Excitatory and inhibitory D-serine binding to the NMDA receptor
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
Activation of NMDA receptors requires two co-agonists: Glutamate which binds to the GluN2 subunit and glycine/D-serine which binds to the GluN1 subunit. In the present manuscript, the authors address the interaction of D-serine, which is a less studied co-agonist than glycine, with the GluN1 and GluN2A subunits using molecular simulations as well as electrophysiology experiments. Surprisingly they find that D-serine interacts with the GluN2 subunit, further expanding our molecular understanding of NMDA receptor structure-function.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their names with the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Structural Biology and Molecular Biophysics (eLife)
- Reading list (BiophysicsColab)
Abstract
N-methyl-D-aspartate receptors (NMDARs) uniquely require binding of two different neurotransmitter agonists for synaptic transmission. D-serine and glycine bind to one subunit, GluN1, while glutamate binds to the other, GluN2. These agonists bind to the receptor’s bi-lobed ligand-binding domains (LBDs), which close around the agonist during receptor activation. To better understand the unexplored mechanisms by which D-serine contributes to receptor activation, we performed multi-microsecond molecular dynamics simulations of the GluN1/GluN2A LBD dimer with free D-serine and glutamate agonists. Surprisingly, we observed D-serine binding to both GluN1 and GluN2A LBDs, suggesting that D-serine competes with glutamate for binding to GluN2A. This mechanism is confirmed by our electrophysiology experiments, which show that D-serine is indeed inhibitory at high concentrations. Although free energy calculations indicate that D-serine stabilizes the closed GluN2A LBD, its inhibitory behavior suggests that it either does not remain bound long enough or does not generate sufficient force for ion channel gating. We developed a workflow using pathway similarity analysis to identify groups of residues working together to promote binding. These conformation-dependent pathways were not significantly impacted by the presence of N-linked glycans, which act primarily by interacting with the LBD bottom lobe to stabilize the closed LBD.
Article activity feed
-
-
Author Response
Reviewer #1 (Public Review):
This is a study that is aimed at understanding the binding mechanism of D-serine to the two different binding lobes of the NMDA receptor. D-serine is a known agonist and binder of the GluN1 ligand-binding domain, but its interaction with the GluN2A is unknown. Using long time-scale conventional molecular dynamics simulations, the researchers observe that D-serine interacts and associates readily with both binding domains, often via protein surface pathways referred to as a guided-diffusion mechanism. As observed previously, free-energy calculations show that D-serine stabilizes the closure of both binding domains. Finally, analysis of the effect of glycans shows that these modifications play a role in further stabilizing the closed state of the ligand-binding domains.
Amongst this broad …
Author Response
Reviewer #1 (Public Review):
This is a study that is aimed at understanding the binding mechanism of D-serine to the two different binding lobes of the NMDA receptor. D-serine is a known agonist and binder of the GluN1 ligand-binding domain, but its interaction with the GluN2A is unknown. Using long time-scale conventional molecular dynamics simulations, the researchers observe that D-serine interacts and associates readily with both binding domains, often via protein surface pathways referred to as a guided-diffusion mechanism. As observed previously, free-energy calculations show that D-serine stabilizes the closure of both binding domains. Finally, analysis of the effect of glycans shows that these modifications play a role in further stabilizing the closed state of the ligand-binding domains.
Amongst this broad and careful analysis, the major finding from this work is that D-serine surprisingly associates with GluN2A, which has been known to bind glutamate to enable activation of the channel. Since the binding of D-serine to GluN2A had not been observed previously, they proposed that D-serine acts as an inhibitor for glutamate at high concentrations. This hypothesis was investigated and supported by electrophysiological experiments, yielding a novel result that presents new interpretations for the field. However, the guided-diffusion mechanism still remains hypothetical and is unclear as to whether this is in fact a driving force, or requirement, for the binding. Specifically, the following questions warrant further investigation:
- Specific or non-specific association? It is possible that non-specific association events of ligands to the protein could be an intrinsic artifact of the MD simulations. To investigate this, it would be informative to compare the current results with a negative control simulation where the ligand was replaced with a similar amino acid or molecule that has been verified as a non-binder for NMDAR.
To address this, we quantified the non-specific association signal by comparing the number of successful binding events to random association (see response to Essential Revisions #4). In theory, any appropriately small amino acid could associate with the conserved arginine of each LBD through its C-terminus (as evidenced by our PMF of glycine bound to GluN2A). However, an amino acid’s ability to remain bound long enough to induce LBD closure is largely dependent on the presence of interactions with the LBD bottom lobe.
- Dissociation events? Further clarification is required to understand whether any dissociation events are observed in these simulations to the non-specific sites or the final binding site. If dissociation is not observed, how does this impact the interpretation of the binding mechanisms that characterize only the association events?
Association and dissociation are both observed and documented in Datasets S2-S4. We added clarification to the text on page 5 about the nature of both processes and how pathways are defined by residues that allow the agonist to enter and leave the binding site. As illustrated in the clustering dendrograms, association (even-numbered events) and dissociation (odd-numbered events) pathways are present in all clusters.
- Testing the hypothesis of guided diffusion. It is proposed that guided diffusion drives serine binding to its site. This would imply that the residues on this path are important, and if mutated, would decrease the association rate and the ability to compete with glutamate. Additional electrophysiological experiments or direct binding experiments would be useful in understanding the relevance of guided diffusion in the ligand-binding mechanism of NMDARs.
To address this point, we performed additional TEVC experiments generating D-serine dose-response curves for GluN1a Arg694Ala and Arg695Ala, and GluN2A Arg692Ala and Arg695Ala. The curves for both GluN2A mutants support our guided diffusion mechanism, as they lowered the D-serine inhibition potency (These mutants also likely also alter glutamate binding, but since D-serine and glutamate bind through the same residues, it is not possible to separate out individual contributions.) The GluN1a mutants did not show altered behavior, supporting the increased diffusiveness of D-serine binding to GluN1 compared to GluN2A. These additional findings are included in the main text on page 12 and in Fig. 4D.
Reviewer #2 (Public Review):
In this manuscript, Yovanno et. al. did a comprehensive mechanistic study of D-serine binding to NMDAR ligand-binding domains (LBDs). The framework of the current investigation is built upon this research group's previous studies of NMDAR agonists glutamate and glycine binding. Using an aggregated 51 microseconds of all-atom MD simulations of spontaneous binding, the authors applied rigorous pathway similarity analysis to cluster the paths through which D-serine enters the LBDs from the bulk solution. The most interesting and unexpected result from this study is the spontaneous binding of D-serine to the GluN2A LBD, which was previously known to be the glutamate binding site.
By computing the overlap coefficient for all binding pathways, the authors concluded that D-serine binding to GluN2A LBD through "guided" diffusion, while to GluN1 through random diffusion (the clustered pathways comprise random contacts rather than specific, conserved residue contacts). A "guided" binding pathway further suggests that the agonist binding could be sensitive to the conformational change within and around the binding pocket, and vice versa.
To investigate whether D-serine binding events are able to modulate the GluN2A LBD conformation, the authors then computed a series of LBD conformational free energy landscapes (2D-PMF) using 2D-umbrella sampling simulations. The 2D-PMF profiles confirmed that D-serine stabilizes the closed LBD conformation, just like glutamate. Because the D-serine 2D-PMF shows a metastable state that was absent in glutamate 2D-PMF, the authors argue that D-serine may not stabilize the closed conformation to the same extent as glutamate. Likewise, based on the 2D-PMF of GluN1 LBD, the authors suggest that D-serine has a higher potency than glycine, in part due to its ability to more strongly stabilize a closed LBD conformation.
The simulations above generated the hypothesis that D-serine could function as a competitive antagonist of glutamate at high concentrations. This computationally derived hypothesis is beautifully tested by the authors' dose-response curves and the Schild plot.
One question that would merit further clarification is whether the binding affinity of D-serine to the two LBDs is stronger or weaker in comparison with glutamate and glycine. The difference in agonist potency could be due to the difference in binding affinity and/or efficacy. Stabilizing the closed LBD conformation may indicate the efficacy of the agonist, but affinity (Kd) will still play a role in the final potency.
Indeed, as Reviewer 2 pointed out, affinity should play a role since the D-serine inhibition here is attributed to the competitive binding of D-serine against glutamate as we showed with our Schild plot. The bona fide binding site for D-serine is GluN1 LBD where D-serine binds more strongly than glycine (Furukawa/Gouaux 2003). In the GluN1 LBD, D-serine is a full agonist. The D-serine binding to the GluN2A LBD (the finding here) is substantially weaker (mM) than glutamate (~1 uM).
While a glycosylated GluN1/GluN2A dimer was used for the majority of MD simulations, the authors also checked the "reality" by mapping the pathway residues onto the NMDAR heterotetramer structure. The role of glycans in D-serine binding pathways was further investigated by conducting an additional 30 microseconds simulations of the non-glycosylated dimer. It was found that glycans introduced small kinetic "traps" that slow down the binding process. Glycan was also found to stabilize LBD closure from 1D-PMF profiles.
The detailed mechanistic insight and D-serine's inhibitory effect on NMDAR, unraveled by this study, may play an important role in therapeutic strategies, and thus is likely to have a broad impact in the field.
-
Evaluation Summary:
Activation of NMDA receptors requires two co-agonists: Glutamate which binds to the GluN2 subunit and glycine/D-serine which binds to the GluN1 subunit. In the present manuscript, the authors address the interaction of D-serine, which is a less studied co-agonist than glycine, with the GluN1 and GluN2A subunits using molecular simulations as well as electrophysiology experiments. Surprisingly they find that D-serine interacts with the GluN2 subunit, further expanding our molecular understanding of NMDA receptor structure-function.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their names with the authors.)
-
Reviewer #1 (Public Review):
This is a study that is aimed at understanding the binding mechanism of D-serine to the two different binding lobes of the NMDA receptor. D-serine is a known agonist and binder of the GluN1 ligand-binding domain, but its interaction with the GluN2A is unknown. Using long time-scale conventional molecular dynamics simulations, the researchers observe that D-serine interacts and associates readily with both binding domains, often via protein surface pathways referred to as a guided-diffusion mechanism. As observed previously, free-energy calculations show that D-serine stabilizes the closure of both binding domains. Finally, analysis of the effect of glycans shows that these modifications play a role in further stabilizing the closed state of the ligand-binding domains.
Amongst this broad and careful analysis, …
Reviewer #1 (Public Review):
This is a study that is aimed at understanding the binding mechanism of D-serine to the two different binding lobes of the NMDA receptor. D-serine is a known agonist and binder of the GluN1 ligand-binding domain, but its interaction with the GluN2A is unknown. Using long time-scale conventional molecular dynamics simulations, the researchers observe that D-serine interacts and associates readily with both binding domains, often via protein surface pathways referred to as a guided-diffusion mechanism. As observed previously, free-energy calculations show that D-serine stabilizes the closure of both binding domains. Finally, analysis of the effect of glycans shows that these modifications play a role in further stabilizing the closed state of the ligand-binding domains.
Amongst this broad and careful analysis, the major finding from this work is that D-serine surprisingly associates with GluN2A, which has been known to bind glutamate to enable activation of the channel. Since the binding of D-serine to GluN2A had not been observed previously, they proposed that D-serine acts as an inhibitor for glutamate at high concentrations. This hypothesis was investigated and supported by electrophysiological experiments, yielding a novel result that presents new interpretations for the field. However, the guided-diffusion mechanism still remains hypothetical and is unclear as to whether this is in fact a driving force, or requirement, for the binding. Specifically, the following questions warrant further investigation:
1. Specific or non-specific association? It is possible that non-specific association events of ligands to the protein could be an intrinsic artifact of the MD simulations. To investigate this, it would be informative to compare the current results with a negative control simulation where the ligand was replaced with a similar amino acid or molecule that has been verified as a non-binder for NMDAR.
2. Dissociation events? Further clarification is required to understand whether any dissociation events are observed in these simulations to the non-specific sites or the final binding site. If dissociation is not observed, how does this impact the interpretation of the binding mechanisms that characterize only the association events?
3. Testing the hypothesis of guided diffusion. It is proposed that guided diffusion drives serine binding to its site. This would imply that the residues on this path are important, and if mutated, would decrease the association rate and the ability to compete with glutamate. Additional electrophysiological experiments or direct binding experiments would be useful in understanding the relevance of guided diffusion in the ligand-binding mechanism of NMDARs.
-
Reviewer #2 (Public Review):
In this manuscript, Yovanno et. al. did a comprehensive mechanistic study of D-serine binding to NMDAR ligand-binding domains (LBDs). The framework of the current investigation is built upon this research group's previous studies of NMDAR agonists glutamate and glycine binding. Using an aggregated 51 microseconds of all-atom MD simulations of spontaneous binding, the authors applied rigorous pathway similarity analysis to cluster the paths through which D-serine enters the LBDs from the bulk solution. The most interesting and unexpected result from this study is the spontaneous binding of D-serine to the GluN2A LBD, which was previously known to be the glutamate binding site.
By computing the overlap coefficient for all binding pathways, the authors concluded that D-serine binding to GluN2A LBD through …
Reviewer #2 (Public Review):
In this manuscript, Yovanno et. al. did a comprehensive mechanistic study of D-serine binding to NMDAR ligand-binding domains (LBDs). The framework of the current investigation is built upon this research group's previous studies of NMDAR agonists glutamate and glycine binding. Using an aggregated 51 microseconds of all-atom MD simulations of spontaneous binding, the authors applied rigorous pathway similarity analysis to cluster the paths through which D-serine enters the LBDs from the bulk solution. The most interesting and unexpected result from this study is the spontaneous binding of D-serine to the GluN2A LBD, which was previously known to be the glutamate binding site.
By computing the overlap coefficient for all binding pathways, the authors concluded that D-serine binding to GluN2A LBD through "guided" diffusion, while to GluN1 through random diffusion (the clustered pathways comprise random contacts rather than specific, conserved residue contacts). A "guided" binding pathway further suggests that the agonist binding could be sensitive to the conformational change within and around the binding pocket, and vice versa.
To investigate whether D-serine binding events are able to modulate the GluN2A LBD conformation, the authors then computed a series of LBD conformational free energy landscapes (2D-PMF) using 2D-umbrella sampling simulations. The 2D-PMF profiles confirmed that D-serine stabilizes the closed LBD conformation, just like glutamate. Because the D-serine 2D-PMF shows a metastable state that was absent in glutamate 2D-PMF, the authors argue that D-serine may not stabilize the closed conformation to the same extent as glutamate. Likewise, based on the 2D-PMF of GluN1 LBD, the authors suggest that D-serine has a higher potency than glycine, in part due to its ability to more strongly stabilize a closed LBD conformation.
The simulations above generated the hypothesis that D-serine could function as a competitive antagonist of glutamate at high concentrations. This computationally derived hypothesis is beautifully tested by the authors' dose-response curves and the Schild plot.
One question that would merit further clarification is whether the binding affinity of D-serine to the two LBDs is stronger or weaker in comparison with glutamate and glycine. The difference in agonist potency could be due to the difference in binding affinity and/or efficacy. Stabilizing the closed LBD conformation may indicate the efficacy of the agonist, but affinity (Kd) will still play a role in the final potency.
While a glycosylated GluN1/GluN2A dimer was used for the majority of MD simulations, the authors also checked the "reality" by mapping the pathway residues onto the NMDAR heterotetramer structure. The role of glycans in D-serine binding pathways was further investigated by conducting an additional 30 microseconds simulations of the non-glycosylated dimer. It was found that glycans introduced small kinetic "traps" that slow down the binding process. Glycan was also found to stabilize LBD closure from 1D-PMF profiles.
The detailed mechanistic insight and D-serine's inhibitory effect on NMDAR, unraveled by this study, may play an important role in therapeutic strategies, and thus is likely to have a broad impact in the field.
-
Reviewer #3 (Public Review):
Activation of NMDA receptors requires two co-agonists: Glutamate which binds to the GluN2 subunit and glycine/D-serine which binds to the GluN1 subunit. In the present manuscript, the authors address the interaction of D-serine, which is a less studied co-agonist than glycine, with the GluN1 and GluN2A subunits using molecular simulations as well as electrophysiology experiments.
Initial molecular dynamic simulations surprisingly reveal that D-serine interacts not only with the GluN1 agonist-binding domain but also with that of the GluN2A subunit. The authors characterize mechanisms associated with the GluN1 and GluN2A binding including assaying, using pathway similarity analysis, whether free diffusion or guided diffusion is predominant for the two subunits. Electrophysiological experiments are used to test …
Reviewer #3 (Public Review):
Activation of NMDA receptors requires two co-agonists: Glutamate which binds to the GluN2 subunit and glycine/D-serine which binds to the GluN1 subunit. In the present manuscript, the authors address the interaction of D-serine, which is a less studied co-agonist than glycine, with the GluN1 and GluN2A subunits using molecular simulations as well as electrophysiology experiments.
Initial molecular dynamic simulations surprisingly reveal that D-serine interacts not only with the GluN1 agonist-binding domain but also with that of the GluN2A subunit. The authors characterize mechanisms associated with the GluN1 and GluN2A binding including assaying, using pathway similarity analysis, whether free diffusion or guided diffusion is predominant for the two subunits. Electrophysiological experiments are used to test the idea that D-serine inhibits glutamate activity at the GluN2A subunit. The authors also address how N-glycans positioned around the binding cleft for GluN1 and GluN2A impact agonist binding.
Overall, the results add to our molecular understanding of agonist binding to the GluN1 and GluN2 agonist binding pocket. The results for D-serine interacting with the GluN2A agonist pocket are surprising but probably should not have been (no one has addressed these questions). The conclusions of the manuscript are generally supported by the simulations and functional experiments. The manuscript is well written and laid out quite clearly.
-
-