Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Jul 23;182(2):357-371.e13.
doi: 10.1016/j.cell.2020.05.052. Epub 2020 Jun 30.

Structural Basis of Functional Transitions in Mammalian NMDA Receptors

Affiliations

Structural Basis of Functional Transitions in Mammalian NMDA Receptors

Tsung-Han Chou et al. Cell. .

Abstract

Excitatory neurotransmission meditated by glutamate receptors including N-methyl-D-aspartate receptors (NMDARs) is pivotal to brain development and function. NMDARs are heterotetramers composed of GluN1 and GluN2 subunits, which bind glycine and glutamate, respectively, to activate their ion channels. Despite importance in brain physiology, the precise mechanisms by which activation and inhibition occur via subunit-specific binding of agonists and antagonists remain largely unknown. Here, we show the detailed patterns of conformational changes and inter-subunit and -domain reorientation leading to agonist-gating and subunit-dependent competitive inhibition by providing multiple structures in distinct ligand states at 4 Å or better. The structures reveal that activation and competitive inhibition by both GluN1 and GluN2 antagonists occur by controlling the tension of the linker between the ligand-binding domain and the transmembrane ion channel of the GluN2 subunit. Our results provide detailed mechanistic insights into NMDAR pharmacology, activation, and inhibition, which are fundamental to the brain physiology.

Keywords: N-methyl-D-aspartate receptors; NMDAR antagonists; NMDARs; X-ray crystallography; channel activation and inhibition mechanisms; cryo-EM; electron cryo-microscopy; ligand-gated ion channels.

PubMed Disclaimer

Conflict of interest statement

Declaration of Interests The authors declare no competing interests.

Figures

Figure 1.
Figure 1.. Domain organization and architecture of NMDARs.
(A) Domain organization of NMDAR subunits. (B) Cryo-EM density of the GluN1b-GluN2B NMDARs bound to glycine and glutamate (the ‘Non-active1’ 3D class) showing the GluN1b-GluN2B ATD heterodimer (orange dashed lines), the GluN1b-GluN2B LBD heterodimer (black dashed lines), and the dimer of the GluN1b-GluN2B LBD heterodimer (cyan dashed lines). (C-E) The structural model of the intact heterotetramer (C), the extracellular domains (D) (ATD top and LBD bottom), and the gate region focusing on the pore forming M3 and M3’ helices (top) and the channel (bottom) (E). A line and a dotted line in (D) show distances between GluN1 α5 and GluN2B α4’ and between GluN1 L2 and GluN2B L1’, which are used to define conformational states of the intact NMDARs. See also Figure S1.
Figure 2.
Figure 2.. Conformational variability of agonist-bound NMDARs.
3D classes of the GluN1b-GluN2B NMDAR bound to glycine and glutamate dubbed ‘Non-active1,’Non-active 2,’ and ‘Active’ and the structure of the ‘Active-SS’ construct in magenta (GluN1b) and dark green (GluN2B). Structures of ‘Non-active2,’ “Active,’ and ‘Active-SS’ are aligned to that of ‘Non-active1’ (gray) in panels A-C. (A-B) The conformational states can be defined by the α4’-α5 distance (spheres and lines) between GluN1b and GluN2B ATDs controlled by ‘rolling’ motions of GluN2B ATDs (A), extent of opening of the GluN2B ATD (A), and the L1’-L2 distance (spheres and arrowed dots) between the GluN1b-Glu2B heterodimers controlled by ‘rolling’ motion of the GluN1b-GluN2B LBD heterodimers (B). (C-E) Tension of the GluN2B LBD-M3’ linkers measured by the distances between the two Gln662 Cαs (spheres; panel C) is the major determent for opening or closing of the channel gate (D-E). See also Figures S1–S3, Table S1, and Video S1.
Figure 3
Figure 3. Antagonist binding and structures of GluN1-GluN2A LBDs.
(A-B) A GluN1-antagonist L689,560 and a GluN2-antagonist SDZ-220-040 can inhibit NMDARs in the presence of glutamate and glycine, respectively. TEVC recording on the WT and the EM constructs show the similar pattern of inhibition (error bars ±SD) (B). (C-E) Crystal structure of the GluN1-GluN2A LBD complexed to L689,560 and glutamate showing domain opening (D1/D2) of the GluN1 LBD bi-lobe by 28° compared to the glycine-bound GluN1 LBD (D). (F-H) Crystal structure of the GluN1-GluN2A LBD complexed to glycine and SDZ-220-040 showing domain opening (D1/D2) of the GluN2A LBD bi-lobe by 23° compared to the glutamate-bound GluN2A LBD (G). In both structures, electron density of ligands (Fo-Fc omit map contoured at 3σ in green mesh) is sufficiently resolved to pinpoint polar interactions (dotted lines) and hydrophobic interactions (E and H). See also Table S2–S3.
Figure 4
Figure 4. Cryo-EM structure of GluN1b-GluN2B NMDAR complexed to glycine and SDZ-220-040 (Gly/SDZ).
(A) Cryo-EM density of the Gly/SDZ (Class 1) in the same color code as in Fig. 1. (B) Comparison of the GluN2B LBD bi-lobe in the Gly/SDZ (dark green) to that in the ‘Active-SS’ bound to glutamate (gray) showing the antagonist induced domain opening by 24°. (C) Cryo-EM density for SDZ-022-040 (cyan mesh) at the bi-lobe cleft. (D-E) In this structure, the subunit and domain arrangements are similar to those of the ‘Active-SS’ (gray) as represented by the ‘short’ α4’-α5 distance (spheres and a line) in the ATDs (D) and the ‘short’ L1’-L2 distance (spheres and arrowed dots) in the LBDs (E). (F) However, the GluN2B LBD-M3’ linkers are more relaxed compared to those in the ‘Active-SS’ (gray) due to opening of the GluN2B LBD bi-lobes as in panel B. See also Figure S4, Table S1, and Video S2.
Figure 5
Figure 5. Cryo-EM structure of GluN1b-GluN2B NMDAR complexed to L689,560 and glutamate (L689/Glu).
(A) Cryo-EM density of the L689/Glu (Class 1) in the same color code as in Fig. 1. (B) Comparison of the GluN1b LBD bi-lobe in the L689/Glu (magenta) compared to that in the ‘Active-SS’ bound to glutamate (gray) showing the antagonist induced domain opening by 12-15°. (C) Cryo-EM density for L689,560 (green mesh) at the bi-lobe cleft. (D-F) The subunit and domain arrangements of L689/Glu are distinct from ‘Active-SS’ as shown by rolling motions (arrows), longer α4’-α5 (D) and L1’-L2 (E) distances, and more closed GluN2B ATD bi-lobes (D). (F) Consequently, the GluN2B LBD-M3’ linkers are relaxed. (G-H) The subunit and domain arrangement of L689/Glu are similar to those in the ‘Non-active1’ (light gray) as represented by the similar α4’-α5 (G) and L1’-L2 (H) distances. (I) The GluN1b LBD-M3 linkers are similar between L689/Glu and ‘Active-SS.’ See also Figure S5–S6, Table S1 and Video S3.
Figure 6
Figure 6. Cryo-EM structure of GluN1b-GluN2B NMDAR complexed to L689,560 and SDZ-220-040 (L689/SDZ).
(A) Cryo-EM density of the L689/SDZ (Class 1) in the same color code as in Fig. 1. (B) Comparison of the GluN1b LBD bi-lobe complexed to L689,560 (magenta) and the GluN2B LBD bi-lobe complexed to SDZ-220-040 (dark green) against the ‘Active-SS’ complexed to glycine and glutamate (gray) showing antagonist-induced domain opening. (C) Cryo-EM density for L689,560 (green mesh) and SDZ-220-040 (cyan mesh) at the bi-lobe clefts. (D-E) In this structure, the subunit arrangements and domain arrangement is similar to that of the ‘Active-SS’ (gray) as represented by the ‘short’ α4’-α5 distance in the ATDs (D) and the ‘short’ L1’-L2 distance in the LBDs (E). (F) As in the Gly/SDZ, the GluN2B LBD-M3’ linkers are relaxed compared to the ‘Active-SS’ (gray) due to opening of the GluN2B LBD bi-lobe. Little or no change in the GluN1b LBD-M3 linkers but a change in the orientation of the GluN2B LBD-M4’ linkers is observed. (G) Consequently, the hydrophobic interaction at the channel gate is strengthened by GluN2B Val808 (yellow ovals). See also Figure S7, Table S1, and Video S4.
Figure 7
Figure 7. Schematic presentation of conformational transitions.
(A-B) The agonist-bound GluN1b-GluN2B NMDAR reside in ‘Non-active’ and ‘Active’ conformations where they differ in the subunit arrangement within the GluN1b-GluN2B ATD heterodimer and the orientation between the GluN1b-GluN2B LBD heterodimers as a result of their rigid-body movement (arrows indicating the movement from ‘Non-active1’ to ‘Active-SS’). These conformations can be defined by the distances between GluN1 α5 (red rectangle) and GluN2B α4’ (yellow rectangle) at the ATDs and/or the distances between GluN1 L2 (orange oval) and GluN2B L1’ (cyan oval) at the LBDs where they are ‘long’ and ‘short’ in ‘Non-active’ and ‘Active,’ respectively. For clarity, only ‘Non-active1’ and ‘Active-SS’ are displayed here. (C) Binding of L689,560 and glutamate to GluN1 and GluN2B, respectively, leads to ‘Non-active1’-like conformation where the inter-subunit distances above are ‘long.’ (D) In contrast, binding of SDZ-220-040 in the presence of glutamate or L689,560 opens the GluN2B LBD cleft and stabilizes ‘Active-SS’-like conformation where the inter-subunit distances above are ‘short.’ In the L689/SDZ, the gate is further closed by extra hydrophobic interactions. Black arrows indicate the rigid-body movement of the ATD and LBD heterodimers from ‘Non-active1’ to ‘Gly/SDZ’ or ‘L689/SDZ.’ Red arrows indicate LBD bi-lobe opening by antagonist binding.

References

    1. Adams PD, Afonine PV, Bunkoczi G, Chen VB, Davis IW, Echols N, Headd JJ, Hung LW, Kapral GJ, Grosse-Kunstleve RW, et al. (2010). PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr 66, 213–221. - PMC - PubMed
    1. Amico-Ruvio SA, and Popescu GK (2010). Stationary gating of GluN1/GluN2B receptors in intact membrane patches. Biophys J 98, 1160–1169. - PMC - PubMed
    1. Banke TG, Dravid SM, and Traynelis SF (2005). Protons trap NR1/NR2B NMDA receptors in a nonconducting state. J Neurosci 25, 42–51. - PMC - PubMed
    1. Banke TG, and Traynelis SF (2003). Activation of NR1/NR2B NMDA receptors. Nat Neurosci 6, 144–152. - PubMed
    1. Benveniste M, and Mayer ML (1991). Kinetic analysis of antagonist action at N-methyl-D-aspartic acid receptors. Two binding sites each for glutamate and glycine. Biophys J 59, 560–573. - PMC - PubMed

Publication types

MeSH terms