Mice that carry a heterozygous, autism spectrum disorder-risk C456Y mutation in the NMDA receptor (NMDAR) subunit GluN2B show decreased protein levels, hippocampal NMDAR currents, and NMDAR-dependent long-term depression and have abnormal anxiolytic-like behavior. Early, but not late, treatment of the young mice with the NMDAR agonist D-cycloserine rescues these phenotypes.
Autism spectrum disorders (ASDs) are neurodevelopmental disorders characterized by social deficits and repetitive behaviors. Although a large number of ASD-risk mutations have been reported , the mechanisms underlying ASD remain largely unclear. An emerging ASD-related mechanism is dysfunction of N-methyl-D-aspartate (NMDA) receptors (NMDARs) , a key, multisubunit regulator of brain development and function that is subject to various forms of receptor modulation [3–7]. Many known ASD-risk genetic variants have been shown to cause NMDAR dysfunction in animal models of ASD  that are causally associated with ASD-like abnormal behaviors [8–10]. However, a better animal model of ASD for the NMDAR dysfunction hypothesis would presumably be one carrying mutations in Glutamate receptor, ionotropic, NMDA1 (GRIN1); GRIN2A; or GRIN2B genes encoding the main NMDAR subunits GluN1/NR1, GluN2A/NR2A, and GluN2B/NR2B, respectively.
Among known NMDAR subunit genes, GRIN2B is one of the most frequently mutated ASD-risk genes, belonging to category 1 in the Simons Foundation Autism Research Initiative (SFARI) gene list, and shows stronger impacts on ASD than mutations in GRIN1 or GRIN2A [1,11–16]. In addition to ASD, GRIN2B has been extensively associated with various neurodevelopmental disorders, including developmental delay, intellectual disability, attention-deficit/hyperactivity disorder, epilepsy, schizophrenia, obsessive-compulsive disorder, and encephalopathy [5,15,17,18].
In line with the strong involvement of GRIN2B in diverse brain disorders, mice carrying a conventional homozygous deletion of Grin2b display impaired suckling, neonatal death during postnatal day (P) 1–3, and impaired hippocampal long-term depression (LTD) in neonates . Similarly, a homozygous truncation of the intracellular C-terminal region of GluN2B causes perinatal lethality in mice . These early studies were followed by those restricting homozygous Grin2b deletion to specific cell types and developmental stages to circumvent the strong developmental impacts of Grin2b deletion, which revealed the important roles of GluN2B in the regulation of long-term potentiation (LTP), LTD, and cognitive behaviors [21–23]. Notably, an early study investigated mice with heterozygous (not homozygous) Grin2b deletion and reported impaired LTP at the mutant hippocampal fimbrial-CA3 synapses , although associated behaviors were not investigated. Conversely, Grin2b overexpression has been shown to enhance LTP and learning and memory in mice . These results suggest that GluN2B is important for normal brain development, synaptic plasticity, and cognitive behaviors.
However, dissimilar to the previous studies on Grin2b mice largely analyzing the synaptic and behavioral impacts of a homozygous Grin2b deletion, GRIN2B mutations identified in human brain disorders are preponderantly heterozygous mutations, and synaptic and behavioral phenotypes of heterozygous Grin2b-mutant mice remain largely unexplored. In addition, human GRIN2B mutations are often missense mutations that induce a single amino acid change in the encoded protein, again distinct from the null or truncation mutations previously studied in mice. Although many of the missense mutations of NMDAR subunits have been characterized in vitro , their in vivo impacts have been minimally studied.
In the present study, we generated and characterized a knock-in mouse line carrying an ASD-risk mutation (GluN2B-C456Y) in the Grin2b gene, a de novo mutation identified in a male individual with ASD and intellectual disability . These heterozygous GluN2B-C456Y mutant mice (Grin2b+/C456Y) showed substantial decreases in GluN2B protein levels, suggestive of mutation-induced protein degradation in vivo. Currents of GluN2B-containing NMDARs and NMDAR-dependent LTD (but not LTP) were also decreased, revealing sensitivity of LTD to GluN2B haploinsufficiency. Behaviorally, these mice showed normal social interaction but enhanced anxiety-like behavior in pups and contrasting anxiolytic-like behavior in juveniles and adults. These synaptic and behavioral effects were largely mimicked by an independent mouse line carrying a conventional heterozygous Grin2b deletion (Grin2b+/–). Importantly, early, but not late, treatment of young mice (P7–16) with the NMDAR agonist D-cycloserine normalized NMDAR currents and LTD in juvenile Grin2b+/C456Y mice and improved anxiolytic-like behavior in adult Grin2b+/C456Y mice, supporting the emerging concept in the field of neurodevelopmental and neuropsychiatric disorders that early and timely correction of key pathophysiological deficits is important for efficient and long-lasting beneficial effects.
Structural analysis has suggested that a missense mutation in the GRIN2B gene leading to a C456Y mutation in the GluN2B subunit of NMDARs disrupts a disulfide bond within a loop residing at the interface between the amino-terminal domain (ATD) and ligand-binding domain (LBD) . In addition, experiments using Xenopus oocytes and human embryonic kidney 293 (HEK-293) cells have shown that the GluN2B-C456Y mutation induces multiple changes in the GluN2B protein, including protein degradation, limited surface trafficking, and gating alterations of GluN2B-containing NMDARs .
Our own structural investigation and functional characterization of the GluN2B-C456Y protein using Xenopus oocytes yielded overall similar results. Specifically, structural modeling, based on the known structure of GluN1/GluN2B NMDARs [28–30], revealed that the GluN2B-C456Y mutation in the LBD region alters the structure of a large loop protruding from the GluN2B LBD by disrupting the formation of an intraloop disulfide bond (S1A and S1B Fig). This loop, which is absent in alpha-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and kainate receptors, makes extensive intra- and intersubunit interactions within neighboring domains pointing to an important role in receptor assembly [28–30]. It is therefore likely that the GluN2B-C456Y mutation alters the receptor quaternary structure and in turn its function.
Diheteromers produced by coexpressing GluN2B-C456Y and wild-type (WT) GluN1 yielded NMDAR currents that were <1% of the currents produced by WT GluN1/GluN2B diheteromers (S2A Fig), thus revealing a strong expression phenotype, as previously observed . Functional characterization of the small-amplitude mutant receptor currents showed an approximately 30% increase in receptor channel maximal open probability, as assessed by MK-801 inhibition kinetics , likely due to a decreased inhibition by ambient protons, as indicated by full proton dose-response curves (S2B and S2C Fig). In agreement with a decreased pH sensitivity, potentiation by spermine, a GluN2B-specific positive allosteric modulator , was also significantly reduced (S2D Fig). In addition, sensitivity to glycine was decreased, whereas sensitivities to glutamate and zinc, an endogenous allosteric inhibitor of NMDARs , were minimally affected (S2E–S2G Fig). Overall, these results indicate that the C456Y mutation in GluN2B drastically reduces expression and alter channel functions of recombinant NMDARs most likely because of GluN2B misfolding and degradation.
To explore the impacts of the C456Y mutation on the stability or function of GluN2B in mice, we generated a knock-in mouse line carrying the C456Y mutation in the Grin2b gene (Fig 1A; S3A and S3B Fig). Homozygous C456Y-mutant (Grin2bC456Y/C456Y) mice showed neonatal death at P7 (approximately 0% survival), similar to the case for mice with a conventional homozygous Grin2b deletion [19,20]. In contrast, our heterozygous mutant (Grin2b+/C456Y) mice were produced with the expected mendelian ratios and showed normal survival and growth.
Immunoblot analyses of whole-brain lysates and crude synaptosomal fractions from Grin2b+/C456Y mice at embryonic day 20 (E20) and several postnatal stages (P14, P21, P28, and P56) indicated approximately 30%–50% reductions in the levels of GluN2B protein (Fig 1B; S4A Fig). Notably, the levels of GluN1, but not GluN2A, were also reduced, although to a lesser extent than those of GluN2B, indicating that the stability of GluN1 strongly depends on GluN2B, whereas that of GluN2A does not. However, the C456Y mutation had no effect on mRNA levels of Grin2b or Grin1 (encoding GluN1) (S4B Fig). These results indicate that the GluN2B-C456Y mutation induces a strong reduction in the levels of the GluN2B protein as well as a concomitant reduction in GluN1 protein levels in vivo, without affecting their mRNA levels.
We next tested whether the GluN2B-C456Y mutation affects NMDAR currents in the hippocampus of Grin2b+/C456Y mice. This mutation caused significant decreases in the ratio of NMDAR/AMPA receptor (AMPAR)-mediated evoked excitatory postsynaptic currents (EPSCs), tau of NMDAR current decay, and the amount of ifenprodil-sensitive current of GluN2B-containing NMDARs at Schaffer collateral-CA1 pyramidal (SC-CA1) synapses (Fig 1C and 1D). In contrast, Grin2b+/C456Y SC-CA1 synapses showed a normal input-output relationship of evoked EPSCs and paired-pulse facilitation (Fig 1E and 1F), indicative of normal AMPAR-mediated basal excitatory synaptic transmission and presynaptic release. These results suggest that the GluN2B-C456Y mutation selectively suppresses currents of GluN2B-containing NMDARs at hippocampal SC-CA1 synapses.
Previous studies on Grin2b -mutant mice have demonstrated the critical roles of GluN2B in the regulation of synaptic plasticity such as LTP and LTD [19–24], although the majority of these studies used mice carrying homozygous Grin2b deletion.
However, these observations are not congruent with human cases of GRIN2B mutations and related brain dysfunctions, in which heterozygous GRIN2B mutations are prevalent [1,11–17]. Thus, whether Grin2b haploinsufficiency in Grin2b -mutant mice would affect various forms of hippocampal synaptic plasticity or other synaptic functions is an important question that needs to be addressed. This question becomes more complicated when we consider the juvenile and adult stages, when both GluN2B and GluN2A are expressed and contribute to the formation of multiple forms of NMDARs with different subunit compositions, including diheteromeric (1/2A or 1/2B) and triheteromeric (1/2A/2B) NMDAR complexes [5,34,35].
To address this question, we measured several forms of synaptic plasticity in addition to low-frequency stimulation (LFS)-LTD, including LTP induced by high-frequency stimulation (HFS-LTP), theta burst stimulation–induced LTP (TBS-LTP), and metabotropic glutamate receptor (mGluR)-dependent LTD (mGluR-LTD) in the CA1 region of the Grin2b+/C456Y hippocampus at juvenile stages (P16–33).
The GluN2B-C456Y mutation reduced LFS-LTD by about 55% at Grin2b+/C456Y SC-CA1 synapses compared with WT mice (Fig 2A), a result similar to that obtained in neonatal mice with a conventional homozygous Grin2b deletion . A similar decrease (approximately 84%) in LFS-LTD was observed in the prelimbic layer 1 region of the medial prefrontal cortex (mPFC) (Fig 2B). This result provides genetic evidence that LFS-LTD in the hippocampus is sensitive to Grin2b haploinsufficiency.
In contrast, the GluN2B-C456Y mutation had no effect on mGluR-LTD induced by the group I mGluR agonist dihydroxyphenylglycine (DHPG) at SC-CA1 synapses of Grin2b+/C456Y mice (Fig 2C). It also had no effect on HFS-LTP or TBS-LTP at Grin2b+/C456Y SC-CA1 synapses (Fig 2D and 2E). These results suggest that the heterozygous C456Y mutation and consequent decreases in GluN2B protein levels, GluN2B-dependent NMDAR currents, and LFS-LTD have no effect on other forms of synaptic plasticity in the hippocampus.
Because LTD is implicated in the regulation of synapse shrinkage and pruning , we attempted an electron microscopic (EM) analysis to see whether Grin2b+/C456Y mice display altered density or morphology of excitatory synapses. However, there were no genotype differences in the density and morphology (length, depth, and perforation [a measure of maturation]) of postsynaptic density (PSD) structures in the CA1 region of the WT and Grin2b+/C456Y hippocampus (P21) (Fig 2F), electron-dense multiprotein complexes at excitatory postsynaptic sites [37,38], suggesting that a moderate (approximately 50%) decrease in LTD does not induce morphological changes of excitatory synapses.
A previous study employed single-neuron gene deletion to show that GluN2A and GluN2B distinctly regulate the number and strength of functional excitatory synapses . In addition, GluN2B is expressed in GABAergic interneurons  and NMDARs can function at presynaptic sites . It is therefore possible that a heterozygous GluN2B-C456Y mutation might influence synaptic features unrelated to synaptic plasticity, such as synapse development and spontaneous synaptic transmission, at both excitatory and inhibitory synapses. In addition, mutations expected to mainly affect excitatory synapses are frequently associated with changes in intrinsic neuronal properties, such as neuronal excitability , suggesting that the GluN2B-C456Y mutation might also affect neuronal properties. To test these possibilities, we first measured spontaneous synaptic transmission at excitatory and inhibitory Grin2b+/C456Y synapses.
The frequency and amplitude of miniature EPSCs (mEPSCs) and miniature inhibitory postsynaptic currents (mIPSCs) did not differ in CA1 pyramidal neurons in the hippocampus of Grin2b+/C456Y mice compared with those of WT animals (S5A and S5B Fig), suggestive of normal development and efficacy of excitatory and inhibitory synapses. Moreover, there were no differences between genotypes in spontaneous EPSCs (sEPSCs) or spontaneous IPSCs (sIPSCs), measured in the absence of tetrodotoxin to allow network activities (S5C and S5D Fig), suggesting that excitatory network activity is unaltered in the hippocampus of Grin2b+/C456Y mice. We also measured the ratio of evoked EPSCs and IPSCs in the CA1 hippocampal region and found no genotype difference (S5E Fig).
In addition to spontaneous and evoked synaptic transmission, neural excitability was unaltered in Grin2b+/C456Y CA1 pyramidal neurons, as shown by current-firing curves (S5F Fig). These results collectively suggest that, in contrast to its effects on LFS-LTD, the heterozygous GluN2B-C456Y mutation does not affect neuronal excitability or excitatory or inhibitory synapse development or function in the hippocampus in the presence or absence of network activity.
To explore behavioral impacts of the GluN2B-C456Y mutation, we subjected Grin2b+/C456Y mice to a battery of behavioral tests. Adult male Grin2b+/C456Y mice displayed hypoactivity in the open-field test (Fig 3A and 3B) but spent normal amounts of time in the center region of the open-field arena, indicative of largely normal anxiety-like behavior (Fig 3C and 3D). These mice, however, displayed anxiolytic-like behavior during the first 10 minutes in the arena, likely reflecting a modified response to a novel environment.
In the elevated plus-maze, Grin2b+/C456Y mice also showed anxiolytic-like behavior, as shown by both the number of entries into and time spent in closed/open arms (Fig 3E–3H). In contrast, these mice showed normal anxiety-like behaviors in the light-dark test, as shown by time spent in the light chamber (Fig 3I). These results suggest that Grin2b+/C456Y mice display anxiolytic-like behavior in the elevated plus-maze.
In tests measuring learning and memory, Grin2b+/C456Y mice showed normal levels of learning and memory in the learning and probe phases of both initial- and reversal-learning sessions of the Morris water maze (S6A–S6E Fig). In addition, they showed a normal preference for a novel object over a familiar object in the novel object–recognition test (S6F Fig).
Contrary to our expectation, Grin2b+/C456Y mice showed largely normal social behaviors, including social approach and social novelty recognition in the three-chamber test ; social interaction between freely moving mice in the direct social-interaction test; and ultrasonic vocalizations (USVs), a form of social communication in rodents, upon encountering a female (S7A–S7N Fig) [42–44].
Furthermore, these mice showed enhanced self-grooming (but normal digging) in home cages with bedding but showed no repetitive self-grooming in a novel chamber without bedding (S7O–S7Q Fig), indicative of a moderate increase in self-grooming. These results collectively suggest that the GluN2B-C456Y mutation leads to hypoactivity, anxiolytic-like behavior, and moderately enhanced self-grooming, without affecting social interaction, social communication, or learning and memory in mice.
We next employed an independent mouse line carrying a conventional heterozygous Grin2b deletion (Grin2b+/–) to see whether the behavioral phenotypes observed in Grin2b+/C456Y mice could be reproduced. This mouse line has been used previously to demonstrate that a homozygous null Grin2b mutation entirely eliminates GluN2B protein and causes severe phenotypes, including impaired suckling and neonatal death .
Grin2b+/– mice showed decreased (approximately 50%) whole-brain levels of the GluN2B, but not GluN1 or GluN2A, subunit of NMDARs at P14 and P21 (S8A and S8B Fig), partly similar to the results from Grin2b+/C456Y mice in which both GluN2B and GluN1 levels were decreased (P14 and P21). Behaviorally, adult male Grin2b+/–mice showed phenotypes that were largely similar to those observed in adult male Grin2b+/C456Y mice, including hypoactivity and moderately anxiolytic-like behavior (S8C–S8K Fig), as well as normal social interaction and communication and object-recognition memory (S9A–S9K and S9N Fig). Unlike Grin2b+/C456Y mice, which showed modestly enhanced self-grooming, Grin2b+/– mice showed no repetitive self-grooming (S9L and S9M Fig).
These results indicate that the heterozygous C456Y mutation and conventional Grin2b heterozygosis lead to similar, although not identical, biochemical and behavioral phenotypes and suggest that the phenotypes observed in Grin2b+/C456Y mice are likely consequences of the loss (not gain) of GluN2B function. The small differences in the behaviors of the two mouse lines may reflect minor effects attributable to the specific mutation/deletion in the Grin2b gene.
ASD is characterized by early onset of core and comorbid symptoms. When Grin2b+/C456Y pups (P4–12) were tested for the emission of USVs upon mother separation, a measure of anxiety in rodents responsive to anxiolytic medications , these pups showed strongly enhanced USVs, as determined by the total number of USV calls, duration of each USV calls, and latency to first calls (Fig 3J–3L). This suggests that Grin2b+/C456Y pups display anxiety-like behaviors, similar to the anxiety symptoms comorbid with human ASD .
When Grin2b+/C456Y juveniles (P18–26) were subjected to a battery of behavioral tests, they displayed hypoactivity, similar to adult mice, and, notably, strong anxiolytic-like behavior in the center region of the open-field arena (Fig 3M–3P). However, upon mother separation and reunification in a mother-homing test, Grin2b+/C456Y juveniles spent normal amounts of time with the reunited mothers (S10A and S10B Fig), suggestive of normal anxiety-like behaviors. Therefore, anxiety-like behavior in Grin2b+/C456Y pups seems to be strongly weakened or changed into anxiolytic-like behavior at juvenile stages, similar to the anxiolytic-like behaviors in adults.
Grin2b+/C456Y juveniles showed normal social interaction, as shown by the juvenile play test (S10C Fig), similar to adult mice. In addition, these mice showed normal self-grooming and digging in home cages with bedding (S10D and S10E Fig), partly dissimilar to the adult mutant mice that show enhanced self-grooming but normal digging in home cages with bedding. These results suggest that self-grooming in Grin2b+/C456Y mice develops slowly in late life after the juvenile stage.
The reduced NMDAR function and LTD observed in young (2–3-week-old) Grin2b+/C456Y mice might be associated with the behavioral abnormalities (hypoactivity and anxiolytic-like behavior) observed in adult (2–4-month-old) Grin2b+/C456Y mice. This hypothesis could be tested by normalizing the reduced NMDAR function and NMDAR-dependent LTD in early stages and examining whether these corrections are associated with behavioral rescues at late stages. To this end, we used D-cycloserine, a partial agonist at the glycine-binding site of NMDARs with increasing potential for the treatment of neurological and neuropsychiatric disorders .
We first tested whether the reduced LFS-LTD is attributable to decreased NMDAR currents in Grin2b+/C456Y mice. In hippocampal slices from young Grin2b+/C456Y mice, application of D-cycloserine (10 μM), which can still activate mutant GluN2B-C456Y receptors (S2H Fig), fully normalized the reduced LFS-LTD at Grin2b+/C456Y SC-CA1 synapses, without affecting WT synapses (Fig 4A). These results suggest that abnormal NMDAR currents are associated with reduced LFS-LTD at Grin2b+/C456Y hippocampal SC-CA1 synapses in juvenile mice.
We next attempted early, chronic treatment of young Grin2b+/C456Y mice with D-cycloserine (40 mg/kg), administered orally twice daily for 10 days (P7–16), followed by measurements of NMDA/AMPA ratio, paired-pulse facilitation, and LFS-LTD in juvenile mice (P17–21) and behavioral experiments in adult mice (>P56) (Fig 4B). Early D-cycloserine treatment fully normalized the reduced NMDA/AMPA ratio and LFS-LTD at Grin2b+/C456Y SC-CA1 synapses in juvenile mice (P16–21), without affecting paired-pulse facilitation (Fig 4C–4E). WT synapses were unaffected in these measurements.
In addition, early D-cycloserine treatment improved anxiolytic-like behavior in Grin2b+/C456Y mice in the elevated plus-maze test, without affecting WT mice (Fig 4F–4I). In contrast, D-cycloserine had no effect on hypoactivity in Grin2b+/C456Y mice (Fig 4J–4M). We could not test whether early D-cycloserine treatment (P7–16) could affect pup USVs in Grin2b+/C456Y mice because the time window for the treatment fell after the early time window for pup USV (P4–12).
When Grin2b+/–mice carrying conventional heterozygous Grin2b deletion were tested for D-cycloserine-dependent rescue of synaptic and behavioral deficits, early chronic D-cycloserine treatment (P7–16, 40 mg/kg, twice daily for 10 days; oral) normalized the decreased LFS-LTD at Grin2b+/C456Y SC-CA1 synapses, without affecting WT synapses (S11A and S11B Fig). Early D-cycloserine treatment also had no effect on the hypoactivity of Grin2b+/– mice (S11C Fig), similar to the results from Grin2b+/C456Y mice. We could not determine whether D-cycloserine has any effect on the anxiety-like behavior in Grin2b+/– mice (S11C Fig) because the anxiolytic-like behavior in Grin2b+/–mice was weaker than that in Grin2b+/C456Y mice, and the small baseline difference between WT and Grin2b+/–mice became insignificant (between vehicle-treated WT and Grin2b+/–mice) by the chronic drug treatment procedures (oral, twice/day for 10 days).
Grin2b+/C456Y mice display decreased GluN2B levels at an adult stage (P56), suggesting that the continuing decrease in GluN2B levels in adult mutant mice, in addition to the reduced LFS-LTD in young mutant mice, might be associated with anxiety-like behavior or hypoactivity. To test this, we attempted to enhance NMDAR function in adult Grin2b+/C456Y mice by treating with D-cycloserine. In these experiments, we first used an acute treatment paradigm because acute D-cycloserine treatment has been previously shown to rescue ASD-like behaviors in many mouse models of ASD [9,10,48–50].
In contrast to early chronic treatment, late acute D-cycloserine treatment (20 mg/kg; intraperitoneal [i.p.]) in the adult stage had no effect on anxiolytic-like behavior or hypoactivity in Grin2b+/C456Y mice in open-field or elevated plus-maze tests (Fig 5A–5I); it also had no effect on WT mice. In addition, late chronic D-cycloserine treatment (P57–66, 40 mg/kg, twice/day, oral) had no effect on the anxiolytic-like behavior or hypoactivity in Grin2b+/C456Y mice (S12 Fig). However, the late chronic drug treatment procedures using a restrainer substantially blunted the baseline differences in elevated plus-maze variables (but not locomotion) between vehicle-treated WT and mutant mice, making it difficult to assess the drug effects on these values. These results collectively suggest that late treatment of Grin2b+/C456Y mice with D-cycloserine to enhance NMDAR function has no effect on anxiolytic-like behavior or hypoactivity, highlighting the importance of early treatments.
In this study, we demonstrated that mice carrying a heterozygous C456Y mutation in the GluN2B subunit of NMDARs, an ASD-risk mutation in humans, exhibit decreased GluN2B and GluN1 protein levels, diminished currents of GluN2B-containing NMDARs, and reduced LFS-LTD. This mutation also induced anxiolytic-like behavior that can be corrected by early, but not late, D-cycloserine treatment that restores NMDAR function and NMDAR-dependent LTD.
A key finding in our study is that the GluN2B-C456Y mutation induces substantial degradation of the GluN2B protein in mice. This conclusion is supported by the measurement of GluN2B protein levels in the Grin2b+/C456Y brain at various developmental stages. A previous study on multiple GluN2A/B mutations using structural analysis and oocyte/HEK cell experiments reported similar findings on the impact of the GluN2B-C456Y mutation . Interestingly, a similar expression phenotype was observed for the patient-derived GluN2A-C436R mutation, which also disrupts a disulfide bond within LBD loop 1 . Our study extends these previous findings by providing in vivo evidence of the importance of the C456Y mutation and the proper folding of LBD loop 1 for GluN2B protein levels.
Our results further reveal an impact of reduced GluN2B protein levels on GluN1 protein levels, although the magnitude of this latter decrease was less than that of GluN2B. This further supports the previously reported importance of GluN2B in the maintenance of normal levels of GluN1 [21,23]. This decrease in the GluN1 subunit in our study does not seem to involve changes in Grin1 mRNA levels. It may occur because the reduction in GluN2B protein levels may lead to a situation in which some GluN1 proteins that can no longer associate with GluN2B to form heteromeric NMDAR complexes become destabilized and degraded. It is possible that some GluN1 proteins may fail to associate with mutant GluN2B proteins beginning in the endoplasmic reticulum  and are degraded via the ubiquitin-proteasomal pathway following retrograde transport to the cytoplasm [51–54]. Alternatively, the two proteins may initially associate with each other and reach the plasma membrane surface and synaptic sites but gradually dissociate from one another, leaving GluN1 subject to endocytosis and degradation through the late endosomal-lysosomal pathway involving a conserved membrane-proximal signal present in GluN1 [55,56].
Complicating the situation is the fact that GluN2B can form a triheteromeric complex with GluN1 and GluN2A [5,34,35,57,58] that is known to be the major NMDAR population in the adult hippocampus [39,59,60]. Although further details remain to be elucidated, the concomitant reduction in GluN1 levels creates a situation in which GluN1 protein is produced normally but not used.
In addition, given that diheteromic and triheteromeric NMDAR complexes display distinct biophysical and pharmacological properties in different spatiotemporal contexts [35,61–63], the reduced levels of GluN2B and GluN1 in Grin2b+/C456Y mice would affect both diheteromic and triheteromeric NMDAR complexes differentially in different brain regions, cell types, and developmental stages.
Another key finding of our study is the reduced NMDAR-dependent LFS-LTD by about 50% at Grin2b+/C456Y hippocampal SC-CA1 synapses. Previous studies on genetic Grin2b deletion and its impacts on LTD found near-complete impairments of LTD in neonate mice (P1–3) carrying a conventional homozygous Grin2b deletion  and in adult mice (14–22 weeks) carrying conditional homozygous Grin2b deletion restricted to Ca2+ /calmodulin dependent protein kinase II (CaMKII)-positive principal neurons in the cortex and hippocampal CA1 region . The design of our Grin2b-mutant mouse study differs from those of the previous studies in the following respects: (1) use of a patient-derived knock-in mutation rather than a conventional, or conditional, gene deletion; (2) use of heterozygous instead of homozygous mutant mice (an early study on heterozygous Grin2b mice examined only LTP but not LTD ); and (3) analysis of LTD at a juvenile stage rather than at a neonatal or adult stage. In addition, our results indicate that the heterozygous GluN2B-C456Y mutation has no effect on other synaptic and neuronal variables, such as spontaneous synaptic transmission in CA1 neurons (mEPSCs, mIPSCs, sEPSCs, sIPSCs), basal transmission at SC-CA1 synapses (evoked EPSCs), the ratio of evoked IPSCs and EPSCs, and intrinsic neuronal excitability of CA1 neurons. Moreover, Grin2b+/C456Y hippocampal SC-CA1 synapses displayed normal HFS-LTP, TBS-LTP, or mGluR-LTD. The lack of changes in LTP (HFS and TBS) measured during a late juvenile stage (P27–33) could be because the postnatal switch from GluN2B to GluN2A by neuronal activity may be largely complete at this stage . Together, these results support the established notion that genetic deletion of Grin2b suppresses LTD and extends it by demonstrating that the patient-derived heterozygous GluN2B-C456Y mutation induces a selective reduction in LTD by approximately 50% in juvenile mice.
A straightforward mechanism underlying the decreased LFS-LTD at Grin2b+/C456Y hippocampal SC-CA1 synapses would be decreased GluN2B function. Previous studies employing pharmacological inhibitors of GluN2B, however, yielded conflicting results, with their significant effects on LTD [64–67] or insignificant effects on LTD [68,69] (reviewed in ). This difference could be attributable to multiple factors, including the limited selectivity of the GluN2B inhibitors , differential actions of GluN2B inhibitors on di- and triheteromeric NMDARs [61,62], and influences of GluN2B inhibitors on glutamate dissociation rate [39,72,73].
Notably, a recent study employing single-neuron gene knockout (KO) has reported that GluN2A or GluN2B is not critically required for ionotropic or non-ionotropic (not involving NMDAR-dependent ion flow [74–76]) NMDAR-dependent LTD, whereas GluN1 is required for non-ionotropic NMDAR-dependent LTD . It is thus possible that the reduced levels of GluN1 in Grin2b+/C456Y mice may contribute to the reduced LTD at SC-CA1 synapses. However, the previous single-neuron KO study employed AAV-dependent gene KO at the mouse age of P0–1, leaving GluN2B expression and function at embryonic stages unaffected. In addition, the single-neuron KO study would lead to homozygous (not heterozygous) Grin2b deletion, which might also affect the results.
Grin2b+/C456Y mice showed moderately enhanced self-grooming, a core ASD-like behavior, in home cages with bedding, but normal self-grooming in a novel chamber without bedding, suggesting that these mice display moderately enhanced self-grooming that is suppressed by a novel environment. Moreover, enhanced self-grooming in home cages with bedding was observed in adult but not juvenile Grin2b+/C456Y mice, suggesting that repetitive self-grooming develops late in life in Grin2b+/C456Y mice and thus is unlikely to be ameliorated by early D-cycloserine treatment.
Contrary to our expectations, Grin2b+/C456Y mice showed normal social approach, social novelty recognition, and social interaction in three-chamber and direct social-interaction tests. In addition, these animals showed normal social communication (USVs) during courtship. Juvenile Grin2b+/C456Y mice also displayed normal social interaction in the juvenile play test and spent normal amounts of time with reunited mothers. It is possible, however, that the social tests and variables that we employed in the present study may not be sensitive enough to detect certain social deficits.
Both adult and juvenile Grin2b+/C456Y mice showed hypoactivity in the open-field test, suggesting that this phenotype is established early (in juvenile or earlier stages) and persist into adulthood. Adult Grin2b+/C456Y mice show anxiolytic-like behavior in the elevated plus-maze test but normal anxiety-like behaviors in open-field and light-dark tests. Juvenile Grin2b+/C456Y mice show anxiolytic-like behavior in the open-field test, suggesting that adult and juvenile Grin2b+/C456Y mice show normal or anxiolytic-like behaviors. In contrast, Grin2b+/C456Y pups display strongly increased USV calls upon mother separation, suggestive of anxiety-like behavior. Therefore, the anxiety-like behavior of Grin2b+/C456Y pups seems to be rapidly weakened as these mice grow up, whereas self-grooming slowly develops at an adult stage, pointing to the contrasting trajectories of two important ASD-related phenotypes (anxiety-like behavior and self-grooming). How the early anxiety-like behavior in Grin2b+/C456Y pups is weakened or reversed as the pups grow into juveniles and adults remain unclear. This age-dependent reversal might reflect compensatory changes trying to overcome the over-activation of anxiety-related neural circuits. Although further details remain to be determined, our results are in line with the fact that anxiety is one of the key comorbidities of ASD [46,78] and that many mouse models of ASD display anxiety-like behaviors [42,79,80]. Notably, anxiolytic-like behavior has also been observed in mice lacking oxytocin , implicated in ASD .
Grin2b+/–mice (carrying conventional heterozygous Grin2b deletion) mice showed largely similar behaviors compared with those observed in Grin2b+/C456Y mice. Similar behaviors include hypoactivity in the open-field test and normal anxiety-like behavior in open-field and light-dark tests, but anxiolytic-like behavior of Grin2b+/–mice in the elevated plus-maze was much weaker than that in Grin2b+/C456Y mice. Biochemically, Grin2b+/–mice showed decreased levels of GluN2B but not GluN1 at P14 and P21, unlike the concomitant decreases in GluN2B and GluN1 in Grin2b+/C456Y mice, which may lead to subtle differences in synaptic and behavioral dysfunctions in these two mouse lines.
The behavioral phenotypes of Grin2b+/C456Y mice could not be compared with the symptoms of the human individual carrying GluN2B-C456Y mutation as they were minimally described in the previous study other than the fact that the mutation is a de novo mutation from a male individual with autism and intellectual disability . However, the abnormal behaviors (i.e., anxiolytic-like behavior) of the mutant mice are important biologically because spending more time in the center region of a novel open-field arena or in the open arms of the elevated plus-maze reflects behaviors that would pose a significant threat for the survival of a mouse in its natural environment, implicating substantial deficits in cognitive functions. The anxiolytic-like behavior may not reflect increased fear of a dark or closed environment, because these mice exhibited a normal preference for light and dark chambers in the light-dark test. In addition, the anxiolytic-like behavior of the mutant mice in the elevated plus-maze does not seem to involve suppressed cognition of the fact that a darker and closed place is generally safe at least based on their normal learning and memory in Morris water-maze and novel object–recognition tests.
Importantly, our study suggests synaptic mechanisms that may be associated with the anxiolytic-like behavior, namely suppressed NMDAR functions and LFS-LTD at an early stage. In support of this hypothesis, early chronic D-cycloserine treatment of young Grin2b+/C456Y mice normalizes NMDAR function and LTD in juvenile Grin2b+/C456Y mice and anxiolytic-like behavior in adult Grin2b+/C456Y mice. In addition, late acute treatment of adult Grin2b+/C456Y mice with D-cycloserine has no effect on abnormal behaviors probably because GluN2B expression is decreased at adult stages, and NMDAR-dependent LTD is difficult to induce at adult stages likely due to the switch of GluN2B to GluN2A [5,83,84]. The effect of late chronic D-cycloserine treatment in adult Grin2b+/C456Y mice could not be tested because the chronic treatment procedure seemed to increase anxiety levels in these mice, blunting the baseline difference between WT and mutant mice. In addition, we could not test whether the increased USV calls in the mutant pups are associated with the reduced NMDAR function because the time window for early D-cycloserine treatment (P7–16) fell behind that for pup USV testing (P4–12). Together, these results suggest that early correction of NMDAR function and NMDAR-dependent LTD in young mice leads to long-lasting improvement of anxiolytic-like behavior in adult mice. Early treatment seems to be particularly important, not only because it has long-lasting effects, eliminating the necessity of repeated drug administration, but also because the small time window during which treatment is efficient appears to occur only during early developmental stages. Indeed, LTD is known to be most prominent during an early period (approximately 2–3 weeks) of postnatal brain development in mice and becomes weaker as the brain progressively matures and the ratio of GluN2B/GluN2A expression decreases [34,85–87].
Our findings are in line with the emerging concept that early and timely correction of key pathophysiological deficits in young mice is critical for the long-lasting and efficient rescue of synaptic and behavioral phenotypes in adult mice. For instance, early chronic fluoxetine treatment to restore reduced serotonin levels in young mice carrying a 15q11-13 duplication, a human ASD-risk mutation, has been shown to induce long-lasting normalization of serotonin levels and abnormal behaviors in adult mice . Similarly, Shank2 -mutant mice show increased NMDAR function (in contrast to decreased NMDAR function in later stages) , and early, but not late, chronic memantine treatment to suppress the abnormal NMDAR hyperfunction improves late synaptic and social phenotypes in Shank2 -mutant mice .
Our data indicate that LFS-LTD are similarly decreased in the hippocampus and mPFC. These results suggest that the decreased NMDAR function and LFS-LTD in the hippocampus may represent a proxy for changes occurring in other brain regions and that decreased NMDAR function and LFS-LTD in many brain areas, additional to the hippocampus, could contribute to the behavioral changes (i.e., anxiolytic-like behavior) observed in Grin2b+/C456Y mice. In line with this idea, Grin2b is widely expressed in the brain , and anxiety-like behavior has been associated with various brain regions, including the hippocampus, anterior cingulate cortex, lateral septum, bed nuclei of the stria terminalis, paraventricular nucleus, and basolateral amygdala [90–97].
Lastly, GluN2B-C456Y is a strong ASD-risk mutation . How might the abnormal synaptic and behavioral phenotypes of Grin2b+/C456Y mice be related to ASD pathophysiology? GluN2B-containing NMDARs that mediate large calcium influx are strongly expressed during early brain developmental stages to promote synapse and neuronal maturation through mechanisms, including posttranslational modification and gene expression . Therefore, the decreased levels of GluN2B and GluN2B-containing NMDARs in Grin2b+/C456Y mice would suppress these critical molecular and cellular early events. In addition, NMDAR-dependent LTD during early brain development is well known to sharpen neuronal circuits by promoting weakening of less active synapses and strengthening of more active synapses through redistribution of synaptic protein resources between these synapses [83,84]. Therefore, reduced LTD in the developing brain of young Grin2b+/C456Y mice would suppress LTD-dependent synapse-pruning and circuit-sharpening processes, leading to brain malfunctions and abnormal behaviors. This prediction, based on in vivo results, might apply not only to GluN2B-C456Y-related cases of ASD [1,11–16] but also to various GRIN2B-related brain dysfunctions, including developmental delay, intellectual disability, attention-deficit/hyperactivity disorder, epilepsy, schizophrenia, obsessive-compulsive disorder, and encephalopathy [15,17]. How these predicted changes manifest into synaptic and circuit properties in the mutant brain remains to be determined. Previous studies have shown that NMDAR antagonists, including the GluN2B-specific antagonist ifenprodil, can induce anxiolytic-like behaviors in both humans and experimental animals . However, a decrease in NMDAR function in the adult mutant brain is an unlikely possibility because both acute and chronic D-cycloserine treatment failed to rescue the anxiolytic-like behavior in adult Grin2b+/C456Y mice.
In conclusion, the heterozygous ASD-risk mutation, GluN2B-C456Y, leads to decreased GluN2B protein levels, diminished currents of GluN2B-containing NMDARs, and reduced NMDAR-dependent LTD in young mice, as well as abnormal, anxiolytic-like behavior in adult mice. In addition, early D-cycloserine treatment of young mutant mice correcting NMDAR function and NMDAR-dependent LTD leads to long-lasting improvement of anxiolytic-like behaviors in adult mice.
All animals were bred and maintained according to the Requirements of Animal Research at KAIST and all procedures were approved by the Committees of Animal Research at KAIST (KA2016-31).
Grin2b knock-in mice under the genetic background of C57BL/6J carrying C456Y mutation in exon 6 with Frt sites and cassette were designed and generated by Biocytogen (Grin2b+/cassette , S1A Fig). To remove the neomycin cassette, Grin2b+/cassette mice were crossed with Protamine-Flp mice (C57BL/6J), which yielded floxed heterozygote mice (Grin2b+/C456Y).
Statistical analyses were performed using Prism GraphPad 7 and SigmaPlot 11. The data with nonparametric distribution were analyzed by Mann-Whitney test, and those with parametric distribution were analyzed by Student t test. If the data are parametric but have significant difference in variance in the F-test, Welch’s correction was used. Including gender, age, and number of mice, all the details of the statistical analyses are described in S1 Data.
Details on other methods, including those for experiments on recombinant GluN1/GluN2B receptors, can be found in S2 Data.
The numerical data used in all figures can be found in S3 Data.
The original images for blots and gels can be found in S4 Data.
We would like to thank Dr. Yeonseung Chung in the Department of Mathematical Sciences at KAIST for help with the statistical analyses.
autism spectrum disorder
Ca2+/calmodulin dependent protein kinase II
excitatory postsynaptic current
field excitatory postsynaptic potential
Glutamate receptor, ionotropic, NMDA1
Glutamate receptor, ionotropic, NMDA2B
human embryonic kidney 293
inhibitory postsynaptic current
metabotropic glutamate receptor
medial prefrontal cortex
Schaffer collateral-CA1 pyramidal
Simons Foundation Autism Research Initiative
theta burst stimulation
27 Feb 2020
Thank you for submitting your revised manuscript entitled "Early correction of synaptic long-term depression improves abnormal anxiety-like behavior in adult GluN2B-C456Y-mutant mice" for consideration as a Research Article by PLOS Biology.
Your revision has now been evaluated by the PLOS Biology editorial staff, as well as by the original Academic Editor, and I am writing to let you know that we would like to send your submission out for external peer review.
However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.
Please re-submit your manuscript within two working days, i.e. by Mar 02 2020 11:59PM.
Login to Editorial Manager here: https://www.editorialmanager.com/pbiology
During resubmission, you will be invited to opt-in to posting your pre-review manuscript as a bioRxiv preprint. Visit http://journals.plos.org/plosbiology/s/preprints for full details. If you consent to posting your current manuscript as a preprint, please upload a single Preprint PDF when you re-submit.
Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.
Feel free to email us at firstname.lastname@example.org if you have any queries relating to your submission.
Gabriel Gasque, Ph.D.,
26 Mar 2020
Thank you for submitting your revised Research Article entitled "Early correction of synaptic long-term depression improves abnormal anxiety-like behavior in adult GluN2B-C456Y-mutant mice" for publication in PLOS Biology. I have now obtained advice from the original reviewers and have discussed their comments with the Academic Editor.
Based on the reviews, we will probably accept this manuscript for publication, assuming that you will modify the manuscript to address the remaining points raised by reviewer 2. Please also make sure to address the data and other policy-related requests noted at the end of this email.
We expect to receive your revised manuscript within two weeks. Your revisions should address the specific points made by reviewer 2. Please submit the following files along with your revised manuscript:
1. A 'Response to Reviewers' file - this should detail your responses to the editorial requests, present a point-by-point response to all of the reviewers' comments, and indicate the changes made to the manuscript.
*NOTE: In your point by point response to the reviewers, please provide the full context of each review. Do not selectively quote paragraphs or sentences to reply to. The entire set of reviewer comments should be present in full and each specific point should be responded to individually, point by point.
2. In addition to a clean copy of the manuscript, please also upload a 'track-changes' version of your manuscript that specifies the edits made. This should be uploaded as a "Related" file type.
In addition to the remaining revisions and before we will be able to formally accept your manuscript and consider it "in press", we also need to ensure that your article conforms to our guidelines. A member of our team will be in touch shortly with a set of requests. As we can't proceed until these requirements are met, your swift response will help prevent delays to publication.
Upon acceptance of your article, your final files will be copyedited and typeset into the final PDF. While you will have an opportunity to review these files as proofs, PLOS will only permit corrections to spelling or significant scientific errors. Therefore, please take this final revision time to assess and make any remaining major changes to your manuscript.
NOTE: If Supporting Information files are included with your article, note that these are not copyedited and will be published as they are submitted. Please ensure that these files are legible and of high quality (at least 300 dpi) in an easily accessible file format. For this reason, please be aware that any references listed in an SI file will not be indexed. For more information, see our Supporting Information guidelines:
*Published Peer Review History*
Please note that you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out. Please see here for more details:
Please note that an uncorrected proof of your manuscript will be published online ahead of the final version, unless you opted out when submitting your manuscript. If, for any reason, you do not want an earlier version of your manuscript published online, uncheck the box. Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us as soon as possible if you or your institution is planning to press release the article.
To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosbiology/s/submission-guidelines#loc-materials-and-methods
*Submitting Your Revision*
To submit your revision, please go to https://www.editorialmanager.com/pbiology/ and log in as an Author. Click the link labelled 'Submissions Needing Revision' to find your submission record. Your revised submission must include a cover letter, a Response to Reviewers file that provides a detailed response to the reviewers' comments (if applicable), and a track-changes file indicating any changes that you have made to the manuscript.
Please do not hesitate to contact me should you have any questions.
Gabriel Gasque, Ph.D.,
--Please update your S3 Data file to include Figure S2C.
-- Please also ensure that figure legends in your manuscript include information on where the underlying data can be found, and ensure your supplemental data file/s has a legend.
BLOT AND GEL REPORTING REQUIREMENTS:
For manuscripts submitted on or after 1st July 2019, we require the original, uncropped and minimally adjusted images supporting all blot and gel results reported in an article's figures or Supporting Information files. We will require these files before a manuscript can be accepted so please prepare and upload them now. Please carefully read our guidelines for how to prepare and upload this data: https://journals.plos.org/plosbiology/s/figures#loc-blot-and-gel-reporting-requirements
Reviewer #1: The authors have adequately addressed the majority of this reviewer's concerns. With the significant changes made to the manuscript during the revision, the paper is suitable for publication.
Reviewer #2, Marc Fuccillo: The authors have attempted to address many of my requests. In doing so, they have strengthened the idea that the C456Y point mutant functions like the NR2B heterozygous LOF mutation, both at the synaptic, protein and behavioral level. While perhaps expected, these data provide clarity to the role of ASD-associated NMDAR mutations. Another strong plus is the addition of data showing a similar plasticity effect in the mPFC. these data suggest that the data in hippo are a proxy for other areas (something that should be mentioned in the text).
Overall, I feel like this is a strong contribution to the asd pathophysiology literature. while not incredibly surprising, the data is clear and makes a strong point supporting this specific disease-associated mutation as GRIN2B haploinsufficiency. the reproducibility of DCS rescue across assays is also a strength.
A few short textual things should be added:
1. discuss the relationship between anxiogenic and anxiolytic behaviors as they relate to ASD. these are not part of the core behavioral phenotype but are the most clearly altered in this work. the literature on ASD and anxiety-related behaviors should be discussed.
2. please discuss that in light of the mPFC data, it is unclear which brain regions are contributing to the mutation-associated behavioral (and rescue-related) changes.
3. "suppressed recognition of the fact that a dark and closed place is generally safe" - this isn't any better than before - perhaps just describe the phenotype.
Reviewer #3: The authors performed extra experiments and gave explanations to address all the questions raised by the reviewers. The manuscript became even more data heavy and the conclusions are now better discussed to support the authors ideas. Although there are some technical issues, the authors do demonstrate GluN2B-C456Y haploinsufficiency decreases GluN2B protein levels, LTD, and anxiety-like behavior. The overall behavioral and physiological experiments provide an insight for the importance of early correction of pathophysiological deficits.
The findings are important in the field for the treatment of neurodevelopmental or psychiatric diseases caused by NMDAR mutations.
Reviewer #4: Satisfied with the manuscript as edited. Again, it is an important contribution to the literature as I stated in my initial review, though there are not major new insights here. Addition of the PFC data at least provides some additional evidence for more general deficits.
15 Apr 2020
Dear Dr Kim,
On behalf of my colleagues and the Academic Editor, Thomas C Südhof, I am pleased to inform you that we will be delighted to publish your Research Article in PLOS Biology.
The files will now enter our production system. You will receive a copyedited version of the manuscript, along with your figures for a final review. You will be given two business days to review and approve the copyedit. Then, within a week, you will receive a PDF proof of your typeset article. You will have two days to review the PDF and make any final corrections. If there is a chance that you'll be unavailable during the copy editing/proof review period, please provide us with contact details of one of the other authors whom you nominate to handle these stages on your behalf. This will ensure that any requested corrections reach the production department in time for publication.
The version of your manuscript submitted at the copyedit stage will be posted online ahead of the final proof version, unless you have already opted out of the process. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.
We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with email@example.com. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.
We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.
Thank you again for submitting your manuscript to PLOS Biology and for your support of Open Access publishing. Please do not hesitate to contact me if I can provide any assistance during the production process.
on behalf of