Postnatal hypofunction of N‐methyl‐D‐aspartate receptors alters perforant path synaptic plasticity and filtering and impairs dentate gyrus‐mediated spatial discrimination

Transient hypofunction of the NMDA receptor represents a convergence point for the onset and further development of psychiatric disorders, including schizophrenia. Although the cumulative evidence indicates dysregulation of the hippocampal formation in schizophrenia, the integrity of the synaptic transmission and plasticity conveyed by the somatosensorial inputs to the dentate gyrus, the perforant pathway synapses, have barely been explored in this pathological condition.


| INTRODUCTION
A growing body of evidence indicates that hypofunction of N-methyl-D-aspartate (NMDA) receptors represents a convergence point for the onset and further development of psychiatric disorders, including autism and schizophrenia (Forsyth & Lewis, 2017;Nakazawa & Sapkota, 2020;Snyder & Gao, 2013).The genetic/chemogenetic ablation or pharmacological blockade of NMDA receptors during critical periods of brain development has successfully reproduced the negative, cognitive and psychotic symptoms associated with schizophrenia (Jeevakumar et al., 2015;Kjaerby et al., 2017;Nakao et al., 2019;Segev et al., 2020;Seshadri et al., 2018).Specifically, hypofunction of NMDA receptor by postnatal blockade with antagonists, such as MK-801 (dizocilpine) leads to alterations in the release of dopamine, glutamate and GABA (Márquez et al., 2023;Nakao et al., 2019;Segev et al., 2020) and in the functional expression of ion channels and postsynaptic receptors in the hippocampus (Griego et al., 2022;Márquez et al., 2023), along with behavioural deficits that resemble the negative and cognitive symptoms of schizophrenia (Kjaerby et al., 2017;Segev et al., 2020;Seshadri et al., 2018).
It is now accepted that the hippocampus, a brain region essential for cognitive processing, the formation of new memories (Eichenbaum, 2004;Nadel et al., 2012) and the development of social behaviours (Lopez-Rojas et al., 2022), is central in the pathophysiology of schizophrenia (Segev et al., 2020;Tamminga et al., 2010).In line with this tenet, the dentate gyrus (DG), the somatosensorial gate of the hippocampus, shows the greatest volumetric loss, neuroanatomic disorganization and altered glutamate/ GABA neurotransmission in schizophrenic individuals and animal models (Griego et al., 2022;Li et al., 2015;Nakahara et al., 2020;Stan et al., 2015).The DG receives and processes somatosensorial and spatial information from the entorhinal cortex via the lateral perforant pathway (LPP) and the medial perforant pathway (MPP), respectively (Fernández-Ruiz et al., 2021;Hunsaker et al., 2007), minimizes the overlapping of new memories with highly similar content via a theoretical mechanism known as pattern separation (Yassa & Stark, 2011).
Additionally, multiple works have shown the relevance of the cannabinoid 1 (CB 1 ) receptor for the induction of long-term depression and potentiation in the perforant path-DG synapses (Fontaine et al., 2020;Peñasco et al., 2019;Wang et al., 2016;Wang, Jia, et al., 2018;Wu et al., 2006).Interestingly, dysregulation of NMDA and CB 1 receptors is present in animal models and schizophrenics themselves (Forsyth & Lewis, 2017;Márquez et al., 2023;Osborne et al., 2019;Szűcs et al., 2016).Despite the potential repercussions of the altered functionality of NMDA and CB 1 receptors within the LPP and MPP-DG synapses, little is known about the possibly altered functionality of the DG in schizophrenia.
This study identified neurophysiological alterations experienced by LPP and MPP-DG synapses in response to the transient hypofunction of NMDA receptors during early postnatal development.
The changes in the neurotransmitter release, altered synaptic strength and dysregulated synaptic filtering found in these synapses, are accompanied by changes in the induction and expression of long-term potentiation (LTP) and long-term depression (LTD) and dysfunction of the presynaptic activity of the CB 1 receptor.We have also demonstrated that induction of LTD is restored by enhancing the 2-arachidonoylglycerol (2-AG) signalling, the endogenous ligand for the CB 1 receptor, via negative modulation of monoacylglycerol lipase (MAGL).Finally, we demonstrated for the first time that spatial discrimination, a cognitive task in which the DG takes part, deteriorates in response to postnatal hypofunction of NMDA receptors.

| Animals and postnatal treatment with MK-801
All our experimental procedures were carried out in rigorous accordance with the 'NOM-062-ZOO-1999' local regulation for the use and care of laboratory animals and the protocols authorized by the internal ethics committee (CICUAL) of our institution (CINVESTAV; Protocol number 0090-14), which mandate the minimization of What is already known?
• NMDA receptor hypofunction is a mechanism for development of psychiatric illnesses involving hippocampal dysregulation.
• Evidence suggests physiological and behavioural dysregulation of the dentate gyrus, the hippocampus sensorial gate.
What does this study add?
• Postnatal administration of MK-801 modifies LTP, LTD in the dentate gyrus and reduces spatial discrimination.
What is the clinical significance?
• Mnemonic discrimination performance could help identify individuals at high risk for developing schizophrenia.
• Cannabinoid system modulation may improve dentate gyrus-dependent synaptic and behavioural capabilities in schizophrenic individuals.
suffering and the number of experimental animals used.Animal studies are reported in compliance with the ARRIVE guidelines (Percie du Sert et al., 2020) and with the recommendations made by the British Journal of Pharmacology (Lilley et al., 2020), and the National Institutes of Health (US) guidelines for animal care.
Pregnant Wistar rats were supplied by our vivarium and given continuous veterinary care.The rats were housed in acrylic cages with a density of 5-6 rats per cage following the next conditions: an inverted light-dark cycle (12 h; the light was turned off at 10:00 h), a room temperature of 23 ± 1 C, and free access to food (Rat chow, LabDiet 5008) and water.The pups birth was designated as postnatal day 0 (PD0) and animals were kept with their mothers until PD21.
The electrophysiological and behavioural procedures were performed in independent groups of male animals ranging from 30 to 38 days old.Each experimental manipulation was carried out in at least three different litters.From PD7 to PD11, the animals received daily subcutaneous injections of either saline solution (SS, 0.9%) or MK-801 (0.2 mgÁkg À1 in a 2-mlÁkg À1 volume) (Griego et al., 2022;Márquez et al., 2023).The behavioural procedures were carried out between 10:00 and 16:00 h.The experimental animals were treated with saline solution or MK-801, following a single sequence of random assignments.The MK-801 was dissolved in isotonic saline solution.Animal suffering was minimized during and after the injections to reduce behavioural or neurophysiological alterations that might affect the experimental results.In addition, the random assignment of animals to control and MK-801 groups was blind to prevent any potential bias.

| Extracellular recordings
Extracellular recordings were used to determine the synaptic properties of the MPP-DG synapse or the LPP-DG synapse acquired from the supra-pyramidal blade of the DG (see schematic representation in Figure 1a).The MPP field excitatory postsynaptic potentials (MPP fEPSPs) were evoked with a bipolar nichrome electrode placed in the middle one-third section of the DG molecular layer.The LPP fEPSPs were evoked with an electrode placed in the outer one-third of the DG's molecular layer.The evoked fEPSPs were recorded with a borosilicate pipette (1-2 MΩ of tip resistance, when filled with NaCl solution, 3 M) and placed 200-300 μm away from the stimulation electrode (see Figure 1a).The current pulses were delivered via a highvoltage isolation unit (A365D; World Precision Instruments, Sarasota, FL, USA), commanded with a Master-8 pulse generator (AMPI, Jerusalem, Israel).Responses were amplified with a Dagan BVC-700A amplifier (Minneapolis, MN, USA) connected with a 100Â gain headstage (Dagan, model 8024) and high-pass filtered at 0.3 Hz.Electrical noise suppression was accomplished with a Humbug noise eliminator (Quest Scientific Instruments; North Vancouver, BC, Canada).The evoked fEPSPs were displayed on a computer-based oscilloscope and digitalized for storage and offline analysis with LabVIEW 7.1 software (National Instruments, Austin, TX, USA).The A/D converter device used was the BNC-2110 (National Instrument, Austin, TX).

| fEPSP analyses
The fEPSP slope (10% to 60% of the fEPSP) was measured for synaptic plasticity experiments, given that the slope measurement reflects the synaptic strength (or synaptic drive), that is the number and properties of ionotropic glutamate receptors that open at the synapses.
The fEPSP slope values were normalized to the baseline response and expressed as percent values (%).For the input-output relationships, the amplitude of the fEPSP was measured as a parameter of global synaptic activity, given that this measurement not only includes the synaptic drive but also reflects the activity of multiple dendritic voltage-gated ion channels that influence the intrinsic excitability at the dendritic level or influence the waveform of the fEPSP (see Oulé et al., 2021;Wigström & Gustafsson, 1983).The fEPSP amplitude values were expressed in millivolts.

| Determination of basal synaptic strength and paired-pulse ratio
A series of input-output (I-O) curves of current versus synaptic responses was constructed by delivering increasing current pulses at 0.067 Hz from 0 to 800-μA (100-μA steeps, 100-μs duration) and measuring the evoked fEPSP amplitude (in mV).From the inputoutput curves, we also analysed the fibre volley (FV) amplitude in response to current injection to determine presynaptic excitability.
The basal synaptic strength was determined by analyses of the fibre volley amplitude versus fEPSP slope fitted with a Boltzmann function, and the differences between slices from control and MK-801-treated animals were compared using the resulting slope of the Boltzmann equation.For short-term plasticity determined by paired-pulse ratio (PPR) analysis, a paired stimulation at 0.067 Hz (15%-20% of evoked maximal fEPSP amplitude) was obtained, with inter-stimulus intervals (ISI) of 40, 60, 100, 200 and 500 ms.The PPR was calculated as the ratio between the amplitude of the second response (stimulus S2) and the amplitude of the first response (stimulus S1) (PPR = S2/S1).The input-specific origins of synaptic responses from the MPP and the LPP were corroborated by the preferential pharmacological sensitivity to DCG-IV (5 μM) and L-AP4 (20 μM), respectively (Macek et al., 1996).

| Long-term plasticity: depression and synaptic potentiation
Because homosynaptic long-term depression (LTD) has scarcely been examined in the two anatomical divisions of the perforant path to the DG, different patterns of low-frequency stimulation (LFS) were used to examine the LPP and MPP susceptibility to express a stable LTD.A baseline response of fEPSP slope (50%-70% of maximal fEPSP amplitude) was acquired with paired stimulation (60-ms inter-stimulus interval, 100-μs duration of current pulse) at 0.067 Hz for 20 min.Then, 900 unitary current pulses at 1-3 Hz were delivered at the LPP or the MPP, and the synaptic responses were recorded for 50 or 90 min.This was followed by pharmacological identification with L-AP4 or DCG-IV (Macek et al., 1996).In the case of MPP-DG synapse, LFS at 3 Hz was examined in the presence of the inverse agonist AM251 (5 μM) to determine the dependence of CB 1 receptors during the induction of LTD.For LTP experiments, the baseline conditions and pharmacological identification were identical to those in the LTD experiments, except that the baseline response of the fEPSP slope was configured at 25%-35% of its maximal amplitude.Then, a theta-burst stimulation (TBS) protocol was delivered to the LPP or the MPP, and synaptic responses were recorded for 90 minutes.The TBS protocol (based on Larson & Munkácsy, 2015) consisted of three episodes repeated at 10 s, each with 10 bursts at 5 Hz and five current pulses at 100 Hz, according to exploratory results (see Figure S2).The decay of post-tetanic potentiation (PTP) induced by TBS was expressed as τ value, obtained from adjusting the best fit of individual fEPSP slope values with a nonlinear regression function of one phase decay.Heatmaps were constructed to depict the magnitude of LTP, and cumulative probability charts were constructed using post-TBS fEPSP slope values from minute 11 to 90.

| Pharmacological modulation of 2-AG signalling pathway
In a subset of experiments focused on the MPP-DG synapse in slices from control and MK-801-treated animals, WIN55212-2, a CB 1 agonist (5 μM), was bath perfused for 15 min to examine the presynaptic induction of LTD mediated by CB 1 receptor activation, as previously reported (Fontaine et al., 2020).Additionally, physostigmine (10 μM), an acetylcholinesterase inhibitor, was perfused for 15 min to evaluate endogenous 2-AG production in the MPP-DG synapse, as previously demonstrated in acute hippocampal slices (Wang, Cox, et al., 2018;Wang, Jia, et al., 2018).Finally, in another subset of experiments, JZL 184 (1 μM), an irreversible inhibitor of MAGL that increases the 2-AG levels (Wang et al., 2016) was perfused for 15 min in slices from MK-801-treated animals to examine the effects of MAGL inhibition during the delivery of LFS at 3 Hz.

| Determination of frequency-dependent filtering in the LPP-DG synapse
Synaptic frequency-dependent filtering was examined in the presence of DCG-IV (5 μM), and a baseline response of fEPSP slope configured at 50% of its maximal amplitude was acquired for 10-12 min.Then, a train of 10 current pulses (100 μs of duration) at 5, 20 and 50 Hz was delivered to the LPP, with an interval of 10 min between trains, as previously reported (Quintanilla et al., 2022).For each stimulation frequency, the train's evoked synaptic responses were normalized to the slope of its first synaptic response and plotted as the number of stimuli versus normalized fEPSP slope (%).

| Spatial pattern separation task
The spatial object pattern separation (OPS) task (based on van Goethem et al., 2018) was performed to evaluate the animals capacity to discriminate minimal changes in the spatial position of identical objects in a familiar arena.To reduce the animals stress from the experimental manipulation during behavioural evaluations, the experimenter handled the rats twice per day (2-3 min) for five consecutive days before starting the behavioural evaluations.A circular arena (40 cm high and 83 cm in diameter, grey wall) was placed on a black acrylic platform inside an experimental room with red light.On this platform, a series of reference points denominated positions (P) were designated as P1 to P5, with 6 cm between the Ps.P1 was aligned at the centre of the platform, both left and right, and then P2 to P5 were designated both upward and downward, as depicted in Figure 9a.
The spatial object pattern separation task consisted of two trials videotaped (3 min duration each): the learning trial (T1) and the discrimination trial (T2).Before T1 in each evaluation, a habituation phase was carried out, during which the animal was exposed to the empty arena for 5 min.In T1, the animal was introduced to the arena with two identical objects placed in P1 from the left and right of the platform.One hour later, in T2, the animal was re-exposed to the same objects, but one object was displaced to a new position; this was then carried on to include all other positions (P2-P5).The epoch duration was 3 min for T1 and T2.Object exploration was considered for analysis when the animal explored an object with its nose, with a minimal distance of 2 cm.The time of object exploration was analyzed offline with Optimouse, a graphical user interface, executed in Matlab (Ben-Shaul, 2017).From the time of exploration for each object, we calculated the discrimination index (DI), a quantitative measurement of the animals preference for exploring the displaced object over the stationary object, using the following formula: DI ¼ displaced object ´s exploration time À stationary object 0 s exploration time two objects ´total exploration time To estimate the DI, a minimum exploration time of 7 and 10 s was required for T1 and T2, respectively (van Goethem et al., 2018).
Likewise, the performance of each animal for all positions was determined.Therefore, the animals were exposed 5 times to the spatial object pattern separation task with an interval of 2 days and different par objects, which prevented familiarization (van Goethem et al., 2018).The order of evaluation of object positions was random: while one animal began with P1 (without displacement of objects), another animal began with P5 (maximal displacement of one object).
Likewise, the allocation of object location for five spatial configurations was random, as presented in Table S1.After each trial, the platform and the objects were cleaned with ethanol (70%) to eliminate odour residues.The first evaluation of the spatial object pattern separation task was carried out in animals of P30.

| Statistical analysis
Group measures are numerically expressed as mean ± SEM.The violin plots show the median, mean, first, and third quartiles.For plasticity and synaptic filtering experiments, the data were normalized to baseline response (%).All statistical comparisons were performed between control values and those obtained from MK-801-treated animals, following the editorial guidelines on experimental design and analysis in pharmacology (Curtis et al., 2022).We have complied with the recom-

| Materials
Except for (+)-MK-801 maleate, DCG-IV, L-AP4 and AM251, which were purchased from Tocris Biosciences (Minneapolis, MN, USA), the drugs and chemicals used in this study were purchased from Sigma Aldrich (St. Louis, MO, USA).Details of other materials and suppliers were provided in the specific sections.

| Nomenclature of targets and ligands
Key protein targets and ligands in this article are hyperlinked to corresponding entries in the IUPHAR/BPS Guide to PHARMACOLOGY http://www.guidetopharmacology.org, and are permanently archived in the Concise Guide to PHARMACOLOGY 2021/22 (Alexander, Christopoulos et al., 2023;Alexander, Fabbro et al., 2023).

| Criteria for isolating synaptic responses from the medial and lateral perforant paths onto granule cells of the dentate gyrus
The dentate gyrus (DG) granule cells receive glutamatergic inputs via the axons of stellate cells and pyramidal neurons of the lateral and medial EC.These axons, collectively named perforant paths (PP), exhibit unique pharmacological and synaptic properties depending on whether they originate from the lateral or medial entorhinal cortex (LEC and MEC, respectively).We first isolated and characterized their synaptic responses before exploring whether postnatal treatment with MK-801 alters the LPP and MPP synapses.A stimulation electrode was positioned in the outer one-third (≈200-220 μm above the cell layer) or middle one-third (≈100-120 μm above) of the supra-pyramidal DG molecular layer (Petersen et al., 2013), and the recording pipettes were positioned 200-300 μm away from the stimulation electrode to record a fEPSPs from the LPP or MPP, respectively (Figure 1a).Presynaptic terminals of lateral entorhinal cortex-contacting granule cells selectively express group III metabotropic glutamate (mGlu) receptors, while those of medial entorhinal cortex-contacting granule cells express group II mGlu receptors (Macek et al., 1996;Shigemoto et al., 1997).We used these pharmacological criteria to verify the origin of the evoked responses.Bath perfusion of DCG-IV (5 μM), a group II mGlu agonist, did not depress the LPP fEPSP (92.9 ± 3.77% of baseline response).hHowever, the subsequent perfusion of L-AP4 (20 μM), a group III mGlu agonist, abolished the synaptic response (17.18 ± 2.8% of baseline response; n = 7 slices per six animals, upper traces in Figure 1b).On the other hand, bath perfusion of L-AP4 did not depress the MPP fEPSP (94.8 ± 3.28% of baseline response).However, the synaptic response was abolished during the subsequent application of DCG-IV (20.35 ± 3.2% of baseline response; n = 7 slices per six animals; lower traces in Figure 1b).The pharmacological activation of mGlu receptors shows that responses evoked from the LPP and the MPP converging on DG granule cells can be reliably isolated.Therefore, we systematically used these criteria to corroborate the synaptic origins of the evoked responses included in this study.

| The transient hypofunction of NMDA receptors during early postnatal development differentially affects the synaptic strength of the perforant path-DG synapses
It is already established that postnatal treatment with MK-801 or related antagonists of NMDA receptors such as phencyclidine alters the glutamatergic transmission of the hippocampus (Griego et al., 2022;Kjaerby et al., 2017;Márquez et al., 2023).Therefore, we explored the effects of MK-801 on the synaptic strength of the LPP and the MPP inputs to DG granule cells.In the control condition, the averaged input-output curve of the LPP fEPSP (0-800 μA; 100-μA steps, at 0.067 Hz) reached a maximal amplitude of 4.16 ± 0.46 mV (black symbols in Figure 1c; n = 9 slices per sic animals).Remarkably, postnatal treatment with MK-801 did not alter the maximal amplitude of the LPP fEPSP (maximal amplitude: 3.54 ± 0.34 mV; green symbols in Figure 1c, n = 10 slices per six animals).Contrary to this observation, the MPP fEPSP reached a maximal amplitude of 7.5 ± 0.68 mV (black symbols in Figure 1d; n = 8 slices per six animals) and postnatal treatment with MK-801 reduced the maximal MPP fEPSP amplitude to 3.87 ± 0.52 mV (P < 0.05 vs. control, two-way RM ANOVA treatment effect: F [1, 16] = 8.31; red symbols in Figure 1d; n = 10 slices per six animals).Our findings indicate that postnatal treatment with MK-801 selectively dysregulates the MPP inputs to DG granule cells.
While the MPP exhibits a ≈ 50% reduction in its maximal amplitude, the LPP transmission is not affected by MK-801.Likewise, the pharmacological selectivity of MK-801 suggests a differential composition in the presynaptic and postsynaptic components mediating glutamatergic transmission of the LPP and the MPP onto DG granule cells.In addition, we performed an analysis of the fEPSP kinetics from MPP and LPP synapses; these parameters corroborate the differences in the strength of both synapses (Table 1).
Because the fEPSPs acquired for the input-output curves were preceded by presynaptic fibre volleys (FV), we also analysed the relationships between stimulus intensity, fibre volley amplitude and fEPSP amplitude.The left panel in Figure 1e shows the relationship between stimulus intensity (I) and LPP fibre volley amplitude.The right panel depicts the relationship between fibre volley amplitude and LPP fEPSP slope in both experimental conditions (black and green symbols).We found that postnatal treatment with MK-801 did not alter either relationship at the LPP-DG synapse.However, the same analysis at the MPP synapse uncovered a different scenario.The left panel in Figure 1f contrasts the fibre volley response in control versus MK-801-treated animals (black and red symbols).Compared to the control slices, the slices from MK-801-treated animals exhibited a significant decrease in fibre volley amplitude and faster response saturation (two-way RM ANOVA, interaction effect [treatment Â current intensity]: F (8, 88) = 2.57, P < 0.05; red symbols; left panel in Figure 1f).
Likewise, the coupling analysis of the fibre volley versus MPP fEPSP slope shows a dysregulation between fibre volley and MPP fEPSP (slope in control: 0.87 ± 0.24 mVÁms À1 ; in MK-801: 0.16 ± 0.04 mVÁms À1 ; Mann-Whitney test, P < 0.05; right panel in Figure 1f).These results suggest that postnatal treatment with MK-801 alters the propagation of presynaptic action potentials and the concomitant synaptic response in the MPP but not the LPP synapse.2f).The increased PPF observed in the LPP and the MPP synapses suggests a dysregulation in the presynaptic machinery that controls glutamate release on the DG granule cells.

| Low-frequency stimulation induces the MPP long-term depression dependent on the cannabinoid receptor type 1
Thus far, our data show decreased synaptic strength and altered control of glutamate release of the perforant path terminals synapsing DG granule cells.Previous studies demonstrated that those MK-801-driven changes negatively impact the induction of long-term synaptic plasticity in the hippocampus (Griego et al., 2022;Márquez et al., 2023).Therefore, it is reasonable to assume that postnatal treatment with MK-801 will negatively impact the induction of long-term plasticity at the PP synapses on DG granule cells of the hippocampus.
However, since long-term depression has been scarcely explored in the perforant path-DG synapses, we first determined a reliable stimulation protocol to induce LTD.Our exploratory experiments show that low-frequency stimulation (LFS, 900 pulses, 1 Hz), a protocol commonly used to induce glutamatergic LTD in the hippocampus (Dudek & Bear, 1992) Stellar cells and pyramidal neurons of EC layer II discharge volleys in the delta and theta range (0.4-10 Hz) to DG granule cells (Deshmukh et al., 2010;Gloveli et al., 1997).Therefore, we explored whether a stimulation paradigm within this physiological range  1.54 ± 0.17; Wilcoxon test, P < 0.05; black bars in Figure 3f), indicating presynaptic locus for expression of LTD.
According to a previous study, induction of LTD in the MPP synapse requires the delivery of 6000 pulses at 10 Hz, while simultaneously blocking GABA A receptors (Peñasco et al., 2019).Under these experimental conditions, the authors reported that this form of LTD requires activation of the CB 1 receptor.Therefore, our next experiment aimed to determine if CB 1 receptor is required for LTD induced with 3 Hz, while maintaining active GABAergic transmission.Likewise, the reduced synaptic depression observed in the MK-801 in the presence of physostigmine may be ascribed to dysregulated CB 1 receptor functionality.This possibility requires further exploration.

|
The inhibition of monoacylglycerol lipase (MAGL) responsible for 2-AG breakdown restores the impaired CB 1 receptor-dependent LTD in the slices from MK-801-treated animals We reasoned that the impaired LTD may be partially explained by insufficient production or accelerated breakdown of 2-AG during synaptic stimulation.Previous studies have documented increased activity of MAGL, the enzyme responsible for 2-AG degradation, in schizophrenic individuals and experimental models of schizophrenia (Du et al., 2013;Kaya et al., 2019).Therefore, we examined whether  signalling of 2-AG in other animal and human studies of schizophrenia (Kaminitz et al., 2014;Kaya et al., 2019;Szűcs et al., 2016).It is hypothesized that after the arrival of cortical information conveying two similar experiences or two events close in time, the DG uses a computational process called pattern separation to orthogonalize (or maximize) the differences in the incoming information onto area CA3 (Santoro, 2013;Yassa & Stark, 2011).This neural mechanism is believed to facilitate the proper storage of similar neuronal information as independent events and prevent the overlapping of new memories.While strictly evaluating the pattern separation in humans and rodents is a challenging task (Santoro, 2013), its assessment as a behavioural phenomenon (referred to as mnemonic discrimination) is more accessible.Compared to humans, assessing mnemonic discrimination in rodents is more challenging due to the limited number of validated behavioural tasks.One exception is the spatial object pattern separation task developed by van Goethem et al. (2018), which mainly evaluates mnemonic discrimination of spatial type.We hypothesized that if the DG integrity is compromised due to transient hypofunction of NMDA receptors, the behavioural activity in which the DG participates will also be compromised.To explore this possibility, we evaluated MK-801-treated animals ability to discriminate small changes in the spatial configuration of identical objects maintained in a familiar environment (Santoro, 2013).The discrimination index of this behavioural test was evaluated by alternating the spatial configuration of identical objects in a familiar environment (see Figure 9a for a schematic representation).A series of spatial positions (P) was used to determine the minimal displacement position of one object versus another at which the animal perceived the change in spatial position.
This cognitive ability increases the demand for spatial pattern separation activity (van Goethem et al., 2018).
In control animals, we corroborated that the discrimination index (DI) depends on the magnitude of displacement of one object (Figure 9b), a phenomenon previously reported in adult animals (van Goethem et al., 2018) Saggu et al., 2013) and schizophrenic individuals (Egbujo et al., 2016).
Finally, the altered neurotransmitter release in the MPP compared to the LPP synapse of MK-801-treated animals may suggest a marked dysregulation in transferring spatial information but not nonspatial information from the perforant path to the DG, a phenomenon that requires additional investigation.

| Changes in the induction and expression of synaptic plasticity
Because glutamatergic LTD has been scarcely explored in the perforant path-DG synapses, a relevant finding of this study was the establishment of a protocol to induce reliable LTD.The exploratory experiments designed to this end showed that electrical stimulation of MPP, but not LPP, induces a stable LTD in response to 900 pulses delivered at 3 Hz, a stimulation frequency that mimics the delta-theta range of volley activity of stellar cells and pyramidal neurons of EC layer II synapsing DG granule cells (Deshmukh et al., 2010;Gloveli et al., 1997).previously reported in patients with schizophrenia and experimental models (Du et al., 2013;Kaya et al., 2019).Although appealing, this surmise requires further investigation.
Another intriguing observation is that LFS triggered anomalous synaptic potentiation in the slices from MK-801-treated animals.In a previous study, it was documented that the constitutive genic deletion of CB 1 receptor or its pharmacological blockade induces LFSmediated potentiation of the MPP synapse, and this phenomenon is dependent on NMDA receptors.One possible explanation for the potentiation induced with LFS is that MK-801 may alter the functional expression of CB 1 receptor (Peñasco et al., 2019), thus facilitating synaptic potentiation of the NMDA receptor transmission.In support of this possibility, the potentiation observed in the MK-801 slices did not exhibit changes in the paired-pulse ratio, suggesting a postsynaptic imbalance.
We also documented that transient hypofunction of NMDA receptors hinders the induction and expression of theta-burst stimulation (TBS)-induced LTP in the LPP synapse and attenuates the magnitude but not the expression of LTP in the MPP synapse.While MPP LTP may be induced in slices from MK-801-treated animals through a series of redundant signalling pathways, including CaMKII and PKA-ERK1/ERK2 (Welsby et al., 2009;Wu et al., 2006), that assures its postsynaptic expression.In the LPP-DG synapse, LTP requires postsynaptic synthesis of 2-AG (Wang et al., 2016) Rojas et al., 2016;Schreurs et al., 2017;Welsby et al., 2009).
Together, these factors can modify the magnitude of the LTP.Under the present experimental conditions, we observed a larger LTP in the MPP synapse than the observed in the LPP synapse.More importantly, the magnitude of LTP in both synapses is consistent with values previously reported for the dorsal hippocampus maintaining its GABAergic transmission active (Oulé et al., 2021).

| Changes in the synaptic filtering at the lateral perforant path (PP)
It has been hypothesized that low-pass filtering of synaptic inputs to the DG underlies the encoding information of auditory, olfactory and visual cues (Madar et al., 2019;Scullin & Partridge, 2012).In this regard, frequency-dependent synaptic filtering of the LPP has been thoroughly characterized in adult mice (Quintanilla et al., 2022).We first confirmed that juvenile rats possess this physiological property; then, we demonstrated that synaptic filtering in the LPP is drastically altered in slices from MK-801-treated animals.Our experimental findings and multiple theoretical models (Chance et al., 1998;Quintanilla et al., 2022) imply that the diminished filtering capacity of the LPP synapse may result from a decreased presynaptic release.The impaired low-pass filtering capability of the LPP could represent the cellular substrate that explains the abnormal processing of sensory information conveyed by the LPP, which is excessively amplified in the DG of schizophrenic individuals (Arnold, 1999;Behrendt, 2016;Prasad et al., 2004) instead of being properly filtered.Although appealing, this idea requires experimental examination.

| Behavioural dysregulation
Our spatial discrimination tests suggest that transient hypofunction of NMDA receptors alters the functional integrity of the DG and thus pattern separation activity, a model of neuronal activity that enables the recognition of independent events despite highly similar information content (Yassa & Stark, 2011).Previous studies have documented reduced volume and alterations in the cellular architecture of the hippocampus and surrounding areas of schizophrenic individuals, with the greatest volume loss in the DG (Nakahara et al., 2020).Volumetric dysregulation also implies changes in the information processing in which the DG participates, including pattern separation activity (Faghihi & Moustafa, 2015).In line with this, a recent study has shown deficits in mnemonic discrimination in schizotypal and first-episode psychotic individuals (Kraguljac et al., 2021).We demonstrate for the first time that transient hypofunction of NMDA receptors during early postnatal development impairs spatial discrimination in juvenile male rats.This finding substantiates the idea that impaired mnemonic discrimination is present during the early or prodromal phase of schizophrenia.Consistent with this, recent work showed that children and adolescents at high risk of developing schizophrenia exhibited impaired mnemonic discrimination compared to a control group ( _ Imamo glu et al., 2023).Given the similar impairments in the mnemonic discrimination task found in our animal model and individuals at high risk for developing schizophrenia, we propose that mnemonic discrimination can be used as a clinical hallmark of the cognitive status of individuals with a high risk of developing schizophrenia.Early screening could allow early therapeutic interventions and modify the course of the illness, as previously suggested (Insel, 2010).At the preclinical level, the animal models that presumably mimic the phenotype schizophrenia should reproduce this impaired mnemonic discrimination in the juvenile stage of animals.
Lastly, it is worth noting that while this study identified several changes in the electrophysiological properties of the MPP-and LPP-DG synapses, only the modifications observed in the MPP synapse contribute to the understanding of the dysregulation observed in the spatial object pattern separation task, primarily due to its spatial nature (Hunsaker et al., 2007).This finding does not rule out the contribution of the LPP synapse in spatial discrimination, which specializes in multisensorial information processing (Hunsaker et al., 2007).
The findings in the LPP-DG synapses suggest a potential dysregulation in the nonspatial discrimination process.The subsequent validation of behavioural tasks that specifically assess nonspatial discrimination will enable us to substantiate this hypothesis.In addition, whereas postnatal blockade of NMDA receptors did not interfere with the expression of MPP LTP, it did impair the induction of MPP CB 1 receptor dependent LTD.Given that both LTP and LTD are forms of long-lasting synaptic plasticity that collaborate to generate neural representations of memory in the hippocampus (Stacho & Manahan-Vaughan, 2022), these findings are extremely encouraging.For instance, while LTP is proposed to create a general schema for spatial representation (Kemp & Manahan-Vaughan, 2008;Stacho & Manahan-Vaughan, 2022), the expression of MPP LTD has been associated with changes in the spatial position of objects inside of familiar environment (Kemp & Manahan-Vaughan, 2008).According to these observations, the lack of MPP LTD may explain why MK-801-treated animals exhibited deficits in recognizing new spatial configurations of identic objects.This deficit in spatial discrimination may not be attributed to LTP, given that postnatal treatment with MK-801 did not interfere with its expression (see Figure 7) at 34 C for 25 to 30 min in an artificial cerebrospinal fluid solution (ACSF; pH ≈ 7.30-7.35)containing (in mM):-125 NaCl, 2.5 KCl, 1.25 Na 2 HPO 4 , 25 NaHCO 3 , 4 MgCl 2 , 1 CaCl 2 and 10 D-glucose.Next, the slices were kept at room temperature for at least 1 hour before any experimental procedure was performed.An individual slice was transferred to a submerged chamber (total volume: 400 μl).The slice was continuously perfused with modified artificial cerebrospinal fluid (ACSF) at a rate of 2.2-2.5 mlÁmin À1 with the help of a peristaltic pump (Sci Q400, Watson-Marlow, Wilmington, MA, USA).The modified ACSF (with continuous carbogen bubbling) included the following components (in mM): 125 NaCl, 2.5 KCl, 1.25 Na 2 HPO 4 , 25 NaHCO 3 , 1.5 MgCl 2 , 2.5 CaCl 2 and 10 D-glucose.The recordings were performed at 32.5 ± 1 C.

F
I G U R E 1 Postnatal treatment with MK-801 selectively reduces synaptic strength in the medial perforant path (MPP)-dentate gyrus (DG) synapse.(a) Top panel: schematic representation of the experimental timeline.Neonatal Wistar rats were daily injected with MK-801 (0.2 mgÁkg À1 ) during postnatal days 7-11.Acute brain slices from dorsal hippocampus for extracellular recordings or behavioural evaluations were performed at postnatal days 30-38.Bottom panel: schematic representation of the EC-DG circuit and placement of stimulation and recording electrodes for extracellular recordings.Stimulation of the lateral perforant path (LPP) and the resulting LPP field excitatory postsynaptic potential (fEPSP) were acquired in the outer one-third section of the molecular layer of the DG (green axons).MPP fEPSPs were acquired in the middle one third of the molecular layer of the DG (red axons).(b) Pharmacological sensitivity of the LPP fEPSPs and the MPP fEPSP to different agonists of metabotropic glutamate receptors.Upper panel: LPP fEPSP are sensitive to activation of group III mGlu receptors with L-AP4 (20 μM) and insensitive to stimulation of group II mGlu receptors with DCG-IV (5 μM).Lower panel: MPP fEPSP are insensitive to activation of group III mGlu eceptors.The opposite occurs in the presence of group II mGluRs agonists.The pharmacological sensitivity of the fEPSPs to the different mGluRs agonists was rigorously used as an inclusion criterion for all the experiments included in this study.(c) Input-output graphs of the LPP fEPSPs and (d) the MPP fEPSPs in response to increasing current pulses (100 μA steps at 0.067 Hz).Postnatal treatment with MK-801 did not alter the magnitude of the LPP fEPSP (upper trace and green symbols in panel c) but reduced the magnitude of the MPP fEPSP (upper trace and red symbols in panel d), *P < 0.05.(e) Left panel: input-output curve of the LPP fibre volleys (FV) in response to increasing current pulses; right panel: coupling analysis of the FV versus LPP fEPSP slope, showing no difference in synaptic strength.(f) Left panel: input-output curve of the MPP FV in response to increasing current pulses, showing decreased presynaptic excitability in the MK-801 group; right panel: coupling analysis of FV versus MPP fEPSP slope.*P < 0.05.For LPP: n = 9 slices per six animals for the control group and n = 10 slices per six animals for the MK-801 group.For MPP: eight slices per six animals for the control group and n = 10 slices per six animals for the MK-801 group.
mendations of the British Journal of Pharmacology on experimental design and analysis.Statistical analysis was conducted when each group size was at least n = 5.The normality distribution of data was validated with the Kolmogorov-Smirnov test (P = 0.05).No outliers were removed from the data.In a subset of exploratory experiments (see supporting information figures), we conducted statistical analysis in groups with size of n = 3-4; these results are referred as exploratory results.Each group size represents independent values, and the statistical analysis was done on these independent values.In electrophysiological experiments, the group size is expressed as the number of slices obtained from n animals (n = number of slices per number of animals).The comparability among experimental conditions was assessed by two-tailed paired (or unpaired) Student's t test, Mann-Whitney test, Wilcoxon test, Kruskal-Wallis test or two-way repeated-measures (RM) analysis of variance (ANOVA), as appropriate.The Holm-Šidák post hoc or Dunn's post hoc test was used for multiple comparisons among experimental groups, only when F achieved minimal statistical significance.For all the experiments, data were considered significant if P < 0.05.

3. 3 |
The transient hypofunction of NMDA receptors during early postnatal development modifies the paired-pulse ratio at the LPP and the MPP synapses Consistent with previous studies (Collitti-Klausnitzer et al., 2021; Petersen et al., 2013), the paired-pulse facilitation (PPF; 60 ms inter-T A B L E 1 Kinetic profile of dentate gyrus (DG) field excitatory postsynaptic potentials (fEPSP) evoked by stimulation of the medial perforant path (MPP) and the lateral perforant path (LPP) from control and MK-801 slices.Student's t tests; n = 10 slices per five animals for MPP in each experimental condition; n = 9 slices per five animals for LPP in each experimental condition.
stimulus interval, ISI) differs between the LPP and MPP synapses.We corroborated that the LPP has a prominent PPF (PPF at the LPP-DG = 1.72 ± 0.03; n = 28 slices per 15 animals) compared to the MPP (PPF at the MPP-DG = 1.22 ± 0.03; n = 29 slices per 15 animals; Student's t test: t [55] = 12.01, P < 0.05).Figure2ashows the PPF distribution of the LPP and the MPP using a violin graph and the responses obtained from 57 independent dorsal hippocampal slices.Figure2bshows representative examples of facilitation from the LPP and the MPP and their corresponding depression in response to selective activation of group III and II mGlu receptors, respectively.Because postnatal treatment with MK-801 enhances the PPF at the Mossy Fibre-CA3 pyramidal cell synapse of the rat hippocampus(Márquez et al., 2023;Segev et al., 2020), we performed a similar exploration on the LPP and the MPP using a broader inter-stimulus interval range(40, 60, 100, 200  and 500ms).The exploration of the LPP PPF revealed increased facilitation in the MK-801 slices compared to control slices (two-way RM ANOVA, treatment effect: F [1, 16] = 5.19, P < 0.05; n = 9 slices per seven animals, for each condition, Figure 2c).The LPP PPF level was statistically increased at 40 and 60 ms (LPP PPF at 40 ms in control: 1.64 ± 0.07; in MK-801: 2.01 ± 0.01; t (16) = 2.57, P < 0.05; at 60 ms: 1.74 ± 0.06; in MK-801: 1.99 ± 0.06; t (16) = 2.9, P < 0.05; at 100 ms: 1.71 ± 0.02; in MK-F I G U R E 2 Postnatal treatment with MK-801 reduces presynaptic neurotransmitter release at the lateral perforant path (LPP) and the medial perforant path (MPP).(a) Violin plots contrasting the paired-pulse facilitation values (PPF; inter-stimulus interval (ISI) 60 ms) of the LPP (n = 28 slices per 15 animals; green symbols) and the MPP (n = 29 slices per 15 animals; red symbols) synapses in the control condition.LPP PPF was greater than MPP PPF, *P < 0.05.(b) Representative traces showing the pharmacological sensitivity of the LPP and the MPP synapses.LPP PPF decreased in the presence of L-AP4 (20 μM, upper traces), whereas the MPP PPF was abolished by perfusion of DCG-IV (5 μM; lower traces).(c) Time course of the LPP PPF exploring different ISI (40, 60, 100, 200 and 500 ms).Postnatal treatment with MK-801 increased the LPP PPF.*P < 0.05.For LPP PPF: n = 8 slices per six animals for control and n = 9 slices per seven animals for MK-801.(d) Representative PPF of the LPP synapses contrasting the values of slices from control and MK-801-treated animals.(e) Time course and (f) representative traces for MPP PPF in slices from control (red symbols and traces) and MK-801-treated (black symbols and traces) animals.MPP PPF: n = 9 slices per seven animals for each condition.*P < 0.05.801: 1.79 ± 0.04; t (16) = 1.5, P > 0.05).The temporal change in the PPF is depicted in the representative traces of Figure 2d.The MPP PPF also showed increased facilitation in the slices from MK-801-treated animals (n = 9 slices per seven animals) compared to control slices (n = 8 slices per six animals; two-way RM ANOVA, treatment effect: F [1, 15] = 4.63, P < 0.05; Figure 2e).The facilitation was limited to the inter-stimulus interval of 200 ms (PPF at 60 ms in control: 1. FigureS1d,e).
FigureS1a,b).This form of synaptic potentiation requires additional investigation beyond the scope of the present study.

Figure 3a ,
Figure 3a,b summarizes the induction of LTD in the MPP-DG synapse and its sensitivity to perfusion DCG-IV (fEPSP at 90 min post-LFS: 69 ± 6.25% of baseline response, n = 6 slices per six animals; fEPSP in the presence of DGC-IV: 15.66 ± 2.55% of baseline response, Wilcoxon test, P < 0.05 vs. baseline response).Moreover, the induction of MPP LTD was accompanied by an increment in the PPF (MPP PPF in control condition: 1.18 ± 0.08; at 90 min post-LFS: Consistent with Peñasco et al., we found that perfusion of the CB 1 antagonist, AM251 (5 μM), blocks the induction of LTD (MPP fEPSP at 90 min post-LFS: 119.7 ± 10.94% of baseline response, n = 4 slices per four animals; lower traces and white symbols in Figure 3a,b and white bars in Figure 3e).The subsequent perfusion of the DCG-IV (5 μM) at the end of experiments corroborated the MPP origin of the synaptic responses (fEPSP in the presence of DCG-IV: 15.7 ± 5.22%of baseline; white bar in Figure3e).Likewise, perfusion of AM251 prevented the changes in the MPP PPF observed after the induction of LTD (MPP PPF in AM 251 treated slices in control condition: 1.22 ± 0.1; at 90 min post-LFS: 1.2 ± 0.08; Wilcoxon test, P > 0.05; white bars in Figure3f).Together, these results confirm that the MPP-DG synapse exhibits a stable LTD with a presynaptic locus of expression and requires CB 1 receptor activation.

increasing 2 -
AG signalling by inhibiting MAGL activity might restore the electric induction of LTD in the slices from MK-801-treated animals.For these experiments, the slices from MK-801-treated animals were perfused with the irreversible MAGL inhibitor, JZL 184 (1 μM, perfused for the last 10 minutes of the baseline response and the first 5 minutes during LFS).Before LFS, the application of JZL 184 did not modify the baseline response.More importantly, JZL 184 restored LTD at the MPP-DG synapse of the slices from MK-801-treated animals for up to 90 min (fEPSP 90 min post-LFS: 58.46 ± 11.68% of baseline, n = 5 slices per five animals, Student's t test: t [9] = 2.94, P < 0.05 vs. MK-801; traces and blue bars in Figure 5a-c).Perfusion of DCG-IV at the end of the recordings corroborated the MPP origin of the synaptic response (fEPSP in the presence of DCG-IV: 27 ± 7.43% of baseline).Likewise, induction of MPP LTD in the slices from MK-801-treated animals was accompanied by increased MPP PPF (PPR in control condition: 1.38 ± 1.11; at 90 min post-LFS: 1.77 ± 0.09; paired Student's t test: t (4) = 2.97, P < 0.05; blue bars in Figure 5d), which exhibited similar facilitation magnitude to the control slices (see black bars in Figure3f).The cumulative probability chart in Figure5e

F
Figure 6e shows the overall potentiation for control slices (black symbols) versus slices from MK-801-treated animals (green symbols).Together, these experiments show that the potentiation of the LPP fEPSP is sensitive to the transient hypofunction of NMDA receptors during early postnatal development, which is consistent with impaired functional expression of CB 1 receptors (see Figure 4a,b) and reduced

3. 9 |
Figure7a,b).Interestingly, MPP TBS delivered on the slices from MK-801-treated animals also triggered PTP, followed by a sustained increase in the slope of the fEPSP not statistically different from the synaptic potentiation observed in the control slices (fEPSP at PTP: 198.9 ± 29.47% of baseline; fEPSP 90 min post-TBS: 135 ± 9.43% of

3. 11 |
Transient hypofunction of NMDA receptors impairs spatial discrimination in rats . On the other hand, MK-801-treated animals efficiently differentiated the change in the spatial position of objects when displacement was maximal (DI in P5 from control vs. MK-801: 0.41 ± 0.035 vs. 0.43 ± 0.07; n = 10 for the control group and n = 12 for MK-801 group), and exhibited comparable DI values in the absence of object displacement (DI in P1 from control vs. MK-801: 0.001 ± 0.021 vs. À0.01 ± 0.011).However, when the magnitude of one object's displacement was gradually reduced from P4 to P2, the MK-801-treated animals discrimination ability was reduced (DI in P4 from control vs. MK-801: 0.38 ± 0.04 vs. 0.17 ± 0.05; DI in P3 from control vs. MK-801: 0.25 ± 0.03 vs. 0.1 ± 0.04; DI in P2 from control vs. MK-801: 0.07 ± 0.03 vs. 0.001 ± 0.01).These differences in DI values were significant at P3 (two-way RM ANOVA, treatment effect: F [1, 20] = 8.225, P < 0.05; Holm-Šidák post hoc test, P < 0.05).The heatmaps in Figure9cshow the spatial discrimination performance for both experimental conditions.Although MK-801-treated animals efficiently discriminate new object positions when the magnitude of the displacement is maximal (P5), their mnemonic ability for spatial discrimination is reduced when the magnitude of displacement of one object is gradually narrowed, a condition that increases the demand of spatial pattern separation (Figure9d).Together, these results suggest that impaired spatial discrimination in response to transient blockade of NMDA receptors may reflect impaired pattern separation associated with psychiatric disorders such as schizophrenia(Faghihi & Moustafa, 2015).The individual values of exploration time for objects in all spatial positions are depicted in FigureS3.4 | DISCUSSION This study provides experimental evidence of a series of synaptic alterations of the lateral and the medial perforant paths in response to the transient hypofunction of NMDA receptors during early postnatal development.We documented persistent dysregulation in the glutamate release from the lateral perforant path (LPP) synapse, blunted induction of LPP long-term potentiation (LTP), and decreased synaptic filtering capability.In the medial perforant path-dentate gyrus (MPP-DG) synapse, the altered glutamate release was accompanied by impaired CB 1 -dependent long-term depression (LTD) and without significant changes in the LTP expression.Mechanistically, the impairment of MPP LTD was partly due to decreased functional expression of the CB 1 receptor.More importantly, enhancing the 2-AG signalling pathway via pharmacological inhibition of the MAGL enzyme restored the strength of the MPP LTD.At the behavioural level, we show for the first time that transient hypofunction of NMDA receptors impairs spatial discrimination, a cognitive task in which dentate gyrus (DG) plays a critical role.4.1 | Changes in synaptic strength and presynaptic release at the perforant path (PP) synapsesFunctionally, the glutamatergic inputs to the DG convey distinct types of somatosensorial information.The MPP conveys spatial information from the medial entorhinal cortex and the LPP transfers multisensorial information from the lateral entorhinal cortex(Fernández-Ruiz et al., 2021;Hunsaker et al., 2007).Moreover, the synaptic transfer is sustained by axons with unique electrophysiological properties that determine plasticity capabilities and the pace and strength of neurotransmitter release(Collitti-Klausnitzer et al., 2021;Petersen et al., 2013).Those synaptic features have been robustly demonstrated in this study.Consistent with previous works(Márquez et al., 2023;Segev et al., 2020), our fibre volley and paired-pulse facilitation (PPF) analyses revealed that MK-801 alters the fibre volley amplitude (or presynaptic action potentials) and glutamate release.Dysregulation in the propagation of the fibre volley and the subsequent activation of the molecular machinery underlying glutamate release may explain the reduced strength of the glutamatetransmission found in the MPP synapse.In line with this possibility, transient hypofunction of NMDA receptors interferes with the functional expression of presynaptic proteins that control neurotransmitter release in animal models of schizophrenia(Maher & LoTurco, 2012; In a previous study,Peñasco et al. demonstrated  that MPP LTD can be induced with 6000 pulses at 10 Hz, but this required simultaneous blockade of GABA A receptors and involved activation of CB 1 receptors(Peñasco et al., 2019).In the present study, LTD was induced with less electrical stimulation and GABAergic transmission remained active.The resulting LTD was stable for up to 90 minutes without visible signs of a return to the baseline fEPSP value.The LTD was accompanied by increased PPR, indicating a presynaptic locus for its expression.As in Peñasco et al., the MPP LTD reported in our study requires postsynaptic production of 2-AG.The absence of LTD may result from presynaptic dysregulation in CB 1 receptor activity, accelerated breakdown of 2-AG, or altered synthesis of 2-AG at the postsynaptic level.The latter is unlikely since physostigmine induced a synaptic depression like that observed with activation of CB 1 receptor in the MK-801 condition.These findings indicate that MK-801-treated animals maintain functional postsynaptic production of 2-AG, suggesting a potential dysregulation locus at the presynaptic MPP terminals.In this regard, a critical finding of this study was that the blockade of the MAGL enzyme restored LTD in MK-801-treated animals.MAGL is a presynaptically expressed enzyme that metabolizes 2-AG and is essential in multiple physiological processes, including neuroplasticity, cognitive performance, and behaviour(Wang, Cox, et al., 2018;Zanfirescu et al., 2021).Our results demonstrate that reduced functional expression of CB 1 receptor contributes to the failure in the induction of MPP LTD in MK-801-treated animals.Another possibility contributing to this failure is that transient hypofunction of NMDA receptors could favour the accelerated breakdown of 2-AG via increased MAGL expression, as , suggesting a limited impact on the first neuronal template of spatial representation (i.e. during T1 from the spatial object pattern separation task).If this idea is true, restoring the MPP LTD or promoting its associated mechanisms, such as 2-AG signalling, could potentially improve spatial discrimination in MK-801-treated animals.5 | CONCLUSION Transient hypofunction of NMDA receptors with MK-801 alters the synaptic transfer of information from the entorhinal cortex to the dentate gyrus.These changes impact the lateral and medial synapses, dysregulating presynaptic glutamate release, induction of LTD and LTP, and synaptic filtering.Consistent with these alterations, spatial discrimination that depends on pattern separation activity is hindered in the experimental group.These physiological and behavioural dysregulations might account for the reduced cognitive performance observed in schizophrenia.Given its therapeutic potential, future research should consider the modulation of the endocannabinoid system (i.e.CB 1 receptor, 2-AG, or MAGL activity) to restore the synaptic strength and cognitive abilities of individuals at higher risk of developing schizophrenia.AUTHOR CONTRIBUTIONS L. A. A. Márquez: Conceptualization (equal); data curation (equal); formal analysis (equal); investigation (equal); methodology (equal); software (equal); visualization (equal); writing-original draft (equal).C. Lopez-Rubalcava: Resources (equal).E. J. Galvan: Conceptualization (equal); data curation (equal); funding acquisition (equal); investigation (equal); methodology (equal); project administration (equal); resources (equal); software (equal); supervision (equal); validation (equal); visualization (equal); writing-review and editing (lead).