ABSTRACT Inborn errors of metabolism (IEMs) are common causes of neurodevelopmental disorders, including microcephaly, hyperactivity, and intellectual disability. However, the synaptic

mechanisms of and pharmacological interventions for the neurological complications of most IEMs are unclear. Here, we report that metabolic dysfunction perturbs neuronal NMDA receptor

(NMDAR) homeostasis and that the restoration of NMDAR signaling ameliorates neurodevelopmental and cognitive deficits in IEM model mice that lack aminopeptidase P1. Aminopeptidase

P1-deficient (Xpnpep1–/–) mice, with a disruption of the proline-specific metalloprotease gene _Xpnpep1_, exhibit hippocampal neurodegeneration, behavioral hyperactivity, and impaired

administration of memantine reversed hyperactivity in adult Xpnpep1–/– mice without improving learning and memory. Furthermore, chronic administration of memantine ameliorated hippocampal

neurodegeneration, hyperactivity, and impaired learning and memory in Xpnpep1–/– mice. In addition, abnormally enhanced NMDAR-dependent LTP and NMDAR downstream signaling in the hippocampi

of Xpnpep1–/– mice were reversed by chronic memantine treatment. These results suggest that the metabolic dysfunction caused by aminopeptidase P1 deficiency leads to synaptic dysfunction

with excessive NMDAR activity, and the restoration of synaptic function may be a potential therapeutic strategy for the treatment of neurological complications related to IEMs. SIMILAR

CONTENT BEING VIEWED BY OTHERS ALTERED HIPPOCAMPAL GENE EXPRESSION, GLIAL CELL POPULATION, AND NEURONAL EXCITABILITY IN AMINOPEPTIDASE P1 DEFICIENCY Article Open access 13 January 2021

ACTIVATION OF D2-LIKE DOPAMINE RECEPTORS IMPROVES THE NEURONAL NETWORK AND COGNITIVE FUNCTION OF PPT1KI MICE Article 16 September 2024 ROLE OF PHOSPHODIESTERASES IN THE PATHOPHYSIOLOGY OF

NEURODEVELOPMENTAL DISORDERS Article Open access 07 January 2021 INTRODUCTION Inborn errors of metabolism (IEMs), also known as inheritable metabolic diseases, are caused mainly by mutations

in a single gene that encodes an enzyme in a specific metabolic pathway1. Although an individual IEM is rare (incidence < 1:100,000) owing largely to a recessive inheritance pattern,

IEMs are collectively common disorders with an incidence of 1:800–2500 births and account for more than 15% of single-gene disorders2,3,4. To date, more than 1000 distinct IEMs have been

identified5, and the most serious and common outcomes in IEMs are neurodevelopmental disorders, such as developmental delay, microcephaly, hyperactivity, attention deficit, autism, and

intellectual disability6,7,8,9,10,11. The neural circuit mechanisms underlying neurological complications in most IEMs are currently unknown. Therefore, there are no pharmacological

treatments for neurodevelopmental disorders associated with IEMs. Although causal therapy, dietary restrictions or the supplementation of enzyme cofactors, improves clinical outcomes in some

IEM patients, the therapeutic effect on neurodevelopmental disorders is often limited6,7,12,13,14. Moreover, dietary restriction is ineffective for IEMs in which harmful metabolites are

generated by endogenous sources, and these treatments are not indicated for the majority of IEMs, for which the exact biochemical basis of the disease is unknown. Pharmacological

interventions to restore neural circuits may therefore have broad utility in the treatment of neurological disorders that result from various IEMs. However, an understanding of the altered

neural circuitry in each IEM is essential for pharmacological intervention. In this respect, animal models provide valuable opportunities for the investigation of disease mechanisms in

IEMs15,16. Aminopeptidase P1 deficiency is an IEM caused by mutations in the _Xpnpep1_ gene. As aminopeptidase P1 is a widely distributed metallopeptidase that cleaves the first residue from

peptides containing a penultimate proline in various tissues17,18, the lack of aminopeptidase P1 activity results in massive urinary excretion of undigested peptides containing a

penultimate proline in both humans and mice19,20. In addition, neurodevelopmental disorders, such as developmental delay, microcephaly, and epilepsy, have been observed in patients with

aminopeptidase P1 deficiency19,20. We previously reported that aminopeptidase P1 is predominantly expressed in neurons, compared to glial cells, in the hippocampus, and a disruption of

aminopeptidase P1 in mice results in neurodegeneration in the hippocampal CA3 area, hyperactivity, and impaired hippocampus-dependent learning and memory21,22. However, the identity and

biochemical actions of undigested imino-oligo peptides responsible for neurological complications in Xpnpep1–/– mice are unknown. Despite this, we hypothesized that the characterization and

pharmacological restoration of altered neural circuitry would reverse neurological symptoms in the mice. This approach may provide an opportunity to develop more effective treatments for

neurological complications in various IEMs as well as valuable insight into the pathological mechanism of IEMs. In this study, we found that the metabolic dysfunction in aminopeptidase P1

deficiency perturbs NMDAR homeostasis in brain neurons, thereby leading to synaptopathy in the hippocampi of Xpnpep1–/– mice. In addition, chronic treatment with memantine, an NMDAR

antagonist approved by the US Food and Drug Administration (FDA) for the treatment of Alzheimer’s disease23, improved neurological defects in Xpnpep1–/– mice at the cellular and behavioral

levels. These observations indicate that neurological complications in IEMs are treatable by pharmacological intervention, and the restoration of neural circuitry may be an effective

treatment for neurological symptoms in patients with IEMs. MATERIALS AND METHODS ANIMALS The generation of Xpnpep1 mutant mice and the genotyping of the Xpnpep1 allele have been previously

described20. Mice were backcrossed with two different inbred strains, C57BL/6J and 129S4/SvJae, for 8–16 generations before use. All experiments were performed on age-matched pairs of

Xpnpep1+/+ and Xpnpep1–/– mice generated by intercrossing C57BL/6J and 129S4/SvJae heterozygous parents. Animals were housed 4–5 per cage in an animal facility and maintained in a

climate-controlled room with free access to food and water under a 12-h/12-h light/dark cycle (lights on at 7:00 AM). Animal maintenance and experiments were conducted in accordance with the

guidelines of and approved by the Institutional Animal Care and Use Committee at Seoul National University. HISTOLOGY Histochemical analyses were performed as described previously21,22.

Briefly, mice were deeply anesthetized with a mixture of Zoletil (50 mg/kg, intraperitoneally [i.p.]) and xylazine (1 mg/kg, i.p.), transcardially perfused with heparinized (10 U/ml)

phosphate-buffered saline (PBS), and fixed with 4% (w/v) paraformaldehyde in PBS. Mouse brains were postfixed in the same fixative for 48 h at 4 °C and cut into 60 μm coronal sections using

a vibratome (VT1200S, Leica, Germany). The sections were postfixed in the same fixative for 1 h and permeabilized with 0.3% (v/v) Triton X-100 in PBS for 3 h. The sections were incubated in

blocking buffer (5% normal goat serum, 5% horse serum, 5% donkey serum, and 0.5% BSA in PBS) for 2 h, incubated overnight at 4 °C with primary antibodies, and incubated with Cy3- or

fluorescein isothiocyanate (FITC)-conjugated secondary antibodies for 2 h. After each step, the sections were rinsed three times for 10 min with PBS. Images were acquired using a confocal

laser scanning microscope (LSM510, Zeiss, Germany). For double immunohistochemical/X-gal staining, formalin-fixed mouse brains were postfixed for 12 h at 4 °C and cut into 100 μm-thick

sections using a vibratome. The sections were incubated in X-gal staining solution (5 mM K3Fe(CN)6, 5 mM K4Fe(CN)6, 2 mM MgCl2, 0.01% deoxycholate, 0.02% NP-40, and 1 mg/mL X-gal in PBS) at

37 °C for 5–8 h and then postfixed for 1 h at 4 °C. The sections were rinsed, permeabilized, blocked, and immunostained with primary and secondary antibodies as described above. For

hematoxylin and eosin (HE) staining, fixed brains were embedded in paraffin using an embedding module (Shandon Histocentre 3, Thermo Scientific, USA) and dissected into 4 μm sections using a

rotary microtome (RM2145, Leica), and sections were mounted on glass slides. The sections were deparaffinized and rehydrated by immersing successively in xylene (three times for 10 min),

100% ethanol (twice for 5 min), 95% ethanol (twice for 5 min), and running tap water for 5 min. The sections were then stained with hematoxylin (Merck) and eosin (Sigma–Aldrich). The stained

sections were successively rinsed with 95% ethanol, 100% ethanol, and xylene. Images were acquired using a light microscope (BX-51, Olympus) with a digital imaging system (DFC280, Leica).

ANTIBODIES AND WESTERN BLOTTING GluA1 and GluA2 antibodies have been previously described24. The following antibodies were purchased commercially: PSD-95 (Thermo Scientific, MA1-045),

synapsin I (Chemicon, AB1543), GluN1 (BD Biosciences, 556308), GluN2A (BD Biosciences, 612286), GluN2B (BD Biosciences, 610416), VGLUT1 (Synaptic Systems, 135 303), α-tubulin (Sigma, T5168),

NeuN (Millipore, ABN78), MAP2 (Sigma, M9942), p-CaMKII (Abcam, ab32678), and Calpain-1 (Cell Signaling Technology, 2556S). For western blotting, mouse forebrains or hippocampi were

homogenized in homogenization buffer (320 mM sucrose, 10 mM Tris-HCl, 5 mM EDTA, pH 7.4) containing a protease inhibitor cocktail (Sigma–Aldrich, MO, USA, Cat. # P8340) and a phosphatase

inhibitor cocktail (GenDEPOT, TX, USA, Cat. # P3200). Homogenates were separated using sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and proteins were transferred to

nitrocellulose membranes. After protein transfer, the membranes were incubated in blocking buffer [5% skim milk in Tris-buffered saline with 0.1% Tween 20 (TBST)] for 30 min at room

temperature and then successively were incubated with primary antibodies and horseradish peroxidase (HRP)-conjugated secondary antibodies (Jackson ImmunoResearch, West Grove, USA). After

each step, the membranes were rinsed three times for 10 min with TBST. The HRP signals were developed using enhanced chemiluminescence (GE Healthcare, UK) and detected by exposing the

membrane to X-ray film. Western blot signals were quantified using MetaMorph software (Molecular Devices). SLICE ELECTROPHYSIOLOGY Electrophysiological recordings from hippocampal slices

were performed as previously described22. Hippocampal sections (400 μm) from 5-week-old mice of both sexes were prepared using a vibratome (Leica, Germany) in ice-cold dissection buffer (230

 mM sucrose; 25 mM NaHCO3; 2.5 mM KCl; 1.25 mM NaH2PO4; 10 mM D-glucose; 1.3 mM Na-ascorbate; 3 mM MgCl2; 0.5 mM CaCl2, pH 7.4 with 95% O2/5% CO2). The slices were recovered for at least 1 h

at 36 °C in an aerated (95% O2, 5% CO2) artificial cerebrospinal fluid (ACSF) solution (125 mM NaCl, 26 mM NaHCO3, 2.5 mM KCl, 1.25 mM NaH2PO4, 1.3 mM MgCl2, 2.5 mM CaCl2, 10 mM D-glucose)

and then maintained at room temperature. All electrophysiological experiments were performed in a submerged-type recording chamber, which was perfused with heated (29–30 °C) ACSF. The

signals were filtered at 2.8 kHz and digitized at 10 kHz using a MultiClamp 700B amplifier and a Digidata 1440 A interface (Molecular Devices, CA, USA). During the whole-cell patch clamp

recording, the series resistance (< 10 MΩ) and seal resistance (> 1 GΩ) were monitored by applying a short (50 ms) hyperpolarization voltage pulse (−5 mV), and the data were discarded

if the resistance changed by more than 20%. AMPA receptor (AMPAR)-mediated miniature excitatory postsynaptic currents (mEPSCs) were recorded at –70 mV using a pipette (3–4 MΩ) solution

containing 100 mM CsMeSO4, 10 mM TEA-Cl, 8 mM NaCl, 10 mM HEPES, 5 mM QX-314-Cl, 2 mM Mg-ATP, 0.3 mM Na-GTP, and 10 mM EGTA (pH adjusted to 7.25 with CsOH, 290 mOsm) in the presence of

picrotoxin (50 μM), AP-5 (50 μM), and TTX (0.5 μM) in ACSF. NMDAR-mEPSCs in normal ACSF were recorded at a holding potential of +40 mV (Fig. 1). Picrotoxin, TTX, NBQX (10 μM), and MPEP (10 

μM) were added to the bathing solution to inhibit IPSCs, Na+ channels, AMPARs, and subtype 5 metabotropic glutamate receptors (mGluR5s), respectively. To determine the effect of memantine on

NMDAR activity, NMDAR-mEPSCs were measured at –70 mV, and slices were perfused with Mg2+-free ACSF to allow the activation of NMDARs (Fig. 4). For mIPSC recordings at –70 mV, CsMeSO4 in the

pipette solution was replaced with equimolar CsCl, and NBQX, AP-5, and TTX were added to the normal ACSF. To measure NMDA/AMPA ratios, synaptic responses were evoked by stimulating Schaffer

collaterals with a broken glass pipette (0.3–0.5 MΩ) filled with ACSF. EPSCs were recorded using the same pipette solution used for the measurement of mEPSCs, and stimulation intensity was

adjusted to obtain AMPAR-EPSCs with peak amplitudes of 100–250 pA at –70 mV. After recording stable AMPAR-EPSCs for at least 10 min, the AMPAR inhibitor NBQX (10 μM) was added to ACSF, and

NMDAR-EPSCs in the same neuron were isolated by +40 mV depolarization. During the AMPAR- and NMDAR-EPSC recordings, picrotoxin (10 μM) was included in the ACSF. Field excitatory postsynaptic

potentials (fEPSPs) were recorded using an ACSF-filled recording pipette placed in the stratum radiatum. Schaffer collaterals were stimulated with an ACSF-filled broken glass pipette

(0.3–0.5 MΩ), and the stimulation intensity was adjusted to produce one-third of the maximal synaptic responses. LTP was induced by four episodes of theta burst stimulation (TBS) with

interepisode intervals of 10 s. An episode of TBS consisted of ten stimulus trains at 5 Hz, with each stimulus train consisting of four pulses at 100 Hz. Low-frequency stimulation (LFS)

consists of 900 stimuli at 1 Hz. Slices displaying an unstable (>10%) baseline (20 min) or changes in the fiber volley were discarded. All electrophysiology data were analyzed using

Clampfit (Molecular Devices, USA) and custom macros written in Igor Pro (WaveMetrics). All chemicals were purchased from Sigma–Aldrich (USA), except for picrotoxin, NBQX, AP-5, MPEP, and

DHPG, which were purchased from Tocris (UK). BEHAVIOR ANALYSES The open-field test, novel object recognition (NOR) test, and contextual fear conditioning test were performed, as previously

described21, in male mice between 11:00 AM and 6:00 PM. The object location memory (OLM) test was performed using female mice. Animals were transferred to the behavior testing room for at

least 1 h before testing for acclimation. All testing apparatuses were sprayed with 70% ethanol and wiped before the start of each trial. In the open-field test, each mouse was placed in the

center of the open-field apparatus with opaque walls (40 × 40 × 40 cm) and allowed to freely explore for 30 min (Fig. 5) or 1 h (Fig. 8) in a dimly lit room. The behavior of each mouse was

video recorded, and the distance traveled in the open field box was calculated using video tracking software (Ethovision XT, Noldus, Netherlands). The NOR and OLM tests were performed in the

same chamber used for the open-field test. During the training and test sessions, the animals were allowed to explore the objects for 10 min, with a 24-h intersession interval. Two

identical objects were placed in the chamber during the training session, and one of the objects was replaced with a new object in the test session of the NOR test. The test phase of the OLM

test was conducted by moving one of the two familiar objects to a different location in the chamber. The behavior of the animals was recorded, and the duration of exploring objects in each

session was manually scored by an experienced experimenter blinded to the mouse genotype and treatment. The preference index (%) in the NOR test was calculated as follows: (time spent

exploring the new object)/(total time spent exploring both new and familiar objects) × 100. The preference index in the OLM test was calculated as follows: (time spent exploring the moved

object)/(total time spent exploring both moved and unmoved objects) × 100. For the contextual fear conditioning test, animals were allowed to explore the fear conditioning chamber (Coulbourn

Instruments) for 5 min, during which (Pre-CS) the activity of each mouse was monitored. The mice were then exposed to a 2-s foot shock (0.7 mA) and returned to their home cage after 60 s.

The next day, the animals were returned to the same fear conditioning chamber, and the activity of each mouse was monitored for 5 min (CS). The freezing time (Fig. 5) was manually scored by

an experimenter blinded to the mouse genotype and treatment. The extent of activity suppression (Fig. 8) was measured using video tracking software (Ethovision XT, Noldus, Netherlands) as

follows: [(distance moved during pre-CS – distance moved during CS)/(distance moved during pre-CS)]. STATISTICAL ANALYSIS Statistical analyses were performed using Igor Pro (WaveMetrics) and

SPSS (Apache Software Foundation). The collected data were compared using parametric two-tailed Student’s _t_ tests or nonparametric Mann–Whitney tests. A one- or two-way analysis of

variance (ANOVA) with the Tukey multiple comparison test was used to compare multiple groups. All bar graphs in the figures show the mean ± standard error of the mean (SEM). RESULTS

DEFICIENCY OF AMINOPEPTIDASE P1 CAUSES ENHANCED NMDAR EXPRESSION AND ACTIVITY IN THE HIPPOCAMPUS To investigate the synaptic mechanisms underlying neurological and behavioral deficits in

Xpnpep1–/– mice, we first examined the expression levels of excitatory synaptic proteins in Xpnpep1–/– mouse brains (Fig. 1a, b). We observed a significant increase in the expression levels

of the NMDAR subunits GluN1 (_t_(6) = –3.84 and _p_ = 0.0086 by Student’s _t_ test) and GluN2A (_t_(6) = –3.90 and _p_ = 0.0080 by Student’s _t_ test) in the Xpnpep1–/– mouse hippocampal

homogenates, while forebrain homogenates from Xpnpep1–/– and wild-type (WT, Xpnpep1+/+) mice showed similar levels of GluN1 and GluN2A proteins. Interestingly, the expression levels of other

excitatory synaptic proteins, including GluN2B, GluA1, GluA2, PSD-95, VGLUT1, and synapsin I, did not change in either homogenate (Fig. 1a, b and Supplementary Fig. 1). Enhanced expression

of GluN1 and GluN2A suggests that a deficiency of aminopeptidase P1 increases NMDAR-mediated signaling in the hippocampus. Indeed, a significant enhancement in NMDAR-mediated synaptic

transmission was detected in the selected subregions of the hippocampus. NMDAR-mediated miniature excitatory postsynaptic currents (NMDA-mEPSCs) measured in the Xpnpep1–/– CA3 pyramidal

neurons at a holding potential of +40 mV in the presence of blockers of AMPARs, GABAA receptors (GABAARs), voltage-gated Na+-channels, and mGluR5s were significantly larger than those

measured in WT CA3 neurons, whereas there was no difference in the frequencies of NMDA-mEPSCs between the two genotypes (Fig. 1c–e). A larger amplitude with a normal frequency of

NMDA-mEPSCs, in addition to the enhanced hippocampal expression of GluN1 and GluN2A, indicates an increased NMDAR content in each excitatory synapse rather than an increase in the number of

silent synapses in Xpnpep1–/– CA3 neurons. Similar to the large NMDA-mEPSCs in CA3 neurons, synaptic NMDA/AMPA ratios, as determined from the evoked NMDAR-EPSCs and AMPAR-EPSCs, at Schaffer

collateral (SC)-CA1 synapses were significantly increased in Xpnpep1–/– CA1 pyramidal neurons (Fig. 1f–h). Interestingly, however, the NMDA/AMPA ratios in dentate gyrus (DG) granule cells

were not changed by the genetic disruption of aminopeptidase P1 (Fig. 1i–k). To determine whether subregion-specific changes in NMDAR-mediated neurotransmission in the hippocampi of

Xpnpep1–/– mice are associated with the expression level of aminopeptidase P1 in each hippocampal subregion, we investigated the expression pattern of aminopeptidase P1 in the hippocampus

using X-gal staining, as a β-galactosidase (lacZ) reporter is expressed in Xpnpep1 mutant mice under the control of the Xpnpep1 promoter. Immunohistochemical staining of the neuronal marker

NeuN on an X-gal-stained Xpnpep1+/– mouse hippocampal section revealed a distinct pattern of aminopeptidase P1 expression in the principal layer of each hippocampal subregion. In contrast to

the CA1 and CA3 areas, in which X-gal signals were detected throughout the neuronal somata in the principal cell layer, the DG exhibited prominent X-gal signals in the middle to outer

molecular layers and in the deep layer of the granule cell layer (Supplementary Fig. 2). Intriguingly, most granule cells with somata located in the superficial layer of the granule cell

layer did not exhibit X-gal signals in their somata (Supplementary Fig. 2e), indicating that aminopeptidase P1 deficiency is less likely to affect NMDAR activity in these neurons. We further

examined whether aminopeptidase P1 deficiency affected AMPAR- or GABAAR-mediated synaptic transmission in the hippocampus. AMPAR-mediated synaptic transmission in the CA3, CA1 and DG

principal neurons did not change in Xpnpep1–/– mice, as revealed by the normal amplitudes and frequencies of AMPAR-mEPSCs (Fig. 2a–d). This observation suggests that enhanced NMDA/AMPA

ratios at SC-CA1 synapses are unlikely to originate from decreased AMPAR currents. In addition, Xpnpep1–/– mice exhibited normal miniature inhibitory postsynaptic currents (mIPSCs) in the

CA3 and DG principal neurons, whereas the frequency of mIPSCs, but not the amplitude, in CA1 pyramidal neurons was increased in Xpnpep1–/– mice (Fig. 2e–h). Collectively, these results

indicate that a deficiency of aminopeptidase P1 results in synaptic dysfunction, mainly characterized by exaggerated NMDAR signaling, in hippocampal CA3 and CA1 pyramidal neurons.

AMINOPEPTIDASE P1-DEFICIENT MICE EXHIBIT ENHANCED NMDAR-DEPENDENT LONG-TERM POTENTIATION AT SC-CA1 SYNAPSES NMDARs play pivotal roles in the induction of activity-dependent modification of

synaptic strength, also known as synaptic plasticity, in hippocampal neurons, and alterations in NMDAR activity are often associated with neuropsychiatric disorders. We hypothesized that

abnormal NMDAR activity would influence NMDAR-dependent long-term potentiation (LTP) and/or long-term depression (LTD) in Xpnpep1–/– neurons. Therefore, we recorded field excitatory

postsynaptic potentials (fEPSPs) at the hippocampal SC-CA1 synapse. The amplitude of the fiber volley (FV) and the slope of the fEPSP evoked by increasing stimulation intensities were not

different between WT and Xpnpep1–/– mice (Fig. 3a–c). As AMPARs mediate the fast component of fEPSPs in hippocampal slices, the normal initial slopes of fEPSPs, together with the normal

AMPA-mEPSCs, further suggest intact AMPAR function in Xpnpep1–/– CA1 neurons under basal conditions. In addition, paired-pulse ratios determined from the slopes of fEPSPs evoked by two

successive stimuli with varying interstimulus intervals did not change at Xpnpep1–/– SC-CA1 synapses (Fig. 3d–e), indicating that basal presynaptic functions remain normal in Xpnpep1–/–

mice. However, theta burst stimulation (TBS) at SC-CA1 synapses produced significantly enhanced LTP in Xpnpep1–/– mice (Fig. 3f), although Xpnpep1–/– mice exhibit impaired

hippocampus-dependent learning and memory21. The enhancement in the magnitude of LTP was observed within 20 min after TBS and was maintained throughout the recording (Fig. 3f). Next, we

examined the effect of aminopeptidase P1 deficiency on the LTD of synaptic transmission at the same synapse. Intriguingly, LTD induced by low-frequency stimulation (LFS), a form of LTD

dependent on either the nonionotropic activation or the ionotropic activation of NMDARs, was indistinguishable between Xpnpep1+/+ mice and Xpnpep1–/– mice (Fig. 3g). In addition, mice of

both genotypes exhibited similar magnitudes of synaptic depression during and after bath application of the group I metabotropic glutamate receptor (mGluR) agonist DHPG (50 μM; Fig. 3h).

Together, these results suggest that a deficiency of aminopeptidase P1 results in enhanced LTP at SC-CA1 synapses (Fig. 3i). SUPPRESSION OF NMDARS RESTORES EXAGGERATED LTP AND NMDAR ACTIVITY

IN XPNPEP1–/– MICE As Xpnpep1–/– mice exhibit enhanced NMDAR activity and TBS-induced LTP, we wondered whether the exaggerated NMDAR activity is associated with abnormal LTP in the

Xpnpep1–/– hippocampus. Therefore, we examined the effect of NMDAR inhibition on LTP at SC-CA1 synapses. Bath application of the NMDAR antagonist AP-5 (50 μM) 5 min before and after TBS

completely blocked the induction of LTP in mice of both genotypes (Fig. 4a), suggesting that the long-lasting potentiation induced by TBS in our experiments was dependent on NMDAR

activation. If exaggerated NMDAR activity during TBS is responsible for the enhanced LTP and initial potentiation in Xpnpep1–/– mice, the suppression of excessive NMDAR activity would

restore abnormal synaptic potentiation. We tested this prediction using the uncompetitive NMDAR antagonist memantine. As memantine has a low affinity for NMDARs and binds only to open

channels, this drug, unlike competitive or noncompetitive NMDAR antagonists, preferentially inhibits the pathological overactivation of NMDARs without disturbing their normal

activity25,26,27. Indeed, bath application of memantine (2 μM) during electrophysiological recordings normalized the exaggerated LTP, whereas these concentrations of memantine had no effect

on TBS-induced LTP in control mice (Fig. 4b–d). We further examined whether memantine could restore exaggerated NMDAR activity in Xpnpep1–/– CA3 neurons (Fig. 4e–g). Memantine bears a single

positive charge under physiological conditions and has a primary binding site overlapping with that of Mg2+ in NMDARs, which is thought to be associated with the voltage-dependent effects

of memantine on NMDAR inhibition26,28,29. Therefore, we examined the effect of memantine (2 μM) on NMDA-mEPSCs at a holding potential of –70 mV with Mg2+-free bathing solution. Under these

experimental conditions, NMDA-mEPSCs were detected in the CA3 pyramidal neurons of both genotypes with a significantly (+/+, _t_(27) = 7.77, _p_ < 0.001; –/–, _t_(27) = 3.19, _p_ = 

0.0035) reduced frequency (Fig. 4f) compared to depolarization (+40 mV) and normal Mg2+ (1.3 mM) in the bathing solution (Fig. 1e). However, genotype differences in the amplitude of

NMDA-mEPSCs were still observed. In addition, bath application of memantine (2 μM) during electrophysiological recordings normalized the increased NMDAR-mediated synaptic transmission in

Xpnpep1–/– CA3 neurons (Fig. 4g). Collectively, these results indicate that memantine restores exaggerated NMDAR activity and NMDAR-mediated LTP in Xpnpep1–/– mice. ACUTE MEMANTINE

ADMINISTRATION REVERSES HYPERACTIVITY IN XPNPEP1–/– MICE Xpnpep1–/– mice exhibit hyperactivity and impaired hippocampus-dependent learning21. We reasoned that if exaggerated NMDAR signaling

is a key mechanism in triggering hyperactivity and cognitive disorders, the suppression of NMDAR activity will reverse the behavioral and cognitive symptoms observed in adult Xpnpep1–/–

mice. Indeed, a single intraperitoneal administration of 10 mg/kg memantine induced a significant reduction in both male and female Xpnpep1–/– mouse locomotor activity during the open-field

test performed 30 min after drug administration (Fig. 5a–f). This concentration of memantine is known to result in a peak brain concentration of 1–2 μM in rodents28. However, a single

administration of memantine did not have significant effects on learning and long-term memory in Xpnpep1–/– mice (Fig. 5g–l). Both memantine- and saline-treated Xpnpep1–/– mice showed

similar levels of impairment in the NOR and OLM tests, whereas memantine- and saline-treated WT mice preferred novel or moved objects when examined 24 h after administration (Fig. 5h–k).

Similarly, acute memantine administration had no effect on the freezing time of Xpnpep1–/– mice during the fear conditioning test (Fig. 5l). CHRONIC MEMANTINE TREATMENT REVERSES

NEURODEGENERATION IN THE HIPPOCAMPI OF XPNPEP1–/– MICE Based on the observation that acute memantine administration did not improve learning and long-term memory in Xpnpep1–/– mice, we

hypothesized that protection from hippocampal neurodegeneration is required to improve cognitive dysfunction in these mice. We sought to determine whether exaggerated NMDAR activity is

associated with CA3 neurodegeneration in Xpnpep1–/– mice21. Therefore, we intraperitoneally injected memantine (10 mg/kg) into mice twice daily for five weeks beginning on postnatal day 3

(Fig. 6a). Unexpectedly, chronic memantine treatment improved microcephaly in Xpnpep1–/– mice (Fig. 6b). In contrast to saline-treated Xpnpep1–/– mice, which exhibited prominent reductions

in the length and thickness of the forebrain, the brain size of memantine-treated Xpnpep1–/– mice was comparable with that of WT mice that received saline or memantine. In addition, repeated

administration of memantine for five weeks prevented early (4–5 weeks old21) neuronal cell death in Xpnpep1–/– mice, while saline-treated Xpnpep1–/– mice exhibited severe neurodegeneration

with a significant loss of neurons in the hippocampal CA3 but not in the DG and CA1 areas (Fig. 6c-f). These results indicate that perturbations in NMDAR signaling are associated with CA3

neurodegeneration in Xpnpep1–/– mice. As neurodegeneration in the CA3 region of Xpnpep1–/– mice is accompanied by confluent vacuoles of varying sizes (5–40 μm) with dark surrounding

neurons21, we further tested whether the restoration of NMDAR activity would suppress vacuolation in the CA3 regions of Xpnpep1–/– mice. Consistent with our previous histochemical

observations in the hippocampi of naïve Xpnpep1–/– mice, numerous confluent vacuoles were detected in the hippocampi of saline-treated Xpnpep1–/– mice (Fig. 7a). However, vacuolization was

significantly attenuated by repeated memantine administration, as revealed by the reduction in vacuole numbers and size in hematoxylin and eosin (HE)-stained brain sections from

memantine-treated Xpnpep1–/– mice (Fig. 7b–d). These results demonstrate that the suppression of exaggerated NMDAR signaling reverses paraptosis-like cell death in CA3 neurons caused by

aminopeptidase P1 deficiency. CHRONIC MEMANTINE TREATMENT IMPROVES THE NEURODEVELOPMENTAL AND COGNITIVE DEFICITS OF XPNPEP1–/– MICE As chronic memantine treatment reversed neurodegeneration

in the hippocampi of Xpnpep1–/– mice, we investigated whether the behavioral and cognitive deficits in Xpnpep1–/– mice were reversed by chronic memantine treatment. Intriguingly, repeated

administration (twice daily) of memantine (10 mg/kg, i.p.) for five weeks ameliorated the developmental delay observed in Xpnpep1–/– mice (Fig. 8a–c). Although memantine-treated Xpnpep1–/–

mice gained less body weight than WT mice, a treatment-specific difference in body weight was observed in Xpnpep1–/– mice (Fig. 8b). Memantine had a more profound effect on the body growth

of Xpnpep1–/– mice, such that memantine-treated Xpnpep1–/– mice had body lengths similar to those of saline- or memantine-treated Xpnpep1+/+ mice at 5 weeks. In addition, long-term treatment

with memantine improved hyperactivity in 6-week-old Xpnpep1–/– mice, as shown by the significantly reduced distance traveled in the open-field tests (Fig. 8d, e). Notably, the distance

traveled in the open-field arena was inversely correlated (_R_ = –0.825) with body weight (Fig. 8f), indicating that hyperactivity during and after weaning might have influenced feeding and

energy consumption in Xpnpep1–/– mice. We next examined the effects of chronic memantine treatment on cognitive dysfunction. In the NOR test, memantine-treated Xpnpep1–/– mice, but not

saline-treated Xpnpep1–/– mice, showed levels of new object preference that were comparable to those of WT mice (Fig. 8g, h). Moreover, chronic memantine treatment significantly improved

deficits in contextual fear learning in Xpnpep1–/– mice (Fig. 8i, j). As the improved performance of memantine-treated Xpnpep1–/– mice in the fear conditioning test may stem from reduced

hyperactivity and enhanced learning, we evaluated the effects of chronic memantine treatment on the performance of Xpnpep1–/– mice in the fear conditioning test using the suppression of

activity, an index of adaptive defense reaction30,31. There was a significant difference in the suppression of activity between saline- and memantine-treated Xpnpep1–/– mice, indicating that

fear learning was improved following chronic memantine treatment (Fig. 8i, j). We further investigated whether chronic memantine treatment restored NMDAR-dependent LTP and NMDAR expression

in the hippocampi of Xpnpep1–/– mice. Intriguingly, chronic memantine treatment abolished genotype-dependent differences in NMDAR-dependent LTP at SC-CA1 synapses (Fig. 8k, l), while this

treatment had no effect on the expression levels of hippocampal GluN1 or GluN2A in Xpnpep1–/– mice (Supplementary Fig. 3). However, the downstream signaling activity of NMDARs associated

with NMDAR-dependent LTP and neurodegeneration was significantly reduced in Xpnpep1–/– mice that received chronic memantine treatment (Supplementary Fig. 4). Collectively, these results

suggest that chronic memantine treatment improves learning and memory in Xpnpep1–/– mice through the suppression of exaggerated NMDAR activity and the restoration of signaling downstream of

NMDAR activation. DISCUSSION The present study demonstrates the pathological mechanisms of and a therapeutic strategy for neurological and cognitive disorders in a mouse model of an inborn

error of metabolism (IEM) caused by aminopeptidase P1 deficiency. Excessive NMDAR activity in aminopeptidase P1 deficiency induces exaggerated LTP in the hippocampus and triggers

neurodevelopmental disorders involving neurodegeneration, hyperactivity, and cognitive dysfunction. These observations indicate that the metabolic dysfunction in aminopeptidase P1 deficiency

perturbs NMDAR homeostasis in neurons, thereby leading to synaptopathy. Although aminopeptidase P1 deficiency is a rare IEM in humans19, our study provides experimental evidence showing

that neurodevelopmental and cognitive deficits in IEMs are treatable and preventable by pharmacological intervention restoring neural circuits. It has been reported that abnormal NMDAR

signaling is also implicated in other IEMs, such as phenylketonuria and homocystinuria32,33. Notably, mild to severe intellectual disability and hyperactivity are frequently observed in

patients with these IEMs34,35,36. Although the effects of L-phenylalanine at the concentration (0.1–0.4 mM) observed in the brains of patients with phenylketonuria on NMDAR activity are

unknown, the plasma concentration (> 1 mM) of L-phenylalanine in patients with phenylketonuria inhibits NMDARs, and chronic exposure to high concentrations of L-phenylalanine upregulates

the density of NMDARs in Pahenu2 brains with enhanced expression of GluN1 and GluN2A subunits and decreased GluN2B expression33,37,38. Similarly, homocysteine modulates both the amplitudes

and the desensitization of NMDAR currents dependent on GluN2 subunit composition39. Consistent with this, the NMDAR antagonists memantine and MK-801 block the homocysteine-induced cell death

of cultured neurons and glia32. Whether the restoration of NMDAR signaling, alone or in combination with dietary management, improves intellectual disability and hyperactivity in patients

with these diseases warrants further investigation. Based on the enzymatic action and hippocampal expression pattern of aminopeptidase P1, the metabolic substrates that cause hippocampal

dysfunction in Xpnpep1–/– mice are presumably oligopeptides with a penultimate proline17,19,40, and they are likely to be cleared in neuronal somatodendritic compartments under normal

conditions22. It has been reported that the tripeptide glycine-proline-glutamate (GPE) and the tetrapeptide threonine-proline-proline-threonine (TPPT or GLYX-13) act as an NMDAR agonist and

modulator, respectively41,42. Endogenous cleavage of insulin-like growth factor I (IGF-I) in the brain produces truncated IGF-1 and the N-terminal tripeptide GPE, which binds to the

glutamate binding site in NMDARs43. High concentrations of GPE activate NMDARs, and the GPE-induced currents are blocked by the competitive NMDAR antagonist

3-((R)-2-carboxypiperazin-4-yl)-propyl-1-phosphonic acid [(R)-CPP] or by extracellular Mg2+ at −70 mV42. Meanwhile, the tetrapeptide TPPT binds to the glycine modulatory site in NMDARs,

thereby acting as a partial agonist of NMDARs, similar to glycine, D-serine, and D-cycloserine41. Interestingly, GPE and TPPT possess a proline residue in the second position from the

N-terminus and are potential substrates of aminopeptidase P1. The proteolytic degradation of proteins, including proline-rich protein collagen, which consists of

glycine-proline-hydroxyproline tripeptide repeats44, may produce diverse oligopeptides with a penultimate proline by the action of many proteases or peptidases40. It is likely that the

accumulation of these peptides in the aminopeptidase P1 deficiency may stimulate or modulate NMDARs in the brain. However, the identity of harmful substrates that perturb NMDAR homeostasis

in Xpnpep1–/– mice remains unknown. Xpnpep1–/– mice exhibited enhanced NMDAR-dependent LTP at SC-CA1 synapses, possibly because of the exaggerated NMDAR activity in CA1 neurons (Figs. 3 and

4), but these mice displayed a significant impairment in hippocampus-dependent learning and memory21. The enhanced LTP in Xpnpep1–/– mice appears to be contradictory to their behavioral

outcomes (Figs. 5 and 8), as LTP is considered to be one of multiple neuronal learning and memory mechanisms45,46. Similar to our study, multiple studies observed an inverse correlation

between LTP and learning and memory in genetically engineered mice. Genetic disruption of AMPA receptor 2 (GluA2), postsynaptic density-95 protein (PSD-95), protein tyrosine phosphatase δ

(PTPδ), fragile XE-associated familial mental retardation protein 2 (FMR2), and insulin receptor tyrosine kinase substrate p53 (IRSp53) results in enhanced LTP but impaired

hippocampus-dependent learning and memory in mice24,47,48,49,50. As excessive LTP may cause subnormal place cell function51, the saturation of LTP52, or alterations in the spatial pattern of

synaptic weights53, abnormally enhanced LTP in the hippocampi of Xpnpep1–/– mice might affect new memory encoding or the accuracy of memory retrieval45. Although a low concentration of

memantine reversed abnormally enhanced NMDAR activity and NMDAR-dependent LTP in hippocampal slices from Xpnpep1–/– mice (Fig. 4), a single administration of memantine did not improve

learning and long-term memory but reversed hyperactivity in Xpnpep1–/– mice (Fig. 5). These observations imply that the impaired hippocampus-dependent learning in Xpnpep1–/– mice is

attributable to multiple defects, including severe neurodegeneration in the CA3 area. In support of this idea, chronic memantine administration ameliorated neurodegeneration (Fig. 6) and the

impairment of learning and memory (Fig. 8) in Xpnpep1–/– mice. The reduced activation of signaling downstream of NMDARs (Supplementary Fig. 4) and attenuated vacuolation (Fig. 7) in the

hippocampi of Xpnpep1–/– mice through chronic memantine treatment further indicate that excessive NMDAR activity during the developmental period is associated with neurodegeneration in CA3

neurons. As CA3 pyramidal neurons receive robust excitatory synaptic inputs, compared to DG or CA1 neurons (Fig. 2), through the mossy fiber-CA3, associational/commissural-CA3, and

entorhinal cortex-CA3 pathways54,55, excessive NMDAR activity might result in excitotoxic cell death and a disruption of normal ensemble patterns of CA3 neurons by abnormal amplification of

synchronous synaptic inputs in dendrites56,57. Therefore, excessive NMDAR activity as well as CA3 neurodegeneration and enhanced LTP seem to be associated with impaired learning and memory

in Xpnpep1–/– mice. Interestingly, while the neuronal cell death in the CA3 area of Xpnpep1–/– mice was almost completely prevented by chronic memantine treatment, vacuolation and behavioral

outcomes did not fully recover. One possible explanation for this incomplete recovery may be the fast clearance of memantine in mice. The half-life of memantine in mice (<4 h) is much

shorter than that in humans (60–80 h), and mice exhibit a steep Cmax/Cmin ratio of up to 100 for memantine58. Accordingly, memantine treatment twice daily might be insufficient for the

complete correction of behavioral abnormalities in mice. Nevertheless, chronic memantine treatment significantly improved behavioral hyperactivity and impaired learning in Xpnpep1–/– mice.

Although the detailed mechanism by which a deficiency of aminopeptidase P1 perturbs NMDAR homeostasis remains to be elucidated in future studies, the results of the present study suggest

that the restoration of neural circuits by pharmacological intervention may be an effective strategy for the treatment of neurodevelopmental, behavioral, and cognitive deficits in patients

with IEMs. REFERENCES * Lanpher, B., Brunetti-Pierri, N. & Lee, B. Inborn errors of metabolism: the flux from Mendelian to complex diseases. _Nat. Rev. Genet._ 7, 449–460 (2006). Article

  CAS  PubMed  Google Scholar  * Applegarth, D. A., Toone, J. R. & Lowry, R. B. Incidence of inborn errors of metabolism in British Columbia, 1969-1996. _Pediatrics_ 105, e10 (2000).

Article  CAS  PubMed  Google Scholar  * Sanderson, S., Green, A., Preece, M. A. & Burton, H. The incidence of inherited metabolic disorders in the West Midlands, UK. _Arch. Dis. Child._

91, 896–899 (2006). Article  CAS  PubMed  PubMed Central  Google Scholar  * Pampols, T. Inherited metabolic rare disease. _Adv. Exp. Med. Biol._ 686, 397–431 (2010). Article  PubMed  Google

Scholar  * Ferreira, C. R., van Karnebeek, C. D. M., Vockley, J. & Blau, N. A proposed nosology of inborn errors of metabolism. _Genet. Med._ 21, 102–106 (2019). Article  PubMed  Google

Scholar  * Kahler, S. G. & Fahey, M. C. Metabolic disorders and mental retardation. _Am. J. Med. Genet. C. Semin. Med. Genet._ 117C, 31–41 (2003). Article  PubMed  Google Scholar  * van

Karnebeek, C. D. & Stockler, S. Treatable inborn errors of metabolism causing intellectual disability: a systematic literature review. _Mol. Genet. Metab._ 105, 368–381 (2012). Article 

PubMed  CAS  Google Scholar  * Pierre, G. Neurodegenerative disorders and metabolic disease. _Arch. Dis. Child._ 98, 618–624 (2013). Article  PubMed  Google Scholar  * Sedel, F. et al.

Psychiatric manifestations revealing inborn errors of metabolism in adolescents and adults. _J. Inherit. Metab. Dis._ 30, 631–641 (2007). Article  CAS  PubMed  Google Scholar  * Ghaziuddin,

M. & Al-Owain, M. Autism spectrum disorders and inborn errors of metabolism: an update. _Pediatr. Neurol._ 49, 232–236 (2013). Article  PubMed  Google Scholar  * Saudubray, J. M. &

Garcia-Cazorla, A. An overview of inborn errors of metabolism affecting the brain: from neurodevelopment to neurodegenerative disorders. _Dialogues Clin. Neurosci._ 20, 301–325 (2018).

Article  PubMed  PubMed Central  Google Scholar  * Schweitzer, S., Shin, Y., Jakobs, C. & Brodehl, J. Long-term outcome in 134 patients with galactosaemia. _Eur. J. Pediatr._ 152, 36–43

(1993). Article  CAS  PubMed  Google Scholar  * Smith, I., Beasley, M. G. & Ades, A. E. Intelligence and quality of dietary treatment in phenylketonuria. _Arch. Dis. Child._ 65, 472–478

(1990). Article  CAS  PubMed  PubMed Central  Google Scholar  * Mudd, S. H. et al. The natural history of homocystinuria due to cystathionine beta-synthase deficiency. _Am. J. Hum. Genet._

37, 1–31 (1985). CAS  PubMed  PubMed Central  Google Scholar  * McDonald, J. D. Production of mouse models for the study of human inborn errors of metabolism. _Mol. Genet. Metab._ 71,

240–244 (2000). Article  CAS  PubMed  Google Scholar  * Moro, C. A. & Hanna-Rose, W. Animal model contributions to congenital metabolic disease. _Adv. Exp. Med. Biol._ 1236, 225–244

(2020). Article  CAS  PubMed  PubMed Central  Google Scholar  * Cottrell, G. S., Hooper, N. M. & Turner, A. J. Cloning, expression, and characterization of human cytosolic aminopeptidase

P: a single manganese(II)-dependent enzyme. _Biochemistry_ 39, 15121–15128 (2000). Article  CAS  PubMed  Google Scholar  * Ersahin, C., Szpaderska, A. M., Orawski, A. T. & Simmons, W.

H. Aminopeptidase P isozyme expression in human tissues and peripheral blood mononuclear cell fractions. _Arch. Biochem. Biophys._ 435, 303–310 (2005). Article  CAS  PubMed  Google Scholar 

* Blau, N., Niederwieser, A. & Shmerling, D. H. Peptiduria presumably caused by aminopeptidase-P deficiency. A new inborn error of metabolism. _J. Inherit. Metab. Dis._ 11, 240–242

(1988). Article  PubMed  Google Scholar  * Yoon, S. H. et al. Developmental retardation, microcephaly, and peptiduria in mice without aminopeptidase P1. _Biochem. Biophys. Res. Commun._ 429,

204–209 (2012). Article  CAS  PubMed  Google Scholar  * Bae, Y. S. et al. Deficiency of aminopeptidase P1 causes behavioral hyperactivity, cognitive deficits, and hippocampal

neurodegeneration. _Genes Brain Behav._ 17, 126–138 (2018). Article  CAS  PubMed  Google Scholar  * Yoon, S. H. et al. Altered hippocampal gene expression, glial cell population, and

neuronal excitability in aminopeptidase P1 deficiency. _Sci. Rep._ 11, 932 (2021). Article  CAS  PubMed  PubMed Central  Google Scholar  * Witt, A., Macdonald, N. & Kirkpatrick, P.

Memantine hydrochloride. _Nat. Rev. Drug Discov._ 3, 109–110 (2004). Article  CAS  PubMed  Google Scholar  * Kim, M. H. et al. Enhanced NMDA receptor-mediated synaptic transmission, enhanced

long-term potentiation, and impaired learning and memory in mice lacking IRSp53. _J. Neurosci._ 29, 1586–1595 (2009). Article  CAS  PubMed  PubMed Central  Google Scholar  * Chen, H. S.

& Lipton, S. A. Mechanism of memantine block of NMDA-activated channels in rat retinal ganglion cells: uncompetitive antagonism. _J. Physiol._ 499, 27–46 (1997). Article  CAS  PubMed 

PubMed Central  Google Scholar  * Martina, M., Comas, T. & Mealing, G. A. Selective pharmacological modulation of pyramidal neurons and interneurons in the CA1 region of the rat

hippocampus. _Front. Pharmacol._ 4, 24 (2013). Article  PubMed  PubMed Central  Google Scholar  * Lipton, S. A. & Chen, H. S. Paradigm shift in neuroprotective drug development:

clinically tolerated NMDA receptor inhibition by memantine. _Cell Death Differ._ 11, 18–20 (2004). Article  CAS  PubMed  Google Scholar  * Parsons, C. G., Stoffler, A. & Danysz, W.

Memantine: a NMDA receptor antagonist that improves memory by restoration of homeostasis in the glutamatergic system−too little activation is bad, too much is even worse. _Neuropharmacology_

53, 699–723 (2007). Article  CAS  PubMed  Google Scholar  * Chen, H. S. & Lipton, S. A. Pharmacological implications of two distinct mechanisms of interaction of memantine with

N-methyl-D-aspartate-gated channels. _J. Pharmacol. Exp. Ther._ 314, 961–971 (2005). Article  CAS  PubMed  Google Scholar  * Anagnostaras, S. G. et al. Selective cognitive dysfunction in

acetylcholine M1 muscarinic receptor mutant mice. _Nat. Neurosci._ 6, 51–58 (2003). Article  CAS  PubMed  Google Scholar  * Anagnostaras, S. G., Josselyn, S. A., Frankland, P. W. &

Silva, A. J. Computer-assisted behavioral assessment of Pavlovian fear conditioning in mice. _Learn. Mem._ 7, 58–72 (2000). Article  CAS  PubMed  PubMed Central  Google Scholar  * Lipton, S.

A. et al. Neurotoxicity associated with dual actions of homocysteine at the N-methyl-D-aspartate receptor. _Proc. Natl Acad. Sci. USA_ 94, 5923–5928 (1997). Article  CAS  PubMed  PubMed

Central  Google Scholar  * Glushakov, A. V. et al. Specific inhibition of N-methyl-D-aspartate receptor function in rat hippocampal neurons by L-phenylalanine at concentrations observed

during phenylketonuria. _Mol. Psychiatry_ 7, 359–367 (2002). Article  CAS  PubMed  Google Scholar  * Garcia-Cazorla, A. et al. Mental retardation and inborn errors of metabolism. _J.

Inherit. Metab. Dis._ 32, 597–608 (2009). Article  CAS  PubMed  Google Scholar  * Carson, N. A., Dent, C. E., Field, C. M. & Gaull, G. E. Homocystinuria: clinical and pathological review

of ten cases. _J. Pediatr._ 66, 565–583 (1965). Article  CAS  PubMed  Google Scholar  * Murphy, G. H. et al. Adults with untreated phenylketonuria: out of sight, out of mind. _Br. J.

Psychiatry_ 193, 501–502 (2008). Article  PubMed  PubMed Central  Google Scholar  * Glushakov, A. V. et al. Long-term changes in glutamatergic synaptic transmission in phenylketonuria.

_Brain_ 128, 300–307 (2005). Article  CAS  PubMed  Google Scholar  * Rupp, A. et al. Variability of blood-brain ratios of phenylalanine in typical patients with phenylketonuria. _J. Cereb.

Blood Flow. Metab._ 21, 276–284 (2001). Article  CAS  PubMed  Google Scholar  * Bolton, A. D., Phillips, M. A. & Constantine-Paton, M. Homocysteine reduces NMDAR desensitization and

differentially modulates peak amplitude of NMDAR currents, depending on GluN2 subunit composition. _J. Neurophysiol._ 110, 1567–1582 (2013). Article  CAS  PubMed  PubMed Central  Google

Scholar  * Vanhoof, G., Goossens, F., De Meester, I., Hendriks, D. & Scharpe, S. Proline motifs in peptides and their biological processing. _FASEB J._ 9, 736–744 (1995). Article  CAS 

PubMed  Google Scholar  * Moskal, J. R. et al. GLYX-13: a monoclonal antibody-derived peptide that acts as an N-methyl-D-aspartate receptor modulator. _Neuropharmacology_ 49, 1077–1087

(2005). Article  CAS  PubMed  Google Scholar  * Vaaga, C. E., Tovar, K. R. & Westbrook, G. L. The IGF-derived tripeptide Gly-Pro-Glu is a weak NMDA receptor agonist. _J. Neurophysiol._

112, 1241–1245 (2014). Article  CAS  PubMed  PubMed Central  Google Scholar  * Sara, V. R. et al. The biological role of truncated insulin-like growth factor-1 and the tripeptide GPE in the

central nervous system. _Ann. N. Y. Acad. Sci._ 692, 183–191 (1993). Article  CAS  PubMed  Google Scholar  * Krane, S. M. The importance of proline residues in the structure, stability and

susceptibility to proteolytic degradation of collagens. _Amino Acids_ 35, 703–710 (2008). Article  CAS  PubMed  Google Scholar  * Dringenberg, H. C. The history of long-term potentiation as

a memory mechanism: controversies, confirmation, and some lessons to remember. _Hippocampus_ 30, 987–1012 (2020). Article  PubMed  Google Scholar  * Neves, G., Cooke, S. F. & Bliss, T.

V. Synaptic plasticity, memory and the hippocampus: a neural network approach to causality. _Nat. Rev. Neurosci._ 9, 65–75 (2008). Article  CAS  PubMed  Google Scholar  * Migaud, M. et al.

Enhanced long-term potentiation and impaired learning in mice with mutant postsynaptic density-95 protein. _Nature_ 396, 433–439 (1998). Article  CAS  PubMed  Google Scholar  * Jia, Z. et

al. Enhanced LTP in mice deficient in the AMPA receptor GluR2. _Neuron_ 17, 945–956 (1996). Article  CAS  PubMed  Google Scholar  * Uetani, N. et al. Impaired learning with enhanced

hippocampal long-term potentiation in PTPdelta-deficient mice. _EMBO J._ 19, 2775–2785 (2000). Article  CAS  PubMed  PubMed Central  Google Scholar  * Gu, Y. et al. Impaired conditioned fear

and enhanced long-term potentiation in Fmr2 knock-out mice. _J. Neurosci._ 22, 2753–2763 (2002). Article  CAS  PubMed  PubMed Central  Google Scholar  * Taverna, F. A. et al. Defective

place cell activity in nociceptin receptor knockout mice with elevated NMDA receptor-dependent long-term potentiation. _J. Physiol._ 565, 579–591 (2005). Article  CAS  PubMed  PubMed Central

  Google Scholar  * Moser, E. I., Krobert, K. A., Moser, M. B. & Morris, R. G. Impaired spatial learning after saturation of long-term potentiation. _Science_ 281, 2038–2042 (1998).

Article  CAS  PubMed  Google Scholar  * Brun, V. H., Ytterbo, K., Morris, R. G., Moser, M. B. & Moser, E. I. Retrograde amnesia for spatial memory induced by NMDA receptor-mediated

long-term potentiation. _J. Neurosci._ 21, 356–362 (2001). Article  CAS  PubMed  PubMed Central  Google Scholar  * Henze, D. A., McMahon, D. B., Harris, K. M. & Barrionuevo, G. Giant

miniature EPSCs at the hippocampal mossy fiber to CA3 pyramidal cell synapse are monoquantal. _J. Neurophysiol._ 87, 15–29 (2002). Article  PubMed  Google Scholar  * Rebola, N., Carta, M.

& Mulle, C. Operation and plasticity of hippocampal CA3 circuits: implications for memory encoding. _Nat. Rev. Neurosci._ 18, 208–220 (2017). Article  CAS  PubMed  Google Scholar  *

Dong, X. X., Wang, Y. & Qin, Z. H. Molecular mechanisms of excitotoxicity and their relevance to pathogenesis of neurodegenerative diseases. _Acta Pharmacol. Sin._ 30, 379–387 (2009).

Article  CAS  PubMed  PubMed Central  Google Scholar  * Makara, J. K. & Magee, J. C. Variable dendritic integration in hippocampal CA3 pyramidal neurons. _Neuron_ 80, 1438–1450 (2013).

Article  CAS  PubMed  PubMed Central  Google Scholar  * Beconi, M. G. et al. Pharmacokinetics of memantine in rats and mice. _PLoS Curr._ 3, RRN1291 (2011). Article  PubMed  Google Scholar 

Download references ACKNOWLEDGEMENTS This study was supported by National Research Foundation of Korea (NRF) grants funded by the Korean government (NRF-2017R1D1A1B03032935,

2018R1A5A2025964, 2017M3C7A1029609, 2020R1A2C1014372, and 2022R1I1A1A01063296). AUTHOR INFORMATION Author notes * These authors contributed equally: Young-Soo Bae, Sang Ho Yoon, Young Sook

Kim. AUTHORS AND AFFILIATIONS * Department of Physiology and Biomedical Sciences, Seoul National University College of Medicine, Seoul, 03080, Korea Young-Soo Bae, Sang Ho Yoon, Young Sook

Kim, Sung Pyo Oh, Woo Seok Song, Jin Hee Cha & Myoung-Hwan Kim * Neuroscience Research Institute, Seoul National University Medical Research Center, Seoul, 03080, Korea Sang Ho Yoon, Woo

Seok Song & Myoung-Hwan Kim * Seoul National University Bundang Hospital, Seongnam, Gyeonggi, 13620, Korea Myoung-Hwan Kim Authors * Young-Soo Bae View author publications You can also

search for this author inPubMed Google Scholar * Sang Ho Yoon View author publications You can also search for this author inPubMed Google Scholar * Young Sook Kim View author publications

You can also search for this author inPubMed Google Scholar * Sung Pyo Oh View author publications You can also search for this author inPubMed Google Scholar * Woo Seok Song View author

publications You can also search for this author inPubMed Google Scholar * Jin Hee Cha View author publications You can also search for this author inPubMed Google Scholar * Myoung-Hwan Kim

View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS Y.-S.B. and M.-H.K. conceived the project. Y.-S.B., S.H.Y., Y.S.K., and M.-H.K. designed

the experiments. Y.-S.B., S.H.Y., Y.S.K., S.P.O., S.W.S., J.H.C., and M.-H.K. performed the experiments. Y.-S.B., S.H.Y., Y.S.K., and M.-H.K. analyzed the data. M.-H.K. wrote the manuscript,

with input from all other authors. CORRESPONDING AUTHOR Correspondence to Myoung-Hwan Kim. ETHICS DECLARATIONS COMPETING INTERESTS The authors declare no competing interests. ADDITIONAL

INFORMATION PUBLISHER’S NOTE Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. SUPPLEMENTARY INFORMATION SUPPLEMENTARY

INFORMATION RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution

and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if

changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the

material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to

obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS

ARTICLE Bae, YS., Yoon, S.H., Kim, Y.S. _et al._ Suppression of exaggerated NMDAR activity by memantine treatment ameliorates neurological and behavioral deficits in aminopeptidase

P1-deficient mice. _Exp Mol Med_ 54, 1109–1124 (2022). https://doi.org/10.1038/s12276-022-00818-9 Download citation * Received: 21 May 2021 * Revised: 12 May 2022 * Accepted: 17 May 2022 *

Published: 03 August 2022 * Issue Date: August 2022 * DOI: https://doi.org/10.1038/s12276-022-00818-9 SHARE THIS ARTICLE Anyone you share the following link with will be able to read this

content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative