GluN2A-NMDA receptor inhibition disinhibits the prefrontal cortex, reduces forced swim immobility, and impairs sensorimotor gating

Abstract

Recent investigations into the rapid antidepressant effects of ketamine, along with studies on schizophrenia-related susceptibility genes, have highlighted the GluN2A subunit as a critical regulator of both emotion and cognition. However, the specific impacts of acute pharmacological inhibition of GluN2A-containing NMDA receptors on brain microcircuits and the subsequent behavioral consequences remain poorly understood. In this study, we first examined the effects of MPX-004, a selective GluN2A NMDA receptor inhibitor, on behavior within the dorsomedial prefrontal cortex (dmPFC). Local administration of MPX-004 in the dmPFC led to a reduced immobility duration in the forced swim test, an acute antidepressant-like effect, impairments in sensorimotor gating, and a schizophrenia-like phenotype. In vivo multiple-channel recordings and c-Fos staining revealed that MPX-004 decreases the activity of parvalbumin-expressing interneurons (PV-INs) and increases the activity of pyramidal neurons (PYNs). In vivo patch-clamp recordings further confirmed that PV-IN inactivation leads to an elevated PYN firing rate in the PFC. In vitro whole-cell recordings demonstrated that PV-INs receive stronger excitatory synaptic input and respond more robustly to presynaptic stimulation than do somatostatin-expressing interneurons (SST-INs) and PYNs, rendering them susceptible to GluN2A inhibition. Finally, the specific knockdown of GluN2A in prefrontal PV-INs abolished the behavioral effects of MPX-004, underscoring a critical role of the GluN2A-mediated modulation of PV-INs in these phenotypes. Together, these findings reveal that PV-INs are particularly vulnerable to GluN2A inhibition, leading to disinhibition of prefrontal circuits and resulting in both antidepressant-like and schizophrenia-like behaviors.

Keywords: NMDA receptor; disinhibition; microcircuit; prefrontal cortex; psychiatry.

Fig. 1. Inhibitory efficacy and selectivity of MPX-004 and TCN-201 on NMDA receptor subtypes.

a Schematic diagram showing the expression of various NMDA receptor subtypes in HEK293T cells. NMDA receptor-mediated currents were activated by brief (5 ms) exposure to 1 mM glutamate in the continuous presence of glycine at concentrations of 1, 3, 10, or 30 μM. b, c TCN-201 treatment resulted in complete blockade of currents mediated by GluN1/2 A receptors, partial blockade of currents mediated by GluN1/2 A/2B receptors, and no effect on currents mediated by GluN1/2B receptors. c MPX-004 treatment completely blocked currents mediated by both GluN1/2 A and GluN1/2 A/2B but did not affect currents mediated by GluN1/2B receptors. d TCN and MPX-004 did not affect the deactivation time constant of GluN1/2 A, GluN1/2 A/2B or GluN1/2B NMDA receptors (see Supplemental Table 3 for complete statistical information).

Fig. 2. MPX-004 infusion in the dmPFC impairs sensorimotor gating and reduces immobility duration in the forced swim test.

a Schematic diagram illustrating the guide cannula implantation procedure in the dmPFC, followed by vehicle or MPX-004 infusion and subsequent behavioral tests. b An image displaying the implantation coordinates of the guide cannula, accompanied by a schematic illustrating the infusion site position, confirmed through post hoc verification. c Time spent in the center of the open field (10 min). d Traveled distance (10 min). e, f Time course of traveled distance during the locomotion test (30 min). g, h Bar graphs showing no difference in the startle response measured at 65 dB or 120 dB among the vehicle control and treatment groups. i Bar graph demonstrating a significant reduction in prepulse inhibition in the MPX-004 (10 μM) treatment group compared with the vehicle control group. The data were analyzed by two-way ANOVA followed by multiple comparisons using the two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli, *P < 0.05, ***P < 0.001. (j) Immobility time during the forced swim test, showing a significant reduction in immobility when the animals were treated with MPX-004. The data were analyzed by unpaired t test, *P < 0.05 (see Supplemental Table 4 for complete statistical information).

Fig. 3. Local MPX-004 infusion regulates dmPFC neuronal activity.

a Schematic illustration of tetrode recording in the dmPFC with a cannula for drug administration. b Bar graph showing that MPX-004 suppresses the overall firing rate of recorded neurons (2.72 ± 0.22 vs. 1.76 ± 0.15 Hz, Wilcoxon matched-pairs signed-rank test, ****P < 0.0001). c Scatter plot showing the firing rates (FRs) of recorded cells in the vehicle and MPX-004 groups. d Bar graph showing that MPX-004 enhances the firing rates of recorded neurons with basal firing rates lower than 1 Hz (0.55 ± 0.05 vs. 1.10 ± 0.22 Hz, Wilcoxon matched-pairs signed-rank test, *P = 0.017). e Bar graph showing that MPX-004 reduces the firing rates of recorded neurons with basal firing rates higher than 1 Hz (3.39 ± 0.25 vs. 1.96 ± 0.18 Hz, Wilcoxon matched-pairs signed-rank test, ****P < 0.0001). f Classification of a total of 130 recorded cells from six mice based on the trough-to-peak time (0.4 ms). g Z-scored waveforms of narrow-spiking (red) and wide-spiking neurons (blue). h, l MPX-004 significantly reduces the overall firing rates of both narrow- (2.7 ± 0.42 vs. 1.9 ± 0.30 Hz, Wilcoxon matched-pairs signed-rank test, *P = 0.033) and wide-spiking neurons (2.72 ± 0.26 vs. 1.66 ± 0.17 Hz, Wilcoxon matched-pairs signed-rank test, ****P < 0.0001) in the dmPFC. i, m The effects of MPX-004 on the neuron firing rate are activity dependent. j, n When the basal firing rate is less than 1 Hz, MPX-004 enhances the firing rate of wide-spiking neurons (0.54 ± 0.08 vs. 1.18 ± 0.32 Hz, Wilcoxon matched-pairs signed-rank test, *P = 0.036) but not narrow-spiking neurons (0.57 ± 0.06 vs. 0.97 ± 0.27 Hz, Wilcoxon matched-pairs signed-rank test, P = 0.27). k, o Neurons with basal firing rates above 1 Hz show reduced activity in response to MPX-004 in both narrow-spiking (3.54 ± 0.51 vs. 2.32 ± 0.39 Hz, Wilcoxon matched-pairs signed-rank test, **P = 0.003) and wide-spiking neurons (3.33 ± 0.29 vs. 1.80 ± 0.20 Hz, Wilcoxon matched-pairs signed-rank test, ****P < 0.0001). See Supplemental Table 5 for complete statistical information.

Fig. 4. MPX-004 reduces PV-IN activity and enhances PYN activity in the dmPFC.

a Fluorescence images of the dmPFC showing staining for c-Fos (green), PV::Ai14 (red), and DAPI (blue), with zoomed-in regions of layers II/III b and V c. The white arrows indicate cells costained with anti-c-Fos and anti-Ai14 antibodies. d Fluorescence images of the dmPFC showing staining for c-Fos (red), SST::YFP (green), and DAPI (blue), with zoomed-in regions of layers II/III e and V f. The white arrows indicate cells costained with anti-c-Fos and anti-GFP antibodies. g Fluorescence images of the dmPFC showing staining for c-Fos (red), CaMKII (green), and DAPI (blue), with zoomed-in regions of layers II/III h and V i. The white arrows indicate cells costained with anti-c-Fos and anti-CaMKII antibodies. j In layer II/III of the dmPFC, MPX-004 treatment did not significantly alter the percentage of c-Fos⁺ PYNs in WT mice, the percentage of c-Fos⁺ PV-INs in PV::Ai14 mice or the percentage of c-Fos⁺ SST-INs in SST::YFP mice. PV-INs and SST-INs were visualized using reporter fluorescence. The data were analyzed by unpaired t test. k In layer V of the dmPFC, MPX-004 treatment led to a significant increase in the percentage of c-Fos⁺ PYNs, a decrease in the percentage of c-Fos⁺ PV-INs, and no change in the percentage of c-Fos+ SST-INs. The data were analyzed by unpaired t test, *P < 0.05, **P < 0.01. Each condition was assessed in 7–13 slices obtained from 3 mice a–k. l Representative traces of spontaneous excitatory postsynaptic currents (sIPSCs) recorded from PYNs, illustrating baseline activity and activity following the application of 10 μM MPX-004. m Cumulative probability plots of sIPSC amplitude showing no significant change with MPX-004 treatment. However, the cumulative probability plot of sIPSC interevent intervals exhibited a rightward shift, indicating a decreased frequency of sIPSCs. n Bar graph quantifying the sIPSCs showing no change in amplitude but a significant decrease in frequency with MPX-004 application. The data were analyzed by unpaired t test, **P < 0.01. o Representative traces of spontaneous inhibitory postsynaptic currents (sEPSCs) recorded from PYNs at baseline and in the presence of MPX-004. p Cumulative probability plots of sEPSC amplitude showing no significant change with MPX-004 treatment. The cumulative probability plot of sEPSC interevent intervals demonstrated a leftward shift, reflecting an increased frequency of sEPSCs. q Bar graph depicting sEPSCs showing no change in amplitude but a significant decrease in frequency following MPX-004 application. Each condition was assessed in 8–10 cells obtained from 3 mice l–q. See Supplemental Table 6 for complete statistical information.

Fig. 5. PV and SST neurons express similar GluN2A-containing NMDA receptors.

Representative traces of NMDAR-EPSCs recorded from a PV-IN a and SST-IN d at +40 mV, showing baseline activity (blue) and activity in the presence of MPX-004 (red). b, e Time-course graphs depicting the amplitude of NMDAR-EPSCs (blue) and changes in the holding current (red) in PV-INs b and SST-INs e following treatment with 10 μM MPX-004 and 50 μM DL-APV. c, f Graphs showing a decrease in the amplitude of NMDAR-EPSCs in both PV-INs c and SST-INs f following MPX-004 treatment. The data were analyzed by paired t test c and Wilcoxon matched-pairs signed-rank test f, *P < 0.05, **P < 0.01. g Bar graph indicating no significant difference in the percentage of MPX-004-sensitive NMDAR-EPSCs between PV-INs and SST-INs. The data were analyzed by unpaired t test. h Holding current was significantly reduced by MPX-004 treatment in PV-INs but not in SST-INs. The data were analyzed by two-way ANOVA, followed by Sidak’s multiple-comparisons test, *P < 0.05 i The decay time constant of NMDAR-EPSCs was prolonged during MPX-004 treatment in both PV-INs and SST-INs. The data were analyzed by two-way ANOVA, followed by Sidak’s multiple-comparisons test, *P < 0.05. Each condition was assessed in 6–7 slices obtained from 4 mice. See Supplemental Table 7 for complete statistical information.

Fig. 6. Compared with SST-INs, PV-INs receive stronger excitatory input and are more responsive.

a–c Representative traces of spontaneous postsynaptic potentials (PSPs) recorded at the resting membrane potential from a PYN, a PV-IN, and an SST-IN in layer V of the dmPFC. d–f Cumulative probability distributions of the PSP amplitude, interevent interval (I-E-I), and frequency for PYNs, PV-INs, and SST-INs. g Bar graph showing larger PSP amplitudes in PV-INs and SST-INs than in PYNs. h, i Bar graphs illustrating that the PSP frequency is highest in PV-INs among PYNs, PV-INs, and SST-INs (n = 11, 8, and 13 from 3 PV-AI14, 3 SST-YPF and 4 WT mice, respectively). The data were analyzed by one-way ANOVA followed by Tukey’s multiple-comparisons test/Dunnett’s T3 multiple-comparisons test, *P < 0.05, **P < 0.01 for g–i. j Voltage responses (−65 mV) in layer V PV-INs and SST-INs following a train of stimulation (80 μA, 20 Hz, 10 pulses) delivered via a bipolar electrode positioned at layer II/III. k Bar graph showing that, compared with SST-INs and PYNs, PV-INs are more responsive to the same stimulation train (*P < 0.05, Kruskal‒Wallis test followed by Dunn’s multiple-comparisons test, n = 9, 10, and 10 cells from 3 PV-AI14, 3 SST-YFP and 3 WT mice, respectively). l Demo traces of a typical PV-IN response to the same stimulation train before and after MPX-004 treatment. MPX-004 treatment significantly decreased the spike number induced by the pulse train with the same m or different stimulus intensities n (n = 7 from 3 mice). o Demo traces of a typical PYN response to the same stimulation train before and after MPX-004 treatment. MPX-004 treatment significantly decreased the spike number induced by the pulse train with the same p or different stimulus intensities q. Paired t tests m, p and two-way ANOVA followed by Sidak’s multiple-comparisons tests n, q were used to determine significant differences (n = 6 from 3 mice). See Supplemental Table 8 for complete statistical information. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 7. GluN2A Knockdown in prefrontal PV-INs blocks the antidepressant-like effects of MPX-004 and the deficit in sensorimotor gating.

a Schematic diagram illustrating the viral infection and guide cannula implantation procedure in the dmPFC, followed by vehicle or MPX-004 infusion and subsequent behavioral tests. b An image displaying the viral infection (green) and infusion position, confirmed through post hoc verification. c Time spent in the center of the open field (10 min). d Traveled distance (10 min). e, f Time course of traveled distance during the locomotion test (30 min). g, h Bar graphs showing no difference in the startle response measured at 65 dB or 120 dB among the vehicle control and treatment groups. i Bar graph demonstrating a mildly significant increase in prepulse inhibition in the MPX-004 (10 μM) treatment group compared with the vehicle control group. The data were analyzed by two-way ANOVA followed by multiple comparisons using the two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli, *P < 0.05. j Immobility time during the forced swim test, showing no significant change in immobility when the animals were treated with MPX-004. The data were analyzed by unpaired t test (see Supplemental Table 9 for complete statistical information). k In layer V of the dmPFC, PV-INs receive more functional excitatory presynaptic input than do SST-INs, which allows them to exert strong inhibitory control over pyramidal cell firing rates. When GluN2A-containing NMDA receptor function is impaired, such as by blockade with MPX-004 (1), PV-INs become hyperpolarized. This hyperpolarization reduces the activity of PV-INs (2), leading to diminished inhibitory control over PYNs (3) and resulting in the disinhibition of these PYNs (4).