All-atom simulations of the trimeric spike protein of SARS-CoV-2 in aqueous medium: Nature of interactions, conformational stability and free energy diagrams for conformational transition of the protein.

The spike protein of SARS-CoV-2 exists in two major conformational states, namely the 'open' and 'closed' states which are also known as the 'up' and 'down' states, respectively. In its open state, the receptor binding domain (RBD) of the protein is exposed for binding with ACE2, whereas the spike RBD is inaccessible to ACE2 in the closed state of the protein. In the current work, we have performed all-atom microsecond simulations of the full-length trimeric spike protein solvated in explicit aqueous medium with an average system size of ~0.7 million atoms to understand the molecular nature of intra- and inter-chain interactions, water-bridged interactions between different residues that contribute to the stability of the open and closed states of the protein, and also the free energy landscape for transition between the open and closed states of the protein. We have also examined the changes of such interactions that are associated with switching from one state to the other through both unbiased and biased simulations at all-atom level with total run length of 4 µs. Interestingly, after about 0.8 µs of unbiased molecular dynamics run of the spike system in the open state, we observed a gradual transition of the monomeric chain (B) from open to its partially closed or down state. Initially the residues at the interface of chain B RBD in the open state spike protein were at non-hydrogen-bonding distances from the residues of chain C RBD. However, the two RBDs gradually came closer and finally the residue S459 of the RBD of chain B made a hydrogen bond with F374 of chain C in the last 200 ns of the simulation along with formation of a few more hydrogen bonds involving other residues. Since no transition from closed to the open state of the protein is observed in the present 1 µs unbiased simulation of the closed state protein, the current study seems to suggest that the closed conformational state is preferred for the spike protein of SARS-CoV-2 in aqueous medium. Furthermore, calculations of the free energy surface of the conformational transition from open (up) to the closed (down) state using a biased simulation method reveal a free energy barrier of ~3.20 kcal/mol for the transition of RBD from open to the closed state, whereas the barrier for the reverse process is found to be significantly higher.

A Multiple Proton Transfer Mechanism for the Charging Step of the Aminoacylation Reaction at the Active Site of Aspartyl tRNA Synthetase.

Aspartyl-tRNA synthetase catalyzes the attachment of aspartic acid to its cognate tRNA by the aminoacylation reaction during the initiation of the protein biosynthesis process. In the second step of the aminoacylation reaction, known as the charging step, the aspartate moiety is transferred from aspartyl-adenylate to the 3'-OH of A76 of tRNA through a proton transfer process. We have investigated different pathways for the charging step through three separate QM/MM simulations combined with the enhanced sampling method of well-sliced metadynamics and found out the most feasible pathway for the reaction at the active site of the enzyme. In the charging reaction, both the phosphate group and the ammonium group after deprotonation can potentially act as a base for proton transfer in the substrate-assisted mechanism. We have considered three possible mechanisms involving different pathways of proton transfer, and only one of them is determined to be enzymatically feasible. The free energy landscape along reaction coordinates where the phosphate group acts as the general base showed that, in the absence of water, the barrier height is 52.6 kcal/mol. The free energy barrier is reduced to 39.7 kcal/mol when the active site water molecules are also treated quantum mechanically, thus allowing a water mediated proton transfer. The charging reaction involving the ammonium group of the aspartyl adenylate is found to follow a path where first a proton from the ammonium group moves to a water in the vicinity forming a hydronium ion (H3O+) and NH2 group. The hydronium ion subsequently passes the proton to the Asp233 residue, thus minimizing the chance of back proton transfer from hydronium to the NH2 group. The neutral NH2 group subsequently takes the proton from the O3' of A76 with a free energy barrier of 10.7 kcal/mol. In the next step, the deprotonated O3' makes a nucleophilic attack to the carbonyl carbon forming a tetrahedral transition state with a free energy barrier of 24.8 kcal/mol. Thus, the present work shows that the charging step proceeds through a multiple proton transfer mechanism where the amino group formed after deprotonation acts as the base to capture a proton from O3' of A76 rather than the phosphate group. The current study also shows the important role played by Asp233 in the proton transfer process.

Free Energy Landscape of the Adenylation Reaction of the Aminoacylation Process at the Active Site of Aspartyl tRNA Synthetase.

The process of protein biosynthesis is initiated by the aminoacylation process where a transfer ribonucleic acid (tRNA) is charged by the attachment of its cognate amino acid at the active site of the corresponding aminoacyl tRNA synthetase enzyme. The first step of the aminoacylation process, known as the adenylation reaction, involves activation of the cognate amino acid where it reacts with a molecule of adenosine triphosphate (ATP) at the active site of the enzyme to form the aminoacyl adenylate and inorganic pyrophosphate. In the current work, we have investigated the adenylation reaction between aspartic acid and ATP at the active site of the fully solvated aspartyl tRNA synthetase (AspRS) from Escherichia coli in aqueous medium at room temperature through hybrid quantum mechanical/molecular mechanical (QM/MM) simulations combined with enhanced sampling methods of well-tempered and well-sliced metadynamics. The objective of the present work is to study the associated free energy landscape and reaction barrier and also to explore the effects of active site mutation on the free energy surface of the reaction. The current calculations include finite temperature effects on free energy profiles. In particular, apart from contributions of interaction energies, the current calculations also include contributions of conformational, vibrational, and translational entropy of active site residues, substrates, and also the rest of the solvated protein and surrounding water into the free energy calculations. The present QM/MM metadynamics simulations predict a free energy barrier of 23.35 and 23.5 kcal/mol for two different metadynamics methods used to perform the reaction at the active site of the wild type enzyme. The free energy barrier increases to 30.6 kcal/mol when Arg217, which is an important conserved residue of the wild type enzyme at its active site, is mutated by alanine. These free energy results including the effect of mutation compare reasonably well with those of kinetic experiments that are available in the literature. The current work also provides molecular details of structural changes of the reactants and surroundings as the system dynamically evolves along the reaction pathway from reactant to the product state through QM/MM metadynamics simulations.

All-Atom Simulations of Human ACE2-Spike Protein RBD Complexes for SARS-CoV-2 and Some of its Variants: Nature of Interactions and Free Energy Diagrams for Dissociation of the Protein Complexes.

The spike protein of SARS-CoV-2 is known to interact with the human ACE2 protein via its receptor binding domain (RBD). We have investigated the molecular nature of this interprotein interaction and the associated free energy diagrams for the unbinding of the two proteins for SARS-CoV-2 and some of its known variants through all-atom simulations. The present work involves generation and analysis of 2.5 µs of unbiased and 4.2 µs of biased molecular dynamics trajectories in total for five explicitly solvated RBD-ACE2 systems at full atomic level. First, we have made a comparative analysis of the details of residue-wise specific interactions of the spike protein with ACE2 for SARS-CoV-1 and SARS-CoV-2. It is found that the average numbers of both direct interprotein and water-bridged hydrogen bonds between the RBD and ACE2 are higher for SARS-CoV-2 than SARS-CoV-1. These higher hydrogen bonded interactions are further aided by enhanced nonspecific electrostatic attractions between the two protein surfaces for SARS-CoV-2. The free energy calculations reveal that there is an increase in the free energy barrier by ∼1.5 kcal/mol for the unbinding of RBD from ACE2 for SARS-CoV-2 compared to that for SARS-CoV-1. Subsequently, we considered the RBDs of three variants of SARS-CoV-2, namely N501Y, E484Q/L452R, and N440K. The free energy barrier of protein unbinding for the N501Y variant is found to be ∼4 kcal/mol higher than the wild type SARS-CoV-2 which can be attributed to additional specific interactions involving Tyr501 of RBD and Lys353 and Tyr42 of ACE2 and also enhanced nonspecific electrostatic interaction between the protein surfaces. For the other two mutant variants of E484Q/L452R and N440K, the free energy barrier for protein unbinding increases by ∼2 and ∼1 kcal/mol, respectively, compared with the wild type SARS-CoV-2, which can be attributed to an increase in the number of interprotein hydrogen bonds for the former and also to enhanced positive electrostatic potential on the RBD surfaces for both of the variants. The successive breaking of interprotein hydrogen bonds along the free energy pathway of the unbinding process is also found out for all five systems studied here.

Inhibition of seryl tRNA synthetase by seryl nucleoside moiety (SB-217452) of albomycin antibiotic.

The seryl nucleoside moiety (SB-217452) of the Trojan horse antibiotic albomycin exhibits broad spectrum antibiotic activity against various bacterial pathogens by targeting seryl tRNA synthetase (SerRS). The aim of the present study is to understand how the SB-217452 inhibits SerRSs of different species. First, the binding efficacy of SB-217452 in the dimeric SerRS from Thermus thermophilus (TtSerRS) in complex with tRNASer is compared with the binding of seryl adenylate (Ser-AMP). Multiple reasons for inhibition action of SB-217452 are revealed. In the next part, we have compared the binding event of SB-217452 in SerRS from Staphylococcus aureus (SaSerRS) and from Streptomyces sp. (SpSerRS1). First, quantum mechanical study (AIM analysis) shows that the network of interaction is stronger in SaSerRS:tRNA complex compared to the SpSerRS1:tRNA complex. This conclusion is in fair agreement with the observed IC50 values which show that the binding free energy of SB-217452 in the active site of SaSerRS is more favorable compared to that in SpSerRS1. The interactions of antibiotic with ß sheets contribute to the differences in the binding behavior. Secondly, the classical simulation results corroborate the results of AIM analysis. Finally, metadynamics calculation of the free energy surface of the conformational change of the SB-217452 shows that the antibiotic binds in a unique catalytically non competent organization in SaSerRS:tRNA. In contrast, the antibiotic can bind in the active site of SpSerRS1:tRNA complex with multiple catalytically incompetent conformations. The present study provides a comprehensive molecular perspective of the inhibition mechanism of the antibiotic.Communicated by Ramaswamy H. Sarma.

Classical molecular dynamics simulation of seryl tRNA synthetase and threonyl tRNA synthetase bound with tRNA and aminoacyl adenylate.

Lacunae of understanding exist concerning the active site organization during the charging step of the aminoacylation reaction. We present here a molecular dynamics simulation study of the dynamics of the active site organization during charging step of subclass IIa dimeric SerRS from Thermus thermophilus (ttSerRS) bound with tttRNASer and dimeric ThrRS from Escherichia coli (ecThrRS) bound with ectRNAThr. The interactions between the catalytically important loops and tRNA contribute to the change in dynamics of tRNA in free and bound states, respectively. These interactions help in the development of catalytically effective organization of the active site. The A76 end of the tttRNASer exhibits fast dynamics in free State, which is significantly slowed down within the active site bound with adenylate. The loops change their conformation via multimodal dynamics (a slow diffusive mode of nanosecond time scale and fast librational mode of dynamics in picosecond time scale). The active site residues of the motif 2 loop approach the proximal bases of tRNA and adenylate by slow diffusive motion (in nanosecond time scale) and make conformational changes of the respective side chains via ultrafast librational motion to develop precise hydrogen bond geometry. Presence of bound Mg2+ ions around tRNA and dynamically slow bound water are other common features of both aaRSs. The presence of dynamically rigid Zinc ion coordination sphere and bipartite mode of recognition of ectRNAThr are observed.

Dynamics of the active site loops in catalyzing aminoacylation reaction in seryl and histidyl tRNA synthetases.

Aminoacylation reaction is the first step of protein biosynthesis. The catalytic reorganization at the active site of aminoacyl tRNA synthetases (aaRSs) is driven by the loop motions. There remain lacunae of understanding concerning the catalytic loop dynamics in aaRSs. We analyzed the functional loop dynamics in seryl tRNA synthetase from Methanopyrus kandleri (mkSerRS) and histidyl tRNA synthetases from Thermus thermophilus (ttHisRS), respectively, using molecular dynamics. Results confirm that the motif 2 loop and other active site loops are flexible spots within the catalytic domain. Catalytic residues of the loops form a network of interaction with the substrates to form a reactive state. The loops undergo transitions between closed state and open state and the relaxation of the constituent residues occurs in femtosecond to nanosecond time scale. Order parameters are higher for constituent catalytic residues which form a specific network of interaction with the substrates to form a reactive state compared to the Gly residues within the loop. The development of interaction is supported from mutation studies where the catalytic domain with mutated loop exhibits unfavorable binding energy with the substrates. During the open-close motion of the loops, the catalytic residues make relaxation by ultrafast librational motion as well as fast diffusive motion and subsequently relax rather slowly via slower diffusive motion. The Gly residues act as a hinge to facilitate the loop closing and opening by their faster relaxation behavior. The role of bound water is analyzed by comparing implicit solvent-based and explicit solvent-based simulations. Loops fail to form catalytically competent geometry in absence of water. The present result, for the first time reveals the nature of the active site loop dynamics in aaRS and their influence on catalysis.

Dynamics of the Active Sites of Dimeric Seryl tRNA Synthetase from Methanopyrus kandleri.

Aminoacyl tRNA synthetases (aaRSs) carry out the first step of protein biosynthesis. Several aaRSs are multimeric, and coordination between the dynamics of active sites present in each monomer is a prerequisite for the fast and accurate aminoacylation. However, important lacunae of understanding exist concerning the conformational dynamics of multimeric aaRSs. Questions remained unanswered pertaining to the dynamics of the active site. Little is known concerning the conformational dynamics of the active sites in response to the substrate binding, reorganization of the catalytic residues around reactants, time-dependent changes at the reaction center, which are essential for facilitating the nucleophilic attack, and interactions at the interface of neighboring monomers. In the present work, we carried out all-atom molecular dynamics simulation of dimeric (mk)SerRS from Methanopyrus kandleri bound with tRNA using an explicit solvent system. Two dimeric states of seryl tRNA synthetase (open, substrate bound, and adenylate bound) and two monomeric states (open and substrate bound) are simulated with bound tRNA. The aim is to understand the conformational dynamics of (mk)SerRS during its reaction cycle. While the present results provide a clear dynamical perspective of the active sites of (mk)SerRS, they corroborate with the results from the time-averaged experimental data such as crystallographic and mutation analysis of methanogenic SerRS from M. kandleri and M. barkeri. It is observed from the present simulation that the motif 2 loop gates the active site and its Glu351 and Arg360 stabilizes ATP in a bent state favorable for nucleophilic attack. The flexibility of the walls of the active site gradually reduces near reaction center, which is a more organized region compared to the lid region. The motif 2 loop anchors Ser and ATP using Arg349 in a hydrogen bonded geometry crucial for nucleophilic attack and favorably influences the electrostatic potential at the reaction center. Synchronously, Arg366 of the ß sheet at the base holds the syn oxygen of the attacking carboxylic group so that the attack by the anti oxygen is feasible. This residue also contributes to the reduction of the unfavorable electrostatic potential at the reaction center. Present simulation clearly shows the catalytic role of the residues at reaction center. A precise and stable geometry of hydrogen bonded network develops within the active site, which is essential for the development of an optimum transition state geometry. All loops move away from the platform of active site in the open or adenylate bound state and the network of hydrogen bond disappears. The serine binding site is most rigid among all three subsites. The Ser is held here in a highly organized geometry bound by Zn(2+) and Cys residues. Present simulation further suggests that the helix-turn-helix motif connecting the monomers might have important role in coordinating the functional dynamics of the two active sites. The N-terminal domain is involved in long-range electrostatic interaction and specific hydrogen bond interaction (both direct and water mediated) with tRNA. Overall conformational fluctuation is less in the N terminal compared to the catalytic domain due to the presence of a motif 2 loop, loop f, and serine ordering loop, which change conformation in the later domain during the reaction cycle. The dynamic perspective of the active site of (mk)SerRS with the mobile loop acting as the gate and dynamically silent ß sheets performing as the base has similarity with the perception of the active site in various other enzymes.

Active site nanospace of aminoacyl tRNA synthetase: difference between the class I and class II synthetases.

The present work is aimed at understanding the origin of the difference in the molecular organization of the active site nanospaces of the class I and class II aminoacyl tRNA synthetases (aaRSs) which are tunnel-like structures. The active site encloses the cognate amino acid (AA) and the adenosine triphosphate (ATP) to carry out aminoacylation reaction. Comparison of the structures of the active site of the class I and class II (aaRSs) shows that the nanodimensional tunnels are curved in opposite directions in the two classes. We investigated the origin of this difference using quantum mechanical computation of electrostatic potential (ESP) of substrates, surrounding residues and ions, using Atoms in Molecule (AIM) Theory and charge population analysis. We show that the difference is principally due to the variation in the spatial charge distribution of ATP in the two classes which correspond to extended and bent conformations of ATP. The present computation shows that the most feasible pathway for nucleophilic attack to alphaP is oppositely directed for class I and class II aaRSs. The available crystal structures show that the cognate AA is indeed located along the channel favorable for nucleophilic attack as predicted by the ESP analysis. It is also shown that the direction of the channel changes its orientation when the orientation of ATP is changed from extended to a bent like structure. We further used the AIM theory to confirm the direction of the approach of AA in each case and the results corroborate the results from the ESP analysis. The opposite curvatures of the active site nanospaces in class I and class II aaRSs are related with the influence of the charge distributions of the extended and bent conformations of ATP, respectively. The results of the computation of electrostatic potential by successive addition of active site residues show that their roles on the reaction are similar in both classes despite the difference in the organization of the active sites of class I and class II aaRSs. The difference in mechanism in two classes as pointed out in recent study (S. Dutta Banik and N. Nandi, J. Biomol. Struct. Dyn. 30, 701 (2012)) is related with the fact that the relative arrangement of the ATP with respect to the AA is opposite in class I and class II aaRSs as explained in the present work. The charge population difference between the reacting centers (which are the alphaP atom of ATP (q(p)) and the attacking oxygen atom of carboxylic acid group (q(o)), respectively) denoted by delta(q), is a measure of the propensity of nucleophilic attack. The population analysis of the substrate AA shows that a non-negligible difference exists between the attacking oxygens of AA in class I (syn) and in class II (anti) which is one reason for the lower value of delta(q) in class II relative to class I. The population analysis of the AA, ATP, Mg+2 ions and active site residues shows that the difference in delta(q) values of the two classes is substantially reduced. When ions and residues are considered. Thus, the bent state of ATP, Mg+2 ions and active site residues complements it cognate AA to carry out the nucleophilic reaction in class I as efficiently as occurs in class I (with the extended state of ATP, single Mg+2 ion and active site residues). This could be one reason for the two different conformations of ATP in the two classes. The mutual arrangements of AA and ATP in each aaRS are guided by the spatial charge distribution of ATP (extended and bent). The present work shows that the construction of nanospace complements the arrangement of the substrate (AA and ATP).