Binding and dynamics demonstrate the destabilisation of ligand binding for the S688Y mutation in the NMDA receptor GluN1 subunit

doi:10.21203/rs.3.rs-1825308/v1

Encephalopathies are brain dysfunctions which leads to cognitive, sensory, and motor development impairments. Recently, the identification of several mutations within the N-methyl-D-aspartate receptor (NMDAR) have been identified as significant in the aetiology of this group of conditions. However, a complete understanding of the underlying molecular mechanism and changes to the receptor due to these mutations has been elusive. We studied the molecular mechanisms by which one of the first mutation within the NMDAR GluN1 ligand binding domain, Ser688Tyr, cause the encephalopathies. We performed molecular docking, randomly seeded molecular dynamics simulations, and binding free energy calculations to determine the behaviour of the 2 major co-agonists: glycine and D-serine in both the wild-type and S688Y receptor. We observed that the Ser688Tyr mutation leads to instability of both ligands within the ligand binding site due to structural changes associated with the mutation. The binding free energy for both ligands was significantly more unfavourable in the mutated receptor. These results explain previously observed in vitro electrophysiological data and provide detailed aspects of ligand association, and its effects on receptor activity. Our study provides valuable insight into the consequences of mutations within the NMDAR GluN1 ligand binding domain.

Molecular Dynamics

NMDA

Membrane Protein

Ligand Binding

Early infantile epileptic encephalopathies are some of the most devastating early onset epilepsies and can lead to cognitive, sensory, and motor development impairments. The aetiology of these syndromes are wide-ranging and can be structural, metabolic, neurological or genetic in nature [1]. Recently due to advances in high throughput sequencing technologies there has been an increase in the identification of genetic mutations associated with early infantile epileptic encephalopathies, including mutations in the N-methyl-D-aspartate receptor (NMDAR) [2].

NMDARs are ionotropic glutamate receptors (iGluR) and consist of four subunits, which are typically two GluN1 subunits and two GluN2 subunits (Fig. 1A), however the existence of other triheterotetrameric arrangements have been reported [3–5]. They are unique in their mode of activation, requiring simultaneous binding of glutamate at the GluN2 subunit and a co-agonist such as glycine [6], D-serine [7], D-alanine [6, 8] or D-cycloserine [7] at the GluN1 subunits [9–11], along with a membrane depolarisation to displace a Mg²⁺ ion block present in the channel [12, 13]. Together, these events activate the receptor and enable Ca²⁺ influx and further downstream effects within the cell [14].

Structurally all GluN subunits consist of an amino terminal domain (ATD), a ligand binding domain (LBD), a transmembrane domain (TMD) and an intracellular carboxyl-terminal domain (CTD) (Fig. 1A). X-ray structures of the NMDAR GluN1 LBD were originally reported in 2003 [7], and advances in cryogenic electron microscopy have since led to membrane and extracellular receptor structures being solved [15–19]. Significantly aided by crystallographic work [20] the individual domains appear largely physically discreet. The NMDAR is normally thought to be involved in learning and memory formation [21, 22]. However, dysfunction of the NMDAR has been associated with complex, multifactorial conditions such as schizophrenia [23–26] and Alzheimer’s disease [27, 28].

Generally, mutations within the GluN1 subunit are considered to be associated with conditions such as hypotonia [29], hyperkinetic and stereotyped movements as well as other conditions [30, 31]. Predominantly, mutations in the GluN1 subunit have been identified within the TMD [29–37], however mutations within the ATD have also been reported [31, 37]. The first mutation within the LBD linked to infantile epileptic encephalopathies, Ser688Tyr (S688Y) was described by Zehavi et al, in 2017. Skrenkova et al. established the effects of the S688Y mutation in terms of a significant decrease in functional potency for both glycine and D-serine along with a differential regulation in surface delivery of the mutant NMDAR in HEK293 cells. However, the molecular mechanisms by which the S688Y mutation leads to decreased potency is yet to be established. We used a combination of docking and molecular dynamics simulations to assist in understanding the ligand dynamics within the binding site. The binding thermodynamics relevant to protein-ligand binding was also estimated in the study. Our results reveal how the receptor functions, and specifically how the S688Y mutation leads to impaired receptor function. These calculations demonstrate changes in the receptor-ligand complex structure with important concomitant alterations of the electrostatic and water-bridge network present within the binding site. This gives a deeper insight into the effects of mutations within the NMDAR, and specifically those crucial within the LBD and for ligand binding.

Figures in this paper were prepared using PyMOL [38] and the Maestro visualisation interface [39]. Calculations were performed on Intel Xeon E5-2680 V3 CPUs and Nvidia V100 SXM2 GPUs. Graphs were generated using the Simulation Interactions Diagram module in Desmond 2018-4 [39, 40].

Receptor and Ligand Preparation

The crystal structure of the GluN1/GluN2A ligand-binding domain in complex with glycine and glutamate (PDB: 4NF8) [41] was selected due to its high resolution of 1.86 Å and absence of allosteric modulators or large inhibitors.

The structure was prepared using the protein preparation module in Schrödinger [39, 42]. In brief, this involved assignment of bond orders, adding missing hydrogens and connection of disulphide bonds. Protonation states were assigned corresponding to pH 7.0 ± 2.0 using PROPKA [43, 44]. All water molecules in the PDB entry were removed and the protein structure was minimised with positional restraints as per default settings using the OPLS3e force field [45].

The S688Y mutation was introduced at the GluN1 binding site using the Maestro interface and the rotamer, with highest standard population and probability based on their respective dihedral angles determined using the rotamer selection tool. Minimisation was subsequently carried out with position restraints on Cα atoms.

Molecular docking

Glycine and D-serine were docked into the wild type and mutant GluN1 binding pockets in their fully ionised form using the Glide docking module [39, 46-48] with flexible ligand sampling and the extra precision (XP) scoring function. Van der Waals radii were scaled by a factor of 0.80 and a partial charge cut-off of 0.15 eV for non-polar ligand atoms applied. The receptor grid to guide the docking (Supp. Fig. 1) was defined using the co-crystallised glycine from the original PDB structure as the centroid of the receptor site. The van der Waals radius scale factor was set to 1.0 and a partial charge cut-off to 0.25 eV. The maximum number of conformations and output per ligand was set to 10. Glide utilises a relative docking energy for each docking pose with lower energies being more favourable.

Molecular Dynamics (MD) Simulation

The top binding pose for each of the ligand/receptor combinations were subjected to molecular dynamics (MD) simulations for further analysis. Simulation systems were built using the System Builder tool in Desmond 2018-4 [40] using the OPLS_2005 force field [49] and a cubic system with a 10 Å x 10 Å x 10 Å buffer around the protein. Each system was solvated using the explicit SPC water model [50]. The overall charge of the system was neutralised by substituting water molecules with Cl^- and further NaCl was added to a concentration of 0.15 M. Each system was subject to the default relaxation protocol which involved the relaxation of the simulation system using default NVT and NPT ensembles.

Simulations were conducted for each system using identical starting parameters, differing only in their seeding. Simulations were performed for a period of 1 μs with a sampling interval of 20 ps and a default seed of “2007”. Subsequent simulations were 200ns in duration with a sampling interval of 5 ps using the default seed of “2007” and 5 randomised seeds (Supplementary Table 1).

Binding Free Energy Calculations

The final 50 ns from the 200 ns MD trajectories (fixed seed “2007”) were used to estimate the affinity of glycine and D-serine for the GluN1 and mutant receptor using generalised Born and surface area (MM‑GBSA) [51, 52] method. The Gibbs binding free energy (ΔG_bind) was computed based on the equation shown below:

$${\varDelta G}_{bind}= {G}_{complex}-{G}_{protein} -{G}_{ligand}$$

ΔG_bindis a measure of estimating the binding affinity of a ligand to a protein with negative values corresponding to energetically favourable reactions and a positive value being energetically unfavourable.

The MM‑GBSA calculations were performed using the thermal_mmgbsa.py plugin of Prime [53, 54] as supplied [39]. Calculations were performed using the OPLS3e [45] force field and the VSGB 2.1 implicit solvation model [55]. In brief, the script extracts the final 50 ns and performs geometry optimisation to extract energies of the complex, the protein, and the ligand to calculate ΔG_bind.

Ligand docking to GluN1 LBD

Docking of both glycine and D-serine in the wild type GluN1 LBD (Fig. 2A, 2B, 3A, 3B) revealed key salt bridge interactions between the carboxylate group of both ligands and the guanidium group of Arg523 of the protein as well as between the ammonium groups and the side chain carboxylate group of Asp732. Further interactions involved π-cation interactions with the phenyl group of Phe484 and D-serine displayed π-cation interactions with Trp731. Hydrogen bonding was also present in the docked conformations with the backbone of residues Pro516, Thr518, Ser688 and side chains of residues Arg523 and Trp732. These interactions are consistent with those observed in the X-ray structures of the NMDAR GluN1 LBD co-crystallised with glycine and D-serine by Furukawa and Gouaux [7]. This demonstrates that our docking protocol is valid and appropriate to study the S688Y mutant. The root mean square deviation (RMSD) of all heavy atoms was <1 Å between the X-ray structure containing the glycine ligand (PDB: 1PB7) [7] and the docked ligand when superimposed on the protein backbone.

In the docked poses, the carboxylate and ammonium groups of glycine and D-serine were present in a polar pocket of the binding site with the hydroxymethyl group of D-serine present in a hydrophobic sub-pocket. However, in the S688Y mutant receptor, the environment is changed with fewer polar residues present in the environment of both glycine and D-serine. Furthermore, the key Arg523 residue is not present in the binding pocket due to steric hindrance from the bulky Tyr688 hydroxyphenyl side chain (Fig. 3C, 3D) which upon docking leads to an associated loss of a salt bridge and hydrogen bond in comparison with the wild-type configuration with serine (Fig 2C, 2D). Other interactions however remained intact. Of interest is the main chain nitrogen atom of Tyr688 which continues to facilitate hydrogen bonding with the carboxylate group of glycine and D-serine (Figure 3C, 3D) with the formation of an additional aromatic hydrogen bond between the side chain of Tyr688 and glycine. Positioning of the carboxylate group of both ligands is similar in the wild-type protein, however due to the absence of the guanidium group of Arg523 inside the binding pocket in the mutant receptor there is a conformational difference between glycine and D-serine whereby the carboxylate group on glycine has a “downward” rotated orientation relative to the carboxylate group in D-serine. Meanwhile the ammonium group is held in place by Asp732 and is consistent between both the wild-type protein and the S688Y mutant (Fig 2, 3).

Analysis demonstrated that there was only one binding pose for glycine with a docking score of -5.28 kcal/mol whereas there were 2 poses present for D-serine with a Glide XP score of -7.37 kcal/mol and -6.53 kcal/mol, respectively. However, for the S688Y GluN1 LBD there were 6 possible docked conformations for glycine with 2 of the poses present with a positive docking score of 2.58 kcal/mol and 2.62 kcal/mol which indicates unfavourable interactions. There was a total of 9 output poses for D-serine with docking scores ranging from -4.67 kcal/mol to -1.79 kcal/mol. In both ligands the S688Y docking score was less negative compared to the wild-type, which can be indicative of ligand affinity in the receptor site.

Microsecond MD analysis

To assess the stability of the system and its behaviours we conducted 1 μs MD simulations on the apo state, top binding poses for each ligand and wild-type, and S688Y mutant. Overall, the 1 μs MD simulations demonstrated that the simulation system consisting of the GluN1 LBD in solvent was stable with ligand RMSD of both glycine and D-serine staying below 1 Å.

Simulations performed with equivalent conditions but without the ligands of interest demonstrated no significant fluctuations with protein RMSD fluctuating and stabilising around 4 Å from the input structure. Furthermore, RMSD of flexible loop regions within the protein are consistent between the simulations (Supp. Fig. 2).

D-serine demonstrated a slightly higher RMSD value in the range of 0.6-0.9 Å compared to that of glycine which remained around 0.2 Å for most of the simulation (Supp. Fig. 3). Ligand-receptor interactions during the entire simulation for all 4 simulation systems remained consistent with our observations from the docking experiments. Solvent accessibility of the ligand in the wild-type LBD during the entire course of the simulation was primarily limited to the ammonium group of glycine and D-serine. In the case of D-serine, the side chain hydroxyl group was also solvent accessible for 5% of simulation time. The solvent accessible surface area (SASA) (Fig. 4) for glycine bound to wild-type protein reached a peak of ~12 Å² and stabilised in the range of 0-4 Å² for most of the simulation while the SASA for D-serine bound to the wild-type protein fluctuated between ~6 Å² and a more solvent accessible conformation of ~12 Å².

Analysis of the trajectories of the S688Y receptor showed that there was extensive water accessibility of the ligand compared to the wild-type protein with solvent contact of the ligand reaching as high as 71% of simulation time. The SASA for glycine bound to the S688Y LBD demonstrated stability around 4 Å² until ~450 ns where the loss of interactions with Thr518 and Asp732 (Supp. Figure 4) led to a large increase in SASA reaching as high as 45 Å². For D-serine bound to the S688Y LBD there was an initial period of high solvent accessibility however SASA stabilised around 10-15 Å². There were no large changes in the SASA due to changes in the Thr518 and Asp732 interactions, as they remained stable for the duration of the simulation.

Ligand contact with the guanidinium side chain of the key residue Arg523 was present in 99% of frames in the wild-type receptor. Ligand contact with Ser688 revealed a preference for the main chain nitrogen but not the side chain as seen in the docking poses with hydrogen bonds with the main chain nitrogen present for 76% of simulation time for glycine and 94% in D-serine while hydrogen bonds for the side chain was at 15% and 27% of simulation time for glycine and D-serine, respectively. Ligand interactions with the guanidium side chain of Arg523 were lost upon mutation for glycine however D-serine was able to maintain hydrogen bonds with the guanidium side chain of Arg523 in the S688Y GluN1 LBD.

Randomly seeded MD simulations

To determine the reproducibility of the simulations and to potentially observe any other plausible protein-ligand interactions we performed five 200 ns simulations that were randomly seeded to have alternative starting velocities, and one that used the identical seed as in the microsecond production runs, for each of the 4 simulation systems. (Supp. Table 1). When using the same seed of the μs simulations, the 200 ns simulation gave results identical to the first 200 ns of the longer production run. As anticipated, there were no variations within the same seed calculations. Compared to the μs simulation the randomly seeded simulations for both glycine and D-serine bound to the wild-type protein were consistent and had similar protein-ligand contacts for all simulations. Solvent contacts were present in 3 out of 6 simulations for glycine and present in all simulations for D-serine. In the wild-type protein all interactions observed were consistent with those observed in the microsecond simulations. SASA for both glycine and D-serine stayed consistent around 1Å² with a maximum of 10Å².

In the S688Y mutated protein solvent exposure was observed in all simulations with both glycine and D-serine. Both glycine and D-serine bound to the S688Y mutated protein experienced an average SASA in the range of 15-30Å² for all seeds, contrasting with observations of the wild-type protein. Interestingly, seed 1293 of glycine with S688Y mutated protein demonstrated a complete dissociation of glycine from the receptor binding site.

Binding Free Energy Calculations

To estimate the binding free energy of glycine and D-serine we performed MM‑GBSA free energy calculations to approximate the ΔG_bind to both the wild-type and S688Y mutant receptors for an estimation of the relative overall binding affinity of the ligand to the receptor and the individual components (Table 2).

Table 1 Binding free energy, ΔG_bind,results for (A) glycine bound to wild-type receptor (B) glycine bound to S688Y mutant receptor (C) D-serine bound to wild-type receptor and (D) D-serine bound to S688Y mutant receptor.

(A)

	ΔG_bind	ΔG_bind coulomb	ΔG_bindcovalent	ΔG_bindH‑bond	ΔG_bindLipophilicity	ΔG_bindSolvent	ΔG_bindvan der Waals
Mean (kcal/mol)	-30.0	-5.62	0.09	-5.05	-2.43	-7.04	-9.97
Standard Deviation	3.08	4.27	0.42	0.54	0.15	3.56	2.64

(B)

	ΔG_bind	ΔG_bind coulomb	ΔG_bindcovalent	ΔG_bindH‑bond	ΔG_bindLipophilicity	ΔG_bindSolvent	ΔG_bindvan der Waals
Mean (kcal/mol)	-26.3	11.8	-0.109	-3.08	-2.80	-18.9	-13.2
Standard Deviation	1.94	3.44	0.59	0.26	0.10	3.33	1.78

(C)

ΔG_bind

ΔG_bind coulomb

ΔG_bindcovalent

ΔG_bind

H-bond

ΔG_bind

lipophilicity

ΔG_bindSolvent

ΔG_bindvan der Waals

Mean (kcal/mol)

-34.6

-6.62

1.09

-5.00

-4.03

-6.18

-13.8

Standard Deviation

3.21

4.90

0.54

0.36

0.18

3.47

2.45

(D)

ΔG_bind

ΔG_bind coulomb

ΔG_bindcovalent

ΔG_bindH‑bond

ΔG_bindLipophilicity

ΔG_bindSolvent

ΔG_bindvan der Waals

Mean

(kcal/mol)

-22.7

7.58

0.84

-3.32

-3.18

-9.65

-14.9

Standard Deviation

2.03

3.42

0.63

0.29

0.31

2.92

1.32

Compared to in the wild-type receptor, the calculated binding free energy in the S688Y mutant receptor was increasedfrom -30.0 kcal/mol to -26.3 kcal/mol for glycine and -34.6 kcal/mol to -22.7 kcal/mol for D-serine. This result is consistent with previous electrophysiology studies on the S688Y mutant where the EC₅₀ of the ligands were increased significantly compared to the native receptor [56]. The contribution from electrostatic interactions (ΔG_bindcoulomb) increased from -5.62 kcal/mol to 11.8 kcal/mol for glycine and -6.62 kcal/mol to 7.58 kcal/mol for D-serine. This strongly supports that it is no longer energetically favourable for the system to form salt bridges which is consistent with results from the MD trajectories where electrostatic interactions between the ligand and Arg523 were present in far fewer frames in the S688Y protein compared to the wild-type, especially for the glycine ligand where Arg523 interactions were almost completely absent. There were minor changes to other binding affinities which includes an increased contribution from hydrogen bonding (ΔG_bindhydrogen bonding) in both glycine and D-serine post-mutation which is consistent with our docking results. Lipophilic interactions (ΔG_bindlipophilicity) decreased slightly for glycine which reflects our previous observations where the ligand environment is slightly more hydrophobic (Fig 2A, 2C), however this did not change significantly for D-serine.

Solvent interactions (ΔG_bindsolvent) became more favourable, changing from -7.04 kcal/mol to -18.9 kcal/mol for glycine and similarly a drop from -6.18 kcal/mol to -9.65 kcal/mol for D-serine which reflects our observations from the simulations where there was a significant increase in solvent exposure in the mutated protein. This greater extent of ΔG_bindsolvent decrease for glycine is also consistent with our observations where complexes containing glycine had a higher number of water bridges in the S688Y receptor compared to D-serine which had a modest increase. Van der Waals interactions (ΔG_bindvan der Waals) also decreased for glycine from -9.97 kcal/mol to -13.2 kcal/mol and -13.8 kcal/mol to -14.9 kcal/mol for D-serine. Which may be due to the change in environment around the glycine to one which is more hydrophobic.

In the microsecond simulations of the NMDAR ligand binding domain and ligand there was a demonstrable increase in the presence of water bridges interacting with both glycine and D-serine in the mutated receptor containing the S688Y mutation compared to wild-type. As seen in previous X-ray structures [7] a small quantity of 5 water molecules is normally in contact with the ligand and is involved in binding for ligands at the GluN1 LBD and this is also observed in our microsecond simulations of wild-type protein with both glycine and D-serine (Fig. 5A, 5B). However, in the mutated receptor the water in contact with the ligand increased (Fig. 5C, 5D) relative to that in the wild-type receptor. There was also instability observed with the SASA of the ligands where values hovered around 5–10 Å for both glycine and D-serine in the wild-type receptor but increased significantly in the mutated receptor. This is important to note as the bulk water in contact water in contact with the ligand can behave differently to structured, isolated water within the binding site [57] and instead contribute to instability of the system.

Loss and weakening of key interactions are also observed in the simulations which contributes to an increase of ΔG_bind and especially the key interaction of the guanidium group Arg523 with the ligand. This interaction is absent in the S688Y receptor with glycine in the microsecond simulation, and present in fewer than 20% of frames in the random seeded simulations. The loss of this key interaction, combined with the change in the binding pocket, contributed to the increase in ΔG_bind for both glycine and D-serine.

Results from the ΔG_bind calculations show that there is a significant decrease in binding affinity for the S688Y mutant LBD compared to that of the wild-type protein with an increase of 11.9 kcal/mol in ΔG_bind for D-serine, and an increase of 3.75 kcal/mol in ΔG_bind for glycine. This is of particular importance due to GluN1/GluN2A’s higher affinity for D-serine compared to glycine in the forebrain regions [58] and that synaptic NMDARs are gated by D-serine [59–62]. ΔG_bind coulomb values increase to 11.83 kcal/mol and 7.58 kcal/mol for glycine and D-serine respectively upon mutation shows that the interference of Tyr688 between the guanidium group of Arg523 and the ligand was able to disrupt this key interaction present in the wild-type NMDA GluN1 LBD. However, the rotamer state for Tyr688 in this study differs from those previously reported [56]. This difference leads to a change in conformation for other residues in the receptor binding site and is significant due to the interference with Arg523. Due to the small size of the glycine and D-serine ligands this observed difference is potentially very significant for understanding the binding of the ligand.

Traditionally MD simulations are carried out on the nanosecond scale, however we have performed microsecond simulations which gives us confidence that the possible ligand poses are well sampled, and this combined with randomly seeded 200 ns simulations helps to overcome issues of interpretation from reproducibility of MD simulations, and has ensured that the initial system is stable under multiple conditions. Therefore, we are confident of our approach of using the refined docked structure as a starting point for the simulation. Use of the OPLS3e force field is widespread in simulations [63, 64]. Our choice of the SPC water model is appropriate, despite potential advantages of other models [65, 66].

Our docked poses agree with previously published X-ray structures and shows that this approach is suitable for studying the detailed changes which may occur between the wild-type and mutated receptor. The use of the XP scoring function allows the elimination of false positive docking poses to ensure confidence in the generated poses. Furthermore, differences in docking scores between the wild-type receptor and the S688Y receptor reported by the XP scoring function are consistent with calculated free energy of binding from the MD simulation trajectories.

From our results the S688Y mutation leads to poor binding of both glycine and D-serine to the NMDAR GluN1 subunit. This may result in partial or non-activation of the receptor, which results in issues during development.

Early infantile epileptic encephalopathies are a debilitating group of conditions which have multiple causes. The existence of mutations within the NMDAR GluN1 subunit is well documented in the literature, however mutations within the LBD have only been discovered very recently, and a complete understanding of the impacts of these mutations on receptor function is yet to be achieved. Our current approach attempts to bridge this gap by utilising computational methods, however future work would benefit from the modelling and simulation of the full receptor assembly to determine how the binding of the ligand affects other interactions activating the receptor. Experimental determination of the mutated S688Y receptor structure would also be welcome, as would other mutant NMDAR structures, and assist the study of the available ligand binding modes, and ultimately the design of agents that can modulate the activity of this receptor.

Acknowledgements

This project was undertaken with the assistance of resources and services from the National Computational Infrastructure (NCI), which is supported by the Australian Government. The authors would also like to acknowledge the technical assistance and computing resources provided by the Sydney Informatics Hub, a Core Research Facility at The University of Sydney. We would also like to acknowledge Mr Ali Kusay for his maintenance of the Balle MD cluster at the Brain and Mind Centre at The University of Sydney.

The National Deuteration Facility is partly supported by the National Collaborative Research Infrastructure Strategy – an initiative of the Australian Government.

Funding

This work was supported by the National Computational Merit Allocation Scheme (NCMAS) 2022 (Grant number NCMAS 2022-154) and the Sydney Informatics Hub HPC Access Scheme 2022. ZC is in receipt of an Australian Government Research Training Program Scholarship.

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Author Contributions

ZC contributed to the conceptualization, validation, formal analysis, investigation, resources, writing (original draft, review and editing), data curation, visualisation, supervision, and project administration. WBC contributed to conceptualization, resources, preparation, and writing (original draft, review and editing) of the manuscript, supervision, project administration, and funding acquisition. KB contributed to the conceptualization, validation and writing (review and editing). AD contributed to validation, and writing (review and editing) of the manuscript. TB contributed to conceptualization, validation, resources, writing (review and editing), supervision, project administration, and funding acquisition.

Notes

Simulation initialisation files are publicly available on our GitHub repository: jak-chen/s688y_2021.

Simulation data and other materials will be provided upon request.

Hwang S-K, Kwon S. Early-onset epileptic encephalopathies and the diagnostic approach to underlying causes. Korean J Pediatr. 2015;58(11):407-14. https://doi.org/10.3345/kjp.2015.58.11.407.
Allen A S, Berkovic S F, Cossette P, Delanty N, Dlugos D, Eichler E E, et al. De novo mutations in epileptic encephalopathies. Nature. 2013;501(7466):217-21. https://doi.org/10.1038/nature12439.
Stroebel D, Casado M, Paoletti P. Triheteromeric NMDA receptors: from structure to synaptic physiology. Current Opinion in Physiology. 2018;2:1-12. https://doi.org/https://doi.org/10.1016/j.cophys.2017.12.004.
Traynelis S F, Wollmuth L P, Mcbain C J, Menniti F S, Vance K M, Ogden K K, et al. Glutamate Receptor Ion Channels: Structure, Regulation, and Function. Pharmacol Rev. 2010;62(3):405. https://doi.org/10.1124/pr.109.002451.
Paoletti P, Bellone C, Zhou Q. NMDA receptor subunit diversity: impact on receptor properties, synaptic plasticity and disease. Nature Reviews Neuroscience. 2013;14(6):383-400. https://doi.org/10.1038/nrn3504.
Kleckner N W, Dingledine R. Requirement for glycine in activation of NMDA-receptors expressed in Xenopus oocytes. Science. 1988;241(4867):835. https://doi.org/10.1126/science.2841759.
Furukawa H, Gouaux E. Mechanisms of activation, inhibition and specificity: crystal structures of the NMDA receptor NR1 ligand-binding core. The EMBO journal. 2003;22(12):2873-85. https://doi.org/10.1093/emboj/cdg303.
Tsai G E, Yang P, Chang Y-C, Chong M-Y. D-Alanine Added to Antipsychotics for the Treatment of Schizophrenia. Biol Psychiatry. 2006;59(3):230-4. https://doi.org/https://doi.org/10.1016/j.biopsych.2005.06.032.
Johnson J W, Ascher P. Glycine potentiates the NMDA response in cultured mouse brain neurons. Nature. 1987;325(6104):529-31. https://doi.org/10.1038/325529a0.
Benveniste M, Mayer M L. Structure-activity analysis of binding kinetics for NMDA receptor competitive antagonists: the influence of conformational restriction. Br J Pharmacol. 1991;104(1):207-21. https://doi.org/https://doi.org/10.1111/j.1476-5381.1991.tb12409.x.
Clements J D, Westbrook G L. Activation kinetics reveal the number of glutamate and glycine binding sites on the N-methyl-d-aspartate receptor. Neuron. 1991;7(4):605-13. https://doi.org/10.1016/0896-6273(91)90373-8.
Mayer M L, Westbrook G L, Guthrie P B. Voltage-dependent block by Mg2+ of NMDA responses in spinal cord neurones. Nature. 1984;309(5965):261-3. https://doi.org/10.1038/309261a0.
Nowak L, Bregestovski P, Ascher P, Herbet A, Prochiantz A. Magnesium gates glutamate-activated channels in mouse central neurones. Nature. 1984;307(5950):462-5. https://doi.org/10.1038/307462a0.
Pagano J, Giona F, Beretta S, Verpelli C, Sala C. N-methyl-d-aspartate receptor function in neuronal and synaptic development and signaling. Curr Opin Pharmacol. 2021;56:93-101. https://doi.org/https://doi.org/10.1016/j.coph.2020.12.006.
Chou T-H, Tajima N, Romero-Hernandez A, Furukawa H. Structural Basis of Functional Transitions in Mammalian NMDA Receptors. Cell. 2020;182(2):357-71.e13. https://doi.org/https://doi.org/10.1016/j.cell.2020.05.052.
Jalali-Yazdi F, Chowdhury S, Yoshioka C, Gouaux E. Mechanisms for Zinc and Proton Inhibition of the GluN1/GluN2A NMDA Receptor. Cell. 2018;175(6):1520-32.e15. https://doi.org/10.1016/j.cell.2018.10.043.
Regan M C, Grant T, Mcdaniel M J, Karakas E, Zhang J, Traynelis S F, et al. Structural Mechanism of Functional Modulation by Gene Splicing in NMDA Receptors. Neuron. 2018;98(3):521-9.e3. https://doi.org/10.1016/j.neuron.2018.03.034.
Tajima N, Karakas E, Grant T, Simorowski N, Diaz-Avalos R, Grigorieff N, et al. Activation of NMDA receptors and the mechanism of inhibition by ifenprodil. Nature. 2016;534(7605):63-8. https://doi.org/10.1038/nature17679.
Meyerson J R, Kumar J, Chittori S, Rao P, Pierson J, Bartesaghi A, et al. Structural mechanism of glutamate receptor activation and desensitization. Nature. 2014;514(7522):328-34. https://doi.org/10.1038/nature13603.
Karakas E, Furukawa H. Crystal structure of a heterotetrameric NMDA receptor ion channel. Science. 2014;344(6187):992. https://doi.org/10.1126/science.1251915.
Papouin T, Oliet S H R. Synaptic and Extra-Synaptic NMDA Receptors in the CNS. In: Hashimoto K, editor. The NMDA Receptors. Cham: Springer International Publishing; 2017. p. 19-49.
Furukawa H, Singh S K, Mancusso R, Gouaux E. Subunit arrangement and function in NMDA receptors. Nature. 2005;438(7065):185-92. https://doi.org/10.1038/nature04089.
Balu D T. The NMDA Receptor and Schizophrenia: From Pathophysiology to Treatment. Advances in pharmacology (San Diego, Calif). 2016;76:351-82. https://doi.org/10.1016/bs.apha.2016.01.006.
Jayawickrama G S, Sadig R R, Sun G, Nematollahi A, Nadvi N A, Hanrahan J R, et al. Kynurenine Aminotransferases and the Prospects of Inhibitors for the Treatment of Schizophrenia. Curr Med Chem. 2015;22(24):2902-18. https://doi.org/http://dx.doi.org/10.2174/0929867322666150608094054.
Coyle J T. NMDA receptor and schizophrenia: a brief history. Schizophr Bull. 2012;38(5):920-6. https://doi.org/10.1093/schbul/sbs076.
Guochuan E T, Pao-Yen L. Strategies to Enhance N-Methyl-D-Aspartate Receptor-Mediated Neurotransmission in Schizophrenia, a Critical Review and Meta-Analysis. Curr Pharm Des. 2010;16(5):522-37. https://doi.org/http://dx.doi.org/10.2174/138161210790361452.
Majláth Z, Török N, Toldi J, Vécsei L. Memantine and Kynurenic Acid: Current Neuropharmacological Aspects. Curr Neuropharmacol. 2016;14(2):200-9. https://doi.org/10.2174/1570159x14666151113123221.
Lau C G, Zukin R S. NMDA receptor trafficking in synaptic plasticity and neuropsychiatric disorders. Nature Reviews Neuroscience. 2007;8(6):413-26. https://doi.org/10.1038/nrn2153.
Zehavi Y, Mandel H, Zehavi A, Rashid M A, Straussberg R, Jabur B, et al. De novo GRIN1 mutations: An emerging cause of severe early infantile encephalopathy. Eur J Med Genet. 2017;60(6):317-20. https://doi.org/https://doi.org/10.1016/j.ejmg.2017.04.001.
Ohba C, Shiina M, Tohyama J, Haginoya K, Lerman-Sagie T, Okamoto N, et al. GRIN1 mutations cause encephalopathy with infantile-onset epilepsy, and hyperkinetic and stereotyped movement disorders. Epilepsia. 2015;56(6):841-8. https://doi.org/https://doi.org/10.1111/epi.12987.
Lemke J R, Geider K, Helbig K L, Heyne H O, Schütz H, Hentschel J, et al. Delineating the GRIN1 phenotypic spectrum. Neurology. 2016;86(23):2171. https://doi.org/10.1212/WNL.0000000000002740.
Hamdan Fadi f, Gauthier J, Araki Y, Lin D-T, Yoshizawa Y, Higashi K, et al. Excess of De Novo Deleterious Mutations in Genes Associated with Glutamatergic Systems in Nonsyndromic Intellectual Disability. The American Journal of Human Genetics. 2011;88(3):306-16. https://doi.org/10.1016/j.ajhg.2011.02.001.
Redin C, Gérard B, Lauer J, Herenger Y, Muller J, Quartier A, et al. Efficient strategy for the molecular diagnosis of intellectual disability using targeted high-throughput sequencing. J Med Genet. 2014;51(11):724. https://doi.org/10.1136/jmedgenet-2014-102554.
Chen W, Shieh C, Swanger S A, Tankovic A, Au M, Mcguire M, et al. GRIN1 mutation associated with intellectual disability alters NMDA receptor trafficking and function. J Hum Genet. 2017;62(6):589-97. https://doi.org/10.1038/jhg.2017.19.
Chen W, Tankovic A, Burger P B, Kusumoto H, Traynelis S F, Yuan H. Functional Evaluation of a De Novo <em>GRIN2A</em> Mutation Identified in a Patient with Profound Global Developmental Delay and Refractory Epilepsy. Mol Pharmacol. 2017;91(4):317. https://doi.org/10.1124/mol.116.106781.
Ogden K K, Chen W, Swanger S A, Mcdaniel M J, Fan L Z, Hu C, et al. Molecular Mechanism of Disease-Associated Mutations in the Pre-M1 Helix of NMDA Receptors and Potential Rescue Pharmacology. PLoS Genet. 2017;13(1):e1006536. https://doi.org/10.1371/journal.pgen.1006536.
Rossi M, Chatron N, Labalme A, Ville D, Carneiro M, Edery P, et al. Novel homozygous missense variant of GRIN1 in two sibs with intellectual disability and autistic features without epilepsy. European journal of human genetics : EJHG. 2017;25(3):376-80. https://doi.org/10.1038/ejhg.2016.163.
Delano W L. The PyMOL Molecular Graphics System. http://wwwpymolorg. 2002. doi:https://ci.nii.ac.jp/naid/10020095229/en/.
Schrödinger. Schrödinger Release 2020-3. New York, NY: Schrödinger LLC; 2020.
Bowers K J, Chow E, Xu H, Dror R O, Eastwood M P, Gregersen B A, et al. Scalable algorithms for molecular dynamics simulations on commodity clusters. Proceedings of the 2006 ACM/IEEE conference on Supercomputing. Tampa, Florida: Association for Computing Machinery; 2006. p. 84–es.
Jespersen A, Tajima N, Fernandez-Cuervo G, Garnier-Amblard Ethel c, Furukawa H. Structural Insights into Competitive Antagonism in NMDA Receptors. Neuron. 2014;81(2):366-78. https://doi.org/10.1016/j.neuron.2013.11.033.
Madhavi Sastry G, Adzhigirey M, Day T, Annabhimoju R, Sherman W. Protein and ligand preparation: parameters, protocols, and influence on virtual screening enrichments. J Comput Aided Mol Des. 2013;27(3):221-34. https://doi.org/10.1007/s10822-013-9644-8.
Olsson M H M, Søndergaard C R, Rostkowski M, Jensen J H. PROPKA3: Consistent Treatment of Internal and Surface Residues in Empirical pKa Predictions. J Chem Theory Comput. 2011;7(2):525-37. https://doi.org/10.1021/ct100578z.
Søndergaard C R, Olsson M H M, Rostkowski M, Jensen J H. Improved Treatment of Ligands and Coupling Effects in Empirical Calculation and Rationalization of pKa Values. J Chem Theory Comput. 2011;7(7):2284-95. https://doi.org/10.1021/ct200133y.
Roos K, Wu C, Damm W, Reboul M, Stevenson J M, Lu C, et al. OPLS3e: Extending Force Field Coverage for Drug-Like Small Molecules. J Chem Theory Comput. 2019;15(3):1863-74. https://doi.org/10.1021/acs.jctc.8b01026.
Friesner R A, Murphy R B, Repasky M P, Frye L L, Greenwood J R, Halgren T A, et al. Extra Precision Glide: Docking and Scoring Incorporating a Model of Hydrophobic Enclosure for Protein−Ligand Complexes. J Med Chem. 2006;49(21):6177-96. https://doi.org/10.1021/jm051256o.
Halgren T A, Murphy R B, Friesner R A, Beard H S, Frye L L, Pollard W T, et al. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 2. Enrichment Factors in Database Screening. J Med Chem. 2004;47(7):1750-9. https://doi.org/10.1021/jm030644s.
Friesner R A, Banks J L, Murphy R B, Halgren T A, Klicic J J, Mainz D T, et al. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 1. Method and Assessment of Docking Accuracy. J Med Chem. 2004;47(7):1739-49. https://doi.org/10.1021/jm0306430.
Banks J L, Beard H S, Cao Y, Cho A E, Damm W, Farid R, et al. Integrated Modeling Program, Applied Chemical Theory (IMPACT). J Comput Chem. 2005;26(16):1752-80. https://doi.org/10.1002/jcc.20292.
Berendsen H J C, Postma J P M, Van Gunsteren W F, Hermans J. Interaction Models for Water in Relation to Protein Hydration. In: Pullman B, editor. Intermolecular Forces: Proceedings of the Fourteenth Jerusalem Symposium on Quantum Chemistry and Biochemistry Held in Jerusalem, Israel, April 13–16, 1981. Dordrecht: Springer Netherlands; 1981. p. 331-42.
Genheden S, Ryde U. The MM/PBSA and MM/GBSA methods to estimate ligand-binding affinities. Expert Opinion on Drug Discovery. 2015;10(5):449-61. https://doi.org/10.1517/17460441.2015.1032936.
Hou T, Wang J, Li Y, Wang W. Assessing the Performance of the MM/PBSA and MM/GBSA Methods. 1. The Accuracy of Binding Free Energy Calculations Based on Molecular Dynamics Simulations. J Chem Inf Model. 2011;51(1):69-82. https://doi.org/10.1021/ci100275a.
Jacobson M P, Pincus D L, Rapp C S, Day T J F, Honig B, Shaw D E, et al. A hierarchical approach to all-atom protein loop prediction. Proteins: Structure, Function, and Bioinformatics. 2004;55(2):351-67. https://doi.org/https://doi.org/10.1002/prot.10613.
Jacobson M P, Friesner R A, Xiang Z, Honig B. On the Role of the Crystal Environment in Determining Protein Side-chain Conformations. J Mol Biol. 2002;320(3):597-608. https://doi.org/https://doi.org/10.1016/S0022-2836(02)00470-9.
Li J, Abel R, Zhu K, Cao Y, Zhao S, Friesner R A. The VSGB 2.0 model: a next generation energy model for high resolution protein structure modeling. Proteins. 2011;79(10):2794-812. https://doi.org/10.1002/prot.23106.
Skrenkova K, Song J-M, Kortus S, Kolcheva M, Netolicky J, Hemelikova K, et al. The pathogenic S688Y mutation in the ligand-binding domain of the GluN1 subunit regulates the properties of NMDA receptors. Sci Rep. 2020;10(1):18576. https://doi.org/10.1038/s41598-020-75646-w.
Maurer M, Oostenbrink C. Water in protein hydration and ligand recognition. Journal of molecular recognition : JMR. 2019;32(12):e2810-e. https://doi.org/10.1002/jmr.2810.
Le Bail M, Martineau M, Sacchi S, Yatsenko N, Radzishevsky I, Conrod S, et al. Identity of the NMDA receptor coagonist is synapse specific and developmentally regulated in the hippocampus. Proceedings of the National Academy of Sciences. 2015;112(2):E204. https://doi.org/10.1073/pnas.1416668112.
Papouin T, Ladépêche L, Ruel J, Sacchi S, Labasque M, Hanini M, et al. Synaptic and Extrasynaptic NMDA Receptors Are Gated by Different Endogenous Coagonists. Cell. 2012;150(3):633-46. https://doi.org/10.1016/j.cell.2012.06.029.
Fossat P, Turpin F R, Sacchi S, Dulong J, Shi T, Rivet J-M, et al. Glial D-Serine Gates NMDA Receptors at Excitatory Synapses in Prefrontal Cortex. Cereb Cortex. 2012;22(3):595-606. https://doi.org/10.1093/cercor/bhr130.
Curcio L, Podda M V, Leone L, Piacentini R, Mastrodonato A, Cappelletti P, et al. Reduced d-serine levels in the nucleus accumbens of cocaine-treated rats hinder the induction of NMDA receptor-dependent synaptic plasticity. Brain. 2013;136(4):1216-30. https://doi.org/10.1093/brain/awt036.
Basu A C, Tsai G E, Ma C L, Ehmsen J T, Mustafa A K, Han L, et al. Targeted disruption of serine racemase affects glutamatergic neurotransmission and behavior. Mol Psychiatry. 2009;14(7):719-27. https://doi.org/10.1038/mp.2008.130.
Pekel H, Ilter M, Sensoy O. Inhibition of SARS-CoV-2 main protease: a repurposing study that targets the dimer interface of the protein. J Biomol Struct Dyn. 2021:1-16. https://doi.org/10.1080/07391102.2021.1910571.
Chakraborti S, Chakraborty M, Bose A, Srinivasan N, Visweswariah S S. Identification of Potential Binders of Mtb Universal Stress Protein (Rv1636) Through an in silico Approach and Insights Into Compound Selection for Experimental Validation. Frontiers in Molecular Biosciences. 2021;8:177. https://doi.org/https://doi.org/10.3389/fmolb.2021.599221.
Zielkiewicz J. Structural properties of water: Comparison of the SPC, SPCE, TIP4P, and TIP5P models of water. The Journal of Chemical Physics. 2005;123(10):104501. https://doi.org/10.1063/1.2018637.
Mark P, Nilsson L. Structure and Dynamics of the TIP3P, SPC, and SPC/E Water Models at 298 K. The Journal of Physical Chemistry A. 2001;105(43):9954-60. https://doi.org/10.1021/jp003020w.

No competing interests reported.

SupplementaryChenZetal2022.docx

Binding and dynamics demonstrate the destabilisation of ligand binding for the S688Y mutation in the NMDA receptor GluN1 subunit

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1