuation is also observed for the percentage of helix, in which partial unfolding−refolding of helices takes place. The partial unfolding−refolding event, which involves mostly helices H4 (residues T62−F72), H8 (residues E146−M155), and H9 (residues L166−K174), can be seen from the secondary structure profile in Figure S1a−c. For example, in the trajectory of WT3 (Figure S1c), helix H8 partially unfolds and changes into a turn (in yellow) during the time period of t = 100−340 ns, and refolds into a helix after t = 340 ns. The partial unfolding−refolding of helices is accompanied by the break and reformation of the backbone hydrogen bonds, as hydrogenbond formation is the prerequisite of helix formation. The secondary structure profile in Figure S1a−c also shows that NCS-1 largely maintains its secondary structure. Thus, an RMSD value of ∼0.4 nm in Figure 2 indicates that the tertiary structure can undergo certain conformational rearrangements with time, while the secondary structure and the overall folding of WT NCS-1 remain mostly unchanged. We also calculate the backbone RMSD value for full-length NCS-1, and the results are given in Figure S2. The RMSD values during the last 250 ns in WT1, WT2, and WT3 MD

trajectories are respectively stabilized around 0.45, 0.65, and 0.55 nm. Such a large backbone RMSD value is also seen for some of the 20 NMR structures (PDB ID: 2LCP). For example, the pairwise backbone RMSD values are 0.50 nm for model-2−20 pair, 0.64 nm for model-6−10 pair, and 0.43 nm for model-11−12 pair. The pairwise backbone RMSD values for the 20 NMR structures range from 0.22 to 0.64 nm. The large RMSD values of ≥0.5 nm between the NMR structures indicate that the isolated human NCS-1 protein is very flexible in aqueous solution, and their tertiary structures can change with time. The flexible and dynamical feature of WT NCS-1 is also observed in a recent MD study by Bellucci et al.26 It was reported that, with respect to the NMR structure (model-1 in 2LCP), helices H8, H9, and H1 display large partial RMSD values of 0.5, 0.5, and 0.73 nm, respectively, while all helices are largely maintained (Figure 6 and Table 1 in the reference).26 We then compare the final structures (t = 450 ns) generated in the three different MD trajectories for both WT and R102Q MT by calculating the pairwise backbone RMSD values of the final structures (t = 450 ns) in the three different MD trajectories of WT NCS-1 (WT1, WT2, WT3) and R102Q MT (MT1, MT2, MT3). The RMSD values are given in Table S2, calculated for both the core structure (residues E11−K174) and the full-length NCS-1 (M1−V190). Table S2 shows that for the WT species, the pairwise RMSD value spans from 0.34 to 0.49 nm for E11−K174 and from 0.44 to 0.62 nm for M1− V190. For the R102Q MT, the pairwise RMSD value ranges from 0.38 to 0.47 nm for E11−K174 and from 0.38 to 0.53 nm for M1−V190. Considering that the pairwise RMSD values of the 20 solution NMR structures can reach 0.64 nm, the RMSD values in Table S2 indicate that the WT/MT NCS-1 converges to a similar state in the three different MD runs. We also calculated the backbone RMSD values of the relaxed and final structures of WT/MT species with respect to the Xray structure. The RMSD values are calculated for both the core segment spanning residues E11−K174 and the full-length NCS-1 (M1−V190). The calculated RMSD values are given in Table S3. We see that the RMSD values of the relaxed WT/MT structures with respect to the X-ray structure span from ∼0.37 to 0.44 nm for E11−K174 and from ∼0.50 to 0.57 nm for M1− V190. For the final structures (t = 450 ns), the RMSD values of E11−K174 with respect to the X-ray structure span from ∼0.36 to 0.40 nm for WT species, while they span from 0.33 to 0.37 nm for MT species, slightly lower than the WT species. The RMSD values of M1−V190 are quite similar for WT and MT species, both in the range from 0.43 to 0.53 nm. Overall, these RMSD values are comparable to the RMSDs (Table S1) of the 20 NMR structure models with respect to the X-ray crystal structure. R102Q Variant Displays Similar Global Conformational Dynamics and Structural Stability as WT Species. After confirming that the force-field used for WT NCS-1 can describe the conformational dynamics similar to those reported in recent experimental and simulation studies,26,33 we turn to investigate the structural properties of R102 mutant using the same protein force field. To this aim, we have performed three independent MD runs (MT1, MT2, and MT3) initiated from three different initial states (iMT-1, iMT-2, and iMT-3). The three initial conformations of the R102Q mutant were modeled using the three initial states (iWT-1, iWT-2, and iWT-3) of the WT species and then were energy-minimized. The detailed conformational dynamics of R102Q mutant can be seen from the time evolution of the backbone RMSD, number of 13115

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

phobic Crevice by L3. A recent experimental study reported that in the R102Q mutant, residues I179, V180, L183, and Y186 of L3 have large chemical shift changes, suggesting that the HC is either more exposed to water or more occupied by L3.33 To examine the influence of R102Q mutation on the L3 occupancy in the HC at an atomistic level, we calculated the Cα−Cα distance between starting and ending residues of L3 (D176 and V190) in both WT species and R102Q MT. Figure 4 presents the probability distributions of the D176−V190 distance in the three different MD runs for both systems. We have found that the distance probability distribution peaks are centered at 2.8 nm (Figure 4a), 1.0 and 1.6 nm (Figure 4b), and 1.4 nm (Figure 4c) in the three WT MD runs, indicating that L3 in WT NCS-1 can be either in extended state (Figure 4a) or in collapsed state (Figure 4b,c). In contrast, the D176− V190 distance distribution peaks in R102Q mutant are located at 3.0 nm (Figure 4d), 2.4 nm (Figure 4e), and 2.2 nm (Figure 4f), implying that L3 primarily adopts an extended state (Figure 4d,e,f). The high probability of the collapsed state of L3 in WT species indicates that HC is only partially occupied by L3, while the more extended state of L3 in R102Q mutant indicates that HC is more occupied by L3 with respect to that in WT species. Our deduction is supported by the representative snapshots of WT NCS-1 (Figure 4g) and the R102Q MT (Figure 4h). This finding supports the hypothesis by Bellucci et al. that a mutation that consists in replacing Arg102 with an uncharged residue could affect the mobility of H9 and L3, compromising the docking process of L3 into the HC.26 Noting that this hypothesis was based on their finding that the stabilization of inter- versus intrahelix salt bridges involving D98 (i.e., D98K174 versus D98-R102) relies on the possibility of D98 interacting with R102 or K174.26 Of particular interest is that a short α-helix spanning residues P177−L185 is observed in the R102Q mutant. Secondary structure profiles in Figure S1d,e show that this helical structure is transient, changing between a disordered coil and an α-helix. It could be predicted that the formation of this short helix would reduce the flexibility of L3, as demonstrated in the RMSF plot in Figure 3. It is known that the NCS-1 plays an important role in neurotransmitter release,10 and it interacts with a broad spectrum of interaction partners mostly through a large solvent-exposed HC.29,30,32 The different extents of L3 occupancy in the HC of WT species and of R102Q MT observed in Figure 4 reveal different extents of exposure of the HC. The partially bound HC by L3 in WT species would facilitate release of L3 to allow incoming ligand binding, while the more occupied HC by L3 in R102Q mutant could hamper L3 release and inhibit the ligand binding, which cause functional deficits. Recently, R102Q mutation was found in a patient suffering from autistic spectrum disorder,8 and this variant was demonstrated to have functional deficits.9 Our results provide am atomistic mechanism of the R102Q mutation-associated functional deficits of human NCS-1 protein. R102Q Mutation Significantly Affects the Intersegment Salt Bridges and Alters the Subspace of the Free Energy Landscape of Human NCS-1 Protein. Salt bridges play an important role in stabilizing the structures of proteins,48,49 in both buried50,51 and solvent-exposed locations.52−55 Salt bridges can also control protein allosteric dynamics and functions.56−58 NCS-1 is a highly charged protein, containing 24 positively charged and 33 negatively

backbone H-bonds, and the percentage of helix (red curves in Figure 2). It can be seen from Figure 2a,b,c (red curve) that the backbone RMSDs for the core structure of R102Q mutant with respect to the initial structure of WT NCS-1 stabilize around 0.35, 0.4, and 0.35 nm, respectively, slightly smaller than that of WT species, indicating similar structural stabilities as WT species. The curves of the number of backbone H-bonds and of the percentage of helix as a function of time for R102Q mutant overlap well with those of WT species, revealing similar global conformational dynamics of R102Q MT with WT NCS-1 protein. We also plot the secondary structure profile of R102Q mutant (Figure S1d,e,f). It can be seen from this figure that the secondary structures of R102Q MT are mostly maintained in all of three MD trajectories. Interestingly, the formation of a short helix in the C-terminal tail is observed in two (MT1 and MT2) out of the three MD trajectories, particularly in trajectory MT2. R102Q Mutation Results in a Reduced Flexibility of Loops L2 and L3. After examining the global structural changes of NCS-1 protein by the R102Q mutation, we further investigate the local conformational alternations by calculating the Cα root-mean-square fluctuation (RMSF) relative to the average MD-generated structure. Figure 3 compares the Cα-

Figure 3. Cα-RMSF of each residue in WT species and R102Q mutant in the three independent MD runs (a,b,c). For comparison, the average Cα-RMSF over the three MD runs is given in panel d. CαRMSF values are calculated using the last 200 ns data of each MD trajectory for both systems.

RMSF of R102Q MT against the WT species in the three independent MD runs (a,b,c) as well as the average Cα-RMSF over the three MD runs (d). Except for the N- and C-terminal regions, loop L2 (residues G133−P145) exhibits overall larger fluctuations than other regions of the protein, indicating that L2 has the largest local flexibility. Interestingly, loop L2 in the mutant displays (red curves) smaller RMSF values than that in the WT species, indicating that R102Q mutation results in a reduced flexibility of L2. A dampened flexibility is also seen for the L3 segment (D176−V190), although different extent of reduction in flexibility is observed in the three MD runs (MT1, MT2, and MT3). This finding is consistent with a recent NMR study showing that L3 has dampened mobility in R102Q mutant.33 R102Q Mutation Induces L3 to Adopt a More Extended State, Leading to a More Occupied Hydro13116

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

Figure 4. Probability distributions of the Cα−Cα distance between the starting and ending residues of L3 (D176 and V190) for WT species (a−c) and R102Q mutant (d−f) in the three different MD runs. The D176−V190 distances are calculated for the last 250 ns of each MD trajectory for both systems. The representative structures of WT NCS-1 (g) and its R102Q mutant (h) in surface representations are given, where L3 is in cartoon representation (orange) and the HC is shown in white. In panels g and h, hydrophobic residues are in white, polar residues are in green, positively charged residues are in blue, and negatively charged residues are in red.

Figure 5. Time evolution of the three most-affected intersegment salt bridges D98−K174, D176−R148, and D176−R151 between H6, H8, H9, and L3 segments. Smoothed RMSD data are reported as bright red, bright blue, and black curves.

both D176-R148 and D176-R151 salt bridges almost disappear after 200 ns, while in the R102Q MT, the two salt bridges coexist or exist exclusively. These results suggest that R102Q mutation significantly influences the D98−K174, D176−R148, and D176−R151 salt bridges in H6, H8, H9, and L3 segments, and these salt bridges are dynamically stable. The influence of R102Q mutation on the three salt bridges (D98−K174, D176−R148, and D176−R151) can be seen clearly from the free energy landscape (or potential of mean force) of WT NCS-1 (Figure 6a) and its R102Q MT (Figure 6b), projected on two reaction coordinates: the D98−K174 distance and the minimum distance of R148−D176 and R151− D176. Here, the distance between the two oppositely charged residues refers to the distance between the centroids of the side-chain charged groups of the two residues, and the minimum distance of R148−D176 and R151−D176 refers to the smaller one between the R148−D176 distance and the R151−D176 distance. Three well-separated minimum-energy

charged residues on protein surface. Thus, it is particularly intriguing to examine whether R102Q mutation affects the salt bridges formed in NCS-1 protein. By comparing the chemical shifts of WT NCS-1 and its R102Q MT, Kragelund et al. reported that residues with severe chemical shift perturbations were located around the mutation site in H6, H9, and L3. Thus, we start to probe the salt bridges located in those three segments and take three of the most-affected salt bridges for further analysis. The residue pairs forming the three salt bridges are D98 in H6 and K174 in H9, D176 in L3 and R148 in H8, and D176 in L3 and R151 in H8. We investigate the dynamics of these salt bridges by monitoring their time evolution in each MD trajectory (Figure 5). It can be seen that in WT species (see Figure 5a,b,c, the interhelix salt bridge D98−K174 (black curve) between H6 and H9 can break and reform with simulation time, indicating a dynamic feature of this salt bridge, whereas in the R102Q MT (see Figure 5d,e,f, it almost vanishes during the full duration of MD simulations. In WT NCS-1, 13117

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

Figure 6. Free energy landscape of WT NCS-1 and its R102Q mutant, projected on the D98−K174 distance and the minimum distance of R148− D176 and R151−D176. The unit of free energy is kcal/mol.

Figure 7. R102Q mutation shifts the salt bridge network of WT NCS-1 and its R102Q mutant. The salt-bridge probability maps are given in panel a for WT and in panel b for R102Q MT. Structures in panels c and d show the locations of the residues that involve the formation of the salt bridges that were mostly affected by R102Q mutation. The two structures were respectively taken from the MD trajectory of WT1 and MT1, with L3 in an extended state in the mutant. The Cα-atoms of the residues involving the formation of some of the most-affected salt bridges were represented by blue and red beads: blue for positively charged residues, and red for negatively charged residues.

bridge between helices H6 and H9. Basin B located at (1.1 nm, 0.3 nm) corresponds to the formation of R148−D176 or/and R151−D176 salt bridges, while basin C located at (1.2 nm, 1.0 nm) corresponds to the break of both D98−K174 and R148− D176 or/and R151−D176 salt bridges. With respect to the WT system, significant changes are observed in the free energy

basins (labeled as A, B, C for WT, and D, E, F for R102Q mutant) are observed in the free energy surface of WT/MT NCS-1. They are centered around (dD98−K174, min[dR148−D176, and dR151−D176]) values of (0.25 nm, 0.9 nm), (1.1 nm, 0.3 nm), and (1.2 nm, 1.0 nm). The deepest basin (basin A) at (0.25 nm, 0.9 nm) corresponds to the formation of the D98−K174 salt 13118

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

Table 1. List of the Salt Bridges (along with Their Populated Percentage (P) of Time) That Were Influenced Most Dramatically by the R102Q Mutationa salt bridge

E99− R(Q)102 (H6− H6)

R118−D150 (H7− H8)

D98−K174 (H6− H9)

E74−K106 (L45− H6)

K63−D126 (H4− H7)

P

86.4% 0.0% E26−R94 (H2−L56)

62.1% 20.0% K100−D187 (H6−L3)

P

62.0% 31.9% K9−E24 (LN−L12)

28.9% 0.0% K53−D60 (H3−L1)

P

28.6% 15.1% K158−E171 (L89−H9)

8.5% 54.8% E140−R148 (L2−H8)

4.8% 47.6% D44−R79 (L23−L45)

0.0% 31.8% K3−E81 (LN−L45)

1.2% 31.7% K63−D187 (H4−L3)

46.6% 74.9%

9.0% 33.2%

54.0% 78.0%

0.0% 18.5%

0.0% 15.5%

WT MT salt bridge WT MT salt bridge

WT MT salt bridge P

WT MT

E74−R(Q)102 (L45− H6)

40.4% 0.4% R79−E81 (L45−L45)

57.0% 22.4% E142−R148 (L2−H8)

34.2% 1.1% R148−D187 (H8−L3)

32.7% 0.0% K63−D123 (H4−H7)

23.9% 0.0% R151−D176 (H8−L3)

24.0% 1.1% R94−D187 (L56−L3)

18.3% 0.0% R148−D176 (H8−L3)

19.3% 5.3% R94−E99 (L56−H6) 0.0% 28.7%

a

The salt-bridge populations were calculated using the last 200 ns data from three different MD trajectories for both WT NCS-1 and the R102Q MT. The segments where the two salt-bridging residues locate are given in the parentheses. We use LN to denote the disordered loop in the Nterminus, and Lmn to denote the loop between helices Hm and Hn, where m and n are integers.

surface of the R102Q MT (Figure 6b). The basin (basin D) located at (0.25 nm, 0.9 nm) almost vanishes, indicating the disappearance of the D98−K174 salt bridge, while the basin (basin E) centered at (1.1 nm, 0.3 nm) becomes much deeper, reflecting more populated R148−D176 or/and R151−D176 salt bridges. The basin centered at (1.2 nm, 1.0 nm) shifts to (0.9 nm, 0.7 nm) (basin F) and becomes shallow. The comparison of the free energy surface in Figure 6a,b demonstrates that R102Q mutation changes the potential of mean force of NCS-1 protein. After examining the free energy landscape and dynamics of the three key salt bridges, we investigate all the salt bridges formed in WT and R102Q MT. To this aim, we calculated the formation probabilities of salt bridges between all positively and negatively charged residues in WT NCS-1 and the R102Q MT. The salt-bridge probability maps are given in Figure 7 for WT species (a) and R102Q MT (b), and significant differences are seen in these two maps. To find out the salt bridges that were influenced most dramatically by the R102 mutation, we compare the formation probability of each salt bridge in WT and R12Q MT. A salt bridge is considered as the most affected one if the difference between its formation probability in WT NCS-1 and that in the R102Q MT is greater than the average value (13.4%) of salt bridge probability difference of all salt bridges. Table 1 lists all the most affected salt bridges as well as their population in both WT NCS-1 and the R102Q MT. Some of them (E99−R102, R118−D150, D98−K174, K100−D187, E26−R94, K63−D123, K63−D126, and D44−R79) were also reported in a recent MD study by Bellucci et al.26 The population for the three above-mentioned key salt bridges (D98−K174, R148−D176, and R151−D176 in H6, H8, H9, and L3; see Figure 7) are 40.4% versus 0.4%, 1.2% versus 31.7%, and 4.8% versus 47.6% in WT and MT species, respectively. It can be seen from Table 1 that, in addition to these three salt bridges, there are another 20 most-affected salt bridges located in the protein segments including N-terminal domain helices H3 and H4, C-terminal domain helices H6, H7, H8, and H9, and L2 connecting helices H7 and H8. Except for E99−R102 and R79−E81 salt bridges formed by residues that belong to the same protein segment, all the other salt bridges are intersegment salt bridges, revealing that R102Q mutation predominantly influences the salt bridges formed by residues

that belong to distinct protein segments. The folding mechanism of human NCS-1 protein has been studied recently by Heidarsson et al. using the optical tweezers technique.59 This experimental study reported that the complete folding of the C domain (containing helices H6, H7, H8 and H9) is crucial for subsequent folding of the N domain of NCS-1. Our multiple MD simulations demonstrated that the interhelix salt bridges D98−K174 (between helices H6 and H9) and R118− D150 (between helices H7 and H8) play an important role on the conformational dynamics of WT NCS-1, revealing the importance of helices H6−H9 on the overall folding of NCS-1 protein, consistent with the study by Heidarsson et al.59 NCS-1 protein consists of two domains: N domain (residues M1−S93) and C domain (residues D98−V190). Of particular interest is that three of the most-affected salt bridges are interdomain salt bridges, including K63−D123 and K63−D126 salt bridges between H4 and H7, and the K63−D187 salt bridge between H4 and L3 (see Figure 7c,d for their locations in the protein). Their populations in WT NCS-1 are respectively 34.2%, 19.3%, and 0.0%, whereas they shift to 1.1%, 5.3%, and 15.5% in the R102Q variant, respectively. The significant change of the population of these interdomain salt bridges would influence the interdomain interactions, which may affect protein dynamics and the ligand binding selectivity of NCS-1. It can be seen from Table 1 that the formation of six out of the total 23 most-affected salt bridges involves residues D176 (the first residue of L3) and D187 (the C-terminal residue of L3) (see Figure 7). The six salt bridges are K100−D187, R148−D187, R151−D176, R94−D187, R148−D176, and K63−D187 (see Figure 7). Their population in WT NCS-1 and in the R102Q MT are 28.9% versus 0.0%, 18.3% versus 0.0%, 4.8% versus 47.6%, 0.0% versus 31.8%, 1.2% versus 31.7%, and 0.0% versus 15.5%, respectively. These data indicate that R102Q mutation disabled K100−D187 and R148−D187 salt bridges, while it dramatically enhanced the population of R151−D176 and R148−D176 salt bridges and created two new salt bridges (R94−D187 and K63−D187). In the R102Q mutant, the break of K100−D187 and R148−D187 salt bridges would diminish the attraction between the C-terminal residue D187 of L3 and the C-domain helices H6 and H8 (see Figure 7), while the increased population of R151−D176 and R148− D176 would strengthen the interaction between the N-terminal 13119

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

NCS-1. Two figures present the secondary structure profiles of WT/MT NCS-1 and the backbone RMSD of the full-length WT NCS-1 protein as a function of simulation time in three different MD trajectories. This material is available free of charge via the Internet at http://pubs.acs.org.

residue D176 of L3 and the C-domain helix H8. The observation of new salt bridge formation between D187 and K63/R94 reflects the increased interaction between the Cterminal residue D187 of L3 and the N-domain segments of the mutant, which would facilitate L3 to adopt an extended and restricted conformation. The reduced flexibility of L2 in the R102Q mutant can also be partially attributed to the mutationinduced salt bridge alternations. As can be seen in Table 1, two of most-affected salt bridges (E142−R148 and E140−R148) involves two L2 residues E140 and E142. Their populations are respectively 24.0% and 9.0% in WT NCS-1, but shifted to 1.1% and 33.2% in the R102Q MT. As residue E140 is located closer to the middle of L2 than E142, the E140−R148 salt bridge with an increased formation probability in the mutant would limit the motion of L2, thus reducing its flexibility. On the other hand, the formation of transient helix spanning residues E140− T144 in R102Q MT (see Figure 7d) would also play a role on the decreased flexibility of L2.

■

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

■

REFERENCES

We thank Drs. Guanghong Wei and Cong Guo for helpful discussion. B.M acknowledges the financial support from NCI, NIH, under contract number HHSN261200800001E.

4. CONCLUSIONS NCS-1 protein can control physical activity and cognitive function and is implicated in central neuronal system disorders. The R102Q mutation was identified in an autistic patient. We have investigated the influence of R102Q mutation on the conformational dynamics, the structural stability, and the free energy landscape of human NCS-1 protein by performing six independent 450 ns MD simulations with explicit solvent at 310 K. Our multiple extensive MD simulations initiated from three different states shows that R102Q variant displays similar structural stability as the WT species, consistent with experimental study.33 R102Q variant is found to exhibit similar global stability as WT species. However, a significantly reduced structural flexibility is observed in loops L2 and L3. We found that L3 has a preference to adopt a collapsed sate in the WT species, while it prefers to have a relatively extended state in the mutant. The adoption of extended state by L3 in R102Q mutant facilitates L3 to occupy the hydrophobic crevice to a large extent, which may inhibit the binding of interaction partners of NCS-1 protein. Analysis of salt bridge population in WT NCS-1 and in the R102Q mutant demonstrates that R102Q mutation dramatically changes the intersegment and interdomain salt bridge network. The shift of salt bridge network in R102Q mutant also affects the free energy surface of NCS-1 protein. The increased salt bridge populations of D176 of L3 with the C-domain residues R148/R151 and that of D187 of L3 with the N-domain residues K63/R94 would help L3 in an extended state and thus restrict its mobility. The formation of salt bridge E140-R148 in the mutant may limit the flexibility of L2. The significantly changed salt bridge network induced by R102Q mutation would affect the intersegment and interdomain interaction, which may lead to structural and functional deficits of NCS-1 protein. Our study suggests a possible way to restore the function of R102Q mutant of NCS-1 protein is to make the C-terminus more flexible to allow NCS-1 to interact with its binding ligands.

■

AUTHOR INFORMATION

(1) Loprinzi, P. D.; Herod, S. M.; Cardinal, B. J.; Noakes, T. D. Physical activity and the brain: A review of this dynamic, bi-directional relationship. Brain Res. 2013, 1539, 95−104. (2) White, L. J.; Castellano, V. Exercise and brain health Implications for multiple sclerosis: Part 1Neuronal growth factors. Sports Med. 2008, 38, 91−100. (3) Drumond, L. E.; Mourao, F. A.; Leite, H. R.; Abreu, R. V.; Reis, H. J.; Moraes, M. F.; Pereira, G. S.; Massensini, A. R. Differential effects of swimming training on neuronal calcium sensor-1 expression in rat hippocampus/cortex and in object recognition memory tasks. Brain Res. Bull. 2012, 88, 385−391. (4) Saab, B. J.; Georgiou, J.; Nath, A.; Lee, F. J.; Wang, M.; Michalon, A.; Liu, F.; Mansuy, I. M.; Roder, J. C. NCS-1 in the dentate gyrus promotes exploration, synaptic plasticity, and rapid acquisition of spatial memory. Neuron 2009, 63, 643−656. (5) Dason, J. S.; Romero-Pozuelo, J.; Atwood, H. L.; Ferrus, A. Multiple roles for frequenin/NCS-1 in synaptic function and development. Mol. Neurobiol. 2012, 45, 388−402. (6) Suijo, K.; Inoue, S.; Ohya, Y.; Odagiri, Y.; Takamiya, T.; Ishibashi, H.; Itoh, M.; Fujieda, Y.; Shimomitsu, T. Resistance exercise enhances cognitive function in mouse. Int. J. Sports Med. 2013, 34, 368−375. (7) Braunewell, K. H. The darker side of Ca2+ signaling by neuronal Ca2+-sensor proteins: from Alzheimer’s disease to cancer. Trends Pharmacol. Sci. 2005, 26, 345−351. (8) Piton, A.; Michaud, J. L.; Peng, H.; Aradhya, S.; Gauthier, J.; Mottron, L.; Champagne, N.; Lafreniere, R. G.; Hamdan, F. F.; Joober, R.; Fombonne, E.; Marineau, C.; Cossette, P.; Dube, M. P.; Haghighi, P.; Drapeau, P.; Barker, P. A.; Carbonetto, S.; Rouleau, G. A. Mutations in the calcium-related gene IL1RAPL1 are associated with autism. Hum. Mol. Genet. 2008, 17, 3965−3974. (9) Handley, M. T.; Lian, L. Y.; Haynes, L. P.; Burgoyne, R. D. Structural and functional deficits in a neuronal calcium sensor-1 mutant identified in a case of autistic spectrum disorder. PLoS One 2010, 5, e10534. (10) Tsujimoto, T.; Jeromin, A.; Saitoh, N.; Roder, J. C.; Takahashi, T. Neuronal calcium sensor 1 and activity-dependent facilitation of P/ Q-type calcium currents at presynaptic nerve terminals. Science 2002, 295, 2276−2279. (11) Iketani, M.; Imaizumi, C.; Nakamura, F.; Jeromin, A.; Mikoshiba, K.; Goshima, Y.; Takei, K. Regulation of neurite outgrowth mediated by neuronal calcium sensor-1 and inositol 1,4,5-trisphosphate receptor in nerve growth cones. Neuroscience 2009, 161, 743−752. (12) Romero-Pozuelo, J.; Dason, J. S.; Atwood, H. L.; Ferrus, A. Chronic and acute alterations in the functional levels of Frequenins 1 and 2 reveal their roles in synaptic transmission and axon terminal morphology. Eur. J. Neurosci. 2007, 26, 2428−2443.

ASSOCIATED CONTENT

S Supporting Information *

Three tables show the backbone RMSD values of the 20 solution NMR structure models, the three relaxed and final structures of WT/MT NCS-1 with respect to the X-ray crystal structure, and the pairwise RMSD values of the final structures generated in the three different MD trajectories of WT/MT 13120

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

(13) Sippy, T.; Cruz-Martin, A.; Jeromin, A.; Schweizer, F. E. Acute changes in short-term plasticity at synapses with elevated levels of neuronal calcium sensor-1. Nat. Neurosci. 2003, 6, 1031−1038. (14) Gomez, M.; De Castro, E.; Guarin, E.; Sasakura, H.; Kuhara, A.; Mori, I.; Bartfai, T.; Bargmann, C. I.; Nef, P. Ca2+ signaling via the neuronal calcium sensor-1 regulates associative learning and memory in C. elegans. Neuron 2001, 30, 241−248. (15) Haynes, L. P.; Thomas, G. M.; Burgoyne, R. D. Interaction of neuronal calcium sensor-1 and ADP-ribosylation factor 1 allows bidirectional control of phosphatidylinositol 4-kinase beta and transGolgi network-plasma membrane traffic. J. Biol. Chem. 2005, 280, 6047−6054. (16) Burgoyne, R. D. Neuronal calcium sensor proteins: Generating diversity in neuronal Ca2+ signalling. Nat. Rev. Neurosci. 2007, 8, 182− 193. (17) Burgoyne, R. D.; Haynes, L. P. Understanding the physiological roles of the neuronal calcium sensor proteins. Mol. Brain 2012, 5, 2. (18) Weiss, J. L.; Hui, H.; Burgoyne, R. D. Neuronal calcium sensor-1 regulation of calcium channels, secretion, and neuronal outgrowth. Cell Mol. Neurobiol. 2010, 30, 1283−1292. (19) Bourne, Y.; Dannenberg, J.; Pollmann, V.; Marchot, P.; Pongs, O. Immunocytochemical localization and crystal structure of human frequenin (neuronal calcium sensor 1). J. Biol. Chem. 2001, 276, 11949−11955. (20) Aravind, P.; Chandra, K.; Reddy, P. P.; Jeromin, A.; Chary, K. V.; Sharma, Y. Regulatory and structural EF-hand motifs of neuronal calcium sensor-1: Mg2+ modulates Ca2+ binding, Ca2+induced conformational changes, and equilibrium unfolding transitions. J. Mol. Biol. 2008, 376, 1100−1115. (21) Ames, J. B.; Porumb, T.; Tanaka, T.; Ikura, M.; Stryer, L. Amino-terminal myristoylation induces cooperative calcium binding to recoverin. J. Biol. Chem. 1995, 270, 4526−4533. (22) Li, C.; Pan, W.; Braunewell, K. H.; Ames, J. B. Structural analysis of Mg2+ and Ca2+ binding, myristoylation, and dimerization of the neuronal calcium sensor and visinin-like protein 1 (VILIP-1). J. Biol. Chem. 2011, 286, 6354−6366. (23) Burgoyne, R. D. The neuronal calcium-sensor proteins. Biochim. Biophys. Acta 2004, 1742, 59−68. (24) Gifford, J. L.; Walsh, M. P.; Vogel, H. J. Structures and metalion-binding properties of the Ca2+-binding helix−loop−helix EF-hand motifs. Biochem. J. 2007, 405, 199−221. (25) Haynes, L. P.; Fitzgerald, D. J.; Wareing, B.; O’Callaghan, D. W.; Morgan, A.; Burgoyne, R. D. Analysis of the interacting partners of the neuronal calcium-binding proteins L-CaBP1, hippocalcin, NCS-1 and neurocalcin delta. Proteomics 2006, 6, 1822−1832. (26) Luca Bellucci, S. C.; Rosa Di Felice; Emanuele Paci. The Structure of Neuronal Calcium Sensor-1 in SolutionRevealed by Molecular Dynamics Simulations. PLoS One 2013, 8, e74383. (27) Ikura, M. Calcium binding and conformational response in EFhand proteins. Trends Biochem. Sci. 1996, 21, 14−17. (28) Johnson, D. R.; Bhatnagar, R. S.; Knoll, L. J.; Gordon, J. I. Genetic and biochemical studies of protein N-myristoylation. Annu. Rev. Biochem. 1994, 63, 869−914. (29) Hendricks, K. B.; Wang, B. Q.; Schnieders, E. A.; Thorner, J. Yeast homologue of neuronal frequenin is a regulator of phosphatidylinositol-4-OH kinase. Nat. Cell Biol. 1999, 1, 234−241. (30) Strahl, T.; Huttner, I. G.; Lusin, J. D.; Osawa, M.; King, D.; Thorner, J.; Ames, J. B. Structural insights into activation of phosphatidylinositol 4-kinase (Pik1) by yeast frequenin (Frq1). J. Biol. Chem. 2007, 282, 30949−30959. (31) Peitzsch, R. M.; McLaughlin, S. Binding of acylated peptides and fatty acids to phospholipid vesicles: Pertinence to myristoylated proteins. Biochemistry 1993, 32, 10436−10443. (32) Kabbani, N.; Negyessy, L.; Lin, R.; Goldman-Rakic, P.; Levenson, R. Interaction with neuronal calcium sensor NCS-1 mediates desensitization of the D2 dopamine receptor. J. Neurosci. 2002, 22, 8476−8486. (33) Heidarsson, P. O.; Bjerrum-Bohr, I. J.; Jensen, G. A.; Pongs, O.; Finn, B. E.; Poulsen, F. M.; Kragelund, B. B. The C-terminal tail of

human neuronal calcium sensor 1 regulates the conformational stability of the Ca2+-activated state. J. Mol. Biol. 2012, 417, 51−64. (34) Martin, V. M.; Johnson, J. R.; Haynes, L. P.; Barclay, J. W.; Burgoyne, R. D. Identification of key structural elements for neuronal calcium sensor-1 function in the regulation of the temperaturedependency of locomotion in C. elegans. Mol. Brain 2013, 6, 39. (35) Gunasekaran, K.; Ma, B.; Nussinov, R. Is allostery an intrinsic property of all dynamic proteins? Proteins 2004, 57, 433−443. (36) Ma, B.; Tsai, C. J.; Haliloglu, T.; Nussinov, R. Dynamic allostery: Linkers are not merely flexible. Structure 2011, 19, 907−917. (37) Kabbani, N.; Woll, M. P.; Nordman, J. C.; Levenson, R. Dopamine receptor interacting proteins: Targeting neuronal calcium sensor-1/D2 dopamine receptor interaction for antipsychotic drug development. Curr. Drug Targets 2012, 13, 72−79. (38) Hess, B.; Carsten, K.; van der Spoel, D.; Lindahl, E. GROMACS 4: Algorithms for highly efficient, load-balanced, and scalable molecular simulation. J. Chem. Theory Comput. 2008, 4, 435−447. (39) Bjelkmar, P.; Larsson, P.; Cuendet, M. A.; Hess, B.; Lindahl, E. Implementation of the CHARMM force field in GROMACS: Analysis of protein stability effects from correction maps, virtual interaction sites, and water models. J. Chem. Theory Comput. 2010, 6, 459−466. (40) Lindorff-Larsen, K.; Maragakis, P.; Piana, S.; Eastwood, M. P.; Dror, R. O.; Shaw, D. E. Systematic validation of protein force fields against experimental data. PLoS One 2012, 7, e32131. (41) Lange, O. F.; Lakomek, N. A.; Fares, C.; Schroder, G. F.; Walter, K. F.; Becker, S.; Meiler, J.; Grubmuller, H.; Griesinger, C.; de Groot, B. L. Recognition dynamics up to microseconds revealed from an RDC-derived ubiquitin ensemble in solution. Science 2008, 320, 1471− 1475. (42) Bussi, G.; Donadio, D.; Parrinello, M. Canonical sampling through velocity rescaling. J. Chem. Phys. 2007, 126, 014101. (43) Parrinello, M. R. A. Polymotphic transitions in single-crystals A new molecular-dynamics method. J. Appl. Phys. 1981, 52, 7182− 7190. (44) Hess, B.; Bekker, H.; Berendsen, H. J. C.; Fraaije, J. G. E. M. LINCS: A linear constraint solver for molecular simulations. J. Comput. Chem. 1997, 18, 1463−1472. (45) Miyamoto, S. K.; PA. Settle- an analytical version of the shake and rattle algorithm for rigid water models. J. Comput. Chem. 1992, 13, 952−962. (46) Kabsch, W. S. C. Dictionary of potein secondary structure pattern -recognition of hydrogen -bonded and geometrical features. Biopolymers 1983, 22, 2577−2637. (47) Ma, B.; K, S.; Tsai, C. J.; Nussinov, R. Folding funnels and binding mechanisms. Protein Eng. 1999, 12, 713−720. (48) Kumar, S.; Ma, B.; Tsai, C. J.; Nussinov, R. Contribution of Salt Bridges Toward Protein Thermostability. J. Biomol. Struct. Dyn. 2000, 11, 79−85 (conversation). (49) Kumar, S.; Ma, B.; Tsai, C. J.; Nussinov, R. Electrostatic strengths of salt bridges in thermophilic and mesophilic glutamate dehydrogenase monomers. Proteins 2000, 38, 368−383. (50) Waldburger, C. D.; S, J.; Sauer, R. T. Are buried salt bridges important for protein stability and conformational specificity? Nat. Struct. Biol. 1995, 2, 122−128. (51) Tissot, A. C.; Vuilleumier, S.; Fersht, A. R. Importance of two buried salt bridges in the stability and folding pathway of barnase. Biochemistry 1996, 35, 6786−6794. (52) Makhatadze, G. I.; Loladze, V. V.; Ermolenko, D. N.; Chen, X.; Thomas, S. T. Contribution of surface salt bridges to protein stability: Guidelines for protein engineering. J. Mol. Biol. 2003, 327, 1135−1148. (53) Ibarra-Molero, B.; Zitzewitz, J. A.; Matthews, C. R. Salt-bridges can stabilize but do not accelerate the folding of the homodimeric coiled-coil peptide GCN4-p1. J. Mol. Biol. 2004, 336, 989−996. (54) Williams, D. V.; Byrne, A.; Stewart, J.; Andersen, N. H. Optimal salt bridge for Trp-cage stabilization. Biochemistry 2011, 50, 1143− 1152. (55) Andersson, H. S.; Figueredo, S. M.; Haugaard-Kedstrom, L. M.; Bengtsson, E.; Daly, N. L.; Qu, X.; Craik, D. J.; Ouellette, A. J.; Rosengren, K. J. The alpha-defensin salt-bridge induces backbone 13121

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122

The Journal of Physical Chemistry B

Article

stability to facilitate folding and confer proteolytic resistance. Amino Acids 2012, 43, 1471−1483. (56) van Wijk, S. J.; Melquiond, A. S.; de Vries, S. J.; Timmers, H. T.; Bonvin, A. M. Dynamic control of selectivity in the ubiquitination pathway revealed by an ASP to GLU substitution in an intra-molecular salt-bridge network. PLoS Comput. Biol. 2012, 8, e1002754. (57) Mohan, S.; Sheena, A.; Poulose, N.; Anilkumar, G. Molecular dynamics simulation studies of GLUT4: substrate-free and substrateinduced dynamics and ATP-mediated glucose transport inhibition. PloS One 2010, 5, e14217. (58) Chebaro, Y.; Amal, I.; Rochel, N.; Rochette-Egly, C.; Stote, R. H.; Dejaegere, A. Phosphorylation of the retinoic acid receptor alpha induces a mechanical allosteric regulation and changes in internal dynamics. PLoS Comput. Biol. 2013, 9, e1003012. (59) Heidarsson, P. O.; Otazo, M. R.; Bellucci, L.; Mossa, A.; Imparato, A.; Paci, E.; Corni, S.; Di Felice, R.; Kragelund, B. B.; Cecconi, C. Single-molecule folding mechanism of an EF-hand neuronal calcium sensor. Structure 2013, 21, 1812−1821.

13122

dx.doi.org/10.1021/jp507936a | J. Phys. Chem. B 2014, 118, 13112−13122