(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Role of water-bridged interactions in metal ion coupled protein allostery [1]

['Xingyue Guan', 'Department Of Physics', 'National Laboratory Of Solid State Microstructure', 'Nanjing University', 'Nanjing', 'Wenzhou Institute', 'University Of Chinese Academy Of Sciences', 'Wenzhou', 'Zhejiang', 'Cheng Tan']

Date: 2022-07

Allosteric communication between distant parts of proteins controls many cellular functions, in which metal ions are widely utilized as effectors to trigger the allosteric cascade. Due to the involvement of strong coordination interactions, the energy landscape dictating the metal ion binding is intrinsically rugged. How metal ions achieve fast binding by overcoming the landscape ruggedness and thereby efficiently mediate protein allostery is elusive. By performing molecular dynamics simulations for the Ca 2+ binding mediated allostery of the calmodulin (CaM) domains, each containing two Ca 2+ binding helix-loop-helix motifs (EF-hands), we revealed the key role of water-bridged interactions in Ca 2+ binding and protein allostery. The bridging water molecules between Ca 2+ and binding residue reduces the ruggedness of ligand exchange landscape by acting as a lubricant, facilitating the Ca 2+ coupled protein allostery. Calcium-induced rotation of the helices in the EF-hands, with the hydrophobic core serving as the pivot, leads to exposure of hydrophobic sites for target binding. Intriguingly, despite being structurally similar, the response of the two symmetrically arranged EF-hands upon Ca 2+ binding is asymmetric. Breakage of symmetry is needed for efficient allosteric communication between the EF-hands. The key roles that water molecules play in driving allosteric transitions are likely to be general in other metal ion mediated protein allostery.

Natural proteins often utilize allostery in executing a variety of functions. Metal ions are typical cofactors to trigger the allosteric cascade. In this work, using the Ca 2+ sensor protein calmodulin as the model system, we revealed crucial roles of water-bridged interactions in the metal ion coupled protein allostery. The coordination of the Ca 2+ to the binding site involves an intermediate in which the water molecule bridges the Ca 2+ and the liganding residue. The bridging water reduces the free energy barrier height of ligand exchange, therefore facilitating the ligand exchange and allosteric coupling by acting as a lubricant. We also showed that the response of the two symmetrically arranged EF-hand motifs of CaM domains upon Ca 2+ binding is asymmetric, which is directly attributed to the differing dehydration process of the Ca 2+ ions and is needed for efficient allosteric communication.

Funding: This work was supported by the National Natural Science Foundation of China (Nos. 11974173 (WL), 11934008(WW), 11574132(WL)). Part of this work was done while DT was a visiting Professor in Nanjing University. DT acknowledges the NSF (CHE09--14033) for partial support of this work. The computing resources were provided by the High Performance Computing Center of Nanjing University. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Here, we used the bias-exchange metadynamics method [ 48 ] to investigate the Ca 2+ binding coupled conformational changes in CaM domains by explicitly modeling Ca 2+ in the simulations. We extracted the free energy landscape of the Ca 2+ binding coupled conformational changes of CaM domains at atomic level and demonstrated the full picture of the allosteric motions of CaM [ 32 , 49 – 51 ]. Our results revealed the key role of water-bridged interactions during Ca 2+ binding and protein allostery. The bridging water molecules between Ca 2+ and binding residues reduce the free energy barrier of ligand exchange landscape, and therefore the landscape ruggedness, by acting as a lubricant. This enables efficient Ca 2+ -ligand coordination and allosteric motions in the CaM. We propose that the water-bridged coordination is a general mechanism utilized in metal-coupled folding and allosteric communication in proteins. We also showed that Ca 2+ binding leads to rotation of the EF-hand helices with the hydrophobic core as the pivot, a structural change that should precede recognition by target proteins for signal transduction. In addition, the molecular simulations revealed obvious asymmetry in the allosteric coupling of the symmetrically paired EF-hands: the EF-hand 1 (EF 1 ) binds Ca 2+ in a sequential manner by chelating to the N-terminal residues followed by coordination with the central residues, and finally to the C-terminal residues. In contrast, chelation of Ca 2+ to the EF-hand 2 (EF 2 ) is initiated either by interactions with the residues in the N- and/or C-terminal residues followed by coordination to the central residues. Similar results were obtained for the cCaM. Such a symmetry breaking process is likely involved in larger multi-domain complexes, such as the bacterial chaperonin, GroEL, in which ligand binding induces asymmetric response in different subunits.

In a more general context, Ca 2+ -triggered conformational changes in CaM is an example of allostery, which describes the structural changes that occur in enzymes and molecular machines at distances far from the site at which a ligand binds [ 14 – 16 ]. Nature utilizes such functional motions, which are triggered by binding of cofactors to specific sites in proteins and RNA, to execute a variety of functions [ 17 – 27 ]. Binding of cofactors at a certain site (allosteric site) induces structural rearrangements at a distant site (regulated site) in the enzyme, thereby modulating the downstream activity. By such allosteric (action at a distance) movements, small local changes are amplified over long distances, which allows for control and regulation of cellular signaling and functions [ 28 ]. Divalent metal ions, e.g., Ca 2+ and Zn 2+ , have been widely utilized as effectors to trigger the allosteric cascade. However, due to the involvement of strong coordination interactions, the energy landscape dictating the metal ion binding and unbinding is intrinsically rugged. Despite the biological importance of metal ion coupled protein allostery, two key questions remain elusive, including: i) What strategy do metal ions utilize to overcome the ruggedness of the binding landscape and achieve rapid binding/unbinding? and ii) What physical interactions enable the propagations of the metal ion triggered allosteric signal to regulate the downstream processes? Because of the availability of rich experimental data and the typical allosteric features, CaM is an ideal system for answering these questions computationally [ 27 , 29 – 32 , 32 – 47 ].

(A, B) Structure of nCaM in the apo state (A) and holo state (B). Helix-loop-helix motifs α1–EF-loop1–α2 (cyan) and α3–EF-loop2–α4 (violet) comprise two EF-hands in the N-terminal domain of Calmodulin. Yellow spheres correspond to Ca 2+ ions. The details for one of the allosteric site in the loop region were also shown (right). (C) The sequence of nCaM and corresponding secondary structures. Ca 2+ binding loops are underlined with blue wavy lines. The native ligands are marked with “ln” to indicate their position in the EF-loop.

Calmodulin (CaM), a versatile calcium (Ca 2+ ) sensing protein expressed in all eukaryotic cells, is involved in a bewildering range of intracellular signaling processes [ 1 ]. Examples include activation of kinases [ 2 , 3 ], muscle contraction [ 4 ], gene regulation [ 5 ], signal transduction [ 6 – 8 ], and apoptosis [ 9 ]. Binding of Ca 2+ , whose intracellular and extracellular concentrations differ by several orders of magnitude, to CaM results in a large conformational change, leading to the exposure of hydrophobic residues that serve as recognition sites for target proteins. CaM is composed of two nearly symmetric globular domains connected by a flexible central helix. Each domain consists of two helix-loop-helix motifs, termed EF-hands ( Fig 1A and 1B ), which are found in a large number of calcium-sensing proteins [ 10 , 11 ]. The EF-hands chelate Ca 2+ , resulting in coordination to seven ligands arranged in a pentagonal bipyramid geometry (often involving one water molecule and five residues) with the negatively charged residues in the loop [ 11 ] ( Fig 1B ). In the apo state of CaM ( Fig 1A ), the helices in the EF-hand motif are arranged in an anti-parallel manner. Upon binding of Ca 2+ , the helices undergo substantial rearrangement into a nearly perpendicular conformation, exposing the hydrophobic sites which enable recognition and subsequent activation of target proteins [ 10 , 12 , 13 ]. Through such a mechanism, CaM initiates a variety of cellular processes.

2 Results and discussions

Ca2+ binding is coupled to conformational changes of the EF-hand motifs We use well-tempered bias-exchange metadynamics [48, 52] and the corresponding reweighting techniques [53] to extract the free energy landscape projected onto physically motivated multi-dimensional reaction coordinates characterizing Ca2+ binding and the conformational change. Quantitative analysis of the changes in these coordinates allows us to infer the mechanism of coupling between Ca2+ binding, the role of water, and the allosteric conformational changes in calmodulin. In particular, we use the “path collective variables” S α (α = 1, 2, 3, and 4), defined in the Methods and Materials section, to describe the conformational changes of the two EF-hand motifs. Small and large values of S α correspond to open and closed states, respectively. In order to assess the consequences of Ca2+ binding, we use the native coordination numbers N α (Methods and materials), representing the number of native ligands (oxygen atoms of residues that are coordinated with Ca2+ in the native holo structure) that bind to Ca2+ during the allosteric transitions for the two EF-hand motifs, respectively. Fig 2 shows the free energy landscapes, F(N 1 , S 1 ) and F(N 2 , S 2 ), of the nCaM. For both EF 1 and EF 2 , the conformational change of nCaM is tightly coupled to the extent of Ca2+ binding as indicated by the conformational distributions at different N 1 and N 2 values. When S α (α = 1 or 2) is large and N α (α = 1 or 2) is small, the closed state is more stable. Whereas when the native ligands are fully coordinated to Ca2+ the open structure is more stable (Fig 2 and Fig A in S1 Text). Interestingly, even with full coordination of native ligands (N 1 = 6) to Ca2+, the EF 1 samples a wide range of conformations as assessed by S 1 fluctuations, suggesting the conformational plasticity of the open state. Conformational plasticity for the Ca2+ bound state, which was also observed in experiments, could be the major reason that the EF-hand motifs recognize and bind a variety of target proteins [54, 55]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. Free energy profiles of the Ca2+ coupled conformational transitions of EF 1 (A) and EF 2 (B) projected onto the collective variables (N 1 , S 1 ) and (N 2 , S 2 ), respectively. Representative conformations of the major basins in the landscape are also shown. The unit of the free energy is kJ/mol. https://doi.org/10.1371/journal.pcbi.1010195.g002 Although the two EF-hand motifs have similar structures, comparison of Fig 2A and 2B shows that there are noticeable differences in the free energy landscapes upon Ca2+ binding. For example, when three or four ligands bind to Ca2+, the conformation of EF 2 is poised to make a transition to the open-like structure, whereas EF 1 remains in the closed-like structure. It is worth noting that the two EF-hand motifs were simulated as a whole in this work, and the conformational changes of the two EF-hands are not independent of each other. In the calculations of the free energy landscape of EF 2 , the EF 1 is allowed to change its conformation. Therefore, the cooperativity between the two EF-hand motifs was included in the simulations. The free energy landscapes of the two EF-hand motifs shown in Fig 2 correspond to the projections of the overall free energy, but not the free energy landscapes of the isolated EF-hands. Detailed analysis shows that during the conformational transition, the two EF-hand motifs are tightly coupled. The open conformation of EF 2 is stabilized when EF 1 is in the open conformation with most of the associated native ligands bound to Ca2+. Thus, the behavior of EF 2 is affected by the conformation of EF 1 through long range allosteric interactions or action at a distance between the two EF-hands, which will be discussed further in later subsection. The observed differences in the allosteric communication are reasonable since the two EF-hand motifs have different amino-acid sequences in the Ca2+ binding loop. Consequently, the ensembles of structures with three or four native ligands bound to the Ca2+ should exhibit substantial differences in the two EF-motifs. Such an asymmetry in the Ca2+ binding and conformational transitions of the symmetrically arranged EF-hand pair in CaM domains likely plays an important role in signaling. It is worth mentioning that conventional classic force fields often encounter difficulties in quantitatively reproducing the thermodynamic properties of the coordination bonds involving metal ions (especially Ca2+ or Mg2+) due to the lack of explicit consideration of charge transfer, polarization, and protonation/deprotonation effects [56–59]. For example, a recent study by Zhang and coworkers showed that the force field parameters derived from ab initio quantum calculations for the Ca2+ coordination in calmodulin depend on the loop conformations [56]. Meanwhile, a number of works have been devoted to improving existing force fields for better description of metal ion coordination [57, 58, 60, 61]. In an early work by Project and coworkers, a new set of Lennard-Jones (LJ) parameters for the calcium-oxygen pair were proposed to improve the quantitative description of the experimental Ca2+-formate affinity using the GROMOS96 and OPLS-AA force fields [57]. By refining the LJ parameters of calcium-oxygen pair for the OPLS-AA force field, Kahlen and coworkers [58] not only reproduced the experimental dissociation constant of calcium acetate, but also provided a better description of the key characteristics of the potential of mean force for the calcium acetate ion pair, including the relative probabilities of the contact ion pairs and the solvent-shared ion pairs. In this work, the OPLS-AA force field was used for the MD simulations of the Ca2+ coupled conformational change of calmodulin domains. According to the discussions in Ref. [57], the depths of the free energy minima involving the Ca2+ mediated interactions and the barrier separating them in the free energy landscapes of this work may depend on the details of the force field parameters. However, as will be discussed in the end of this section, the qualitative features of the free energy landscapes are insensitive to the details of the used force fields.

Ca2+ binding-induced rotation of EF-hand helices Structures of nCaM show that there are several hydrophobic residues located around Ile27 and Ile63 whose non-polar side-chains stack against each other to form the center of the hydrophobic core [54, 68] (Fig 5). During the Ca2+ binding and conformational transition, the hydrophobic cluster formed by these residues, including Phe16, Phe19, Ile27, Leu32, Val35, Ile52, Val55, Ile63 and Phe68, are almost rigid (Fig F in S1 Text). Comparison of the structures in the apo- and holo- states suggests that the two helices of the EF-hand rotate around the hydrophobic cluster during the conformational transition. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. Ca2+ binding induced rotation of EF-hand helices. (A) The distance between close ends of EF-loop 1 (represented by the C α atoms from Asp20 and Glu31) and the dihedral angle characterizing the rotation of the two helices of EF 1 as a function of N 1 (upper panel). The same results for the EF 2 are shown in the lower panel of (A). (B) Schematic diagrams illustrating the end-end distance of EF-loop1 and (C) the inter-helical angle and diheral angle of EF 1 . The end-end distance of EF-loop is represented by the distance between C α atoms: EF-loop1 by Asp20–Glu31, and EF-loop2 by Asp56–Glu67. The direction of helices are represented by vectors pointing from one residue’s C α atom to another residue’s C α atom: α1 by Phe12–Phe19, α2 by Glu31–Leu39, α3 by Leu48–Val55, and α4 by Phe68–Lys75. The dihedral angles Ψ are defined by the C α atoms of the residues Phe12–Phe19–Glu31–Leu39 for EF 1 and Leu48–Val55–Phe68–Lys75 for EF 2 . (D) Schematic diagram of the Ca2+ binding induced EF-hand helix rotation. Black spheres indicate the hydrophobic core in EF 1 . (E) Free energy profile of nCaM conformational transition projected onto S 1 and S 2 . The free energy scale is in kJ/mol. https://doi.org/10.1371/journal.pcbi.1010195.g005 To demonstrate the coupling between the Ca2+ binding and the EF-hand opening, we calculated the distance between the EF-loop ends (R loop , distance between the C α atoms of the first and twelfth residues, see Fig 5B) and the dihedral angle for the rotation of the two helices (Ψ, defined by the C α atoms of four residues selected from the α-helices, as shown in Fig 5C) of the two EF-hands, respectively. The results in Fig 5A show that R loop decreases upon Ca2+ binding, thus pulling the ends of the EF-hand helices close together (blue lines). In particular, coordination of the fourth (third) ligand to Ca2+, which corresponds to the binding of the terminal residues of the EF-hand loops to the Ca2+ by water bridged interactions (direct interactions) in EF 1 (EF 2 ), has the most prominent effect in changing the distance between the two close ends. Meanwhile, the dihedral angles formed by the two helices increase (red lines), indicating the rotation of the EF-hand helices. In the process of rotation, the hydrophobic cluster, which is stable during the Ca2+ binding, acts as the pivot (Fig 5D). We also defined the angle η to characterize the in plane rotation of the two helices, and similar results can be observed (Fig G in S1 Text). The results described above show that the large conformational change in the two EF-hand motif occurs at different stages of the coordination (Fig 2). We simulated the two EF-hand motifs of the individual CaM domain as a whole system, which implies that cooperative interactions between the two EF-hand motifs are automatically included. The differences in the free energy profiles for the two EF-hand motifs is related to the high cooperativity. Analysis of the correlation between the coordination processes of the two EF-hand motifs demonstrates that the Ca2+ binding in the EF 1 proceeds before interaction with EF 2 (Fig H in S1 Text). When N 1 is less than 5, N 2 predominantly takes on the values of 1 or 2, and both the EF 1 and EF 2 mainly stay in the closed conformation (Fig 2). As shown in Fig 5E, the conformational changes in the two EF-hands mostly follow the diagonal line in the two-dimensional free energy landscape, which suggests tight coupling and cooperativity between the two EF-hand motifs. With the full coordination of the EF 1 (N 1 = 5 and 6), it switches to the open conformation. Due to the tight coupling between the two EF-hand motifs, Ca2+ coordination induced stabilization of the open conformation of the EF 1 tends to promote the closed to open switching of the EF 2 conformation, even though the coordination of the EF 2 is not fully completed. Consequently, we observe that the major conformational switching occurs at N 1 = 5 for the EF 1 , but it occurs at N 2 = 3 for the EF 2 as shown in the free energy profiles (Fig 2). The tight coupling between the two EF-hands is consistent with previous experimental observations [29, 41]. Taken together, these results demonstrate that the rotation of the EF-hand helices is coupled to Ca2+ binding, which is consistent with the free energy landscapes shown in Fig 2. Ca2+ binding induced rotation of the EF-hand helices was also noted in a simulation study [69].

[END]
---
[1] Url: https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1010195

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.