Non-classical NMDAR subunit GluN3A in a specialized hippocampal region regulates stress-coping strategies in male mice

Wei Zhang; Mingyue Zhao; Jiajie Dai; Linan Zhuo; Haimou Ye; Jiesi Wang; Weiwen Wang

doi:10.1038/s42003-026-10457-9

. 2026 Jun 22;9:1309. doi: 10.1038/s42003-026-10457-9

Non-classical NMDAR subunit GluN3A in a specialized hippocampal region regulates stress-coping strategies in male mice

Wei Zhang ^1,^2,³, Mingyue Zhao ^1,², Jiajie Dai ^1,^2,⁴, Linan Zhuo ^1,², Haimou Ye ^1,², Jiesi Wang ^1,^✉, Weiwen Wang ^1,^2,^✉

PMCID: PMC13645741 PMID: 42332071

Abstract

The hippocampus exhibits marked structural, functional, and molecular heterogeneity. GluN3A, a non-classical N-methyl-D-aspartate receptor subunit enriched in hippocampal CA1, has a poorly understood role in stress-related behavior. Here, using chronic social defeat stress (CSDS), GluN3A knockout mice, and behavioral, molecular, genetic, pharmacological, and circuit analyses, we investigated the role of hippocampal GluN3A in stress coping. CSDS selectively reduced GluN3A expression in intermediate CA1 (CA1i), and both stress-induced and genetic GluN3A loss promoted passive coping, whereas CA1i-specific GluN3A overexpression reversed this phenotype. c-Fos staining, fiber photometry, and chemogenetic manipulations showed that GluN3A is required for CA1i pyramidal neuron activation during coping. In addition, local D-serine administration rapidly shifted coping from passive to active in a GluN3A-dependent manner. Circuit tracing showed that CA1i neurons project broadly, and GluN3A overexpression selectively enhanced downstream activity in CA3 and infralimbic cortex. Together, these findings identify GluN3A in CA1i as a key regulator of stress-coping behavior.

Subject terms: Stress and resilience, Limbic system, Molecular neuroscience

Stress selectively lowers GluN3A in hippocampal CA1i, biasing mice toward passive coping, whereas restoring GluN3A or engaging D-serine-sensitive CA1i signaling promotes active coping.

Introduction

The hippocampus is a structurally and functionally heterogeneous structure that plays essential roles in learning, memory, emotional regulation, and adaptive responses to stress^1–5. It is broadly divided into multiple subfields, such as CA1, CA2/3, and the dentate gyrus (DG)⁶. Beyond these classical anatomical partitions, traditional studies have also revealed prominent structural and functional heterogeneity along the longitudinal axis of the hippocampus^3,4,7. For example, the dorsal hippocampus is primarily involved in spatial learning and cognition, whereas the ventral hippocampus is more associated with emotional regulation and affective behaviors^3,4,8. Furthermore, advances in molecular profiling, such as spatial transcriptomics, have revealed an even finer organization, showing that the classical dorsal-ventral dichotomy of subfields CA1, CA3, and DG can be refined into dorsal, intermediate, and ventral portions^4,9. These finer subregions differ in molecular expression, morphology, and projection patterns^9,10. However, how these distinct neuronal populations and molecular profiles, particularly within the intermediate subregions, execute specific behavioral functions remains largely unclear.

Notably, gene expression profiling has shown a concentration of GluN3A expression in the intermediate and ventral regions of CA1; this specific expression pattern identifies it as a key molecular marker distinguishing the intermediate from the dorsal CA1 subregion⁹. This subunit forms non-classical N-methyl-D-aspartate receptors (NMDARs) with GluN1/N2, exhibiting unique biophysical and pharmacological properties distinct from those of conventional GluN1/GluN2 NMDARs^11,12. Unlike canonical GluN2-dependent receptors that require glutamate for activation, GluN1/GluN3A assemblies can be directly activated by sensing ambient extracellular glycine or D-serine, generating excitatory glycine receptor-like currents^13–15. Beyond being glutamate-insensitive, GluN3A-containing receptors also exhibit low Ca²⁺ permeability, significantly reduced Mg²⁺ blockade, and resistance to classical NMDAR antagonists such as D-AP5 and MK-801^13,16–18. Consistent with these unique features, GluN3A-containing NMDARs have been shown to regulate synaptic maturation and plasticity in several brain regions, including the hippocampus, and have been implicated in multiple affective behavioral phenotypes^{16,17,19–22}. Furthermore, growing evidence from preclinical and clinical studies suggests that exogenous D-serine can exert rapid antidepressant-like effects and modulate emotional and cognitive behaviors, underscoring the GluN3A subunit’s potential as a viable target for future clinical translation^23–29. Collectively, these advances suggest that the subregion-specific expression of GluN3A, combined with the unique properties of the receptors it forms, may represent a key molecular mechanism underlying the hippocampus’s broader spatial and functional heterogeneity; however, its specific hippocampal role remains to be identified, especially in stress-related behavioral regulation.

Chronic stress promotes various negative affective states, including passive coping, anhedonia, and anxiety^8,30–32. In the present study, passive coping was operationally defined by immobility-related behavioral measures in the tail suspension test (TST) and forced swim test (FST). Increasing evidence indicates that chronic stress disrupts hippocampal synaptic transmission and intrinsic excitability, contributing to cognitive deficits, emotional dysregulation, and changes in the stress response^2,32–35. Furthermore, hippocampal subregions, such as the distinct segments of the CA1, play differential roles in mediating these adverse stress-induced outcomes. For example, the dorsal CA1, as traditionally delineated, is primarily linked to deficits of spatial cognition, working memory, and social function under chronic stress, whereas the ventral CA1 is more closely associated with affective disturbances and anxiety-related behaviors^36–40. Although the roles of canonical GluN2-containing NMDARs in these processes have been extensively studied^41–43, very little is known regarding the function of non-classical GluN3A-containing NMDARs, which exhibit a distinct spatial expression pattern within the hippocampus, under chronic stress conditions. A recent study has shown that GluN3A deletion in the traditional ventral CA1 region, which encompasses both the CA1 intermediate (CA1i) and CA1 ventral (CA1v) areas, induces anxiety-like behavior and that these excitatory glycine receptors (eGlyR) contribute to corticosterone-dependent modulation of neuronal plasticity, including long-term potentiation¹⁵. These findings hint that GluN3A may influence stress-related behavioral outcomes and contribute to the functional heterogeneity of CA1. However, GluN3A serves as a molecular marker distinguishing the CA1 dorsal (CA1d) and CA1i subregions⁹, it remains unclear which specific hippocampal CA1 loci and behavioral domains GluN3A controls, and how it functionally contributes to stress regulation, particularly under chronic stress exposure.

To address these critical questions, we employed chronic social defeat stress (CSDS) and GluN3A knockout (KO) mouse models, combining comprehensive approaches encompassing behavior, molecular biology, genetic manipulation, microscopic imaging, fiber photometry, pharmacology, and single-cell transcriptomic analysis. This integrative strategy allowed us to delineate the specific contribution of hippocampal GluN3A to passive coping behavior, clarify its role relative to anhedonia, social interaction, and anxiety, and identify the neural mechanisms and specific loci underlying this regulation. Our findings provide mechanistic insight into how non-classical NMDARs contribute to stress adaptation, revealing a previously unrecognized role of hippocampal GluN3A in shaping coping strategies under adverse conditions.

Results

Effects of CSDS and GluN3A KO on stress-related behaviors and GluN3A expression in the hippocampus

We systematically evaluated various stress-related behaviors induced by CSDS in mice (Fig. 1a, b). Regarding the anhedonia indicator, CSDS mice showed a significant reduction in social interaction ratio (t₁₆ = 5.802, ^***P <0.001, Fig. 1c, d) in the social interaction test (SIT) but no significant change in sucrose preference ratio (P > 0.05, Supplementary Fig. 1a) in the sucrose preference test (SPT) compared to the control group. In terms of coping behaviors in response to inescapable stress, CSDS significantly increased the immobile time of mice in the TST (t₁₄ = 2.877, ^*P = 0.012, Fig. 1e) and FST (t₁₁ = 2.240, ^*P = 0.047, Fig. 1f). In locomotor activity, compared to the control group, CSDS did not affect the total distance traveled by mice (P > 0.05, Fig. 1g, h). Regarding anxiety-related indicators, CSDS markedly induced anxiety-like behavior in mice, evidenced by a decrease in the number of entries of the central area (t₁₃ = 2.801, P = 0.015, Fig. 1i) and the time spent in the central area (Welch-corrected t_9.157 = 3.290, ^*P = 0.009) in the open field test (OFT) (Fig. 1g, j). Additionally, in the elevated plus maze (EPM), CSDS mice exhibited a reduction in the number of open arm entries (Welch-corrected t_11.43 = 2.232, ^*P = 0.047, Fig. 1k, l), and the time in open arms showed a non-significant trend toward reduction (Welch-corrected t_9.299 = 2.174, P = 0.057, Fig. 1k, m). These results indicated that CSDS successfully induced various negative affective states in mice.

Subsequently, we examined the changes in GluN3A expression levels in various emotion-related brain regions, particularly the hippocampus, in mice following CSDS (Supplementary Fig. 1b, c). The quantitative real-time PCR (qRT-PCR) results showed a significant reduction in GluN3A expression in the CA1i region of the hippocampus in CSDS mice compared to the control group (t₂₃ = 2.692, ^*P = 0.013, Fig. 1n), while there were no significant changes in the other subregions of the hippocampus, including CA1d, CA1v, CA3, and DG (P > 0.05, Fig. 1n). The western blot results similarly demonstrated a significant decrease of GluN3A protein expression in the CA1i in CSDS mice (t₉ = 2.866, ^*P = 0.024), with no significant changes observed in CA1v and CA3 (P > 0.05, Fig. 1o, p, Supplementary Fig. 1d). These data indicated that CSDS downregulated GluN3A expression in the hippocampal CA1i region in mice.

To further examine the relationship between GluN3A and emotion-related behaviors, we conducted behavioral tests on stress-naïve GluN3A KO mice under physiological conditions (Fig. 1q). On the anhedonia indicators, GluN3A KO mice did not exhibit an abnormal ratio of social interactions (P > 0.05, Fig. 1r) and sucrose preference (P > 0.05, Supplementary Fig. 2a) compared to wild-type (WT) mice. In contrast, the results showed that GluN3A KO mice exhibited a passive coping phenotype compared to WT mice, as evidenced by significantly increased immobile time in TST (t₁₆ = 2.778, ^*P = 0.013, Fig. 1s) and FST (t₁₄ = 2.457, ^*P = 0.028, Fig. 1t). In addition, GluN3A KO did not influence the locomotor activity (P > 0.05, Fig. 1u) and anxiety-like behavior (P > 0.05, Fig. 1v-y).

To ascertain whether GluN3A KO is associated with an increased susceptibility to social defeat stress (SDS), a systematic assessment was conducted to evaluate the effects of subthreshold SDS on emotion-related behaviors in GluN3A KO mice. The results showed GluN3A KO mice subjected to subthreshold SDS exhibited a similar significant increase in immobile time in the FST (Welch-corrected t_2.520 = 9.677, ^*P = 0.031, Supplementary Fig. 3d). Still, no substantial changes observed in social interest, sucrose preference, locomotor activity, and anxiety-like behavior (P > 0.05, Supplementary Fig. 3b, c, e, g–i), except a significant reduction in the number of entries of central area in the OFT (t₁₃ = 2.815, ^*P = 0.015, Supplementary Fig. 3f). The data indicated that the GluN3A KO does not influence the susceptibility of mice to SDS.

Taken together, these results suggested that the decreased expression of GluN3A in the brain was involved in the manifestation of passive coping behaviors, and that the hippocampal CA1i may be a potential brain region for the regulation of stress-coping behaviors by GluN3A.

Overexpression of the GluN3A decline in CA1i neurons reversed passive coping behaviors

We further investigated the regulatory role of CA1i GluN3A in emotion-related behaviors by using Adeno-associated virus (AAV) to overexpress recombinant GluN3A in the CA1i of CSDS mice (Fig. 2a, b). The spatial distribution of viral expression across animals is summarized in Supplementary Fig. 4. We found that overexpression of exogenous GluN3A did not significantly affect anhedonia behaviors in CSDS mice (SIT, Fig. 2c; SPT, Supplementary Fig. 5a). In contrast, CA1i-targeted GluN3A overexpression normalized the CSDS-elevated immobility in the TST and reduced FST immobility more broadly in both control and stress mice. Specifically, two-way ANOVA for TST showed a significant interaction effect (F _{(1, 28)} = 5.568, ^*P = 0.026, Fig. 2d), with post-hoc analysis revealing a significant increase in immobile time in the stress: AAV-ZsGreen group compared to the control: AAV-ZsGreen group (False discovery rate-corrected ^*P = 0.024), while the stress: AAV-Grin3A group showed a significant decrease in immobile time compared to the stress: AAV-ZsGreen group (FDR-corrected ^*P = 0.016). Two-way ANOVA for FST showed significant main effects of stress (F _{(1, 28)} = 5.010, ^*P = 0.033, Fig. 2e) and viral treatment (i.e., AAV-ZsGreen or AAV-Grin3A) (F _{(1, 28)} = 12.99, ^**P = 0.0012) on the immobile time, with post-hoc analysis showing a significant increase in immobile time in the stress: AAV-ZsGreen group compared to the control: AAV-ZsGreen group (FDR-corrected ^*P = 0.026) and a significant decrease in immobile time in both the control: AAV-Grin3A group (FDR-corrected ^*P = 0.036) and stress: AAV-Grin3A group (FDR-corrected ^*P = 0.016) compared to their respective AAV-ZsGreen groups. Furthermore, consistent with the previous experiments, CSDS induced the changes in locomotor indicators and multiple anxiety indicators in OFT and EPM (Fig. 2f–j), but these were not affected by the overexpression of the exogenous GluN3A subunit. These results indicated that CA1i GluN3A modulates passive coping behavior across conditions, contributing to stress-induced passive coping while also influencing baseline coping performance.

Fig. 2 — a Experimental scheme of CSDS modeling, virus injection, and behavior tests. b Schematic target locations and immunofluorescence images of ZsGreen or flag expression of Virus in the CA1i. Scale bar, 1 mm. c Social interaction ratio of SIT (n = 7, 8, 7 and 7 in control:AAV-ZsGreen, control:AAV-Grin3A, stress:AAV-ZsGreen and stress:AAV-Grin3A, respectively). d Immobility time in the TST (n = 8 per group). e Immobility time in the FST (n = 8, 8, 7 and 9 in control:AAV-ZsGreen, control:AAV-Grin3A, stress:AAV-ZsGreen and stress:AAV-Grin3A, respectively). f The total distance traveled in OFT (n = 7, 8, 8 and 9 in control:AAV-ZsGreen, control:AAV-Grin3A, stress:AAV-ZsGreen and stress:AAV-Grin3A, respectively). g The number of entries to centre area of OFT. h The time in the centre area of OFT. i The number of entries to open arms of EPM (n = 8 per group). j The time spent in the open arms of EPM. k Experimental scheme of virus injection and behavior tests in stress-naïve GluN3A KO and WT mice. l Immobility time in the TST (n = 8, 6 and 6 in WT + AAV-ZsGreen, GluN3A KO + AAV-ZsGreen and GluN3A KO + AAV-Grin3A, respectively). m Immobility time in the FST (n = 7, 6 and 6 in WT + AAV-ZsGreen, GluN3A KO + AAV-ZsGreen and GluN3A KO + AAV-Grin3A, respectively). n The total distance traveled in OFT (n = 8, 6 and 7 in WT + AAV-ZsGreen, GluN3A KO + AAV-ZsGreen and GluN3A KO + AAV-Grin3A, respectively). * P < 0.05, ** P < 0.01.

To further confirm the causal relationship between CA1i GluN3A and stress-coping behaviors, we overexpress recombinant GluN3A in the CA1i region of stress-naïve GluN3A KO mice using AAV (Fig. 2k). Consistent with the findings above, GluN3A overexpression normalized passive coping behaviors without affecting social interaction, locomotor activity, or anxiety-like behavior (P > 0.05, Fig. 2l–n and Supplementary Fig. 6a–c). Specifically, GluN3A KO mice showed a significant increase in immobile time in TST (t₁₂ = 2.212, ^*P = 0.047, Fig. 2l) and FST (t₁₁ = 2.309, ^*P = 0.041, Fig. 2m) compared to WT mice, while GluN3A KO: AAV-Grin3A mice exhibited a significant restoration of immobile time to normal levels in both TST (t₁₀ = 2.696, ^*P = 0.023) and FST (t₁₀ = 2.912, ^*P = 0.016) compared to GluN3A KO mice. In summary, hippocampal CA1i GluN3A played a crucial role in regulating coping strategies.

Activity of CA1i pyramidal neurons mediated by GluN3A-dependent regulation of stress-coping behaviors

Having established that GluN3A in the hippocampal CA1i region regulates passive coping behavior in CSDS mice, we next sought to identify the neuronal types that might contribute to this effect. Through reanalysis of single-nuclei RNAseq (snRNAseq) data from mice subjected to repeated social defeat (RSD) in a previously published study (Fig. 3a)⁴⁴, we found that GluN3A is abundantly expressed in excitatory neurons in the hippocampal CA1 region, and its expression pattern changes after RSD (Fig. 3b, c). Moreover, compared to the Control: WT group, GluN3A expression significantly decreased in excitatory neurons of WT subjected to RSD (^*P = 0.037, Fig. 3d, e).

Fig. 3 — a Experimental scheme of repeated social defeat (RSD) modeling and snRNA-seq in control and stress mice: Mice were subjected to RSD or left undisturbed (control group). 14 h after the final RSD session, hippocampi were dissected, pooled from 3 mice per group, and processed for nuclei isolation followed by snRNA-seq to define nuclear transcriptomic profiles. The figure was created in BioRender (https://BioRender.com/6d5pops). b Identification of the various cell types in the hippocampus using UMAP clustering. c Visualization of the distribution of *Grin3A* expression in excitatory neurons in CA1 for each condition: Control WT, RSD WT. d Dot plot shows the expression of *Grin3A* in excitatory neurons of each group. The size of the circle represents the percentage of neurons expressing the gene. The color of the circle represents the average level of gene expression in neurons. e Violin plots show the expression level of *Grin3A* in excitatory neurons of each group. f Statistical region of c-Fos-positive cells in CA1i (ZsGreen: green, c-Fos: red, Dapi: blue). Scale bar, 1 mm. g Representative images of c-Fos-positive cells in the CA1i of each group of mice (c-Fos: red, Dapi: blue). Scale bar, 100 μm. h Quantification of c-Fos-positive cells in CA1i pyramidal layer neurons of each group of mice. (n = 6, 6, 5 and 7 in control:AAV-ZsGreen, control:AAV-Grin3A, stress:AAV-ZsGreen and stress:AAV-Grin3A, respectively, 4 slices for each dot). i Quantification of c-Fos-positive cells in CA1i pyramidal layer neurons of each group of stress-naïve mice. (n = 5, 6 and 6 in WT+ No-FST, WT + FST and GluN3A KO + FST, respectively, 4 slices for each dot). j Representative images of c-Fos-positive cells in the CA1i of each group of stress-naïve mice (c-Fos: red, Dapi: blue). Scale bar, 100 μm. k Experimental scheme for fiber photometry in stress-naïve GluN3A KO and WT mice. Schematic illustrating fiber placement and representative images of GCaMP8s expression. Scale bar, 100 μm. The figure was created in BioRender (https://BioRender.com/6d5pops). l Examples of fluorescence tracing and fluorescence intensity heatmaps of GCaMP signals in WT and GluN3A KO mice during TST. Gray shaded areas indicate struggling behavior; white areas indicate immobile behavior. m Mean heatmaps of GCaMP8s transient ΔF/F signals aligned to struggling onset. Each row represents one mouse, and a total of 11 mice are shown. n Dynamics of GCamp8s signaling in mice 5 seconds before and 10 seconds after struggling in the TST (mean ΔF/F ± s.e.m.). o Changes in peak ΔF/F of WT and KO mice during the struggle (n = 5 in WT, n = 6 in KO). * P < 0.05, ** P < 0.01.

Therefore, we further investigated the functional activity of CA1i pyramidal neurons during stress-coping behaviors in the CSDS model under conditions of GluN3A modulation. The results of c-Fos immunostaining showed reduced neuron activity of the pyramidal layer in the CA1i region after the FST in stressed mice compared to controls (Fig. 3f–h), which was restored by overexpressing recombinant GluN3A. Specifically, two-way ANOVA analysis of the number of c-Fos-positive cells showed significant interaction (F _{(1, 20)} = 4.820, ^*P = 0.040, Fig. 3h) and stress main effects (F _{(1, 20)} = 6.015, ^*P = 0.024), with a marginally significant effect for viral treatment (F _{(1, 20)} = 3.679, P = 0.070). Post-hoc analysis revealed a significant decrease in the number of c-Fos-positive cells in the stress: AAV-ZsGreen group compared to the control: AAV-ZsGreen group (FDR-corrected ^**P = 0.009), while the stress: AAV-Grin3A group showed a significant increase in the number of c-Fos-positive cells compared to the stress: AAV-ZsGreen group (FDR-corrected ^**P = 0.009).

Furthermore, we examined the direct relationship between the GluN3A subunit and the functional activity of CA1i pyramidal neurons after the FST in stress-naïve GluN3A KO mice. First, immunostaining results showed no significant difference in the number of CA1i c-Fos-positive cells between WT and GluN3A KO mice at baseline (no FST experience, P > 0.05, Supplementary Fig. 7a–c). Notably, the number of c-Fos-positive cells was significantly increased after FST compared with the basal condition in WT mice (t₉ = 2.884, ^*P = 0.018), and such activation did not occur in FST-experienced GluN3A KO mice (t₁₀ = 3.232, ^**P = 0.009, Fig. 3i, j), indicating that the activation of pyramidal neurons in the CA1i region was involved in FST behavior, while the absence of GluN3A resulted in this kind of functional defect.

Moreover, we validated the necessity of GluN3A in the activation of excitatory neurons in the CA1i during the TST using fiber photometry (Fig. 3k). During TST, the excitatory-specific GCaMP signal, which characterizes calcium levels, showed a significant correlation with the active state of the stress-naïve WT mice. The increase in signal lagged behind the transition from immobility to struggling (Fig. 3l). This correlation was weaker in stress-naïve GluN3A KO mice. Subsequently, we observed that the average peak amplitude of neuronal responses significantly increased during events when WT mice transitioned from immobility to struggling (Fig. 3m, n), with GluN3A KO mice showing significantly lower struggle-related neuronal activation compared to WT mice (t₉ = 3.398, ^**P = 0.008, Fig. 3o).

Collectively, these results indicated that the activity of CA1i pyramidal neurons mediated the GluN3A-dependent regulation of stress-coping behavior in mice.

Bidirectional chemogenetic modulation of CA1i pyramidal neurons directly mimicked or reversed the passive coping behaviors

To directly test the causal role of CA1i pyramidal neuron activity in stress-coping behavior, we employed bidirectional chemogenetic manipulation. We first increased the activity of CA1i neurons using AAV-hM3D(Gq) in combination with different concentrations of clozapine-N-oxide (CNO) (Fig. 4a, Supplementary Fig. 8a). The results showed that saline, 0.05 mg/kg, 0.2 mg/kg, and 0.5 mg/kg CNO did not affect the locomotor activity and anxiety-like behavior of stress-naïve mice in the OFT and rotarod test (RT) (P > 0.05, Supplementary Fig. 8b–e). However, one-way ANOVA showed that concentrations of 0.2 mg/kg (FDR-corrected ^**P = 0.009, Fig. 4b, c, Supplementary Fig. 8f) and 0.5 mg/kg (FDR-corrected ^**P = 0.009) CNO effectively increased the number of c-Fos-positive cells at the injection site (F _{(2, 3)} = 16.89, ^*P = 0.023). Therefore, we chose a concentration of 0.2 mg/kg CNO for subsequent chemogenetic activation experiments (Fig. 4d). The results indicated that the activation of CA1i pyramidal neurons reversed passive coping behavior induced by CSDS in the TST and reduced passive coping behavior in both control and CSDS mice in the FST, but did not significantly alter social interaction in the SIT, locomotor activity, or anxiety-like behavior in the OFT (Fig. 4e–j). Specifically, two-way ANOVA for the social interaction ratio showed a significant main effect of viral treatment (F _{(1, 20)} = 6.983, ^*P = 0.016, Fig. 4e), whereas the post hoc comparisons were not significant (P > 0.05). Two-way ANOVA for immobile time in the TST showed a significant interaction effect (F _{(1, 22)} = 6.923, ^*P = 0.015, Fig. 4f), with post-hoc analysis showing a significant increase in immobile time in the stress: AAV-mCherry group compared to the control: AAV-mCherry group (FDR-corrected ^*P = 0.011), while the stress: AAV-hM3D(Gq) group exhibited a significant decrease in immobile time to normal levels compared to the stress: AAV-mCherry group (FDR-corrected ^*P = 0.011). Two-way ANOVA for immobile time in the FST showed significant stress (F _{(1, 21)} = 5.350, ^*P = 0.031, Fig. 4g) and viral treatment main effects (F _{(1, 21)} = 12.34, ^**P = 0.002), with post-hoc analysis showing a significant increase in immobile time in the stress: AAV-mCherry group compared to the control: AAV-mCherry group (FDR-corrected ^*P = 0.023), and a significant decrease in immobile time in both the control: AAV-hM3D(Gq) group (FDR-corrected ^*P = 0.026) and stress: AAV-hM3D(Gq) group (FDR-corrected ^*P = 0.009) compared to their respective AAV-mCherry groups. These results indicated that increased excitatory activity in CA1i neurons is sufficient to bias coping from passive toward active responses, with the FST being more sensitive to baseline modulation.

On the other hand, we inhibited the activity of CA1i pyramidal neurons in stress-naïve mice using AAV-hM4D(Gi) (Fig. 4k, l). We found that reducing the activity of the neurons significantly increased immobile time in the TST (t₁₂ = 3.261, ^**P = 0.007, Fig. 4m) and FST (t₁₃ = 2.622, ^*P = 0.021, Fig. 4n), but did not induce abnormalities in locomotor activity and anxiety-like behavior in the OFT (P > 0.05, Fig. 4o–q), indicating that the inhibition of neuronal activity in CA1i can directly induce passive coping phenotype.

Overall, our findings suggest that the activity of CA1i pyramidal neurons may underlie GluN3A’s regulation of stress-coping behavior.

Potentiation of GluN3A-mediated CA1i pyramidal neuron activity was crucial for the active coping strategies induced by D-serine

We further explored whether GluN3A participates in the rapid D-serine–induced shift of stress-coping strategies within the hippocampus. By injecting D-serine into the bilateral CA1i region of CSDS mice and stress-naïve GluN3A KO mice before behavioral testing (Fig. 5a, b), we found that D-serine could significantly reverse the passive coping behavior induced by CSDS (Fig. 5c, d), while it had no significant effect on locomotor activity and anxiety-like behavior in the OFT in all groups (P > 0.05, Fig. 5e–g). Specifically, two-way ANOVA for immobile time in the TST and FST showed significant interaction (Control-cerebrospinal fluid (CSF), Control-D-serine, Stress-CSF and Stress-D-serine group, TST: F _{(1, 27)} = 6.084, ^*P = 0.020; FST: F _{(1, 27)} = 5.106, ^*P = 0.032, Fig. 5c, d) and drug treatment main effects (TST: F _{(1, 27)} = 11.60, ^**P = 0.002; FST: F _{(1, 27)} = 11.19, ^**P = 0.002), with post-hoc analysis showing a significant increase in immobile time in the stress: CSF group compared to the control: CSF group in both tests (TST: FDR-corrected ^*P = 0.017; FST: FDR-corrected ^*P = 0.032), while the stress: D-serine group exhibited a significant decrease in immobile time compared to the stress: CSF group (TST: FDR-corrected ^***P <0.001; FST: FDR-corrected ^**P = 0.002). Interestingly, compared to the control (WT):D-serine group, the GluN3A KO:D-serine group still exhibited significantly increased immobile time in both the TST (t₁₃ = 2.470, ⁺P = 0.028) and FST (t₁₃ = 3.225, ⁺⁺P = 0.007), indicating that GluN3A is required for D-serine–induced behavioral improvement. Additionally, simultaneously inhibiting the activity of CA1i neurons in CSDS mice while administering D-serine, we found that the stress-hM4D(Gi) group showed a significant increase in immobile time in both the TST (t₁₂ = 4.628, ^###P <0.001) and FST (t₁₂ = 2.671, ^#P = 0.020) compared to the Stress-mCherry group, suggesting that suppressing CA1i pyramidal neuronal activity blocked the active coping effect of D-serine.

Furthermore, we performed immunostaining for c-Fos in the CA1i pyramidal layer neurons of mice from the groups above after the FST (Fig. 5h). The results showed that D-serine significantly increased the number of c-Fos-positive cells in the CA1i pyramidal layer (Fig. 5i, j). Specifically, two-way ANOVA for the number of c-Fos-positive cells showed significant stress (Control-CSF, Control-D-serine, Stress-CSF, and Stress-D-serine group, F _{(1, 16)} = 11.23, ^**P = 0.004, Fig. 5j) and drug treatment effects (F _{(1, 16)} = 13.78, ^**P = 0.002). Post-hoc analysis showed a significant decrease in the number of c-Fos-positive cells in the stress: CSF group compared to the control: CSF group (FDR-corrected ^*P = 0.037), while both the control: D-serine group (FDR-corrected ^*P = 0.016) and stress: D-serine group (FDR-corrected ^*P = 0.045) showed a significant increase in the number of c-Fos-positive cells compared to their respective CSF groups. However, in GluN3A KO:D-serine mice, the number of c-Fos-positive cells in the CA1i pyramidal layer did not rise to the level of the control (WT):D-serine group (t₈ = 2.967, ⁺P = 0.018). In summary, D-serine facilitates a rapid shift from passive to active coping behavior by potentiating GluN3A-dependent CA1i pyramidal neuron activity, highlighting GluN3A signaling in this hippocampal subregion as a crucial modulator of adaptive stress responses.

Projections of CA1i neurons and downstream circuits may be involved in GluN3A-mediated regulation of stress-coping behaviors

To identify the downstream targets of hippocampal CA1i neurons potentially involved in GluN3A-mediated behavioral regulation, we first performed projection-mapping analyses using the Allen Brain Atlas database and AAV-mediated tracing experiments. According to the Allen Brain connectivity data, CA1i neurons send prominent projections to several sub-hippocampal and limbic related brain regions, including the CA3, nucleus accumbens (NAc), infralimbic cortex (IL), BLA, and lateral entorhinal cortex (LEC) (Fig. 6a). Consistent with these database results, injection of AAV2/9-syn-ZsGreen into the CA1i region resulted in robust fluorescent projection signals in these downstream areas, particularly in the CA3, NAc core (AcbC) and shell (AcbSh), IL, BLA, and LEC (Fig. 6b), confirming that CA1i pyramidal neurons form extensive efferent connections with emotion-related circuits.

Fig. 6 — a Projection map of the hippocampal CA1i region generated from the Allen Brain Atlas database. b Representative images of virus tracing results following AAV2/9-syn-ZsGreen injection into the CA1i region. Scale bar, 1000 μm. c Statistical region of c-Fos-positive cells in CA3 (ZsGreen: green, c-Fos: red, DAPI: blue). Scale bar, 500 μm. d Quantification of c-Fos-positive cells in CA3 neurons of each group of mice (n = 6, 7, 5 and 7 in control:AAV-ZsGreen, control:AAV-Grin3A, stress:AAV-ZsGreen and stress:AAV-Grin3A, respectively, 4 slices for each dot). e Representative images of c-Fos-positive cells in the CA3 of each group of mice (c-Fos: red, Dapi: blue). Scale bar, 500 μm. f Statistical region of c-Fos-positive cells in IL (ZsGreen: green, c-Fos: red, DAPI: blue). Scale bar, 500 μm. g Quantification of c-Fos-positive cells in IL neurons of each group of mice (n = 4, 5, 5 and 5 in control:AAV-ZsGreen, control:AAV-Grin3A, stress:AAV-ZsGreen and stress:AAV-Grin3A, respectively, 4 slices for each dot). h Representative images of c-Fos-positive cells in the IL of each group of mice (c-Fos: red, Dapi: blue). Scale bar, 50 μm. * P < 0.05, ** P < 0.01.

Next, we examined whether overexpression of the GluN3A subunit in CA1i neurons could influence neuronal activation in these projection targets following chronic stress. c-Fos immunostaining was performed in representative downstream regions after the FST in control and CSDS mice receiving either AAV-ZsGreen or AAV-Grin3A injections in the CA1i region (Fig. 6c–h, Supplementary Fig. 9a–d). In the CA3 region, a two-way ANOVA revealed a significant stress × viral treatment interaction (F _{(1, 21)} = 5.641, ^*P = 0.027, Fig. 6d) and a significant stress main effect (F _{(1, 21)} = 5.816, ^*P = 0.025). Post hoc analysis showed that in the AAV-ZsGreen group, CSDS significantly decreased the number of c-Fos–positive neurons in the CA3 region compared with controls (FDR-corrected ^**P = 0.009). In contrast, overexpression of GluN3A in CA1i neurons reversed this reduction, restoring c-Fos expression to control levels (FDR-corrected ^*P = 0.017). In the IL, both stress (F _{(1, 15)} = 5.368, ^*P = 0.035, Fig. 6g) and viral treatment (F _{(1, 15)} = 5.182, ^*P = 0.038) exerted significant main effects. Post hoc analysis indicated that in the AAV-ZsGreen group, CSDS significantly reduced c-Fos–positive cell counts in the IL (FDR-corrected ^*P = 0.049), whereas GluN3A overexpression markedly increased c-Fos expression levels relative to the stress: AAV-ZsGreen group (FDR-corrected ^*P = 0.049). In contrast, in both the AcbSh and AcbC regions, there were no significant effects observed (P > 0.05, Supplementary Fig. 9b, c). These findings suggest that GluN3A modulation in the CA1i primarily affects activity in specific CA3 and IL targets, rather than in the NAC.

Together, these results indicate that CA1i pyramidal neurons establish strong anatomical and functional connections with emotion-related structures, and that GluN3A-dependent regulation of stress-coping behaviors may occur through selective modulation of CA1i-CA3 or CA1i-IL circuit activity rather than global changes across all projection targets.

Discussion

In the present study, we systematically investigated a previously unrecognized role of the GluN3A-containing non-classical NMDARs within the hippocampal CA1i region in regulating stress-coping strategies. Our results demonstrated that chronic stress induced a specific decrease in GluN3A expression in the hippocampal CA1i region, and that CA1i GluN3A explicitly regulates the passive coping behaviors in the TST and FST. Mechanistically, the effects of GluN3A on stress-coping may be mediated through the activation of CA1i pyramidal neurons. Moreover, we found that D-serine administration facilitated a rapid shift from passive to active coping by activating GluN3A-dependent CA1i neurons. Finally, circuit-level analyses revealed that the selective, GluN3A-mediated activation of CA1i-CA3 and CA1i-IL pathways, but not CA1i-NAc, may serve as a potential circuit substrate for stress adaptation. Collectively, these findings highlight a critical role for GluN3A in response to chronic stress, and its ability to shape adaptive coping through hippocampal CA1i pyramidal neurons and specific projection circuits, and reveal its contribution to the rapid shift in coping strategy induced by D-serine.

First, by focusing on the non-classical NMDAR subunit GluN3A and its distinctive expression pattern in refined hippocampal subregions, we found that mice exhibited a selective reduction in GluN3A expression in the CA1i region, but not in the traditional, more broadly dorsal and ventral subareas, after CSDS. This finding provides functional significance for the Hippocampal Gene Expression Atlas (HGEA) constructed by Bienkowski et al., which divides the CA1 into four subregions: CA1d, CA1i, CA1v, and the ventral tip (CA1vv), and identifies GluN3A as a key molecular marker distinguishing CA1i from CA1d⁹. Moreover, our causal manipulations (both genetic knockout and selective CA1i overexpression in CSDS or KO animals) confirmed that CA1i GluN3A bidirectionally regulates passive coping behaviors in the TST and FST, without affecting social interest or anxiety-like behaviors. This demonstrates a regionally and behaviorally specific regulatory role for CA1i GluN3A-expressing neurons in stress coping. This behavioral specificity for coping refines previous reports. For instance, while one study linked global hippocampal GluN3A influences to both helplessness and anxiety⁴⁵, another suggested that GluN3A in the traditional ventral hippocampus modulates anxiety-like behaviors and is essential for neurons’ response to corticosterone¹⁵. Our results reveal a more precise function, identifying GluN3A as a key molecular determinant of stress response and coping flexibility, and pinpointing CA1i as a critical functional locus linking molecular heterogeneity to behavioral specificity during stress adaptation.

Coping behavior represents a core dimension of the stress response, reflecting the individual’s ability to adjust strategies dynamically when facing adversity^46,47. Active coping is associated with resilience and positive affect, whereas passive coping (helplessness) is linked to depression vulnerability^48–51. Clinically, impairments in coping flexibility are a hallmark of major depressive disorder and other stress-related psychopathologies^52–54. In human studies, genetic variations in the GRIN3A locus have been implicated in bipolar disorder and schizophrenia^55,56, lending clinical relevance to our findings and further suggesting that the CA1i GluN3A-mediated pathway we identified may contribute to the pathophysiology of related mental illnesses. Thus, identifying CA1i GluN3A as a regulator of coping strategies may provide novel therapeutic entry points for disorders characterized by maladaptive stress responses.

Our study further elucidates the neuronal mechanism by which GluN3A modulates stress-coping strategies, suggesting that modulation of CA1i neuronal excitability, primarily in pyramidal neurons, is the key mediator. Several convergent lines of evidence support this conclusion. First, snRNAseq data revealed a significant downregulation of GluN3A expression in CA1 excitatory neurons of CSDS-exposed mice⁴⁴, a finding consistent with the identified neuronal hypoactivity observed in CSDS and GluN3A KO mice in the current study. Second, both stress exposure and GluN3A KO decreased CA1i c-Fos expression after coping behaviors, and KO animals further showed attenuated struggling-related calcium transients in pyramidal neurons during TST. Third, overexpression of CA1i GluN3A restored c-fos-represented neuronal activity and modified passive coping behavior. Critically, Chemogenetic manipulation further established a causal link between CA1i neuron excitability and coping behavior: activation promoted active coping, while inhibition induced passive coping. Although we cannot exclude the potential contribution of GluN3A-expressing inhibitory neurons, our results suggest that pyramidal neurons are the main effector population translating GluN3A signaling into behavioral outcomes. Furthermore, across manipulations, we observed a consistent pattern of behavioral detection outcomes: CA1i GluN3A overexpression and CA1i activation produced a stress-dependent rescue effect in the TST, whereas both manipulations reduced FST immobility more broadly, including in stress-naïve mice. One possible explanation is a difference in behavioral dynamic range between the two assays. In our paradigm, the baseline immobility level of the TST was relatively low, potentially approaching the floor effect. In contrast, the FST exhibited a higher baseline immobility level, thereby offering greater dynamic range and thus higher sensitivity to operationalized interventions. Moreover, previous work has demonstrated physiological and pharmacological distinctions between the TST and FST^47,57,58, including differential sensitivity to NMDAR-related manipulations⁵⁹, which may further account for the separation pattern observed in our data. Notably, we observed a dissociation between CA1i c-Fos levels and FST immobility in control mice. This may be because c-Fos is a thresholded, time-integrated marker with limited sensitivity to subtle but behaviorally relevant activity changes. In control mice, CA1i c-Fos levels were already relatively high, making further increases difficult to detect, whereas in stressed mice with lower baseline c-Fos, the effect of GluN3A manipulation was more readily observed.

One plausible mechanism underlying this regulatory role in CA1i neuron activity may involve the unique properties of GluN3A-containing NMDARs, which are inherently glutamate-insensitive but responsive to extracellular glycine or D-serine, enabling CA1i pyramidal neurons to sense glial-derived glycine or D-serine released under stress conditions^13,15,16. Chronic stress-induced downregulation of GluN3A may disrupt this alternative excitatory pathway, thereby dampening pyramidal neuron excitability and biasing behavioral output toward passive coping. Recent electrophysiological evidence supports the core premise of this hypothesis, which indicates that excitatory glycine receptors are present on ventral hippocampal pyramidal neurons and mediate distinct synaptic dynamics¹⁵. Furthermore, the proposed link between GluN3A and neuronal excitability is further strengthened by findings in other brain regions, such as the olfactory bulb, where GluN3A KO also reduces c-Fos expression and CaMKII phosphorylation⁶⁰. Finally, the causal link between CA1 pyramidal neuron hypoactivity and passive coping is consistent with previous studies, as inhibition of these neurons has been shown to promote depression-like behaviors, including stress-coping behaviors represented by the FST and TST^34,35. This convergence of evidence strengthens the hypothesis that this non-classical GluN3A-containing NMDAR pathway modulates CA1i pyramidal neuron excitability.

Our findings on D-serine further underscore the translational significance of GluN3A signaling. As a high-affinity ligand for the glycine-binding site of GluN3 subunits^13,61, D-serine and its derivatives have been reported to exert rapid antidepressant effects in both preclinical and clinical studies, with the hippocampus being a critical site of action^23,28,62. Given that passive coping is a behavioral endophenotype targeted by antidepressant treatment^63–65, we explored the mechanistic link between D-serine, GluN3A, and stress-coping behaviors. We found that D-serine microinjection into the CA1i rapidly shifted coping behavior from passive to active, accompanied by increased CA1i pyramidal neuron activity. Remarkably, this effect was abolished in GluN3A KO mice or when CA1i neuron activity was suppressed, indicating that GluN3A-dependent excitation of CA1i pyramidal neurons is essential for D-serine’s behavioral efficacy. Notably, D-serine normalized stress-elevated immobility but did not further reduce baseline FST immobility, despite the effects of CA1i GluN3A overexpression and CA1i activation on baseline FST. This may indicate that D-serine primarily compensates for stress-induced reductions in CA1i GluN3A expression and associated impairment of excitatory function in GluN3A-containing NMDAR neurons. Thus, its behavioral effect may reflect restoration of a disrupted excitatory state rather than supraphysiological enhancement under baseline conditions. Although D-serine increased CA1i c-Fos expression in stress-naïve mice, this effect may reflect the activation of neurons expressing conventional GluN1/GluN2-containing NMDARs, since D-serine acts as an agonist at the GluN1 subunit and can synergize with glutamate independent of GluN3A⁶⁶. These neuronal populations may not directly participate in coping behavior. Given that D-serine itself has also been reported to have extensive links, including endogenous changes and exogenous administration, to symptom or cognitive improvements in clinical and pre-clinical settings^67–70, our findings suggest that the GluN3A-dependent pathway in CA1i is not only critical for adaptive coping but may also offer a promising and mechanistically targetable path for the development of therapeutic interventions or cognitive enhancements in related conditions.

Finally, our initial exploration of CA1i projection analysis points toward potential circuit-level mechanisms to explain the significant role of GluN3A. We demonstrated that CA1i neurons project extensively to sub-hippocampal and limbic-related regions, including CA3, IL, and NAc. Furthermore, our results showed that GluN3A overexpression selectively enhanced downstream neuronal activation in CA3 and IL, but not NAc. These projection circuits, although not specific to CA1i, have previously been reported to be associated with stress-coping strategies of mice^30,71,72. This alignment with previous reports strongly suggests that CA1i-CA3 and CA1i-IL projections may be the functional circuits underlying GluN3A-dependent regulation of stress coping by CA1i neurons.

An important limitation of the present study is that we did not directly determine how endogenous neurotransmission engages GluN3A-containing receptors in CA1i neurons during stress-related behaviors. Although our gain- and loss-of-function experiments, together with the fiber photometry and D-serine results, support a causal contribution of GluN3A to stress-coping-related phenotypes, the precise physiological mechanism by which endogenous ligands and synaptic activity recruit GluN3A under behavioral conditions remains unresolved. In addition, the mechanisms underlying the experience-dependent regulation of GluN3A expression in this specialized hippocampal region are still unclear. Addressing these questions will require future methodological advances that can define how endogenous GluN3A signaling is engaged and regulated during stress adaptation. Another limitation of the present study is that the chemogenetic manipulations were not restricted exclusively to pyramidal neurons. Therefore, although convergent evidence supports pyramidal neurons as the principal mediators of the observed behavioral effects, a contribution from local inhibitory neurons cannot be fully excluded. Future studies using cell-type-specific approaches will be needed to address this issue more precisely. Further work is also warranted to determine the exact roles of GluN3A-expressing pyramidal neurons and their downstream circuit targets.

In conclusion, our study identifies non-classical GluN1/GluN3A NMDARs in the hippocampal CA1i as key regulators of adaptive stress responses by selectively activating pyramidal neurons and defined projection circuits. These findings bridge hippocampal structural and transcriptional heterogeneity with behavioral specialization, offering new insight into the neural basis of coping and resilience. Furthermore, the discovery that D-serine exerts GluN3A-dependent rapid modulation of coping behavior highlights a translational opportunity for targeting non-classical NMDARs in clinical settings, such as the treatment of depression and the amelioration of stress-related symptoms.

Methods

Animals

All animal procedures were approved by the Institutional Animal Care and Use Committee of the Institute of Psychology, Chinese Academy of Sciences. C57BL/6 N male mice and CD-1 (ICR) male mice were purchased from Charles River, Beijing, China. Grin3A^-/- (GluN3A KO) mice (C57BL/6 J background) were provided by Tifei Yuan’s laboratory. This model was established by the construction of a specific small guide RNA (sgRNA) vector to direct Cas9 to delete exon 2 of the GluN3A gene. The GluN3A KO and WT littermates used in this study were bred from the Grin3A^+/- mice in the individually ventilated cages (IVCs). All mice used in the experiments were adults at 8 weeks of age, and the GluN3A KO mice were age-matched to the WT mice used in the CSDS experiments. All mice were housed in a specific-pathogen-free (SPF) environment with a temperature of 22 ± 2°C, relative humidity of 40-70%, and a 12-hour (h) light-dark cycle (7:00 am-19:00 pm). The experimental mice had free access to food and water throughout the experimental period. We have complied with all relevant ethical regulations for animal use.

CSDS model

CSDS was performed as previously described^73,74. Male CD1 mice, singly housed for 10-24 weeks, were screened based on aggression intensity. CD1 mice displaying attacks toward unfamiliar “intruders” within 3 minutes (min) for 3 consecutive days were designated as “residents” for subsequent stress experiments. 3 days before the CSDS experiment, “resident” CD1 mice were placed into the CSDS cage to establish territorial awareness. During each day of the 10-day CSDS period, adult C57BL/6 N mice were introduced into the “resident” CD1 mouse home cage and subjected to physical attacks for 10 min. For the remainder of the day, a transparent perforated PVC divider was used to separate them while maintaining visual, auditory, and olfactory contact. For 10 successive days, a C57BL/6 N mouse was introduced to a novel aggressive CD1 mouse home cage every day. C57BL/6 N control mice were housed in the CSDS cage with a divider to avoid body contact, and a new cage mate was introduced every 24 hours (h) for 10 days. After the social defeat stress, the experimental mice were individually housed until all tests were completed.

Subthreshold SDS

Experiments were performed with GluN3A KO mice involving three consecutive 5 min physical interactions with a novel CD-1 mice, with a 15 min interval between each interaction⁷⁵.

RSD model

Mice were subjected to RSD as previously described⁷⁶. Briefly, one male CD-1 aggressor mouse was introduced into the home cage of experimental mice (three male mice per cage) for 2 h per night (16:00-18:00) over six consecutive nights. Control mice were left undisturbed.

Behavioral assessments

This study assessed the anhedonia symptom using the SPT (sensory anhedonia) and the three-chamber SIT (social anhedonia). Locomotor activity was evaluated through the OFT and RT, while coping behavior was measured using the TST and FST. Anxiety symptoms were examined through the OFT and EPM.

All behavioral tests were conducted during the light cycle, with at least a 24 h interval between different tests. 30 minutes before each behavioral test, experimental mice were placed in a dimly lit testing room to acclimate to the environment. After each mouse completed the behavioral test, the apparatus was cleaned with 70% alcohol to eliminate residual mouse odors and traces. All test procedures were video-recorded.

SPT. During the adaptation phase, mice were acclimatized for 48 h by simultaneously giving one bottle of 1% sucrose solution and one bottle of pure water, and the bottle positions were switched after 24 h of acclimatization in order to eliminate the position preference effect. During testing, the mice were deprived of water and food for 12 h, followed by the placement of 1% sucrose solution and pure water for another 12 h, and the amount of the two solutions consumed in the testing phase was measured. The sucrose preference index was calculated as follows: Sucrose preference ratio (%) = Sucrose solution intake (g) / [Sucrose solution intake (g) + Pure water intake (g)] × 100%.

OFT. Mice were placed in the center of an open field arena (40×40×40 cm) to freely explore for 10 min. The central square area (24×24 cm) was designated as the center zone. The total distance traveled in the open field, entries into the center zone, and time spent in the center zone were automatically tracked using Anymaze software.

EPM. The EPM apparatus consists of a cross 50 cm above the ground (single side arm 35 cm long, 5 cm wide, closed arm 15 cm high). Mice were placed in the central area with their heads facing the open arm and allowed to freely explore for 5 min. The number of entries into the open arms and the time spent in the open arms were automatically tracked using Anymaze software. The percentage of time spent in the open arms (Open Arm Time Ratio) was calculated as follows: time spent in open arms (s)/[time spent in open arms (s) + time spent in closed arms (s)] × 100%.

SIT. The three-chamber test apparatus is an area (40 cm wide, 60 cm long, 20 cm high) divided into three equal areas by two transparent dividers with a shuttle channel at the bottom of the centre. A metal mesh cage is placed in the area on each side of the distal end. The testing process comprised two 10 min phases. During Phase 1 (adaptation phase), the test mouse is placed in the central area with the shuttle passage closed, allowing it to freely explore and familiarize itself with the environment. In Phase 2 (social test phase), an unfamiliar conspecific social mouse is placed in a metal cage on a random side area. The test mouse is then placed in the central area with the shuttle passage open, allowing free exploration of the three areas. Anymaze software automatically tracked the time the mouse spent in each area. The social interaction ratio was calculated as follows: Social interaction ratio (%) = Time spent in the social area (s)/[Time spent in the social area (s) + Time spent in the empty cage area (s)] × 100%.

TST. During testing, tape is attached to the tail tip of the experimental mouse, 1 cm from the end. A plastic tube is placed over the mouse’s tail to prevent climbing. The mouse’s tail is then secured to the top of the tail suspension apparatus, suspending the mouse approximately 25 cm above the ground for 6 minutes. The duration of immobility (s) during the 6 min video recording was recorded and analyzed.

FST. The experimental mouse is placed in a transparent cylindrical container (10 cm in diameter, 30 cm high) with water (25 cm deep, 25 ± 1°C) to swim for 6 min. The duration of immobility (s) during the last 4 minutes of the 6 min video recording was quantified and analyzed.

RT. During the training phase, mice were placed on a rotating rod at a constant speed of 5 rpm for 5 min, and the mice were placed back on the rod immediately after falling off the rod to minimize accidental falls from the rod during the test phase. Then, the mice were rested for 1 h. During the test phase, the rotating rod was set to run at an acceleration of 20 rpm/min from 4 rpm to a maximum of 40 rpm for 5 min, and the mice were placed on the rotating rod at the beginning of the initial speed. The formal test consisted of 3 replicate trials with a 30 min rest interval between each trial. The average value of the fall latency (s) of the 3 repeated trials was calculated.

Molecular biology test

This study employed qRT-PCR and Western blot to assess the mRNA and protein expression levels of the GluN3A subunit in different brain regions of mice. Immunofluorescence staining, coupled with confocal imaging, was utilized for staining analysis of neurons in the CA1 region and their projection areas. Additionally, the spatial location of viral expression was confirmed by observing the fluorescence of the injected viruses in each experiment.

qRT-PCR. After rapidly decapitating the experimental mice, the brain was promptly dissected and placed in a freezing microtome. Specific brain region boundaries were delineated based on the mouse brain atlas, and target brain regions were extracted using a 1-mm-diameter tissue extraction needle⁷⁷, as shown in the Supplementary fig1b. Total RNA was extracted from the tissue samples using TRIzol reagent (Cat# 15596018CN, Invitrogen, USA) following the manufacturer’s instructions, and the concentration was analyzed using NanoDrop (Thermo Scientific, Massachusetts, USA). Subsequently, 500 ng of RNA samples were reverse transcribed using the M-MLV Reverse Transcriptase (Cat# M1701, Promega, USA) following the reverse transcription kit instructions, with a two-step reaction condition: 70°C for 5 min, 37°C for 60 min, to generate cDNA. SYBR Green PCR Master Mix (Cat# CW2601M, Cowin Biotech, China) was used for fluorescence qPCR reactions. The real-time PCR instrument reaction steps were: a single cycle (95°C/10 min); 40 cycles (95°C/15 s; 60°C/1 min); a single cycle (95°C/15 s; 60°C/30 s; 95°C/15 s). Analysis was performed using the ΔΔC(t) method, with GAPDH as the reference gene for standardization. The primer sequences used were: Grin3A: forward: 5’-AGTACCTGAAGAATGATCCAGAGT-3’, reverse: 5’-TCAGATATGTTAGAGGTCAACGG-3’; Gapdh: forward: 5’-GCATGGCCTTCCGTGTTCCT-3’, reverse: 5’-CCCTGTTGCTGTAGCCGTAT-3’.

Western blot. The target brain area tissue was homogenized in RIPA lysis buffer containing 1% PMSF (Cat# P0100, Solarbio, China). The collected protein supernatant was quantified using a BCA protein quantification kit (Cat# PC0020, Solarbio, China), and then protein samples were mixed with 4× protein loading buffer (Cat# P1015, Solarbio, China) for denaturation. An equal number of samples were loaded onto a 10% SDS-polyacrylamide gel (Cat# CW0022M, Cowin Biotech, China) for separation and then transferred to a 0.2 μm PVDF membrane (Cat# 03010040001, Millipore, Germany). The PVDF membrane was blocked in TBST solution containing 5% milk powder and then incubated overnight at 4°C with diluted primary antibodies (5% milk powder TBST): Grin3A (1:500; Cat# orb694249, Biorbyt, UK); Gapdh (1:3000; Cat# 2118, Cell Signaling, USA). After washing with TBST, the membrane was incubated at room temperature for 2 h with secondary antibodies (5% milk powder TBST) containing horseradish peroxidase: Goat Anti-Rabbit IgG H&L (1:3000; Cat# ZB-2301, Zhongshan Golden Bridge, China). After washing again, the ECL luminescent liquid (Cat# P90719, Millipore, Germany) was added to the PVDF membrane, and the immunoblot signal was captured using an imaging system (FluorChem E, Protein Simple, California, USA).

Immunofluorescence staining. 90 minutes after the completion of the FST or CNO injection, mice were intraperitoneally injected with pentobarbital (100 mg/kg) for anesthesia. Following anesthesia, mice were perfused with 30 ml of phosphate-buffered saline (PBS) through the heart, followed by 30 ml of 4% paraformaldehyde (PFA). The brain was removed, fixed in 4% PFA for 24 h at 4°C, and then transferred to PBS with 30% sucrose for dehydration. After the brain settled, 40 μm thick coronal brain slices were obtained using a freezing microtome (CM3050S, Leica, Wetzlar, Germany) and stored at -80°C in an anti-freezing solution (50% PBS, 30% ethylene glycol, 20% glycerol). For immunofluorescence, the brain slices to be stained were washed with PBS, placed in PBST (0.1% TritonX-100, PBS) containing 5% goat serum at room temperature for 1.5 h, and then transferred to the primary antibody solution for overnight incubation at 4°C. The primary antibodies used were c-Fos (1:500; Cat# SYSY.226003, Synaptic Systems, Germany) and DYKDDDDK Tag (1:500; Cat#14793, Cell Signaling, USA). After washing with PBS, the slices were transferred to the secondary antibody solution at room temperature away from light for 1.5 h. The secondary antibodies used were Goat Anti-Rabbit IgG H&L (Alexa Fluor® 488) (1:500; Cat# ab150077, Abcam, UK) and Goat anti-Rabbit IgG (H + L), Alexa Fluor™ 546 (1:500; Cat#A11010, Invitrogen, USA). After washing with PBS, the slices were incubated in PBS with DAPI (1:500; Cat# C0060, Solarbio, China) at room temperature away from light for 10 min. Following another PBS wash, the slices were mounted on glass slides using a mounting medium (Cat# ab104139, Abcam, UK).

Confocal imaging and analyses. Sliced brains or immunofluorescent-stained brain slices were imaged using a high-speed confocal imaging system (SpinSR, Olympus, Tokyo, Japan) and a digital slice scanning and analysis system (DMI8, Leica, Wetzlar, Germany) under 10× and 20× objectives. Automated cell counting for c-Fos-positive neurons was performed using the ImageJ Fiji plugin (http://imagej.nih.gov/ij/). Two to four images were selected in each animal for statistical analysis of the brain regions.

In vivo manipulation

AAV. AAV2/9-syn-m-Grin3A-3×flag-Null (titer 1.6×1012 vg/mL) and AAV2/9-syn-ZsGreen (titer 1.6×1012 vg/mL), AAV2/9- syn-hM3D(Gq)-mCherry (titer 1.2×1012 vg/mL), AAV2/9-syn-hM4D(Gi)-mCherry (titer 1.4×1012 vg/mL), and AAV2/9-syn-mCherry (titer 1.5×1012 vg/mL) were purchased from Hanbio Biotechnology, Shanghai, China. AAV2/PHP.eB-mCaMKIIα-GCaMP8s-WPRE-pA (titer 1.1 × 1013 vg/mL) was purchased from Taitool Bio, Shanghai, China. The viral vectors were aliquoted and stored at −80 °C.

Stereotaxic surgery. Mice were anesthetized with isoflurane and secured in a stereotaxic apparatus (RWD Life Science, Shenzhen, China). After exposing the skull, the corresponding virus was slowly injected into the target location (AP: −2.90 mm; ML ± 3.30 mm; DV: −2.90 mm) using a Stoelting pump (QSI storage pump, Stoelting, Wisconsin, USA) attached to a Hamilton syringe. After injection, the syringe was left in place for 10 min before slow withdrawal. The mice were allowed to recover for 14 days before behavioral testing.

Regulation of the GluN3A subunit and virus tracing experiments. 0.15 µL (unilateral) of AAV2/9-syn-m-Grin3A-3×flag-Null or AAV2/9-syn-ZsGreen was bilaterally injected into CA1i at a rate of 0.05 µL/min.

Chemogenetic experiments. 0.15 µL (unilateral) of AAV2/9-syn- hM3D(Gq)-mCherry or AAV2/9-syn-hM4D(Gi)-mCherry (or AAV2/9-syn- mCherry) was injected into each hemisphere of CA1i at a rate of 0.05 µL/min. 14 days later, CNO (Cat# HY-17366, MCE, USA) was dissolved in dimethyl sulfoxide (DMSO, Cat# D8418, Sigma, Germany) to create a 100 mg/mL stock solution. Before injection, further dilution was made with 0.9% saline to achieve a CNO injection quality of 0.2 mg/kg (injection volume: 10 mL/kg) for chemogenetic activation experiments and 2 mg/kg for chemogenetic inhibition experiments. CNO was intraperitoneally injected 30 minutes before behavioral testing.

Fiber photometry experiments. AAV2/PHP.eB-mCaMKIIα-GCaMP8s-WPRE-pA was diluted 1:4 in cerebrospinal fluid. 0.2 µL (unilateral) of the viral diluted solution was injected into the bilateral CA1i at a rate of 0.05 µL/min. Then, a ceramic optical fiber (200 µm in diameter, 3 mm length, 0.37 NA, Inper Technology Co., Ltd., Hangzhou, China) was implanted above the CA1i region (AP: −2.90 mm; ML ± 3.30 mm; DV: −2.85 mm). The fiber was secured to the skull with dental cement. 14 days later, power levels of 470 nm excitation light and 410 nm excitation light in the Inper Signal fiber photometry system (Inper Technology Co., Ltd., Hangzhou, China) were measured using a photometer and adjusted to 25 µW and 10-15 µW, respectively. Prior to the TST, the bundled fiber was connected to the optical fiber implant in the mouse’s head for a two-min adaptation. The Inper Signal system recorded Ca2+ activity in the CA1i region of mice suspended and during tail suspension tests, synchronizing with a camera to record the behavioral state. Inper Plot software was used to analyze and statistically process the fluorescence change data (F) during hanging and struggling events, calculating the relative fluorescence change ΔF/F0 (ΔF/F0 = (F-F0)/F0). Mice with fiber implantation positions confirming the expected target were included in data analysis.

Drug delivery experiments. After injecting chemogenetic viruses into bilateral CA1i, 22-gauge single-guide cannulas (RWD Life Science, Shenzhen, China) were bilaterally implanted 0.5 mm above the target positions (AP: -2.90 mm; ML: ±3.30 mm; DV: −-2.40 mm). The cannulas were fixed on the skull surface with dental cement. 14 days later, 10 mg/mL D-serine was prepared in CSF. 30 min before the behavioral test, mice were anesthetized with isoflurane, and an injection inner tube exceeding the cannula target position by 0.5 mm was inserted into the mouse’s head cannula. Subsequently, using a dual-channel microinjection pump (RWD Life Science, Shenzhen, China), 0.5 µL (unilateral) of D-serine or CSF was injected into the bilateral CA1i region at a rate of 0.2 µL/min. After completion, the inner tube was left in place for 5 min before slow withdrawal.

SnRNAseq analysis

The samples from “Control WT” and “RSD WT” in GSE253687⁷⁸ were reanalyzed using the standard analysis pipeline of Seurat (version 4.4.0)⁷⁹. Briefly, only cells expressing at least 200 genes and genes containing non-zero values as well as associated with at least 3 cells, were maintained for subsequent analysis. “NormalizeData” functions were used to normalize data. Highly variable genes were observed using the “FindVariableFeatures” function with default parameter settings, and these highly variable genes were used to perform principal components analysis (PCA). Dimensionality reduction was performed using the “RunUMAP” function, where reduction was set to ‘pca’ and dims to 1:50. Default resolution was used for clustering, using “FindNeighbors” and “FindClusters” functions. Next, we extracted the cells from the CA1 excitatory neuron subpopulation (Mpped1 positive) and performed the aforementioned analysis again. Due to the lack of effective marker genes for screening the CA1 inhibitory neuron subpopulation, we could not extract and analyze the inhibitory neuron subpopulation of CA1⁸⁰. The R package ggplot2⁸¹ was used for visualization.

Statistics and reproducibility

All data are presented as mean ± standard error of the mean (M ± SEM) and analyzed using GraphPad Prism 9.0 software. Normal distribution was assessed using the Anderson-Darling test. Differences between the two groups were evaluated using an unpaired two-tailed Student’s t-test, and when variances were unequal, Welch’s correction was applied. Analysis of variance with one or two factors was used for comparisons among three groups or more, followed by FDR: Benjamini-Krieger-Yekutieli correction for multiple comparisons. A p-value < 0.05 was considered statistically significant.

Some animals or samples were excluded based on predefined experimental inclusion criteria or experimental failure, including failure of behavioral testing procedures, body injuries that could interfere with behavioral performance, accidental death during the study, or failure of drug delivery caused by cannula loss or blockage. In addition, some molecular measurements were obtained from independent batches of animals owing to anatomical and tissue-collection constraints.

Data reporting

No statistical methods were used to predetermine sample size. Sample sizes were estimated based on our previous experience and with consideration of the 3 R principles. Animals were randomly assigned to control and stress/intervention groups. The order of behavioral testing, interventions, and animal placement was alternated across groups. Data were analyzed by investigators blinded to group allocation.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Transparent Peer Review file^{(2.3MB, pdf)}

Supplementary information^{(2.6MB, pdf)}

42003_2026_10457_MOESM3_ESM.pdf^{(72.3KB, pdf)}

Description of Additional Supplementary Materials

Supplementary Data 1^{(119KB, xls)}

Reporting summary^{(2.1MB, pdf)}

Acknowledgements

We are grateful to the lab of Tifei Yuan (Shanghai Mental Health Center, Shanghai, China) for providing the GluN3A KO mice. We are grateful to Wenyu Tian (Shandong University, Shandong, China) for database technologies and analysis assistance. We greatly appreciate the excellent work of the technical support staff at the Institutional Center for Shared Technologies and Facilities of the Institute of Psychology, Chinese Academy of Sciences.

Author contributions

Methodology, data curation and investigation, writing-original draft&editing: W.Z.; Research conceptualization, project administration, funding acquisition and editing: J.W.; Research conceptualization, resources, project administration, funding acquisition, supervision and writing-review&editing: W.W.; Methodology: M.Z., J.D., L.Z., and H.Y. participated.

Peer review

Peer review information

Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editors: Christoph Anacker and Benjamin Bessieres. A peer review file is available.

Funding

W.W. discloses support for the research of this work from the National Natural Science Foundation of China [grant number 82071517, U21A20364], National Key Research and Development Program of China [Grant number 2017YFE0126500], and the Scientific Foundation of the Institute of Psychology, Chinese Academy of Sciences (Grant number E2CX4115CX). J.W. discloses support for the research of this work from the National Natural Science Foundation of China (Grant number 31900731).

Data availability

The data from the SnRNAseq analysis employed here is available at https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE253687, Series GSE253687: Sequencing to determine the effect of neuronal IL-1 receptor knockout and microglia depletion on stress-induced changes to the neuronal transcriptome: Goodman E, Biltz R, Packer J, DiSabato D, Swanson S, Oliver B, Quan N, Sheridan J, Godbout J, 2024. The numerical source data underlying all graphs are provided in Supplementary Data 1. Additional data that support the findings of this study are available from the corresponding author upon reasonable request.

Code availability

All packages and software used in this study were obtained from publicly available sources, as detailed in the “Methods” section. The code necessary to reproduce the primary results of this study is available at https://github.com/chiefzhangw/CA1-GluN3A-IN-RSD-MICE-from-GSE253687.git.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors jointly supervised this work: Jiesi Wang and Weiwen Wang.

Contributor Information

Jiesi Wang, Email: wangjs@psych.ac.cn.

Weiwen Wang, Email: wangww@psych.ac.cn.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s42003-026-10457-9.

References

1.Roddy, D. W. et al. The Hippocampus in depression: more than the sum of its parts? Advanced hippocampal substructure segmentation in depression. Biol. Psychiatry85, 487–497 (2019). [DOI] [PubMed] [Google Scholar]
2.Anacker, C. et al. Hippocampal neurogenesis confers stress resilience by inhibiting the ventral dentate gyrus. Nature559, 98–102 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Strange, B. A., Witter, M. P., Lein, E. S. & Moser, E. I. Functional organization of the hippocampal longitudinal axis. Nat. Rev. Neurosci.15, 655–669 (2014). [DOI] [PubMed] [Google Scholar]
4.Fanselow, M. S. & Dong, H. W. Are the dorsal and ventral hippocampus functionally distinct structures? Neuron65, 7–19 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Tzilivaki, A. et al. Hippocampal GABAergic interneurons and memory. Neuron111, 3154–3175 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Schultz, C. & Engelhardt, M. Anatomy of the hippocampal formation. Front Neurol. Neurosci.34, 6–17 (2014). [DOI] [PubMed] [Google Scholar]
7.Ayhan, F. et al. Resolving cellular and molecular diversity along the hippocampal anterior-to-posterior axis in humans. Neuron109, 2091–2105.e2096 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Levone, B. R., Cryan, J. F. & O’Leary, O. F. Specific sub-regions of the longitudinal axis of the hippocampus mediate behavioural responses to chronic psychosocial stress. Neuropharmacology201, 108843 (2021). [DOI] [PubMed] [Google Scholar]
9.Bienkowski, M. S. et al. Integration of gene expression and brain-wide connectivity reveals the multiscale organization of mouse hippocampal networks. Nat. Neurosci.21, 1628–1643 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Okuyama, T., Kitamura, T., Roy, D. S., Itohara, S. & Tonegawa, S. Ventral CA1 neurons store social memory. Science353, 1536–1541 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Pachernegg, S., Strutz-Seebohm, N. & Hollmann, M. GluN3 subunit-containing NMDA receptors: not just one-trick ponies. Trends Neurosci.35, 240–249 (2012). [DOI] [PubMed] [Google Scholar]
12.Perez-Otano, I., Larsen, R. S. & Wesseling, J. F. Emerging roles of GluN3-containing NMDA receptors in the CNS. Nat. Rev. Neurosci.17, 623–635 (2016). [DOI] [PubMed] [Google Scholar]
13.Bossi, S., Pizzamiglio, L. & Paoletti, P. Excitatory GluN1/GluN3A glycine receptors (eGlyRs) in brain signaling. Trends Neurosci.46, 667–681 (2023). [DOI] [PubMed] [Google Scholar]
14.Grand, T., Abi Gerges, S., David, M., Diana, M. A. & Paoletti, P. Unmasking GluN1/GluN3A excitatory glycine NMDA receptors. Nat. Commun.9, 4769 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Pizzamiglio, L. et al. Excitatory glycine receptors control ventral hippocampus synaptic plasticity and anxiety-related behaviors. Proc. Natl. Acad. Sci. USA122, e2501118122 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Bossi, S. et al. GluN3A excitatory glycine receptors control adult cortical and amygdalar circuits. Neuron110, 2438–2454.e2438 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Conde-Dusman, M. J. et al. Control of protein synthesis and memory by GluN3A-NMDA receptors through inhibition of GIT1/mTORC1 assembly. Elife10, e71575 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Chatterton, J. E. et al. Excitatory glycine receptors containing the NR3 family of NMDA receptor subunits. Nature415, 793–798 (2002). [DOI] [PubMed] [Google Scholar]
19.Murillo, A. et al. Temporal dynamics and neuronal specificity of Grin3a expression in the mouse forebrain. Cereb. Cortex31, 1914–1926 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Otsu, Y. et al. Control of aversion by glycine-gated GluN1/GluN3A NMDA receptors in the adult medial habenula. Science366, 250–254 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Zhong, W. et al. Pathogenesis of sporadic Alzheimer’s disease by deficiency of NMDA receptor subunit GluN3A. Alzheimer's Dement18, 222–239 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Roberts, A. C. et al. Downregulation of NR3A-containing NMDARs is required for synapse maturation and memory consolidation. Neuron63, 342–356 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Wang, J. et al. Epigenetic activation of ASCT2 in the Hippocampus contributes to depression-like behavior by regulating D-Serine in Mice. Front. Mol. Neurosci.10, 139 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
24.MacKay, M. B. et al. D-Serine: potential therapeutic agent and/or biomarker in schizophrenia and depression? Front Psychiatry10, 25 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Otte, D. M. et al. Effects of Chronic D-serine elevation on animal models of depression and anxiety-related behavior. PLoS One8, e67131 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Zhang, W., Wang, W., Zhou, Y. & Wang, J. A comparative study of blood and hippocampal D-serine change patterns in drug-naive patients and animal models of depression. Psychiatry Res.348, 116453 (2025). [DOI] [PubMed] [Google Scholar]
27.Chen, Z. et al. D-Serine produces antidepressant-like effects in mice through suppression of BDNF signaling pathway and regulation of synaptic adaptations in the nucleus accumbens. Mol. Med.27, 127 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Malkesman, O. et al. Acute D-serine treatment produces antidepressant-like effects in rodents. Int J. Neuropsychopharmacol.15, 1135–1148 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Wei, I. H. et al. Acute Amino Acid d-Serine Administration, Similar to Ketamine, Produces Antidepressant-like Effects through Identical Mechanisms. J. Agric. Food Chem.65, 10792–10803 (2017). [DOI] [PubMed] [Google Scholar]
30.Pignatelli, M. et al. Cooperative synaptic and intrinsic plasticity in a disynaptic limbic circuit drive stress-induced anhedonia and passive coping in mice. Mol. Psychiatry26, 1860–1879 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Lu, J. J. et al. BNIP3L/NIX-mediated mitophagy alleviates passive stress-coping behaviors induced by tumor necrosis factor-alpha. Mol. Psychiatry28, 5062–5076 (2023). [DOI] [PubMed] [Google Scholar]
32.Jing, W. et al. A circuit of COCH neurons encodes social-stress-induced anxiety via MTF1 activation of Cacna1h. Cell Rep.37, 110177 (2021). [DOI] [PubMed] [Google Scholar]
33.Duman, R. S., Aghajanian, G. K., Sanacora, G. & Krystal, J. H. Synaptic plasticity and depression: new insights from stress and rapid-acting antidepressants. Nat. Med22, 238–249 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Li, Y. et al. Stress-induced reduction of Na(+)/H(+) exchanger isoform 1 promotes maladaptation of neuroplasticity and exacerbates depressive behaviors. Sci. Adv.8, eadd7063 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Su, T., Lu, Y., Fu, C., Geng, Y. & Chen, Y. GluN2A mediates ketamine-induced rapid antidepressant-like responses. Nat. Neurosci.26, 1751–1761 (2023). [DOI] [PubMed] [Google Scholar]
36.Xia, F. et al. Understanding the neural code of stress to control anhedonia. Nature637, 654–662 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Ma, Y. N. et al. Dorsal CA1 NECTIN3 reduction mediates early-life stress-induced object recognition memory deficits in adolescent female mice. Neurosci. Bull.41, 243–260 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Kim, J., Lei, Y., Lu, X. Y. & Kim, C. S. Glucocorticoid-glucocorticoid receptor-HCN1 channels reduce neuronal excitability in dorsal hippocampal CA1 neurons. Mol. Psychiatry27, 4035–4049 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Ma, H. et al. Amygdala-hippocampal innervation modulates stress-induced depressive-like behaviors through AMPA receptors. Proc. Natl. Acad. Sci. USA118, e2019409118 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Maras, P. M. et al. Preferential loss of dorsal-hippocampus synapses underlies memory impairments provoked by short, multi-modal stress. Mol. Psychiatry19, 745 (2014). [DOI] [PubMed] [Google Scholar]
41.Huang, M. et al. Ventral Hippocampal CA1 GADD45B Regulates Susceptibility to Social Stress by Influencing NMDA Receptor-Mediated Synaptic Plasticity. Neurosci. Bull.41, 406–420 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Diniz, C. R., Casarotto, P. C. & Joca, S. R. NMDA-NO signaling in the dorsal and ventral hippocampus time-dependently modulates the behavioral responses to forced swimming stress. Behav. Brain Res.307, 126–136 (2016). [DOI] [PubMed] [Google Scholar]
43.Chang, C. H., Su, C. L. & Gean, P. W. Mechanism underlying NMDA blockade-induced inhibition of aggression in post-weaning socially isolated mice. Neuropharmacology143, 95–105 (2018). [DOI] [PubMed] [Google Scholar]
44.Goodman, E. J. et al. Enhanced fear memory after social defeat in mice is dependent on interleukin-1 receptor signaling in glutamatergic neurons. Mol. Psychiatry (2024). [DOI] [PMC free article] [PubMed]
45.Zhang, M. et al. Dysfunction of GluN3A subunit is involved in depression-like behaviors through synaptic deficits. J. Affect. Disord.332, 72–82 (2023). [DOI] [PubMed] [Google Scholar]
46.Commons, K. G., Cholanians, A. B., Babb, J. A. & Ehlinger, D. G. The rodent forced swim test measures stress-coping strategy, not depression-like behavior. ACS Chem. Neurosci.8, 955–960 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Cryan, J. F., Mombereau, C. & Vassout, A. The tail suspension test as a model for assessing antidepressant activity: review of pharmacological and genetic studies in mice. Neurosci. Biobehav. Rev.29, 571–625 (2005). [DOI] [PubMed] [Google Scholar]
48.Zapater-Fajari, M., Crespo-Sanmiguel, I., Pulopulos, M. M., Hidalgo, V. & Salvador, A. Resilience and psychobiological response to stress in older people: the mediating role of coping strategies. Front. Aging Neurosci.13, 632141 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Andalman, A. S. et al. Neuronal dynamics regulating brain and behavioral state transitions. Cell177, 970–985.e920 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Kent, M. et al. Profiling coping strategies in male and female rats: Potential neurobehavioral markers of increased resilience to depressive symptoms. Horm. Behav.95, 33–43 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Wood, S. K. & Bhatnagar, S. Resilience to the effects of social stress: evidence from clinical and preclinical studies on the role of coping strategies. Neurobiol. Stress1, 164–173 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Kato, T. Testing of the coping flexibility hypothesis based on the dual-process theory: Relationships between coping flexibility and depressive Symptoms. Psychiatry Res.230, 137–142 (2015). [DOI] [PubMed] [Google Scholar]
53.Kato, T. The impact of coping flexibility on the risk of depressive symptoms. PLoS One10, e0128307 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Kato, T. Development of the Coping Flexibility Scale: evidence for the coping flexibility hypothesis. J. Couns. Psychol.59, 262–273 (2012). [DOI] [PubMed] [Google Scholar]
55.Rao, J. S., Harry, G. J., Rapoport, S. I. & Kim, H. W. Increased excitotoxicity and neuroinflammatory markers in postmortem frontal cortex from bipolar disorder patients. Mol. Psychiatry15, 384–392 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Crawley, O., Conde-Dusman, M. J. & Perez-Otano, I. GluN3A NMDA receptor subunits: more enigmatic than ever? J. Physiol.600, 261–276 (2022). [DOI] [PubMed] [Google Scholar]
57.Kulkarni, S. K. & Dhir, A. Effect of various classes of antidepressants in behavioral paradigms of despair. Prog. Neuropsychopharmacol. Biol. Psychiatry31, 1248–1254 (2007). [DOI] [PubMed] [Google Scholar]
58.Renard, C. E., Dailly, E., David, D. J., Hascoet, M. & Bourin, M. Monoamine metabolism changes following the mouse forced swimming test but not the tail suspension test. Fundam. Clin. Pharm.17, 449–455 (2003). [DOI] [PubMed] [Google Scholar]
59.Chatterjee, M., Jaiswal, M. & Palit, G. Comparative evaluation of forced swim test and tail suspension test as models of negative symptom of schizophrenia in rodents. ISRN Psychiatry2012, 595141 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Lee, J. H., Wei, L., Deveau, T. C., Gu, X. & Yu, S. P. Expression of the NMDA receptor subunit GluN3A (NR3A) in the olfactory system and its regulatory role on olfaction in the adult mouse. Brain Struct. Funct.221, 3259–3273 (2016). [DOI] [PubMed] [Google Scholar]
61.Yao, Y. & Mayer, M. L. Characterization of a soluble ligand binding domain of the NMDA receptor regulatory subunit NR3A. J. Neurosci.26, 4559–4566 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Ragguett, R. M., Rong, C., Kratiuk, K. & McIntyre, R. S. Rapastinel - an investigational NMDA-R modulator for major depressive disorder: evidence to date. Expert Opin. Investig. Drugs28, 113–119 (2019). [DOI] [PubMed] [Google Scholar]
63.Keck, M. E. et al. Reduction of hypothalamic vasopressinergic hyperdrive contributes to clinically relevant behavioral and neuroendocrine effects of chronic paroxetine treatment in a psychopathological rat model. Neuropsychopharmacology28, 235–243 (2003). [DOI] [PubMed] [Google Scholar]
64.Carhart-Harris, R. L. & Nutt, D. J. Serotonin and brain function: a tale of two receptors. J. Psychopharmacol.31, 1091–1120 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Mahar, I., Bambico, F. R., Mechawar, N. & Nobrega, J. N. Stress, serotonin, and hippocampal neurogenesis in relation to depression and antidepressant effects. Neurosci. Biobehav Rev.38, 173–192 (2014). [DOI] [PubMed] [Google Scholar]
66.Wolosker, H. & Balu, D. T. D-Serine as the gatekeeper of NMDA receptor activity: implications for the pharmacologic management of anxiety disorders. Transl. Psychiatry10, 184 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Chen, L. et al. Ezrin-mediated astrocyte-synapse signaling regulates cognitive function via astrocyte morphological changes in fine processes in male mice. Brain Behav. Immun.124, 177–191 (2024). [DOI] [PubMed] [Google Scholar]
68.Gomez-Galan, M., De Bundel, D., Van Eeckhaut, A., Smolders, I. & Lindskog, M. Dysfunctional astrocytic regulation of glutamate transmission in a rat model of depression. Mol. Psychiatry18, 582–594 (2013). [DOI] [PubMed] [Google Scholar]
69.Basu, A. C. et al. Targeted disruption of serine racemase affects glutamatergic neurotransmission and behavior. Mol. Psychiatry14, 719–727 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Levin, R. et al. Behavioral and cognitive effects of the N-methyl-D-aspartate receptor co-agonist D-serine in healthy humans: initial findings. J. Psychiatr. Res.61, 188–195 (2015). [DOI] [PubMed] [Google Scholar]
71.Yoon, S. H., Song, W. S., Chung, G., Kim, S. J. & Kim, M. H. Activity in the dorsal hippocampus-mPFC circuit modulates stress-coping strategies during inescapable stress. Exp. Mol. Med.56, 1921–1935 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Bagot, R. C. et al. Ventral hippocampal afferents to the nucleus accumbens regulate susceptibility to depression. Nat. Commun.6, 7062 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
73.Xu, H. et al. mPFC GABAergic transmission mediated the role of BDNF signaling in cognitive impairment but not anxiety induced by adolescent social stress. Neuropharmacology184, 108412 (2021). [DOI] [PubMed] [Google Scholar]
74.Golden, S. A., Covington, H. E., 3rd, Berton, O. & Russo, S. J. A standardized protocol for repeated social defeat stress in mice. Nat.Protoc.6, 1183-1191 (2011). [DOI] [PMC free article] [PubMed]
75.Menard, C. et al. Social stress induces neurovascular pathology promoting depression. Nat. Neurosci.20, 1752–1760 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Wohleb, E. S. et al. beta-Adrenergic receptor antagonism prevents anxiety-like behavior and microglial reactivity induced by repeated social defeat. J. Neurosci.31, 6277–6288 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
77.Franklin, K. B. J. & Paxinos, G. Paxinos and Franklin’s The mouse brain in stereotaxic coordinates, Academic Press, an imprint of Elsevier, Amsterdam (2013).
78.Goodman, E. J. et al. Sequencing to determine the effect of neuronal IL-1 receptor knockout and microglia depletion on stress-induced changes to the neuronal transcriptome. GEO, V1.https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE253687 (2024).
79.Satija, R., Farrell, J. A., Gennert, D., Schier, A. F. & Regev, A. Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol.33, 495–502 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Habib, N. et al. Div-Seq: Single-nucleus RNA-Seq reveals dynamics of rare adult newborn neurons. Science353, 925–928 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
81.Wickham, H. & Wickham, H. Data analysis, (Springer, 2016).

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Transparent Peer Review file^{(2.3MB, pdf)}

Supplementary information^{(2.6MB, pdf)}

42003_2026_10457_MOESM3_ESM.pdf^{(72.3KB, pdf)}

Description of Additional Supplementary Materials

Supplementary Data 1^{(119KB, xls)}

Reporting summary^{(2.1MB, pdf)}

Data Availability Statement

[CR1] 1.Roddy, D. W. et al. The Hippocampus in depression: more than the sum of its parts? Advanced hippocampal substructure segmentation in depression. Biol. Psychiatry85, 487–497 (2019). [DOI] [PubMed] [Google Scholar]

[CR2] 2.Anacker, C. et al. Hippocampal neurogenesis confers stress resilience by inhibiting the ventral dentate gyrus. Nature559, 98–102 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.Strange, B. A., Witter, M. P., Lein, E. S. & Moser, E. I. Functional organization of the hippocampal longitudinal axis. Nat. Rev. Neurosci.15, 655–669 (2014). [DOI] [PubMed] [Google Scholar]

[CR4] 4.Fanselow, M. S. & Dong, H. W. Are the dorsal and ventral hippocampus functionally distinct structures? Neuron65, 7–19 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Tzilivaki, A. et al. Hippocampal GABAergic interneurons and memory. Neuron111, 3154–3175 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR6] 6.Schultz, C. & Engelhardt, M. Anatomy of the hippocampal formation. Front Neurol. Neurosci.34, 6–17 (2014). [DOI] [PubMed] [Google Scholar]

[CR7] 7.Ayhan, F. et al. Resolving cellular and molecular diversity along the hippocampal anterior-to-posterior axis in humans. Neuron109, 2091–2105.e2096 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Levone, B. R., Cryan, J. F. & O’Leary, O. F. Specific sub-regions of the longitudinal axis of the hippocampus mediate behavioural responses to chronic psychosocial stress. Neuropharmacology201, 108843 (2021). [DOI] [PubMed] [Google Scholar]

[CR9] 9.Bienkowski, M. S. et al. Integration of gene expression and brain-wide connectivity reveals the multiscale organization of mouse hippocampal networks. Nat. Neurosci.21, 1628–1643 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Okuyama, T., Kitamura, T., Roy, D. S., Itohara, S. & Tonegawa, S. Ventral CA1 neurons store social memory. Science353, 1536–1541 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Pachernegg, S., Strutz-Seebohm, N. & Hollmann, M. GluN3 subunit-containing NMDA receptors: not just one-trick ponies. Trends Neurosci.35, 240–249 (2012). [DOI] [PubMed] [Google Scholar]

[CR12] 12.Perez-Otano, I., Larsen, R. S. & Wesseling, J. F. Emerging roles of GluN3-containing NMDA receptors in the CNS. Nat. Rev. Neurosci.17, 623–635 (2016). [DOI] [PubMed] [Google Scholar]

[CR13] 13.Bossi, S., Pizzamiglio, L. & Paoletti, P. Excitatory GluN1/GluN3A glycine receptors (eGlyRs) in brain signaling. Trends Neurosci.46, 667–681 (2023). [DOI] [PubMed] [Google Scholar]

[CR14] 14.Grand, T., Abi Gerges, S., David, M., Diana, M. A. & Paoletti, P. Unmasking GluN1/GluN3A excitatory glycine NMDA receptors. Nat. Commun.9, 4769 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Pizzamiglio, L. et al. Excitatory glycine receptors control ventral hippocampus synaptic plasticity and anxiety-related behaviors. Proc. Natl. Acad. Sci. USA122, e2501118122 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Bossi, S. et al. GluN3A excitatory glycine receptors control adult cortical and amygdalar circuits. Neuron110, 2438–2454.e2438 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Conde-Dusman, M. J. et al. Control of protein synthesis and memory by GluN3A-NMDA receptors through inhibition of GIT1/mTORC1 assembly. Elife10, e71575 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Chatterton, J. E. et al. Excitatory glycine receptors containing the NR3 family of NMDA receptor subunits. Nature415, 793–798 (2002). [DOI] [PubMed] [Google Scholar]

[CR19] 19.Murillo, A. et al. Temporal dynamics and neuronal specificity of Grin3a expression in the mouse forebrain. Cereb. Cortex31, 1914–1926 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Otsu, Y. et al. Control of aversion by glycine-gated GluN1/GluN3A NMDA receptors in the adult medial habenula. Science366, 250–254 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Zhong, W. et al. Pathogenesis of sporadic Alzheimer’s disease by deficiency of NMDA receptor subunit GluN3A. Alzheimer's Dement18, 222–239 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Roberts, A. C. et al. Downregulation of NR3A-containing NMDARs is required for synapse maturation and memory consolidation. Neuron63, 342–356 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Wang, J. et al. Epigenetic activation of ASCT2 in the Hippocampus contributes to depression-like behavior by regulating D-Serine in Mice. Front. Mol. Neurosci.10, 139 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.MacKay, M. B. et al. D-Serine: potential therapeutic agent and/or biomarker in schizophrenia and depression? Front Psychiatry10, 25 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Otte, D. M. et al. Effects of Chronic D-serine elevation on animal models of depression and anxiety-related behavior. PLoS One8, e67131 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Zhang, W., Wang, W., Zhou, Y. & Wang, J. A comparative study of blood and hippocampal D-serine change patterns in drug-naive patients and animal models of depression. Psychiatry Res.348, 116453 (2025). [DOI] [PubMed] [Google Scholar]

[CR27] 27.Chen, Z. et al. D-Serine produces antidepressant-like effects in mice through suppression of BDNF signaling pathway and regulation of synaptic adaptations in the nucleus accumbens. Mol. Med.27, 127 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Malkesman, O. et al. Acute D-serine treatment produces antidepressant-like effects in rodents. Int J. Neuropsychopharmacol.15, 1135–1148 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Wei, I. H. et al. Acute Amino Acid d-Serine Administration, Similar to Ketamine, Produces Antidepressant-like Effects through Identical Mechanisms. J. Agric. Food Chem.65, 10792–10803 (2017). [DOI] [PubMed] [Google Scholar]

[CR30] 30.Pignatelli, M. et al. Cooperative synaptic and intrinsic plasticity in a disynaptic limbic circuit drive stress-induced anhedonia and passive coping in mice. Mol. Psychiatry26, 1860–1879 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Lu, J. J. et al. BNIP3L/NIX-mediated mitophagy alleviates passive stress-coping behaviors induced by tumor necrosis factor-alpha. Mol. Psychiatry28, 5062–5076 (2023). [DOI] [PubMed] [Google Scholar]

[CR32] 32.Jing, W. et al. A circuit of COCH neurons encodes social-stress-induced anxiety via MTF1 activation of Cacna1h. Cell Rep.37, 110177 (2021). [DOI] [PubMed] [Google Scholar]

[CR33] 33.Duman, R. S., Aghajanian, G. K., Sanacora, G. & Krystal, J. H. Synaptic plasticity and depression: new insights from stress and rapid-acting antidepressants. Nat. Med22, 238–249 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Li, Y. et al. Stress-induced reduction of Na(+)/H(+) exchanger isoform 1 promotes maladaptation of neuroplasticity and exacerbates depressive behaviors. Sci. Adv.8, eadd7063 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Su, T., Lu, Y., Fu, C., Geng, Y. & Chen, Y. GluN2A mediates ketamine-induced rapid antidepressant-like responses. Nat. Neurosci.26, 1751–1761 (2023). [DOI] [PubMed] [Google Scholar]

[CR36] 36.Xia, F. et al. Understanding the neural code of stress to control anhedonia. Nature637, 654–662 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Ma, Y. N. et al. Dorsal CA1 NECTIN3 reduction mediates early-life stress-induced object recognition memory deficits in adolescent female mice. Neurosci. Bull.41, 243–260 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Kim, J., Lei, Y., Lu, X. Y. & Kim, C. S. Glucocorticoid-glucocorticoid receptor-HCN1 channels reduce neuronal excitability in dorsal hippocampal CA1 neurons. Mol. Psychiatry27, 4035–4049 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR39] 39.Ma, H. et al. Amygdala-hippocampal innervation modulates stress-induced depressive-like behaviors through AMPA receptors. Proc. Natl. Acad. Sci. USA118, e2019409118 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Maras, P. M. et al. Preferential loss of dorsal-hippocampus synapses underlies memory impairments provoked by short, multi-modal stress. Mol. Psychiatry19, 745 (2014). [DOI] [PubMed] [Google Scholar]

[CR41] 41.Huang, M. et al. Ventral Hippocampal CA1 GADD45B Regulates Susceptibility to Social Stress by Influencing NMDA Receptor-Mediated Synaptic Plasticity. Neurosci. Bull.41, 406–420 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Diniz, C. R., Casarotto, P. C. & Joca, S. R. NMDA-NO signaling in the dorsal and ventral hippocampus time-dependently modulates the behavioral responses to forced swimming stress. Behav. Brain Res.307, 126–136 (2016). [DOI] [PubMed] [Google Scholar]

[CR43] 43.Chang, C. H., Su, C. L. & Gean, P. W. Mechanism underlying NMDA blockade-induced inhibition of aggression in post-weaning socially isolated mice. Neuropharmacology143, 95–105 (2018). [DOI] [PubMed] [Google Scholar]

[CR44] 44.Goodman, E. J. et al. Enhanced fear memory after social defeat in mice is dependent on interleukin-1 receptor signaling in glutamatergic neurons. Mol. Psychiatry (2024). [DOI] [PMC free article] [PubMed]

[CR45] 45.Zhang, M. et al. Dysfunction of GluN3A subunit is involved in depression-like behaviors through synaptic deficits. J. Affect. Disord.332, 72–82 (2023). [DOI] [PubMed] [Google Scholar]

[CR46] 46.Commons, K. G., Cholanians, A. B., Babb, J. A. & Ehlinger, D. G. The rodent forced swim test measures stress-coping strategy, not depression-like behavior. ACS Chem. Neurosci.8, 955–960 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR47] 47.Cryan, J. F., Mombereau, C. & Vassout, A. The tail suspension test as a model for assessing antidepressant activity: review of pharmacological and genetic studies in mice. Neurosci. Biobehav. Rev.29, 571–625 (2005). [DOI] [PubMed] [Google Scholar]

[CR48] 48.Zapater-Fajari, M., Crespo-Sanmiguel, I., Pulopulos, M. M., Hidalgo, V. & Salvador, A. Resilience and psychobiological response to stress in older people: the mediating role of coping strategies. Front. Aging Neurosci.13, 632141 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Andalman, A. S. et al. Neuronal dynamics regulating brain and behavioral state transitions. Cell177, 970–985.e920 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR50] 50.Kent, M. et al. Profiling coping strategies in male and female rats: Potential neurobehavioral markers of increased resilience to depressive symptoms. Horm. Behav.95, 33–43 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Wood, S. K. & Bhatnagar, S. Resilience to the effects of social stress: evidence from clinical and preclinical studies on the role of coping strategies. Neurobiol. Stress1, 164–173 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Kato, T. Testing of the coping flexibility hypothesis based on the dual-process theory: Relationships between coping flexibility and depressive Symptoms. Psychiatry Res.230, 137–142 (2015). [DOI] [PubMed] [Google Scholar]

[CR53] 53.Kato, T. The impact of coping flexibility on the risk of depressive symptoms. PLoS One10, e0128307 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Kato, T. Development of the Coping Flexibility Scale: evidence for the coping flexibility hypothesis. J. Couns. Psychol.59, 262–273 (2012). [DOI] [PubMed] [Google Scholar]

[CR55] 55.Rao, J. S., Harry, G. J., Rapoport, S. I. & Kim, H. W. Increased excitotoxicity and neuroinflammatory markers in postmortem frontal cortex from bipolar disorder patients. Mol. Psychiatry15, 384–392 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Crawley, O., Conde-Dusman, M. J. & Perez-Otano, I. GluN3A NMDA receptor subunits: more enigmatic than ever? J. Physiol.600, 261–276 (2022). [DOI] [PubMed] [Google Scholar]

[CR57] 57.Kulkarni, S. K. & Dhir, A. Effect of various classes of antidepressants in behavioral paradigms of despair. Prog. Neuropsychopharmacol. Biol. Psychiatry31, 1248–1254 (2007). [DOI] [PubMed] [Google Scholar]

[CR58] 58.Renard, C. E., Dailly, E., David, D. J., Hascoet, M. & Bourin, M. Monoamine metabolism changes following the mouse forced swimming test but not the tail suspension test. Fundam. Clin. Pharm.17, 449–455 (2003). [DOI] [PubMed] [Google Scholar]

[CR59] 59.Chatterjee, M., Jaiswal, M. & Palit, G. Comparative evaluation of forced swim test and tail suspension test as models of negative symptom of schizophrenia in rodents. ISRN Psychiatry2012, 595141 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR60] 60.Lee, J. H., Wei, L., Deveau, T. C., Gu, X. & Yu, S. P. Expression of the NMDA receptor subunit GluN3A (NR3A) in the olfactory system and its regulatory role on olfaction in the adult mouse. Brain Struct. Funct.221, 3259–3273 (2016). [DOI] [PubMed] [Google Scholar]

[CR61] 61.Yao, Y. & Mayer, M. L. Characterization of a soluble ligand binding domain of the NMDA receptor regulatory subunit NR3A. J. Neurosci.26, 4559–4566 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR62] 62.Ragguett, R. M., Rong, C., Kratiuk, K. & McIntyre, R. S. Rapastinel - an investigational NMDA-R modulator for major depressive disorder: evidence to date. Expert Opin. Investig. Drugs28, 113–119 (2019). [DOI] [PubMed] [Google Scholar]

[CR63] 63.Keck, M. E. et al. Reduction of hypothalamic vasopressinergic hyperdrive contributes to clinically relevant behavioral and neuroendocrine effects of chronic paroxetine treatment in a psychopathological rat model. Neuropsychopharmacology28, 235–243 (2003). [DOI] [PubMed] [Google Scholar]

[CR64] 64.Carhart-Harris, R. L. & Nutt, D. J. Serotonin and brain function: a tale of two receptors. J. Psychopharmacol.31, 1091–1120 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR65] 65.Mahar, I., Bambico, F. R., Mechawar, N. & Nobrega, J. N. Stress, serotonin, and hippocampal neurogenesis in relation to depression and antidepressant effects. Neurosci. Biobehav Rev.38, 173–192 (2014). [DOI] [PubMed] [Google Scholar]

[CR66] 66.Wolosker, H. & Balu, D. T. D-Serine as the gatekeeper of NMDA receptor activity: implications for the pharmacologic management of anxiety disorders. Transl. Psychiatry10, 184 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR67] 67.Chen, L. et al. Ezrin-mediated astrocyte-synapse signaling regulates cognitive function via astrocyte morphological changes in fine processes in male mice. Brain Behav. Immun.124, 177–191 (2024). [DOI] [PubMed] [Google Scholar]

[CR68] 68.Gomez-Galan, M., De Bundel, D., Van Eeckhaut, A., Smolders, I. & Lindskog, M. Dysfunctional astrocytic regulation of glutamate transmission in a rat model of depression. Mol. Psychiatry18, 582–594 (2013). [DOI] [PubMed] [Google Scholar]

[CR69] 69.Basu, A. C. et al. Targeted disruption of serine racemase affects glutamatergic neurotransmission and behavior. Mol. Psychiatry14, 719–727 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR70] 70.Levin, R. et al. Behavioral and cognitive effects of the N-methyl-D-aspartate receptor co-agonist D-serine in healthy humans: initial findings. J. Psychiatr. Res.61, 188–195 (2015). [DOI] [PubMed] [Google Scholar]

[CR71] 71.Yoon, S. H., Song, W. S., Chung, G., Kim, S. J. & Kim, M. H. Activity in the dorsal hippocampus-mPFC circuit modulates stress-coping strategies during inescapable stress. Exp. Mol. Med.56, 1921–1935 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR72] 72.Bagot, R. C. et al. Ventral hippocampal afferents to the nucleus accumbens regulate susceptibility to depression. Nat. Commun.6, 7062 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR73] 73.Xu, H. et al. mPFC GABAergic transmission mediated the role of BDNF signaling in cognitive impairment but not anxiety induced by adolescent social stress. Neuropharmacology184, 108412 (2021). [DOI] [PubMed] [Google Scholar]

[CR74] 74.Golden, S. A., Covington, H. E., 3rd, Berton, O. & Russo, S. J. A standardized protocol for repeated social defeat stress in mice. Nat.Protoc.6, 1183-1191 (2011). [DOI] [PMC free article] [PubMed]

[CR75] 75.Menard, C. et al. Social stress induces neurovascular pathology promoting depression. Nat. Neurosci.20, 1752–1760 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR76] 76.Wohleb, E. S. et al. beta-Adrenergic receptor antagonism prevents anxiety-like behavior and microglial reactivity induced by repeated social defeat. J. Neurosci.31, 6277–6288 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR77] 77.Franklin, K. B. J. & Paxinos, G. Paxinos and Franklin’s The mouse brain in stereotaxic coordinates, Academic Press, an imprint of Elsevier, Amsterdam (2013).

[CR78] 78.Goodman, E. J. et al. Sequencing to determine the effect of neuronal IL-1 receptor knockout and microglia depletion on stress-induced changes to the neuronal transcriptome. GEO, V1.https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE253687 (2024).

[CR79] 79.Satija, R., Farrell, J. A., Gennert, D., Schier, A. F. & Regev, A. Spatial reconstruction of single-cell gene expression data. Nat. Biotechnol.33, 495–502 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR80] 80.Habib, N. et al. Div-Seq: Single-nucleus RNA-Seq reveals dynamics of rare adult newborn neurons. Science353, 925–928 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR81] 81.Wickham, H. & Wickham, H. Data analysis, (Springer, 2016).

PERMALINK

Non-classical NMDAR subunit GluN3A in a specialized hippocampal region regulates stress-coping strategies in male mice

Wei Zhang

Mingyue Zhao

Jiajie Dai

Linan Zhuo

Haimou Ye

Jiesi Wang

Weiwen Wang

Abstract

Introduction

Results

Effects of CSDS and GluN3A KO on stress-related behaviors and GluN3A expression in the hippocampus

Fig. 1. Chronic social defeat stress (CSDS) induces multiple stress-related behaviors and specifically reduces GluN3A expression in the intermediate CA1 (CA1i) region, while GluN3A knockout (KO) specifically promotes passive coping.

Overexpression of the GluN3A decline in CA1i neurons reversed passive coping behaviors

Fig. 2. Overexpression of GluN3A in CA1i reverses passive coping behavior in CSDS mice and GluN3A KO mice.

Activity of CA1i pyramidal neurons mediated by GluN3A-dependent regulation of stress-coping behaviors

Fig. 3. Activity of CA1i pyramidal neurons mediates the GluN3A-dependent regulation of passive coping behavior.

Bidirectional chemogenetic modulation of CA1i pyramidal neurons directly mimicked or reversed the passive coping behaviors

Fig. 4. Bidirectional chemogenetic modulation of CA1i pyramidal neurons directly mimics or reverses passive coping behavior.

Potentiation of GluN3A-mediated CA1i pyramidal neuron activity was crucial for the active coping strategies induced by D-serine

Fig. 5. GluN3A-dependent potentiation of CA1i pyramidal neuronal activity is essential for the rapid shift in coping strategy by D-serine.

Projections of CA1i neurons and downstream circuits may be involved in GluN3A-mediated regulation of stress-coping behaviors

Fig. 6. GluN3A-mediated activation of CA1i pyramidal neurons selectively influenced neuronal activity in downstream CA3 and IL circuits.

Discussion

Methods

Animals

CSDS model

Subthreshold SDS

RSD model

Behavioral assessments

Molecular biology test

In vivo manipulation

SnRNAseq analysis

Statistics and reproducibility

Data reporting

Reporting summary

Supplementary information

Acknowledgements

Author contributions

Peer review

Peer review information

Funding

Data availability

Code availability

Competing interests

Footnotes

Contributor Information

Supplementary information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases