UCM-A86 is a selective positive allosteric modulator of GluN1/GluN3 NMDA receptors

Abstract

N-methyl-d-aspartate (NMDA) receptors are ionotropic glutamate receptors that mediate excitatory neurotransmission in the central nervous system (CNS) where they play critical roles in normal and pathological brain functions and neurodevelopment. While the glutamate/glycine-activated GluN2-containing NMDA receptors (GluN1/GluN2) have been extensively studied, the physiological roles and pharmacology of glycine-activated GluN3-containing receptors (GluN1/GluN3) remain less understood. Although GluN1/GluN3 receptors exhibit unique functional properties and play distinct roles in neuronal development and synapse maturation, studies of their precise roles in neurophysiology and circuit function are impeded by limited availability of GluN3-selective pharmacological tools. This study describes UCM-A86, a novel GluN3-selective positive allosteric modulator, with EC₅₀ values of 21 μM and 19 μM at GluN1/GluN3A and GluN1/GluN3B receptors, respectively. UCM-A86 selectively potentiates recombinant GluN1/GluN3A and GluN1/GluN3B receptors by 436% and 174%, respectively, relative to activation by glycine, with no activity at recombinant GluN1/GluN2A-D receptors. Furthermore, UCM-A86 selectively potentiates responses from native GluN1/GluN3A receptors expressed in somatostatin-expressing interneurons of the somatosensory cortex with no modulation of hippocampal AMPA receptor- and GluN1/2 NMDA receptor-mediated excitatory postsynaptic currents. Mechanistic studies suggest that UCM-A86 modulation is facilitated by agonist binding (or channel gating) and that UCM-A86 primarily potentiates GluN1/GluN3A by increasing open probability with no effects on mean channel conductance. These findings advance the synthetic pharmacology of GluN1/GluN3 receptors and provide a novel tool for modulation of native GluN3-containing NMDA receptors.

Keywords: Ionotropic glutamate receptors; allosteric modulation; drug discovery; ligand-gated ion channels; neuropharmacology; synaptic transmission.

Figure 1.

Effects of UCM-A86 on recombinant GluN3A-containing NMDA receptors. (A) Chemical structure of UCM-A86. (B) Representative two-electrode voltage-clamp recording of modulation by 30 μM UCM-A86 at recombinant GluN1–1a^FATL/3A receptors. (C) The bar graph summarizes the responses shown in B. Responses were normalized to the glycine only response (100%). Data are mean ± SEM from 7 oocytes. Significance determined by one-sample t-test compared to 100% (****, p < 0.0001). (D) Representative recordings of UCM-A86 concentration-responses data at GluN1–4a^CGP/3A and GluN1–4a^CGP/3B receptors. (E) GluN1–4a^FATL/3A-B and wildtype GluN1–4a^CGP/3A-B receptors were activated by glycine (30 μM) in the absence and presence of increasing concentrations of UCM-A86. Data are mean ± SEM from 10 to 16 oocytes. (F) Representative recordings evaluating glycine concentration-response data in the presence and absence of 30 μM UCM-A86. (G) Glycine concentration-response data ± 30 μM UCM-A86 for GluN1–1a^FATL/3A receptors normalized to initial response to 1 mM glycine (100% control) in the absence of UCM-A86. Data are mean ± SEM from 12 to 16 oocytes. (H) Bar graphs show the maximum responses and EC₅₀ values in the absence (control) and presence of UCM-A86. Statistical significance was analyzed using unpaired t-test (****, p < 0.0001). See Table 1 for maximum responses and EC₅₀ values.

Figure 2.

GluN1 splice variant effects on UCM-A86 modulation. (A) Schematic representation of the polypeptide chain for relevant GluN1 splice variants co-expressed with GluN3A. (B) Concentration-response data measured using two-electrode voltage-clamp electrophysiology demonstrating modulation by UCM-A86 of GluN1 splice variants co-expressed with GluN3A. F484A + T518L (FATL) mutations or CGP were used to prevent desensitization and receptors were activated by 30 μM glycine followed by increasing concentrations of UCM-A86. Data are mean ± SEM from 10 to 16 oocytes. (C-D) Summary of EC₅₀ values and maximum potentiation for UCM-A86 at GluN1 splice variants co-expressed with GluN3A. See Table 1 for statistical analysis.

Figure 3.

Selectivity of UCM-A86 for GluN3- over GluN2-containing NMDA receptors. (A-E) Representative two-electrode voltage-clamp recordings for GluN1–1a/2A-D receptors activated by 100 μM glycine and 100 μM glutamate (control) and GluN1–1a^FATL/3A receptors activated by 30 μM glycine (control), followed by the addition of 30 μM UCM-A86. (F) Summary of UCM-A86 responses relative to control. Statistical significance between GluN1–1a/2A-D receptors and GluN1–1a^FATL/3A receptors was determined using one-way ANOVA (****, p < 0.0001).

Figure 4.

UCM-A86 potentiates glycine-mediated responses in mouse somatostatin (SST) neurons. (A) Coronal brain slice showing the primary somatosensory cortex (S1) from which whole-cell patch clamp recordings were made from GFP-expressing SST cells. (B) Representative recording demonstrating current responses from SST cells in response to pressure applied 10 mM glycine at baseline (pink), in the presence of 1 μM CGP-78608 (black), and following bath application of 20 μM AIMS 86 (blue). (C) Representative recording demonstrating current responses from SST cells in response to pressure applied 10 mM glycine at baseline (pink), in the presence of 1 μM CGP-78608 (black), and following bath application of 0.15% DMSO vehicle (purple). (D) Cumulative time course showing the effect of AIMS 86 on glycine-mediated responses. (E) Summary data demonstrates a significant potentiation of glycine-mediated currents after bath application of AIMS 86. (***, p < 0.001; unpaired t-test).

Figure 5.

UCM-A86 lacks activity at neuronal AMPA and GluN2-containing NMDA receptors. (A) Schematic representation of the dorsal hippocampus from mouse brain. (B) Representative NMDAR-EPSCs recorded from CA1 pyramidal cells at baseline and in the presence of 20 μM UCM-A86. (C-D) UCM-A86 (20 μM) does not modulate evoked NMDAR-EPSCs in the mouse hippocampus. (ns, p>0.05; paired t-test) (E) Representative AMPAR-EPSCs recorded from CA1 pyramidal cells at baseline and in the presence of 20 μM UCM-A86. (F-G) UCM-A86 (20 μM) does not modulate evoked AMPAR-EPSCs in the mouse hippocampus. Significance determined using a paired t-test (ns, p > 0.05). (H) Cumulative time course data showing the lack of effect of 20 μM UCM- on AMPAR- and NMDAR-EPSCs.

Figure 6.

UCM-A86 modulation is facilitated by glycine binding or channel gating. (A) Representative whole-cell patch-clamp recordings demonstrating differences in the time course of responses from GluN1–1a/3A receptors activated by long (3 s) exposures to glycine and in the absence (control) or continous presence of UCM-A86. CGP-78608 (1 μM) was continously present during the recordings. (B) Representative recordings of glycine responses from GluN1–1a/3A receptors in the absence of UCM-A86 or with UCM-A86 present pre and post exposure to glycine. CGP-78608 (1 μM) was continously present during the recordings. (C) Summary of rise times for glycine responses in control, pre/post UCM-A86 exposure, and continous UCM-A86 exposure conditions. Significance determined by one-way ANOVA with Tukey post-test (****, p < 0.0001; ns, p > 0.05). (D) Summary potentiation of glycine responses in pre/post UCM-A86 exposure and continous UCM-A86 exposure conditions relative to control. Significance determined by unpaired t-test (**, p < 0.01). (E) Summary of desensitization for glycine responses in control, pre/post UCM-A86 exposure, and continous UCM-A86 exposure conditions. Significance determined by a one-way ANOVA with Tukey post-test (****, p < 0.0001; ns, p > 0.05). (F) Representative recordings of responses to brief (5 ms) exposures to glycine in the absence (control) or presence of UCM-A86. (G-H) Summaries of rise times and deactivation times for responses to brief (5 ms) glycine exposures in the absence (control) or presence of UCM-A86. Significance determined by unpaired t-test (****, p < 0.0001). See Table 2 for summary of all kinetic parameters.

Figure 7.

UCM-A86 modulation is mediated by a multistep mechanism with no effect on mean channel conductance. (A) Representative whole-cell patch-clamp recordings in the continuous presence of 1 μM CGP. Responses were activated by 10 mM glycine and modulated by co-application of either 10 μM or 30 μM UCM-A86. (B) Bar graph summarizing the weighted time constants for onset and offset of UCM-A86 potentiation. Significance determined by two-way ANOVA with Tukey post-test (ns, p > 0.05). (C) Bar graph summarizing potentiation relative to the response to glycine. Significance determined by unpaired t-test (ns, p > 0.05). (D) Representative whole-cell patch-clamp recordings of responses to bath-applied 10 mM glycine the continous presence of 1 μM CGP, but in the absence (vehicle, 0.15% DMSO) or presence of 30 μM UCM-A86. The same recording filtered at 1 Hz are shown above to illustrate noise arising from channel gating. (E) Current-variance plots from two different exposed to glycine alone (vehicle control) or glycine + UCM-A86. (F) Bar graph summarizing mean channel conductance measured from cells exposed to glycine alone (vehicle control) or glycine + UCM-A86. Significance determined by unpaired t-test (ns, p > 0.05).