Distinct GluN1 and GluN2 Structural Determinants for Subunit-Selective Positive Allosteric Modulation of N-Methyl-d-aspartate Receptors

Abstract

N-Methyl-d-aspartate receptors (NMDARs) are ionotropic ligand-gated glutamate receptors that mediate fast excitatory synaptic transmission in the central nervous system (CNS). Several neurological disorders may involve NMDAR hypofunction, which has driven therapeutic interest in positive allosteric modulators (PAMs) of NMDAR function. Here we describe modest changes to the tetrahydroisoquinoline scaffold of GluN2C/GluN2D-selective PAMs that expands activity to include GluN2A- and GluN2B-containing recombinant and synaptic NMDARs. These new analogues are distinct from GluN2C/GluN2D-selective compounds like (+)-(3-chlorophenyl)(6,7-dimethoxy-1-((4-methoxyphenoxy)methyl)-3,4-dihydroisoquinolin-2(1H)-yl)methanone (CIQ) by virtue of their subunit selectivity, molecular determinants of action, and allosteric regulation of agonist potency. The (S)-enantiomers of two analogues (EU1180-55, EU1180-154) showed activity at NMDARs containing all subunits (GluN2A, GluN2B, GluN2C, GluN2D), whereas the (R)-enantiomers were primarily active at GluN2C- and GluN2D-containing NMDARs. Determination of the actions of enantiomers on triheteromeric receptors confirms their unique pharmacology, with greater activity of (S) enantiomers at GluN2A/GluN2D and GluN2B/GluN2D subunit combinations than (R) enantiomers. Evaluation of the (S)-EU1180-55 and EU1180-154 response of chimeric kainate/NMDA receptors revealed structural determinants of action within the pore-forming region and associated linkers. Scanning mutagenesis identified structural determinants within the GluN1 pre-M1 and M1 regions that alter the activity of (S)-EU1180-55 but not (R)-EU1180-55. By contrast, mutations in pre-M1 and M1 regions of GluN2D perturb the actions of only the (R)-EU1180-55 but not the (S) enantiomer. Molecular modeling supports the idea that the (S) and (R) enantiomers interact distinctly with GluN1 and GluN2 pre-M1 regions, suggesting that two distinct sites exist for these NMDAR PAMs, each of which has different functional effects.

Keywords: EPSC; Electrophysiology; molecular dynamics; positive allosteric modulator.

Figure 1.

EU1180–55 potentiates the response of NMDARs. (A) CIQ (compound 1) selectively potentiates the maximal response of only GluN2C- and GluN2D-containing NMDARs following activation with saturating concentrations of glutamate and glycine (left panel) or EC₃₀ agonist concentrations (right panel), which were (for glutamate and glycine) 1.5 and 0.8 μM for GluN1/GluN2A, 1.5 and 0.45 μM for GluN1/GluN2B, 0.85 and 0.2 μM for GluN1/GluN2C, and 0.3 and 0.07 μM for GluN1/GluN2D. The fitted EC₅₀ value for CIQ potentiation of GluN1/GluN2C responses (in saturating agonist) to 340% ± 27% of control was 17 μM (12–18 μM 95% CI); the fitted EC₅₀ value was 15 μM (14–17 μM) for potentiation of GluN1/GluN2D to 281% ± 10% of control. The EC₅₀ value for CIQ potentiation of NMDAR activation by EC₃₀ concentrations of glutamate and glycine (right panel) could not be determined. (B) EU1180–55 (compound 2) potentiates NMDARs during activation by saturating concentrations of agonists (left panel) to 226% ± 10% of control with an EC₅₀ of 4.6 μM (4.0–5.1 μM) for GluN1/GluN2B, to 259% ± 27% of control with an EC₅₀ of 1.9 μM (1.4–2.2 μM) for GluN1/GluN2C, to 366% ± 27% of control with an EC₅₀ of 4.1 μM (2.5–4.6 μM) for GluN1/GluN2D. The EC₅₀ values for EU1180–55 potentiation of current responses activated by EC₃₀ concentrations of agonists ranged between 4.3 and 7.9 μM (right panel; Supplemental Table S1). For all experiments, data are from 5 to 22 oocytes from 2 to 3 batches of oocytes. (C) Representative example of two electrode voltage clamp recording showing that 1, 3, 10, 30, and 100 μM EU1180–55 potentiates the response of GluN2B-, GluN2C-, and GluN2D-containing NMDARs expressed in Xenopus oocytes activated by 100 μM glutamate and 30 μM glycine. (D, E) EU1180–55 (30 μM) increases the potency for glutamate and glycine at GluN1/GluN2A NMDA receptors; fitted EC₅₀ values are given in Table 1. (F) GluN1/GluN2A NMDAR response time courses are superimposed in the absence or presence of 30 μM EU1180–55; for EU1180–55 experiments, the compound was included in both the wash and the glutamate solution; 30 μM glycine was present in all solutions. The fitted time constants describing the deactivation time course are in Table 2. Plot of Tau_FAST vs concentration for 3, 10, and 30 μM EU1180–55 (n = 3 cells). (G) EU1180–55 enhances glutamate potency for GluN1/GluN2A at concentrations above 3 μM, as assessed by the concentration–response curve for EU1180–55 enhancement of the fractional current response to a submaximal concentration of glutamate (1 μM) in the presence of saturating glycine (30 μM), expressed as a percent of the response to 100 μM glutamate and 30 μM glycine (n = 5–6 oocytes for each point).

Figure 2.

Subunit-selective actions of EU1180–55 enantiomers. (A_1,2) Left panel, representative whole cell current response from a HEK cell expressing GluN1/GluN2A to 100 μM glutamate and 30 μM glycine (plus vehicle, 0.075% DMSO), followed by an immediate switch to glutamate/glycine plus 15 μM (S)-EU1180–55 (A₁) or (R)-EU1180–55 (A₂). Middle panel, normalized mean whole cell current time course recorded in response to 1.5 s application of 100 μM glutamate and 30 μM glycine in the absence (black) and presence of 15 μM (S)-EU1180–55 (red) or (R)-EU1180–55 (blue); EU1180–55 and glycine were in the wash solution prior to application of glutamate. Right panel, Plot of weighted mean tau in vehicle or test compound. The same experiment is shown for GluN1/GluN2B (B_1,2), GluN1/GluN2C (C_1,2), and GluN1/GluN2D (D_1,2). (E, F) Fold change in steady-state amplitude (E) or weighted tau (F) for (S)-EU1180–55 or (R)-EU1180–55. Fitted time constants for all experiments are given in Table 2 (weighted tau) and Supplemental Table S2 (tau_FAST, tau_SLOW, relative contribution for tau_FAST). * p < 0.05, paired t test.

Figure 3.

Potentiation of triheteromeric NMDARs in response to (S)-EU1180–55 or (R)-EU1180–55 in the presence of 100 μM glutamate and 30 μM glycine in Xenopus oocytes: concentration–response relationships for potentiation of diheteromeric NMDARs (A, C) and triheteromeric NMDARs (B, D) by (S)-EU1180–55 or (R)-EU1180–55. Experiments were performed 2–3 times from different batches of oocytes. Each data point is represented as mean ± SEM (see Table 3 for EC₅₀ values; n ≥ 8 oocytes for all data points).

Figure 4.

Effect of EU1180–55 on NMDARs in hippocampal neurons. (A) Representative photomicrograph of a neuron maintained 12 days in culture (calibration 20 μm). The patch electrode can be seen on the right side. (B) Typical current response to 100 μM NMDA and 3 μM glycine (black bar). After the current relaxed to a steady-state level, EU1180–55 (10 μM, dark gray bar) was coapplied with NMDA and glycine, which increased the current response to NMDA in the presence of bicuculline (10 μM), CNQX (10 μM), and TTX (0.5 μM) to 200% ± 10% of control (p = 0.001, paired t test, n = 5). After wash out of EU1180–55, 3 μM ifenprodil, the GluN2B-selective inhibitor (light gray bar), was applied. Ifenprodil inhibited the current to 30% ± 10% of control (p = 0.005, paired t test, n = 5). The calibration bar is 50 pA, 20 s. (C) NMDAR-mediated component of evoked EPSCs recorded at −30 mV in the presence of 0.2 mM Mg²⁺ from 300 μm hippocampal slices from P7–14 mice. (D) Left panel, representative evoked NMDAR-mediated EPSC before (black) and after (red) the addition of 20 μM (S)-EU1180–55. Right panel, normalized evoked NMDAR-mediated EPSC before (black) and after (red) the addition of 20 μM (S)-EU1180–55. (E, F) Effect of (S)-EU1180–55 on weighted tau and response amplitude (n = 8, red). (G–I) Similar data as in panels D–F but with DMSO vehicle (n = 8, green). See Supplemental Table S5 for summary of fitted parameters. *p < 0.025, paired t test.

Figure 5.

(S)-EU1180–154 potentiation requires the NMDA receptor TMD. (A) Cartoon representations showing construction of chimeric subunits from GluK2/GluN1 and GluK2/GluN2B. (B–E) Whole-cell currents evoked by 10 μM NMDA and 10 μM glycine (N+g) or 10 μM kainate (K) to transfected HEK cells expressing chimeric receptors. (S)-EU1180–154 was applied as indicated by the gray bars. The effects of 3 μM (S)-EU1180–154 (mean, SEM, N) are summarized in Supplemental Table S6. The holding potential was −80 mV.

Figure 6.

Structural determinants of EU1180–55 actions on GluN1/GluN2B reside in the preM1-M1 region of GluN1. (A) Response in the presence of modulator for GluN1 coexpressed with mutant GluN2B receptors in oocytes relative to the control response in the absence of modulator. (B) Response in the presence of modulator of mutant GluN1 coexpressed with GluN2B relative to the control response in the absence of modulator. Responses were activated by maximally effective concentrations of glutamate and glycine (100/30 μM). For all panels, solid and dashed lines show the mean ratio for WT receptors ±99% confidence intervals. Bars are colored when they diverged more than 20% from the mean response and had nonoverlapping confidence intervals. An alanine was substituted for each residue starting with GluN1 Leu551 and GluN2 Pro547, except when alanine was the original amino acid, in which case it was changed to a cysteine. Data are from 4 to 33 oocytes from 1 to 4 batches of oocytes. Error bars are 99% confidence intervals. Positions at which mutations altered the response to EU1180–55 are mapped onto a spacefill model of the pore forming region (GluN1 green, GluN2 yellow).

Figure 7.

Structural determinants of action for (S)-EU1180–55 and (R)-EU1180–55 on GluN1/GluN2D receptors. (A) Results of scanning mutagenesis on racemic 30 μM EU1180–55 modulation of the response to maximally effective glutamate and glycine (100/30 μM) for pre-M1 and M1 residues in GluN2D, starting at Ser570. Alanine was substituted at all residues, except when alanine was the original amino acid, in which case it was changed to a cysteine. Data are from 3 to 24 oocytes from 1 to 5 batches of oocytes for each mutation. The star below indicates residues in GluN2D at which mutations alter the actions of the positive allosteric modulator CIQ. (B) Quantitative evaluation was performed on the effects of 8 GluN2D mutations (B₁, L575A, E576A, P577A, V582A, W583A, V584A, V588A, L591A) and 6 GluN1 mutations (B₂, D552A, S553A, M555A, L562A, G567A, V570A) on 10 μM (S)-EU1180–55 or (R)-EU1180–55 potentiation. For all panels, the solid line shows the mean potentiation of WT GluN1/GluN2D recorded on the same day; dashed lines and error bars are the 99% confidence intervals. Bars are filled red when EU1180–55 reduced current by more than 20% and confidence intervals were nonoverlapping or blue when EU1180–55 enhanced the response by 20% and confidence intervals were nonoverlapping. Data are from 4 to 21 oocytes from 1 to 2 batches of oocytes for each mutation. (C) A model of the transmembrane domain of GluN1/GluN2D with pre-M1 and M1 helices represented as spacefill (GluN1 is green, GluN2 is yellow). Residues are colored to illustrate the actions of the two enantiomers tested at GluN1 and GluN2 mutations shown in panel B.

Figure 8.

(A) Cartoon representation of a GluN1/GluN2D NMDAR homology model bound with EU1180–55 enantiomers in the GluN1 (green box) and GluN2 (blue box) pre-M1 pockets. Blue represents the GluN2 subunits and gray the GluN1 subunits. (left) Representation of the fold symmetry in the NMDAR of the NTD, ABD, and TMD. (B) Rotation (90°) and removal of the ABD and NTD reveals the pseudo-4-fold symmetry of the transmembrane region with the (S) and (R) enantiomers of EU1180–55 bound. Green shows the (S)-EU1180–55 enantiomer, which mutagenesis suggests binds to a pocket on GluN1. White shows (R)-EU1180–55 bound to the analogous pocket on GluN2. (C, D) Expanded view of (S)-EU1180–55 (green) within the GluN1 binding pocket and (R)-EU1180–55 (white) bound within a pocket on GluN2D. If residues were tested in either Figure 6 or 7, the residues numbers are colored using the color scheme in Figures 6 and 7. (E) Overlay of the GluN1 and GluN2D binding pockets with blue showing identical residues and red showing nonidentical residues; only a single ribbon is shown. The binding pocket is conserved with most of the variation within pre-M1 and the linker region between the TMD and the ABD. See Supplemental Figure S8 for more information.