Nanobody therapy rescues behavioural deficits of NMDA receptor hypofunction

Abstract

There is an urgent need for efficient and innovative therapies to treat brain disorders such as psychiatric and neurodegenerative diseases. Immunotherapies have proved to be efficient in many medical areas, but have not been considered to treat brain diseases due to the poor brain penetration of immunoglobulins^1,2. Here we developed a bivalent biparatopic antibody, made of two camelid heavy-chain antibodies (called nanobodies)³, one binding to, and the other potentiating the activity of, homodimeric metabotropic glutamate receptor 2. We show that this bivalent nanobody, given peripherally, reaches the brain and corrects cognitive deficits in two preclinical mouse models with endophenotypes resulting from NMDA receptor hypofunction. Notably, these in vivo effects last for at least 7 days after a single intraperitoneal injection and are maintained after subchronic treatment. Our results establish a proof of concept that nanobodies can target brain receptors, and pave the way for nanobody-based therapeutic strategies for the treatment of brain disorders.

Fig. 1. The mGlu2-specific DN13–DN1 nanobody potentiates receptor activation and penetrates the brain.

a, Schematic of DN13 and the bivalent biparatopic DN13–DN1 in which the two nanobodies are linked using the hinge from llama heavy-chain antibodies (HcAb). b, Cartoon of the mGlu2 receptor fused to an extracellular SNAP tag (ST) for labelling and the 6×His-tagged DN13–DN1 binding to the receptor. Intracellular IP1 accumulation is a functional readout of mGlu2 receptor activation. c, Saturation binding experiments of DN13 and DN13–DN1 on cells expressing SNAP–mGlu2 in the presence of a saturating concentration of the agonist LY379268 (1 µM). Data are mean ± s.d. of three individual experiments each performed in quadruplicates. d, Potentiation of IP1 accumulation induced by the indicated nanobodies in the presence of LY379268 (EC₂₀, 0.6 nM). Data are mean ± s.d. of three individual experiments each performed in triplicates. e, Ex vivo immunofluorescence analysis of brain sagittal sections from WT (left) or mGlu2^−/− (right) mice using d2-labelled DN13–DN1. Scale bar, 1.1 mm. White levels were modified uniformly (Zen, Zeiss). Representative slices of two technical replicates from three animals. f, Quantification of the radioactivity present in the mouse brain 4 h, 24 h, 4 days or 7 days after i.p. administration of [³H]DN13–DN1. Data are mean ± s.d. of the indicated number of animals (n). Statistical analysis was performed using Brown–Forsythe and Welch analysis of variance (ANOVA) with Dunnett’s multiple-comparison test compared with the vehicle treatment; *P < 0.05, **P < 0.01, ***P < 0.001. g, Immunohistochemical detection of DN13–DN1 in the ventral tegmental area (VTA) (top) and in the cerebral cortex (bottom) 24 h after i.p. administration of DN13–DN1 (right) versus vehicle (left). Representative slices of four technical replicates from two animals. Scale bar, 50 μm. The diagram in f was partly generated using Servier Medical Art, provided by Servier, under a CC BY 4.0 license. Source Data

Fig. 2. DN13–DN1 nanobody rescues recognition memory in a neurodevelopmental PCP-induced mouse model of schizophrenia.

a, Schematic of the experimental timeline. b, Schematic of the NOR protocol with either a 24 h or 7 day retention period between the training and testing phases (for c and d, respectively). c, The discrimination index was determined using the NOR task 24 h after training in control (black, n = 34) or PCP-treated (green, n = 20) mice additionally treated with vehicle, and in PCP-treated animals additionally treated with DN13–DN1 i.c.v. (4 pmol, 5 μl; 48 h after administration; orange; n = 9) or i.p. (10 mg per kg; 24 h after administration; red; n = 21), the negative control DN1 (4 pmol, 5 μl i.c.v.; brown; n = 13) and the reference mGlu2 agonist (LY379268; 1 mg per kg i.p.; 1 h after the second administration; blue; n = 11). The effect of the mGlu2 antagonist LY341495 (3 mg per kg i.p.) alone (grey; n = 12) or with DN13–DN1 (purple; n = 13) was tested. d, The discrimination index was determined using the NOR task 7 days after training in mice treated with DN13–DN1 i.p. (10 mg per kg; red; n = 24) and LY379268 (1 mg per kg i.p.; blue; n = 10) compared with mice treated with vehicle and PCP (green; n = 23) and mice treated with vehicle control (black; n = 23). For c and d, data are mean ± s.d. Statistical analysis was performed using one-way ANOVA followed by Dunnett’s multiple-comparison test compared with the control mice (black; hash symbols), or compared with the PCP-treated mice (green) or the indicated groups (all other comparisons). NS, not significant; ^####P < 0.0001, ***P < 0.001. Details on statistics and P values are provided as source data. Source Data

Fig. 3. DN13–DN1 nanobody improves working memory and sensory gating of GluN1-KD mice.

a–c, Y-maze experiment. a, Schematic of the experiment. Two independent cohorts of mice were used in the Y-maze to examine the effect of drugs at 3 h (b) and 7 days (c) after acute injection. The colour key in a applies to b, c, e and f. b,c The percentage of spontaneous alternations 3 h (b) or 7 days (c) after treatment in vehicle-treated (black), LY379268-treated (blue) and DN13–DN1-treated (red) mice. None of the studied drugs affected the ambulation of the experimental mice (Extended Data Fig. 4c–e (number of entries)). Data are mean ± s.d. The number of animals is indicated at the bottom of the bar. Statistical analysis was performed using two-way multivariate ANOVA (MANOVA) followed by Tukey’s honest significant difference (HSD) test compared with vehicle-treated WT mice (hash symbols) and compared with vehicle-treated GluN1-KD mice (asterisks); ^###P < 0.0001, ***P < 0.001. d–f, The PPI experiment. d, Schematic of the experiment. One cohort of mice was tested in PPI at 3 h after drug treatment (e) and retested after 7 days (f). e,f, The percentage of PPI was determined in vehicle-treated (black), LY379268-treated (blue) and DN13–DN1-treated (red) mice, assessed 3 h (e) or 7 days (f) after treatment. DN13–DN1 had no effect on the acoustic startle response (ASR; Extended Data Fig. 4f–h). Data are mean ± s.d. The number of animals is indicated above the x axis. Statistical analysis was performed using three-way repeated-measures MANOVA (rm-MANOVA) followed by Tukey’s HSD test compared with vehicle-treated WT mice (hash symbols) and compared with vehicle-treated GluN1-KD mice (asterisks); ^#P < 0.0001, **P < 0.01, ***P < 0.001. Details of the statistical analysis and P values are provided as source data. Source Data

Fig. 4. Absence of acute behavioural effects of DN13–Fc in the two mouse models of schizophrenia.

a, Schematic of DN13–Fc, which is a fusion of two DN13 nanobodies to the human IgG1 Fc domain. b, The discrimination index was determined using the NOR task 24 h after training in vehicle control mice (black; n = 18) or in PCP-treated mice that were additionally treated with vehicle (green; n = 32), DN13–Fc (10 mg per kg, i.p.; 24 h after administration; blue; n = 23) and DN13–DN1 (10 mg per kg, i.p.; 24 h after administration; red; n = 10). Data are mean ± s.d. Statistical analysis was performed using Kruskal–Wallis tests followed by Dunn’s multiple-comparison test compared with the control mice (hash symbols) or compared with the other groups (other comparisons); ##P < 0.01, **P < 0.01. c, The percentage of spontaneous alternations 3 h after treatment in vehicle-treated (black), DN13–Fc-treated (blue) and DN13–DN1-treated (red) WT and GluN1-KD mice. Data are mean ± s.d. The number of animals is indicated at the bottom of the bar. Statistical analysis was performed using two-way MANOVA followed by Tukey’s HSD test compared with vehicle-treated WT mice (hash symbols) and compared with vehicle-treated GluN1-KD mice (asterisks); ^###P ≤ 0.001, ***P < 0.001. d, The percentage of PPI was assessed 3 h after i.p. injection of vehicle-treated (black), DN13–Fc-treated (blue) and DN13–DN1-treated (red) mice. Data are mean ± s.d. The number of animals is indicated above the x axis. Statistical analysis was performed using three-way rm-MANOVA followed by Tukey’s HSD test compared with vehicle-treated WT mice (hash symbols) and compared with vehicle-treated GluN1-KD mice (asterisks); ^#P < 0.05, *P < 0.05, ***P < 0.001. Details of the statistical analysis and P values are provided as source data. Source Data

Fig. 5. Subchronic DN13–DN1 treatment restores behavioural deficits in the two mouse models of schizophrenia.

a, The timeline of the 4-week subchronic DN13–DN1 administration protocol. After an initial dose of 10 mg per kg i.p., three additional administrations at 1 mg per kg i.p. were performed 1 week apart. b, The discrimination index in the NOR task performed 24 h after training in vehicle-treated (black; n = 11) or DN13–DN1-treated (red; n = 7) control mice, and in vehicle-treated (green; n = 21) or DN13–DN1-treated (red; n = 13) PCP-treated mice 2 days after the last injection. Data are mean ± s.d. Statistical analysis was performed using one-way ANOVA followed by Dunnett’s multiple-comparison test compared with PCP-treated mice (bottom); **P < 0.01 compared with PCP-treated mice. c, The percentage of spontaneous alternations was assessed using the Y-maze, 3 h after the last injection of the subchronic treatment in vehicle-treated (black) and DN13–DN1-treated (red) WT and GluN1-KD mice. Data are mean ± s.d. The number of animals is indicated at the bottom of the bar. Statistical analysis was performed using two-way MANOVA followed by Tukey’s HSD test compared with vehicle-treated WT mice (hash symbols) and compared with vehicle-treated GluN1-KD mice (asterisks); ^###P < 0.001, **P < 0.01. d, The percentage of PPI was assessed 3 h after the last injection of the subchronic treatment in vehicle-treated (black) and DN13–DN1-treated (red) WT and GluN1-KD mice. Data are mean ± s.d. The number of animals is indicated above the x axis. Statistical analysis was performed using three-way rm-ANOVA followed by Tukey’s HSD test compared with vehicle-treated WT mice (n = 8) (hash symbols) and compared with vehicle-treated GluN1-KD mice (n = 10) (asterisks); ^##P < 0.01, **P < 0.01, ***P < 0.001. Details of the statistical analysis and P values are provided as source data. The diagram in a was partly generated using Servier Medical Art, provided by Servier, under a CC 4.0 license. Source Data

Extended Data Fig. 1. Comparison of mGlu₂ bivalent nanobodies.

a-b, Schematic representation of the indicated mGlu₂ nanobodies and linkers (a) used to generate a library of mGlu₂ bivalent nanobodies (b). b, Binding affinity on active (left) or inactive (centre) conformation of the mGlu₂ receptor and PAM potency (right) of all the bivalent nanobodies. Binding affinity for mGlu₂ receptor was determined in a FRET-based assay where the active and inactive conformations of the receptor were stabilized with saturating concentrations of LY379268 (1 μM) or LY341495 (1 μM), respectively. The dotted lines indicate the Kd values of DN13 binding (left panel) and of DN1 binding previously reported (middle panel) and the bars highlight the difference from these reference values. The PAM potency was evaluated by measuring the accumulation of IP₁ in the presence of an EC₂₀ concentration of LY379268 (0.6 nM). Each data point represents the mean ± SD of the indicated number (n) of independent experiments performed in triplicates. Statistics are calculated using one-way ANOVA followed by a Dunnett multiple comparison test. * p < 0.05, ** p < 0.01 and *** p < 0.001 compared to DN13. NB, no binding. c, Potentiation of IP₁ accumulation induced by the indicated nanobodies in presence of EC₂₀ of glutamate (5.42 μM). Data are mean ± SD of three individual experiments each performed in triplicates. d, Binding of DN13-DN1 on the different SNAP-mGlu receptor subtypes transfected in HEK293 cells in the presence of a saturating concentration of their respective agonists. Data are mean ± SD of three individual experiments each performed in triplicates. e, Cartoon depicting DN13-DN1 labelling on native lysine residues using a N-hydroxysuccinimide fluorophore (d2-NHS). Saturation binding experiments for the labelled DN13-DN1-d2 on mGlu₂ expressing HEK293 cells, and comparison with nanobodies. Data are mean ± SD of four individual experiments each performed in triplicates. f, Cartoon depicting the bacterial transglutaminase (BTG) derivatization of DN13-DN1 using and CBz-Gln-Gly-Lys-OH peptide under controlled conditions to reach derivatization on a single lysine 121 in the linker HcAb as determined by mass spectrometry. Saturation binding experiments for the DN13-DN1 cold, derivatized as for the tritiated version, and comparison with the other unlabelled nanobodies. Values are a mean ± SD of two experiments each performed in quadruplicate. For details on statistics and p-values, see Source data. Source Data

Extended Data Fig. 2. Pharmacological properties of DN1, DN13, DN13-DN1 and DN13-Fc on different mGlu₂ orthologues.

a, Affinity of DN1, DN13, DN13-DN1 and DN13-Fc for various mGlu₂ orthologues. pKd values determined by binding experiments of the indicated nanobodies-derivatives on HEK293 cells expressing the indicated SNAP-mGlu₂ orthologue in the presence of a saturating concentration of the agonist LY379268 (1 μM) or the antagonist LY341495 (1 μM). Data are mean ± SD of three individual experiments each performed in triplicates. b, PAM potency of DN1, DN13, DN13-DN1 and DN13-Fc for various mGlu₂ orthologues in the presence of an EC₂₀ of agonist. PAM potency (pEC₅₀) values determined in an IP₁ accumulation assay on cells expressing the indicated SNAP-mGlu₂ orthologue. Potentiation was assessed in the presence of EC₂₀ of LY379268 or EC₂₀ of glutamate. Data are mean ± SD of three individual experiments each performed in triplicates. For details on statistics and p-values, see Source data. Source Data

Extended Data Fig. 3. Pharmacokinetics information on mGlu₂ nanobodies.

a, Schematic representation of the FRET-based quantification assay of 6His-tagged nanobodies. Cells expressing Lumi4Tb-labelled mGlu₂ receptors were incubated with a 6His-tagged nanobody in the presence of an anti-6His antibody coupled to the FRET acceptor d2. Under binding of the nanobody, FRET occurs. b, Linearity range of the assay for DN13-DN1 assessed using a known concentration of nanobody. c, Schematic representation of the nanobody quantification in the blood stream. After i.p. administration of the nanobody, the blood was collected over time and the plasma extracted. Samples were then used in the FRET-based quantification assay. d, Time course determination of DN13-DN1 (red) in the blood stream after i.p administration of 10 mg/kg. Each data point is the mean ± s.e.m. from three animals. e, Kinetics of DN13-DN1 (red) and DN13 (black) elimination from the blood stream. Each data point is the mean ± s.e.m. from three animals. f, Representative immunohistochemical labelling with an anti-6His antibody of a sagittal slice from a mouse intraperitoneally administrated with either PBS (top) or 10 mg/kg of DN13-DN1 (bottom) and sacrificed 24 h after administration. DN13-DN1 contains a C-terminal 6His tag. Boxes indicate areas that were magnified in Fig. 1g. Representative slices of four technical replicates from two animals. VTA, ventral tegmental area. g, Autoradiographic signal in mouse brain slices 4 h after i.p. administration of [³H]-DN13-DN1 at 10 mg/kg (3 μCi / mouse). Autoradiographic images were superimposed with transmission images of the same brain slices. Experiment performed in one animal. For details on statistics and p-values, see Source data. Panel a was partly generated using Servier Medical Art, provided by Servier, licensed under a CC 4.0. Source Data

Extended Data Fig. 4. Acute DN13-DN1 treatment does not alter discrimination index of control animal in NOR and ambulation or acoustic startle response in wild-type or GluN1 KD mice.

a, b, Discrimination index in the NOR test performed 24 h (a) or 7 days (b) after the training session. a, Index obtained in control mice (black, n = 34) further treated with DN13-DN1 i.c.v (orange, n = 8) or i.p. (red, n = 12), DN1 i.c.v. (brown, n = 8), mGlu₂ agonist (LY379268, blue, n = 6), mGlu₂ antagonist alone (LY341495, grey, n = 11) or with DN13-DN1 i.p. (purple, n = 13) and in PCP-treated mice (green, n = 20). b, Index obtained in control animals (black, n = 23) further treated with DN13-DN1 i.p. (red, n = 19) and LY379268 (blue, n = 8) and PCP-vehicle animals (green, n = 23). Bars represent means ± SD. Statistics are calculated using one-way ANOVA. ns, non-significant and #### p < 0.0001 compared to control mice. Note that vehicle- (black) and PCP-injected mice (green) are the same as those used in experiments illustrated in Fig. 2. c-e, Y-maze. Number of entries in Y-maze after acute injection at 3 h (d) and 7 days (e) as post-treatment period in vehicle (black), LY379268 (blue) and DN13-DN1 (red)-treated mice. Bars represent mean ± SD. The number of animals is indicated at the bottom of the bar. ### p < 0.0001 in comparison with vehicle-treated WT; ***p < 0.001 in comparison with vehicle-treated GluN1 KD mice in two-way MANOVAs followed by a Tukey HDS test. f-h, Acoustic startle response (ASR) determined on vehicle (black), LY379268 (blue) and DN13-DN1-treated (red) mice assessed 3 h (g) or 7 days (h) after treatments. LY379268 (blue) increased startle response in both WT and GluN1 KD mice after 3 h of treatment but not seven days after treatment. Bars represent mean ± SD. The number of animals is indicated above the X-axis. ### represents p < 0.01 in comparison with vehicle-treated WT mice; *p < 0.05 and **p < 0.01 in comparison with vehicle-treated mice within each genotype in two-way MANOVAs followed by a Tukey HDS test. For details on statistics and p-values, see Source data. Source Data

Extended Data Fig. 5. Locomotor activity measured in the two mouse models after acute or subchronic treatment.

a-b, DN13-DN1 acute treatment (red) does not modify the locomotor activity across 5-min bins (a) or total activity over 60 min (b) of control or PCP-treated mice, using a cyclotron. Data are presented as mean ± s.e.m. (a) or mean ± SD (b) and analysed using the one-way ANOVA followed by a Dunnet post-hoc analysis, ns: non-significant in comparison with vehicle-treated mice c-f, Both DN13-DN1 acute (c, d) and subchronic (e, f) treatment decrease the hyperlocomotor activity of GluN1 KD mice assessed in an open-field using the travelled distance across 5-min bins (c, e) or total distance (d, f). Data are presented as mean ± s.e.m. (c, e) or mean ± SD (d, f). ### p < 0.001 and #### p < 0.0001 in comparison with vehicle-treated WT; ***p < 0.001 in comparison with vehicle-treated GluN1 KD mice; ns: non-significant; using two-way MANOVAs followed by a Tukey HDS test. The number of animals is indicated above the X-axis. For details on statistics and p-values, see Source data. Source Data

Extended Data Fig. 6. mGlu₂ relative expression measured after acute treatment with DN13-DN1 in the two mouse models of schizophrenia.

a, Schematic representation of the FRET-based assay for quantification of mGlu₂ receptors expression using DN1 nanobody coupled to a donor (DN1-Tb) and DN10 coupled to an acceptor (DN10-d2). b, No significant differences were observed on the relative quantification of mGlu₂ receptors in the cortex, hippocampus, striatum, cerebellum and midbrain of control and PCP-treated mice injected with either LY379268 (1 mg/kg, light blue), DN13-DN1 (10 mg/kg, red), DN13-Fc (10 mg/kg, dark blue), or with vehicle (black). c, No variation in the mGlu₂ relative expression was also observed between WT and GluN1 KD mice in the prefrontal cortex, cortex, hippocampus and midbrain, after an acute treatment with DN13-DN1 (10 mg/kg, red), DN13-Fc (10 mg/kg, blue) or with vehicle (black). The signal indicates the slope values of the relative FRET linear range in the quantification experiments. Data are presented as mean ± SD and analysed using the one-way ANOVA followed by a Dunnet post-hoc analysis. The number of animals is indicated above the X-axis. For details on statistics and p-values, see Source data. Parts of panel a were created in Biorender. Lafon, P. (2025) https://BioRender.com/w77j520. Source Data

Extended Data Fig. 7. Absence of sex-dependency of both acute and subchronic DN13-DN1 positive behavioural effect.

Animals used in the experiments illustrated in Fig. 2c (a), Fig. 4b (b), Fig. 4c,d (d), Fig. 5b (c) and Fig. 5c,d (e) were separated according to their sex and analysed for potential sex difference. No significant differences were observed when comparing males (M) and females (F) for the neurodevelopmental model (a-c) and GluN1 KD mice (d,e), upon treatment with DN13-DN1 (acute or subchronic), DN13-Fc, LY379268 or vehicle. Statistics were calculated using unpaired Student’s t-test or Mann-Whitney test (a-c) or three-way MANOVA (Y-maze) or three-way rm-MANOVA (PPI) followed by a Tukey HDS test (d-e), ns: non-significant. The number of animals is indicated above the X-axis. All data are mean ± SD. For details on statistics and p-values, see Source data. Source Data

Extended Data Fig. 8. Absence of body weight influence on PPI results and absence of DN13-DN1 effect on locomotion and cataleptic behaviour in control mice.

a, Recapitulation of body weight of animals used in experiments illustrated in Figs. 3, 4 and 5. *** p < 0.001 in comparison with PBS-treated males within genotype; ^### p < 0.001 in comparison with PBS-treated WT mice. b, Pearson’s correlations among body weight, percentage of pre-pulse inhibition (PPI, %) at various pre-pulses and acoustic startle response (ASR) across all experiments. No correlation was found between body weight and PPI, however, a negative association was found between body weight and ASR, attributed to the increased startle response in GluN1 KD mice with lower body weight. c, Effects of DN13-DN1 acute treatment (10 mg/kg) on motor coordination and balance (inclined platform and rotarod), and cataleptic behaviour (pinch-induced catalepsy) in WT mice. Data are presented as mean ± s.e.m. and analysed by a two-way MANOVA (a) or three-way MANOVA (c) followed by a Tukey HDS test and Pearson’s correlation (b). The numbers of animals are indicated in the panels. For more details on statistics, see Source data. Source Data

Extended Data Fig. 9. Acute DN13-Fc and subchronic DN13-DN1 treatment do not alter discrimination index of control animals in NOR and ambulation or acoustic startle response in wild-type or GluN1 KD mice.

a, Discrimination index in the NOR test performed 24 h after the training session. DN13-DN1 and DN13-Fc were administered 3 h before the training session, respectively. Neither DN13-DN1 (10 mg/kg, i.p., red) nor DN13-Fc (10 mg/kg, i.p., blue) modified the discrimination score compared to vehicle treated animal (black). Bars represent means ± SD. Statistics were calculated using Kruskal-Wallis test followed by a Dunn’s multiple comparison test, ns: non-significant and ### p < 0.001 compared to control mice. Note that vehicle- (black) and PCP-injected mice (green) are the same as those used in experiments illustrated in Fig. 5b. b, Effect of acute administration of DN13-Fc (10 mg/kg, i.p., blue) or DN13-DN1 (10 mg/kg, i.p., red). Left panel. Number of entries in Y-maze: DN13-DN1 (red) but not DN13-Fc (blue) reduced ambulation in Y-maze. Right panel. Acoustic startle response (ASR) was not altered by treatment. Bars represent mean ± SD. c, Effect of subchronic administration of DN13-DN1 (10 mg/kg, 3x 1 mg/kg, i.p., red) on number of entries in Y-maze (left panel) or ASR (right panel). Bars represent mean ± SD. In b,c, ### p < 0.01 in comparison with vehicle-treated WT mice; ** p < 0.01 in comparison with vehicle-treated mice within each genotype in two-way MANOVA followed by a Tukey HDS test. In a,b,c the numbers of animals are indicated in the panels. For details on statistics and p-values see Source data. Source Data

Extended Data Fig. 10. Low LPS contamination of DN13-DN1 is not responsible for nanobody brain-penetration or behavioural effect.

a, Evaluation of LPS amount injected into animals during DN13-DN1 administration. The content of LPS in each nanobody sample was assessed by mass spectrometry (n = 3 biological samples). Along DN13-DN1 purified in the standard protocol (Immobilized-metal affinity chromatography – IMAC) administration, around 0.25 mg/kg LPS are injected into animals which is below the threshold reported to induce toxicity (0.5 mg/kg). A further size exclusion chromatography (SEC) purification step reduces this amount to 0.002 mg/kg. Results in DN13-DN1 nanobody with almost no endotoxin (below 0.19 μg/mg nanobody) still have an acute effect in the genetic mouse model of schizophrenia. b, Y-maze. Percentage of spontaneous alternations and number of entries 3 h after the acute injection of 10 mg/kg (i.p.) of DN13-DN1 either purified in one-step IMAC or purified in two steps IMAC followed by SEC. No difference was observed between DN13-DN1 samples, and both samples significantly corrected the working memory deficit in GluN1 KD mice. Bars represent mean ± SD. ### p < 0.001 in comparison with vehicle-treated WT mice; ***p < 0.001 in comparison with vehicle-treated GluN1 KD mice in two-way MANOVA followed by a Tukey HDS test. The numbers of animals are indicated in the panels. For details on statistics and p-values, see Source data. Source Data

Extended Data Fig. 11. mGlu₂ relative expression measured after subchronic treatment with DN13-DN1 in the two mouse models of schizophrenia.

a, No significant differences were observed on the relative quantification of mGlu₂ receptors in the prefrontal cortex, cortex, hippocampus, striatum, cerebellum and midbrain of control and PCP-treated mice treated with DN13-DN1 (red) or vehicle (black) subchronic protocol detailed in Fig. 5a (10 mg/kg, 3 ×1 mg/kg). b, No variation in the mGlu₂ relative expression was also observed between subchronic administration of DN13-DN1 (10 mg/kg, 3 ×1 mg/kg, red) or vehicle (black) in both WT and GluN1 KD mice. The signal indicates the slope values of the relative FRET linear range in the quantification experiments. Data are presented as mean ± SD and analysed using the one-way ANOVA followed by a Dunnet post-hoc analysis. The number of animals is indicated above the X-axis. For details on statistics and p-values, see Source data. Source Data